Download From: www.EasyEngineering.net ww w.E asy E ngi nee rin g.n et **Note: Other Websites/Blogs Owners Please do not Copy (or) Republish this Materials, Students & Graduates if You Find the Same Materials with EasyEngineering.net Watermarks or Logo, Kindly report us to easyengineeringnet@gmail.com Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH INDEX UNITS PAGE NO. ww I. Introduction 06 II. One Dimensional Problems w.E asy 46 En gin III. Two Dimensional Continuum IV. V. 101 eer i ng. Axisymmetric Continuum 150 net Isoparametric Elements For Two Dimensional Continuum 190 VEL TECH VEL TECH MULTI TECH 2 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH MH1003 FINITE ELEMENT ANALYSIS (Common to Mechanical, Automobile, Mechatronics (Elective) and Metallurgical Engineering (Elective)) 1. INTRODUCTION 9 Historical background – Matrix approach – Application to the continuum – Discretisation – Matrix algebra – Gaussian elimination – Governing equations for continuum – Classical Techniques in FEM – Weighted residual method – Ritz method 2. ONE DIMENSIONAL PROBLEMS 9 Finite element modeling – Coordinates and shape functions- Potential energy approach – Galarkin approach – Assembly of stiffness matrix and load vector – Finite element equations – Quadratic shape functions – Applications to plane trusses ww w.E 3. TWO DIMENSIONAL CONTINUUM 9 Introduction – Finite element modelling – Scalar valued problem – Poisson equation –Laplace equation – Triangular elements – Element stiffness matrix – Force vector – Galarkin approach - Stress calculation – Temperature effects asy 4. AXISYMMETRIC CONTINUUM En 9 gin Axisymmetric formulation – Element stiffness matrix and force vector – Galarkin approach – Body forces and temperature effects – Stress calculations – Boundary conditions – Applications to cylinders under internal or external pressures – Rotating discs 5. eer i ISOPARAMETRIC ELEMENTS FOR TWO DIMENSIONAL CONTINUUM 9 ng. The four node quadrilateral – Shape functions – Element stiffness matrix and force vector – Numerical integration Stiffness integration – Stress calculations – Four node quadrilateral for axisymmetric problems. TEXT BOOKS 1. 2. net Chandrupatla T.R., and Belegundu A.D., ‚Introduction to Finite Elements in Engineering‛, Pearson Education 2002, 3rd Edition. David V Hutton ‚Fundamentals of Finite Element Analysis‛2004. McGraw-Hill Int. Ed. REFERENCES 1. 2. 3. 4. 5. Rao S.S., ‚The Finite Element Method in Engineering‛, Pergammon Press, 1989 Logan D.L., ‚A First course in the Finite Element Method‛, Third Edition, Thomson Learning, 2002. Robert D.Cook., David.S, Malkucs Michael E Plesha, ‚Concepts and Applications of Finite Element Analysis‛ 4 Ed. Wiley, 2003. Reddy J.N., ‚An Introduction to Finite Element Method‛, McGraw-Hill International Student Edition, 1985 O.C.Zienkiewicz and R.L.Taylor, ‚The Finite Element Methods, Vol.1‛, ‚The basic formulation and linear problems, Vol.1‛, Butterworth Heineman, 5th Edition, 2000. VEL TECH VEL TECH MULTI TECH 5 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH UNIT – I PART – A 1. Define the term finite element. A complex region defining a continuum is discredited into simple geometric shapes called finite elements. ww 2. Why study FEA? w.E a) To design products that is safe & cost effective. b) To analyze cause of failure in engineering structures 3. Why FEA is so important? asy En FEA is numerical method, which can be used to find location and magnitude of critical stress and reflection in a structure. FEA method can be applied to structure that have no theoretical solution available, and without FEA we will have to use experimental techniques, which can be consuming and expensive. gin ng. F Solid plate –theoretical Solution is possible eer i Plates with notes-No theoretical solution Available. f net 4. Define nodes: The finite element procedure reduces such unknown to a finite number by dividing the solution region into small parts called elements and by expressing the unknown field variables interms of assumed approximating functions within each element. The approximating functions are defined interms of field variables of specified points called nodes or nodal points. VEL TECH VEL TECH MULTI TECH 6 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 5. List out the various steps involved in finite element analysis. i) ii) iii) iv) v) vi) vii) viii) Select suitable field variables and the elements. Discritise the continuum. Select interpolation functions. Find the element properties. Assemble elements properties to get global properties. Impose the bounding condition. Solve the system equations to get the nodal unknowns. Make the additional calculations to get the required values. ww 6. Define Matrix: w.E A rectangular array of numbers with a definite number of rows and columns is a matrix. a11 a12 ...a1n a21 a22 ... a2n e.g : A ................... am1 am2 ... amn asy En gin 7. How does FEA work? eer i In FEA, an engineering structure is divided into smaller regions, which have simpler geometry and theoretical solution. Collectively the regions represent the entire structure, and the individual element contributes to the solution of the structure. Challenge lies in representing the exact geometry of the structure. Especially, the sharp curves. Generally, a multi-degree polynomial is approximated by a high Number of straight edges. ng. 8. Explain the term transposition in matrix. net If A = [aij], then the transpose of A, denoted as AT, is given by AT = [aji]. Thus the rows of a are the columns of AT. 1 5 0 6 then A T 1 0 2 4 e.g : A 5 6 3 2 2 3 4 2 VEL TECH VEL TECH MULTI TECH 7 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 9. Write the global matrix equation. { F} = [K] {u} Where, {F} = External force matrix. [K] = Global stiffness matrix. {u} = Displacement matrix. 10. List out the major steps used in FEA. i) ii) iii) ww Preprocessing or modeling the structure. Analysis Post processing. w.E 11. Briefly explain the term Discretization:- asy Discretization is the process of dividing an engineering structure into small elements. En In FEA, Discretization of a structural model is another name for mesh generation. gin 12. List out the basic elements used in FEA. i) Line elements: Elements consisting of two nodes. eer i ng. In computers, a line, connecting two nodes at its ends as shown, represents a line element. The cross, sectional area is assumed constant throughout the elements. e.g: Truss and beam elements. net Line elements ii) 2-D solid elements: Elements that have geometry similar to a flat plate. 2-D solid elements are plane elements, with constant thickness, and have either a triangular or quadrilateral shape, with 3 nodes or 4 nodes. e.g: plane stress, plain strain, plates shells and axisymmetric elements. VEL TECH VEL TECH MULTI TECH 8 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH 2D solid: Triangular iii) VEL TECH HIGH TECH 2-D solid: Quadrilateral. 3-D solid elements: elements that have a 3-D geometry. ww The basic 3-D solid elements have either a tetrahedral (4 focus) or hexahedral (6 faces) shape. w.E Tetrahedral - 4 nodes. Hexahedral – 8 nodes. asy 13. Give the relation ship between matrix & Algebra, algebraic equation: En a11x1 a12 x 2 a13 x 3 b1 a21x1 a22 x 2 a23 x 3 b2 a31x1 a32 x 2 a33 x3 b3 gin Matrix form: a11 a12 a13 x1 b1 a21 a22 a23 x 2 b2 a a a x b 31 32 33 3 3 eer i ng. (or) [A] {x} = {b}. net 14. Write a short note on RA Z-method. (or) Raleigh Ri+z method. The Rayleigh – Ri+z method of expressing field variables by approximate method clubbed with minimization of potential energy has made a big break through in finite element analysis. The Rayleigh – Ri+z method involves the construction of an assumed displacement field, u ai i (x,y,z) i 1 tol v a j j (x,y,z) j l 1 to m. w ak k (x,y,z) k m 1 to n n m l. VEL TECH VEL TECH MULTI TECH 9 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH The functions I are usually taken as polynomials. Displacements u,v,w must be cinematically admissible. That is u,v,w must satisfy specified boundary conditions. Introducing stress-strain and strain – displacement relations, and substituting above equation into the equation 1 T T T T v dv v u f d v s u Tds ui pi 2 i it gives, (a1,a2 ,...,ar) Where r = number of independent unknowns. Now, the extremum with respect to ai, (I = 1 to r) yields the set of r equations. ww 0 ai i 1,2, ...,r. w.E from the solutions of r equation, we get these values of all ‘a’. With these values of ai and I satisfying boundary conditions, the displacements are obtained. asy 15. What is the principally virtual work? En gin A body is in equilibrium if the internal virtual work equals the external virtual work for every kinematically admissible displacement field (,()). 16. What is the principle of minimum potential energy? eer i ng. For conservative systems, of all the kinematically admissible displacements fields, these corresponding to equilibrium extremize the total potential energy. If the extremum condition is a minimum, the equilibrium state is stable. 17. Distinguish between Finite element method is classical methods: i) ii) iii) net In classifial methods exact equations are formed and exact solutions are obtained where as in finite element analysis exact equations are formed but approximate solutions are obtained . Solutions have been obtained for few standard cases by classical methods, where as solution can be obtained for all problems by finite element analysis. When material property is not isotropic, solutions for the problems become very difficult in classical method. Only few simple cases have been tried successfully by researchers. FEM can handle structures with anisotropic properties also without any difficulty. VEL TECH VEL TECH MULTI TECH 10 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH iv) v) VEL TECH MULTI TECH VEL TECH HIGH TECH If structure consists of more than one material, element can be used without any difficulty. Problems with material and geometric non-linearities cannot be handled by classical methods but there is no difficulty in FEM. 18. Distinguish between finite Element method (FEM) vs Finite Difference method (FEM): i) FDM makes point wise approximation to the governing equations i.e it ensures continuity only at the node points. Continuity along the sides of grid lines are not ensured. ww FEM makes piecewise approximation i.e it ensures the continuity at node points as well as along the sides of the element. ii) iii) w.E FDM needs larger number of nodes to get good results while FEM needs fewer nodes. With FDM fairly complicated problems can be handled where as FEM can handle all complicated problems. asy En 19. Write a short note on plane stress models and plane strain models: Plane stress models: gin eer i i) ii) iii) No loading Normal to the plane. No stress Normal to the plane. z 0, xz 0, yz 0. iv) z 0. v) If the change in length (T) to the original length is greater, i.e., original length(T) is T o. small, hence T ng. net Plane strain models. i) ii) iii) Strain occurs only in the xy – plane. z=0 and shear strains yxz and yyz are also equal to zero. z0. iv) If the plate with a hole is thick, the change in thickness compared to the original thickness will be small, and therefore, z=0. VEL TECH VEL TECH MULTI TECH 11 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 20. Define Aspect ratio. State its significance. Aspect ratio is defined as the ratio of largest to smallest size in an element. Aspect ratio should be as close to unity as possible. For a two dimensional rectangular element, the aspect ratio is conveniently defined as length to breadth ratio. Aspect ratio closer to unity yields better results. 21. What is the post processing in finite element Analysis? ww This is the last step in a finite element analysis. Results obtained in Analysis after the preprocessing are usually in the form of raw data and difficult to interpret. In post analysis, a CAD program is utilized to manipulate the data for generating deflected shape of the structure, creating stress plots, animation, etc. A graphical representation of the results is very useful in understanding the behaviour of the structure. w.E asy 22. List out the applications of FEA. FEA can be used in. i) ii) iii) iv) v) vi) vii) En Heat transfer Fluid mechanics (Two dimensional flow). Solid mechanics. Boeing 747 aircraft. Nuclear reaction vessel. Bio-mechanics Reinforced concrete beam. gin eer i ng. 23. Write some advantages and disadvantages of finite element method. net Advantages: i) ii) iii) iv) VEL TECH The method can efficiently be applied to cater irregular geometry. It can take care of any type of boundary. Material anisotropy and in homogeneity can be treated without much difficulty. Any type of loading can be handled. VEL TECH MULTI TECH 12 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Disadvantages:i) ii) iii) There are many types of problems where some other method of analysis may probe efficient then the finite element method. Cost involved in the solution of the problem. For vibration and stability problems in many cases the cost of analysis by finite element method may be prohibitive. 24. Use the Gaussian elimination method to solve the simultaneous equations. ww 2a + b + 2c – 3d = 0 2a – 2b + c – 4d = 5 a +2c – 3d = -4 4a + 4b – 4c + d = -6 Matrix from is w.E asy En 2 1 2 3 a 0 2 2 1 4 b 5 1 0 2 3 c 4 4 4 4 1 d 6 gin The upper triangular matrix is eer i ng. 3 a 0 1 0.5 1 0 1 0.33 0.33 b 1.66 0 0 1 1.14 c 4.14 0 1 d 10.80 0 0 net solving the equation, it gives, a = 12, b = -8, c =-8.2, d = 10.2. 25. Define stiffness matrix. The term stiffness matrix originates from structural analysis. The matrix relation between temperature and heat flux is called the stiffness matrix. VEL TECH VEL TECH MULTI TECH 13 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH PART – B 3 1 4 1. A= -1 4 2 -2 2 2 3 1 4 : 1 0 0 -1 4 2 : 0 1 0 -2 2 -2 : 0 0 1 ww 3 1 4 : 1 0 0 R2 3R2 + R1 0 13 10 : 1 3 0 R3 3R3 + 2R2 0 8 2 : 2 0 3 w.E 0 0 3 1 4 : 1 0 13 10 : 1 3 0 R 3 13R3 8R2 0 0 -54 : 18 -24 39 asy 39 0 42 12 3 0 0 13 10 1 3 0 R1 13R1 8R2 0 0 54 18 24 39 En 0 1404 1170 1638 2106 0 0 13 10 1 3 0 R1 54R1 42R3 0 0 54 18 24 39 gin 0 1404 1170 1638 2106 0 0 13 10 1 3 0 R1 54R1 42R3 0 0 54 18 24 39 eer i ng. net 0.55 0.77 1 0 0 0.66 0 1 0 0.179 0.11 0.55 0 0 1 0.33 0.44 0.722 0.55 0.77 0.66 A 1 0.179 0.11 0.55 0.33 0.44 0.722 VEL TECH VEL TECH MULTI TECH 14 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 2. A. Briefly explain about i) Rayleigh – Ri+z method. ii) Weighted residual method. iii) steps involved in FEA. B. Give the comments on Rayleigh – Ri+z method. Solution:A. i) Rayleigh – Ri+z method:- ww For continua, the total potential energy in the following equation, w.E 1 v T dv v uT fdv s uT Tds uiTpi 2 i asy can be used for finding an approximate solution The Rayleigh – Ri+z method involves the construction of an assumed displacement field, i.e., u aii(x,y,z) i 1 to l v a ji (x,y,z) j l 1 to m w akk (x,y,z) k m 1 to n En gin nml eer i ng. The functions I are usually taken as polynomials. Displacements u,v,w must be kinematically admissible. That is u,v,w must satisfy specified boundary conditions. Introducing stress-strain and strain displacement relations and substituting the above equations (i.e values of u,v,w) gives = ( a1, a2, <,ar) net Where r = number of independent unknowns Now, the extremum with respect to ai, (I = 1 to r). yields the set of r equations 0, ai i 1,2,...,r. (ii) Weighted residual method: Weighted residual methods are another way to develop approximate solutions. In weighted residual method, first assume the form of the global solution and then adjust parameters to obtain the best global fit to the actual solution. VEL TECH VEL TECH MULTI TECH 15 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH The following figure contains a body B with boundary S. The boundary is divided into two regions su with essential (Dirichlet) boundary conditions and a region sf with natural (Neumann) boundary conditions. The essential boundary conditions are specifications of the solution on the boundary (for example, known boundary displacement), while the natural boundary conditions are specifications of derivatives of the solution (for e.g. surface tractions). All points on the boundary must have one or the other type of specified boundary conditions. General body with boundary. The basic step in weighted residual methods is to assume a solution of the form: n ww un a j j j1 w.E In that aj value should be find out and that gives a best approximation to the exact solution. asy iii) Steps involved in FEA. En FEA solution of engineering problems, such as finding deflections stresses in a structure, requires three steps: gin 1. Pre-processing or modeling the structure 2. Analysis 3. Post processing A brief description of each of these steps follows. Step 1: Pre-process of modeling the structure eer i ng. net The structure is modeled using a CAD program that either comes with the FEA software or provided by another software vendor. The final FEA model consists of several elements that collectively represent the entire structure. The elements not only represent segments of the structure, they also simulate its mechanical behaviour and properties. Regions where geometry is complex (curve, notches, holes, etc.) require increased number of elements to accurately represent the shape; where as, the regions with simple geometry can be represented by coarser mesh (or fewer elements). The selection of proper elements requires prior experience with FEA, knowledge of structure’s behaviour, available elements in the software and their characteristics, etc. The elements are joined at the nodes, or common post. In the pre-processing phase, along with the geometry of the structure, the constraints, loads and mechanical properties of the structure are defined. Thus, in pre-processing, the entire VEL TECH VEL TECH MULTI TECH 16 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH structure is completely defined by the geometric model. The structure represented by nodes and elements is called ‚mesh‛. Step 2: Analysis In this step, the geometry, constraints, mechanical properties and loads are applied to generate matrix equations for each element, which are then assembled to generate a global matrix equation of the structure. The from of the individual equations, as well as the structural equation is always, {F} = [K] {u} Where ww w.E {F} = External force matrix [K] = Global stiffness matrix {u} = Displacement matrix asy En The equation is then solved for deflections. Using the deflection values, strain, stress, and reactions are calculated. All the results are stored and can be used to graphic plots and charts in the post analysis. gin Step 3: Post processing eer i ng. This is the last step in a finite element analysis. Results obtained in step 2 are usually in the form of raw data and difficult to interpret. In post analysis, a CAD program is utilized to manipulate the data for generating deflected shape of the structure, creating stress plots, animation, etc. A graphical representation of the results is very useful in understanding behaviour of the structure. net 2. b. Comments on Rayleigh-Ritz Method In this method the approximating functions must satisfy the boundary conditions and should be easy to use. Polynomials are normally used. Some times sine-cosine terms are also used. Results can be obtained for complex problems. But for complex problem it is difficult to say whether the results obtained are accurate enough to use. The doubt will arise due to the following two reasons (i) Whether this is the only function which can be used (ii) How many terms in the function are to be used. VEL TECH VEL TECH MULTI TECH 17 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH The best way to ensure the accuracy is to get result using a certain number of terms and then use additional terms to get the results. If the difference is negligible, we can conclude that the satisfactory result is obtained. However it may be noted that the lowest terms in the series should not be omitted in the approximating functions. 3.a. How weighted residuals work? Let us first demonstrate how weighted residuals work using a bar subjected to body and end loads (Figure). ww w.E asy Figure. Uniform bar with body and tip loads En For static equilibrium, the summation of the forces is zero: F 0 or: x gin A x f B (x)x A xx 0 eer i ng. Rearranging and assuming constant area: A x x. x x f (x) 0 B Taking the limit as x0, A d f B (x) 0 dx net (1) For an elastic material, the stress is related to the strain by, E (2) Where E is Young’s modulus. The strain is related to the displacements by: du dx VEL TECH (3) VEL TECH MULTI TECH 18 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Substituting (3) and (2) into (1) , A d du B E f (x) 0 dx dx Assuming young’s modulus E is constant, with fB(x) =b, gives, AE d2u b 0 dx 2 for 0 x L (4) with boundary conditions: u x0 0 EA ww (5) du x L P dx (6) w.E Equation (4), along with the boundary conditions (5) and (6), forms the differential equation for the problem at hand. They can be solved, by direct integration, for the exact solution. asy For the weight residual formulation, we first choose a weighting function w(x), multiply (4) by the weighting function: En d2u w EA 2 b 0 dx (7) gin and then integrate over the entire body: d2u w EA 0 dx2 b dx 0 L (8) eer i ng. net This is called the weighted residual formulation. It is called this because if we assume an approximate solution Un (that satisfies all boundary conditions) then, d2un AE 2 b R(x) 0 dx (9) Instead, we have an error (residual) that is a function of x. Thus (8) is really a weighting of the residual over the body: L wRdx 0 (10) 0 We have taken the error (residual), multiplied by a weighting function and set the weighted integral to zero. VEL TECH VEL TECH MULTI TECH 19 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 3.b. Using Rayleigh – Ri+z method determine the expression for deflection and bending moments in a simply supported beam subjected to uniformly distributed load over entire span. Find the deflection and moment at miss pan and compare with exact solutions. Solution: The following figure shows the typical beam. The Fourier series y a sin m 1,3 i m x is the ideal l 2 function for simply supported beams since y = o and M = EI dy 0 at x = 0 and x = l are satisfied. x 2 For the simplicity ww w.E asy En Figure. Let us consider only two terms in the series i.e. let y a1 sin x l a2 sin 2 3 x l eer i <(a) EI d2 y 2 dx wy dx 2 dx 0 0 l gin ng. l <(b) Substituting y in equation (b) we get 2 net EI 2 x 9 2 3 x x 3 x dx = 2 a1 sin 2 a2 sin dx w a1 sin a2 sin 2 l l l l l l 0 0 l l EI 2 x 3 x l l 3 x a sin 9a2 sin dx w a1 a2 cos 2 1 2 l 0 l l 3 l 0 1 2 l l 2a EI 4 x 3 x 3 x wl a sin2 18a1a2 sin 81a22 sin2 dx 2a1 2 4 1 2 l 0 l l l 3 VEL TECH VEL TECH MULTI TECH 20 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH l Nothing that sin2 0 x VEL TECH HIGH TECH 1 2 x 1 dx 1 cos dx l 2 l 2 0 x l 3 x 2 x 4 x 0 sin l sin l dx 0 cos l cos l dx 0 l l l and sin2 0 3 x 1 6 x 1 cos dx l 2 l 2 0 l y a EI 4 2 1 1 2wl a 81 a22 a1 2 4 1 2 l 2 2 3 ww we get, a EI 4 2wl a 12 81a22 a1 2 4 4 l 3 w.E to be minimum, 0 and 0. a1 a2 asy EI 4 2wl 2a1 0 3 4l 4wl4 a1 3 5 EI 4 2wl 81 x 2a2 0 3 4l 3 En i.e., or and a2 y 4wl4 x 4wl4 3 x sin sin 5 5 EI l 243EI l Max, deflection which occurs at x eer i ng. 4wl4 243EI 5 or ymax gin net l is 2 4wl4 4wl4 wl4 EI 4 243EI 5 76.82EI we know the exact solution is 5 wl4 wl4 ymax 384 EI 76.8EI VEL TECH VEL TECH MULTI TECH 21 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Thus the deflection is almost exact. Now, Mx EI d2 y 2 x 9 2 3 x EI a sin a sin 1 2 2 2 2 dx l l l l 4wl2 x 4wl2 x9 3 x = EI sin sin 3 3 l 243EI l EI 4wl2 4wl2 9 wl2 Mcentre 3 243EI 3 8.05 EI ww we know the exact value is wl2 . 8 w.E By taking more terms is Fourier series accurate results can be obtained. 4. asy a. (i) Explain in detail H – elements versus P – elements. (ii)Distinguish between Bottom up and 4.a.(i) H – versus P – elements En Top-down approach in FEA. gin In FEA, there are two types of elements: 1. h-elements and, eer i ng. 2. P-elements net H-element is the original and ‚classic‛ element. The name is derived from the field of numerical analysis, where the letter ‘h’ is used for the step size, to achieve convergence in the analysis. The h-element is always of low order, usually, linear or quadratic. When a finite element mesh is refined to achieve convergence, the procedure is called h-convergence. For helements, convergence is accomplished regions require a very fine mesh, thereby increasing the number of elements. Finite elements used by commercial programs in the 1970s and 80s, well helements. However, with improvement in computer power and efficiency, a much more useful, pelements were developed. P-elements are relatively new, developed in late 1980s and offer not only the traditional static analysis, they provides option of optimizing a structure. In Pro/M, P-elements can have edgeploynomial as high as 9th order, unlike the low order polynomials of h-elements. The high polynomial edge order of p-elements makes it possible to model a curved edge of a structure with accuracy. Therefore, fewer elements can be used to achieve convergence. When p-elements are used, the number of elements in the mesh usually remains fixed; convergence is achieved by VEL TECH VEL TECH MULTI TECH 22 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH increasing the polynomial order of the p-elements, rather than refinement of the mesh. For optimization, as the dimensions of the structure being analyzed are changed, the number of elements remains constant. Only the polynomial order of the elements is changed as needed. 4. a. (ii) Bottom-up and Top-down approach When modeling a structure (creating an FEA model), bottom – up approach refers to creation of model by defining the geometry of the structure with nodes and elements. These nodes and elements represent the physical structure. When an FEA model is created by this procedure, it is known as a bottom-up approach. This is the original procedure for creating FEA mesh, and requires a substantial investment in time and skill. When this method is employed, most of analyst’s time is devoted to creation of the mesh, and only a fraction of time is spent for analysis and results interpretation. ww w.E In FEA, a top-down procedure refers to certain of FEA mesh by first building a solid model, using a 3-D CAD program, and then dividing the model into nodes and elements. Thus, the top-down method requires building of geometric model of the structure, which is then used to create an FEA mesh. The advantages of the top-down approach are obvious; we don’t have to define the geometry of individual elements in the structure, which can be very time-consuming. Obviously, a 3-D model requires high-end computer hardware, along with familiarity with the modeling software. asy En gin eer i 4.b. Using Rayleigh – Ri+z method, determine the expressions for displacement and stress in a fixed bar subjected to axial force P as shown in the below fig. Draw the displacement and stress variation diagram. Take 3 terms in displacement function. ng. net Solution: Let the displacement at load point be u1. Then the strain energy of the bar 2 1 du U EA dx 20 dx l And potential energy due to external forces = Pu1 2 1 du EA dx Pu1 20 dx l VEL TECH VEL TECH MULTI TECH 23 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Let the displacement at any point be given by, u = a1 + a2x + a3 x2 This function has to satisfy the boundary conditions (i) at x = 0, u = o (ii) at x = 1, u = 0 From Boundary condition (i), we get 0 = a1 (1) ww From Boundary conditions (ii), we get, 0 a1 a2l a3l2 w.E (2) asy From equations (1) and (2) we get 0 = a2 l + a3 l2 or a2 = -a3 l u a3lx a3 x2 a3 lx x 2 x At En (3) gin 1 2 ng. l l2 u u1 a3 l 2 4 u1 i.e., eer i 2 a 3l 4 (4) net du a3 l 2x dx Now l 1 l2 2 2 EA a l 2x dx Pa 3 3 2 0 4 l 1 l2 2 2 2 EA a 3 l 4lx 4x dx Pa3 2 4 0 l 1 4x 3 l2 2 2 2 EA a3 l x 2lx Pa3 2 3 0 4 VEL TECH VEL TECH MULTI TECH 24 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH l2 1 l2 EA a23 Pa3 2 4 3 d l3 l2 0 EAa3 P 0 da3 3 4 a3 3P 4EAl u 3P lx x 2 4EAl ww a3l2 3pl u1 4 16 w.E du E Ea3 ( l 2x) dx 3p 3p E [ l 2x] [l 2x] 4EAl 4Al 0 x 0 1 x l 2 3p 4A asy En gin 0 eer i ng. 3p 2 x l 4A The variation of displacement and stresses are shown in Figure. net Fig.(a) Variation of u (b) Variation of stress. VEL TECH VEL TECH MULTI TECH 25 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 5.a. In the given spring structure, k1 = 20 lb/in, k2 = 25 lb/in, K3 = 30 ib/in, F = 5 lb. Determine deflection at all the nodes. Solution: Step 1: Derive the Element Equations ww As derived earlier, the stiffness matrix equations for an elements e is, k e k e K (e) k e k e w.E asy Therefore, stiffness matrix equations for an element e is, 1 2 20 20 1 Element1: k (1) 20 20 2 2 3 En gin eer i 25 25 2 Element1: k (2) 25 25 3 Element 3 : k (3) 3 ng. 4 30 30 3 30 30 4 net Step 2: Assemble element equations into a global equation Assembling the terms according to their row and column position. We get K g 1 2 3 4 20 0 0 1 20 20 20 25 25 0 2 0 25 25 30 30 3 0 30 30 4 0 VEL TECH VEL TECH MULTI TECH 26 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Or, by simplifying 0 20 20 0 20 45 25 0 K g 0 25 55 30 0 30 30 0 The global structural equation is, F1 20 20 0 0 u1 F2 20 45 25 0 u2 F3 0 25 55 30 u3 F 0 0 30 30 4 u4 ww w.E Step 3 : Solve for deflections asy First, applying the boundary conditions u1 = 0, the first column will drop out. Net, F1 =F2=F3=0, and F4 = 5 lb. The final form of the equation becomes. 0 45 25 0 u2 0 25 55 30 u3 5 0 30 30 u 4 En gin eer i ng. This is the final structural matrix with all the boundary conditions being applied. Since the size of the final matrices is small, deflections can be calculated by hand. It should be noted that in a real structure the size of a stiffness matrix is rather large and can only be solved with the help of a computer. Solving the above matrix equation by hand we get, 0 = 45 u2 – 25 u3 0 = -25 u2 + 55 u3 – 30 u4 net u2 0.2500 Or u3 0.4500 u 0.6167 4 5 = -30 u3 + 30 u4 5. b. In the spring structure shown, k1 = 10 lb/in, k2 = 15 lb/in, k3 = 20 lb/in, P = 5 lb. Determine the deflections at node 2 and 3. VEL TECH VEL TECH MULTI TECH 27 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Solution: Step 1: Find the Element Stiffness Equations 1 2 10 10 1 k (1) 10 10 2 Element 1 : 2 3 15 15 2 k (2) 15 15 3 Element 2 : 3 ww 4 20 20 3 k (3) 20 20 4 Element 3 : w.E Step 2 : Find the Global stiffness matrix 1 2 3 4 asy En 1 10 10 0 0 10 10 0 0 2 10 10 15 15 0 10 25 15 0 3 0 15 15 20 20 0 15 35 20 4 0 0 20 20 0 0 20 20 gin Now the global structure equation can be written as, F1 10 10 0 0 u1 F2 10 25 15 0 u2 F3 0 15 35 20 u3 F 0 0 20 20 4 u4 eer i ng. net Step 3 : Solve for Deflections The known boundary conditions are : u1= u4 = 0, F1= P = 31b. Thus, rows and columns 1 and 4 will out, resulting in the following matrix equation, 0 25 15 u2 3 15 35 u3 Solving, we get u2 = 0.0692 & u3 = 0.1154. VEL TECH VEL TECH MULTI TECH 28 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH 6. VEL TECH MULTI TECH VEL TECH HIGH TECH In the spring structure shown, k1 = 10n/mm, k2 = 15n/mm, k3 = 20 n/mm, k4 = 25 n/mm, k5= 30 n/mm, k6 = 35 N/mm, F2 = 100N. Find the deflections in all springs. ww Solution : w.E 1 Element 1 : asy 4 10 10 1 k (1) 10 10 4 1 Element 2 : 15 15 1 k(2) 15 15 2 2 Element 3 : VEL TECH eer i ng. 3 3 net 4 30 30 2 k (5) 30 30 4 3 Element 6 : gin 25 25 2 k (4) 25 25 3 2 Element 5 : En 20 20 2 k (3) 20 20 3 2 Element 4 : 2 4 35 35 3 k (6) 35 35 4 VEL TECH MULTI TECH 29 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH The global stiffness matrix is, 1 2 3 4 15 0 10 10 15 1 15 15 20 25 30 20 25 2 30 k g 0 3 20 25 20 25 35 35 30 35 10 30 35 4 10 And simplifying, we get 0 10 25 15 15 90 45 30 k g 0 15 80 35 10 30 35 75 ww w.E And the structural equation is, asy F1 25 15 0 10 u1 F2 15 90 45 30 u2 F3 0 45 80 35 u3 F 10 30 35 75 u 4 4 En gin eer i Now, apply the boundary conditions, u1 = U4 = 0, F2 = 100N. The is carried out by deleting the rows 1 and 4, columns 1 and 4, and replacing F2 by 100N. The final matrix equation is, ng. 100 90 45 u2 0 45 80 u3 Which gives net u2 1.5459 u3 0.8696 Deflections: Spring 1 : u4 – u1 = 0 Spring 2 : u2 – u1 = 1.54590 Spring 3 : u3 – u2 = 0.6763 Spring 4 : u3 – u2 = 0. 6763 VEL TECH VEL TECH MULTI TECH 30 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Spring 5 : u4 – u2 = 1.5459 Spring 6 : u4 – u3 = 0.8696. 7. a. State and explain with an example Gaussian elimination. Gaussian Elimination: It is the name given to a well. Known method of solving simultaneous equations by successively eliminating the unknowns. x1 2x 2 6x3 0 ( ) 2x1 2x 2 3x 3 3 ( ) x1 3x 2 ( ) ww 2 (1) w.E The equations are labeled as ,, and . Now, we wish to eliminate x1 from and. We have, from Eq. , x1 = + 2x2 – 6x3. Substituting for x1 into Eqs. and yields x1 2x 2 6x 3 0 ( ) 0 6x 2 9x 3 3 ((1) ) 0 x 2 6x 2 2 ( ) (1) asy En (2) gin eer i It is important to realize that Eq. Can also be obtained from Eq. by row operations. Specifically, in Eq., to eliminate x1 from II, we subtract 2 times I from II, and to eliminate x1 from III we subtract – 1 times I from III. The result is Eq. . Notice the zeroes blow the main diagonal in column 1, representing the fact that x1 has been eliminated from Eqs. II and III. The superscript (1) on the labels in Eqs. Denotes the fact that the equations have been modified once. ng. net We now proceed to eliminate x2 from III in Eqs. For this, we subtract 1/6 times II from III. The resulting system is x1 2x 2 6x 3 0 ( ) (1) 0 6x 2 9x 3 3 ( ) (2) 15 3 ( ) 0 0 x3 2 2 (3) The coefficient matrix on the left side of Eqs. Is upper triangular. The solution now is virtually complete, since the last equation yields x3 = 1/5, which, upon substitution into the second equation, yields x2 = 4/5, and then x1 = 1/5, from the first equation. This process of obtaining the unknowns in reverse order is called back-substitution VEL TECH VEL TECH MULTI TECH 31 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH These operations can be expressed more concisely in matrix form as follows: Working with the augmented matrix [A,B], the Gaussian elimination process is 1 2 6 0 1 2 6 0 1 2 6 0 2 2 3 3 0 6 9 3 0 6 9 3 (4) 1 3 0 2 0 1 6 2 0 0 15 / 2 3 / 2 Which, upon back-substitution, yields x3 1 5 x2 4 5 ww x1 2 . 5 (5) 7. B. Use the Gaussian elimination method to solve the simultaneous equations. w.E 4x1 + 2x2 – 2x3 – 8x4 = 4 x1 + 2x2 + x3 =2 0.5x1 – x2 + 4x3 + 4x4 = 10 -4x1 - 2x2 – x4 = 0 asy En Solution: gin 2 2 8 x1 4 4 1 2 1 0 x 2 2 0.5 1 4 4 x3 10 1 x 4 0 4 2 0 (a) eer i ng. net Divide row 1 by 4. Subtract the new row 1 from row. Multiply the new row 1 by 0.5 and subtract it from row 3. Multiply row 1 by – 4 and subtract it from row 4. The result is 1 0.5 0.5 2 x1 1 0 1.5 1.5 2 x 2 1 0 1.25 4.25 5 x3 9.5 0 2 7 0 x 4 4 Divide row 2 by 1.5. Multiply the new row 2 by – 1.25 and subtract it from row 3. A zero already appears in row 4, and no modification is required. The result is VEL TECH VEL TECH MULTI TECH 32 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 2 x1 1 1 0.5 0.5 0 1 1 1.3333 x 2 0.6667 0 0 5.5 6.6667 x3 10.3333 2 7 0 0 x 4 4 Divide row 3 by .5.5. Multiply the new row 3 by -2 and subtract it from row 4: 2 X1 1 1 0.5 0.5 0 1 1 1.3333 X2 0.6667 0 0 1 1.2121 X3 1.8788 0 4.5758 0 0 X4 7.7576 ww w.E Divide row 4 by – 4.5758 and solve for the unknowns by substitution: X1 = 0.0794 x2 = 1.0066 x3 = 3.9338 x4 = -1.6954 asy 8.a. Explain the term convergence & briefly explain about its types. En Convergence refers to this process, where we optimize the mesh to arrive at the desired results. In general, there are three main types of convergences: i) gin eer i Von-Misses Stress (VMS) convergence: Mesh is refined until the percentage variation in VMS is less than 1,5,10 or any given value selected by the user. VMS convergence should be avoided if there are stress concentrations points, convergence will be difficult to achieve. ng. net ii) Strain Energy Convergence: Mesh is refined until the percentage variation in the average strain of elements is less then a chosen value. Strain convergence is a better criterion for optimizing an FEA mesh. Stress concentrations points do not significantly influence the average strain energy of elements and variation in strain energy is influenced by mesh size or polynomial order of the elements only. iii) Deflection Convergence: It is similar to the above convergences, except, node deflection values are used for the convergence criterion. 8.b. How does FEA works within the software? The following steps can summarize FEA procedure that works inside software: i) Using the user’s input, the given structure is graphically divided into small elements (sections or regions) so that every element’s mechanical behaviour can be defined by as set of differential equations. VEL TECH VEL TECH MULTI TECH 33 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH ii) The differential equations are converted into algebraic equation, and then into matrix equations, suitable for a computer-aided solution. iii) The element equations are combined and a global structure equation is obtained. iv) Appropriate load and boundary conditions, supplied by the user, are incorporated into the structure matrix. v) The structure matrix is solved and deflections of all the nodes are calculated. vi) A node can be shared by several elements and the deflection at the shared node represents deflection of the sharing elements at the location of the node. vii) ww Deflection at any other point in the element is calculated by interpolation of all the node points in the elements. viii) w.E An element can have linear or higher order interpolation function. asy The individual element matrix equations are assembled into a combined structure equation, {F} = [k] {u}. En As defined earlier gin eer i {F} = Column matrix of the externally applied loads. [k] = Stiffness matrix of the structure, which is always a symmetric matrix. analogues to n equivalent spring constant of several of connected springs. This matrix is ng. net {u} = Column matrix representing the deflection of all the node points, that results when the load {F} is applied. 9. Explain in detail about various stress-strain relation ship. Liner elastic materials, the stress-strain relations come from the generalized Hooke’s for isotropic materials, the two material properties are Young’s modulus (or models of elasticity) E and Poisson’s ratio v. Considering an elemental cube inside the, Hooke’s law gives Ex x v E E y v x Ez v VEL TECH E x E y v E y E v z y z E v z E E (1) E VEL TECH MULTI TECH 34 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Txz G T xz xz G T xy xy G yz The shear modulus (or modulus rigidity), G, is given by G E 2(1 v) (2) ww From Hooke’s law relationships (Eq.(1), node that Ex Ey Ez w.E (1 2v) ( x y z ) E asy (3) Substituting for ( y z ) and so on into Eq. 1, we get the inverse relations En DE (4) D is the symmetric (6 x6) material matrix given by gin v 0 0 0 1 v v v 1 v v 0 0 0 v v 1 v 0 0 0 E D (5) (1 v)(1 2v) 0 0 0 0.5 v 0 0 0 0 0 0 0.5 v 0 0 0 0 0 0.5 v 0 Special Cases eer i ng. net One dimension. In one dimension, we have normal stress along x and the corresponding normal strain. Stress-strain relations (Eq.4) are simply = E (6) Two dimensions. In two dimensions, the problems are modeled as plane stress and plane strain. Plane Stress. A thin planar body subjected to in-plane loading on its edge surface is said to be in plane stress. A ring press fitted on a shaft, Fig. a, an example. Here stresses z, Tyz are set as zero. The Hooke’s law relations (Eq.1) then give us VEL TECH VEL TECH MULTI TECH 35 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH x x E y v v x VEL TECH MULTI TECH VEL TECH HIGH TECH y E y E E 2(1 v) xy Txy E v z ( x y ) E (7) ww w.E asy En gin eer i ng. Inverse relations are given by x 1 v 0 x 0 y y v 1 1 v xy Txy 0 0 2 Which is used as net (8) D . Plane strain. If a long body of uniform cross section is subjected to transverse banding along its length, a small thickness in the loaded area, as shown in Fig. b, can are treated as subjected to plane strain. Here z, zx, yz are taken as zero. Stress may not be zero in this case. The stressstrain relations can be obtained directly from Eqs., and , : VEL TECH VEL TECH MULTI TECH 36 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH x 1 v v E 1 v y v (1 v)(1 2v) xy 0 0 0 x 0 y 1 v xy 2 VEL TECH HIGH TECH (9) D here is a (3x3) matrix, which relates stresses and three strains. Anisotropic bodies, with uniform orientation, can be considered by using the appropriate D matrix for the material. 10. ww The total potential energy is given by w.E 1 1 1 1 k1 12 k 2 2 2 k 3 3 2 k 4 4 2 F1q1 F3 q3 2 2 2 2 asy where 1, 2 , 3 , and 4 are extensions of the four springs. En Since 1 q1 q2 , 2 q2 , 3 q3 q2 , and 4 q3 , we have gin 1 1 1 1 k1(q1 q2 )2 k 2q2 2 k 3 (q3 q2 )2 k 4 q3 2 F1q1 F3 q3 2 2 2 2 eer i where q1,q2 and q3 are the displacements of nodes 1,2 and 3, respectively. ng. net Figure. A. For equilibrium of this three degrees of freedom system, we need to minimize with respect to q1,q2, and q3. The three equations are given by 0 qi i 1,2,3 VEL TECH (1) VEL TECH MULTI TECH 37 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Which are k1(q1 q2 ) F1 0 q1 k1 q1 q2 k 2q2 k 3 (q3 q2 ) 0 q2 k 3 (q3 q2 ) k 4q4 F3 0 q3 These equilibrium equations can be put in the form of kq = F as follows: ww k1 k1 k k k k 3 1 1 2 0 k 3 0 q1 F1 k 3 q2 0 k 3 k 4 q3 F3 (2) w.E asy If , on the other hand, we proceed to write the equilibrium equations of the system by considering the equilibrium of each separate node, as shown in Fig. b, we can write. En 11. gin The potential energy for the elastic one-dimensional rod (Fig), with body force neglected, is 2 1 du EA dx 2u1 20 dx L eer i ng. Where u1 = u(x = 1). Let us consider a polynomial function U = a1 + a2x + a3x2 net This must satisfy u = 0 at x = 0 and u = 2. Thus, 0 = a1 0 = a1 + 2a2 + 4a3 Hence, A2 = -2a3 U = a3 (-2x + x2) VEL TECH u1 = -a3 VEL TECH MULTI TECH 38 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH ww w.E asy En gin Figure. E eer i 12. Give the brief explanation of FEA for a stress analysis problem. ng. net The steps involved in finite elements analysis are clarified by taking the stress analysis of a tension strip with fillets (refer Fig.). In this problem stress concentration is to be studies in the fillet zone. Since the problem is having symmetry about both x and y axes, only one quarter of the tension strip may be considered as shown in Fig. . About the symmetric axes, transverse displacements of all nodes are to be made zero. The various steps involved in the finite element analysis of this problem are discussed below: Step 1 : Four nodded isoparametric element (refer Fig) is selected for the analysis (However note that 8 noded isoparametric element is ideal for this analysis). The four noded isoparametric element can take quadrilateral shape also as required for elements 12,15,18, etc. As there is no bending of strip, only displacement continuity is be ensured but not the slope continuity. Hence displacement of nodes in x and y directions are taken as basic unknowns in the problem. VEL TECH VEL TECH MULTI TECH 39 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH ww w.E asy En gin eer i ng. Step 2 : The portion to be analyzed is to be discretised. Fig. shows discretised portion. For this 33 elements have been used. There are 48 nodes. At each node unknowns are x and y components of displacements. Hence in this problem total unknowns (displacements) to be determined are 48 x 2 = 96. net Step 3 : The displacement of any point inside the element is approximated by suitable functions in term of the nodal displacements of the elements. For the typical element (Fig.b), displacements at P are u Nu i i N1u1 N2u2 N3u3 N4u4 (1) and v Nu i i N1u1 N2u2 N3u3 N4u4 The approximating functions Ni are called shape functions or interpolation functions. Usually they are derived using polynomials. Step 4 : Now the stiffness characters and consistent loads are to be found for each element. There are four nodes and at each node degree of freedom is 2. Hence degree of freedom in each element VEL TECH VEL TECH MULTI TECH 40 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH is 4 x 2 = 8. The relationship between the nodal displacements and nodal forces is called elements stiffness characteristics. It is of the form ke e Fe' as explained earlier. For the element under consideration, ke is 8 x 8 matrix and Fe are vectors of 8 values. In solid mechanics element stiffness matrix is assembled using variational approach i.e. by minimizing potential energy. If the load is acting is the body of element or on the surface of element, its equivalent at nodal points are to be found using variational approach, so that right hand side of the above expression is assembled. This process is called finding consistent loads. ww Step 5 : The structure is having 48 x 2 = 96 displacement and load vector components to be determined. Hence global stiffness equation is of the form [k] 96 x 96 {} w.E 96 x 1 = {F} asy 96 x 1 En Each element stiffness matrix is to be placed in the global stiffness matrix appropriately. This process is called assembling global stiffness matrix. In this problem force vector F is zero at all nodes except at nodes 45, 46, 47 and 48 in x direction. For the given loading nodal equivalent forces are found and the force vector F is assembled. gin eer i Step 6 : In this problem, due to symmetry transverse displacements along AB and BC are zero. The system equation [k] {} = {F} is modified to see that the solution for {} comes out with the above values. This modification of system equation is called imposing the boundary conditions. ng. net Step 7: The above 96 simultaneous equations are solved using the standard numerical procedures like Gausselimination or Choleski’s decomposition techniques to get the 96 nodal displacements. Step 8: Now the interest of the analysis to study the stresses at various points. In solid mechanics the relationship between the displacements and stresses are well established. The stresses at various points of interest may be found by using shape functions and the nodal displacements and then stresses calculated. The stress concentrations may be studies by comparing the values at various points in the fillet zone with the values at uniform zone, far away from the filet (which is equal to P/b2 t). 13. Determine the displacement and stress in a bar of uniform cross section due to self weight only when held as shown in Fig. Use (i) two terms (ii) three terms, for approximating polynomial. Verify the expression for total extension with the exact value. VEL TECH VEL TECH MULTI TECH 41 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH ww Figure. w.E Solution: Let ‘p’ be unit weight and E Young’s modulus of the material of the bar. If A is the cross section of the bar then, U asy 1 T dV 2 v En T 1 du du E A dx 2 dx dx 0 l 2 1 du EA dx 20 dx l gin (1) We u Xb dV T v and (2) l u A dx 0 eer i ng. net Polynomial function for displacements may be taken as U = a0 + a1x + a2x2 + a3x3 + < + an xnn The boundary condition to be satisfied is At X = 0, u = 0 From this we get 0 = a0 u a1x a2 x 2 a3 x 3 ... an x n VEL TECH VEL TECH MULTI TECH 42 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH (i)When only two terms of polynomial equation are used, u a1x du a1 dx 2 1 1 1 du U EA dx EA a12dx EA a12l 20 20 2 dx l l l Wp u A dx 0 l x2 a1x A dx Aa1 2 0 0 l ww Aa1 l2 2 w.E 1 l2 2 EA a1 l A a1 2 2 asy From minimization condition, we get d l2 0 i.e.,0 EAa1l A da1 2 or and a1 l 2E l u x 2E du l E dx 2E En (4) gin (5) (6) eer i ng. net The displacement and stress variations are shown in Fig. Extension of the bar = u1 – u0 (a) VEL TECH VEL TECH MULTI TECH 43 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Fig. (a) Variation of displacement (b) Variation of stress ww (ii)When three terms are considered for displacement in equation 3 : u a1x a2 x 2 U Wp 2 du a1 2a2 x dx w.E asy 1 du EA dx u A dx 20 dx 0 l l l l En 1 2 EA a1 2a2 x dx A a1x a 2 x 2 dx 2 0 0 gin eer i l l a x2 a x3 1 x2 x3 EA a12 4a1a2 4a2 2 A 1 2 2 2 3 0 3 0 2 ng. l2 1 4 2 3 l3 2 2 EA a1 l 2a1a2l a2 l A a1 a 2 2 3 3 2 d 1 1 EA 2a1l 2a2l2 A 0 da1 2 2 l 2E d 1 2 8 3 l3 0 EA 2a1l a2l A 0 da2 2 3 3 4 l i.e., a1 a2l 3 3E From equation 8 and 9 we get, 1 l l l a2l 3 3E 2E 6E a2 2E i.e., VEL TECH a1 a2l VEL TECH MULTI TECH 44 net (8) (9) VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Substituting it in equation 8, we get a1 l l l 2E 2E E u a1x a2 x 2 E l E x 2E x2 lx E 2 x2 du l x dx The variations of displacements and stress in this case also are shown in Fig. ww Extension of the bar u1 u0 2 l2 l2 l E 2 2E w.E asy Actual extension of the bar [refer Fig.] En gin eer i ng. net Figure. l Px dx l Axdx x 2 l2 AE 0 AE E 2 0 2E 0 l Thus total extension of the bar obtained is exact in both the cases. VEL TECH VEL TECH MULTI TECH 45 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH UNIT – II PART – A 1. Use the lagrangian interpolation formula for deriving one-dimensional three-node shape functions for the element illustrated in the below fig specialize the shape function for an element of length L with a note at its center. ww w.E Sub the values of xi,xj,xk in lagrange polynomial formula. n LK m0 K m asy (x x0 )(x x1 ).............(x xk 1 )(x xk 1 )......(x xn ) x xm We get, xk xm (xk x0 )(xk x1 ).....(xk xk 1 )(xk xk 1 )....(xk xn ) Li Ni En (x x j )(x xk ) (xi x j )(xi xk ) L j Nj Lk Nk gin (x xi )(x xk ) (x j xi )(x j xk ) (x xi )(x x j ) (xk xi )(xk x j ) eer i Let xi=0, xj=L/2 and xk=L and let corresponding node numbers ,12, and 3: N1 (x L / 2)(x L) x(x L) ;N2 ( L / 2)( L) (L / 2)(L / 2 L) N3 x(x L / 2) L(L L / 2) ng. net 2. Define shape function. In the finite element analysis aims is to find the field variables at nodal points by rigorous analysis, assuming at any point inside the element basic variable is a function of values at nodal points of the element. This function which relates the field variable at any point within the element to the field variables of nodal points is called shape functions. VEL TECH VEL TECH MULTI TECH 46 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 3. Derive a one-dimensional linear interpolation formula for a function u=u(x) that is valid in the range u1 through u2 as shown in fig. The function u(x) is shown in above figure. A linear equation that would approximate u(x) between u1 and u2 is assumed as u= A+Bx (a) where A and B are constants, substituting the boundary conditions u(x 1)= u1and u(x2)=u2 gives two equations that can be solved for A and B. ww u1=A+Bx1 u2=A+Bx2 w.E Solving for A and BB substituting into Eq. (a) gives the interpolating polynomial: A u1x 2 u2 x1 x3 x1 u u1 B= asy u2 u1 x 2 x1 (b) x2 x x x1 u2 x 2 x1 x 2 x1 En gin 4. Derive the shape functions for a one-dimensional linear finite element. u x u 2 x1 x 2 x1 The results of A= 1 2 B ng. u2 u1 x 2 x1 u u1 eer i x2 x x x1 u2 x 2 x1 x 2 x1 net Can be used to derive the shape function at node 1 is the coefficient of u 1 in the above said equation or N1=(x2-x)/(x2-x1). Similarly the shape function for node 2 is N2=(x-x1). Note that N1=1 for x =x1 and N1=0 for x=x2. N2 is zero at node 1 and 1 at node 2. 5. Derive a linear one-dimensional interpolation formula in terms of shape functions and nodal point variables and write the result as a matrix equation. u x u 2 x1 x 2 x1 The interpolation formula A= 1 2 B u2 u1 ; x 2 x1 u u1 VEL TECH x2 x x x1 u2 x 2 x1 x 2 x1 VEL TECH MULTI TECH 47 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH and the corresponding shape functions N1=(x2-x)/(x2-x1) N2=(x-x1)/(x2-x1) And using the notations of the below fig results in u=N1u1+N2u2 ww [N]=[N1 N2] {u}=[u1 u2]T w.E u=[N]{u} asy 6. Write down three types of loading used in one dimensional problems. (i) (ii) (iii) body force (f) Traction force (T) Point load (Pi) En gin eer i 7. Briefly explain about finite element modeling in one dimensional problems. ng. Major steps are i) element division ii) Node numbering scheme Element division: net The first step is to model the bar as stepped shaft, consisting of a discrete number of elements, each having a uniform cross section. Specifically, let us mode the bar using four finite elements. A simple scheme for doing this is to ivied the bar into four regions, as shown in figure. the average cross-sectional area within each region is evaluated and then used to define an element with uniform cross section. The resulting four-element, five node finite element model is shown in fig. In the finite element mode, every element connects to two nodes. In fig the element numbers are circled to distinguish them from one numbers. In additional to the cross section, traction and body forces are also (normally0 treated at constant within each element. However, cross-sectional area, traction and body forces can differs in magnitude from element to element. Better approximation are obtained by VEL TECH VEL TECH MULTI TECH 48 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH increasing the number of elements. It is convenient to define a node at each locations where a point load is applied. For easy implementation, an orderly numbering scheme for the model has to be adopted. ww w.E asy En In a one-dimensional problem, every node is permitted to displace only in the x direction. Thus, each node has only one degree of freedom (dof). The five-node finite element model in fig has five dofs. The displacements along each dof are denoted by Q1,Q2<<.Q5. In fact, the column vector Q=[Q1,Q2,<<.Q5]T is called the global displacement vector. The global load vector is denoted by F=[F1,F2,<<F5]T. The vectors Q and F are shown in fig. The sign convention used is that a displacement or load has a positive value if acting along the +x direction. At this stage, conditions at the boundary are not imposed. For example, node 1 in fig is fixed, which implies Q1=0. these conditions are discussed later. gin eer i ng. net Each element has two nodes; therefore the element connectivity information can be conveniently represented as shown in fig. further the element connectivity table is also given. In the connectivity table, the headings 1 and 2 refer to local node numbers of an element, and the corresponding node number on the body are called global numbers. Connectivity thus establishes the local –global correspondence. In this simple example, the connectivity can be easily generated since local node 1 is the same as the element number e, and local node 2 is e+1. Other ways of numbering nodes or more complex geometries suggest the need for a connectivity table. The connectivity is introduced in the program using the array NOC. VEL TECH VEL TECH MULTI TECH 49 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH The concepts off dof, nodal displacement, nodal loads and element connectivity are central to the finite element method and should be clearly understood. ww w.E 8. What are all the conditions to be satisfied in general shape function? (i) (ii) (iii) First derivative must be finite within an element Displacement must be continuous across the element boundary. Rigid body motion should not be introduce at any stresses in the element. asy En 9. Briefly explain about one-dimensional potential energy approach method. gin The general expression for the potential energy 1 is 1 T Adx uT f Adx uT Tdx ui Pi L L 2 L i ...(1) eer i ng. The quantities ,u,f and T in equation (1). In the last term, Pi represents a force acting at point I, and ui is the x displacement at that point. The summation on I gives the potential energy the potential energy due to all point load. net Since the continuum has been discretized into finite elements, the expression for becomes T T T 1 A dx u fA dx- u T dx- QiPi e e e e 2 e e i ...(2) the last term in Eq. assumes that point loads Pi are applied at the nodes. This assumption makes the present derivation simpler with respect to notation and is also a common modeling practice, Equation can be written as Ue u fA dx- u T dx- QP i i T e e e Ue = T e e ...(3) i 1 T A dx 2 is the element strain energy. VEL TECH VEL TECH MULTI TECH 50 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 10. Briefly explain about Galerkin Approach. Let us introduce a virtual displacement field =(x) <(1) and associated virtual strain ( ) d dx .... (2) Where ia an arbitrary or virtual displacement consistent with the boundary conditions. Galerkin’s variational form, given in equation for the one dimension problem considered here is, ww w.E ( )Adx fAdx Tdx P 0 T T L L T L ...(3a) i i this equation should hold for every consistent with the boundary condition. The first term represent the internal virtual work, while the load terms represents the external virtual work. asy On the discretized region, EQ.(3) becomes En gin E ( )A dx fAdx Tdx P 0 T e T e e T e e ...(3b) i e i eer i Note that is the strain due to the actual loads in the problem, while is the strain due to the actual loads in the problem, while () is a virtual strain. Similar to the interpolation steps in Eqs. We express N ( ) B ng. <(4) net Where =[ 1, 2]T represents the arbitrary nodal displacement of element e. Also, the global virtual displacements at the nodes are represented by [1, 2 ,..... N ]T ....(5) 11. Write a short note on element stiffness: Consider the first term, representing internal virtual work in the equation. E ( )Adx fAdx dx P 0 T e e VEL TECH T e e T e e i i ..(1) i VEL TECH MULTI TECH 51 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH substituting the values of =N & ()=B in the above equation & nothing that =Bq then we get, E ( )Adx q B EB Adx T T e T ..(2) e in the finite element model, the cross sectional area of element e, denoted by Ae, is constant. Also, B is a constant matrix, Further, dx =( ( e / 2)d. thus, e 1 E ( )A dx=q [E A 2 B B d ] ...(3a) =qT k e ...(3b) T T e e T e 1 ww = k q T e Where Ke is the (symmetric) element stiffness matrix given by w.E Ke E Ae eBTB ...(4) asy substituting B from eq. we have ke Ee A e 1 1 e 1 1 En gin eer i 12. How the global stiffness matrix can be assembled in one dimensional problem: The total potential energy written in the form 1 qT k e q qT f e qT f e qT T e PQ i i e 2 e e e i <(1) can be written in the form = ng. net 1 T Q KQ QTF 2 by taking element connectivity into account. This step involves assembling K and F from element stiffness and force matrices. The assembly of the structural stiffness matrix K from element stiffness matrices ke+ will first be shown here. Referring to the finite element model in fig. (0.1b) let us consider the strain energy in say, element 3. We have VEL TECH VEL TECH MULTI TECH 52 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH 1 T 3 qk q 2 U3 = VEL TECH HIGH TECH (0.1 a) or, substituting for k3, U3 = 1 T E3 A 3 1 1 q q 2 3 1 1 (o.1b) For, element 3, we have q = [Q3, Q4]T. Thus, we can write U3 as 0 0 1 0 U3 Q1,Q2 ,Q3 ,Q 4 ,Q5 2 0 0 ww 0 0 0 0 E3 A 3 3 0 0 E3 A 3 3 0 E3 A 3 3 E3 A 3 3 0 0 0 0 w.E 0 Q1 0 Q 2 0 Q3 0 Q4 Q5 0 asy (0.3) En From the previous equations, we see that elements of the matrix k3 occupy the third and forth rows and columns of the K matrix. Consequently, when adding element-strain energies, the elements of ke are placed in the appropriate locations of the global K matrix, based on the element connectivity; overlapping elements are simply added. We can denote this assembly symbolically as K ke (0.4a) gin eer i ng. e net Similarly, the global load vector F is assembled from element-force vectors and point loads are F (f e + T e ) P (0.4b) e The Galerkin approach also gives us the same assembly procedure. 13. List shown the Properties of global stiffness matrix (K). 1. The dimension of the global stiffness K is (N X N), where N is the number of nodes. This follows from the fact that each node has only one degree of freedom. 2. K is a symmetric. 3. K is a banded matrix. That is, all elements outside of the band are zero. This can be seen in Example, just considered. In this example, K can be compactly represented in banded form as VEL TECH VEL TECH MULTI TECH 53 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH A1 1 A1 A 2 1 2 A A Kbanded = E 2 3 2 3 A A 3 4 4 3 A 4 4 VEL TECH HIGH TECH A1 1 A2 2 A 3 3 A 4 4 0 ww 14. List out the steps involved in elimination approach. w.E Elimination Approach: Consider the boundary conditions asy Qp 1 = a1, Qp 2 , = a2 , ...., Qp r = ar En gin Step 1: Store the p1th, p2th,<<and prth rows of the global stiffness matrix K and force vector F. These rows will be used subsequently. eer i Step 2: Delete the p1th row and column, the p2th row and column, <<.,and the prth row and column from the K matrix. The resulting stiffness matrix K is of dimensions (N – r, N – r). Similarly, the corresponding load vector F is of dimension (N – r, 1). Modify each load component as Fi = Fi - (Ki ,p 1 a1 + K i ,p 2 a2 + ........+ K i,p r ar ) ng. (1) For each dof i that is not a support. Solve KQ = F net For the displacement vector Q. Step 3: For each element, extract the element displacement vector q from the Q vector, using element connectivity, and determine element stresses. Step 4: Using the information stored in step 1, evaluate the reaction forces at each support dof from VEL TECH VEL TECH MULTI TECH 54 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Rp1 = K p11Q1 + K p1 2Q2 + ............+K p1NQN - Fp1 Rp2 = K p2 1Q1 + K p2 2Q2 + ............+K p2NQN - Fp2 ....................................................................... Rpr = K pr 1Q1 + K pr 2Q2 + ............+K pr NQN - Fpr (2) 15. Write down the steps involved in penalty approach. Penalty Approach: Consider the boundary conditions ww Qp 1 = a1, Qp 2 , = a2 , ...., Qp r = ar w.E Step 1: Modify the structural stiffness K by adding a large number C to each of the p1th, p2th, <<, prth diagonal elements of K. Also, modify the global load vector F by adding Ca1 to Fp , Ca2 asy 1 to Fp , <<., and Car to Fp , . Solve KQ = F for the displacement Q, where K and F are the modified 2 r En stiffness and load matrices. Step 2: For each element, extract the element displacement vector q from the Q vector, using element connectivity, and determine the element stresses. gin eer i Step 3: Evaluate the reaction force at each support from Rpi = -C(Qpi - ai ) i = 1, 2, ..........r. (1) ng. 16. State minimum potential – energy theorem. net Of all possible displacements that satisfy the boundary conditions of a structural system, those corresponding to equilibrium configurations make the total potential energy assume a minimum value. 17. Write down the general properties of Shape functions. 1 0 i. A shape function has a value of either 1 or 0 at a nodal point : Ni j ij = ie i = j if i j ii. The sum of all shape functions at any point equals 1= Ni () = 1. 18. Write the characteristics of Truss elements. Truss is a slender member (length is much larger than the cross-section). It is a two-force member i.e. it can only support an axial load and cannot support a bending load. Members are joined by pins (no translation at the constrained node, but free to rotate in any directions). VEL TECH VEL TECH MULTI TECH 55 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH The cross-sectional dimensions and elastic properties of each member are constant along its length. The element may interconnect in a 2-D or 3-D configuration in space. The element is mechanically equivalent to a spring, since it has no stiffness against applied loads except those acting along the axis of the member. However, unlike a spring element, a truss element can be oriented in any direction in a plane, and the same element is capable of tension as well as compression. 19. Write down the characteristic features of global stiffness matrix of a truss element. ww The matrix is symmetric Since there are 4 unknown deflections (DOF), the matrix size is a 4 x 4. The matrix represents the stiffness of a single element. The terms c and s represent the sine and cosine values of the orientation of element with the horizontal plane, rotated in a counter clockwise direction (positive direction). w.E asy 20. Write a short note on Treatment of loads in FEA. En For a truss element, loads can be applied on a node only. If loads are distributed on a structure, they must be converted to the equivalent loads can be applied at nodes. Loads can be applied in any direction at the node, however, the element can resist only the axial component, and the component perpendicular to the axis merely causes free rotation at the joint. gin eer i ng. VEL TECH VEL TECH MULTI TECH 56 net VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH PART – B 1. i) (i) Give the relation ship between the following: a). Local and global deflections. b). Local and global forces. (ii) Explain in detail: a). Finite element equation in local co-ordinate system. b). Finite element equation in global co-ordinate. ww (a) Relationship Between Local and Global Deflections: Let us consider the truss member, shown in figure. The element is inclined at an angle , in a counter clock wise direction. The local deflections are 1 and 2. The global deflections are: u1, u2, u3, and u4. We wish to establish a relationship between these deflections in terms of the given trigonometric relations. w.E asy En gin eer i Figure: Local and Global Deflections By trigonometric relations, we have, 1 u1x cos + u2 sin c u1x s u1y ng. net 1 u2x cos + u2y sin c u2x s u2y where, cos = c, and sin =s Writing the above equations in a matrix form, we get, u1x 1 c s 0 0 u1y = 0 0 c s u2x 1 u 2y VEL TECH VEL TECH MULTI TECH 57 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH T u Or, in short form, Along with equation (3, 2), we also need an equation that relates the local and global forces. i) (a) Relationship Between Local and Global Forces: By using trigonometric relations similar to the previous section, we can derive the desired relationship between local and global forces. However, it will be easier to use the work-energy concept for this purpose. The forces in local coordinates are : R1 and R2, and in global coordinates: f1, f2, f3, and f4, see figure for their directions. ww Since, work done is independent of a coordinate system, it will be the same whether we use a local coordinate system or a global one. Thus, work done in the two systems is equal and given as, w.E asy W = T R = uT f, or in an expanded form, R W = 1 2 1 = u1x R2 = T R u1y u2x f1 f u2y 2 f3 f4 En gin = u f T eer i ng. T u R T T T T Substituting = T u in the above equation, we get, u T R u f .dividing by u T T R = f T Equation can be used to convert local forces into global forces and vice versa. net Figure: Local and Global Forces VEL TECH VEL TECH MULTI TECH 58 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH ii) (a) VEL TECH MULTI TECH VEL TECH HIGH TECH Finite Element Equation in Local Coordinate System: Now we will derive the finite element equation in local coordinates system. This equation will be converted to global coordinate system, which can be used to generate a global structural equation for the given structure. Note that, we can not use the element equations in their local coordinate form; they must be converted to a common coordinate system, the global coordinate system. Consider the element shown below, with nodes 1 and 2, spring constant k, deflections 1 and 2 and forces R1 and R2. As established earlier, the finite element equation in local coordinates is given as, ww R1 k -k 1 = R2 k k 1 1 1, R1 w.E figure: Truss Element k 2 asy 2, R2 Recall that, for a truss element, K = AE/L En Let ke = stiffness matrix in local coordinates, then, AE / L AE / L ke = AE / L AE / L ii) (b) gin Stiffness matrix in local coordinates. eer i Finite Element Equation in Global Coordinates: ng. net Using the relationship between local and global deflections and forces, we can convert an element equation from a local coordinate system to a global system. Let kg = Stiffness matrix in global coordinates. In local system, the equation is: R= [kg] {} (A) We want a similar equation, but in global coordinates. We can replace the local force R with the global force f derived earlier and given by the relation: { f } = [TT] {R} VEL TECH VEL TECH MULTI TECH 59 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Replacing R by using equation (A), we get, { f } = [TT] [[ke] {}], and can be replaced by u, using the relation = [T] {u}, therefore, { f } = [TT] [ke] [T] uR} { f } = [kg] {u} ww Where, [kg] = [TT] [ke] [T] w.E Substituting the values of [T]T, [T] and [Ke], we get, c s [Kg] 0 0 asy 0 0 AE / L AE / L C S 0 0 c AE / L AE / L 0 0 C S s Simplifying the above equation, we get En gin c2 cs c 2 cs cs s2 cs s2 K g 2 (AE / L) c cs c 2 cs 2 cs s2 cs s eer i ng. 2. Example For the truss structure shown: net find displacements of joints 2 and 3. find stress, strain & internal force in each member. AAL= 200 mm2, AST=100 mm2 VEL TECH VEL TECH MULTI TECH 60 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH All other dimensions are in mm. Solution: Let the following node pairs form the elements Element (1) (2) (3) Node Pair 1-3 2-1 2-3 ww Using Shigley’s Machine Design book for yield strength values, we have w.E S y( AL) 0.0375kN / mm2 (375Mpa) S y(ST) 0.0586kN / mm2 (586 Mpa) asy E(AL) 69kN / mm2 ,E(ST) 207kN / mm2 A (1) A (2) 200mm ,A 2 (3) 100mm 2 Find the stiffness matrix for each element En gin eer i ng. Element (1) L(1)= 260 mm, E(1)=69kN/mm2 A(1)=200mm2 =0 c=cos =1, s=sin=0, CS=0 net c2=1 s2=0 EAL/L=69 kN/mm2 x 200 mm2 x 1/(260mm)=53.1 kN/mm VEL TECH VEL TECH MULTI TECH 61 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH c2 cs c 2 cs (1) cs s2 cs s2 K g (AE / L)x 2 c cs c 2 cs 2 cs s2 cs s 1 0 (1) K g (53.1)x 1 0 0 1 0 0 0 1 0 0 0 0 0 0 Element 2: ww =90 c=cos 90 =cos 0=1,s2=1 cs=0 EA/L=69 x 200(1/150)=92 kN/mm w.E u2x u2y u1x u1y 0 0 0 1 [kg](2) (92) 0 0 0 1 0 0 u2x 0 1 u2y 0 0 u1x 0 1 u1y asy En gin Element 3 eer i ng. net =30 c=cos 30 =0.866,c2=0.75 s=cos 60 =.5,s2=0.25 cs=0.433 EA/L=207 x 100 (1/300)=69 kN/mm VEL TECH VEL TECH MULTI TECH 62 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH u2x u2y u3x VEL TECH HIGH TECH u3y u2x .75 .433 75 433 .433 .25 u .433 2.5 2y [kg](3) (69) u3x .75 .433 .75 .433 u3y .433 2.5 0.433 .25 Assembling the stiffness matrices: Since there are 6 deflections (or DOF), u1 through u6, the matrix is 6 x6. Now, we will place the individual matrix element from the element stiffness matrices into the global matrices into the global matrix according to their position of row and column members. ww Element [1] u1x u1y u2x w.E u2y u3x u3y 53.1 u1x 53.1 u1y u2x u2y u3x -53.1 u3y 53.1 asy En The blank spaces in the matrix have a zero value u1x u1x u1y u2x u2y u3x u3y u1y u2x u2y 92 -92 -92 92 u3x gin u3y eer i ng. net Element [3] u1x u1x u1y u2x u2y u3x u3y u1y u2x u2y u3x u3y 51.7 29.9 -51.7 -29.9 29.9 17.2 -29.9 -17.2 -51.7 -29.9 51.7 29.9 29.9 17.2 29.9 17.2 VEL TECH VEL TECH MULTI TECH 63 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Assembling all the terms for elements [1],[2] and [3], we get the complex matrix equation of the structure. u1x u1y u2x u2y u3x u3y 0 u1x F1 0 u1y F1 -29.9 u2x F1 -17.2 u2y F1 29.9 u3x F1 17.2 u3y F1 0 0 53.1 53.1 0 0 92 0 -92 0 0 0 51.7 29.9 -51.7 0 -92 29.9 109.2 29.9 -53.1 0 -51.7 -29.9 104.8 0 -29.9 -17.2 29.9 0 ww Boundary conditions: w.E Node 1 is fixed in both x and y directions, where as, node 2 is fixed in x-direction only and free to move in the y-direction. Thus, U1x=u1y=u2x=0 asy En Therefore, all the columns and rows containing these elements should be set to zero. The reduced matrix i: gin 109.2 29.9 17.2 u2y 0 29.9 104.8 29.9 u 0 3x 17.2 29.9 17.2 u3y 0.4 eer i Writing the matrix equation into algebraic linear equations, we get 29.9u2y -29.9u3x-17.2u3y= 0 -29.9u2y+104 u3x+29.9u3y=0 -17.2u2y+ 29.9 u3x+17.2 u3y = -0.4 ng. net solving we get u2y= 0.0043 u3x =0.0131 u3y= 0.0502 Stress, strain and deflections. Element (1) VEL TECH VEL TECH MULTI TECH 64 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Note that u1x, u1y, u2x etc are not coordinates, they are actual displacements. L u3x 0.0131 L / L 0.0131111/ 260 5.02 10-5mm / mm E 69 5.02 105 0.00347kN/ mm2 reaction R= A 0.00347 kN L u3x 0.0043 L / L 0.00043 /150 2.87 10 -5mm / mm E 69 2.87 10 5 1.9803kN/ mm2 Reaction R= A (1.9803 10 3 )(200)=0.396 kN ww Element (3) w.E Since element (30 is at angle 30, the change into the length is found by adding the displacement components of nodes 2 and 3 along the element (at 30). Thus, asy L u3x cos 30 u3y sin30 u2y cos 60 =0.0131 cos 30-0.0502 sin30+0.0043 cos 60 =0.0116 En gin =L/L=-0.0116/300=-3.87 x 10 -5 3.87x10 5 mm / mm E 207 x 387.87 x 10 5 .0080kN / mmm 2 eer i ng. Reaction R A ( 0.0087)(100) 0. 0.800kN Factor of Safety: net Factor of safety ‘n’ is the ration of yield strength to the actual stress found in the part. Element(1) n= Sy Element(2) n= Element(3) n= Sy Sy 0.0375 10.8 0.00347 0.0375 18.9 0.00198 0.0586 7.325 0.0080 The lowest factor of safety is found in element (3), and therefore the steel bar is the most likely to fail before the aluminum bar does. VEL TECH VEL TECH MULTI TECH 65 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 3. Given; Elements 1 and 2 : Aluminum Element 3 : steel A(1) =1.5 in2 A(2)=1.0 in2 A(3)= 1.0 in2 ww Required : Find stresses and displacement using hand calculations: w.E asy En gin eer i ng. Solution: net Calculate the stiffness constants: AE 1.5 10 10h6 lb 7.5 10s L in 6 AE 1 10 10 lb K2 2.5 105 L 40 in 6 AE 30 10 10 lb K3 6.0x105 L 50 in K1 Calculate the element matrix equations VEL TECH VEL TECH MULTI TECH 66 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Element (1) Denoting the Spring constant for element (1) by k1, and the stiffness matrix by K(1), the stiffness matrix in global coordinates is given as, u1x u1y u2x u2y c cs c cs u1x (1) cs s2 cs s2 u1y K g K1 2 c cs c 2 cs u2x 2 cs s2 u2y cs s 2 2 ww w.E for element (1), =0 therefore c=1, c2=1 s=0,s2=0, and cs =0 u1x u1y u2x u2y 1 0 K (1) K1 1 0 0 0 0 0 asy 1 0 u1x 0 0 u1y 1 0 u2x 0 0 u2y En gin eer i ng. Element (2) For this element =90, therefore, net C=cos =0 , c2=0 S=sin =1, s2=1 Cs2=0 VEL TECH VEL TECH MULTI TECH 67 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH u3x u3y u2x VEL TECH HIGH TECH u2y c2 cs c 2 cs u3x (2) cs s2 cs s2 u3y K g K 2 2 c cs c 2 cs u2x 2 cs s2 u2y cs s u3x u3y u2x u2y 0 0 0 1 K (2) K 2 0 0 0 1 0 0 u3x 0 1 u3y 0 0 u2x 0 0 1 u2y ww Element 3 w.E For element (3), =126.9 C=cos (126.9)=-0.6,c2=3.6 S=sin (126.9,s2=.64 Cs=-0.48 u 4x u 4y u2x asy En gin eer i ng. u2y .36 .48 .36 .48 u4x .48 .64 .48 .64 u 4y K (3) K 3 .36 .48 .36 .48 u2x .48 .64 48 .64 u2y net Assembling the global Matrix Following the procedure for assembly described earlier, the assembled matrix is, 1 k1 2 0 3 K1 4 0 K g 5 0 6 0 7 0 8 0 VEL TECH 0 K1 0 0 0 K1 .36K 3 0 .48K 3 0 0 0 0 0 .36K 3 0 .48k 3 0 0 .48K 3 K 2 .64K 3 0 K 2 .48K 3 .64K 3 0 0 0 0 0 0 0 K 2 0 0 0 K2 0 0 0 00 0 0 .36K 3 .48K 3 0 0 .36K 3 .48K 3 0 0 .48K 3 .64K 3 0 0 .48k 3 .64K 3 VEL TECH MULTI TECH 68 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH the boundary conditions are: u1x=u1y=u3x=u3y=u4x=u4y=0 We will suppress the corresponding rows and columns. The reduced matrix is a 2 x2, given below. 48K 3 u2x K1 K g u .48K K 3 2 .64K 3 2y The final equation is, ww 48K 3 u2x 4000 k1 .36K 3 48K K 2 .64K 3 u2x 8000 3 w.E Substituting values for k1,k2 and k3, we get 9.66 2.88 u2x 4000 105 2.88 6.34 u2y 8000 asy u2x 0.0000438 in u2y 0.01241 in . En gin 1 P / A K i u / A1 [(7.5 105 )( 0.0000438)] /(1.5) 214 psi 2 P / A K 2 u / A 2 [(2.5 105 )( 0.012414)] /(1.0) 3015 psi eer i 1 P / A K 3 u / A 3 [(6 105 )( 0.0000438cos53.1 0.012414sin53.1)] /(1.0) 6119 psi 4. Evaluate the following. Referring to fig do the following : ng. net (a) Evaluate ,N1 and N2 at point P. (b) If q1=0.003 in. and q2 =- -0.005 in. determine the value of the displacement q at point p. Solution: (a) Using Eq: the coordinate of point P is given by VEL TECH VEL TECH MULTI TECH 69 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH p= VEL TECH MULTI TECH VEL TECH HIGH TECH 2 (24 20) 1 16 =-.0.5 Now Eqs. and yield N1=0.75 and N2=0.25 (b) using Eq. we get, up =0.75(0.003)+0.25(-00.0005) =0.001 in. ww w.E The strain-displacement relation in equation in Eg is du dx asy Upon using the chain rule of differentiation, we obtain du d d dx (1) En gin From the relation between x and in equation , we have d 2 dx x 2 x1 ng. (2) Also since, U=N1q1+N2q2= eer i 1 1 q2 2 2 net We have du q1 q2 d 2 (3) thus, Eq yield 1 ( q1 q2 ) x2 x2 (4) the equation can be written as =Bq VEL TECH (5) VEL TECH MULTI TECH 70 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH where the (1 x2 ) matrix B, called the element strain-displacement matrix, is given by B 1 [ 1 1] x 2 x1 (6) 5. Explain in detail about co-ordinate and shape junction. COORDINATES AND SHAPE FUNCTIONS: Consider a typical finite element in figure. In the local number scheme, the first node will be numbered 1 and the second node. 2. The notation x1=x-coordinate of node 1,x2=x-coordinate of one 2 is used. We define a Natural or intrinsic coordinate system, denote by , as ww w.E asy 2 (x x1 ) 1 x 2 x1 (1) En gin eer i From fig, we see that =1 at node 2. the length of an element is covered when changes from -1 to 1. We use this system of coordinates in defining shape functions, which are used in interpolating the displacement field. ng. net Now the unknown displacement field within an element will be interpolated by a linear distribution (fig). this approximation becomes increasingly accurate as more element are considered in the model. To implement this linear interpolation, linear shape functions will be introduced as. N1( ) 1 2 (2) N2 ( ) 1 2 (3) The shape function N1 and N2 are shown in fig a and b, respectively. Then graph of the shape function N1 in fig (2a) is obtained from equation (2) by noting that N11 at =-1, N1=0 at =1, and N1 is a straight line between he two points. Similarly, the graph of N2 in fig (4b) is obtained from equation (3) . Once the shape function are defined the linear displacement filed within the element can be written in terms of the nodal displacement q1 and q2 as VEL TECH VEL TECH MULTI TECH 71 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH u N1q1 N2q2 VEL TECH HIGH TECH (4a) Figure. Linear interpolation of the displacement filed with an element ww w.E asy En gin eer i ng. net Figure :4(a) Shape function N1, 9b) shape function N2, and (c) linear interpolation using N1 and N2 Or, in matrix notation as U=Nq (4b) Where N=[N1,N2] and q=[q1,q2]T (5) In these equations , q is referred to as the element displacement vector. It is readily verified from Eq. 4a that u=q1 , at node 1, u= q2 at node 2, and that u varies linearly (fig 4c) VEL TECH VEL TECH MULTI TECH 72 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH It may be noted that the transformation from x to in equation (1) can be written in terms of N1 and N2 as X=N1x1+N2x2 (6) Comparing E qs.(4a) and (6) , we see that both the displacement u and the coordinate x are interpolated within the element using the same shape functions N 1 and N2. this is referred to as isoparametric formation in the literature. Though linear shape function have been used previously, other choices are possible. Shape functions need to satisfy the following: 1. First derivatives must be finite within an element 2. Displacements must be continuous across the element boundary. ww w.E Rigid body motion should not introduce any stress in the element. asy 6. Consider the bar as shown in fig (1). For each element i,Ai and I are the cross-sectional area and length, respectively. Each element i is subjected to a traction force T i per unit length and a body force f per unit volume. The units of Ti,f, Ai and so on are assumed:- be consistent. The Young’s modulus of the material is E. A concentrated load P2 is applied at node 2. The structural stiffness matrix and nodal load vector will bow be assembled. En gin eer i ng. net The element stiffness matrix for each element I is obtained from Equation as [K(1) ] VEL TECH EAi 1 1 i 1 1 VEL TECH MULTI TECH 73 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH The element connectivity table is the following: Element 1 2 3 4 1 1 2 3 4 2 2 3 4 5 The element stiffness matrices can be ‚expanded’ using the connectivity table and then summed (or assembled) to obtain the structural stiffness matrix as follows:* ww 1 1 0 0 0 1 1 EA1 0 K 1 0 0 0 0 EA 3 0 + 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 EA 2 0 1 0 2 0 0 0 0 0 0 w.E 0 0 0 0 0 1 0 1 0 0 0 0 1 1 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 asy 0 0 0 0 EA 4 0 0 4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 1 1 En Which gives A1 1 A 1 1 K E 0 0 0 A 1 1 gin ng. 0 A1 A 2 A2 0 0 2 1 2 A2 A3 A 3 A 2 0 2 3 3 2 A3 A 4 A3 A4 0 3 4 4 3 A 4 A4 0 0 4 4 0 eer i 0 net *This ‚expansion‛ of element stiffness matrices as shown in Examples is merely for illustration purposes and is never explicitly carried out in the computer. Since storing zeroes is inefficient. Instead , K is assembled directly from k’ using connectivity table. VEL TECH VEL TECH MULTI TECH 74 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH The global vector is assembled as A11f 1T1 0 2 2 A 2 2 f 2 T2 A11f 1T1 2 2 2 2 P2 A 2 2 f 2 T2 A 3 3 f 3 T3 F 2 2 0 2 2 A 3 3 f 3 T3 A f T 4 4 4 4 2 2 0 2 2 A f T 4 4 4 4 2 2 0 7. ww w.E Consider the thin (thin) plate in fig.(1a). the plate has a uniform thickness t=1 in. Young’s modulus E=30 x 106 psi, and weight density =100 lb at its midpoint. asy (a) Model the plate with two finite elements. (b) Write down expressions for the element stiffness matrices and element body force vectors. (c) Assemble the structural stiffness matrix K and global load vector F. (d) Using the elimination approach, solve for the global displacement vector Q, (e) Evaluate the stress in each element (f) Determine the reaction force at the support. En gin eer i ng. net Solution: (a) Using two element each of 12 in, in length, we obtain the finite element model in fig. Nodes and elements are numbered as shown. Note that the area at the midpoint of the plate in fig (1a) is 4.5 in2. consequently, the average and of element 1 is A1 =(6+4.5)/2 =5.25 in2, and the average area VEL TECH VEL TECH MULTI TECH 75 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH of element 2 is A2=(4.5+3)/2 =3.75 in2. the boundary condition for this model is Q1=0. (b) From Eq. we can write down expressions for the element stiffness matrices of the two element as 1 K1 2 Global dof 30 106 5.25 1 1 1 1 1 2 12 and 2 3 30 106 3.75 1 1 2 1 1 3 12 ww K2 w.E Using Eq. the element body force vector are asy global dof 5.25 12 0.2836 1 1 f = 2 1 2 1 and f2= 3.75 12 0.2836 1 2 2 1 3 En gin eer i © the global stiffness matrix K is assembled from K1 and k2 as 1 K 2 3 0 1 5.25 5.25 30 106 2 5.25 9.00 3.75 12 0 3.75 3.75 3 ng. net the externally applied global load vector f is assembled from f1,f2, and the point load P=100 lb; as 8.9334 F 15.3144 100 6.3810 (d) In the elimination approach, the stiffness matrix K is obtained by deleting rows and columns corresponding to fixed dofs. In this problem , dof 1 is fixed. Thus, K is obtained by deleting the first row and column of the original f. the resulting equations are VEL TECH VEL TECH MULTI TECH 76 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH 2 VEL TECH HIGH TECH 3 30 10 9.00 3.75 Q2 115.3144 12 3.75 3.75 Q3 6.3810 6 solution of these equations yields Q2=0.9272 x105 in Q3=0.9953 x 10-5 in Thus, Q=[0,0.9272 x 10-5, 0.9953 x10-5]T in. (e) using Eqs, 3.15 and 3.16 , we obtain the stress in each element ww 30 106 1 {1 12 w.E 0 1} -5 0.9272 10 = 23.18 psi and 30 106 = 1.70 psi asy (f) the reaction force R1 at node 1 is obtained from 1 {1 12 0.9272 10 1} -5 0.9953 10 -5 En Eq. This calculation require the first row of K from part ©. Also , from part ©, note that the externally applied load 9due to the self-weight) at note 1 is F1= 8.9334 lb. thus, 30 106 R1 [5.25 -5.25 12 0 0] 0.9272 10-5 0.9953 10 5 gin 8.9334 eer i ng. =-130.6 lb net 8. An axial load P= 300 x 103 N is applied at 20C to the rod as shown in fig . the temperature is then raised to 60 C. (a) Assemble the K and F matrices. (b) Determine the nodal displacement and elements stresses. Solution: (a) the element stiffness matrices are VEL TECH VEL TECH MULTI TECH 77 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH K1 VEL TECH MULTI TECH VEL TECH HIGH TECH 70 103 900 1 1 1 1 N/mm 200 K2 70 103 900 1 1 1 1 N/mm 200 Thus, 0 315 315 K 10 315 1115 800 N / mm 0 800 800 3 ww Now, in assembling F, both temperature and point load effects have to be considered. The element temperature forces due to T=40C are obtained from Eq. as w.E Global dof 1 1 N 1 1 1 70 103 900 23 10 6 40 asy and 1 2 200 10 1200 11.7 10 40 N 1 3 2 3 6 En Upon assembling 1, 2, and the point load , we get gin eer i ng. 57.96 f=103 57.96 112.32 300 112.32 or F 103 [57.96,245.64,112.32]T N net (b) the elimination approach will now be used to sole for the displacements. Since dofs 1 and 3 are fixed , the first and third rows and columns to K, together with the first and third components of F, are deleted. This results in the scalar equation. 103[115]Q2=103 x 245.64 yielding Q2=0.220 mm VEL TECH VEL TECH MULTI TECH 78 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Thus, Q=[0,0.220, 0]Tmm In evaluating element stresses, we have to use Eq. 3.105 b 1 0 70 102 3 6 [ 11] 70 10 23 10 40 0.220 200 ww =12.60 MPa and 2 0.220 3 6 1] 200 10 11.7 10 40 0 200 103 [1 300 w.E =-240.27 MPa asy 9. Consider the bar shown in figure. an axial load P=2500 x 103 N . using the penalty approach for handling boundary conditions, do the following. En gin (a) Determine the nodal displacements (b) Determine the stress in each material (c) Determine the reaction forces eer i ng. net Figure Solution: (a)The element stiffness matrices are 1 2 Global dof K1 70 10 2400 1 1 1 1 300 K2 200 10 600 1 1 1 1 400 3 and 2 VEL TECH 3 3 VEL TECH MULTI TECH 79 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH The structural stiffness matrix that is assembled from k1 and k2 is 1 2 3 0 0.56 -0.56 K 10 -0.56 0.86 -0.30 0 -0.30 0.30 6 The global load vector is C=[0.86 x 106] x 104 ww Thus, the modified stiffness matrix is w.E 0 8600.56 0.56 K 10 0.56 0.86 0.30 0 0.30 8600.30 6 asy the finite element equations are given by En gin 0 0 8600.56 0.56 Q1 6 3 10 0.56 0.86 0.30 Q 2 200 10 0 0.30 8600.30 Q3 0 which yields the solution ng. Q={15.1432 x10-6, 0.23257,8.0027 x10-6]T mm (c) the element stresses (Eq. 3.16 are 1 70 103 1 (1 300 eer i 15.1432 10-6 1) 0.23257 net =54.27 MPa where 1 MPa =106 N/m2 =1N/mm2. also, 2 200 103 VEL TECH 0.23257 1) -6 8.1127 10 =-116.27 MPa 1 ( 1 400 VEL TECH MULTI TECH 80 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH (c) the reaction forces are obtained from Eq, as R1 CQ1 =-[0.86 1010 ] 15.1432 106 =-130.23 103 Also, R3 CQ1 =-[0.86 1010 ] 8.1127 106 =-69.77 103N 10. In the following fig , a load P= 60 x 103 N is applied as shown Determine the displacement field stress, and support reactions in the body. Take E= 20 x103 N/mm2. ww w.E asy En gin eer i ng. Solution: net In this problem, we should first determine whether contact occurs between the bar and the wall , B. to do this, assume that the wall does not exist. Then, the solution to the problem can be verified to be. Mm QB=1.8 mm Where QB is the displacement of point B’ . from this result, we see that contact does occur. The problem has to be re-solved, since the boundary conditions are now different: the displacement at B’ is specified to be 1.2 mm. Consider the two –element finite element model in figure. the boundary conditions are Q1=0 and Q3=1.2 mm. The structural stiffness matrix K is VEL TECH VEL TECH MULTI TECH 81 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 1 1 0 20 103 250 K 1 2 1 150 0 1 1 and the global load vector f is F=[0,60 x 103, 0]T In the penalty approach, the boundary conditions Q1=0 and Q3=1.2 imply the following modification: A large number c chosen here as C= (2/3)x 1010 , is added on to the 1st and 3rd diagonal elements of K. also, the number (C x 1.2) gets added on to the 3rd component of F. thus, the modified equation are. ww w.E 0 Q1 0 20001 1 105 3 1 2 1 Q 60.0 10 2 3 0 1 20001 Q3 80.0 107 asy En the solution is Q=[7.49985 x 10-5, 1.500045, 1.200015]Tmm The element stress are 1 200 103 1 [1 150 gin 1] 7.49985 105 1.500045 ng. =199.996 MPa 1 [1 2 200 10 150 1] 1.500045 1.200015 3 eer i =-40.004 MPa The reaction forces are net R1=-C x 7.49985 x 10-5 = -49.999 x 103 N and R3=-C x (1.200015 -1.2) =-10.001 x 103 N The results obtained from the penalty approach have a small approximation error due to flexibility of the support introduced. In fact, the reader may verify that the elimination approach for handling boundary conditions yields the exact reactions, R1=-50.0 x 103 N and R3 =-10.0 x 103 N. VEL TECH VEL TECH MULTI TECH 82 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 11. Explain in detail about Quadratic Shape functions. QUADRATIC SHAPE FUNCTIONS So far, the unknown displacement field was interpolated by linear shape functions within each element. In some problems, however, use of quadratic interpolation leads to far more accurate results. In this section, quadratic shape functions will be introduced, and the corresponding element stiffness matrix and load vectors will be derived. The reader should note that the basic procedure is the same as that used in the linear one-dimensional element earlier. ww Consider a typical three-node quadratic element, as shown in figure (a). In the local numbering scheme, the left node will be numbered 1, the right node 2, and the midpoint 3. Node 3 has been introduced for the purposes of passing a quadratic fit and is called an internal node. The notation xi = x – coordinate of node i, I = 1, 2, 3, is used. Further, q = [q1, q2, q3]T, where q1, q2, and q3 are the displacements of nodes 1, 2 and 3, respectively. The x-coordinate system is mapped onto a -coordinate system, which is given by the transformation. w.E 2 x x3 x2 x1 asy .......(1) En gin From Equation 1, we see that = -1, 0 and +1 at nodes 1, 3 and 2 (Fig. (b)). Now, in -coordinates, quadratic shape functions N1, N2, and N3 will be introduced as 1 N1 1 ....(2a) 2 1 N2 1 .....(2b) 2 N3 1 1 .....(2c) eer i ng. net The shape function N1 is equal to unity at node 1 and zero at nodes 2 and 3. Similarly, N2 equals unity at node 2 and equals zero at the other two nodes; N 3 equals unity at node 3 and equals zero at nodes 1 and 2. The shape functions N 1, N2 and N3 are graphed in figure. The expressions for these shape functions can be written down by inspection. For example, since N1 = 0 at = 0 and N1 = 0 at = 1, we know that N1 must contain the product (1 - ). That is, N1 is of the form N1 = c(1 - ) <.(3) VEL TECH VEL TECH MULTI TECH 83 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Quadratic element in x- and -coordinates. ww w.E asy En gin eer i Shape functions N1, N2 and N3. ng. net 1 The constant c is now obtained from the condition N1 = 1 at = -1, which yields c = , resulting 2 in the formula given in Equation 2(a). These shape functions are called Lagrange shape functions. Now the displacement field within the element is written in terms of the nodal displacements as u = N1q1 + N2q2 + N3q3 <. (4a) or u = Nq <(4b) where N = [N1, N2, N3] is a (1 3) vector of shape functions and q = [q1, q2, q3]T is the (3 1) element displacement vector. At node 1, we see that N1 = 1, N2 = N3 = 0, and hence u = q1. Similarly, VEL TECH VEL TECH MULTI TECH 84 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH u = q2 at node 2 and u = q3 at node 3. Thus, u in Equation 4(a) is a quadratic interpolation passing through q1, q2 and q3. The strain is now given by du dx du d = d dx 2 du = x 2 x1 d (strain-displacement relation) (chain rule) (using Eq.1) ......(5) ww = 2 dN1 dN2 dN3 , , .q (using Eq. 4) x 2 x1 d d d w.E asy En gin eer i Interpolation using quadratic shape functions ng. We have, 2 1 2 1 2 , , 2 q ......(6) x2 x1 2 2 which is of the form = Bq net <..(7) where B is given by B 2 1 2 1+2 , , 2 x2 x1 2 2 <..(8) Using Hooke’s law, we can write the stress as E Bq VEL TECH <(9) VEL TECH MULTI TECH 85 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Note that since Ni are quadratic shape functions, B in Eq.8 is linear in . This means that the strain and stress can vary linearly within the element. Recall that when using linear shape functions, the strain and stress came out to be constant within the element. We now have expressions for u, , and in Eqs. 4(b), 7, and 9, respectively. Also, we have dx e / 2 d from Eq.1. Again, in the finite element model considered here, it will be assumed that cross-sectional area Ae, body force F, and traction force T are constant within the element. Substituting for u, , , and dx into the potential-energy expression yields. ww T 1 A dx - uT fA dx - uT T dx - QiPi e e e e 2 e e i w.E 1 1 1 = qT Ee Ae e BT B d q qT Ae e f N T d 1 1 2 2 e 2 e asy 1 - q e T N T d Q i Pi 2 1 i e T <.(10) En gin Comparing the above equation with the general form 1 qT k eq qT f e qTT e Qi Pi e 2 e e i yields E A 1 <(11a) k e e e e BT B d 1 2 which, upon substituting for b in Eq.(8), yields 1 2 VEL TECH Ae ef 1 T N d 2 1 ng. net 3 local dof 1 8 1 7 E A K e e e 1 7 8 2 3 e 8 8 16 3 the element body force vector fe is given by fe eer i <(11b) <<(12a) VEL TECH MULTI TECH 86 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH which, upon substituting for N in Eqs,, yields local dof 1/ 6 1 f e Ae ef 1/ 6 2 2 / 3 3 <.. (12b) Similarly, the element traction –force vector Te is given by Te eT 1 T N d 2 1 ww <..(13a) w.E which results in local dof 1/ 6 1 T ef 1/ 6 2 2 / 3 3 e asy <..(13b) En gin 1 T Q KQ QT F , where the structural 2 stiffness matrix k and nodal load vector f are assembled from element stiffness matrices and load vectors, respectively. The total potential energy is again of the form eer i ng. 12. Consider the structure shown in figure. A rigid bar of negligible mass, pinned at one end, is supported by a steel rod and an aluminium rod. A load P = 30 103 N is applied as shown. net (a) Model the structure using two finite elements. What are the boundary conditions for your model? (b) Develop the modified stiffness matrix and modified load vector. Solve the equations for Q. Then determine element stresses. VEL TECH VEL TECH MULTI TECH 87 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Solution (a) The problem is modeled using two elements as shown in the following connectivity table: CONNECTIVITY TABLE Element No. 1 2 Node 1 3 4 Node 2 1 2 ww The boundary conditions at nodes 3 and 4 are obvious: Q3 = 0 and Q4 = 0. Now, since the rigid bar has to remain straight, Q1, Q2, and Q5 are related as shown in Figure(b). The multipoint constraints due to the rigid bar configuration are given by w.E Q1 0.333 Q5 0 Q 0.833 Q 0 asy 2 5 En (b) First, the element stiffness matrices are given by 3 1 gin 53.33 3 200 10 1200 1 1 3 53.33 k1 10 1 1 4500 -53.33 53.33 1 and 4 2 3 k2 21 21 4 70 10 900 1 1 103 3000 1 1 -21 21 2 3 The global stiffness matrix K is 1 2 3 4 0 53.33 0 53.33 0 21 0 21 3 K 10 53.33 0 0 53.33 21 0 21 0 0 0 0 0 eer i ng. net 5 0 1 0 2 0 3 0 4 0 5 The K matrix is modified as follows: a number C = 53.33 103 104 , large in comparison to the stiffness values, is chosen. Since Q3 = Q4 = 0, C is added on to the (3, 3) and (4, 4) locations of K. Next, multipoint constraints given in part (a) are considered. For the first constraint, Q 1 VEL TECH VEL TECH MULTI TECH 88 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 0.333Q5 = 0, we note that 0 = 0, 1 = 1, and 2 = -0.333. The addition to the stiffness matrix is obtained from Equation as 1 5 C 12 C 1 2 17.77 1 7 53.33 10 2 17.77 5.925926 5 C 1 2 C 2 The force addition is zero since 0 = 0. Similarly, the consideration of the second multipoint constraint Q2 – 0.833Q5 = 0 yields the stiffness addition ww 2 5 53.33 44.44 2 107 44.44 37.037037 5 w.E On addition of all the preceding stiffnesses, we obtain the final modified equations as asy 0 53.33 0 177777.7 Q1 0 533386.7 Q 0 0 533354.3 0 21.0 44444.4 2 3 Q3 0 10 53.33 0 533386.7 0 0 0 21.0 0 533354.3 0 Q4 0 3 177777.7 444444.4 0 0 429629.6 Q5 30 10 En gin eer i The solution, obtained from a computer program that solves matrix equations, is Q 0.486 1.215 4.85 105 4.78 105 1.457 mm The element stresses are now recovered from Equations as 4.85 105 200 103 1 1 1 4500 0.486 = 21.60 MPa ng. net and 2 28.35 MPa 13. Consider the rod (a robot arm) in figure. Which is rotating at constant angular velocity = 30 rad/s. Determine the axial stress distribution in the rod, using two quadratic elements. Consider only the centrifugal force. Ignore bending of the rod. Solution: A finite element model of the rod, with two quadratic elements, is shown in figure. The model has a total of five degrees of freedom. The elements stiffness matrices are (from equation). VEL TECH VEL TECH MULTI TECH 89 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH ww w.E asy En 1 3 gin 2 Global dof 1 8 1 7 107 0.6 1 K 1 7 8 3 3 21 8 8 16 2 eer i ng. Figure net And 3 5 4 1 8 3 7 107 0.6 K 1 7 8 5 3 21 8 8 16 4 2 VEL TECH VEL TECH MULTI TECH 90 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Thus, 1 2 3 4 5 7 8 1 0 0 1 8 16 8 0 0 2 107 0.6 K 1 8 14 8 1 3 3 21 0 0 8 16 8 4 0 0 1 8 7 5 The body force (Ib/in.3) is given by f r 2 g Ib / in3 ww w.E Where = weight density and g = 32.2 ft/s2. Note that Is a function of the distance r from the pin. Taking average values of over each element, we have asy and 0.2836 10.5 302 32.2 12 = 6.94 f1 En gin 0.2836 31.5 302 32.2 12 = 20.81 f2 Thus, the elements body force vectors are (from equation) Global dof 1 1 6 1 1 f 0.6 21 f1 3 6 2 3 2 eer i ng. net and Global dof 1 3 6 1 f 1 0.6 21 f2 5 6 2 3 4 VEL TECH VEL TECH MULTI TECH 91 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Assembling f1 and f2, we obtain F = [14.57, 58.26, 58.26, 174.79, 43.70] T Using the elimination method, the finite element equations are 16 8 0 0 Q2 58.26 10 0.6 8 14 8 1 Q3 58.26 0 8 16 8 Q4 174.79 63 1 8 7 Q5 43.7 0 7 ww Which yields w.E Q = 10-3[0, .5735, 1.0706, 1.4147, 1.5294]T mm asy There stress can now be evaluated from equations. The element connectivity table is as follows: Element number 1 1 1 2 3 3 2 2 3 5 4 En Local Node Nos. Global Node Nos. gin Thus, eer i ng. q = [Q1, Q3, Q2] T for element 1, while q = [Q3, Q5, Q4] T for element 2, Using equations, we get net Q1 2 1 2 1 2 1 10 , , 2 Q3 21 2 2 Q2 7 Where – 1 1, and 1 denotes the stress in element 1. The stress at node 1 in element 1 is obtained by substituting = -1 into the previous equation, which results in VEL TECH VEL TECH MULTI TECH 92 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 0 2 3 1 1 10 10 1.5, 0.5, 2.0 1.07076 21 .5735 = 583 psi 7 The stress at node 2 (which corresponds to the midpoint of element 1) is obtained by substituting for = 0: 0 2 3 1 3 10 10 0.5,0.5,0 1.07076 21 .5735 = 510 psi 7 ww w.E Similarly, we obtain 1 2 2 1 437psi asy 2 3 218psi 2 2 0 En The axial distribution is shown in figure. The stresses obtained from the finite element model can be compared with the exact solution, given by exact (x) 2 2g gin (L2 x 2 ) eer i ng. 14. net Consider the four bar truss shown in figure. It is given that E = 29.5 x 10 6 psi and Ae = l in 2.for all elements. Complete the following: a) Determine the element stiffness matrix for each element b) Assemble the structural stiffness matrix K for the entire truss c) Using the elimination approach, solve for the nodal displacement. d) Recover the stresses in each element. e) Calculate the reaction forces. VEL TECH VEL TECH MULTI TECH 93 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH Figure VEL TECH MULTI TECH VEL TECH HIGH TECH ww Solution: w.E asy En a) It is recommended that tabular form be used for representing nodal coordinate data are as follows: Node X 1 2 3 4 0 40 40 0 gin Y 0 0 30 30 ng. The element connectivity table is Element 1 2 3 4 1 1 3 1 4 eer i net 2 2 2 3 3 Note that the user has a choice in defining element connectivity. For example, the connectivity of element 2 can be defined as 2 – 3 instead of 3 – 2 as in the previous table. However, calculations of the direction cosines will be consistent with the adopted connectivity scheme. Using formulas in equations and, together with the nodal coordinate data and the given VEL TECH VEL TECH MULTI TECH 94 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH element connectivity information, we obtain the direction cosine table: Element e M 1 2 3 4 40 30 50 40 1 0 0.8 1 0 -1 0.6 0 For example, the direction cosines of elements 3 are obtained as = (x3 – x1)/ e = (40 – 0)/50 = 0.8 and m = (y3 – y1)/ e = (30 – 0)/50 = 0.6. Now, using equation, the element stiffness matrices for element 1 can be written as ww w.E 1 1 29.5 10 0 K1 1 40 0 6 2 3 4 Global dof 0 1 0 0 0 1 0 0 0 1 0 2 0 3 0 4 asy En gin The global dofs associated with elements 1, which is connected between nodes 1 and 2, are indicated in K1 earlier. These global dofs are shown in figure and assist in assembling the various element stiffness matrices. eer i The element stiffness matrices of elements 2, 3, and 4 are as follows: 5 6 3 4 0 0 29.5 10 0 1 K2 0 0 50 0 1 6 1 0 0 5 0 1 6 0 0 3 0 1 4 2 5 6 .64 .48 29.5 106 .48 .36 3 K .64 .48 50 .48 .36 .64 .48 .64 .48 .48 .36 .36 .36 VEL TECH ng. net 1 2 5 6 VEL TECH MULTI TECH 95 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 7 8 5 6 1 29.5 106 0 4 K -1 40 0 0 -1 0 7 0 0 0 8 0 1 0 5 0 0 0 6 b) The structural stiffness matrix K is now assembled from the element stiffness matrices. By adding the element stiffness contributions, noting the element connectivity, we get 1 ww 2 3 4 0 22.68 5.76 15.0 5.76 4.32 0 0 15.0 0 15.0 0 0 0 20.0 29.5 106 0 K 7.68 5.76 600 0 0 0 20.0 5.76 4.32 0 0 0 0 0 0 0 0 w.E 5 6 7 7.68 5.76 0 5.76 4.32 0 0 0 0 0 20 0 22.68 5.76 15.0 5.76 24.32 0 15.0 0 15.0 0 0 0 asy 8 En 0 1 0 2 0 3 0 4 0 5 0 6 0 7 0 8 c) The structural stiffness matrix K given above needs to be modified to account for the boundary conditions. The elimination approach will be used here. The rows and columns corresponding to dots 1, 2, 4, 7, and 8, which correspond to fixed supports, are deleted form the K matrix. The reduced finite element equations are given as gin 0 0 Q3 20 000 15 29.5 106 0 22.68 5.76 Q5 0 600 0 5.76 24.32 Q6 -25 000 eer i ng. The nodal displacement vector for the entire structure can therefore be written as net Q = [0, 0, 27.12 x 10-3, 0, 5.65 x 10-3, -22.25 x 10-3, 0, 0]T in. D) The stress in each element can now be determined form equation, as shown below. The connectivity of element 1 is 1 – 2. Consequently, the nodal displacement vector for element 1 is given by q = [0, 0, 27.12 x 10-3, 0]T, and equation yields. VEL TECH VEL TECH MULTI TECH 96 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 0 0 29.5 10 1 1 0 1 0 3 40 27.12 10 0 = 20 000.0 psi 6 The stress in member 2 is given by 5.65 10 3 3 22.25 10 29.5 106 2 0 1 0 1 3 30 27.12 10 0 =-21 880.0 psi ww w.E Following similar steps, we get asy En 3 = 5208.0 psi 4 = 4167.0 psi gin d) The final step is to determine the support reactions. We need to determine the reaction forces along dofs 1, 2, 4, 7, and 8, which correspond to fixed supports. These are obtained by substituting for Q into the original finite element equation R = KQ – F. In this substitution, only those rows of K corresponding to the support dofs are needed, and F = O for theses dofs. Thus, we have 22.68 5.76 R1 5.76 4.32 6 R2 29.5 10 0 0 600 R3 0 0 R 4 0 0 15.0 0 0 0 0 0 0 20.0 0 0 7.68 5.76 0 15.0 0 0 0 5.76 0 0 3 27.12 10 4.32 0 0 0 20.0 0 0 5.65 10 3 0 15.0 0 22.25 10 3 0 0 0 0 0 eer i ng. net Which results in R1 15833.0 R 2 3126.0 R 4 21879.0 R 4167.0 7 0 R8 A free body diagram of the truss with reaction forces and applied loads is shown in figure. VEL TECH VEL TECH MULTI TECH 97 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 15. The four bar truss of Example is considered here, but the loading is different. Take E = 29.5 x 106 psi and = 1/150 000 per F. a) There is an increase in temperature of 50F in bars 2 and 3 only (figure). There are no other loads on the structure. Determine the nodal displacements and element stresses as a result of this temperature increase. Use the elimination approach. b) A support settlement effect is considered here. Node 2 settles by 0.12 in. Vertically down, and in addition, two point loads are applied on the structure (figure). Write down (without solving) the equilibrium equations KQ = F, where K and F are the modified structural stiffness matrix and load vector, respectively. Use the penalty approach. c) Use the program TRUSS2 to obtain the solution to part (b). ww Solution: w.E a) The stiffness matrix for the truss structure has already been developed in example. Only the load vector needs to be assembled due to the temperature increase. Using equation, the temperature loads as a result of temperature increases in elements 2 and 3 are, respectively. asy En gin eer i ng. net Figure Global dof 0 5 29.5 106 50 1 6 2 150,000 0 3 1 4 And 0.8 1 6 29.5 10 50 0.6 2 3 150,000 0.8 5 0.6 6 VEL TECH VEL TECH MULTI TECH 98 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH The 2 and 3 vectors contribute to the global load vector F. Using the elimination approach; we can delete all rows and columns corresponding to support dofs in K and F. The resulting finite element equations are 0 0 Q3 0 15.0 29.5 106 0 22.68 5.76 Q5 7866.7 600 0 5.76 24.32 Q6 15733.3 Which yield ww 0 Q3 Q5 0.003951 in. Q 0.01222 6 w.E asy The element stresses can now be obtained from equation. For example, the stress in element 2 is given as En 0.003951 0.01222 29.5 106 50 29.5 106 2 0 1 0 1 30 0 150,000 0 gin = -8631.7 psi eer i ng. The complete stress solution is 1 0 2 2183 psi 3643 3 4 2914 net b) Support 2 settles by 0.12 in. vertically down, and two concentrated forces are applied (figure). In the penalty approach for handling boundary conditions, recall that a large spring constant C is added to the diagonal elements in the structural stiffness matrix at those dofs where the displacements are specified. Typically, C may be chosen 104 times the largest diagonal element of the unmodified stiffness matrix (see equation). Further, a force Ca is added to the force vector, where a is the specified displacement. In this example, for dof 4, a = -0.12 in., and consequently, a force equal to -0.12C gets added to the fourth VEL TECH VEL TECH MULTI TECH 99 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH location in the force vector. Consequently, the modified finite element equations are given by 5.76 15.0 0 7.68 22.68 C 4.32 C 0 0 5.76 15.0 0 0 20.0 C 0 29.5 106 600 22.68 symmetric 5.76 4.32 0 20.0 5.76 24.32 0 0 0 0 15.0 0 15.0 C 0 Q1 0 0 Q2 0 0 Q3 20 000 0 Q4 -0.12C 0 Q 5 0 0 Q6 -25 000.0 0 Q 7 0 C Q8 0 c) Obviously, the equations in (b) are too large for hand calculations. In the program TRUSS, that is provided, these equations are automatically generated and solved from the user’s input data. The output from the program is ww w.E Q3 0.0271200 Q 4 0.1200145 in. Q 0.0323242 5 Q6 0.1272606 And 1 20 000.0 2 -7 125.3 psi 3 -29 791.7 4 23 833.3 VEL TECH asy En gin eer i ng. VEL TECH MULTI TECH 100 net VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH UNIT - III PART - A 1. Write down the strain displacement relation equation. u v u v , , x y y x T 2. Define the term element: ww The points where the cameras of the triangles meet are called nodes, and each triangle formed by three nodes and three series called an element. w.E 3. Write a short note on constant stain elements. asy Plane Stress and Plane strain models have Constant strain The quadrilateral and Triangular elements are known as Constant Strain Elements (CSE). Element Stiffness Matrix En gin eer i ng. net ui Vi u j d v j um v m VEL TECH VEL TECH MULTI TECH 101 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 4. Write a short note on Isoparametric representation. The displacement inside the element are now written using the shape functions and the nodal values of the unknown displacement field. We have u=N1q1+ N2q2+ N3q3 v= N1q2+ N2q4+ N3q6 (5.12a) or, using Eq. 5.10 u q1 q5 q3 q5 qs ww u q2 q6 q4 q6 q6 w.E 5.12b The relations 5.12a can be expressed in a matrix form by defining a shape functions matrix. N 0 N2 0 N3 0 N 1 0 N1 0 N2 0 N3 and asy (5.13) En 5.14 u=Nq gin for the triangular element, the coordinates x, y can also be represented interms of nodal coordinates using the same shape functions. This isoparametric representation. eer i ng. 5. Briefly explain about element stiffness. Element Stiffness net We now substitute for the strain from the element strain – displacement relationship in Eq. into the element strain energy Ue in Eq.,b to obtain 1 T D tdA e 2 1 qTBTDBqt A 2 e Ue (5.29a) Taking the element thickness te as constant over the element and remembering that all terms in the D and B matrices are constants, we have Ue 1 qTBTDBt e dA q e e 2 VEL TECH 5.29b VEL TECH MULTI TECH 102 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Now edA=Ae, where Ae is the area of the element, thus, Ue 1 T q t e A eBTDBq 2 5.29c 1 T e qK q 2 5.29d or Ue where ke is the element stiffness matrix given by ke te AeBTDB 5.30 ww For plane stress or plane strain, the element stiffness matrix can be obtained by taking the appropriate material property matrix D defined 1 and carrying out the previous multiplication on the computer. We note that ke is symmetric since D is symmetric. The element connectivity as established in Table is now used to add the element stiffness values in K e into the corresponding global locations in the global stiffness matrix K, so that w.E 1 U qTk e q e 2 1 = QTKQ 2 5.31 asy En gin 6. Explain the term fraction force. eer i ng. A traction force is a distributed load acting on the surface of the body. Such a force acts on edges connecting boundary nodes. A traction force acting on the edge of an element contributes to the global load vector F. This contribution can be determined by considering the traction force term e uT Ttdl. net 7. Write a short note on Galarkin approach. x , y T 5.46 and x , y , x y x x y y T 5.47 Where is an arbitrary (virtual) displacement vector, consistent with the boundary conditions The variation form is given by VEL TECH VEL TECH MULTI TECH 103 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH e T tdA T ftdA T Ttdl iTPi =0(5.48) A A i where the first term represents the internal virtual work. The expression in parentheses represents the external virtual work. 8. Write short notes on stress calculation. Since strains are constant in a constant – strain triangle (CST) element, the corresponding stresses are constant. The stress values need to be calculated for each element. Using the stress – strain relations in Eq. and element strain-displacement relations in Eq. we have ww =DBq. w.E 9. How the temperature should effect the two dimensional problems? asy If the distribution of the change in temperature T(x,y) is known, the strain due to this change in temperature can be treated as an initial strain 0. From the theory of mechanics of solid, 0 can be represented by En 0 T.T.0 T gin 5.61 for plane stress and 0 1 v T.T,0 T 5.62 eer i for plane strain. the stresses and strains are related by =D 0 5.63 ng. net The effect of temperature can be accounted for by considering the strain energy term. We have 1 T U 0 D 0 tdA 2 1 T D 2 T D 0 0T D 0 tdA 5.64 2 The first term in the previous expansion gives the stiffness matrix derived earlier. The last term is a constant, which has no effect on the minimization process. The middle term. Which yields the temperature load, is now considered in detail. Using the strain displacement relationship =Bq. D tdA q B D t A T A T 0 VEL TECH T 0 e e 5.65 e VEL TECH MULTI TECH 104 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH This step is directly obtained in the Galerkin approach where T will be T() and qT will be T. It is convenient to designate the element temperature load as e=teAeBTD0 e=[1, 2, 3, 4, 5, 6 ]T (5.66) (5.67) The vector 0 is the strain in Eq. 5.61 or 5.62 due to the average temperature change in the element. e represents the element nodal load contributions that must be added to the global force vector using the connectivity. ww The stresses in an element are then obtained by using Eq. 5.63 in the form =D(Bq-) w.E 10. Consider the two-dimensional loaded plate showing Fig. E5.4. In addition to the conditions defined in Example 5.5, there is an increase in temperature of the plate of 80 oF. The coefficient of linear expansion of the material is 10-6 /oF. Determine the additional displacement due to temperature Also, calculate the stresses in element 1. asy En gin Solution: …………..110 / F. and T=80F. So -6 o T 5.6 4 0 T 10 5.6 0 0 eer i ng. net Thickness <<<<..the area of the element A is 3 in 2. The element temperature loads are =tA(DB)T0 where DB1 if officiated in the solution of Example 5.5. On evaluation, we get =[11206 – 16800 0 16800 – 11206 0]T with associated dofs ,2,3,4,7,8, and (2)T=[-11206 16800 -16800 11206 0]T with associated dofs 5,6,7,8,3, and 4. Picking the forces for dofs ,3, and 4 <<the previous equations, we have FT=[F1F3F4]=[16800 11206 16800] VEL TECH VEL TECH MULTI TECH 105 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH On solving KQ = F, we get [Q1 Q3 Q4]=[.862 10-3 1.992 10-3 0.934 10-3] in The displacements of element 1 due to temperature are Q1=[1.862 10-3 0 1.992 10-4 0.934 10-3 0 0]T The stresses are calculated using Eq. 5.68 as 1=(DB1)Tq1 -D0 Substituting for the terms on the right – hand side, we get ww w.E 1=104 [.204 – 2.484 0.78]Tpsi asy 11. Write down the general equation of the stiffness matrix for 2-D element. En By energy method, we can derive the following relationship, [K]= [B] [D] [B] dV T Line Element: gin eer i ng. D is replaced by E dV is replaced by dL [B]=[-I/L I/L] net Plane Element [D] is defined by Equation dV=tdA, For constant cross section dA=A [B] is given by equation VEL TECH VEL TECH MULTI TECH 106 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH For Constant Strain Triangular element [k]=tA[B]T[D][B] (6.2.52) 12. Find the strain – nodal displacement matrices Be for the elements shown in fig. E5.3. Use local numbers given at the corners. ww w.E asy Solution We have y 23 0 y 31 0 y12 0 1 B 0 x 32 0 x13 0 x 21 det J x 32 y 23 x13 y 31 x 21 y12 2 0 0 0 2 0 1 = 0 0 0 6 3 2 0 0 2 1 En gin eer i ng. net where det J is obtain x13y23 – x23y13 = (3) (2) - (3)(0)=6. Using the local numbers at the corners., B2 It can be written using the relationship as 2 0 0 0 2 0 1 B 0 3 0 3 0 0 6 3 2 3 0 0 2 2 13. A CST elements is shown in Fig. E5.5. The element is subjected to a body force f x = x2 N/m3. Determine the nodal force vector fe. Take element thickness =1m. VEL TECH VEL TECH MULTI TECH 107 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH The work potential is - e fTudV, where fT=[fx, 0]. Substituting for u=Nq. We obtain the work potential in the form =qTfe. where fe = e NTtdV, where N is given in Eq. 5.13. All y components of fe are zero. The x components at nodes 1,2,3 are given, respectively, by f dV, f dV, 1 f dV e x x e x e We now make the following substitutions: f=x=x2,x=x1+x2+(1--)x3=4, dV= det J dd, det J = 2Ae, and Ae =6. Now, integration over a triangle is illustrated in Fig. 5.6. Thus, f dV 16 12 dd 3.2N 1 1 e x ww 2 0 0 Similarly, the other integration result in 9.6 N and 3.2N. Thus, w.E Fe=[3.2,0,9.6,0,3.2,0]TN asy 14. Write down the solution of equations for Natal displacements. En The modified equations are solved using Gaussian elimination and the strains may be computed as B u v gin eer i ng. the stresses may be computed as CB u v net 15. What are the two types of Natal load? i) ii) Direct procedure method Variational approach 16. Explain the term lumped load method. A portion is assigned to each node and load on that region as the nodal load. This method is called as lumped load method. VEL TECH VEL TECH MULTI TECH 108 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH PART - B 1. A) Explain in Detail i) Constant –strain triangle (CST) The displacements at points inside an element need to be represented in terms of the nodal displacements of the element. As discussed earlier. The finite element method uses the concept of shape functions in systematically developing these interpolations. For the constant strain triangle, the shape functions are linear over the element. The three shape functions, N1,N2 and N3 corresponding to nodes 1,2, and 3, respectively, are shown in fig. 5.4 shape function N1 is 1 at node 1 and linearity reduces to 0 at nodes 2 and 3, respectively, and dropping to 0 at the opposite edges. Any linear combination of these shape functions also represents a plane surface. In particular, N1+N2 +N3 represents a plane at a height of at nodes ,2,and 3, and, thus, it is parallel to the triangle 123. Consequently for every N1,N2, and N3. ww w.E asy N1+N2+N3= (5.9) En N1, N2, and N3 are therefore not linearity independent; only two of these are independent. The independent shape functions are conveniently represented by the pair ,as N1 = N2= N3=-- (5.10) gin eer i ng. Where, are natural coordinates (Fig.) At this stage, the similarity with the one dimensional element should be noted: in the one – dimensional problem the x-coordinates were mapped onto the coordinates, and shape functions were defined as functions of . Here, in the twodimensional problem, the x-,y-coordinates are mapped onto the -,-coordinates, and shape functions are defined as functions of and . net The shape functions can be physically represented by area coordinates. A point (x,y) in a triangle divides it into three areas, A,A2, and A3, as shown in Fig. The shape functions N1, N2, and N3 are precisely represented by VEL TECH VEL TECH MULTI TECH 109 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH ww N1 A1 A N2 w.E asy A A2 N3 3 A A 5.11 En where A is the area of the element. Clearly, N1+N2+N3=1 at very point inside the triangle. ii) Isoparametric representation. gin eer i The displacements inside the element are now written using the shape functions and the nodal values of the unknown displacement field. We have u N1q1 N2q3 N3q5 ng. 5.12a v N1q2 N2q4 N3q6 or, using Eq. 5.10, u q1 q5 q3 q5 q5 v q2 q6 q4 q6 q6 net 5.12b The relations 5.12a can be expressed in a matrix form by defining a shape function matrix N 0 N2 0 N3 0 N 1 0 N1 0 N2 0 N3 5.13 and u=Nq VEL TECH (5.14) VEL TECH MULTI TECH 110 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH For the triangular element, the coordinates x,y can also be represented in terms of nodal coordinates using the same shape functions. This is isoparametric representation. This approach lends to simplicity of development and retains the uniformly with other complex elements. We have X=N1x1+ N2x2+ N3x3 Y= N1y1+ N2y2+ N3y3 (5.15a) Or X=(x1-x3)+( x2-x3)+x3 Y=(y1-y3)+( y2-y3)+y3 ww (5.15b) Using the notation, xij =xi and yij=yi-yi, we can write Eq. 5.15b as w.E X=x13+x23+x3 Y=y13+y23+y3 asy (5.15c) En This equation relates x- and y- coordinates to the - and - coordinates. Equation 5.12 express u and v as functions of and . gin b) Evaluate the shape functions N1, N2 and N3 at the interior point P for the triangular element. eer i ng. net Solution: Using the isoparametric representation (Eqs.5.15), we have 3.85=1.5N1+7N2+4N3=2.5+3+4 4.8=2N1+3.5N2+7N3=5-3.5+7 These two equations are rearranged in the form 2.5-3=0.15 VEL TECH VEL TECH MULTI TECH 111 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 5+3.5=2.2 Solving the equations, we obtain =0.3 and =0.2, which implies that N1=0.3 N2=0.2 N3=0.5 In evaluating the strains, partial derivatives of u and v are to be taken with respect to x and y. From Eqs. 5.12 and 5.15, we see that u, v and x,y are functions of and . That is, u=u(x( ,),y( ,)) and similarly v=v (x( ,),y( ,)). Using the chain rule for partial derivatives of u, we have u u x u y x y u u x u y x y ww w.E asy which can be written in matrix notation as u x u x y u x y u y (5.16) En gin eer i where the (22) square matrix is denoted as the Jacobian of the transformation. J: x J x y y ng. (5.17) net Some additional properties of the Jacobian are given in the appendix. On taking he derivative of x and y. x J 13 x 23 y13 y 23 VEL TECH (5.18) VEL TECH MULTI TECH 112 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Also, from Eq. 5.16, u u x 1 J u u y (5.19) where J-1 is the inverse of the Jacobian J, given by J1 1 y 23 det J x 23 y13 x13 ww (5.20) detJ=x13 y 23 x 23 y13 (5.21) w.E From the knowledge of the area of the triangle, it can be seen that the magnitude of det J is twice the area of the triangle. If the points 1,2, and 3 are ordered in a counter clockwise manner, det J is positive in sign. We have A 1 | det J | 2 asy En (5.22) gin eer i where | | represents the magnitude. Most computer codes use a counter clockwise order for the nodes and use det J for evaluating the area. ng. 2. Determine the Jacobian of the transformation J for the triangular element shown in fig. E5.1. Solution: We have x J 13 x 23 y13 -2.5 -5.0 = y 23 3.0 3.5 net Thus, detJ =23.75 units. This is twice the area of the triangle. If 1,2,3 are in a clockwise order, then det J will be negative. From Eqs. 5.19 and 5.20, it follows that u u y 23 x 1 u det J x u y 23 VEL TECH u u x13 y13 (5.23a) VEL TECH MULTI TECH 113 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Replacing u by the displacement v, we get a similar expression v v y 23 x 1 v det J x v y 23 v v x13 y13 (5.23b) Using the strain-dis-placement relations (5.5) and Eqs. 5.12b and 5.23, we get u x v y u v y x ww w.E y 23 q1 q5 y13 q3 q5 1 x 23 q2 q6 x13 q4 q6 (5.24a) det J x q q x q q y q q y q q 13 3 5 23 2 6 13 4 6 23 1 5 asy En From the definition of xij and yij we can write y31 =y13 and y12 =y13 –y23, and so on. The foregoing equation can be written in the form gin y 23 q1 y 31q3 y12q5 1 x32q2 x13 q4 x 21q6 (5.24b) det J x32q1 y 23 q2 x13q3 y 31q4 x 21q5 y12q6 This equation can be written in matrix form as =Bq (5.25) eer i ng. net where B is a (36) element strain – displacement matrix relating the three strains to the six nodal displacements and is given by y 23 1 B 0 det J x32 0 x 32 y 23 y13 0 x13 0 y12 0 x13 0 x 21 (5.26) y 31 x 21 y12 It may be noted that all the elements of the B matrix are constants expressed in terms of the nodal coordinates. VEL TECH VEL TECH MULTI TECH 114 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 3. Explain in detail about Galerkin Approach. =[x,y]T (5.46) and x , y , x , y x y y x (5.47) where is an arbitrary (virtual) displacement vector, consistent with the boundary conditions. The variational form is given by ww tdA ftdA Ttdl P 0 5.48 T A T T A T i i A i w.E where the first term represents the internal virtual work. The expression in parentheses represents the external virtual work. On the discretized region, the previous equation becomes. asy D tdA ftdA Ttdl P 0 5.49 T e e T e e T L i T i i En gin Using the interpolation steps of Eqs. 5.12-5.14, we express =N (5.50) ()=B (5.51) where =[1, 2, 3, 4, 5, 6,]T (5.52) eer i ng. net Represents the arbitrary nodal displacements of element e. The global nodal displacement variations are represented by =[1, 2,<<, N]T (5.53) The element internal work term in Eq. 5.49 can be expressed as D tdA q B DBtdA T e T T e Noting that all terms of B and D are constant and denoting t e and Ae as thickness and area of element, respectively, we find that VEL TECH VEL TECH MULTI TECH 115 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH D tdA q B DBt dA T T T e e e =qT t e A eBTDB (5.54) =qTK e where Ke is the element stiffness matrix given by Ke=teAeBTDB (5.55) The material property matrix D is symmetric, and, hence, the element stiffness matrix is also symmetric. The element connectivity as presented in table 5.1 is used in adding the stiffness values of ke to the global locations. Thus, ww w.E D tdA q k K q T e e T e e T e e 5.56 = TKQ asy En The global stiffness matrix K is symmetric and banded. The treatment of external virtual work terms follows the steps involved in the treatment of force terms in the potential energy formulation, where u is replaced by . Thus, T T e ftdA f T (5.57) gin eer i ng. which follows from Eq. 5.33, with fe given by Eq. 5.36. Similarly, the traction and point load treatment follows from Eqs. 5.38 and 5.43. The terms in the variation form are given by Internal virtual work = TKQ External virtual work=TF (5.58a) (5.58b) net The stiffness and force matrices are modified to use the full size (all degrees of freedom). From the Galerkin form (Eq. 5.49). the arbitrariness of gives KQ=F (5.59) Where K and F are modified to account for boundary conditions. VEL TECH VEL TECH MULTI TECH 116 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 4. For the two – dimensional loaded plate shown in fig. E5.6, determine the displacement of nodes 1 and 2 and the element stresses using plane stress conditions. Body force may be neglected in comparison with the external forces. ww w.E asy Solution: For plane stress conditions, the material property matrix is given by 1 v 0 3.2 107 E D v 1 0 0.8 107 2 1 v 1 v 0 0 0 2 En gin 0.8 107 0 7 3.2 10 0 7 0 1.2 10 eer i ng. Using the local numbering pattern used in fig. E5.3, we establish the connectivity as follows: Element No. 1 2 1 1 3 Nodes 2 3 2 4 4 2 net On performing the matrix multiplication DBe, we get 0 1.067 0.4 0 0.4 1.067 DB 10 0.267 1.6 0 1.6 0.267 0 0.6 0.4 0.6 0 0 0.4 1 7 and 0.4 0 0.4 1.067 0 1.067 DB 10 0.267 1.6 0 1.6 0.267 0 0.6 0.4 0.6 0 0 0.4 1 VEL TECH 7 VEL TECH MULTI TECH 117 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH These two relationships will be used later in calculating stresses using e=DBeq. The multiplication teAeBeT DBe gives the element stiffness matrices. 1 2 3 4 7 8 Global of 0.983 0.5 0.45 0.2 0.533 0.3 1.4 0.3 1.2 0.2 0.2 0.45 0 0 0.3 k1 107 1.2 0.2 0 Symmetric 0.533 0 0.2 ww 5 6 7 8 3 4 Global of 0.983 0.5 0.45 0.2 0.533 0.3 1.4 0.3 1.2 0.2 0.2 0.45 0 0 0.3 k 2 107 1.2 0.2 0 Symmetric 0.533 0 0.2 w.E asy En In the previous element matrices, the global dof association is shown on top. In the problem under consideration, Q2,Q5, Q6, Q7, and Q8, are all zero. It is now sufficient to consider the stiffness associated with the degrees of freedom Q1,Q3 and Q4. Since the body forces are neglected, the first vector has the component f4 =-1000Ib. the set of equations is given by the matrix representation gin 0.983 0.45 0.2 Q1 0 10 0.45 0.983 0 Q2 0 0.2 0 1.4 Q3 1000 7 eer i ng. solving for Q1,Q3, and Q4, we get Q1=1.913 10-5 in. Q3=0.875 10-5 in. Q4=-7.436 10-5 in. net For element 1, the element nodal displacement vector is given by Q1=10-5[1.913,0,0.875,-7436,0,0]T The element stresses 1 are calculated from DB1q as 1=[-93.3-1138.7-62.3]Tpsi Similarly, Q2=10-5[0,0,0,0,0.875,-7.436]T 2= [93.4,23.4-297.4]Tpsi VEL TECH VEL TECH MULTI TECH 118 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 5. Derive stiffness matrix for a CST element by direct approach. Turner was first to suggest it and that was real starting point of FEM. Consider the typical element shown in fig. It is subjected to constant stresses along its all the three edges. Let the constant stresses be x, y, xy, =yx . Assembling stiffness matrix means finding nodal equivalent set of forces which are statically equivalent to the constant stress field acting at the edges of the elements. The equivalent nodal forces to be found are F1, F2, F3,<F6 as shown in Fig. we have six unknown nodal forces, but only three equilibrium. Hence it is not possible to determine F1, F2 <.. F6 in terms of x,y,xy mathematically. Turner resolved the uniform stress distribution into an equivalent force system at midsides as shown in Fig. Note side I is the side opposite to node. With this notation, ww w.E Fm1x x y3 y2 t xy x2 x3 t asy where t is the thickness of the element. Fm1y y x3 x 2 t xy y 3 y 2 t Fm2x x y3 y1 t xy x 3 x1 t Fm2y y x3 x1 t xy y 3 y1 t ...(7.1) En gin eer i ng. VEL TECH VEL TECH MULTI TECH 119 net VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Fm3x x y 2 y1 t xy x 2 x1 t Fm3y y x 2 x1 t xy y 2 y1 t After this Turner transferred half of mid side forces to nodes at the end of sides to get equivalent nodal forces. Thus he got ww w.E asy En gin 1 Fm2x Fm3x 2 1 = x y 3 y1 t xy x 3 x1 t x y 2 y1 t xy x 2 x1 t 2 1 = x y 3 y1 y 2 y1 xy x 3 x1 x 2 x1 2 F1 1 = x y 2 y 3 xy x 3 x 2 2 1 Fm1x Fm3x 2 1 = x y 3 y 2 t xy x 2 x 3 x y 2 y1 xy x 2 x1 2 1 = x y 3 y1 xy x1 x 3 2 F2 eer i ng. net 1 Fm1x Fm2x 2 1 = x y 3 y 2 xy x 2 x 3 x y 3 y1 xy x 3 x1 2 1 = x y1 y 2 xy x 2 x1 2 F3 VEL TECH VEL TECH MULTI TECH 120 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 1 Fm2x Fm3x 2 1 = y x 3 x1 xy y 3 y1 y x 2 x1 xy y 2 y1 2 1 = y x 3 x 2 xy y 2 y 3 2 F4 1 Fm1y Fm3y 2 1 = y x 2 x 3 xy y 3 y 2 y x 2 x1 xy y 2 y1 2 1 = y x1 x 3 xy y 3 y1 2 F5 ww w.E 1 Fm1y Fm2y 2 1 = y x 2 x 3 xy y 3 y 2 y x 3 x1 xy y 3 y1 2 1 = y x1 x 3 xy y 3 y1 2 F6 asy b1 b 2 b But, 3 0 0 0 and 0 c1 0 c 2 0 c3 T 2A B c1 b1 c 2 b2 c 3 b3 En gin eer i ....(7.3) ng. D DBe We have got F 1 T T 2A B DBe B DB tA e 2 net ...(7.4) =B DB V e T where V is the volume F k e where k B DB V B DB dV. T T (7.5) since [B] and [D] are constants. VEL TECH VEL TECH MULTI TECH 121 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 6. Explain in detail nodal loads by direct approach. In the stiffness equation [k] []=[F] refers to the nodal forces. Generally, while subdividing a structure, nodal locations are selected to as to coincide with the external forces applied. This can be easily done in case of concentrated loads acting on the structure. But in case of distributed loads like self weight, uniformly distributed load. Uniformly varying load, we need a technique of transferring the load as nodal load. There are two procedures for it. Namely direct procedure and variational approach. This procedure was first to be used in the finite element method. In this procedure classical structural analysis background is utilized or a portion is assigned to each node and load on that region as the nodal load. The latter method is called as lumped load method. ww w.E Consider the self weight of the uniform bar element shown in fig. Half the bar length may be asy En gin eer i ng. Assumed to contribute to each node. Hence at each node vertical downward load is net Al 2 where is unit weight of the material. A the cross sectional area and l is the length of the element. In case of beam element subject to uniformly distributed load. (i) Lumped load procedure: half the region is assigned to each node as shown in fig. (b) Its wl is taken at the centre of gravity of the element as shown in fig. (c). Then at the node, 2 wl2 wl the force are and . Hence Lumped load vector is 8 2 equivalent VEL TECH VEL TECH MULTI TECH 122 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH wl wl2 wl wl2 2 8 2 8 F T ww w.E asy (ii) Classical Structural Analysis Approach: In case of beam the end reactions, for a fixed beam wl2 wl and 12 2 En Hence the equivalent nodal loads are gin VEL TECH eer i ng. wl wl wl wl 2 12 2 12 2 2 VEL TECH MULTI TECH 123 net VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH (iii) In case of a CST element nodal force VEL TECH HIGH TECH 1 area may be assigned to each node and hence equivalent 3rd 1 the self weight as shown in fig. 3rd ww w.E For complex loading and elements this method may not be of much use. The distribution obtained in lumped load approach may be one of the possible distribution. asy En 7. Derive a local stiffness matrix for heat transfer using shape functions for a four –mode quadrilateral element. gin The variational function is written in matrix format. As in Eq. (f) of Prob. 3,4 and will be minimized with respect to the unknown variable that in this case is {T}. It follows that B kBT tdxdy N Qtdxdy T T A ng. (a) A eer i The first three matrices on the left-hand side of Eq. (a) are multiplied together to give a 44 local stiffness matrix. The local stiffness matrix is defined as K B k B tdxdy T A (b) net For illustration, the first term in the stiffness matrix appears as a b N N N N K11 1 k x 1 1 k y 1 tdxdy 0 0 x y y x (c) The derivatives of the shape functions are required before an attempt is made to evaluate the integral and are listed below for ready reference. Refer to Eq. (e) of prob3.1. VEL TECH VEL TECH MULTI TECH 124 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH N1 b y N1 b x x ab y ab N2 b y N2 x x ab y ab N3 N3 y x (d) x ab y ab N4 y N4 a x x ab y ab The proper derivatives are substituted into Eq. © to <<<<<<<<<<..in the stiffness matrix a ww K11 0 b 0 t 2 2 v v k y a x k y dxdy 2 ab 2 w.E integrating and substituting limits gives b k a k t K 2 2 x 11 y 3ab asy En The remaining terms in the stiffness matrix are evaluated in a similar manner. The local stiffness matrix appears as follows and the reader may verify the results as an exercise: 2b2k x 2a2k y 2b2k x a2k y 2b2k x 2a2k y t k 6ab Symmetric b2k x a2k y b k x 2a k y 2 2 2b k x 2a k y 2 2 gin b2k x a2k y b2k x a2k y (e) 2b2k x a2k y 2b2k x 2a2k y eer i ng. net The local stiffness matrix is symmetric, and some terms are repeated in the matrix. For instance, all diagonal terms are the same. The term on the right-hand side of Eq. (a) represents the distribution of the heat – source term. a x b y ab / 4 ab / 4 a b t x b y Qdxdy tQ 0 0 ab xy ab / 4 y a x ab / 4 (f) The heat source is distributed equally to the four nodes. VEL TECH VEL TECH MULTI TECH 125 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 8. Briefly explain about Displacement function for CST element. Unlike Spring and Beam elements. There is no deflection equation available for CST element. The displacement equation is derived by assuming an equation and then boundary conditions are applied to solve the equation. The displacement function is assumed to be a linear equation given by: U(x,y)=a1+a2x+a3y V(x,y)=a4+a5x+a6y (6.2.2) ww w.E asy En Apply the boundary conditions at nod ij, and k gin eer i ng. Ui=u(xi,yi)=a1+a2xi+a3yi Uj=u(xj,yj)=a1+a2xj+a3yj Um=u(xm,ym)=a1+a2xm+a3ym vi=v(xi,yi)=a4+a5xi+a6yi net vj=v(xj,yj)=a4+a5xj+a6yj vm=v(xm,ym)=a4+a5xm+a6ym Writing in matrix form, we get, VEL TECH VEL TECH MULTI TECH 126 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH ui 1 xi u j 1 x j u 1 x m m VEL TECH MULTI TECH VEL TECH HIGH TECH yi a1 y j a2 ym a3 and v i 1 xi v j 1 x j v 1 x m m y i a 4 y j a5 ym a6 The equation has the form ww {a}=[x]-1{u} w.E Solving for the coordinates i 1 x 1/(2A) i i where j j j m m m 1 xi 2A= 1 x j yi yj 1 xm ym asy En gin eer i The values of ,,I are found using the given nodal coordinates (x,y). ng. Now,. The coefficient values can be found in terms of the nodal coordinates and the boundary conditions. i a1 a 1/(2A) i 2 a3 i j j j m ui m u j m um (6.2.11) m v i m v j m v m (6.2.12) net and i a 4 a 1/(2A) i 5 a6 i j j j The deflection function or equation is, VEL TECH VEL TECH MULTI TECH 127 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH i u 1/(2A)1 x y i i and similarly, j i v 1/(2A)1 x y i i j j j j j m ui m u j m um VEL TECH HIGH TECH (6.2.14) m v i m v j m v m 9. Explain the general procedure when CST elements are in the usage. ww Step1: Field Variable and Element: w.E Since plane stress and plane strain problems are two dimensional problems, we need two dimensional elements. Any one from the family or triangular elements (CST/LST/QST) are ideally suited for these problems. Any one element from the family of two dimensional isoparametric elements also may be used. In these elements there are two degree of freedom at each node i.e. the displacement in x direction and displacement in y direction. Hence total degree of freedom in (i) (ii) asy En each element =2 No. of nodes per element Structure = 2 No. of nodes in entire structure. gin eer i For a CST element shown in Fig. the displacement vector may be taken as e 1 2 3 4 5 6 =u1 u2 u3 v1 v 2 v 3 T ng. 12.1a or as =u1 T v1 u2 v2 u3 v 3 ...(12.1b) net In most of the programs the order shown in equation (b) is selected. Hence the displacement vector {} is used in the form of equation (b) Then the x and y displacements of the node in global system are referred as 2n-1th and 2nth displacements. VEL TECH VEL TECH MULTI TECH 128 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH ww w.E Step2: Discritization asy Discritization of the structure should be made keeping in mind all the points listed. For all nodes x and y coordinates are to be supplied/ generated. Then nodal connectivity details is to be supplied. For the dam analysis problem shown in fig. the nodal connectivity detail is of the form shown in Table. Table Nodal connectivity En Element No. 1 2 : 7 8 : 10 : gin 1 1 2 2 2 7 3 7 8 4 4 11 5 10 11 6 11 11 eer i Local numbers Global Numbers ng. net Step 3: Shape/Interpolation Functions As shown in equation the shape function terms are N1 a b3 x c 3 y a1 b1x c1y a b2 x c 2 y ,N2 2 andN3 3 2A 2A 2A where a1 x 2 y 3 x 3 y 2 a2 x 3 y1 x1y 3 a3 x1y 2 x 2 y1 b1 y 2 y 3 b 2 y 3 y1 b3 y1 y 2 c1 x 3 x 2 c 2 x1 x 3 c 3 x 2 x1 VEL TECH VEL TECH MULTI TECH 129 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH 1 x1 2A= 1 x 2 1 x3 and VEL TECH HIGH TECH y1 y2 y3 when we select nodal displacement vector as shown in fig. (b). ux,y N1 0 N2 0 N3 0 ux,y e v x,y 0 N1 0 N2 0 N3 (12.3) Step 4: Element Properties ww Since strain vector w.E u x x v y y 2 u v y x asy En gin and nodal displacement vector is in the form 12.3, the strain displacement vector ({}=[B]{}),[B] is given by b1 0 b2 1 [B] 0 c1 0 2A c1 0 c 2 0 c2 0 b3 0 c3 0 c 3 0 eer i ng. (12.4) According to variational principal [k]e B [D][B]dv T net v Since [B]T,[D] are constant matrices we get [k]e=[B]T[D][B]v (12.5) where V=At This is exactly same as equation which was obtained by Turner by the direct approach. In equation [D] is the elasticity matrix, In case of isotropic materials, for plane stress case, VEL TECH VEL TECH MULTI TECH 130 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 1 0 E [D] 1 0 (12.6) 2 1 1 0 0 2 1 0 E [D] 1 0 (12.7) 1 1 2 1 2 0 0 2 Consistent Loads ww Consistent loads can be derived using the equation w.E Fe N xb dv N T ds T T (9.26) asy If there are nodal forces they are to be added directly to the vector {F}e En Step5: Global Properties gin Using nodal connectivity details the exact position of every term of stiffness matrix and nodal vector must be identified and placed in global stiffness matrix. eer i ng. Step6: Boundary Conditions net Since in most of the problems in plane stress and plane strain degree of freedom is quite high, the computers are to be used. These problems are not suitable for hand calculations. When computer programs are to be developed, imposition of boundary condition is conveniently done by penalty method. Step7: Solution of Simultaneous Equations Gauss elimination method or Cholesky’s decompositions method may be used. In elasticity problems, there exists symmetry and banded nature of stiffness matrix. Hence the programs are developed to store only half the bandwidth of stiffness matrix and solve simultaneous equations using Choleski’s decomposition method. Step8: Additional Calculations After getting nodal displacements stresses and strains in each element is assembled using the VEL TECH VEL TECH MULTI TECH 131 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH relations and Be DBe The calculated value of stress for an element is constant. It is assumed to represent the value at the centroid of the element. As a designer is normally interested in the principal stresses, for each element these values also may be calculated. 10. Find the nodal displacements and element stresses in the propped beam shown in fig. Idealize the beam into two CST elements as shown in the figure. Assume plane stress condition. Take = 0.25, E= 210=5 N/mm2, Thickness = 15mm. ww w.E asy En gin eer i ng. Solution: For element (1), global nodal numbers are 1,3,4. Local numbers 1,2,3 selected are indicated in Fig. Selecting node 4 as the origin of global coordinate system. net 1(0,0),2(750,500)and 3(0,500) VEL TECH VEL TECH MULTI TECH 132 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 1 0 0 2A 1 750 500 750 500 0 750 500 1 0 500 0 500 0 500 0 0 1 [B] 0 750 0 0 0 750 750 750 750 0 0 500 750 500 0 1 0 1 0 0 1 = 0 15 0 0 0 15 750 15 0 0 1 15 1 1 0 E [D] 1 0 1 1 2 1 2 0 0 2 ww w.E asy 0 0.75 0.25 3 1 0 2 105 5 0.25 0.75 0 0.2 10 1 3 0 1.25 0.5 0 0 0 1 0 0.25 En 0 1 0 1 0 3 1 0 0 1 5 [D][B] 0.2 10 1 3 0 0 15 0 0 0 15 75 0 0 1 15 0 0 1 15 1 15 3 0 3 15 0 2 105 0 45 1 0 1 45 750 15 0 0 1 15 1 gin eer i [K]1 tA[B]T [D][B] 0 15 0 0 15 0 0 15 3 0 3 15 0 0 0.2 105 15 750 500 1 1 0 45 1 0 1 45 0 1 750 2 750 0 15 0 0 1 15 1 1 0 15 1 0 15 u1 v1 u3 v3 u4 0 0 15 2.25 225 0 6.75 15 0 15 0 15 3.0 0 3.0 100000 0 0 1 15 15 225 15 3 15 525 6.75 15 1 3.0 15 VEL TECH ng. net v 4 Global 150 u1 6.75 v1 15 u3 1.0 v 3 3.0 u4 7.75 v 4 VEL TECH MULTI TECH 133 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH For element (2), Local and global node numbers are as shown in fig. ww w.E The coordinates of nodes are 1(0,0),2(750,0)3(750,500) asy En gin b1 y 2 y 3 =-500 b 2 y 3 y1=-500 b 3 y1 y 2 0 c1 x 3 x 2 =0 c 2 x1 x 3 =-750 c 3 x 2 x1 750 1 0 0 2A 1 750 0 1 750 500 750 500 1 750 500 ng. 0 500 0 0 0 500 1 [B] 0 0 0 750 0 750 750 500 0 500 750 500 750 0 = eer i net 0 1.0 0 0 0 1.0 1 0 0 0 750 0 750 750 0 500 750 500 750 0 3 1 0 [D] 0.2 10 1 3 0 same as for element 1. 0 0 1 5 VEL TECH VEL TECH MULTI TECH 134 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 0 1.0 0 0 0 3 1 0 1.0 1 5 [D][B] 0.2 10 1 3 0 0 0 0 1.5 0 15 750 1.0 0 1.5 1.0 1.5 0 0 0 1 0 0 3.0 1.5 0 1.5 3.0 2 105 1.0 0 1.0 4.5 0 4.5 750 0 1.0 1.5 1.0 1.5 0 [K]2 tA[B]T [D][B] 0 10 0 0 0 1.0 3.0 1.5 0 1.5 3.0 0 0 1.5 0.2 105 750 500 1 1.0 [k]2 15 1.0 0 1.0 4.5 0 4.5 2 750 0 1.5 1.0 750 0 1.0 1.5 1.0 1.5 0 1 0 1.5 1.5 0 0 ww u1 v1 w.E u2 v2 asy u3 v3 0 3.0 1.5 0 1.5 u1 3.0 0 1.0 1.5 1.0 1.5 0 v1 3.0 1.5 5.25 3.0 2.25 1.5 u2 10000 1.5 6.75 v 2 1.5 1.0 3.0 7.75 0 1.5 2.25 1.1 2.25 0 u3 0 1.5 6.75 0 6.25 v 3 1.5 En [k]=100 000 gin eer i ng. VEL TECH VEL TECH MULTI TECH 135 net VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH {F}T=[0 0 0 0 50000 0 0 0] The equation is [k][]={F} 5.5 0 3.0 1.5 i.e.,100000 0 3.0 2.25 1.5 1.50 1 0 6.25 2 0 0 3 0 0 4 0 1.5 5 50000 1.0 6 0 3.0 7 0 7.75 8 0 0 3.0 1.5 0 3.0 2.25 7.25 1.5 1.0 3.0 0 1.5 1.5 5.25 30 225 1.5 0 1.0 3.0 7.25 1.5 6.75 0 3.0 2.25 1.5 5.25 0 3.0 0 15 6.75 0 7.75 1.5 1.5 0 0 3.0 1.5 5.25 6.75 0 0 1.5 1.0 3.0 ww w.E The boundary conditions are 1 2 4 7 8 0 Reduced equation is, asy En 5.25 2.25 1.5 3 0 100000 2.25 5.25 0 5 50000 1.5 0 7.75 6 0 gin eer i ng. 5.25 2.25 1.5 3 0 2.25 5.25 0 5 0.5 1.5 0 7.75 6 0 1.5 3 0 5.25 2.25 0 4.2857 0.6429 5 0.5 0 0.6429 7.3214 6 0 net 1.5 3 0 5.25 2.25 0 4.2857 0.6429 5 0.5 0 0 7.17139 6 0.075 VEL TECH VEL TECH MULTI TECH 136 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 6 0.010459 4.2857 5 0.6429 0.010 0.5 5 0.118236 5.25 3 2.25 0.118236 1.5 0.010459 0 3 0.053661 0 0 0.53661 0 0.118236 0.010459 0 0 T ww 1 DB 0 0 5.584 0 1.5 3 0 3 1.5 0.053661 2 105 0 4.5 1 0 1 4.5 2.977 0 750 5.000 1.5 0 0 1 1.5 1.0 0.118236 0.010459 w.E asy En 0 0 9.877 3.0 0 3.0 1.5 0 1.5 0.118236 2 105 1.0 0 1.0 4.5 0 4.5 2 4.408 0.010459 750 5.008 0 1.0 1.5 1.0 1.5 0 0 0 gin eer i ng. 11. (i) Derive the expression for consistent load vector due to self weight in a CST element. net Solution: The general expression for the consistent load in any element due to the body force is Fe [N]T xb dv v 0 For self weight x b where is unit weight of the material It is advantageous to take interpolation functions in the natural coordinate systems, since closed form integration formulae can be used. VEL TECH VEL TECH MULTI TECH 137 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH We know for CST element. L 0 1 N 01 L L 2 0 L3 0 0 L2 0 L3 u When nodal vector selected is in the order 1 v1 L1 0 0 0 L L 1 1 L 0 0 0 [F]e 2 hdA hdA 0 L 2 L A A 2 L3 0 0 0 L3 L3 ww w.E asy Nothing that the standard integration formula is p!q!r! L L L dA p q r 2 2A p q r 1 2 3 A we get L1hdA A En gin 1!0!0! 1 hA h2A 2A 23 3 1 0 0 2 hA 3 A hA and L3dA 3 A ng. Similarly L 2dA 0 1 hA 0 Fe Answer 3 1 0 1 VEL TECH eer i VEL TECH MULTI TECH 138 net VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 11. (ii) Find the expression for nodal vector in a CST element subject to pressure Px1, Py1 on side 1. Px2, Py2 on side 2 and Px3, Py3 on side 3 as shown in fig. ww w.E asy En gin eer i Solution: Let us first consider nodal vector due to pressure Pxt and Pyt only. ng. We know in CST element [N]=[L] Along side 1-L2 i.e., L3=1-L2 net ds1=l1 dL2, when s is measured form node 3 towards 2. The surface force are P xs Px1 y1 Hence the line integral form exists for nodal force vector as given below: Fe [N]T xs ds1t VEL TECH VEL TECH MULTI TECH 139 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 0 0 0 0 0 0 t 1 L P L 2 0 Px1 2 x1 t ds t L P ds 0 L 2 Py1 o o 2 y1 L P L 3 0 3 x1 L3Py1 0 L3 Nothing that the standard integration form for natural coordinate system is ww p!q! L L ds p q 1! l p q 1 2 we get L 2x1ds l 1!0! l1Px1 1 Px1 2 1 0 1! w.E l L2Py1ds 21 Py1 l 0!1! L3Px1ds 0 1 1 l1Px1 21 Py1 asy l1 L P ds 2 P and 3 x1 Fe T y1 tl1 0 0 Px1 Py1 Px1 Py1 2 En gin ng. similarly due to forces on side 2 we get Fe tl2 Px2 Py2 T 0 0 Px2 Py2 and due to forces on side 3, Fe tl3 Px3 Py3 Px3 Py3 T eer i net 0 0 Nodal vector due to forces on all the three sides is, l2 Px2 l P y1 2 l P Fe l1 Px1 y1 1 l P x1 1 l1 Py1 VEL TECH l3 Px3 l3 Py3 l3 Px3 l3 Py3 l2 Px2 l2 Py2 VEL TECH MULTI TECH 140 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 12. A three-node triangular element is defined in fig. in an x,y coordinate system with nodes 1,2, and 3 located at (x1,y1), (x2,y2), and (x3,y3), respectively, in the global system. Derive shape functions in terms of global coordinates. ww w.E asy Assume an interpolation function to represent the variation of the unknown quantity =C1+C2x+C3y En (a) gin Write the interpolation function as a matrix equation C1 1 x y C2 or in matrix format =[]{c} (b) C 3 eer i The boundary condition are in terms of nodal point values of : x1,y1 1 x2,y2 2 x3,y3 3 (c) ng. net Substitute Eqs. © into Eq. (a) to obtain three equations that can be solved for C 1,C2, and C3: 1 C1 C2 x1 C3 y1 1 1 x1 2 C1 C2 x 2 C3 y 2 or 2 1 x 2 1 x 3 C1 C2 x3 C3 y3 3 3 y1 C1 y 2 C2 (d) y3 C3 Write Eq. (d) in matrix form as {}=[x]{c} VEL TECH (e) VEL TECH MULTI TECH 141 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH and solve for {c} as {c}[x]-1{} (f) substitute Eq. (f) into Eq. (b): =[][x]-1{}=[N] {} (g) The shape functions are the product of the first two matrics on the right- hand side of Eq. (g) ww [N]=[][x]-1 or [N1 N2 N3]=[1 x y][x]-1 (h) Solving Eq. (h) gives w.E N1 x 2 y 3 x 3 y 2 x y 2 y 3 y x 3 x 2 / 2A asy N2 x 3 y1 x1y 3 x y 3 y1 y x1 x 3 / 2A N3 x1y 2 x 2 y1 x y1 y 2 y x 2 x1 / 2A (i) where 1 x1 1 A= det 1 x 2 2 1 x 3 y1 y 2 (j) y 3 En gin eer i ng. A is the area of the triangular element. net 13. Derive a local stiffness matrix for heat transfer using the three-node triangular element defined in Fig. The local stiffness matrix is defined by Eq. (b) of Prob: T [k] [B] [B]t dx dy (a) The temperature is defined in terms of the unknown temperature at each node and is similar to Eq.(b) of Prob. Except there are only three nodal values: T = [ N1 N2 N3] { T1 T2 T3}T = [N] {T}T (b) The shape function given by Eqs. (i) of Prob. And, for convenience, are usually written in the following form VEL TECH VEL TECH MULTI TECH 142 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH a1 b1x c1y 2A a b2 x c 2 y N2 2 2A a b3 x c 3 y N3 3 2A VEL TECH HIGH TECH N1 where (c) a1 x 2 y3 x3 y 2 b1 y 2 y 3 c1 x 3 x 2 a2 x3 y1 x1y3 b2 y3 y1 c 2 x1 x 3 a3 x1y 2 x 2 y1 b3 y1 y 2 c 3 x 2 x1 ww Define the shape function matrix of Eq. (b), using Eqs. (c) as [N] [1 x w.E a1 a2 a3 y] b1 b2 b3 2A c1 c 2 c 3 asy En The [B] matrix is formed as [L] [N] where the operator matrix [L] is given by Eq. (e) of Prob. For the heat-transfer problem or any physical problem governed by an equation of the form of Eq. when = 0: or gin a1 a2 a3 / x 0 1 0 L N / y N 0 0 1 b1 b2 b3 2A c c c 2 3 1 b1 b2 b3 [B] 2A c1 c 2 c 3 eer i ng. net The material matrix [k] of Eq. (a) is a 2 x 2 matrix and is the same as that in Prob. The area integration is dx dy = A, the area of the triangle, since all terms are constant. Substituting into Eq. (a) gives the symmetric matrix b12k x c12k y b1b2k x c1c 2k y b1b3k x c1c 3k y t (d) [k] b22k x c 22k y b2b3k x c 2c 3k y 4A b3 2k x c 3 2k y Equation (d) can be used for numerical computation, however, for computer implementation it is probably more practical to formulate the individual matrices and use a matrix multiplication routine to compute [k]. VEL TECH VEL TECH MULTI TECH 143 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 14. Derive the shape functions for the nine-node rectangular element. ww Figure. w.E asy En The nine-node element has midside nodes located at the midpoint of each side of the element, and the ninth node located at the center of the element. The node numbering shown is standard, with midside nodes numbered 5 through 8, but the nodes can be numbered in any sequence. In this case the dimensions of the element (the local coordinate system) are assumed as a and b. Other dimensions, such as 2a x 2b could also be used. A local coordinate system similar to –a and –b will be introduced. gin eer i ng. net The shape function can be derived using the results of Prob. And fig. The global x coordinates of Fig. are reproduced in Fig. along with corresponding y coordinates. The results of Prob. Are repeated for reference and given an extra subscript to indicate the x direction in the global system: N1x (x x 2 )(x x 3 ) (x x1)(x x 3 ) (x x1 )(x x 2 ) N5x N2x (x1 x 2 )(x1 x3 ) (x 2 x1 )(x 2 x 3 ) (x 3 x1)(x 3 x 2 ) (a) In the local system let node 1 correspond to x1 = 0 a/2, and x3 = a. The shape function , in the x direction only, along nodes1,5 and 2 is obtained by substituting into Eq.(a): N1x x (a / 2)(x a) (2x a)(x a) a2 0 (a / 2)(0 a) VEL TECH (b) VEL TECH MULTI TECH 144 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Similarly, N5x for the nine-node element can be constructed using N5x of Eqs.(a): N5x (x 0)(x a) x(x a) (a / 2) 0(a / 2) a (a2 / 4) (c) Visualize that the shape function of Prob. Could have been derived along the y axis of the global system using y1,y2, and y3 of Fig. and the result would be Eqs.(a) with all x value replaced by y values. A shape function for node 1 of the nine-node element would be identical to Eq.(b) with x and a replaced by y and b, respectively. The two-dimensional shape function for node 1 is the product of the two one-dimensional shape functions for node 1: ww N1 N1x N1y (2x a)(x a)(2y b)(y b) a 2b 2 w.E (d) substituting local coordinates corresponding to node 1(x = 0, y = 0) into Ew.(d) will give N1 = 1, and the value of N1 all other nodes will be zero. asy En The y contribution of the two-dimensional shape function for node 5 is constructed using the first Eq.(a) and is the same as that used in Eq.(d) or, corresponding to Fig. is written as N5y (y y 9 )(y y 7 ) (2y b)(y b) (y 5 y 9 )(y 5 y 7 b2 gin (e) eer i ng. Obviously, the nine-node rectangular element is made up of combinations of quadratic shape functions. It follows that N5 N5x N5x 4x(x a)(2b y)(y b) a 2b 2 (f) net A significant result is that two- or three-dimensional shape functions can be constructed from one dimensional shape functions. The remaining shape functions are obtained in a similar manner and, as an exercise, the reader should verify the results: VEL TECH VEL TECH MULTI TECH 145 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH x(2x a)(2y b)(y b) a2b2 x(2x a)y(2y b) N3 a2b2 (2x a)(x a)(2y b) N4 a2b2 4xy(2x a)(y b) N6 a2b2 4xy(x a)(2y b) N7 a2b2 4y(2x a)(x a)(y b) N8 a2b2 16xy(x a)(y b) N9 a2b2 N2 ww w.E asy 15. Derive: Two dimensional shape function for a four –model quadrilateral element. En Consider the general quadrilateral element shown in Fig. The local nodes are numbered as 1,2,3, and 4 in a counter clock wise fashion as shown, and (xi,yi) are the coordinates of node i. The vector q=[q1, q2,<q8]T Denotes the element displacement vector. The displacement of an interior point P located at (x,y) is represented as u= [u(x,y),v(x,y)]T. Shape Functions gin eer i ng. We first develop the shape functions on a master element. Shown in fig. the master element is defined in -,-coordinates (or natural coordinates) and is square shaped. The Lagrange shape functions where i=1,2,3. and 4, are defined such that N1 is equal to unity at node I and is zero at other nodes. In particular, consider the definition of Ni: net N1 =1 at node 1 =0 at nodes 2,3, and 4 VEL TECH VEL TECH MULTI TECH 146 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH ww w.E asy En gin Now, the requirement that N1 =0 at nodes 2,3, and 4 is equivalent to requiring that N1=0 along edges =+and =+1 Thus, N1 has to be of the form N1=c(1-)(1-) ng. (7.2) Which yields c=1/4 Thus, N1 1 1 (1 ) 4 eer i (7.4) net all the four shapes functions can be written as 1 1 (1 ) 4 1 N2 1 (1 ) 4 1 N3 1 (1 ) 4 1 N4 1 (1 ) 4 N1 VEL TECH (7.5) VEL TECH MULTI TECH 147 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH While implementing in a computer program, the compact representation of Eqs. Is useful. N1 1 1 1 (1 1 ) (7.6) 4 where (1,1) are the coordinates of node i. We now express the displacement field within the element in terms of the nodal values. Thus, if u=[u,v]T represents the displacement components of a point located at (, ), and q, dimension (8x1), is the element displacement vector, then ww u N1q1 N2q3 N3 q5 N4 q7 v N1q2 N2q4 N3 q6 N4 q8 (7.7a) w.E which can be written in matrix form as u=Nq where (7.7b) asy N 0 N2 0 N3 0 N4 0 N 1 0 N1 0 N1 0 N3 0 N4 En (7.8) gin eer i In the isoparametric formulation, we use the same shape functions N 1 to also express the coordinates of a point within the element in terms of nodal coordinates. Thus, ng. x N1x1 N2 x 2 N3 x 3 N4 x 4 y N1y1 N2 y 2 N3 y 3 N4 y 4 (7.9) net subsequently, we will need to express the derivatives of a function in x-y- coordinates in terms of its derivatives in -,- coordinates. This is done as follows: A function f=f(x,y), in view of Eqs. Can be considered to be an implicit function of and as f=f[x (,),y(,)].Using the chain rule of differentiation, we have f f x f y x y f f x f y x y or VEL TECH VEL TECH MULTI TECH 148 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH f f x J f f y x J x y y VEL TECH HIGH TECH (7.11) (7.12) ww In view of Eqs. 7.5 and 7.9, we have w.E 1 1 x1 1 x 2 1 x 3 1 x 4 1 y1 1 y 2 1 y 3 1 y 4 4 1 x1 1 x 2 1 x 3 1 x 4 1 y1 1 y 2 1 y 3 1 y 4 J J 11 12 J J 21 22 J asy Equation 7.11 can be inverted as f f x J (7.14a) f f y En gin eer i ng. or f f x 1 J22 J12 J (7.14b) f J det J 21 11 f y net These expression will be used in the derivation of the element stiffness matrix. An additional result that will be needed is the relation. Dxdy=detJ dd. VEL TECH VEL TECH MULTI TECH 149 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH UNIT – IV PART – A 1. Write down the potential energy formula in terms of elemental volume. 2 2 2 1 T rdAd uT frdAd uTTrdld uiTPi 2 0 A i 0 A o Where rdld is the elemental surface area and the point load Pi represents a line load distributed around a circle. ww w.E 2. Derive the body force equation. Consider the body force term 2 UT fr dA. We have 2 u fr dA 2 (uf fz )r dA T e e asy e En 2 [(N1q1 N2 q3 N3 q5 )fr (N1q2 N2 q4 N3 q6 )fz ]r dA e gin Once again, approximating the variable quantities by their values at the centroid of the triangle, we get 2 uT fr dA = qT f e e eer i ng. Where the elements body force vector fe us given by fe 2 rA e [f r ,f z ,f r ,f z ,f r ,f z ,]T 3 net The bar on the f terms indicates that they are evaluated at the centroid. Where body force is the primary load, greater accuracy may be obtained by substituting r = N 1r1 + N1r2 + N1r3 into equation and integrating to get nodal loads. 3. Write down the Galerkin formulation of axisymmetric elements. 2 T ( )rdA (2 T frdA 2 T Trdl iTPi A VEL TECH A L VEL TECH MULTI TECH 150 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Where, [r ,z ]T ( ) [ r z r z r , , , r z z r r 4. Where we use the triangular element in axisymmetric problem? The two dimensional region defined by the revolving area is divided into triangular elements. Though each element is completely represented by the area in the rz plane, in reality. It is a ring shaped solid of revolution obtained by revolving the triangle about the z- axis. ww 5. Write down the displacement equation in axisymmetric element. w.E Using the three shape functions N1, N2 and N3. We define u= Nq u= [u, w] T N 0 N2 0 N3 0 N 1 0 N1 0 N2 0 N3 asy En q = [q1, q2 q3, q4, q5, q6] If N1 = and N2 = and note that N3 1 - - gives gin U = q1 + q3 + (1 - - q6) U = q2 + q4 + (1 - - q6) 6. Write down the Jacobian equation. eer i ng. net r r J 13 13 r23 r23 det J r13 z23 r23 z13 7. How did you calculate the force in rotating wheel? Let us consider a rotating flywheel with is axis in the z direction. We consider the flywheel to be stationary and apply the equivalent radial centrifugal (inertial) force per unit volume of pr2, where is the density (mass per unit volume), and the angular velocity in rad/s. In addition, if gravity acts along the negative z-axis, then VEL TECH VEL TECH MULTI TECH 151 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH F = [fr, fz]T = [r2, - g]T and f r r 2 ,f z g For more precise results with coarse meshes, we need to use r = N1r1 + N1r2 + N1r3 and integrate. 8. Drive the surface traction equation. For a uniformly distributed load with components Tr and Tz, shown in figure, on the edge connecting nodes 1 and 2, we get 2 uT Trd qT T e ww w.E Where, q = [q1, q2 q3, q4,] T asy Te 2 12 [aTr ,aTz ,aTr ,aTz ,]T a e En 2r1 r2 r 2r2 b 1 6 6 12 (r2 r1 )2 (z2 z1 )2 gin eer i In this derivation, r is expressed as N1r1 + N1r2 and then integrated. When the line 1-2 is pa rallel to the z-axis, we have r1 = r2, which gives a = b = 0.5r1. 9. Write the equation for internal and external virtual work. The first term representing the internal virtual work gives 2 T ( )r dA ng. net A Internal virtual work = 2 qTBTDB rdA e e q K e T e Where the elements stiffness Ke is given by T K e 2 rA e B DB We note that Ke is symmetric. Using the connectivity of the elements, the internal virtual work can be expressed in the form VEL TECH VEL TECH MULTI TECH 152 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH q K K q T Internal virtual work = VEL TECH HIGH TECH e T e e = TKQ Where K is the global stiffness matrix. The external virtual work terms in equation involving body forces, surface tractions, and point loads can be treated in the same way as in the potential energy approach, by replacing q with . The summation of all the force terms over the elements then yields. External virtual work = TF. ww 10. How the stress can be calculated in axisymmetric problem? w.E From the set of nodal displacements the elements nodal displacements q can be found using the connectivity. Then, using stress strain relation in = d and strain displacement relation in the equation = Bq we get DBq , where B is b and it is evaluated at the centroid of the element. asy En 11. Discuss the effect of temperature on axisymmetric elements. gin Uniform increase in temperature of T introduces initial normal strains 0 given as 0 = [T, T, 0, T]T eer i ng. The stresses are given by = D ( - 0) Where is the total strain. net On substitution into the strain energy, this yields an additional term of -TD0 in the potential energy II. Using the element strain displacement relation in equation we find that 2 T D 0 r dA qT (2 rA e B D0 T A e The consideration of the temperature effect in the Galerkin approach is rather simple. The term T in equation is replaced by T (). The expression in parentheses gives elements nodal contributions. The vector 0 is the initial strain evaluated at the centroid, representing the average temperature rise of the element. We have VEL TECH VEL TECH MULTI TECH 153 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH T e 2 rA e B D0 e = [1, 2, 3, 4, 5, 6]T 12. Explain in brief cylinder subjected to internal pressure. Following figure shows a hollow cylinder of length L subjected to an internal pressure. One end of the cylindrical pipe is attached to a rigid wall. In this, we need to model only the rectangular region of the length L bound between ri and r0. Nodes on the fixed end are constrained in the z and r directions. Stiffness and force modifications will be made for these nodes. ww Figure: Hollow cylinder under internal pressure. w.E asy En gin eer i ng. 13. Briefly explain about infinite cylinder. net The modeling of a cylinder of infinitive length subjected to external pressure is shown below. In that the length dimensions are assumed to remain constant. This plane strain condition is modeled by considering a unit length. Figure: cylinder of infinite length under external pressure. VEL TECH VEL TECH MULTI TECH 154 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 14. Write down the element strain displacement matrix equation. The element strain displacement matrix, of dimension, is given by z 21 z12 z 23 0 0 0 det J det J det J r32 r13 r21 0 0 0 det J det J det J B z 23 r13 z31 r21 z12 r32 det J det J det J det J det J det J N N3 N2 1 0 0 0 r r r ww w.E PART – B asy 1. Derive the equation of axisymmetric formulation. En Considering the elemental volume shown figure, the potential energy can be written in the form gin 2 2 1 2 T r dA d - uT fr dA d uTTr d d - uiT Pi (1) 0 A 0 A 0 L 2 i eer i ng. where r d d is the elemental surface area and the point load Pi represents a line load distributed around a circle, as shown in figure. 1 2 T r dA - uTfr dA uTTr d ui T Pi (2) A L 2 A i where u = [u, w]T <.(3) f = [fr, fz]T <..(4) T = [Tr, Tz]T <..(5) net We can write the relationship between strains and displacements u as VEL TECH VEL TECH MULTI TECH 155 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH r ,z , rz , T T u w u w u = , , , r z z r r .....(6) The stress vector is correspondingly defined as r , z ,Trz , T ....(7) The stress-strain relations are given in the usual form, viz., ww D ...(8) w.E asy En gin Elemental volume eer i ng. net Deformation of elemental volume VEL TECH VEL TECH MULTI TECH 156 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Where the (4 4) matrix D can be written by dropping the appropriate terms from the threedimensional matrix, as 1 v E 1 v 1 v D 1 v (1 2v ) 0 v 1 v ww v 1 v v 0 1 v 1 2v 0 2(1 v ) 0 1 v 1 v 0 1 0 v 1 v w.E ...(9) In the Galerkin formulation, we require 2 T ( )r dA- 2 T fr dA + 2 TTr d + iT Pi 0 ...(10) A A L where r , z T asy .....(11) T ( ) = r , z , r , z , r r z z r r En .....(12) gin eer i 2. Write down the Jacobian matrix equations for axisymmetric element. ng. net The two-dimensional region defined by the revolving area is divided into triangular elements, as shown in figure. Though each element is completely represented by the area in the rz plane, in reality, it is a ring-shaped slid of revolution obtained by revolving the triangle about the z-axis. A typical element is shown in figure. The definition of connectivity of elements and the nodal coordinates follow the steps involved in the CST element discussed in Section. We note here that the r-and z-coordinates, respectively, replace x and y. Using the three shape functions N1, N2, and N3 we define U = Nq VEL TECH <<(1) VEL TECH MULTI TECH 157 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH where u is defined in (1) and N 0 N2 N 1 0 N1 0 0 N2 N3 0 0 .....(2) N3 q = q1, q2 , q3 , q4 , q5 , q6 .....(3) T If we denote N1 = and N2 = , and note that N3 = 1 - - , then Eq. 1 gives u = q1 + q3 + (1 - - )q5 w = q2 + q4 + (1 - - )q6 <(4) ww w.E asy En gin eer i ng. net Axisymmetric triangular element VEL TECH VEL TECH MULTI TECH 158 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH By using the isoparametric representation, we find r r1 r2 1 r3 z z1 z2 1 z3 .....(5) The chain rule of differentiation gives u u r J ....(6) u u z and ww w.E w w r J ....(7) w w z asy where the Jacobian is given by r J 13 r23 En z13 .....(8) z23 gin eer i ng. In the definition of J earlier, we have used the notation rij = ri – ri and zij = zi – zj. The determinant of J is det J = r13z23 – r23 z13 <<. (9) net Recall that |det J| = 2Ae. That is, the absolute value of the determinant of J equals twice the area of the element. The inverse relations for Eqs. 6 and 7 are given by u w u w r r 1 1 J and J .....(10) u u w w z z VEL TECH VEL TECH MULTI TECH 159 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH where J 1 1 z23 det J r23 z13 .....(11) z13 Introducing these transformation relationships into the strain-displacement relations in Eq. (4), we get z23 q1 q5 z13 q3 q5 det J r23 q2 q6 r13 q4 q6 det J r23 q1 q5 r13 (q3 q5 ) z23 (q2 q6 ) z13 q4 q6 det J N1q1 N2q3 N3 q5 r ww w.E asy This can be written in the matrix form as Bq <(12) En gin eer i where the element strain-displacement matrix, of dimension (4 6), is given by z23 det J 0 B r32 det J N1 r 0 r32 det J z23 det J 0 z31 det J 0 r13 det J N2 r 0 r13 det J z31 det J 0 z12 det J 0 r21 det J N3 r 0 r21 det J .....(13) z12 det J 0 ng. net 3. (a) An axisymmetric body with a linearly distributed load on the conical surface is shown in figure. Determine the equivalent point loads at nodes 2, 4, and 6. VEL TECH VEL TECH MULTI TECH 160 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Solution: We approximate the linearly distributed load by the average uniformly distributed loads on the edges 6-4 and 4-2 as shown in figure. Relationships for more precise modeling of a linearly distributed load are provided in Problem. We now consider the two edges 6-4 and 4-2 separately and then merge them. ww For edge 6-4 w.E asy p 0.35 MPa, r1 60mm, z1 40mm, r2 40mm, z 2 55mm 1-2 c= r1 r2 z1 z2 25mm 2 z 2 z1 0.6, 1 2 2 s= r1 r2 0.8 1 2 En Tr pc 0.21, Tz ps 0.28 a= 2r1 r2 r 2r2 26.67, b = 1 23.33 6 6 gin eer i ng. T T1 2 1 2 aTr , aTz,bTr , bTz = -879.65 - 1172.9 -769.69 - 1026.25 N T These loads add to F11, F12, and F8, respectively. For edge 4-2 net p 0.25 MPa, r1 40mm, z1 55mm, r2 20mm, z 2 70mm 1-2 c= r1 r2 z1 z2 25mm 2 z 2 z1 0.6, 1 2 2 s= r1 r2 0.8 1 2 Tr pc 0.15, Tz ps 0.2 a= 2r1 r2 r 2r2 16.67, b = 1 13.33 6 6 T T1 2 1 2 aTr , aTz,bTr , bTz = -392.7 - 523.6 -314.16 - 418.88 N T VEL TECH VEL TECH MULTI TECH 161 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH These loads add to F1, F8, F3, and F4, respectively. Thus, [F3 F4 F7 F8 F11 F12] = [-314.2 -418.9 –1162.4 –1696.5 –879.7 – 1172.9] N 3(b) Derive the potential-energy equation. Potential-Energy Approach The potential energy on the discretized region is given by ww 1 2 T D r dA 2 uT fr dA - 2 uTTr d e e e e 2 T - ui Pi ....(1) w.E a sy The element strain energy Ue given by the first term can be written as En 1 Ue qT 2 BT DBr dA q e 2 .....(2) gin The quantity inside the parentheses is the element stiffness matrix, k e 2 BT DBr dA e ....(3) eer i ng. The fourth row in B has terms of the type Ni/r. Further, this integral also has an additional r in it. As a simple approximation, B and r can be evaluated at the centroid of the triangle and used as representative values for the triangle. At the centroid of the triangle, N1 = N2 = N3 = 1 3 <.(4) net And r r r r = 1 2 3 3 where r is the radius of the centroid. Denoting B as the element strain-displacement matrix B evaluated at the centroid, we get k e 2 r BTDB dA e or k e 2 r A eBT DB VEL TECH ...(5) VEL TECH MULTI TECH 162 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH We note here that 2 r A e is the volume of the ring-shaped element shown in figure. Also, Ae is given by 1 det J 2 Ae ....(6) 4. In figure, a long cylinder of inside diameter 80 mm and outside diameter 120 mm snugly fits in a hole over its full length. The cylinder is then subjected to an internal pressure of 2 MPa. Using two elements on the 10-mm length shown, find the displacements at the inner radius. ww w.E asy En gin Solution Consider the following table: Element 1 2 Connectivity 1 2 3 1 2 4 2 3 4 Node 1 2 3 4 eer i Coordinates r z 40 10 40 0 60 0 60 10 ng. net We will use the units of millimeters for length, newtons for force, and megapascals for stress and E. These units are consistent. On substituting E = 200 000 MPa and v = 0.3, we have 2.69 105 1.15 105 1.15 105 2.69 105 D 0 0 5 5 1.15 10 1.15 110 VEL TECH 0 1.15 105 0 1.15 105 0.77 105 0 0 2.69 10 5 VEL TECH MULTI TECH 163 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH for both elements, det J = 200mm2 and Ae = 100mm2. forces F1 and F3 are given by F1 f3 2 r1 ePi 2 40 10 2 2514N 2 2 The B matrices relating element strains to nodal displacements are obtained first. For element 1, 1 r 40 40 60 46.67 mm and 3 0 0 0 0.05 0 0.05 0 0.1 0 0.1 0 0 1 B 0.1 0.05 0.1 0 0 0.05 0 0.0071 0 0.0071 0 0.0071 ww w.E 1 40 60 60 53.33mm and 3 0 0.05 0 0 0 0.05 0 0 0 0.1 0 0.1 B2 0 0.05 0.1 0.05 0.1 0 0 0.00625 0 0.00625 0 0.00625 For element 2, r= asy En gin eer i The element stress-displacement matrices are obtained by multiplying DB: 1.15 1.26 0.49 2.69 DB1 104 0.77 0.385 0.384 1.15 0.082 1.15 1.43 0 0.082 2.69 0.657 0.1 0.77 0 0 0.385 0.191 1.15 0.766 0 ng. 0 1.42 1.15 0.072 1.15 1.27 0.503 0 0.647 2.69 0.072 2.69 DB2 10 4 0 0.385 0.77 0.385 0.77 0 0 0.743 1.15 0.168 1.15 0.407 net The stiffness matrices are obtained by fixing 2 rAe BT DB for each element: VEL TECH VEL TECH MULTI TECH 164 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH Global of VEL TECH MULTI TECH 1 2 3 4 7 VEL TECH HIGH TECH 8 -2.58 -2.34 1.45 -1.932 1.13 4.03 8.45 1.37 -7.89 1.93 -0.565 2.30 -0.24 0.16 -1.13 k1 107 7.89 -1.93 0 Symmetric 2.25 0 0.565 Global of 3 4 5 6 7 8 0 -2.22 1.69 -0.085 -1.69 2.05 0.645 1.29 -0.645 -1.29 0 5.11 -3.46 -2.42 2.17 k 2 107 9.66 1.05 -9.01 Symmetric 2.62 0.241 9.01 ww w.E asy Using the elimination approach, on assembling the matrices with reference to the degrees of freedom 1 and 3, we get 4.03 2.34 Q1 2514 107 2.34 4.35 Q3 2514 so that En gin Q1 0.014 102 mm eer i ng. Q3 0.0133 102 mm. 5. Explain the following i) Stress distribution in thick cylinder. ii) Business of problem iii) Stress distribution around the thread of a bolt. net 6. Derive the shape functions and the element Stiffness matrix for an axisymmetric annular ring element as shown in figure. ANNULAR RING ELEMENT VEL TECH VEL TECH MULTI TECH 165 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Being an axisymmetric problem the radial displacement u depends on r 1 U 1 2 r = < 1 r > 2 .....(1) 1 U1 1 r1 1 r1 U1 or a 1 r2 1 r2 U2 U2 = Hence r2 r1 U1 1 r2 r1 1 1 U2 ww U 1r > .....(2) w.E r2 r1 U1 1 r2 r1 1 1 U2 asy The strain displacement is written as U / r r U / r 0 1 1 r2 = 1 1 r2 r1 1 r ....(3) r1 U1 1 U 2 En gin ....(4) ng. U1 = B U2 r 1 v r U1 = E C B ....(5) v 1 U2 K = BTC B dv eer i net ......(6) r2 = 2 t BT C B r dr. ......(7) r1 Where ‘t’ is the thickness of the ring element. VEL TECH VEL TECH MULTI TECH 166 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 7. A solid element in the form of a right circular cone is under the action of self weight along its axis. It is proposed to be analyzed by using single axisymmetric ring element of triangular cross section. (a) Determine the shape functions for the element. ww (b) Derive the components of vectors of nodal forces. Assume the height, radius, weight density to be equal to 1. w.E asy 1 U 1 r z 0 0 0 : V 0 0 0 1 1 z 6 so, ....(1) En gin U1 1 : 1 U3 1 = V1 0 0 : 0 V3 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 1 1 1 0 0 0 0 1 0 0 0 0 0 0 1 ....(2) 1 1 1 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 1 0 1 0 1 0 0 0 0 1 0 0 0 0 U 0 V 0 1 ...(3) U 1 r z r z 0 0 0 V VEL TECH eer i ng. net 0 0 0 U ....(4) 1 r z r z V VEL TECH MULTI TECH 167 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH The strain displacement matrix is written as U1 = U3 = 0 [B] matrix is reduced to 4 4 written as 1 0 0 1 0 1 1 0 1 1 0 1 0 0 0 0 1 0 U U or B .....(5) V V K 2 B C B r dr dz .......(6) T ww z 0 r 0 1 1 0 Qx 2 NT r dr dz ......(7) pg z 0 r 0 Qy w.E Qx The integration of the elements of [K] and the consistent load vector are to be carried out Qy using Table. asy En Table of Integrals for Axisymmetric Problems Designation Integral I1 I2 rdrdz dr dz A (ri + rj + rk) /3 A k k K 3 ij I3 dr dz/r 3 jk 3 ki where K ij3 ri z j r j zi r r i I4 z dr dz gin log ri / r j j z z z A i j k 3 k k 5jk k ki5 5 ij Value eer i ng. net where I5 z dr dz r kij5 zi z j zi 3r j ri 4 ri r j z j 3ri r j 1 ri z j r j zi log ri / r j 2 r r 2 2 i VEL TECH j VEL TECH MULTI TECH 168 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH kij6 k 6jk kki6 kij6 I6 zi z j 18 ri r j 2 z 2j 11r j2 - 7ri r j 2r j2 z2 dz dr r + 2 zi z j 2.5 ri2 11ri r j 2.5r j2 + zi2 11rj2 7ri r j 2ri 2 r 1 ri z j r j zi log i 3 3 r r r j 3 i A = Area of the triangle ijk ww j Note: If ri = rj in an element some of the integrals become infinity. In such cases special techniques must be used, if ri = rj for an element the logarithm term becomes very large and loss of accuracy may result. In such cases r i = rj = (ri + rj) /2. w.E 8. Explain in detail Axisymmetric solid element under non-axisymmetric loads. Axisymmetric solid element under non-axisymmetric loads asy If the loads are not axisymmetric but are symmetric about a plane containing the axis of symmetry, such as static equivalent wind load shown in figure. There are three displacements which can be expanded in Fourier series as En Pr Prn (r , z )cos n P P n n, z cos n ...(1) gin Pz Pzn r , z cos n ng. The displacements are in the form U un (r , z )cos n V v n n, z cos n eer i net ...(2) W w n r , z cos n When the amplitude in the three displacements functions are assumed to be linear function in rz plane within each element. Un (n1z ) a1n a2n r a3 n z Vn n1z a4 n a5 n r a6 n z ....(3) Wn n1z a7 n a8 n r a9 n z VEL TECH VEL TECH MULTI TECH 169 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH When n = 0 all values are independent of 1 V = 0. The strain displacement relation is given by U rn cos n r W z zn cos n z 1 V U = n cos n r r U V V v r v r n sin n r r r ww v z V 1 W v z n sin n z r v rz U W v rzn cos n z r .....(4) w.E where for a typical harmonic 0 rn 0 1 zn n r n v r n r v z n 0 v rzn 0 1 0 0 0 z r nz r 0 0 n r 1 r n 0 0 0 0 1 0 1 0 0 0 1 n 0 0 0 0 0 nz r z r asy 0 0 0 0 0 0 0 0 n r 0 n 1 En 0 1 0 a 1n : <(5) 0 a 9n nz r 0 gin or Hence n Gn an and an A qn <<.(6) 1 eer i ng. net Hence n Gn A qn Bn qn .....(7) T .....(8) K B C B dv 1 Determination of Fourier Coefficients Assume wind loading can be expressed as N p an cos n ....(9) n 0 VEL TECH VEL TECH MULTI TECH 170 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Multiplying both sides by cos m and integrating 2 2 N 0 0 n=0 p cos m = an cos n cos m dv <(10) Using orthogonality principle 2 cos m cos n d =2 for m=n=0 0 = for m=n 0 =0 for m n (11) We have ww 2 a0 1 p( ) d 2 0 an 1 2 0 .....(12a) w.E p( )cos n d .....(12b) asy For the above numerical integration can be performed. En 9. Consider the Poisson equation 2u 2u u2 f0 (or ) - 2 2 fo in in a square region . y x gin eer i The boundary condition of the problem is u = 0 on . Solve the problem using finite element method. ng. net A problem possesses symmetry of the solution about a line only when there is symmetry of (a) the geometry, (b) the material properties, (c) the source variation, and (d) the boundary conditions about the line. VEL TECH VEL TECH MULTI TECH 171 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Finite element analysis of the Poisson equation in a square region: (a) geometry and computational domain, and boundary conditions of the problem; (b) a coarse finite element mesh of linear triangular elements; (c) a refined mesh of linear triangular elements. Whenever a portion of the domain is modeled to exploit symmetries available in the problem, a portion of the boundary of the computational domain is a line of symmetry. On lines of symmetry, the normal derivative of the solution (i.e., the derivative with respect to the coordinate normal to the line of symmetry) is zero: u u u qn nx ny 0 .....(1) n x y ww The problem at hand has symmetry about the x = 0 and y = 0 axes; it is also symmetric about the diagonal line x = y. We can exploit such symmetries in modeling the problem. Thus, we can use a quadrant for meshes of rectangular elements and an octant for meshes of triangular elements of the domain to analyze the problem. Of course, it is possible to mix triangular and rectangular elements to represent the domain as well as the solution. w.E asy En Solution by linear triangular elements. Owing to the symmetry along the diagonal x = y, we model the triangular domain shown in figure(a). As a first choice, we use a uniform mesh of four linear triangular elements to represent the domain (fig.(b)), and then a refined mesh (fig. (c)) to compare the solutions. In the present case, there is no discretization error involved in the problem because the geometry is exactly represented. gin eer i ng. Elements 1, 3 and 4 are identical in orientation as well as geometry. Element 2 is geometrically identical with element 1, except that it is oriented differently. If we number the local nodes of element 2 to match those of element 1 then all four elements have the same element matrices, and it is necessary to compute them only for element 1. When the element matrices are calculated on a computer, such considerations are not taken into account. In solving the problem by hand, we use the correspondence between a master element (element 1) and the other elements in the mesh to avoid unnecessary calculations. net x1, y1 0,0, x2, y2 a,0, x3, y3 a, b Hence, the parameters i , i , and i are given by a x2 y 3 x3 y 2 ab, 2 x3 y1 x1y 3 0, 3 x1y 2 x2 y1 0 1 y 2 y 3 b, 2 y 3 y1 b, 3 y1 y 2 0 ......(2) 1 x2 x3 0, 2 x3 x1 a, 3 x1 x2 a VEL TECH VEL TECH MULTI TECH 172 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH The element coefficients Kije and fie are given by b2 1 2 K 1 b 2ab 0 0 1 f0ab 2 1 a , f 1 ....(3a) 6 2 a 1 b 2 a2 b2 a 2 The element matrix in (3a) is valid for the Laplace operator -2 on any right-angled triangle with sides a and b in which the right-angle is at node 2, and the diagonal line of the triangle connects node 3 to node 1. Note that the off-diagonal coefficient associated with the nodes on the diagonal line is zero for a right-angled triangle. These observations can be used to write the element matrix associated with the Laplace operator on any right-angled triangle, i.e., for any element-node numbering system. For example, if the right-angled corner is numbered as node 1, and the diagonal-line nodes are numbered as 2 and 3 (following the counter-clockwise numbering scheme), we have (note that a denotes the length of side connecting nodes 1 and 2) a 2 b 2 b 2 a 2 1 K e b 2 b2 0 ....(3b) 2ab 2 2 a 0 a ww w.E asy En For the mesh shown in figure, we have K 1 K 2 K 3 K 4 , gin f f f f 1 2 3 For a = b, the coefficient matrix in (3a) takes the form 1 1 0 1 e K 1 2 1 2 0 1 1 <(4) 4 eer i ng. net The assembled coefficient matrix for the finite element mesh is 6 6, because there are six global nodes, with one unknown per node. The assembled matrix can be obtained directly by using the correspondence between the global nodes and the local nodes, expressed through the connectivity matrix 1 5 B 2 3 VEL TECH 2 3 4 5 3 2 5 6 <..(5) VEL TECH MULTI TECH 173 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH The assembled system of equations is Q11 1 1 0 0 0 0 U1 1 1 1 4 2 1 0 0 U 3 Q + Q2 + Q3 3 1 2 2 1 2 4 1 0 2 4 0 2 0 U3 f0 3 Q3 + Q2 + Q1 (6) 2 0 1 0 2 1 0 U4 24 1 Q23 0 0 2 1 4 1 U 3 Q 2 + Q 3 + Q 4 3 2 5 1 4 1 0 0 0 0 1 1 U6 Q3 The sums of the secondary variables at global nodes 2, 3 and 5 are ww ˆ Q21 + Q32 + Q13 = Q 2 1 2 4 ˆ Q +Q +Q =Q 3 2 w.E 1 3 ˆ Q12 + Q33 + Q24 = Q 5 ....(7) asy At nodes 1, 4 and 6, we have Q11 Qˆ1,Q23 Qˆ 4 ,and Q34 Qˆ 6. En gin The specified boundary conditions on the primary degrees of freedom of the problem are U4 = U5 = U6 = 0 <(8) eer i The specified secondary degrees of freedom are (all due to symmetry) ˆ 0, Q ˆ 0 <.(9) Qˆ1 0, Q 2 3 ng. net ˆ , are Since U4, U5, and U6 are known, the secondary variables at these nodes, i.e., Qˆ 4 ,Qˆ 5 , and Q 6 unknown and can be obtained in the post-computation. Since the only unknown primary variables are (U1, U2 and U3), and (U4, U5, and U6) are specified to be zero, the condensed equations for the primary unknowns can be obtained by deleting rows and columns 4, 5 and 6 from the system. In retrospect, it would have been sufficient to assemble the element coefficients associated with the global nodes 1, 2 and 3 i.e., writing out equations 1, 2 and 3: 1 1 1 K11 U1 K12 K13 1 1 2 3 1 2 K 23 K32 U2 f21 + K 21 K 22 K33 K11 1 1 2 1 2 4 1 K31 K32 K 23 K 33 K 22 K11 U3 f3 + 0 f + f 0 f + f 0 f11 2 3 2 2 3 1 4 1 <<(10) VEL TECH VEL TECH MULTI TECH 174 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH The unknown secondary variables Qˆ 4 ,Qˆ5 , and Q̂6 can be computed either from the equations (i.e. from equilibrium) 3 Qˆ 4 0 U1 f23 K 21 0 2 3 4 2 3 2 4 ˆ Q f f f 0 K K K K 5 1 3 2 13 31 12 21 U2 <.(11) 4 ˆ 0 f34 0 K 31 U3 Q6 or from their definitions. For example, we have Qˆ 4 Q23 where q qn3 23dx ww 12 3 n 1 2 q 3 n 23 2 3 qn3 23dy qn3 23ds...(12a) 3 1 w.E u u u nx ny 0 nx 0, 0 y 12 y x asy u u u nx ny nx 1, ny 0 y 23 x x y 23 1 , 23 0 2 3 13 h23 En Thus, h 23 u y 3 Qˆ 4 Q22 1 dy 0 x h23 gin where u/x from the finite element interpolation is 3 j u u 3j x j 1 2A3 3 eer i ng. net We obtain h a, a,2A a , U U 0 23 3 1 2 3 4 5 3 ˆ h23 Q u 3j j3 0.5U2 4 4 A3 j 1 ...(12b) Using the numerical values of the coefficients Kije and fie (with f0 =1), we write the condensed equations for U1, U2 and U3 as VEL TECH VEL TECH MULTI TECH 175 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH 0 U1 0.5 0.5 1 0.5 2.0 1.0 U 1 3 2 24 0 1.0 2.0 U3 3 VEL TECH HIGH TECH ....(13) Solving for Ui (I = 1, 2, 3), we obtain 1 0.5 1 U1 3 7.5 0.31250 1 1 1 1 0.5 3 U2 5.5 0.22917 ....(14) U 24 0.5 0.5 0.75 3 24 4.25 0.17708 3 ww and, from (11), we have w.E Q 1 0 0.5 0 U1 -0.197917 1 3 4 0 1 U2 = -0.302083 <(15) Q32 Q22 3 0 24 Q4 1 0 0 0 U3 32 -0.041667 3 22 asy 3 3 By interpolation, Q22 , for example, is equal to -0.5u2, and it differs from Q22 computed from 3 2 equilibrium by the amount f 1 = 24 . En gin 10. Explain in detail. The steps involved in finite element solution of the axisymmetric problem. eer i ng. The finite element solution of the problem is given by the following steps. net Step 1: Replace the solid body of revolution by an assembly of triangular ring elements as shown in figure. Step 2: We use a natural coordinate system and assume linear variation of temperature inside an element e so that the temperature T(e) can be expressed as e e T N q VEL TECH ....(1) VEL TECH MULTI TECH 176 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH ww w.E asy En gin Idealization of an axisymmetric body with triangular ring elements where [N] = [Ni Nj Nk] [L1 L2 L3] <<.(2) ng. and Ti e q T j Tk eer i ......(3) net The natural coordinates L1, L2 and L3 are related to the global cylindrical coordinates (r, z) of nodes I, j and k as 1 1 r = ri z z i 1 rj zj or, equivalently, L1 a1 1 a L2 e 2 2 A L a3 3 VEL TECH 1 L1 rk L2 zk L3 .....(4) b1 c1 1 b2 c2 r ......(5) b3 c3 z VEL TECH MULTI TECH 177 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH where a1 r j zk rk z j a2 rk zi ri zk a3 ri z j r j zi b1 z j zk b2 zk zi ....(6) b3 zi z j ww c1 rk r j c2 ri rk w.E c3 r j ri asy and A is the area of triangle ijk given by (e) A e En 1 ri z j zk r j zk zi rk zi z j ....(7) 2 gin Step 3: The element matrices and vectors can be derived using Eqs. as follows. rk D 0r 0 rkz ......(8) eer i ng. and net Ni N j Nk r r 1 b b b B N Nr N e c1 c2 c3 .......(9) i 2A 1 2 3 j k z z z (e) and by writing dV as 2r. dA where dA is the differential area of the triangle ijk. Eq. gives r B D B dA K1e 2 = VEL TECH e A b12 b1b2 b1b3 2 k r 4 A e 2 + T 2 K z 4 A e 2 b1b2 b22 b2b3 c12 c1c2 c1c3 b1b3 b2b3 r 2dA e b32 A c1c2 c22 c 2c3 c1c3 c2c3 r 2dA e c32 A .....(10) VEL TECH MULTI TECH 178 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net ww w.E asy E ngi nee rin g.n et **Note: Other Websites/Blogs Owners Please do not Copy (or) Republish this Materials, Students & Graduates if You Find the Same Materials with EasyEngineering.net Watermarks or Logo, Kindly report us to easyengineeringnet@gmail.com Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH The radial distance r can be expressed in terms of the natural coordinates L1, L2 and L3 as r = riL1 + rjL2 + rkL3 <<(11) Thus the integral term in equation can be expressed as L21 R 2 r 2dA ri r j rk L1L2 e e A A L1L3 L1L2 L22 L2L3 L1L3 ri L2L3 r j dA .....(12) L23 rk ww By using the integration formula for natural coordinates, can be written as 2 1 1 ri 1 R r dA ri r j rk 1 2 1 r j ......(13) 12 e A 1 1 2 r k 2 w.E 2 asy and hence b12 k R K1e r bb 2Ae 1 2 b1b3 b1b2 b22 b2 b3 2 c12 b1b3 2 k R b2 b3 z e c1c2 2A b32 c1c3 En c1c2 c22 c2c3 gin c1c3 c2c3 ...(14) c32 eer i ng. For isotropic materials with kr = kz = k, becomes b12 c12 2 K1e kR b1b2 c1c2 2A e b b c c 1 3 1 3 b1b2 c1c2 b1b3 c1c3 2 2 ...(15) b c b b c c 2 2 2 3 2 3 b2 b3 c2c3 b32 c32 net To evaluate the surface integral of equation, we assume that the edge ij lies on the surface S 3 from which heat convection takes place. Along this edge, L3 = 0 and dS3 = 2r ds so that equation gives 2 L1 s j rL1 K 2e 2 h s=s L2 L1 L2 0 r ds = 2 h s=s rL1L2 i i 0 0 sj rL1L2 rL22 0 0 0 ds ...(16) 0 By substituting Equation for r and by using the relation sj p! q ! L L ds s p q 1! p q 1 2 ji ....(17) s sI VEL TECH VEL TECH MULTI TECH 179 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH where sji = sj – si = length of the edge ij, 3ri r j hs ji K 2 e ri r j 6 0 r r 0 r 3r 0 ......(18) i j i j 0 0 To evaluate the volume integral for P1e as e T P1 rq N dV <<(19) v e ww We use the approximation rq rc q = constant where rc = (ri + rj + rk)/3 and the relation dV = 2r.dA to obtain w.E rL1 e P1 2 rc q rL2 dA <..(20) e A rL3 asy En With the help of equation can be evaluated to obtain e rcqA e P1 6 2ri r j rk ri 2r j rk ri r j 2rk gin ...(21) eer i ng. e The surface integral involved in the definition of P3 can be evaluated as in the case of equation (18). Thus if the edge ij lies on the surface S3, 2ri r j s j rL1 e hT S T ji P3 hT N dS3 2 hT rL2 ds ri 2r j (22) 3 e s si S3 0 0 e net Similarly expressions for P2 can be obtained as 2ri r j e qS T if the edge ji P2 q N dS2 ri 2r j i j lies on S2 .....(23) 3 e S2 0 Step 4: Once the element matrices and vectors are available, the overall or system equations can be derived as K T P <.(24) VEL TECH VEL TECH MULTI TECH 180 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH where E K K1e K2e <<(25) e 1 and E e e e P P1 P2 P3 e 1 <<..(26) Step 5: The solution of the problem can be obtained by solving Equation (24) after the incorporation of the known boundary conditions. 11. Derive the element matrices and vectors for the elements shown in figure. ww Solution: w.E From the data shown ............... required element properties can be computed as b1 = zj -......... b2 = zk – zi b3 = zi - zj c1 = rk - rj c2 = ri - rk c3 = rj - ri asy En gin A(e) = 1 [4(2 -6) + 7(6-2) + 7(2-2)] = 6 2 2 R eer i ng. 2 1 1 4 1 438 (4 7 7 ) 1 2 1 7 36.5 12 7 12 1 1 2 rc = (ri + rj + rk)/3 = (4 + 7 + 7 )/3 = 6 Skj = [(rk – rj)2 + (zk – zj)2]1/2 = [(7 – 7)2 + (6 – 2)2]1/2 = 4 net Sji = [(rj – ri)2 + (zj – zj)2]1/2 = [(7 – 4)2 + (2 – 2)2]1/2 = 3 Figure: 16.2 VEL TECH VEL TECH MULTI TECH 181 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH [KI(e) ] Can be obtained as 0 9175 9175 [KI(e) 9175 14330 5160 from equation 0 5160 5160 .......... (1) Since convection occurs along the two edges i j and jk, the [KI(e) ] matrix can be written as 0 0 0 0 (h) jk Skj (ri 3rj ) 0 0 (3rj rk ) (rj rk ) 6 0 0 0 (rj rk ) (rj 3rk ) (3ri rj ) [K ] (e) 2 = (ri rj ) (h)ij s ji (ri rj ) 0 6 ww 0 0 (12 7) (4 7) 0 0 (10)(4) 0 (21 7) (7 7) (4 7) (4 21) 0 6 0 0 (7 7) (7 21) 0 0 (15)(3) 6 w.E 0 447.7 259.2 = 259.2 1176.0 293.2 0 293.2 586.4 Equation gives asy En 8 7 7 20734.5 (e) (60)(50)(6) PI 4 14 7 23561.9 6 4 7 14 23561.9 (e) As no boundary heat flux is specified, PI we obtain gin ng. 0 . From equation and a similar equation for the edge jk, (2r r ) 0 (e) (hT )ij S ji i j (hT) jk skj P3 (ri 2rj ) (2rj rk ) 3 3 (ri 2rk ) 0 = eer i net (8 7) 0 (10)(40)(4) (4 14) (14 7) 3 0 (7 14) (15)(40)(3) 3 28274.3 = 69115.0 35185.8 VEL TECH VEL TECH MULTI TECH 182 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Thus, 0 9622.7 8915.8 [K ] [k ] [K ] 8915.8 15506.0 4866.8 0 4866.8 5746.4 (e) (e) I (e) 2 and 49008.8 (e) (e) (e) (e) P PI P2 P3 92676.9 58747.7 12. Derive the three dimensional heat transfer equation for axisymmetric problems. ww The governing differential equation for the steady state heat conduction in a solid body is given by equation with the right hand side term zero and the boundary conditions by equation. The finite element solution of these equations can be obtained by using the following procedure. w.E asy Step 1: Divide the solid body into E tetrahedron elements. En Step 2: We use a natural coordinate system and assume linear variation of temperature inside an element e so that the temperature T(e) can be expressed as (e ) = [N] q T Where [N] = [Ni Nj Nk Nl] = [L1 L2 L3 L4] (e) gin ............(1) ......... (2) And Ti (e) Tj q Tk T l ........... (3) eer i ng. net The natural coordinates L1, L2, L3, and L4 are related to the global Cartesian coordinates of the nodes i, j, l, and l by equation. Step 3: The element matrices and vectors can be derived using equations as follows: VEL TECH VEL TECH MULTI TECH 183 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH k x [D] 0 0 Ni x Ni y N i z VEL TECH MULTI TECH 0 0 k z 0 ky 0 Nj .......... (4) Nk z Nl x b1 b2 Nl 1 c1 c 2 y 6V (e) d1 d2 Nl z ww b1b2 b22 b 2b3 b 2b 4 Nk x Nk y x Nj y Nj z VEL TECH HIGH TECH b12 k x b1b2 [KI(e) ] [B]T [D][B]dv 36V (e) b1b3 V( e ) b1b4 c12 k y c1c 2 + 36V (e) c1c 3 c1c 4 w.E c1c 2 c 22 c 2c 3 c 2c 4 c1c 2 c 2c 3 c 32 c 3c 4 b3 c3 d3 b4 c 4 d4 b1b3 b 2 b3 b32 b3b 4 b1b4 b 2b 4 b3 b 4 b24 d12 c1c 2 c 2c 4 k z d1d2 c 3 c 4 36V (e) d1d3 c 24 d1d4 asy ......(5) d1d2 d22 d2 d3 d2 d4 d1d3 d2 d3 d32 d3 d4 En d1d4 d2 d4 d3 d4 d24 gin For an isotropic material with kx = ky = kz = k, equation becomes eer i (b c d ) (b1b2 c1c 2 d1d2 ) (b1b3 c1c 3 d1d3 ) (b1b 4 c1c 4 d1d4 ) (b22 c 22 d22 ) (b2b3 c 2c 3 d2 d3 ) (b2b 4 c 2c 4 d2 d4 ) k [KI(e) ] 36V (e) (b32 c 32 d32 ) (b3b 4 c 3c 4 d3 d4 ) (b24 c 24 d24 symmetric 2 1 2 1 2 1 ng. The matrix [K(e) 2 ] is given by Ni2 NN NN i j i k 2 Nj NjNk [K (e) 2 ] h Nk2 s(3e ) symmetric NN i l NjNl Nk Nl Nl2 net ......... (6) If the face ijk of the element experiences convection, Nl = 0 along this face and hence equation gives 2 hA ijk 1 [K (e) 2 12 1 0 VEL TECH 1 2 1 0 1 1 2 0 0 0 0 0 ............. (7) VEL TECH MULTI TECH 184 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Where Aijk is the surface area of the face ijk. There are three other forms of equation, one for each of the other faces jkl, kli, and lij. In each case the value of the diagonal terms will be two and the values of the nonzero off-diagonal terms will be one. The coefficients in the row and the column associated with the node not lying on the surface will be zero. Ni 1 N (e) q0 V (e) j PI q dV 1 4 Nk V( e ) 1 N l ......... (8) ww If the face ijk lies on the surface S2 on which heat flux is pacified, Ni L1 1 N L (e) qA ijk 1 j P2 q dS2 q 2 dS2 L 3 1 N S(2e ) k s(2e ) 3 N 0 0 l w.E asy .......... (9) En And similarly, if convection loss occurs form the face ijk, gin Ni L1 1 N L (e) hTAijk 1 j P3 hT dS3 hT 2 dS3 ....... (10) L 3 Nk 1 S(3e ) S(3e ) 3 N 0 0 l eer i ng. There are three other forms of equation (9) and (10). In these equations, the zero coefficients will be located in the row corresponding to the node not lying on the face. 13. Derive the element equations for the elements shown if figure. net Solution: From the given data, the required element properties can be computed as follows: VEL TECH VEL TECH MULTI TECH 185 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH 1 0 11 4 V (e) 6 1 1 1 2 VEL TECH MULTI TECH 1 3 0 2 VEL TECH HIGH TECH 2 1 5 , 0 6 4 1 3 1 4 1 1 b1 1 0 1 10, c1 1 1 0 17, 2 2 1 2 1 4 4 3 1 d1 1 0 1 1 2 2 1 ww w.E 1 0 0 b2 1 2 4 0, 1 1 2 1 1 0 c 2 2 1 4 2, 0 1 2 asy 1 0 1 d2 2 2 1 1 0 1 1 1 2 4 b3 1 1 2 5, 1 3 1 En 2 1 4 c 3 0 1 2 10, 4 1 1 gin ng. 2 2 1 d3 0 1 1 0 4 3 1 1 1 2 b4 1 3 1 5, 1 0 0 eer i net 0 1 2 c 4 4 1 1 11 1 1 0 0 1 1 d4 4 3 1 1 0 1 To compute the area Ajkl. We use the formula Ajkl = [s(s -) (s - ) (s -)] 1/2 VEL TECH VEL TECH MULTI TECH 186 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Where, , are the lengths of the sides of the triangle: = length jk = [(xk – xj)2 + (yk - yj)2 + (zk –zj)2]1/2 = (25+9+1)1/2 = 5.916 = length Kl = [(xl– xk)2 + (yl - yk)2 + (zl –zk)2]1/2 = (9+4+16)(1/2) = 5.385 = length lj = [(xj – xl)2 + (yj - yl)2 + (zj –zl)2]1/2 ww = (4+1+9)(1/2) = 3.742 s w.E 1 1 ( ) (5.916 5.385 3.742) 7.522 2 2 asy A jkl [7.522(7.522 5.916)(7.522 5.385)(7.522 3.742)] 1/ 2 En = 9.877 Equation gives gin eer i (100 289 1) (0 34 1) ( 50 170 0) (50 187 2) (0 4 1) (0 20 0) (0 22 2) 60 6 [KI(e) ] (25 100 0) ( 25 110 0) 36 5 (25 121 4) symmetric -70 -440 478 780 10 40 -48 = 250 -270 300 symmetric ng. net The matrix [K(e) 2 ] will be a modification of equation: 0 h.A jkl 0 [K (e) 2 ] 0 12 0 0 2 1 1 0 1 2 1 0 0 1 (10)(9.877) 0 0 1 12 2 0 0 2 1 1 0 1 2 1 0 1 1 2 0 0 0 0 0 16.462 8.231 8.231 = 0 8.231 16.462 8.231 0 8.231 8.231 16.462 VEL TECH VEL TECH MULTI TECH 187 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 1 10.42 (e) 20 5 PI 1 10.42 64 1 10.42 0 (e) P2 0 since no boundary heat fluxis specified 0 0 0 (e) 10 40 9.877 1 1316.92 P3 3 1 1316.92 1 1316.92 ww w.E [K (e) ] [K1(e) ] [K (e) 2 ] 10.42 (e) (e) (e) (e) 1327.34 P P1 P2 P3 1327.34 1327.34 The element equation are (E1 ) asy En gin (e) (e) [K ]q P eer i ng. (e) (e) where [K (e) ] and P are given by equation (E1 ) and (E 2 ) result and Ti (e) Tj q Tk T l net 14. Solve the Poisson equations by linear triangular elements. 2 2u u2 fo (or) - u2 2 fo in y x in a square region the boundary condition of the problem is u = 0 on . Solution by linear rectangular elements. Note that we cannot exploit the symmetry along the diagonal x = y to our advantage when we use a mesh of rectangular elements. Therefore, we use 2 VEL TECH VEL TECH MULTI TECH 188 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 2 uniform mesh of four linear rectangular elements to discretize a quadrant of the domain. Once again, no discretization error is introduced in the present case. Since all elements in the mesh are identical, we shall compute the element matrices for only one element, say element1. We have 1 1 2x 1 2y , 2 2x 1 2y , 3 4 xy , 4 1 2 x 2y K ije 0.5 0 fie 0.5 0.5 i j x x 0 i j dx dy .......(1) y y 0.5 f dx dy ww 0 0 o i w.E asy En gin eer i ng. net Finite element discretization of the domain of Example 8.3 by linear rectangular elements. VEL TECH VEL TECH MULTI TECH 189 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH UNIT – V PART – A 1. Define: I & O parametric elements. If the shape functions defining the boundary and displacements are the same, the element is called as isoparametric element. O Nodes used for defining geometry. ww Nodes used for defining displacement. w.E asy En ISO parametric element. gin eer i ng. 2. Define: Superparametric element. net The element in which more number of nodes are used to define geometry compared to the number of nodes used to define displacement are known as subparametric element. O Nodes used for defining geometry. Nodes used for defining displacement. Super parametric element. VEL TECH VEL TECH MULTI TECH 190 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 3. Sub parametric element. Sub parametric element in which less number of nodes are used to define geometry compared to the number of nodes used for defining the displacements. Such elements can be used advantageously in case of geometry being simple but stress gradient high. 4. Write the basic theorem of Isoparametric element. Theorem I: ww If two adjacent elements are generated using shape functions, then there is continuity at the common edge. Theorem II: w.E asy It states that if the shape functions used are such that continuity of displacement is represented in the parent co-ordinates, then the continuity requirement will be satisfied in the isoparametric elements. Theorem III: En gin eer i The constant derivative conditions and condition for rigid body are satisfied for all isoparametric elements. 5. Write down the Lagrange Polynomial Formula. ng. net The Lagrange polynomial formula is an interpolation formula that is useful for generating shape functions and is defined as ( x x0 )( x x1 )...( x xk 1 )( x xk 1 )...( x x n ) x xm ( xk x0 )( xk x01 )...( xk xk 1 )( xk xk 1 )...( xk xn ) m 0 x k x m n Lk k m and represents the product of all terms. When x = x k, the product becomes unity. However, when x = xm with m k, the product becomes zero. It follows that Lk of Equation has properties similar to a shape function. VEL TECH VEL TECH MULTI TECH 191 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 6. Write down the Triangular Shape Function Formulae. The shape function can be derived in terms of area coordinates using the formula n F (L , L , L ) Nk m 1 2 3 Fm|L1,L2,L3 m 1 where n is the order of the triangle and is equal to 1 less than the number of nodes along a side. The function Fm is obtained from the equations of n lines that pass through all the nodes except the one of interest. The denominator is the value of Fm when evaluated at the coordinates of node k. ww w.E 7. Use the Lagrangian interpolation formula for deriving one-dimensional three-node shape functions for the element illustrated in Figure. Specialize the shape function for an element of length L with a node at its center. asy Substitute into Eq. (6.3): Li Ni ( x x j )( x xk ) ( xi x j )( xi xk ) Lj N j En ( x xi )( x x j ) ( x xi )( x xk ) Lk Nk ( x j xi )( x j xk ) ( xk xi )( xk x j ) gin eer i Let xi = 0, xj = L/2, and xk = L and let corresponding node numbers be 1, 2, and 3: N1 8. ( x L / 2)( x L) (L / 2)(L) N2 x( x L ) x( x L / 2) N3 (L / 2)(L / 2 L) L(L L / 2) ng. net Derive the interpolation functions for a four-node isoparametric quadrilateral element. Figure : VEL TECH VEL TECH MULTI TECH 192 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Use the Lagrange polynomial formula and Figure. The interpolation function for node 1 is the product of shape functions along the lines = -1 and = -1. Note that the interpolation function is identical to the shape function and that the notation for the shape function is used: 1 1 (1 )(1 ) N1 L1 L1 1 1 1 1 4 The remaining functions are derived in a similar manner. 9. Discuss the evaluation of the jacobian matrix for a four-node isoparametric finite element. ww The interpolation functions are defined in Pro. 6.4, and the jacobian matrix is defined by Equation, (g) of Prob. 6.6. The jacobian matrix can be written as a product of two matrices: w.E (Ni / ) J E [ xi y i ] (Ni / ) Or asy x1 1 (1 ) (1 ) (1 ) (1 ) x2 J 4 (1 ) (1 ) (1 ) (1 ) x3 x4 y1 x2 x3 x4 En gin eer i This definition of J can be extended to any interpolation function. The matrix multiplication would be a [2 x n ] [n x2], where n is the number of nodes to be used for geometry transformation. Equation (a) can be used to evaluate J at any , location for an element defined in the x, y system. In finite element analysis Equation (a) is used in conjunction with the numerical integration and the evaluation of the local stiffness matrix. ng. 10. Discuss the derivation of the [B] matrix for axisymmetric elasticity. net The governing strain – displacement relationships are given by Equations with = / = 0. The matrix of material constants [C] is given by Equation 9d) of . The [B] matrix is derived, as usual as a shape function matrix postmultiplied by an operator matrix. The form of the operator matrix is dictated, in the case of axisymmetric elasticity, by the order of the stresses in the stress matrix or the order of the strains in the strain matrix. In this case, use the same strain matrix given by Equation (a) of problem. Then, {} = [L] [N] {u}, where {u}, in this application, is a matrix of eight unknown displacements corresponding to a four-node quadrilateral element: VEL TECH VEL TECH MULTI TECH 193 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH 0 rr / r 0 N1 0 N2 0 N3 0 N4 0 1/ r u / r 0 N1 0 N2 0 N3 0 N4 zz 0 rz / r / r VEL TECH HIGH TECH ...(a) 0 N2 / r 0 N3 / r 0 N4 / r 0 N1 / r N /r 0 N /r 0 N /r 0 N /r 0 B LN 10 N / z 20 N / z 30 N / z 40 N / z ...(b) 1 2 3 4 N1 / z N1 / r N2 / z N2 / z N3 / z N3 / r N4 / z N4 / r The terms containing partial derivatives are obtained from Equation (b) of problem, and substituted into equation (b) above, with r and z replacing x and y, respectively. The terms containing the shape function divided by r are computed directly for each node (shape function). For instance, let the x coordinate correspond to the radial coordinate r: Equation (b) of problem is used to compute the r in equ. (b) above. The , and of prob. correspond to the coordinates of the integration point in the , system. The shape functions are evaluated by substituting the coordinates of the integration point (Gauss point) into the corresponding shape function equation (see prob. for a four-node quadrilateral). ww w.E asy En gin 11. Discuss higher-order isoparametric element in terms of number of nodes versus the complete polynomial representation that satisfies the Pascal triangle requirement. In Particular, compare the six-node quadratic triangular element, the eight-node serendipity element, and the nine-node Lagrangian element. eer i ng. All these elements can be classified as quadratic elements because they have three nodes along each side. The triangular element was discussed in Problem, and the corresponding interpolation function contains all possible quadratic terms and none higher than quadratic. net A study of the shape functions for the eight-node serendipity element shows that all six constant, linear, and quadratic terms are represented with the addition of two cubic terms. It follows that an eight-node element must have eigth terms in its corresponding interpolation function, and the cubic terms in this case are 2 and 2. The nine-node Lagrangian element shape functions are given in problem, and a study of these shape functions indicates that there are two additional third-order terms and one fourth-order term that can be shown to be 2 , 2, and 22. It can be concluded that quadrilateral elements do not satisfy the Pascal triangle requirement that the shape function be represented by a complete polynomial. However, the idea is to satisfy the Pascal triangle (completeness) requirement in the best possible way. In higherorder elements the excessive interior nodes in the Lagrange family of elements can cause difficulty with convergence and should be avoided. For additional study see Burnett (1987) or Zienkiewicz and Taylor (1989). VEL TECH VEL TECH MULTI TECH 194 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 12. Use area coordinates to determine the shape function N4 for both the quadratic and cubic triangular finite elements of Figure. The quadratic element is shown in figure. The order of the triangle is n = 3 -1, and it follows that two lines should be sufficient to pass through all nodes except node 4. These lines are shown in the figure as L1 = 0 passing through nodes 2, 5, and 3 and L2 = 0 passing through nodes 1, 6, and 1 1 3. The location of node 4 is defined using all three area coordinates as L1 , L2 , L3 0. 2 2 L1 0 L2 0 N4 1 4L1L2 1 2 2 where the numerator in each term is an equation of a line and the denominator is the equation of the line evaluated using the coordinates of node 4. ww w.E asy En PART – B gin 1. (a) A four-element model of a plane area is shown in Figure. Use the interpolation function for a four-node quadrilateral derived in problem and for element III shown that coordinate location (x = 7.0, y = 6.0) corresponds to point (1, 1) in the generalized space. Also, for = 0.5 and = - 0.5, determine the corresponding point in the global system. eer i ng. net Figure : VEL TECH VEL TECH MULTI TECH 195 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH The interpolation functions can be written as follows using the notation of problem. 4 4 x Ni x i i 1 y Ni y i or (a) i 1 1 1 1 1 x= (1 )(1 )x1 (1 )(1 )x2 (1 )(1 )x3 (1 )(1 )x4 4 4 4 4 1 1 1 1 y= (1 )(1 )y1 (1 )(1 )y 2 (1 )(1 )y 3 (1 )(1 )y 4 4 4 4 4 (b) (c) Substitute the global coordinate values for element III: ww 1 1 1 1 x= (1 )(1 )(2) (1 )(1 )(5.5) (1 )(1 )(7) (1 )(1 )(4) 4 4 4 4 1 1 1 1 x= (1 )(1 )(3) (1 )(1 )(3) (1 )(1 )(6) (1 )(1 )(6) 4 4 4 4 w.E (d) (e) Substitute = 1 and = 1 into Equations (d) and (e) and compute the corresponding and x and y. Note that all terms in Equations (d) are zero except the third term and that it corresponds to node 3. Similarly, all terms in equation (e) are zero except the third that corresponds to node 3. It k can be seen that the interpolation function is similar to the shape function. For = 1 and =1, N3 = 1 with N1 = N2 = N3 = 0. asy En gin eer i Substituting = 0.5 and = -0.5 into Eqs. (d) and (e) gives x = 5.0313 and y = 3.75. It is easily verified that these solutions represent a linear approximation for coordinate locations along the element boundaries and on the interior of the element. ng. net 1. (b) An isoparametric parent element is shown in Figure (a), and a corresponding isoparametric distorted element is shown in Figure (b). Discuss the transformation that relates partial derivatives in the original x, y coordinates to the generalized , coordinates. Figure : (a) Parent element VEL TECH VEL TECH MULTI TECH 196 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Figure: (b) Distorted element ww The shape functions, as illustrated in problem, are functions of and . or w.E Ni = Ni (, ) ( the NI are shape functions) (a) asy The x, y coordinates are defined in terms of , coordinates by Eq. (a) of problem, or x = x (Ni) = x (, ) y = y (Ni) = y (, ) En (the Ni are the interpolation functions) (b) gin The derivatives of the shape functions can be written as follows, using the chain rule: Ni Ni x Ni y x x Ni Ni x Ni x x x eer i ng. (c) Equation (c) is written in matrix form as Ni / x / Ni / x / y / Ni / x y / Ni / y net (d) The first matrix on the right-hand side is defined as J and is referred to as the Jacobian matrix: x J x VEL TECH y y (e) VEL TECH MULTI TECH 197 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Multiplying Equ. 9d) by J-1 gives the form of the equation that can be used to compute derivatives of the shape functions in the x, y system: Ni x Ni x = = J-1 Ni x Ni y y y 1 Ni Ni (f) Substituting Equations (a) of Problem into (f) gives the final form ww Ni Ni x x i = Ni Ni y xi N i y i Ni y i 1 w.E Ni Ni (g) asy Note that the Ni within the Jacobian matrix are coordinate interpolation functions, whereas the Ni within the column matrices are shape functions. The formulation of Eq. (g) has been related to a four-node quadrilateral finite element, and that dictates that both the interpolation function and the shape function be linear and that the result be an isoparametric element. A subparametric finite element is formulated in exactly the same manner except that the Ni within the jacobian matrix can be linear (a four-node coordinate approximation) and the Ni in the column matrices can correspond to any higher node shape function approximation (such as the nine-node element of Problem). In addition, for the purpose of performing area integration, an infinitesimal element is defined as En dx dy = |det J| d d. gin eer i ng. (h) net where |det J| is the determinant of the jacobian matrix and is often merely referred to as the jacobian. 2. A quadrilateral element is shown in Figure. Evaluate the stiffness matrix for heat transfer using the definition given by Eq. of problem. Assume the thermal conductivity as k x=ky = k=1 Btu/(hr. in.F). Assume a unit thickness for the element. VEL TECH VEL TECH MULTI TECH 198 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH ww w.E asy En gin eer i A 2 x 2 gaussian quadrature will be used, and the parent element is shown in Figure. The numerical integration procedure is similar to that illustrated in Eq. (a) of Problem. The [B] matrix, in the form given by Eq. (b) of problem, must be evaluated for each Gauss point of Figure. Each time the [B] matrix is computed, the contribution to the stiffness matrix is computed as [B] T [k] [B], and the final stiffness matrix is the sum of the four contributions. In what follows, the computations will be shown in detail, and the reader should keep in mind that they are being done using a computer code with a nested DO loop where I = 1 TO 2 as j = 1 TO 2, where I and J correspond to the 2 x 2 gaussian integration. In order to evaluate [B] at each Gauss point, the jacobian of the transformation for that Gauss point must be computed. Equation (a) of Prob. is used to compute the jacobian in matrix format. The matrix that defines the element in the x, y coordinate system corresponds to the second matrix of that equation and is evaluated using Figure: ng. 2 5 4 1 VEL TECH 1 2 6 4 net (a) VEL TECH MULTI TECH 199 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Gauss point 1: The location of the Gauss point is obtained from Table and is shown in Figure. Let i = 1 = 0.57735 and j = i = -0.57735. Use Eq. (a) of Prob. to evaluate the jacobian matrix: 1 (1 )x1 (1 )x2 (1 )x3 (1 )x4 4 1 J11 (1 0.57735)2 (1 0.57735)(5) (1 0.57735)(4) (1 0.57735)(1) 4 1 Similarly, 3.15470 7.88675 1.69060 0.42265 = 1.50 4 J11 ww w.E 1 J12 (1 0.57735)(1) (1 0.57735)(2) (1 0.57735)(6) (1 0.57735)(4) 4 = J21 = J22 = 1 1.57735 3.15470 2.53590 1.69060 = 0.60566 4 asy 1 (1 0.57735)(2) (1 0.57735)(5) (1 0.57735)(4) (1 0.57735)(1) 4 En 1 3.15470 2.11325 1.69060 1.57735 = -0.5 4 gin 1 (1 0.57735)(1) (1 0.57735)(2) (1 0.57735)(6) (1 0.57735)(4) 4 1 1.57735 0.84530 2.53590 6.3094 = 1.60566 4 eer i 1.5 0.60566 J= -0.5 1.60566 The inverse of J can be computed in an elementary way as. J22 j12 J J11 J 1 = 21 | det EJ | net (b) |det J| = J11J22 J21J21 where ng. (c) Substituting into Eqs. (c) and (b) gives |det J| = 2.71133 and 0.59221 -0.22338 J-1 = 0.18441 0.55324 (d) Let dA=dx dy = |det J| d d = |det J| wi = 1 wj = 1 = 2.71133, and the weights have been included in this calculation. Note that the weight functions for 2 x 2 integration are equal to 1.0. VEL TECH VEL TECH MULTI TECH 200 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH The [B] matrix of Eq. (b) of Problem is evaluated using the derivatives of the shape functions in the , system. These numbers are the same as those used above to compute J. 0.59221 0.22338 -1.57735 1.57735 0.42265 -0.42265 0.55324 -1.57735 -0.42265 0.42265 1.57735 -0.14544 0.25713 0.03897 -0.15066 B 1 -0.29088 0.01426 0.07794 0.19868 B 1 0.18441 (e) The contribution to the stiffness matrix for Gauss point 1 can be written as [K ]1 [B ]T1 [k ][B ]1dA ww where dA is defined above. 0.28676 -0.11265 -0.07684 -0.09728 0.17982 0.03018 -0.09735 K 1 Symmetric 0.02059 0.02606 0.16857 w.E asy (f) The computation for the remaining Gauss points will be shown but with less detail. Gauss point 2: i 1 0.57735 1.5 0.89434 J 0.5 1.60566 and En j=2 0.57735 gin 0.56227 0.31318 J-1 0.17509 0.52527 0.06409 0.09250 0.18863 [ B ]2 0.22564 0.03700 0.12455 0.34522 0 .13809 eer i (g) ng. net 0.04077 0.04573 0.15216 0.15711 0.02834 0.03667 -0.10578 K 2 Symmetric 0.14591 -0.13685 0.39479 VEL TECH VEL TECH MULTI TECH 201 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Gauss point 3: i 2 0.57735 j=1 0.57735 and 0.60566 0.19262 J-1 0.15902 0.47705 1.5 0.60566 J 0.5 1.89434 det J 3.14434 0.30353 0.01230 0.12541 0.20492 -0.21722 [B ]3 0.11311 0.08401 0.03360 (h) ww 0.18859 K 3 -0.16954 0.006448 0.45428 0.35855 0.09293 - 0.09607 Symmetric 0.13251 0.02490 0.02574 Gauss point 4: w.E i 2 0.57735 and 1.5 0.89434 J 0.5 1.89434 asy j=2 0.57735 En gin 0.57602 0.27945 J-1 0.15204 0.45611 det J 3.28868 -0.03213 0.16810 0.11991 0.255 88 [B ] 4 0.06426 0.16380 0.23982 -0.01176 0.01685 0.01697 0.18116 K 4 Symmetric (i) 0.06334 0.02952 - 0.06289 -0.13512 0.23642 -0.11018 0.21577 eer i ng. net The final stiffness matrix is the sum of Eqs. (f) – (i): 0.64945 K 2 VEL TECH -0.22456 0.25040 0.17449 0.74787 0.08897 -0.43433 Symmetric 0.53543 -0.19606 0.80488 (j) VEL TECH MULTI TECH 202 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 3. Assume the eight-node quadrilateral isoparametric element shown in Figure is loaded with a uniform pressure loading, px = 1.0, acting in the x direction along the side defined by nodes 1, 8, and 4. Compute the distribution of the pressure loading to each node. Use a three-point gaussian quadrature. ww w.E asy En gin The formulation is similar to Prob. The surface loading is represented as [N] {T } dS T (a) s eer i ng. where {T} is the surface traction matrix and [N] is the shape function matrix and is similar to Eq. (b) or Prob. except that there are 16 rows rather than 8. The uniform pressure loading along side 1-8-4 corresponds to the isoparametric coordinate = 1, and all shape functions will compute as zero except at nodes 1, 8, and 4. Formally, Eq. (a) can be written as T N1 0 0 0 0 0 N4 0 0 0 0 0 0 0 N8 0 px 1.0 1 0 N1 0 0 0 0 0 N4 0 0 0 0 0 0 0 N8 px 0 d 1 net (b) 1 The term N1px d is evaluated numerically with = -1 and and the weight functions taken 1 from Table. The shape functions are given in Problem. N1 (with = -1) = (1 + 1) (1 - )(1 - - 1) 2 4 2 N p d {[(0.774597) 0.774597)](0.555555) (0 0)(0.888888) 2 1 x + [(0.774597)2 0.774597)](0.555555)} px / 2 0.333333 px VEL TECH VEL TECH MULTI TECH 203 (c) VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Similarly, the integral involving N4 becomes 0.333333px, and the integral representing the contribution to the center node 8 becomes 0.6666667 px. The total length of the side of the isoparametric element is 2, and the results are interpreted to mean that each corner node is assigned one-sixth of the uniform load and the center node is assigned two-thirds of the uniform load. Note that the result given by Eq. (c) could have been obtained by direct integration. While the preceding computations illustrate distributing the force to each node, the analyst should take advantage of the isoparametric formulation to compute the node loading. Consider the plane element of Figure (b). In the isoparametric formulation the term dS should be expressed in terms of the is , system using an interpolation function. Assume a unit thickness (t = 1) or carry t through the derivation and write dS = t dL, where dL defines a curve corresponding to the boundary of the plane area. In the Cartesian coordinate system, ww w.E dL = [(dx)2 + (dy)2]1/2 (d) asy Refer to Eq. (b) of Prob. and relate dx and dy to the , system: x x dx d d dy= En x x d d (e) gin The boundary of the element will always correspond to = 1 or = 1. Let = -1, corresponding to the element of Figure, and it follows that / = 0. Use Eq. (a) of Prob. and rewrite Eq. (e): dx Ni x d xi d (i = 1 to 8) (f) dy Ni y d y i d (i = 1 to 8) (g) eer i ng. net where the Ni are interpolation functions corresponding to the eight-node isoparametric element. Assume that three-point integration is to be used along the curve = -1 for the element of Figure (a). For the first integration point let = -1 abd = -0.774597. Equations (f) and (g) are evaluated, then substituted into Eq. (d) to obtain dL. Equation (a) is evaluated with = -1 and = 0.774597. In the axisymmetric formulation dS is 2r dL to simulate the pressure distributed around the entire circumference of the cylinder. The computations are repeated for the remaining Gauss points, and the final result is the sum of the three computations. For illustration, use the element of Figure in x, y coordinates. Let = -1 and = -0.774597, then substitute into Equations (f) and (g) to obtain dx = 0 and dy = 0.5 and by Eq. (d), dL = 0.5. The first contribution to the pressure loading is computed as follows. VEL TECH VEL TECH MULTI TECH 204 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Node1: (N1)(px)(dL)(w1)=0.687298px(0.5)(0.555555)=0.190916px Node 4: (N4)(px)(dL)(w1) = (-0.087298)px (0.5)(0.555555) = - 0.02424px Node 8: (Ng)(px)(dL)(w1)=(0.400)px(0.5)(0.555555)=0.111111px Similarly, let = -1 and = 0.0, dx = 0.0 and dy = 0.5, and dL = 0.5. The second contribution to the pressure loading is (note that w2 = 0.888888) as follows. ww Node 1 : N1 = 0.0 Node 4: N4 = 0.0 Node 8: (1.0)px= (0.5) (0.88888) = 0.44444px w.E asy The third integration point corresponds to = -1 and = +0.774597 with dx = 0.0, dy = 0.5, dL = 0.5, and w3 = 0.55555, and the contribution to the pressure loading is computed as follows. Node 1: -0.024249px Node 4: 0.190916px Node 8: 0.111111px En gin eer i ng. The total loading is the sum of the three contributions and is computed as 0.166667p x at nodes 1 and 4, and as 0.666666px at node 8. net 4. (a) Discuss node placement and area coordinates for linear and higher-order triangular isoparametric elements. Linear, quadratic, and cubic elements are shown in Figure. The node numbering sequence is arbitrary, but very often the corner nodes are numbered 1, 2, and 3, with the intermediate nodes numbered in sequence starting with 4 along the side between nodes 1 and 2. The areas coordinates then correspond to Figure for triangular elements of any order. For instance, area coordinate L 2 emanates from the side opposite node 2. The intermediate nodes are placed equidistant from the end nodes, and for a cubic element the side nodes are placed at the one-third points along the side. VEL TECH VEL TECH MULTI TECH 205 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH ww Figure: w.E asy En gin eer i ng. The derivation for the linear three-node triangular finite element discussed in Prob. began by assuming an interpolation function that was linear in terms of the x, y coordinates, or = C1 + C2x + C3y (a) net It follows that a quadratic triangular finite element must contain all possible linear and quadratic coordinate functions, or = C1 + C2x + C3y + C4x2 + C5xy + C6y2 (b) This complete polynomial representation for interpolation functions corresponds to the Pascal triangle shown in Figure (d). The cubic element should contain all linear, quadratic, and cubic coordinate terms, and it can be seen from the Pascal triangle that there are 10 such terms. The cubic element, Figure (c), must have 10 corresponding nodes. Therefore, the tenth node is located at the centroid of the triangular element. The derivation of shape functions using interpolation functions and x, y coordinates can become a tedious algebraic task. The use of area coordinates and Eq. simplifies the derivation and formulation of a stiffness matrix. VEL TECH VEL TECH MULTI TECH 206 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH (b). Use area coordinates to determine the shape function N4 for both the quadratic and cubic triangular finite elements of Figure. The quadratic element is shown in Figure (a). The order of the triangle is n = 3 -1, and it follows that two lines should be sufficient to pass through all nodes except node 4. These lines are shown in the figure as L1 = 0 passing through nodes 2, 5, and 3 L2 = 0 passing through nodes 1, 6, and 3. The location of node 4 is defined using all three area coordinates as L1 = ½ , L2 = ½ , L3 = 0. There are two terms in Equation: L 0 L 0 N4 1 1 2 1 = 4L1L2 2 2 ww where the numerator in each term is an equation of a line and the denominator is the equation of the line evaluated using the coordinates of node 4. w.E asy En gin eer i ng. Figure: net The order of the cubic element is 3. The lines L1=0, L1 = 31 , and L2 = 0, shown in Figure (b), interest all nodes except node 4. The coordinate location of node 4 is ( (L1 32 , L2 31 , L3 0). Substituting into Eq. gives L 0 L 0 L 1 L 0 9 N4 1 2 2 1 21 13 2 1 = L1L2 (3L1 1) 2 3 3 3 3 2 VEL TECH VEL TECH MULTI TECH 207 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 5. (a). The isoparametric formulation is advantageous for higher-order triangular elements. Like the quadrilateral element, the distorted element in x, y coordinates can be mapped into a parent element, and standard integration routines can be used to evaluate the stiffness integrals. Discuss the coordinate transformation for isoparametric triangular elements using area coordinates. The method used by Zienkiewicz and Taylor (1989) may help to avoid some confusion in the derivation. There are three area coordinates for a two-dimensional element, but the original coordinates is 2-space (x and y). It follows that the parent element should be 2-space ( and for the quadrilateral). Recall from Prob. that there are only two independent area coordinates, or L 1 +L2 + L3 = 1. Identify ww =L1 = L2 then L3 = 1 - - (a) w.E and then Eqs. (c) – (f) and (h) of Prob. are valid. The shape functions are written in terms of L1, L2, and L3, or asy En Ni = Ni (L1, L2, L3) (b) gin Hold , defined in Eq. (a), independent of L1 in Eq. (b) and write the partial derivative Ni Ni Li Ni L2 Ni L3 L1 L2 L3 (c) eer i The right-hand side of Eq. (c) in view of (a), can be evaluated as Li 1 L2 0 L3 1 ng. net and Eq. (c) becomes Ni Ni Ni L1 L3 (d) A similar analysis gives Ni Ni Ni L2 L3 VEL TECH (e) VEL TECH MULTI TECH 208 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH The jacobian can be written the same as Eq. (g) of Problem: Ni Ni xi y i J= (f) Ni Ni xi y i Equation (f) is evaluated using the definitions given by Eqs. (d) and (e). (b). (i). Assume a five-element model as shown in Figure to solve the long cylinder problem described in Problems. The cylinder has an inside radius of 1 in and outside radius of 2 in. Assume an axisymmetric internal pressure loading of 1000 psi. Show the results for the [B] 1 matrix and the local stiffness matrix for element I. Compare results for radial displacement with the exact solution. Assume E = 1.0 psi and = 0.3. (ii) Compare the results for the 5 – element model with those for a 10-element model . ww w.E (a) The five-element model of Figure is constructed of square elements 0.2 x 0.2 in. In the derivation for the axisymmetric finite element the differential area used was 2r dr dz, and that requires that the internal pressure be distributed around the inside circumference of the cylinder. The pressure loading is converted to nodal point loading as (1000 psi) (2) (1 in) (0.2 in) = 1265.64 lb, and one-half of that is applied to nodes 1 and 2. asy En gin eer i ng. net five-element model VEL TECH VEL TECH MULTI TECH 209 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH The [B]I matrix for element I is a 4 x 8 matrix corresponding to Eq. (b) of Prob. and is computed for Gauss point 1 using i = 1 = -0.57735 and j = 1 = -0.57735; 0 3.9434 0 1.0566 0 1.0566 0 3.9434 0.5968 0 0.1599 0 0.0428 0 0.1599 0 [B]1 0 3.9434 0 1.0566 0 1.0566 0 3.9434 3.9434 3.9434 1.0566 3.9434 1.0566 1.0566 3.9434 1.0566 Note that the complete stiffness matrix is made up of four parts as in Prob. There is a separate B matrix corresponding to each integration point, and the stiffness matrix for element I is the 8 x 8 symmetric matrix 3.74129 ww 1.49024 3.84644 -2.64114 -0.34235 4.30205 0.32221 0.66457 -1.83260 4.12838 w.E -1.98513 -1.71177 0.76148 -0.46318 4.30205 asy -1.61107 -1.99370 0.46318 -2.79924 -1.83260 4.12838 En 0.60193 0.20138 -1.98514 1.61107 -2.64114 -0.32221 3.74129 gin -0.20136 -2.51730 1.71177 -1.99370 0.34235 0.66457 -1.49024 3.84644 The exact solution is computed using the results of Prob. Results for displacement are given in Table. Note that all displacements in the z direction are zero for this problem and were entered into the computer solution as computer solution as zero displacement boundary conditions. eer i ng. (b) The 10 – element solution is computed using square elements 0.1 x 0.1, and the nodal point loading is computed as 314.16 lb at nodes 1 and 2. The results are given in Table. The results for displacement tabulated in Table are for illustration purposes, and the large values are a result of assuming the material constant E = 1 psi. net Table : Radial Displacement for a Thick-walled Cylinder (in) r 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 VEL TECH Five elements Nodes Ten elements Nodes Exact 1894.35 1, 2 1642.63 3, 4 1472.46 5, 6 1353.35 7, 8 1268.72 9, 10 1903.52 1763.64 1649.94 1556.37 1478.64 1413.57 1358.79 1312.50 1273.26 1, 2 3, 4 5, 6 7, 8 9, 10 11, 12 13, 14 15, 16 17, 18 1906.67 1766.42 1652.44 1558.60 1480.76 1415.55 1360.67 1314.27 1274.96 VEL TECH MULTI TECH 210 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH 1.9 2.0 1207.18 11, 12 VEL TECH HIGH TECH 1239.98 1211.76 19, 20 21, 22 1241.61 1213.33 6. A triangular element is shown in Figure. Use the isoparametric formulation to compute the terms in the stiffness matrix corresponding to convection as discussed in Probs. Integration results for triangles are given in Figure (b). Illustrate the use of the both linear and quadratic integration formulas. ww w.E asy En gin eer i ng. VEL TECH VEL TECH MULTI TECH 211 net VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH The function to be integrated is Eq. (b) of Prob. Assume a unit value of t: for reference, the function is N1 A N2 N3 ww N1 ux N2 0 N3 w.E 0 N1 / x uy N1 / y N2 / x N2 / y N3 / x dA N3 / y asy (a) Recall that there is an equality between the linear shape functions and area coordinates: x = [N]{x} = [L1 En L2 L3] [x1 x2 x3]T = L1 +3L2 + L3 y = L1 + 3L2 + 5L3 gin eer i There are partial derivatives in Eq. (a), and that will require evaluating the inverse of the jacobian given by Eq. (f) of Problem. Refer to Eqs. (d) and (e) of Prob. it follows that ng. N1 N1 N1 1 0 L1 L3 N2 N2 N2 00 L1 L3 net N3 N3 N3 0 1 L1 L3 N1 N1 N1 00 L2 L3 N2 N2 N2 1 0 L2 L3 N3 N3 N3 0 1 L2 L3 (b) J11 = (1)(1) +(0)(3)+(-1)(1) = 0 VEL TECH VEL TECH MULTI TECH 212 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH J12 = (1)(1) +(0)(3)+(-1)(5) = -4 J21 = (0)(1) +(1)(3)+(-1)(1) = 2 J22 = (0)(1) +(1)(3)+(-1)(5) = -2 and 1 J-1 41 4 0 4 J 2 2 1 4 0 det J 8 2A © The jacobian for the coordinate transformation is complete, and now the right-hand matrix of Eq. 9a) must be evaluated. The computation is analogous to Eq. (g) of Problem. Using Eqs. (b) and (c) above, ww N1 / x N / y 1 N2 / x N2 / y N3 / x 0 1 41 1 1 J 0 1 1 1 N3 / y 4 3 4 1 4 1 2 w.E 0 asy (d) Combining Eq. (d) with Eq. (a) gives the final from of the equation that is to be integrated numerically. ( L1 / 4)(ux uy ) L1ux / 2 (L1 / 4)(3u x uy ) ( L2 / 4)(ux uy ) L2ux / 2 (L2 / 4)(3u x uy ) ( L3 / 4)(ux uy ) L3ux / 2 (L3 / 4)(3u x uy ) En (e) gin eer i ng. Numerical integration for the stiffness matrix is accomplished using the formulas of Figure (b) corresponding to area coordinates. The limits of integration are written in terms of area coordinates, and the weighted integral approximation follows as 1 1L2 0 0 n 1 F dL1 dL2 w i Fi (L1i L2i L3i ) i 1 2 (f) net where wi are the weights and L1i, L2i, L3i are the sampling points. Keep in mind that the function of Eq. (a) is defined in the x, y system, dA = dx dy, and in the new system, dA = |det J| dL1 dL2. Consider the first term of Eq. (e) and a linear order integration formula. For each L i, substitute 31 and note that wi = 1. Of course, only L1 appears in the first term, or 1 ux uy 1 K11 (1) 3 (u x u y )(8) 2 4 3 VEL TECH VEL TECH MULTI TECH 213 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Similarly, 2u 1 11 K 22 (1) ux (8) x 2 23 3 The remaining terms can be computed in the same manner and should be verified by the reader. The correctness of the linear integration can be checked using the exact solution given in Problem. K11 u x b1 uy c1 ) b1 y 2 y1 2 6 ww K11 c11 xl 3 x2 2 u x uy 3 The linear integration gives an exact result for this term. Similarly, K 22 ux b2 uy c2 ) 6 w.E b2 y 3 y1 4 K 22 c 2 x1 x3 0 asy 2ux 3 Again, the exact result is found. En gin The quadratic integration formula will be illustrated for the K 11 term. Referring to Figure (b), L1 takes the values 21 , 21 and 0. and wi = 31 . 1 1 1 1 1 1 1 1 1 K11 (ux uy ) (ux uy ) (0)(ux uy ) 2 4 2 3 4 2 3 4 3 K11 u x uy 3 (8) eer i ng. net Again, the exact result is obtained. The remaining terms are computed in a similar manner and will not be recorded. Note that linear integration was sufficient to give exact results for the linear triangular finite element. However, the reader should study Prob., where shape functions (no derivatives appear in that problem) are integrated, and discover that linear integration is not sufficient to give exact results in that case. 7. ASSEMBLING STIFFNESS MATRIX. Assembling element stiffness matrix is a major part in finite element analysis. Since it involves coordinate transformation from natural local coordinate system to Cartesian global system, isoperimetric elements need special treatment. In this article assembling of element stiffness matrix for 4 noded quadrilateral elements is explained in detail. The procedure can be easily extended to higher order elements by using suitable functions and noting the increased number of nodes. VEL TECH VEL TECH MULTI TECH 214 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Figure shows the typical parent element and isoparametric quadrilateral element. Figure: Typical isoparametric quadrilateral element ww For parent element, the shape functions are w.E Ni (1 i )(1 )i 4 N1 (1 )(1 ) (1 )(1 ) ,N2 4 4 N3 (1 )(1 ) (1 )(1 ) ,and N4 4 4 ......... (1) asy En gin eer i We use the above functions for defining the displacement as well as for defining the geometry of any point within the element in terms of nodal values. ng. When we use shape functions for the geometry, x N1 0 N2 0 N3 0 N4 y 0 N1 0 N2 0 N3 0 x1 y 1 x2 0 y2 N4 x 4 y 4 net ........ (2) The above relation helps to determine the (x, y) coordinates of any point in the element when the corresponding natural coordinates and are given. We are also using the same functions for defining the displacement at any point in the element. VEL TECH VEL TECH MULTI TECH 215 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH u N1 0 N2 0 N3 0 N4 v 0 N1 0 N2 0 N3 0 u1 v 1 u2 0 v 2 N4 u 4 v 4 VEL TECH HIGH TECH ........ (3) In assembling the stiffness matrix we need the derivatives of displacements with respect to global x, y system. It is easy to find derivatives with respect to local coordinates and but it needs suitable assembly to get the derivatives w.r.t. to global Cartesian system. ww w.E The relationship between the coordinates can be computed using chain rule of partial differentiation. asy Thus, x y x y x y x y En i.e., x x Where, x [J ] x y y gin y x x [J ] y y y ........ (5) eer i ........ (4) ng. net The matrix [J] shown above is called Jacobian matrix. It relates derivative of the function in local coordinate system to derivative in global coordinate system. In case of three dimensional it is given by x x [J ] x y y y VEL TECH z z z VEL TECH MULTI TECH 216 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Now going back to isoparametric quadrilateral element, Let J J [J] 11 12 J21 J22 Where J11 x J21 x ww 4 We know, y y J22 J12 w.E x Ni xi N1x1 N2 x2 N3 x3 N4 x 4 i1 J11 asy N3 N2 N4 x N1 x1 x2 x3 x4 Similarly J12, J21, and J22 can be assembled. Then we get 4 Ni xi i 1 J 4 Ni xi i1 4 Ni i Ni xi i 1 4 gin y i 1 En ...... (6) eer i ng. For any specified point the above matrix can be assembled. Now, net x [J ] y x -1 [J ] y VEL TECH VEL TECH MULTI TECH 217 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH J J * 11 * 21 VEL TECH MULTI TECH J J * 12 * 22 VEL TECH HIGH TECH ...... (7) * * * Where J11 ,J12 ,J21 , and J*22 , are the element of Jacobian inverse matrix. Since for a given point Jacobian matrix is known its inverse can be calculated and jacobian matrix is assembled. With this transformation relation known, we can express derivatives of the displacements as shown below: u x * * u J11 J12 * J* y J 21 22 0 v 0 x 0 0 v y ww 0 0 * J11 J*21 u 0 u 0 * J12 v * J22 v w.E ............. (8) asy En The strain displacement relation is given by u x x 1 0 0 0 u y y 0 0 0 1 0 1 1 0 v x xy v y * * J11 J12 0 0 J*21 J*22 VEL TECH 0 J*21 * J11 u u 0 * J22 * v J12 v gin eer i ng. net ................ (9) VEL TECH MULTI TECH 218 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH 4 4 i1 i1 VEL TECH HIGH TECH But u Nu i i and v= Ni Vi u N1 u N1 v 0 v 0 0 0 N1 N1 N2 N2 N3 N3 0 0 N2 N2 0 0 0 N3 N3 0 0 ww u1 0 v 1 u 0 2 v 2 N4 u3 v 3 N4 u4 v 4 N4 N4 0 0 0 ........ (10) Substituting it in equation (9) strain displacement matrix [B] is obtained as, J * 11 B 0 J*21 * 12 J 0 0 J*21 * J*22 J11 w.E N1 N 0 1 * J22 * J12 0 0 0 0 N1 N1 N2 N2 0 0 0 N3 N3 0 asy 0 N2 N2 0 0 0 N3 N3 N4 N4 0 0 N4 N4 ...... (11) En 0 0 gin K B DB dv eer i K t B DB x y ........... (12) Then element stiffness matrix is given by T ng. In this case, T net Where t is the thickness. It can be shown that elemental area in Cartesian coordinates (x, y) can be expresses in terms of the area in local coordinates (, ) as x y = J ............... (13) Where J is the determinate of the Jacobian. K t B DB J T VEL TECH VEL TECH MULTI TECH 219 ........ (14) VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Integration is to be performed so as to cover entire area i.e. the limit of integration is from is from – 1 to 1 and is also form – 1 to 1. It is difficult to carryout all the multiplications in equation (14) and then the integration. It is convenient to go for numerical integration. 8. Assemble Jacobian matrix and strain displacement matrix corresponding to the Gauss point (0.57735, 0.57735) for the element shown in figure. Then indicate how do you proceed to assemble element stiffness matrix. ww w.E asy En Solution : The coordinates of node points in Cartesian system are (0,0), (60, 0), (65.7735, 10) and (5.7735, 10). gin The shape functions are and eer i ng. 1 Ni 1 i 1 i 4 Ni 1 i 1 i 4 Ni 1 i 1 i 4 net N1 1 1 1 1 0.57735 0.10566 4 4 N2 1 1 1 1 0.57735 0.10566 4 4 N3 1 1 1 1 0.57735 0.39438 4 4 N4 1 1 1 1 0.57735 0.39438 4 4 VEL TECH VEL TECH MULTI TECH 220 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Similarly, N1 1 1 1 1 0.57735 0.10566 4 4 N2 1 1 1 1 0.57735 0.39438 4 4 N3 1 1 1 1 0.57735 0.39438 4 4 N4 1 1 1 1 0.57735 0.10566 4 4 ww The Jacobian Matrix is given by w.E Ni xi J N i xi J11 Ni y i asy Ni y i Ni xi En gin eer i 0.10566 0 0.10566 60 0.39438 65.7735 0.39438 5.7735 30.0000 N J12 i y i 0 0 0.39438 10 0.39438 10 0 N J21 i xi 0 0.39438 60 0.39438 65.7735 0.10566 5.7735 2.88698 N J22 I y i 0 0 0.39438 10 0.10566 10 5.0000 n ng. 0 30.0000 J 288698 5.0000 J * J 1 net 5.0000 288698 0.033333 0.019246 1 30.0000 0 0.166667 30.0000 5.0000 0 The strain displacement matrix is given by * * J11 J12 B 0 0 J * J * 21 22 VEL TECH 0 * J21 * J11 N1 N 0 1 * J 22 0 * J12 0 0 N2 0 0 N2 0 N1 0 N2 N1 0 N2 N3 N3 0 0 0 N4 0 N4 N3 N3 0 0 0 0 N4 N4 VEL TECH MULTI TECH 221 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Where Jij* are the elements of Jacobian inverse matrix, 0 0 0.033333 0.019246 B 0 0 0 0.166667 0 0.166667 0.033333 0.019246 0 0.10566 0 0.39438 0 0.39438 0 0.10566 0.10566 0 0.39438 0 0.39438 0 0.10566 0 0 0.10566 0 0.10566 0 0.39438 0 0.39438 0 0.10566 0 0.39438 0 0.39438 0 0.10566 ww 1.48843 10 3 0 0.01761 0 0.01761 0.011112 0 5.55563 10 3 0 0.06573 0 0.06573 0.011112 0.06573 w.E 1.48843 103 0 0.06573 5.55563 10 3 0.022915 0 0 0.01761 0.01761 0.015179 9. Determine the Cartesian coordinate of the point P( = 0.5, = 0.6) shown in Fig. asy En gin Figure. Solution: It is given that eer i ng. net 0.5 and 0.6 N1 (1 )(1 (1 0.5)(1 0.6) 0.05 4 4 N2 (1 )(1 ) (1 0.5)(1 0.6) 0.15 4 4 N3 (1 )(1 ) (1 0.5)(1 0.6) 0.6 4 4 N4 (1 )(1 ) (1 0.5)(1 0.6) 0.2 4 4 VEL TECH VEL TECH MULTI TECH 222 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH x = Ni xi = 0.05 x 2 +0.15 x 8+ 0.6 x 7 + 0.2 x 3 = 6.1 y = Ni yi = 0.05 x 1 + 0.15 x 3+ 0.6 x 7 + 0.2 x 5 = 5.7 9.b. In the element shown in Fig. P is the point (6.5). On this point the load components in x and y directions are 8 kn and 12 kn respectively. Determine its nodal equivalent forces. Solution: We have to first determine the local natural coordinates of point P. We know for the quadrilateral element (1 i )(1 i ) 4 (1 )(1 ) (1 )(1 ) Ni N2 4 4 (1 )(1 ) (1 )(1 ) Ni N4 4 4 x Ni xi gives ww Ni i.e., w.E asy En 1 (1 )(1 )2 (1 )(1 )8 (1 )(1 )7 (1 )(1 )3 4 24 2(1 ) 8(1 ) 7(1 ) 3(1 ) 20 10 0 2 6 4 10 2 or 2 5 n gin y Ni y i gives 1 1(1 ) 3(1 ) 7(1 ) 5 4 (1 ) eer i ng. net 20 16 4 8 or 4 4 8 or 1 2 From equation (2), 1 2 Substituting it in equation I, we get VEL TECH VEL TECH MULTI TECH 223 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH or 1 2 5 2 4 10 (1 ) 9 2 i.e. 2 9 4 0 9 92 4x4 0.42443 2 1 0.42443 0.28779 2 VEL TECH HIGH TECH ww Now, the equivalent load is given by X Y F N T For Point P. w.E asy (1 0.42443)(1 0.28779) 0.10248 4 (1 0.42443)(1 0.28779) N2 0.25362 4 (1 0.42443)(1 0.28779) N3 0.45859 4 (1 0.42443)(1 0.28779 N4 0.18530 4 N1 En gin eer i ng. {F} = [N] {X} T net Fx1 N1 0.10248 0.81984 Fx2 N2 0.25362 2.02896 {8} 8 Fx3 N3 0.45859 3.66872 F N 0.18530 1.48240 x4 4 Fy1 0.10248 1.22976 3.04344 Fy2 0.25362 12 Fy3 0.45859 5.50308 F 0.18530 2.22360 y4 VEL TECH VEL TECH MULTI TECH 224 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 10. The quadrilateral element shown in Fig. is 20 mm thick and is subjected to surface forces T x and Ty. Determine expressions for its equivalent nodal forces. If Tx = 10N/mm2 Ty = 15N/mm2, determine the numerical values of the nodal forces. ww w.E Figure: asy Solution: The element is subjected to load edge 3-4. We know along edge 3-4, = 1 1 N1 (1 )(1 ) 0 4 1 N2 (1 )(1 ) 0 4 1 1 N3 (1 )(1 ) (1 ) 4 2 1 1 N4 (1 )(1 ) (1 ) 4 2 En gin eer i ng. Nodal forces are given by the expression like Fx [N]T {Tx }ds net t [N]T {Tx }dl We know, 2 l (x) (y) 2 2 x y l and 2 In isoparametric concept, we know x Ni xi and y Ni yi VEL TECH VEL TECH MULTI TECH 225 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH In this, case, along line 3-4, x00 1 1 (1 )x 3 (1 )x 4 2 2 In limiting case, dx x 1 (x 3 x 4 ) d y 2 1 1 Similarly, y 0 0 (1 )y 3 (1 )y 4 2 2 In limiting case dy 1 (y 3 y 4 ) d 2 ww w.E 2 asy 2 dl l 1 1 1 (x 3 x 4 ) (y 3 y 4 ) l34 d 2 2 2 1 dt l34 d 2 {Fx } t [N]2 {Tx }dl En gin 0 0 0 l 1 0 l 1 d t 1 (1 ) {Tx } l34d t 34 2 4 1 (1 )Tx 1 2 (1 )Tx 1 2 (1 ) For uniformly distributed load, Tx is constant, eer i ng. net 0 Fx1 1 Fx2 tl34 0 Since (1 ) d 2 1 Fx3 4 2Tx F 2Ty x4 0 tl34 0 Similarly 2 Tx Ty VEL TECH VEL TECH MULTI TECH 226 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Fy1 0 tl34 0 Fy2 T [N] {Ty }ds 2 Tx Fy3 F Ty y4 In this problem, l34 (700 300)2 (700 500)2 447.21mm t 20 mm ww Tx 10 N/ mm2 w.E Fx1 0 Fx2 0 Fx3 44721.36 Fx4 44721.36 Fy1 0 Fy2 0 Fy3 67082.04 F 67082.04 y4 asy En gin eer i ng. 11. a) State and prove isoparametric theorem concept. (a) BASIC THEOREMS OF ISOPARAMETRIC CONCEPT: net Isoparametric concept is developed based on the following three basic theorems: Theorem 1: If two adjacent elements are generated using shape functions, then there is continuity at the common edge. VEL TECH VEL TECH MULTI TECH 227 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH ww It may be observed that in the parent element , for any point on edge AB, shape functions Ni =0 for nodes not on the edge and N1 exists for node on the edge. Hence the final function is the same for the common edge AB in any two adjacent element, when we give the same coordinate values for the nodes on common edge. Hence edge AB is contiguous in the adjacent elements. w.E asy Theorem II: It states, if the shape functions used are such that continuity of displacement is represented in the parent coordinates, then the continuity requirement, will be satisfied in the isoparametric elements also. The proof is same as for theorem 1. En gin eer i Theorem III: the constant derivative conditions and conditions for rigid body are satisfied for all isoparametric elements if, ng. N 1 i proof: Let the displacement function be u 1 2 x y y 4 z <..(1) net Nodal displacement at ‘I’ the node is given by ui 1 2 xi 3 yi 4 zi In finite element analysis we define nodal displacement at any point in the element in terms of nodal displacement as u Nu i i u Ni (1 2 xi 3 yi 4 zi ) 1 Ni 2 Ni xi 3 Ni yi 4 Ni zi VEL TECH VEL TECH MULTI TECH 228 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH From the isoparametric concepts, we know N x x N y y N z z i i i i <(2) i i u= i Ni 2 x 3 y 4 z N 1 i The shape function developed in natural systems satisfy this requirement. Hence they can be safely used for isoparametric representation. This theorem is known as convergence criteria for isoparametric elements. ww w.E 11.b) Explain in detail co-ordinate transformation of isoparametric element. COORDINATE TRANSFORMATION: asy So far we have used the shape function for defining deflection at any point interms of the nodal displacement Tag[1] suggested use of shape function for coordinate transformation form natural local coordinate system to global Cartesian system and successfully achieved in mapping parent element to required shape in global system. Thus the Cartesian coordinate of a point in an element may be expressed as En x n1x1 N2 x 2 .... Nn xn y n1y1 N2 y 2 .... Nn yn gin eer i ng. z n1z1 N2 z 2 .... Nn zn or in matrix form x Nxe net where N are shape functions and (x)e are the coordinates of nodal points of the elements. The shape functions are to be expressed in natural coordinate system. For example consider mapping of a rectangular parent element into quadrilateral element VEL TECH VEL TECH MULTI TECH 229 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH The parent rectangular element shown in fig (a) has nodes 1,2,3 and 4 and their coordinate are (-1,-1),(-1,1),(1,1) and (1,-1). The shape functions of this element are N1 N3 1 1 4 ,N2 1 1 4 1 1 and N1 4 1 1 4 P is a point with coordinate (,). In global system, the coordinates of the nodal points are (x1,y1), (x2, y2), (x3,y3) and (x4,y4) ww To get this mapping we define the coordinate of point P as w.E x N1x1 N2 x 2 N3 x 3 N4 x 4 and y N1y1 N2 y 2 N3 y 3 N4 y 4 asy Noting that shape functions are such that at node I, Ni=1 and all others are zero, it satisfy the coordinate value at all the nodes. Thus any point in the quadrilateral is defined in terms of nodal coordinates. En gin Similarly other parent elements are mapped suitably when we do coordinate transformation. eer i ng. VEL TECH VEL TECH MULTI TECH 230 net VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH B.E./B.TECH. DEGREE EXAMINATION, NOVEMBER/DECEMBER 2007. SEVENTH SEMESTER MECHANICAL ENGINEERING ME 1401 – INTRODUCTION OF FINITE ELEMENT ANALYSIS (COMMON TO AUTOMOBILE ENGINEERING AND MECHATRONICS ENGINEERING) ww (REGULATION 2004) w.E PART – A asy 1. State the principle of minimum potential energy. 2. Define shape functions. En 3. What do you mean by Constitutive Law? 4. Differentiate CST and LST elements. 5. What are the advantages of natural coordinates? gin 6. How thermal loads are input in finite element analysis? eer i ng. 7. Why polynomial type of interpolation functions are preferred over trigonometric functions? 8. Write short notes on Axisymmetric problems. net 9. What do you mean by isoparametric formulation? 10. What are the types of non linearity? PART – B 11. (a) Compute the value of central deflection (Figure 1) by assuming y sin x L . The beam is uniform throughout and carries a central point load P. Figure 1 VEL TECH VEL TECH MULTI TECH 231 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Or (b) (i) Write short note on Galerkin’s method. (ii) Write briefly about Gaussian Elimination. 12. (a) (i) Derive the shape function for a 2D beam element. (ii) Derive the shape functions for 2D truss element. Or (b) ww Why higher order elements are needed? Determine the shape functions of an eight nodded rectangular element. w.E 13. (a) For the constant Strain Triangular element shown in Figure 2, assemble strain-displacement matrix. Take t = 20 mm and E = 2 x 105 N/mm2. asy En gin Figure 2 Or eer i ng. net (b) The temperature at the four corners of a four-noded rectangle are T1, T2, T3 and T4. Determine the consistent load vector for a 2-D analysis, aimed to determine the thermal stresses. 14. (a) Derive the expression for the element stiffness matrix for an axisymmetric shell element. Or (b) (i) Explain the terms ‚Plane stress‛ and ‚Plane strain‛ problems. Give constitutive laws for these cases. (ii) Derive the equations of equilibrium in case of a three dimensional system. VEL TECH VEL TECH MULTI TECH 232 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 15. (a) Establish the strain-displacement matrix for the linear quadrilateral element as shown in figure 3 at Gauss point r = 0.57735 and s= -57735. Figure 3 ww (b) Derive element stiffness matrix for a Linear Isoperimetric Quadrilateral element. w.E B.E./B.TECH DEGREE EXAMINATION, APRIL / MAY 2008 asy SEVENTH SEMESTER En MECHANICAL ENGINEERING gin ME 1401 – INTRODUCTION OF FINITE ELEMENT ANALYSIS PART – A eer i ng. 1. List any four advantages of finite element method. 2. Write the potential energy for beam of span ‘L’ simply supported at ends, subjected to a concentrated load ‘P’ at mid span. Assume EI constant. 3. What are called higher order elements? 4. Write briefly about CST element. 5. What is the governing differential equation for a one dimensional heat transfer? 6. Differentiate: Local axis and Global axis. 7. What is an equivalent nodal force? 8. Give one example each for plane stress and plane strain problems. 9. What do you mean by Constitutive Law and give Constitutive Law for axi-symmetric problems? 10. State the basic laws on which isoparametric concept is developed. VEL TECH VEL TECH MULTI TECH 233 net VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH PART – B 11. (a) Explain the Gaussian elimination method for solving of simultaneous linear algebraic equations with an example. Or (b) 12. A cantilever beam of length L is loaded with a point load at the free end. Find the maximum deflection and maximum bending moment using Rayleigh-Ritz method using the function Y = A (1 – cos (x / 2L)}. Given: EI is constant. ww (a) (b) (i) (ii) Derive the shape functions for a 2D beam element. Derive the shape functions for 2D truss element. w.E asy Or Each of the five bars of the pin jointed truss shown in figure (b) has a cross sectional area 20 Sq.cm. and E = 200 GPa. En gin eer i ng. (i) (ii) 13. (a) VEL TECH net Form the equation F = KU where K is the assembled stiffness matrix of the structure. Find the forces in all the five members. Find the temperature at a point P(1, 1.5) inside the triangular element shown with the nodal temperatures given as TI = 40C, TJ = 34C, and Tk = 46C. Also determine the location of the 42C contour line for the triangular element shown in figure (a). VEL TECH MULTI TECH 234 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Or (b) Calculate the element stiffness matrix and thermal force vector for the plane stress element shown in figure (b). The element experiences a rise of 10C. ww 14. 15. w.E (a) Derive the constant-strain triangular element’s stiffness matrix and equations. asy Or En (b) Derive the Linear-Strain triangular element’s stiffness matrix and equations. (a) Integrate f(x) = 10 + (20x) – (3x2 / 10) + (4x3 / 100) – (-5x4 / 1000) + (6x5 / 10000) between 8 and 12. Use Gaussian Quadrature Rule. gin Or (b) eer i ng. net Derive element stiffness matrix for a Linear Isoparametric Quadrilateral element. *********** B.E./B.TECH DEGREE EXAMINATION, NOVEMBER/DECEMBER – 2008 SEVENTH SEMESTER MECHANICAL ENGINEERING ME 1401 – INTRODUCTION OF FINITE ELEMENT ANALYSIS PART – A 1. Write the potential energy for beam of span ‘L’ simply supported at ends, subjected to a concentrated load ‘P’ at mid span. Assume EI constant. VEL TECH VEL TECH MULTI TECH 235 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH 2. What do you mean by higher order elements? 3. What is Constitutive Law and give constitutive law for axi-symmetric problems? 4. Explain the important properties of CST element. 5. Give one example each for plane stress and plane strain problems. 6. Write the stiffness matrix for the simple beam element given below. 7. Write the Lagrangean shape functions for a 1D, 2 noded element. 8. What are the advantages of natural coordinates over global co-ordinates? 9. Define Isoparametric elements? 10. Write the natural co-ordinates for the point ‘P’ of the triangular element. The point ‘P’ is the C.G. of the triangle. 11. ww w.E (a) asy PART – B En Determine the expression for deflection and bending moment in a simple supported beam subjected to uniformly distributed load over entire span. Find the deflection and moment at midspan and compare with exact solution using Rayleigh-Ritz method. Use y = a1 sin (x/1) + a2 sin (3 x/1). gin Or 12. eer i ng. net (b) Derive the equation of equilibrium in case of a three dimensional stress system. (a) Derive the shape function for a 2 noded beam element and a 3 noded bar element. Or 13. (b) Write the mathematical formulation for a steady state heat transfer conduction problem and derive the stiffness and force matrices for the same. (a) Find the expression for nodal vector in a CST element shown in figure (a) subject to pressures Px1 on side 1. VEL TECH VEL TECH MULTI TECH 236 VEL TECH HIGH TECH Download From: www.EasyEngineering.net Download From: www.EasyEngineering.net VEL TECH VEL TECH MULTI TECH VEL TECH HIGH TECH Or 14. (b) Determine the shape functions for a constant strain triangular (CST) element in terms of natural coordinate system. (a) For the CST element given below figure (a) assemble strain displacement matrix. Take t = 20 mm, E = 2 105 N/mm2. ww w.E asy En gin Or 15. eer i ng. net (b) Derive the expression for constitutive stress-strain relationship and also reduce it to the problem of plane stress and plane strain. (a) Write short notes on (i) Uniqueness of mapping of isoparametric elements. (ii) Jacobian matrix. (iii) Gaussian Quadrature integration technique. Or (b) (i) (ii) VEL TECH b a 0 0 Use Gauss quadrature rule (n = 2) to numerically integrate xydxdy. Using natural coordinates derive the shape function for a linear quadrilateral element. ***************** VEL TECH MULTI TECH 237 VEL TECH HIGH TECH Download From: www.EasyEngineering.net