Research Journal of Environmental and Earth Sciences 3(3): 187-192, 2011 ISSN: 2041-0492 © Maxwell Scientific Organization, 2011 Received: July 17, 2010 Accepted: September 17, 2010 Published: April 05, 2011 An Analytical Model for Predicting Drag Resistance Caused by Internal Waves in a Stably Stratified Open Channel Flow 1 E.C. Obinabo and 2O.L. Ndubuisi Department of Electrical and Electronic Engineering, 2 Department of Civil Engineering, Ambrose Alli University, P.M.B., 14, Ekpoma, Edo State, Nigeria 1 Abstract: This study explores optimal control via modelling of an open-channel flow, primarily pertaining to the general problem of drag resistance associated with gravity waves in a stably stratified fluid, and the influence of gravity waves on offshore structures. The study applies the results reported in the existing literature to gravity gradient waves using essentially both analytical and numerical techniques, and establishes, among others, detailed characterization of drag resistance effects for application to free and forced conditions in open channel flow. Key words: Drag resistance, internal gravity waves, open channel flow, stratified fluid An earlier study by Scorer (1949) proposed a lee wave term of the type studied by Obinabo (1978) and although his basic equations were based on perturbation theory, he approached his problem by considering a model of two layers instead of one, and in which GO was constant in each. He showed that the length of the waves is such that their speed in the upstream direction relative to the water is equal to the speed of the water in the downstream direction. He also showed that the only quantity which could influence the lee wavelength is the wind speed, and inferred that the wavelength must therefore be determined solely by the terms and Uo and their variations with height. Large amplitude wave theory was developed by Long (1953), and he produced an exact steady state equation of motion and continuity of a perfect liquid moving in two-dimensions with an arbitrary vertical distribution of density and velocity. He integrated the equation once to yield a second order differential equation which he examined with regard to uniqueness and stability of the motion. He developed a criterion giving a sufficient condition for the motion to be uniquely determined by the configuration of the topography over which the fluid moves. He also approached the problem on non-linear solution of the equation of motion of the fluid by considering specific cases in which the non-linear terms were identically zero before he produced analytical results. Groves (2004) regarded the non-linear terms of the equation as "perturbations" on the linearized case and produced a solution which he argued would provide wave solutions not entirely dissimilar to those obtained from the linearized equation. His work also includes the INTRODUCTION The first contributions to the subject of wave forces included valuable results of theoretical and field studies (Baddour and Abbink, 1983) and were similar to those studied by Lukomsky and Gandzha, (2003). the theory assumes that the waves are periodic in the horizontal direction, and independent of time (Jamaloddin et al., 2005,). Mathematical relations were subsequently formulated for computing the forces exerted by breaking waves (Eyo, 2007a; Drennan, 1992). Motion of stratified fluids with laminar flow has been studied extensively in the past several decades. Long (1953) put forward a theory for waves in an airstream which had a uniform stability and velocity up to infinity, and demonstrated that there could be waves characterized by more than one wave crest on each streamline although there is one on the ground. He also showed that these waves have amplitude which increased upwards, and decreased downwards. similar results were computed in which slight modifications were made for the effect of the earth's rotation (Sen-Gupta, 1973; Phillips, 1977). It was shown (Ndubuisi and Obinabo, 2010) that for ordinary sized mountains which had no steep sides the second the subsequent waves had a very small amplitude. The theory of both of these authors is equivalent to the assumption that the wavelengths of the perturbations are small compared to the depth of the atmosphere. The theory, however, was considered inadequate as it was observed that in the troposphere there was only one harmonic wave of large amplitudes known as a lee wave. Corresponding Author: E.C. Obinabo, Department of Electrical and Electronic Engineering, Ambrose Alli University, P.M.B. 14, Ekpoma, Edo State, Nigeria 187 Res. J. Environ. Earth Sci., 3(2): 187-192, 2011 application of the Fourier series in the solution of the wave equation (Oyetunde, 2007; Oyetunde and Okeke, 2007). In fact, although a lot of work has been done on the subject of the internal gravity waves, in stratified fluid (Polton et al., 2004; Naser et al., 1980), none of the published works was devoted to the investigation of the drag resistance due to the internal gravity waves. Consequently this project is, to a great extent, a pioneering venture in this field. The purpose of the study was to develop a twodimensional model of a train of free internal gravity waves in a stably stratified fluid bounded by the four sides of an open tank and free surface. Computation of the model was enhanced by the assumption that the flow in the open channel is linear and stabilizable. Our results are presented for values of the amplitude which was varied from small to large to demonstrate the development of rotor regions, the effect of which was shown to have some considerable influence on the values of the drag (Obinabo, 1978). The measured values of the drag were quite small but they are within the expected magnitude for the laboratory model considered. ∂P ∂ . ∂ x ∂Z ⎛ q2 ⎞ ∂P ∂ ⎛ q2 ⎞ ⎜⎜ ⎟⎟ + . ⎜⎜ ⎟⎟ ⎝ 2 ⎠ ∂z ∂ x ⎝ 2 ⎠ u∂ P ∂2P ∂P +w =− −g ∂x ∂ x∂ z ∂x Subtract (3) from (4) we obtain: ∂P ∂ . ∂ x ∂Z ⎛ q2 ⎞ ∂P ∂ ⎛ q2 ⎞ ⎜⎜ ⎟⎟ − . ⎜⎜ ⎟⎟ ⎝ 2 ⎠ ∂z ∂ x ⎝ 2 ⎠ ⎛ u∂ P w∂ P ⎞ ∂P + w⎜ + ⎟ = −g ∂z ⎠ ∂x ⎝ ∂x Dividing through by P and re-arranging, we obtain: ∂w 1 ∂ P d ⎛ q 2 ⎞ 1 ∂ P d ⎛ q 2 ⎞ ⎜ ⎟− + . ⎜ ⎟ ∂t p ∂ x ∂ z ⎜⎝ 2 ⎟⎠ p ∂ z ∂ x ⎜⎝ 2 ⎟⎠ g ∂P + . =0 p ∂x MATERIALS AND METHODS The theoretical study reported in the literature (Ostrowski and szmytkiemicz, 2006; Eyo, 2007b) considers a streamline wave motion in which an instantaneous displacement of a fluid particle from the undisturbed level applied to Lagrange’s equation to the fluid particle to obtain: ∂ ⎛ q2 ⎞ ∂ ⎜ ⎟ − wPw = − P ∂x⎝ 2 ⎠ ∂x P where, q2 = u2 + w2 by multiplying (1) by by d d d =u +w dt ∂x ∂z Since the fluid is incompressible, there is a streamline functions * (x,z) such that: U =− (2) −w ∂ P w∂ P =− ∂z ∂ x∂ z 2 ∂δ ∂δ , w= ∂z ∂x Also, the density P is conserved so that P = P (*) Equation (5) can be written as: ∂ and (2) ∂Z ∂w 1 ∂ P ⎛ d ⎛ q 2 ⎞ ⎞ d ⎛ q2 ⎞ ⎜ ⎟ u ⎜ ⎟ +w ⎜ ⎟ + ∂t p ∂δ ⎜⎝ ∂ x ⎝ 2 ⎠ ⎟⎠ ∂z ⎝ 2 ⎠ ∂ we obtain: ∂x ∂P ∂ ⎛ q2 ⎞ ∂P ∂ ⎛ q2 ⎞ . ⎜ ⎟+ . ⎜ ⎟ ∂ Z ∂ x ⎜⎝ 2 ⎟⎠ ∂ x ∂ z ⎜⎝ 2 ⎟⎠ (5) where, (1) ∂ ⎛ q2 ⎞ ∂P ⎜ ⎟ + wPw = − − Pg ∂Z ⎝ 2 ⎠ ∂Z (4) (6) g d P dz + . = =0 p dδ dt (3) p and dp are not functions of time, hence (6) may be dδ written as: 188 Res. J. Environ. Earth Sci., 3(2): 187-192, 2011 d dt ⎧ ⎞⎫ 1 ∂p ⎛ q2 ⎪ ⎪ ⎜⎜ + gz⎟⎟ ⎬ = 0 ⎨w + p ∂δ ⎝ 2 ⎠⎪ ⎪ ⎩ ⎭ Clearly this equation is nonlinear and cannot be solved easily by a mathematical analysis. There is, however, one obvious model for which the equation can be linearised, (7) 2 and this is one in which the term U ο p remains constant with depth and the density is linear in Zo. Eq. (12) then reduces to: Integrating (7) and noting that: ∂ w ∂ u ∂ 2δ ∂ 2δ w= − = + ∂ x ∂z ∂ x2 ∂z2 ∇ 2δ + Gοδ = 0 we obtain: g ⎫ ∂ p ⎧⎪ ( ∇δ ) ⎪ + gz ⎨ ⎬ = H (δ ) dδ ⎪ 2 ⎪ 2 ∇ 2δ + ⎩ where, Gο = (8) ⎭ ⎫ ⎪ + g Z0 − Z ⎬ ⎪⎭ ( δ ( x , z ) = δ ( 0, z ) Coskx (9) ) where, k = This result can be written in a different and useful form by substituting Zo for the dependable variable *. If Uo does not vanish anywhere, the relation: d 1 d =− dδ U ο dzο (10) or − k {( )2 − 1} dzd ( LnU ο2 P) 1 ∇ zο 2 (14) (15) U ο2 P ( d Z −Z dzο ο λ δ+ 2 then (15) reduces to: ∂ 2δ Coskx + GοδCoskx = 0 ∂z2 (17) ∂ 2δ + Gοδ = 0 ∂z2 δ ( 0, z ) = ASinν z + BCosν z (11) ) (18) where, A and B are constants. Applying the boundary conditions: With further substitution * = Zo – Z: * = 0, when z = 0 then B = 0 dδ ⎫ d 1⎧ 2 LnU ο2 P ⎨ (∇ δ ) − 2 ⎬ ∂ z ⎭ dzο 2⎩ (12) g ∂d = 2 .δ U ο P dzο ( (16) The solution of this equation is of the form: ο g 2π k 2 Coskx + leads to: ∇ 2δ + gβ = constant U ο2 Assuming a periodic motion of the fluid particle, that is: ( )2 = = ∂ 2δ ∂ 2δ + + Gοδ = 0 ∂ x2 ∂z2 1 ∂p = p dδ ⎧ 1 ∂ p ⎪ Uο w(δ ) + ⎨ p dδ ⎪ 2 ⎩ ∇ 2 zο + U ο2 Solution of the wave equation: We generalized the model (13) as follows: where, H(*) is a function of * to be determined by the conditions upstream. Thus for the height zo(*) of the streamline * = cons tant and T(*) is the vorticity upstream, the final result becomes: ∇ 2δ + 1 dρ ρ dzο (13) ) For z = H, Eq. (18) becomes: A Sin< H = 0 Since A cannot be zero, then: Sin< H = 0 189 Res. J. Environ. Earth Sci., 3(2): 187-192, 2011 that is < H = n B (n = 1, 2, 3,...) we then obtain: (19) υ 2 H 2 = n 2π 2 = Gο − k 2 and (20) i.e., 2π λ= Therefore, Yx = Cos (21) ⎛ n 2π 2 ⎞ ⎜ Gο − ⎟ H2 ⎠ ⎝ 2 gβ ⎧ ⎛ δ ⎞⎫ 1 − υ = ⎨Cos −1 ⎜ ⎟⎬ ⎝ A sin υz ⎠ ⎭ x ⎩ 2 U ο2 2 ⎫ gβ 1 ⎧ δ ⎞⎤ ⎪⎡ 2 ⎪ −1 ⎛ + Cos υ x = ⎟ ⎜ ⎨⎢ ⎬ ⎝ A sin υz ⎠ ⎥⎦ U ο2 x ⎪⎣ ⎪ ⎩ ⎭ Also zο = z − δ = z − δ ( 0, z ) Coskx . Therefore, zο = z − ASinυz Cos kx Therefore, (22) U ο2 = The constant A is an arbitrary amplitude. Significance of n: The expression (----) give the distribution of * at x = 0, From (19)< H = n B, and for n = 1 we obtain: D= <H=B For which the distribution may be obtained to represent the resulting waveform. Drag = The displacement of a fluid particle from an undisturbed layer can be written as * = z ! zo, But zo = z ! A SinLz Cos kx K is a function of the wavelength and of the streamline velocity Uo so that in (23) K was written as: gβ 2 ⎝ Uο 1 CD AS ρU ο2 2 (25) (26) CD AS ρ gβx (27) 2 2 ⎧ δ ⎞⎫ −1 ⎛ 2 ⎟⎬ +υ x ⎨Cos ⎜⎝ A sin υz ⎠ ⎭ ⎩ Solution by Fourier's method: In this solution, a linear model for internal waves in a stratified fluid is considered. The equations are generalized by Fourier's analysis in order to consider motions past an obstacle on the surface. In his approach to this problem Scorer (1949) considered the infinite atmosphere. Sen-Gupta (1973) considered the same problem but applied the theory to a linearized model of stratified water for a laboratory scale experiment. We obtained: (24) Therefore, from (23) ⎛ 2 ⎧ δ ⎞⎫ −1 ⎛ 2 ⎟⎬ +υ x ⎨Cos ⎜⎝ ⎠ A z sin υ ⎩ ⎭ where, A = wave amplitude, As = projected area of the body in the direction of the motion. The values of * can be obtained by applying Fourier's analysis to the linearized wave equation. This solution takes into account the geometry of the mountain model used in generating the waves in the fluid. Therefore, * = A SinLz Cos kx = A sin L z cosk x (23) δ = ASinυzCos⎜⎜ gβx Substituting (25) in (26) we obtain: RESULTS AND DISCUSSION ⎛ gβ ⎞ k = Gο − υ 2 = ⎜⎜ 2 − υ 2 ⎟⎟ ⎝ Uο ⎠ δ ⎞ ⎜ ⎟ ⎝ A sin υz ⎠ −1 ⎛ ⎞ − υ 2 ⎟⎟ x ⎠ ∇ 2δ + Gοδ = 0 Writing the transform; ⎛ gβ ⎞ 1 ⎞ −υ2⎟ ⎜ ⎟ = CosYx , Y = ⎜ A ⎝ Sinυz ⎠ ⎝ Uο ⎠ δ ⎛ δ for δ = δ ( x , z ) 190 (28) Res. J. Environ. Earth Sci., 3(2): 187-192, 2011 δ= ∫ ∞ e −ikxδdx The integral (36) was evaluated by the method of contour integration (Scorer, 1949). From Cauchy’s theorem, the integral round the closed contour C was as follows: (29) 0 hence (28) become transformed to: d 2δ − k δ + 2 + Gο δ = 0 dz 2 ∫ f ( k )dk = 2πi × sum of the residues at the poles (30) e inside the contour The residue at the pole was calculated using the relation: This has the solution: δ = ASinυz + BCosυz f (k) = We now apply the boundary conditions C C At z = 0, At z = H δ = 0 therefore B = 0 and we show that the pole exists where g2(k2). With the assumption that k = k* at the pole, the residue was computed from the relation: M ∫ δ ( H ,k ) = ASinυH = e −ikxδ( N ,k ) dx = F ( k ) (31) R= 0 Therefore, we obtain: δ ( H ,k ) = Hence δ( z , k ) = ∫ ∞ 0 g1 ( K1 ) g2 ( K2 ) F ( K ) sin υz Sinυz (32) F ( K ) sin υz ikz e dk SinυH (33) g1 ( k *) g2′ ( k *) It is assumed that the pole is of the first order, and. that the formula is not general. A more rigorous mathematical argument would not necessarily serve a better purpose as the theory of this problem is at best an approximation to the physical features encountered in practice. From Cauchy's theorem: ∞ ∫ δ( z , k ) = − α 2 e − kb +ibx This integral was solved by choosing a suitable value for F(k). Scorer has used a ridge profile of shape: 0 ∞ δ( H ,k ) = −a b b2 + x 2 (37) ∫ δ( z ,k ) = − α 2 e − kb +ikx 2 (34) + This shape was mountain model used in the experiment (Obinabo, 1978; Ndubuisi and Obinabo, 2010). From above: 0 2 − k *b +ik * x πia e Sinυz dk SinυH Sinυz dk SinυH Sinυ * z ⎛ ∂υ *⎞ HCosυ * H ⎜ ⎟ ⎝ ∂k ⎠ (38) * , ( x > 0) (39) ∞ − a 2b δ( H ,k ) = 2 = − a 2 eikx − kh dk 2 b +x ∫ CONCLUSION (35) 0 The analysis of the result by Fourier's method gave rise to the values of the wave amplitudes which was compared with the experimental results (Ndubuisi and Obinabo, 2010). The results also show that the formation of the rotors is a function of the wave amplitude, and they tend to present some complications in the flow of the fluid which may affect the values of the drag. the drag bears a linear relationship with the wave amplitude. This reduces the problem to the solution of the integral: ∞ δ( Z ,k ) = − a 2 ∫ 0 eikx − kh Sinυz SinυH (36) 191 Res. J. Environ. Earth Sci., 3(2): 187-192, 2011 The study relates the dynamics of the fluid flow to the drag and lift components of the resultant dynamic force exerted on the body immersed in the fluid. Drag, the resistance to motion, is the component of the resultant force in the direction of the relative flow ahead of the body, and the lift is the component normal to this direction. The former, however, is the component we were involved with in this study, and the steady state solution for the displacement of the fluid particle for each velocity variation was obtained by Fourier's analysis. This took into account the geometry of the mountain model developed by Obinabo (1978) to generate the waves in the stratified fluid. The analysis of the drag coefficient of the body was based strictly on the geometry of the mountain model. REFERENCES Baddour, R.E. and H. Abbink, 1983. Turbulent Underflow in a Short Channel of Limited Depth, J. Hydr. Div. ASCE, 109(HY6): 722-740. Drennan, W.M., 1992. Accurate calculations of stokes water waves of large amplitude. Z. Angew. Math. Phys., 43: 367-384. Eyo, A.E., 2007a. Mathematical modelling of uniform flow in three open channels. J. Nig. Ass. Math. Phys., 11: 587-596. Eyo, A.E., 2007b. Mathematical modelling and its impact on dredging a rectangular open channel with hydraulic jump. J. Sci. Eng. Tech., 14(2): 7375-7387. Groves, M.D., 2004. steady water waves. J. Nonlinear Math. Phys., 11(4): 435-460. Jamaloddin, N., I.B. Samsul, S.J. Mohammad, A.M. Waleed and M. Shahrim, 2005. Simulation of waves and current forces on template offshore structure. Suranaree J. Sci. Tech., 12(3): 193-210. Long, R.R., 1953. Some aspects of flow of stratified fluids. 1. a theoretical investigation. Tellus, 5: 42-58. Lukomsky, V.P. and I.S. Gandzha, 2003. fractional fourier approximations for potential gravity waves in deep water. Nonlinear Proc. Geophy., 10: 599-614. Naser, M.S., P. Venkataraman and A.S. Ramamurthy, 1980. Flow in a channel with a slot in the bed. J. Hydr. Res., 18(4): 359-367. Ndubuisi, O.L. and E.C. Obinabo, 2010. Experimental evaluation of gravity wave effects in a stably stratified open channel flow. Res. J. Environ. Earth Sci., 3(3): 193-196. Obinabo, E.C., 1978. Drag resistance caused by gravity waves in a stratified fluid. B.Sc. Thesis, University of Portsmouth, England, United Kingdom. Ostrowski, R. and M. szmytkiemicz, 2006. Modelling longshore sediment transport under asymmetric waves. Oceanologia, 48(3): 395-412. Oyetunde, B.S., 2007. On the effects of wave steepness on higher order stokes waves. J. Nig. Ass. Math. Phys., 11: 369-374. Oyetunde, B.S. and E.O. Okeke, 2007. on the impact of wave-current on stokes waves. J. Nig. Ass. Math. Phys., 8: 363-368. Phillips, D.M., 1977. On the generation of surface waves by turbulent wind. J. Fluid Mech., 2: 417-452. Polton, J.A., D.M. Lewis and S.E. Belcher, 2004. The role of wave-induced coriolis-stokes forcing on the wind driven mixed layer. J. Phys. Oceanogr., 35: 444-457. Scorer, R.S., 1949. Theory of waves in the lee of mountains. Q. J. Royal Met. Soc., 75: 41-56. Sen-Gupta, B.K., 1973. Large amplitude internal gravity waves in stably stratified fluids: linear model. Meccanica, 1(8). ACKNOWLEDGEMENT The authors gratefully acknowledge the cooperation received from their colleagues in the Faculty of Engineering and Technology, Ambrose Alli University, Ekpoma, Edo State, Nigeria, for the success of this research. NOMENCLATURE x, z u, w a, b A AS * *1 Z Uo Po g Space co-ordinates in two-dimensional system Velocities in and directions respectively Dimensions of the mountain model Wave amplitude Projected area of the mountain model in the direction of motion. Vertical displacement from undisturbed position Boundary layer thickness Height of streamline when undisturbed Undisturbed stream velocity Density in undisturbed stream Gravitational constant $o Static stability = G Stability parameter = K Horizontal wave number = H H 8 w )2 CD D Horizontal wave number Depth of fluid model Wavelength Vorticity vector Laplacian operator Coefficient of drag Drag force 1 ∂p p0 ∂ Z0 gβ0 U 02 2π wavelength 192