LAB 3: LIQUID VISCOSITY AND DRAG ON A SPHERE Tuesday April 15th, 2014 Kanchan Bhattacharyya – First Author, Discussion and Conclusion Xie Zheng – Theory, Experimental Procedure, and List of Equipment Matthew Stevens – Introduction and Results Ting Zhang – Abstract, Error Analysis Group 4 ABSTRACT This experiment mainly consists of three parts. Students are firstly required to learn to determine the viscosity of liquid by using the falling ball. Next they are asked to learn to measure the drag force coefficient of a sphere in the same fluid. Finally, students illustrate the way these data can be applied to other fluid and other combinations. In this experiments, a bunch of variables are required to be measured or calculated in order to calculate dynamic viscosity of fluid and drag coefficient of sphere. They are density of sphere ππ , density of liquid ππ , sphere diameter d, container diameter D, velocity of sphere in container as D approaches infinity π∞ , measured ball velocity ππ‘ . From the results we got, the experiment data are in good accuracy when compared to theoretical data. INTRODUCTION The discipline of fluid mechanics covers everything from the aerodynamics of planes and spacecraft to the mechanics of biological organisms. Because fluid flow phenomenon can include a wide variety of media of varying properties Engineers generally subdivide fluid mechanics with regards to two principle parameters, their viscous nature and compressibility. [1] Viscosity is a measure of a fluid’s resistance to flow, and is largely dependent on the temperature of the fluid. The presence of viscous effects explain everything from the drag experienced by a ball throw into the air or when you drop it in a beaker of Glycerol , and many relationships have been derived that give us an ability to quantity the presence of these viscous effects in terms of measurable variables. The most notable of which, the Reynolds Number, describes the relationship between the inertial effects and viscous effects of fluid. A dimensionless quantity, it is named after Osborn Reynolds who after exhaustive experiments in the 1880’s discovered that flow regime depends mainly on the ratio of these effects[2]. The magnitude of the Reynolds Number quantifies the presence of viscous effects in a flow, with small Reynolds numbers less than 0.1 signifying fluids dominated such effects. Small Reynolds numbers have been found to quantify the viscous nature of many fluid phenomena, including those such as a biological virus moving in water[3] or the motion of bubbles in channels and tubes [4]. While fluid motion can be complex, use of parameters such as the Reynolds Number allow for quantification of fluid phenomena regardless of the nature of fluid or object traveling through the fluid. For example, we can use the Reynolds Number to assess the true viscous nature of a fluid via analysis of the motion of an object traveling through the fluid regardless of the nature of the fluid or the traveling object. We find that the power of the Reynolds Number lies in this versatility, and can be used in any scenario to describe the nature of fluid flow. LIST OF EQUIPMENT Stopwatch Electrical Caliper Model #: C-521 Manufacturer: KOBALT Filter to guide the spheres to the center of the tube Specimen: three spheres Testing device Level Model #:7724 Manufacturer: Johnson Level & Tool MPG. Co., Inc THEORY Viscosity Measurements The viscosity of a fluid is a measure of its resistance to gradual deformation by shear stress or tensile stress. Viscosity is due to the friction between neighboring particles in a fluid that are moving at different velocities. In this experiment, we are dropping spheres into the liquid along the center line of a container filled with the test liquid. By measuring the time that spheres travel between lines, the velocity of the sphere will be determined. Once the velocity of the sphere is determined, the dynamics viscosity of fluid could be simply obtained through an expression: µπ = (ππ − ππ )ππ2 (Eq.1) 18π∞ where µf = dynamic viscosity of fluid ρs = density of sphere ρf = density of liquid g = acceleration of gravity d = sphere diameter D = container diameter U∞ = velocity of sphere in container as D goes to infinity Ut = measured ball velocity Since the experiment is processed inside a tube with finite diameter, U∞ should be corrected into Ut. Here, we use the following wall-correction expression: πΆπ = ππ‘ π∞ π 9⁄ 4 = (1 − π·) (Eq.2) Also, the Reynolds number must be less than 0.1 to perform the Eq.1 π ππ‘ = ππ ππ‘ π µπ (Eq.3) Sphere Drag To do the force analysis on the sphere, the balance of the force could be found in the following equation: 4 ππ 3 π (ππ 3 − ππ ) = πΉπ· (Eq.4) By rearranging Eq.1, we can rewrite Eq.4 into: πΉπ· = 6πµπ∞ π (Eq.5) The drag coefficient, CD, is defined as the drag force divided by the cross sectional area of the sphere and the kinetic energy of the ball: πΆπ· = πΉπ· 1 2 π΄πΆπ ππ π∞ 2 (Eq.6) Plug Eq.5 into Eq.6, we can get: πΆπ· = 12µπ π ππ π∞ (Eq.7) If we let d = 2R, the drag coefficient, CD, could be rewrite as: πΆπ· = 24 π π∞ (Eq.8) Alternative Equation for Dynamic Viscosity from Fig. 3 of Ref. 6: (Eq. 9) PROCEDURE Since the experiment was already setup in the lab, we check the connections of the equipment and make sure we have all the instruments we need. Before testing, we used level to make sure the tube is vertical, measured the inside diameter of the tube with caliper and the distances between the circle line on the tube. We also measure the temperature of the glycerol right before and after the experiment. After recording the readings, we placed a filter at the top of the tube to make sure the balls would fall down along the center line of the tube. We dropped the first ball which is largest and record the time duration when it went through the lines and wrote down the readings. And then we dropped the second one and the third one and recorded the time duration as well. Then, we took out all the three balls for other trials by repeat the above steps until we get all the five trials. Once we finished, we cleaned up the lab and put everything back. RESULTS Table 1. Description of Sphere Parameters Number Designation Description Density, ρs (kg/m3) Diameter, d (mm) 1 2 3 Red Plastic - Large Steel - Large Steel - Small 3990 ± 2% 8000 ± 2% 8040 ± 2% 3.173 ± 1% 3.169 ± 1% 1.583 ± 1% Table 1. The assigned number designation, description, given density, and diameter of each of the respective test spheres. Table 2. Description of Glass Tube Testing Segments Number Designation Length, l (cm) 1 24.6 2 25.5 3 25.2 Table 2. Recorded Lengths of each of the segments comprising the testing tube. Table 3. Glass Test Tube Inner Diameter Measurements Measurement Number Diameter, D (mm) 1 46.01 2 46.6 3 46.6 4 46.15 Average 46.34 Table 3. Recorded Diameters for various cross sections of the circular tube. An average value was calculated to be used in analysis. Table 4. Glycerol Temperature Measurements Measurement Number Description Temperature, T (°C) 1 Before Experiment 23 2 After Experiment 24 Table 4. Recorded Glycerol temperatures taken before and after the experiment. 9 ππ‘ π 4 ππππ πΆππππππ‘πππ πΉπππ‘ππ, πΆπ€ = = (1 − ) π∞ π· 9 3.173ππ 4 πππ ππβπππ 1, (1 − ) = .85 46.34ππ 9 3.169ππ 4 πππ ππβπππ 2, (1 − ) = .85 46.34ππ 9 1.583ππ 4 πππ ππβπππ 3, (1 − ) = .92 46.34ππ Fig. 1 Calculation of Wall Correction Factors The calculation of the respective wall correction factors for each sphere using Equation 2. The diameters for each sphere and the Tube diameter can be found in Table 1 and Table 3. Table 5. Average Wall Correction Factors for Respective Spheres Number Designation Wall Correction Factor, CW 1 0.85 2 0.85 3 0.92 Table 5. Calculated wall correction factors to be used to calculate respective terminal velocities for each respective sphere. Table 6. Sphere 1 Velocity Calculations for Segment 2 Trial Time, t (sec) Velocity, Ut (m/s) Terminal Velocity, U∞ (m/s) 1 2 3 4 5 Average 15.47 15.42 15.47 15.50 15.25 15.42 1.65E-02 1.65E-02 1.65E-02 1.65E-02 1.67E-02 1.65E-02 1.93E-02 1.94E-02 1.93E-02 1.93E-02 1.96E-02 1.94E-02 Table 6. Recorded Traversing times and calculated sphere 1 tube and terminal velocities through the second tube segment. Table 7. Sphere 1 Velocity Calculations for Segment 3 Trial Time, t (sec) Velocity, Ut (m/s) Terminal Velocity, U∞ (m/s) 1 16.78 1.50E-02 1.76E-02 2 16.50 1.53E-02 1.79E-02 3 16.66 1.51E-02 1.77E-02 4 16.47 1.53E-02 1.79E-02 5 16.50 1.53E-02 1.79E-02 Average 16.58 1.52E-02 1.78E-02 Table 7. Recorded Traversing times and calculated sphere 1 tube and terminal velocities through the third tube segment. Table 8. Sphere 1 Velocity Calculations Trial Segment Terminal Velocity, U∞ (m/s) 1 2 2 2 1.93E-02 1.94E-02 3 2 1.93E-02 4 2 1.93E-02 5 2 1.96E-02 1 3 1.76E-02 2 3 1.79E-02 3 3 1.77E-02 4 3 1.79E-02 5 3 1.79E-02 Average 1.86E-02 Table 8. A comprehensive listing of all the calculated terminal velocities in Tables 6 and 7. This average terminal velocity will be used in calculation of both the Reynolds Number and the drag coefficient of first sphere. πΈπ₯. ππβπππ πππ, πππππππ‘ ππ€π, πππππ πππ ππ‘ = πππ£ππ πΆπ€ = π· 0.0255 π π = = 1.65 π₯ 10−2 π‘ 15.47 π π ππ‘ π∞ ↔ π∞ = ππ‘ π = 1.93 π₯ 10−2 πΆπ€ π Fig 2. Calculation of Sphere Tube and Terminal Velocities The procedures for calculating tube velocity considering the length of each segment and time taken for the sphere to traverse that distance and the terminal velocity considering the respective wall correction factors. Table 9. Sphere 2 Velocity Calculations for Segment 2 Trial Time, t (sec) Velocity, Ut (m/s) Terminal Velocity, U∞ (m/s) 1 6.38 4.00E-02 4.69E-02 2 6.20 4.11E-02 4.82E-02 3 6.33 4.03E-02 4.72E-02 4 6.22 4.10E-02 4.81E-02 5 6.21 4.11E-02 4.82E-02 Average 6.27 4.07E-02 4.77E-02 Table 9. Recorded Traversing times and calculated sphere 2 tube and terminal velocities through the second tube segment and the procedure from Fig. 2 (Calculation of Tube and Terminal Velocities) Table 10. Sphere 2 Velocity Calculations for Segment 3 Trial Time, t (sec) Velocity, Ut (m/s) Terminal Velocity, U∞ (m/s) 1 6.81 3.74E-02 4.39E-02 2 6.72 3.79E-02 4.45E-02 3 6.62 3.85E-02 4.52E-02 4 6.55 3.89E-02 4.57E-02 5 6.61 3.86E-02 4.52E-02 Average 6.66 3.83E-02 4.49E-02 Table 10. Recorded Traversing times and calculated sphere 2 tube and terminal velocities through the third tube segment and the procedure from Fig 2. (Calculation of Tube and Terminal Velocities) Table 11. Sphere 2 Average Terminal Velocity Calculations Trial Segment Terminal Velocity, U∞ (m/s) 1 2 2 2 4.69E-02 4.82E-02 3 2 4.72E-02 4 2 4.81E-02 5 2 4.82E-02 1 3 4.39E-02 2 3 4.45E-02 3 3 4.52E-02 4 3 4.57E-02 5 3 4.52E-02 Average 4.63E-02 Table 11. A comprehensive listing of all the calculated terminal velocities in Tables 9 and 10. This average terminal velocity will be used in calculation of both the Reynolds Number and the drag coefficient of second sphere. Table 12. Sphere 3 Velocity Calculations for Segment 2 Trial Time, t (sec) Velocity, Ut (m/s) Terminal Velocity, U∞ (m/s) 1 22.94 1.11E-02 1.20E-02 2 22.97 1.11E-02 1.20E-02 3 22.66 1.13E-02 1.22E-02 4 22.72 1.12E-02 1.21E-02 5 22.35 1.14E-02 1.23E-02 Average 22.73 1.12E-02 1.21E-02 Table 12. Recorded Traversing times and calculated sphere 3 tube and terminal velocities through the second tube segment and the procedure from Fig 2. (Calculation of Tube and Terminal Velocities) Table 13. Sphere 3 Velocity Calculations for Segment 3 Trial Time, t (sec) Velocity, Ut (m/s) Terminal Velocity, U∞ (m/s) 1 24.99 1.02E-02 1.10E-02 2 24.99 1.02E-02 1.10E-02 3 24.62 1.04E-02 1.12E-02 4 24.39 1.05E-02 1.13E-02 5 24.01 1.06E-02 1.15E-02 Average 24.60 1.04E-02 1.12E-02 Table 13. Recorded Traversing times and calculated sphere 3 tube and terminal velocities through the third tube segment and the procedure from Fig 2. (Calculation of Tube and Terminal Velocities) Table 14. Sphere 3 Average Terminal Velocity Calculations Trial Segment Terminal Velocity, U∞ (m/s) 1 2 2 2 1.20E-02 1.20E-02 3 2 1.22E-02 4 2 1.21E-02 5 2 1.23E-02 1 3 1.10E-02 2 3 1.10E-02 3 3 1.12E-02 4 3 1.13E-02 5 3 1.15E-02 Average 1.17E-02 Table 14. A comprehensive listing of all the calculated terminal velocities in Tables 12 and 13. This average terminal velocity will be used in calculation of both the Reynolds Number and the drag coefficient of third sphere. Table 15. Average Terminal Velocities for Respective Spheres Number Designation Description Avg. Terminal Velocity, U∞ (m/s) 1 Red Plastic - Large 1.86E-02 2 Steel - Large 4.63E-02 3 Steel - Small 1.17E-02 Table 15. A comprehensive listing of the calculated average terminal velocities for each sphere taken from Tables 8, 11, and 14. These velocities were used in determination of the respective Reynolds Numbers and drag coefficients for each sphere. πΈπ₯. πΆππππ’πππ‘πππ ππ π΄π£πππππ ππππππππ πππππππ‘π¦ Μ ∞ = πππ ππβπππ 1, π ∑ U∞ π πΏππ‘ π = ππ’ππππ ππ ππππ π’ππππππ‘π = 10 ∑ U∞ = U∞ 1 + β― + U∞ π = (1.93 π₯ 10−2 π/π ) + (1.94 π₯ 10−2 π/π ) + β― . +(1.79 π₯ 10−2 π/π ) = 1.86 π₯ 10−1 π/π ∑ U∞ 1.86 π₯ 10 Μ ∞ = π = π 10 −1 π π = 1.86 π₯ 10−2 π/π Fig 3. Procedure for Calculating Average Terminal Velocities for Respective Spheres An example of how the average terminal velocity was calculated for each sphere using the data taken from the sphere 1 trials as an example sample set. The data used can be found in Table 8. The same procedure was extended to the calculation of the terminal velocities for spheres 2 in Tables 11 and 14. πΆπππ ππππ ππππ€π ππππ ππ‘πππ ππ πΊππ¦πππππ ππ‘ ππππππππ‘π’πππ π0 = 20°πΆ πππ π1 = 30°πΆ ππ ππ‘ π0 = 20°πΆ, π0 = 1,264.02 π3 ππ ππ‘ π1 = 30°πΆ, π1 = 1,258.09 π3 π΅π¦ πΏπππππ πΌππ‘πππππππ‘πππ, π€π πππ ππππ π‘βπ π·πππ ππ‘π¦ ππ πΊππ¦πππππ πππ { ππ | 20°πΆ < ππ < 30°πΆ } πΆπππ ππππ ππππππππ‘π’ππ ππ π‘βπ πΊππ¦πππππ ππππππ π‘βπ ππ₯ππππππππ‘, ππ = 23β ππ ππ π1 − π0 ππ 1,258.09 π3 − 1,264.02 π3 (π − π0 ) = 1,264.02 3 + (23°πΆ − 20°πΆ) π = π0 + π1 − π0 π 30°πΆ − 20°πΆ π = 1,262.241 ππ/π3 Fig 4. Procedure for Calculating Glycerol Temperature Tf via Linear Interpolation Given known values of the density of Glycerol for various temperatures from Table A-4 on Page A3-1 of the Manual, the density of the glycerol during the experiment can be calculated via linear interpolation as seen above. Using the recorded temperature of 23°C, we can estimate the corresponding density using known values for 20°C and 30°C. ππ = (ππ −ππ )ππ 2 18π∞ πππ ππβπππ 1: ππ ππ ππ = 3,990 π3 , ππ = 1,262.241 π3 , π = 9.81 π π , π = 3.173 π₯ 10−3 π, π∞ = 1.86 π₯ 10−2 2 π π ππ π (3990 − 1,262.241) ( 3 ) π₯ 9.81 2 π₯ (3.179 π₯ 10−3 π)2 πβπ π π ππ = = 0.804 −2 18 π₯ 1.86 π₯ 10 π/π π2 πππ ππβπππ 2: ππ = 8,000 ππ ,π π3 = 3.169 π₯ 10−3 π, π∞ = 4.63 π₯ 10−2 ππ ππ = πππ ππβπππ 3: π (8000−1,262.241)( 3 ) π₯ 9.81 2 π₯ (3.169 π₯ 10−3 π) π π 18 π₯ 4.63 π₯ 10−2 π/π π π 2 = 0.796 ππ ππ = 8,040 π3 , π = 1.583 π₯ 10−3 π, π∞ = 1.17 π₯ 10−2 πβπ π2 π π ππ π (8000 − 1,262.241) ( 3 ) π₯ 9.81 2 π₯ (1.583π₯ 10−3 π)2 πβπ π π ππ = = 0.793 18 π₯ 1.17 π₯ 10−2 π/π π2 Fig 5. Calculation of Dynamic Fluid Viscosity via Experimental Values The Dynamic viscosity of the Glycerol was calculated using Equation 1 and the respective experimental data collected for each sphere. Table 16. Dynamic Viscosity Calculated from the Stokes Equation Dynamic Viscosity, μ (N·s/m2) Number Designation 1 0.804 2 0.796 3 0.793 Table 16. A comprehensive listing of the Dynamic Viscosity of the Glycerol calculated using the procedure in Fig 5 above. πΆπππ ππππ π(π) = 5.06710 − 0.28266π + 5.4336 π₯ 10−3 π 2 − 3.30493π₯10−5 π 3 πππ π = 23β, π(π) = 5.06710 − 0.28266π₯(23β) + 5.4336 π₯ 10−3 π₯(23β)2 − 3.30493π₯10−5 π₯(23β)3 πβπ = 1.04 π2 πππ π = 24β, π(π) = 5.06710 − 0.28266π₯(24β) + 5.4336 π₯ 10−3 π₯(24β)2 − 3.30493π₯10−5 π₯(24β)3 = 0.96 πβπ π2 Fig 6. Procedure for determining Dynamic Viscosity as a Function of Temperature as Per Equation 9 An illustration of the steps used to calculate the predicted Dynamic Viscosity of Glycerol as a Function of Temperature considering the temperature of the Glycerol both before and after the experiment was conducted. Table 17. Dynamic Viscosity Calculated as a Function of Temperature Measurement Number Temperature, T (°C) Dynamic Viscosity, μ (N·s/m2) 1 2 23 24 1.04 0.96 Table 17. A comprehensive collection of the Dynamic Viscosities of the Glycerol calculated in Fig 6. using Eq. 9 πππ ππβπππ 1 ππ = 1,262.241 ππ π πβπ , π∞ = 1.86 π₯ 10−2 , π = 3.173 π₯ 10−3 π, ππ = 0.804 2 , πΆπ€ = 0.85 3 π π π ππ −2 π −3 ππ πΆπ€ π∞ π (1,262.241 π3 ) π₯ (0.85) π₯ (1.86 π₯ 10 π )(3.173 π₯ 10 ) π ππ‘ = = = 2.50 π₯ 10−1 πβπ ππ . 804 2 π 3 −2 π −3 ππ π∞ π (1,262.241 ππ/π )π₯ (1.86 π₯ 10 π )(3.173 π₯ 10 ) π π∞ = = = 2.93 π₯ 10−1 πβπ ππ 0.804 2 π Fig 7. Calculation of Respective Reynolds Numbers for Sphere 1 An illustration of the procedure used to calculate the Reynolds number for the first sphere using the shown variables calculated from experimental results. πππ ππβπππ 2, ππ = 1,262.241 ππ π πβπ , π∞ = 4.63 π₯ 10−2 , π = 3.169 π₯ 10−3 π, ππ = 0.796 2 , πΆπ€ = 0.85 3 π π π ππ −2 π −3 ππ πΆπ€ π∞ π (1,262.241 π3 ) π₯ (0.85) π₯ (4.63 π₯ 10 π )(3.169 π₯ 10 ) π ππ‘ = = = 1.26 πβπ ππ 0.796 2 π ππ −2 π −3 ππ π∞ π (1,262.241 π3 ) π₯ (4.63 π₯ 10 π )(3.169 π₯ 10 ) π π∞ = = = 1.47 πβπ ππ 0.796 2 π Fig 8. Calculation of Respective Reynolds Numbers for Sphere 2 An illustration of the procedure used to calculate the Reynolds number for the second sphere using the shown variables calculated from experimental results. πππ ππβπππ 3, ππ = 1,262.241 ππ π , π∞ = 1. .17 π₯ 10−2 , π = 1.583 π₯ 10−3 π, 3 π π ππ = 0.793 πβπ , πΆ = 0.92 π2 π€ ππ −2 π −3 ππ πΆπ€ π∞ π (1,262.241 π3 ) π₯ (0.92) π₯ (1.17 π₯ 10 π )(1.583 π₯ 10 ) π ππ‘ = = = 1.73 π₯ 10−1 πβπ ππ 0.793 2 π ππ −2 π −3 ππ π∞ π (1,262.241 π3 ) π₯ (1.17 π₯ 10 π )(1.583 π₯ 10 ) π π∞ = = = 1.87 π₯ 10−1 πβπ ππ 0.793 2 π Fig 9. Calculation of Respective Reynolds Numbers for Sphere 2 An illustration of the procedure used to calculate the Reynolds number for the second sphere using the shown variables calculated from experimental results. Table 18. Reynolds Number Calculations Number Designation Reynolds Number, Ret Reynolds Number, Re∞ 1 2.50E-01 2.93E-01 2 1.26 1.47E 3 1.73E-01 1.87E-01 Table 18. A comprehensive listing of the calculated Reynolds numbers in Figures 7, 8, and 9. 4 πβπ π·πππ πΉππππ ππ π‘βπ ππβπππ πΉπ· = ππ 3 π(ππ − ππ ) 3 πππ ππβπππ 1, π = π 2 = 3.173 π₯ 10−3 2 ππ ππ = 1,262.241 π3 , π∞ = 1.86 π₯ 10−2 πΉπ· = π ππ = 1.5865 π₯ 10−3 π, π = 9.81 π 2 , ππ = 3,990 π3 , π π 4 π ππ π₯ (π) π₯ (1.5865 π₯ 10−3 )3 π₯ (9.81 2 ) π₯ (3,990 − 1,262.241) 3 3 π π = 4.48 π₯ 10−4 π πβπ π·πππ πΆππππππππππ‘ ππ π‘βπ ππβπππ πΆπ· = 2πΉπ· ππ π∞ 2 ππ 2 = 2 π₯ (4.48 π₯ 10−4 π) ππ π (1262.241 3 ) π₯ (1.86 π₯ 10−2 π ) π₯ π π₯ (1.5865 π₯ 10−3 )2 π = 258.91 Fig 10. Calculation of the Drag Force and Drag Coefficient of Sphere 1 An illustration of the procedure used to calculate the drag force and drag coefficient of sphere 1 as per Eq. 5 and 6. 4 πβπ π·πππ πΉππππ ππ π‘βπ ππβπππ πΉπ· = ππ 3 π(ππ − ππ ) 3 πππ ππβπππ 2, π = π 2 = 3.169π₯ 10−3 2 ππ ππ = 1,262.241 π3 , π∞ = 4.63 π₯ 10−2 πΉπ· = π ππ = 1.5845 π₯ 10−3 π, π = 9.81 π 2 , ππ = 8,000 π3 , π π 4 π ππ π₯ (π) π₯ (1.5845 π₯ 10−3 )3 π₯ (9.81 2 ) π₯ (8,000 − 1,262.241) 3 3 π π = 1.10 π₯ 10−3 π πβπ π·πππ πΆππππππππππ‘ ππ π‘βπ ππβπππ πΆπ· = 2πΉπ· ππ π∞ 2 ππ 2 = 2 π₯ (1.10 π₯ 10−4 π) ππ π (1262.241 3 ) π₯ (4.63 π₯ 10−2 π ) π₯ π π₯ (1.5845 π₯ 10−3 )2 π = 103.17 Fig 11. Calculation of the Drag Force and Drag Coefficient of Sphere 2 An illustration of the procedure used to calculate the drag force and drag coefficient of sphere 2 as per Eq. 5 and 6. 4 πβπ π·πππ πΉππππ ππ π‘βπ ππβπππ πΉπ· = ππ 3 π(ππ − ππ ) 3 πππ ππβπππ 1, π = π 2 = 1.583π₯ 10−3 2 ππ ππ = 1,262.241 π3 , π∞ = 1.17 π₯ 10−2 πΉπ· = π ππ = 0.7915 π₯ 10−3 π, π = 9.81 π 2 , ππ = 8,040 π3 , π π 4 π ππ π₯ (π) π₯ (0.7915 π₯ 10−3 )3 π₯ (9.81 2 ) π₯ (8,040 − 1,262.241) 3 3 π π = 1.38 π₯ 10−3 π πβπ π·πππ πΆππππππππππ‘ ππ π‘βπ ππβπππ πΆπ· = 2πΉπ· ππ π∞ 2 ππ 2 = 2 π₯ (1.38 π₯ 10−4 π) ππ π (1262.241 3 ) π₯ (1.17 π₯ 10−2 π ) π₯ π π₯ (0.7915 π₯ 10−3 )2 π = 816.02 Fig 12. Calculation of the Drag Force and Drag Coefficient of Sphere 3 An illustration of the procedure used to calculate the drag force and drag coefficient of sphere 3 as per Eq. 5 and 6. Table 19. Drag Force Coefficient Calculations Number Designation Drag Force, FD (N) Drag Coefficient, CD 1 4.48E-04 258.91 2 1.10E-03 107.12 3 1.38E-04 816.02 Table 19. A Comprehensive listing of the respective drag forces and coefficients calculated for each sphere as per Figs. 10, 11, and 12. Table 20. Comparison of the Drag Coefficient and Reynolds Number Number Designation Reynolds Number, Re∞ 1 2 3 2.93E-01 1.47E 1.87E-01 Drag Coefficient, CD 259.91 103.20 816.02 Table 20. A Comprehensive comparison of the Calculated values for the Reynolds Numbers, Re∞ given in Table 18 and the Drag Coefficients CD given in Table 19 for each respective sphere. Drag Coefficient vs. Reynolds Number 100000 Sphere 3 Sphere 2 10000 Drag Coefficient, CD Sphere 1 1000 100 10 1.00E-03 1.00E-02 1.00E-01 1 1.00E+00 Reynolds Number, Re Fig 13. Log-Log Plot of Drag Coefficient CD vs. Reynolds Number Re∞ for respective spheres The Calculated Drag Coefficients plotted against the calculated Reynolds Numbers for each sphere on a Log-Log Plot. DISCUSSION This experiment compares obtained experimentally using the Stokes Equation formulation in Eqn 1 with those predicted from the best-fit polynomial curve obtained from Eqn 9, where is reported as a function of temperature T. Given that Eqn 1 was derived for the case of a sphere falling in an infinite wide medium, a correction factor from Eqn 2. is used to adapt experimental velocities in the tank to the infinite width velocities needed in Eqn 1. 3 different can be obtained, one for each sphere, each with its own density and diameter. These can then be used to calculate drag forces on each sphere, which are then plotted on a loglog plot against to generalize a curve that can predict either quantity for any sphere of given density and diameter, as long as < 0.1. is met, which is the second objective of this experiment. Looking at Table 1, which lists the parameters for each sphere, two comparisons can be made: (1) the large steel sphere vs. the small steel sphere and (2) the large red plastic sphere vs. the large steel sphere. For (1), since both are steel the densities are roughly equivalent while the diameter is roughly halved going from large to small. The effect of halving the diameter while holding the density constant can be seen in the effect on the average terminal velocities reported in Table 15 for each sphere. In this case, the effect is that halving the diameter while holding the density constant results in a nearly 4x drop (from 4.63E-03 to 1.17E-03 or 3.96x) in velocity, strongly illustrating that velocity is related to the square of “d”, the diameter, which confirms one aspect of the Stokes Equation formulation in Eqn 1. For (2) both have roughly the same diameter but the density of plastic is roughly half that of steel. Referring back to Table 15, the effect is that halving the density while holding the diameter constant results in a nearly 2.5x drop (from 4.63E-03 to 1.86E-03, or 2.48x) in velocity, which is not perfectly linear but far from quadratic in terms of the relation between velocity and density, again strongly confirming the validity of Eqn 1. From this simple analysis of terminal velocities, it appears that the relationships between quantities in Eqn 1 corresponds to experimental results. Once the constant parameters and data for terminal velocities from Tables 6 through 15 are inputted into Eqn 1, average experimental dynamic viscosities for each sphere can be obtained which are tabulated in Table 16. The red plastic sphere had a dynamic viscosity of 0.804 N·s/m2 the large steel sphere had 0.796 N·s/m2, and the small steel sphere had 0.793 N·s/m2. Since the terminal velocities used were all averaged from 2 segments and 5 trials of data, these values have a lot of data backing them up and appear fairly consistent around 0.800 N·s/m2for the experimental dynamic viscosity, across all 3 spheres. Comparing these with the dynamic viscosity calculated from Eqn 9, where the viscosity predicted at the starting room temperature 23β was 1.04 N·s/m2 and at the final room temperature at 24 β was 0.96 N·s/m2. This set of values is also fairly consistent around 1.00 N·s/m2, but most likely because the rise in temperature was only 1 degree. That being said, there is a roughly 20% difference between the experimental dynamic viscosities calculated using Eqn 1 and those calculated from the Fig 3 Eqn in the reference, that needs to be addressed. Discussing minor experimental factors first, looking at Eqn 1 it can be seen that there is an inverse dependence on . This quantity is obtained from using a correction factor for finite width. It may be that the result of averaging velocities from Segment 2 and Segment 3 in the tube resulted in an average velocity higher than the true terminal velocity which may decrease the value of . However, a more plausible experimental issue that also lends to great uncertainty is the use of the correction factor. The correction factor accounts for differences in and with infinite width assumptions. In most trials, the spheres did not fall in a perfectly linear path collinear with the center of the tube. The sphere often descended at a slant approaching the wall of the tube and it’s reasonable to think that coming closer to the wall than expected may result in wall effects that are greater than expected for dropping the sphere in a tube of this width. Qualitatively, the fluid lines around the sphere can abruptly intersect with the wall surface and disrupt movement of the sphere, if it fails to fall down the center. It is not clear if the correction factor used would take into consideration extra wall effects that arise from the ball not falling down the center. In such a case, the correction factor would have to be closer to 1, which yields a smaller and a higher dynamic viscosity. Putting experimental issues aside, Eqn 1 allows us to compute dynamic viscosities in terms of our specific experimental parameters. This is in the spirit of what is known as a Falling Ball Viscometer. Eqn 9 computes dynamic viscosities as a function of temperature, a relation derived from using a “Falling Needle Viscometer” type setup, which tries to avoid some uncertainties of sphere diameter, density, and finite medium problems. The slight difference in the use of needles or thin cylinders as the dropped object over spheres lends a greater degree of accuracy for Eqn 9 than our setup which is more affected by these uncertainties. Therefore, the latter equation may likely yield the better absolute dynamic viscosity values for the fluid, however, with regards to the spheres dropped in our experiment that follow slanted trajectories and such, the experimental dynamic viscosities we obtain are still consistent. Lastly, examining the log-log plot in Fig 13, which plots the log of the against the log of the , the three spheres fall on the linear portion of the graph, with the large steel sphere (Sphere 2) approaching a region that drops sharply. With the experimental data at present, it isn’t clear exactly what combination of and will mark the transition into the turbulent region, however, this curve can be used to interpolate for a variety of sphere diameters and densities that fall between the range of the large steel sphere (Sphere 2) and the small steel sphere. (Sphere 3) Further sphere diameter and density combinations can be used to mark the point where the log-log plot no longer stays linear, at which point our founding equations need to be re-examined. This can also be verified using the constraint on Eqn 1, which is given in Eqn 3 where < 0.1. ERROR ANALYSIS 1. Calculate the uncertainties of measured times: Use data from Table sphere 1 for segment 2 as an example: Governing equation: βx = t π£,π ∗ ππ √π = 2.77 ∗ 0.10035 √5 = 0.1243 We only did only five trials in measuring the time, which is less than 60 trials, so we use values of Student Distribution table in lab manual in order to get the value of βt π£,π . After referring to Table 1, we decided to use 95% as our confidence intervals, besides v =N-1=5-1=4, so t π£,π = 2.77. Sx is value of standard distribution. Example answer for π‘ = 15.42 ± 0.1243(95%) sphere 1 for segment 2 sphere 2 for segment 2 sphere 3 for segment 2 Trial Time, t (sec) Trial Time, t (sec) Trial Time, t (sec) 1 15.47 1 6.38 1 22.94 2 15.42 2 6.2 2 22.97 3 15.47 3 6.33 3 22.66 4 15.50 4 6.22 4 22.72 5 15.25 5 6.21 5 22.35 Average 15.42 Average 6.268 Average 22.73 Sx 0.10035 Sx 0.08167 Sx 0.25054 tv,p 2.77 tv,p 2.77 tv,p 2.77 βx 0.12431174 βx 0.101171298 βx 0.310364357 sphere 1 for segment 3 Trial Time, t (sec) sphere 1 for segment 3 Trial Time, t (sec) sphere 1 for segment 3 Trial Time, t (sec) 1 16.78 1 6.81 1 24.99 2 16.50 2 6.72 2 24.99 3 16.66 3 6.62 3 24.62 4 16.47 4 6.55 4 24.39 5 16.50 5 6.61 5 24.01 Average 16.58 Average 6.66 Average 24.60 Sx 0.13349 Sx 0.10281 Sx 0.10281 tv,p 2.77 tv,p 2.77 tv,p 2.77 βx 0.165364964 βx 0.127359142 βx 0.127359142 Table 1: This table lists the experimental data and their corresponding values which are used to calculate uncertainties for time such as average values, standard deviation values, tv,p and uncertainties for time t for sphere 1, 2 and 3 respectively. 2. Calculate the uncertainties for Ut Governing equation: ππ‘ = βππ‘ = √( π π‘ 2 2 πππ‘ πππ‘ π βπ) + ( βπ‘) = √0 + (− 2 βπ‘)2 ππ ππ‘ π‘ Use sphere 1 for segment 1 as an example ππ 2 ππ 2 π 25.4 βππ‘ = √( πππ‘ βπ) + ( ππ‘π‘ βπ‘) = √0 + (− π‘ 2 βπ‘)2 = √(− 15.422 ∗ 0.1243)2 = 0.1534 sphere1 mean(mm/s) βx(mm/s) sphere2 mean(mm/s) βx (mm/s) v (seg1) 16.5 0.1534 v(seg1) 40.7 0.7317 v(seg2) 15.2 0.1516 v(seg2) 38.3 0.7231 Ut 15.85 0.1525 Ut 39.5 0.7274 sphere 3 mean(mm/s) βx (mm/s) v(seg1) 11.2 0.0537 v(seg2) 10.4 0.0530 Ut 10.8 0.0534 Table 2: This table lists the data that is used to calculate uncertainties for Ut for sphere 1, 2 and 3 respectively. 3.Calculate the uncertainties for π∞ Governing equations: π∞ = ππ‘ π (1 − )9/4 π· a. According to the equation above which is derived by using propagation, we need to find out the value of βπ· Use data from Table sphere 1 for segment 2 as an example: Governing equation: βD = t π£,π ∗ ππ √π = 3.182 ∗ 0.30561 √4 = 0.4862 We only did only four trials in measuring the time, which is less than 60 trials, so we use values of Student Distribution table in lab manual in order to get the value of βt π£,π . After referring to Table 1, we decided to use 95% as our confidence intervals, besides v =N-1=4-1=1, so t π£,π = 3.182. Sx ,which is the value of standard distribution, equals to 0.30561 Example answer for π· = 46.34 ± 0.4862(95%) Measurement Number Diameter, D (mm) 1 46.01 2 46.6 3 46.6 4 46.15 Average 46.34 Standard deviation(Sx) 0.30561 tv,p 3.182 βx 0.48622551 Table 3: This table lists the data that is used to calculate uncertainties for tube diameter Dt b. Calculate uncertainties βπ∞ when βπ, βπ·, βUπ‘ is known, use sphere 1 as an example 2 2 2 ππ∞ ππ∞ ππ∞ βπ∞ = √( βπ) + ( βπ·) + ( βππ‘ ) ππ ππ· πππ‘ 13 13 9 π −4 1 9 π −4 1 (π ∗ π·−2 )βπ·]2 + [ = √[− ππ‘ (1 − ) (− ) βπ]2 + [− ππ‘ (1 − ) βπ ]2 π 9 π‘ 4 π· π· 4 π· (1 − )4 π· 13 13 9 3.173 − 4 1 9 3.173 − 4 1 2 (3.173 ∗ 46.34−2 )β46.34]2 + [ = √[− 815.85 (1 − ) (− ) 0.4862]2 + [− 15.85 (1 − ) 9 0.1524] 4 46.34 46.34 4 46.34 3.173 4 (1 − ) 46.34 = 0.3572 sphere 1 sphere 2 mean βx mean βx d (mm) 3.173 0.03173 d (mm) 3.173 0.03173 D(mm) 46.34 0.48622551 D(mm) 46.34 0.48622551 Ut(mm/s) 15.85 0.152457096 Ut(mm/s) 39.5 0.727442772 U∞(mm/s) 18.59 0.3572 U∞(mm/s) 46.33 0.8598 sphere 3 mean βx d (mm) 1.583 0.01583 D(mm) 46.34 0.48622551 Ut(mm/s) 10.8 0.053350424 U∞(mm/s) 11.67 0.0592 Table 4: This table lists the data that is used to calculate uncertainties for U∞ for sphere 1, 2 and 3 respectively. 4. Calculate the uncertainties for π Governing equations: π= (ππ − ππ )ππ2 18π∞ Use sphere 1 as an example: 2 ππ’ βππ ) πππ βπ’ = √( 2 ππ’ βπ) ππ +( (π −π )π2π ππ 2 βππ )2 + 0 + [ π π βπ]2 18π∞ 18π∞ (3990−1262.24)∗9.81∗2∗0.003173 √( 0+[ 18∗0.01859 2 ππ’ βπ) ππ +( ππ’ βπ∞ )2 ππ∞ +( (ππ −ππ )ππ 2 +[ 18 9.81∗0.0031732 18∗0.01859 (3990−1262.24)9.81∗0.0031732 (−π∞ −2 ) ∗ βπ∞ ]2 = {( ∗ 0.00003173]2 + [ 0.0003572]2 }^0.5=0.028 Sphere 1 = mean βx U∞ 0.01859 0.0003572 d (m) 0.003173 0.00003173 ρ(kg/m3) 3990 79.8 ρf(kg/m3) 1262.24 0 μ(N*s/m2) 0.805125187 0.0285 18 ∗ 79.8)2 + (−0.01859−2 ) ∗ Sphere 2 mean βx U∞ 0.04633 0.0008598 d (m) 0.003169 0.00003169 ρ(kg/m3) 8000 160 ρf(kg/m3) 1262.24 0 μ(N*s/m2) 0.795965451 0.0247 mean βx U∞ 0.01167 0.0000592 d (m) 0.001583 0.00001583 ρ(kg/m3) 8040 160.8 ρf(kg/m3) 1262.24 0 μ(N*s/m2) 0.793183484 0.02461 Sphere 3 Table 5: This table lists the data that is used to calculate uncertainties for dynamic viscosity of fluid μ for sphere 1, sphere 2, sphere 3 repectively. CONCLUSION This experiment illustrates the validity of Eqn 1, where the is stated to be linearly related to the difference between the fluid and sphere densities, and quadratically related to “d”, the diameter of the sphere. This is confirmed earlier by comparing the two steel spheres of roughly similar densities and with one diameter roughly half of the other, which made the correlation of with “d” or the diameter clear, while the large red plastic sphere and the large steel sphere had their densities, one exactly half of the other, which established the correlation of with density. Also, the Stokes Equation formulation for for a sphere in Eqn 1 yielded results closely averaging around 0.8 N·s/m2, which correlated rather well with the predicted dynamic viscosity using Eqn 9, which predicted to be around 1.0 N·s/m2. The 20% is expected for such fluid-type experiments, however, certain sources of uncertainty and deviance were offered. Namely, the effect of averaging velocities from Segment 2 and 3 of the tubes, both of which may not truly reflect the terminal velocity. There is also the issue of the spheres traveling slanted paths closer to the wall and the effect that might have on the correction factor used, which takes into account the finite dimensions of the tube and possibly the wall shear resulting from it but doesn’t account for the ball traveling off-center. Finally, there is the fact that Eqn 9 was specifically developed in a separate experiment that involve adopting a “Falling Needle Viscometer” type setup rather than a “Falling Sphere Viscometer” setup like our experiment, for the purposes of not developing inaccuracy from uncertainties in sphere diameter or density. As a result, the predicted dynamic viscosities for the fluid in Eqn 9 are likely truer values, however, the experimental dynamic viscosities calculated using Eqn 1 do make sense when considering the slanted path the spheres in this experiment took and the other uncertainties that our setup was not free of. With regards to the latter objective of this experiment, a fairly linear log-log plot was established for log of the against the log of the , and this means our experimental results can be used to predict both quantities for a given sphere diameter and density, assuming that this will satisfy the condition of < 0.1 which this experiment required and that the log of and log point falls within the laminar region of the curve that was developed in this experiment. Interpolating in this linear region will yield valid results but outside of this region could be a transition to the turbulent region, where predictions from the current data cannot be made. This experiment would benefit from the use of a slender cylinder model as was done in the paper that established Eqn 9, as this type of an object would be more likely to follow a more linear trajectory down center and avoid issues of different apparent tube “D” and inaccuracies in velocities calculated, since slanted paths would mean the true distance traveled in a segment would not simply be the length of the segment. Although the drag coefficient for such a model may be higher, if a Stokes equation-type formulation exists, it should yield consistent dynamic viscosity values, and will likely match the values predicted from Eqn 9. REFERENCES 1. Pritchard, Philip J, Fox and Mcdonald’s Introduction to Fluid Mechanics, 8th Ed. Wiley, October 2011. P38-40 2. Yunus A. Cengel and Afshin J. Ghajar, Heat and Mass Transfer Fundamentls and Applications 4th Ed., McGraw-Hill 2011, p.373-385 3. Igor N, Serdyuk, Nathan R. Zaccai, and Joseph Zaccai, Methods in Molecular Biophysics: Structure, Dynamics, and Function. Cambridge Unversity Press, March 29th 2007. P255-256 4. Griggs, Andrew J., The Low-Reynolds-number motion of Drops and Bubbles Near Solid Planar Surfaces. Proquest Information and Learning Company, 2008. P102-104 5. Park, N.A. and T.F. Irvine, Jr., “The Falling Needle Viscometer: A New Technique for Viscosity Measurements,” Warme-und Stoffubertragund, Vol. 18, pp. 201-206 (1983).