DECLARATION I hereby declare that this thesis is my original work and it has been written by me in its entirety I have duly acknowledged all the sources of information which have been used in the thesis This thesis has also not been submitted for any degree in any university previously _ Koh Fu Hai Alan July 2013 I ACKNOWLEDGEMENTS The author wishes to express appreciation to the late Associate Professor Luo Siao Chung and Professor Chew Yong Tian for their valuable guidance as supervisors Many thanks also go to the staff of the NUS fluid mechanics laboratory and workshop, especially to Mr James Ng for machining the small parts of the model and Mr Looi Siew Wah for fixing the old computer for the load cell II TABLES OF CONTENTS SUMMARY VII LIST OF TABLES VIII LIST OF FIGURES IX LIST OF SYMBOLS XII INTRODUCTION 1.1 Mechanism of Ground Effect 1.2 History 1.3 Literature Review 1.3.1 Analytical Methods 1.3.1.1 Flat Ground, Image Method, 3D 1.3.1.2 Wavy Ground, Image Method, 3D 13 1.3.2 Computational Methods 14 1.3.2.1 RANS, Moving Ground, 2D 14 1.3.2.2 Vortex Lattice Method, 3D 15 1.3.3 Experimental Methods 18 1.3.3.1 Image Method, 3D 18 1.3.3.2 Flat Plate, 3D 23 1.3.3.3 Flat Plate, 2D, Pressure Distribution 27 1.3.3.4 Moving Ground, 3D, Pressure Distribution 34 1.4 Objectives 38 1.5 Scope and Organisation of the Thesis 39 EXPERIMENTAL DETAILS 2.1 Wind Tunnel 40 2.1.1 Velocity Profile in the y-z plane 41 2.2 Flat Plate 42 2.3 Caution in Simulation of the Ground in Ground Effect Experiments 43 III 2.4 Definition for h/c 47 2.5 Investigated range of α and h/c 47 2.6 Temperature, Dynamic Pressure and Reynolds Number 47 2.7 Model 48 2.8 Measurement of lift 50 2.9 Measurement of drag 50 2.10 Measurement of pressure 52 2.11 Load Cell 52 2.12 Manometer 54 2.13 Challenges in Measurement 55 2.13.1 Drift in Load Cell Datum 55 2.13.2 Downstream Pressure Gradient 55 2.13.3 Blockage Effects 56 RESULTS 3.1 Pressure Distribution 57 3.1.1 Reynolds number Effect on Aerodynamic Coefficients 57 3.1.2 Aerodynamic Coefficients at Reynolds number of 200000 60 3.1.3 Pressure Distribution at α = 0O and 2O 67 3.1.4 Pressure Distribution at α = 4O to 8O 71 3.2 Lift and Drag 77 3.2.1 Datum 3D Lift and Drag 78 3.2.2 Lift and Drag coefficients from Pressure Distribution 80 3.2.3 Effect of α on Lift and Drag coefficients vs h/c 81 3.2.4 Effect of h/c on Lift and Drag coefficient vs α 84 3.3 Uncertainty 86 DISCUSSION 4.1 Effective camber 92 IV 4.2 Effective angle of attack 93 4.3 Behaviour of Air below the wing 94 4.4 Ground effect on circulation on a wing 95 4.5 Drag 96 4.6 Changes in Lift and Drag for a 2D airfoil in Ground Effect 97 4.6.1 Effects of a Flat Moving Ground on lift and drag 98 4.6.2 Effects of a Flat Stationary Ground on lift and drag 99 4.7 Changes in Lift and Drag for a Finite Wing in Ground Effect 100 CONCLUSION AND RECOMMENDATION 5.1 Concluding Remarks 101 5.2 Recommendations 102 BIBLIOGRAPHY 104 APPENDICES A Summary of WIG Vehicles 108 B Summary of Experimental data from NACA TN 67 117 C Summary of Experimental Data from NASA TN D926 125 D Summary of Experimental Data from Chawla, Edwards and Franke 135 E Summary of Experimental Data from Ahmed and Sharma 138 F Summary of Experimental Data from Kwang, Ho and Hee 142 G Summary of Experimental Data from Ahmed 145 H Summary of Experimental Data from Ahmed, Takasaki and Kohama149 I Summary of Pressure Distribution over NACA 4415 section at various Reynolds number 153 J Calculated Lift and Drag of finite NACA 4415 wing 165 K Load cell Characteristics 170 L Drawings of the Model 192 M Downstream Pressure Profile 196 N Y-Z Plane Velocity Profile 200 V O Wind Speed and Effect of Tube Length 208 P Approximating the Leading Edge Pressure Reading from Pitot Reading 209 Q Summary of NACA 4415 Aerodynamic Data Pressure 210 R Summary of NACA 4415 Aerodynamic Data Force 220 S Uncertainty in Force Measurements 227 T Preparation of Wind Tunnel Test Section 229 U Preparation of Model 230 V Preparation of Other Equipment 231 W Formulas 232 X Corrections to Data 236 VI SUMMARY Ground effect is the increase of lift and reduction of drag, on a lift generating body in flight at heights of one chord or less above the ground This beneficial effect has been harnessed for development of ground effect vehicles since in the early 1900s To date, large-scale commercial production of ground effect vehicles has not taken place The literatures consists of investigations reporting ground effect with increased lift and reduced drag, but not discuss the mechanisms behind changes in lift and drag Most ground effect research was carried out with the image method or the flat plate method to simulate the ground Only one recent work used the moving belt method to simulate the ground Due to limited laboratory resources in this study, wind tunnel experiments were carried out with a flat plate simulating the ground Pressure distributions, lift and drag forces were obtained from a finite wing of NACA 4415 section with aspect ratio 2.51 The experiments were carried out at positive angles of attack from 0O to 8O in ground effect As the wing approached the ground, suction reduced on the upper surface when in ground effect than without ground effect and air was slowing down under the wing The resulting higher pressure developing under the wing was the main cause of improved lift at all angles of attack as ground clearance decreased Drag reduction was mainly due to suppressed formation of wing tip vortices, which reduced the induced drag component The circulatory lift contribution decreased slightly due possibly to smaller effective camber as streamlines straightened out about the wing However, without a moving belt facility for ground simulation, the effect of the ground on circulation about the wing is still debatable Generally, ground effect over a flat plate, translated the CL-α curve upward and CD-α curve downward VII LIST OF TABLES Table 3.1 Comparing estimated cl to Abbott and Doenhoff data for similar α at various Reynolds number 63 VIII LIST OF FIGURES Figure 1.1 Changes in WIG vehicles’ Gross Weight Figure 1.2 Changes in WIG vehicles’ Cruising Speed Figure 1.3 Measured and calculated drag polar curves near the ground Figure 1.4 Gradient of Lift coefficient vs α for a 2D flat plate airfoil in ground effect 11 Figure 1.5 Variation of CL|α=0 and ∂CL/∂α with respect to h/b 12 Figure 1.6 Variation of CL with CDi, induced drag 13 Figure 1.7 Span wise loading of a planar wing (aspect ratio 4) at sweep angles 16 Figure 1.8 Effect of ground proximity on CL of rectangular wings 17 Figure 1.9 Lift to DragINDUCED ratios for rectangular wing of maximum 4% camber 17 Figure 1.10 CL vs α (varying h/c) for wing with USA 27 section 20 Figure 1.11 CD vs α (varying h/c) for wing with USA 27 section 20 Figure 1.12 L/D vs h/c (varying α) for wing with USA 27 section 21 Figure 1.13 CL vs α (varying h/c) for wing with modified Glenn Martin 21 section 22 Figure 1.14 CD vs α (varying h/c) for wing with modified Glenn Martin 21 section 22 Figure 1.15 L/D vs h/c (varying α) for wing with modified Glenn Martin 21 section 23 Figure 1.16 CL vs α (varying h/c) for swept wing with NACA 4415 section 24 Figure 1.17 CD vs α (varying h/c) for swept wing with NACA 4415 section 25 Figure 1.18 L/D vs h/c (varying α) for swept wing with NACA 4415 section 25 Figure 1.19 CL vs α (varying h/c) for wing with NACA 6409 section 26 Figure 1.20 CD vs α (varying h/c) for wing with NACA 6409 section 27 Figure 1.21 L/D vs h/c (varying α) for wing with NACA 6409 section 27 Figure 1.22 cl vs α (varying h/c) for airfoil with NACA 0015 section 28 Figure 1.23 cd vs α (varying h/c) for airfoil with NACA 0015 section 29 IX Appendix R Summary of NACA 4415 Aerodynamic Data Force CL vs h/c NACA 4415 | 3D | Re approx 200000 | at various Angles of Attack 1.00 0.90 0.80 CL 0.70 0.60 0.50 0.40 0.30 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 h/c CD vs h/c NACA 4415 | 3D | Re approx 200000 | at various Angles of Attack 0.09 0.08 0.07 CD 0.06 0.05 0.04 0.03 0.02 0.01 0.00 0.10 0.20 0.30 0.40 0.50 h/c 222 0.60 0.70 0.80 0.90 1.00 Appendix R Summary of NACA 4415 Aerodynamic Data Force L/D vs h/c NACA 4415 | 3D | Re approx 200000 | at various Angles of Attack 24 22 20 L/D 18 16 14 12 10 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 h/c CL vs AoA NACA 4415 | 3D | Re approx 200000 | at various Angles of Attack 1.00 0.90 1.00 0.80 0.80 CL 0.70 0.60 y = 0.0518x + 0.2072 0.60 0.40 0.30 0.50 0.20 0.10 0.40 0.05 0.30 OGE 0.20 0.10 AoA [deg] 223 Appendix R Summary of NACA 4415 Aerodynamic Data Force CD vs AoA NACA 4415 | 3D | Re approx 200000 | at various Angles of Attack 0.09 0.08 1.00 0.07 0.80 0.60 CD 0.06 0.40 0.05 0.30 0.20 0.04 0.10 0.05 0.03 OGE 0.02 0.01 AoA [deg] L/D vs AoA NACA 4415 | 3D | Re approx 200000 | at various Angles of Attack 24 22 1.00 20 0.80 0.60 L/D 18 0.40 16 0.30 0.20 14 0.10 0.05 12 OGE 10 AoA [deg] 224 Appendix R Summary of NACA 4415 Aerodynamic Data Force % Increase in CL (points) and Relative Uncertainty (lines) vs h/c NACA 4415 | 3D | Re approx 200000 | at various Angles of Attack 16% 14% Percentage 12% 10% 8% 6% 4% 2% Relative Uncertainty in CL 0% 0.00 -2% 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 -4% h/c % Increase in CD (points) and Relative Uncertainty (lines) vs h/c NACA 4415 | 3D | Re approx 200000 | at various Angles of Attack 20% 15% 10% Percentage 5% 0% 0.00 -5% 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 -10% Relative Uncertainty in CD -15% -20% -25% h/c 225 Appendix R Summary of NACA 4415 Aerodynamic Data Force % Increase in L/D (points) and Relative Uncertainty (lines) vs h/c NACA 4415 | 3D | Re approx 200000 | at various Angles of Attack 40% 35% Percentage 30% 25% 20% 15% 10% Relative Uncertainty in L/D 5% 0% 0.00 -5% 0.10 0.20 0.30 0.40 0.50 0.60 0.70 -10% h/c 226 0.80 0.90 1.00 Appendix S Uncertainty in Force Measurements The following rules were use to determine the uncertainty in force measurements These rules were based on the Evaluation of Measurement data – Guide to the expression of Uncertainty in Measurement (GUM) [reference 36], which was based on the International Organisation of Standardisation Guide to the expression of uncertainty in Measurement 1995 edition For the quantity A, the uncertainties were represented in the following ways Absolute uncertainty - ∆Α Relative uncertainty - ε A = ∆Α Α For readings with the standard deviation available, the standard uncertainty was calculated from the standard deviation according the following equation, ∆Α = σ N where σ is the standard deviation and N is the number of samples For readings with the only tolerances available a ± a , the standard uncertainty was calculated from the tolerances according to the following equation, ∆Α = a n where a is the tolerance and n is determined by the distribution assumed for the samples For measurements where each reading had an equal probability of occurrence, which was a rectangular distribution, n = For measurements where readings had a higher probability of occurrence near the mean, which was a triangular distribution, n = In this work, n = as the rectangular distribution has gives a more conservative estimate of uncertainty According to the Laws of propagation of uncertainty in the GUM, the following rules were used For independent quantities, the following simplified rules apply Addition and Subtraction – absolute uncertainties added Using Χ = Α±Β Uncertainty of X is ∆Χ = ∆Α + ∆Β2 Multiplication and Division – relative uncertainties added Using 227 Appendix S Uncertainty in Force Measurements Χ= Α or Χ = ΑΒ Β Uncertainty of X is ∆Χ  ∆Α   ∆Β  =   +  Χ  Α   Β  Powers Using Χ = Αn Uncertainty of X is ∆Χ n∆A = Χ A The expanded uncertainty was obtained by the following, Expanded uncertainty of X = k∆Χ where k is the coverage factor The coverage factor is taken as for a significance level of 95% 228 Appendix T Preparation of Wind Tunnel Test Section The floor of the wind tunnel test section had several layers of old paint However, the surface was not smooth The entire floor was sanded down gradually with sandpaper on a hand-held buffing machine The grit reduced in stages from 180 to 280 to 400 to 600 to 800 to 1000 Eventually, an almost reflective surface finish was achieved on the remaining layers of paint The floor was made of three large wooden panels but they are not aligned till flat and flush at the lines of joining Since sanding the entire tunnel floor till it became flat was not feasible, the surfaces near the lines of joining were sanded down by a larger amount to blend one panel smoothly to another The cut out within the first half of the tunnel was made of Perspex and joined to the rest of the tunnel floor by wood filler The wood filler was sanded down to blend the adjacent surfaces too However this cut out was located a few mm off the centerline of the tunnel The floor edges along the side did not extend to the tunnel walls Layers of masking tape were applied to fill in the gap This wind tunnel test section wall comprised of four panels on each side Each panel composed of a large, approximately 10mm thick Perspex sheet mounted to a wooden frame by screws Since the wooden frame did not place the Perspex sheet flush to the adjacent panel's sheet, slips of paper were used as shims to bring the Perspex sheets approximately flush to each other The gap introduced between the Perspex sheet and wooden frame was covered by electrical insulation tape Once the panels were mounted into the tunnel walls, the remaining gap between panels and tunnel walls were sealed also by application of electrical insulation tape The tunnel ceiling comprised two large Perspex sheets secured by screws to the frame Both sheets were adjusted such that the joining line matched and both sheets blended one to the next Due to the size and stiffness of the panels and wooden frame, a few mis-aligned screw holes were not used, as the repair was too difficult Other holes were covered with transparent plastic sheets and taped down with scotch tape When operating the tunnel, the other side of each hole was covered to reduce the volume of trapped air behind the plastic sheet Less trapped air caused less bulging, when lower-pressure regions existed inside the tunnel during operation 229 Appendix U Preparation of Model To verify the sectional shape of the wing, a template was created from a 200 point CAD drawing A hundred discrete points defined the upper surface, with more points concentrated within x/c to 0.3 to more accurately define this highly curved portion of the upper surface The same process was used on the lower surface This template was transferred to paper and the paper cutouts were glued to the wing tips while re-shaping the model When the model was received, the section’s shape was oversized at various portions by at most mm The parts of the model were carefully sanded down to almost the profile of the paper cutout Next the gaps in the grains on the surface were filled in with dark wood filler and sanded down in preparation for spray painting Matt white paint was used to fill even smaller gaps on the model surface The spray painted surface was sanded down, ending with grit 1200 sandpaper to give the model a near reflective finish The simple hinges below the wing were replaced with two pieces of plastic with holes These attach to the model by screws fastened into the model Another price of plastic with internal threads was glued into the point of attachment with Araldite, a glue and filler The rest of the hinge mechanism was attached by means of 2mm diameter pins This hinge mechanism pivots the model on its trailing edge Angle of attack adjustments did not change the h/c value with this hinge mechanism The upper half was secured to the lower half by screws and the corresponding internal threads were initially cut into wood Since wooden threads cannot withstand repetitive fastening and unfastening, space was cut within the model to hold a block of plastic that holds the internal threads for the fastening screws 230 Appendix V Preparation of other Equipment Supports Below Test Section A support structure was assembled below the wind tunnel to hold the vertical traverse mechanism for the flat plate This structure was made from 40mm square profile aluminum bars The horizontal bars were aligned to the wind tunnel floor, while the vertical ones were aligned to the wind tunnel pillars To give this structure some stiffness, the vertical bars were secured to both the wind tunnel floor and the ground It was not practical to increase the stiffness by using heavier and stronger materials During wind tunnel operation, the suction fan motor transmitted a small obvious amount of vibration to the ground This vibration was felt on the wooden frame of the test section and the installed supports The vertical traverse mechanism consists of three individually adjustable car jacks that supported a horizontal rectangular frame In turn, four bars of NACA 0012 cross section connect the rectangular frame outside to the elevated ground inside the tunnel Since the model stayed put relative to the tunnel floor and ceiling, the elevated ground was adjusted vertically to the desired ground clearance between itself and the model Cross-section of Support Bars A NACA 0012 cross section was selected for the support bars to reduce their interference to the flow inside the wind tunnel The airfoil shape has a small amount of drag and caused no shedding vortices in steady flow Model Support In this work, the model came from a previous project and was supported from beneath The supports must be rigid to transmit drag, lift and moment to the load cell It must connect rigidly to the model and load cell Tubes from pressure taps were also routed through the support The support had to stay clear of the tunnel walls to give correct readings on the load cell The external shape was streamlined with NACA 0012 cross section, to eliminate separated flow from the supports The support surface was smoothened by sanding Simply taping holes have reduced the drag contribution by half The mechanism to set the angle of attack was also shielded within the support Supporting the model from below is not ideal as the supports possibly disrupted the flow below the model The flow accelerated locally around the supports Using a narrow symmetrical support with airfoil shape cross section mitigated this Alternative positions to mount the support were the upper surface from x/c 0.5 to 1.0 or sting support from the trailing edge Mounting at the top on the rear half of the model placed the support in a region of airflow that does not vary much as the ground approached Having the supports slanted and exiting the wind tunnel at a position aft of the model ensured that if excessive air leaked into the wind tunnel, it did not affect the flow about the model Manometer The bank of manometers was repaired to a state where all tubes responded relatively together to changing reservoir water level Each tube had a short section made of glass at the top All 40 glass tubes were clamped into a holder Due to age and the clamping, most tubes developed minute cracks Once these glass tubes and adjacent stiffened rubber tubes were painstakingly replaced, water level in all tubes rose and fell together 231 Appendix W Formulas The following formulas were used to determine the coefficients and their relative uncertainties For the quantity A, Absolute uncertainty - ∆Α Relative uncertainty - ε A = ∆Α Α Common variables Planform area Dimensions were measured with a vernier rule with the smallest division 0.05mm S = bc Uncertainty εS = (ε b )2 + (ε c )2 Density A simple quadratic curve was fitted to the water and air density values for the temperature range that the experiments were conducted in Temperature was recorded in degree Celsius and the coefficients were considered accurate Air density = 0.00001T2 - 0.0046T + 1.2913 Water density = -0.0047T2 - 0.0191T + 1000.5 ρ AIR = AAIRT + BAIRT + C AIR Uncertainty ∆ρ AIR ∆T   2∆T   =  AAIR T  +  BAIRT  T   T   The uncertainty for water density was determined in this similar way Dynamic Pressure q= ρ AIRU = PSTAGNATION − PSTATIC = hρWATER g where h was the difference in water level heights read on the manometer Uncertainty ∆q  ∆h   ∆ρ =   +  WATER q  h   ρWATER    and h = hSTATIC − hPITOT 2 ∆h = ∆hSTATIC + ∆hPITOT 2 ∆hSTATIC + ∆hPITOT ∆h = h hSTATIC − hPITOT where the smallest division was 1mm on the manometer Force measurements The absolute uncertainty in force measurements was defined by 232 Appendix W Formulas σ ∆F = N where F-force, σ-standard deviation and N-number of samples Drag Since a small drift occurred when establishing the drag force’s baseline and this drift was reasonably assumed linear with time when the time interval was small, the following formula was used DBL = DBL1 + t m − t1 (DBL − DBL1 ) t − t1 where D-drag, t-time, BL-baseline, m-measurement, 1-before, 2-after Uncertainty εD − D BL1 BL εt m εt t − t  ∆  m (DBL − DBL1 ) =  t2 − t1  2 ∆DBL + ∆DBL1 = DBL − DBL1 (ε − t1 − t1 ∆tm2 + ∆tl2 = tm − t1 ∆t22 + ∆t12 = t2 − t1 ) + (ε ) + (ε ) D BL − D BL1 BL1 ∆DBL = ∆D t m − t1 t − t1 tm − t1 (DBL − DBL1 ) t2 − t1 t − t  + ∆  m (DBL − DBL1 )  t2 − t1  Drag experienced by the wing was given by D = Dm − DBL − DS where S-supports Uncertainty ∆D = ∆Dm2 + ∆DBL + ∆DS2 where ∆DS was mean absolute uncertainty for drag on supports Each ∆DS reading was obtained in the same way as ∆DBL Since the load cell gave a slightly different value of Drag force, the corrected drag force is given by DCORRECTED = AD D − DBuoyancy where AD is the correction factor However, since the load cell correction factor is almost 1.0, the load cell was considered to give an accurate reading directly No correction factor is needed for the drag force Buoyancy correction was applied after load cell correction To simplify the analysis, the buoyancy correction was assumed to have no uncertainty As the buoyancy correction is a small quantity compared to the measured drag 233 Appendix W Formulas The drag force was then DCORRECTED = D − DBuoyancy Uncertainty ∆DCORRECTED = ∆D Coefficient of drag DCORRECTED DCORRECTED = qS ρU S CD = Uncertainty ε C = ε D2 D CORRECTED + ε q2 + ε S2 Lift Since a small drift occurred when establishing the lift force’s baseline and this drift was reasonably assumed linear with time when time between readings was small, the following formula was used LBL = LBL1 + tm − t1 (LBL − LBL1 ) t2 − t1 where L-drag, t-time, BL-baseline, m-measurement, 1-before, 2-after Uncertainty εL BL − L BL1 εt m εt t − t  ∆  m (LBL − LBL1 ) =  t2 − t1  (ε ∆L2BL + ∆L2BL1 LBL − LBL1 = − t1 = − t1 = ∆tm2 + ∆tl2 tm − t1 ∆t22 + ∆tl2 t2 − t1 ) + (ε ) + (ε ) LBL − L BL1 BL1 ∆LBL = ∆L t m − t1 t − t1 t − t  +  m (LBL − LBL1 )  t2 − t1  Lift experienced by the wing is given by L = Lm − LBL Uncertainty ∆L = ∆L2m + ∆L2BL 234 tm − t1 (LBL − LBL1 ) t2 − t1 Appendix W Formulas Since the load cell gave a slightly different value of lift force, the corrected drag force is given by LCORRECTED = AL L where AL is the correction factor Since this correction factor is close to 1.0, the load cell’s output can be considered accurate and no correction factor needed The corrected Lift force was then LCORRECTED = L Uncertainty ∆LCORRECTED = ∆L Coefficient of lift CL = LCORRECTED LCORRECTED = qS ρU S Uncertainty ε C = ε L2 L CORRECTED + ε q2 + ε S2 Lift / Drag Since aerodynamic efficiency was represented by L/D The relative uncertainty of L/D was given by ε L / D = ε L2 CORRECTED + ε D2 CORRECTED which improved slightly since the relative uncertainty of dynamic pressure was not required 235 Appendix X Corrections to Data Pressure Taps at the Leading and Trailing edge There was no pressure tap at the leading edge of the wing when all the readings were taken At the end data collection, an additional tap was inserted at the leading edge The metal tube in this tap was of diameter 1.8mm, while the existing tubes are of 1.3mm diameter Three test runs were carried out at α = 0O These show that the rise in water height from the leading edge is 0.94 of the pitot reading rise in water height This pressure reading was inserted into pressure readings of each run While it is direct to obtain stagnation point movement from pressure readings at the leading edge, changes in upper surface suction was used to determine change in circulation This also indicates the stagnation point movement and simplified the construction since no pressure taps were needed at the wing’s leading edge Further details are in the appendix for leading edge pressure measurement Adjustment to Measured Static Pressure on Pitot tube As the pitot static tube was incorrectly positioned ahead of the flat plate, the measured static pressure was corrected using dp/dx The correction effectively shifted the pitot static tube downstream by 0.36m, resulting in lower static pressure readings Corrections to Pressure measurements Static pressure corrected using dp/dx Dynamic pressure corrected using blockage correction Corrections to Force measurements Static pressure corrected using dp/dx Dynamic pressure corrected using blockage correction Lift force corrected using loadcell calibration Drag force corrected using loadcell calibration Drag force corrected again using buoyancy correction Buoyancy correction Downstream pressure gradient was determined from both walls of the tunnel While maximum practical effort was expanded to seal the tunnel walls, leaks were present on the starboard wall The obtained pressure gradient was not linear and its magnitude did not reduce with increasing Reynolds number The pressure gradient from the portside wall was linear and responded to changing Reynolds number of the flow This value of pressure gradient was taken to represent the pressure gradient of the entire free stream region, which in turn was used for buoyancy correction Due to temperature changes, the power required to produce a flow in the test section with Reynolds Number as close to 200000 varied between power inputs indicated by 26hz to 27hz on the control panel of the wind tunnel controller It was not practical to find times where temperature varied between 27 to 32OC to determine the pressure gradient Since temperature caused power inputs to vary, varying power input was used to select the pressure gradient Running the wind tunnel at 26hz and 27hz produced two distinct values of streamwise pressure gradient Further details are in the appendix for pressure gradient The required pressure gradient was obtained by linear interpolation based on the frequency of the power input 236 ... with a large cargo space is expectedly heavy However, payload fraction and fuel fraction are dependent on construction materials and these fractions improve as material advances take place Stronger... of a Flat Moving Ground on lift and drag 98 4.6.2 Effects of a Flat Stationary Ground on lift and drag 99 4.7 Changes in Lift and Drag for a Finite Wing in Ground Effect 100 CONCLUSION... attack AoA Angle of Attack (used only in figures) AMF Model frontal area ATS Cross sectional area of test section A, B,X General symbols for measured or calculated quantities AR Aspect ratio b