Tuesday, December 22, 2015

2004 P04 CFD tests; A Brief Aerodynamic Analysis of a 12 year old design.

P04 2004
When I first designed P04, It was around 1998 my senior year in high school. But I actually modeled the car when I started using rhino, some early 2000s. Car had a basic pre-WWII shape giving it a classic look yet it had some modern features on it such as ground effects. Today, I have just made a simplified CFD model of the car in about 2 hours time and have a 150 iteration run in Ansys CFD software. Results are not perfectly accurate to real world but they are close, such as lowest convergence achieved was around 2.8156e-2 (0.028156%).

detailed rendering of base model used for CFD test

detailed rendering of base model used for CFD test

detailed rendering of base model used for CFD test

Since I am unemployed I thought I go back and evaluate my old designs just to see how well I would have done as a 20 year old me (in this experiment I used a 2004 model of P04 so I was 23year old when I designed it). P04 was designed to be a medium speed sports car not an high end sports car. I was heavily inspired from Morgan Aero 8 (1996 version) and Panoz Esperante GTR-1 while designing it yet it is not completely look a likes of those two cars.

CFD model is was a simplified geometry of late 2004 model of P04 which have a relatively low nose and more realistic engine hood. 150 iterations run in Ansys CFD as a steady state mode. to avoid longer calculation time half model is used so results are based on the half model which need to be multiplied with 2.

Some of the reference values for the test was;
Wind And Ground Speed: 70 m/s (approx. 252kmh)

Frontal Area of the Car: 0.9274703 m^2 (2*0.9274703=1.8549406 m^2 total frontal area)

Rest of the parameters were set according to Ansys CFD sheets which can be found in Program website and over the internet (IF you would like to have full set up parameters you may email or message me).

As a result;

Drag (Half Model Results) :
Forces - Direction Vector (0 1 0)
                          Forces (n)           Coefficients                                  
Zone                    Total                 Total         
bodywork             740.54741        0.26604117    
undertray             77.840038       0.027963983   
wall-solid              253.42488       0.091042722   
------------------------- --------------- ---------------
Net                       1071.8123         0.38504788

Total Drag Generated (complete model) :2143.6246 N (~218.59 kg-f or 481.91 lb-f) at 252 kmh.
Y direction is the direction of drag which is a long the length of the car.

Lift (Half Model Results) :
 Forces - Direction Vector (0 0 1)
                          Forces (n)             Coefficients         
Zone                      Total                 Total         
bodywork               2191.7282       0.78737692    
undertray              -2875.6599     -1.0330789    
wall-solid               -40.74762       -0.014638556  
------------------------- --------------- ---------------
Net                         -724.67929     -0.26034056

Total Lift generated (complete model): -1449.35858 N (~ -147.8 kg-f or -325.8lb-f) at 252 kmh. Z direction is the direction of lift.

Calculated lift to drag ratio: -0.676:1

Moment of lift is rear wards. moment center of the model is; Moment Center according to coordinates axis of X,Y,Z (0, 2.15, 0.36) Moment Axis (1, 0, 0). which is the point that intersects at center of the wheelbase along Z axis and Z height of component of the volume centroid.

Above image shows the moment center of the car graphically

Total moment of lift is equal to 1950.7951 N-m which is in positive direction. In other word rear of the car generates more down force than the front of the car. Direction of the moment is actually determined by the under tray of the car which plays a big role in downforce generation. Aerodynamic balance of the bodywork is actually forward biased.

From these findings we can say that car generates quite a bit drag probably a little higher than now a days sports cars but it generates quite good amount of downforce especially at the rear of the car.

And Finally some Images from the CFD test:

Pressure gradients of the bodywork (-1000Pa to 1000Pa) and streamlines along the floor.

Streamlines along the floor

Pressure gradients on the bodywork

Pressure zone around the car

Vortex regions

Vortex regions

Stream lines along the floor

P04 R 2004

 Racing version of P04 was designed several months earlier than then the road going version (end of 2003). I must say I do have an obsession with racing cars that looks like road cars. P04 road version was designed based on the racing version with slightly less frontal area. Road version was narrower obviously and had slightly higher cockpit height. Racing version had two different cooling configurations yet one was made for rendering and the other was for the analysis. I never had an access to a analysis program back then. So all my designs were experimental until Galmer GT (my first ever design tested with a CFD program).

1. configuration: Side Mounted Radiators; As I have mention car was very much inspired from Panoz GTR-1 for its layout so I made a version with side mounted radiators, where radiators were mounted inside the side pods behind the front wheel. two ducts were feeding those radiators and exhausting from side. Front splitter opening were supplying cool air to the engine bay exhausting partially from forward of wind shield and side pod openings to the rear wheel that's why rear wheel arches are kind of oversized. There also was an opening that supplies air to the radiators front the front splitter opening. I don't know how functional it is but I will be testing it when I have an access to an CFD program.

Side mounted radiators configuration

2. configuration: Front Mounted Radiators; Front bonnet is slightly higher on this version and front wheel poontoons are more of a tear drop shape from top view. This configuration is also very much similar with the road version.

Front mounted radiators configuration

Differences between race version and the road versions;

Front overhang is 50 mm shorter on the racing versions.
Tail angle (Average) is 4 degrees more on the racing versions.
Tail width is slightly less on the racing versions.
Cockpits are 30 mm lower on the racing versions.
Racing versions are 50 mm wider than the road version
Side pods are significantly higher on the racing versions
A pillar start points are more forward on the racing versions.

As for the CFD Test I have run 500 iteration steady state test with same conditions that I have run on the road version. Model detail is more on the racing version so I was expecting a higher drag due to fully functioning radiator inlets and more detailed wheels.

Results for the racing version:

 Test was also carried on at 70 m/s and Frontal area of the racing version is 1.94 m^2 excluding mirrors. Same configuration was applied for the solution such as moving ground and rotating wheels. and half model was considered for the calculation.

Regions of the model:

Yellow: Bodywork
Green: Rear win assy.
Blue: Flat under tray
Purple: Front splitter (diffuser)
Red: Read diffuser
White: Solid wall boundaries and misc.
Dark Gray: Wheels (calculated separately)

Drag (Half Model Results) :

Forces - Direction Vector (0 1 0)
                          Forces (n)             Coefficients                                
Zone                  Total                     Total        
body_work         616.21496        0.21223392  
front_diffuser    30.942650        0.010657125  
front_wheels      107.03279        0.036863739  
rear_diffuser     151.49492        0.052177181  
rear_wheels       110.02186         0.037893223  
rear_wing          82.972048         0.02857685  
undertray          18.214005         0.0062731838
wall-solid           250.88716         0.088219749  
------------------------- --------------- ---------------

Net                     1373.0367         0.47289497  

Total Drag Generated (complete model) :2746.0734 N (~280.022 kg-f or 617.342 lb-f) at 252 km/h.
Y direction is the direction of drag which is a long the length of the car.

Lift (Half Model Results) :

Forces - Direction Vector (0 0 1)
                                    Forces (n)         Coefficients                                   
Zone                           Total                  Total          
body_work                 2457.7375         0.84648261     
front_diffuser            -1559.5129       -0.53712024    
front_wheels              105.76402          0.036426757    
rear_diffuser             -505.90182       -0.17424037    
rear_wheels               82.35793            0.028365339    
rear_wing                 -1712.1224        -0.5896813     
undertray                 -2336.1731        -0.80461397    
wall-solid                  -599.22192        -0.20638125    
------------------------- --------------- --------------- --------------- --------------- --------------- ---------------
Net                            -4067.0727         -1.4007624     

Total Lift generated (complete model): -8134.1454 N (~ -829.452 kg-f or -1828.63 lb-f) at 252 kmh. Z direction is the direction of lift.

Calculated lift to Drag ratio: -2.9621:1

Pressure Contours

Pressure Contours

Stream lines

Stream lines

Stream lines

Honestly I wasn't expecting this much of a lift to drag ratio from this car. there are few faulty areas that needs to be fixed though. car generates high drag which can be reduced and at the end of the front diffuser where it joins with main floor there are pressure zones that possibly create lift and front diffuser looses its efficiency because of that. I think P04 would have been an interesting looking race car, rather than looking modern sports car it looks like 1930s cars after all and it has pretty good lift to drag ratio.

© 2016 Togay Yuvanç

Saturday, December 5, 2015

P49 CFD results (4 runs)

I have made total of 4 CFD runs as I have find time to use Ansys License, Lift and Drag values at 252 km/h (70 m/s) are in order (Units are in kgf):

1. Run ( early model no inlets condition)
Overall Drag: 186.230 kg-f
Overall Lift: -323.158 kg-f

2. Run ( early model inlets are open)
Overall Drag: 177.916 kg-f
Overall Lift: -213.253 kg-f

3. Run ( slightly modified model inlets open)
Overall Drag: 231.436 kg-f
Overall Lift: -374.083 kg-f

4. Run ( model hasn't changed but aerodynamic components were adjusted such as wing height, turning vane angles, and rear diffuser plates... etc)
Overall Drag: 228.747 kg-f
Overall Lift: -407.618 kg-f

As a result with some aerodynamic adjustments on the car drag reduction is possible with some downforce gain. There will be 2 more runs for 2 different models in the future. mos of the downforce on the car is generated by the rear diffuser and front diffuser secondary to the rear diffuser. Some of the parts needs to be optimized such as front diffuser that supplies too much airflow into the body that some of the air spills from front wheel arches which kind of my problem for the time being rest of the car works fine with some excessive drag that i am trying to reduce without loosing too much of the downforce. Downforce levels of the car is quite high relative to its drag, yet still needs to be optimized.

P.S. I must admit this is not a detailed report on aero data of the car But I would like to keep detailed data to myself for the time being.

Friday, August 7, 2015

Some rendering work for a friend

A small rendering projects for a friend of mine's hotel. That he was face lifting his hotel in Mutrah, Sultanate of Oman. Place called Mutrah Hotel which is the oldest hotel in Sultanate of Oman. Face lift project was only phase of the many that will be continued on the future. Rework of the hotel has been done around the lobby and restaurant exteriors.

WIP Render evening

19:00 15th June evening time in Muscat

11:00 15th June morning time in Muscat (filtered)

10:00 15th June morning time in Muscat

18:40 15th June evening time in Muscat

10:00 15th June morning time in Muscat (higher iso number)

Monday, June 1, 2015

P49 Main Frame FEA Update and Rear Frame Preliminary Data

 Main Frame Update:

Comparison of old and updated chassis (deflection wise) under 30000 Nm (22126.86 lbf*ft) torsional load (below). Previous deflection value in degree was 1.08 degrees, now it is 0.861 degree and there is no weight gain on the chassis a part from modification of the beams and some additional mini gussets on some weld points. Current calculated torsional stiffness is ~34480 Nm/degree, previous value was 27540Nm/degree (~20% stiffer chassis). Stress concentration on chassis beams and welds are also less.
Updated chassis from March 2015 (tested in May 2015)
Chassis from late 2014

 Rear Frame 1st FEA Test Obtained Data:

Another FEA study for the rear frame which was done in Autodesk Inventor, Rear Frame is quite similar to Galmer GT rear frame (due to same power and gear train) with some differences and this one is symmetric. Torque range varies from 15000Nm to 80000Nm and torsional stiffness 79225.5221 Nm/degree which is mathematically achievable but failure occurs on some beams (eg. gearbox frame diagonal brace) beyond 50000Nm Torsional load. yet model is not perfect there are some residual stresses due to CAD design of the beam which the don't contact perfectly. calculation was done with 847676 elements (1635684 nodes). Detail of beams increases close to where torsional load was applied. Calculation to about 27 hours on incremental loads and computer crashed 5 times. otherwise calculation would have been less. Material used: 4130 Chrome-Molybdenum alloy steel which has a yield point at 460 MPa.

Safety factor map

Safety factor map

Von Mises stresses

Von Mises stresses
Resultant displacement
Displacement in Y direction