Wind drag on body
- Thread starter Pinewood Outlaw
- Start date
Im no pro but as i understand the question and from what research ive done on the topic its a "ying yang" thing...the bottom (between the track and car) has an area of compresion causing more drag with an identical shape on top....but the longer liner surface (the top) with its angled surface for a wedge or curved contour) has a higher drag than the flat bottom sooo its very tough to say without some kind of math i couldnt possibly comprehend
Can I interpret your question to: which has the most drag, instead of most turbulence?
Of course it depends on the shape of the car, but the stock full width wheels are your largest source of air drag with a low profile car.
Digging out a very old simulation I did years back:
Just considering the body, it looks like the bottom experiences more drag than the top, especially in the nose portion. This first car in the top image has a 0.25" thickness, and the front nose is pointed in the vertical center (symmetric top to bottom). Maybe a better design would be to have the point lower, but the difference will be very small.
My track in the model has a 0.25" flat raised portion for a rail, which is different than the besttrack cross section. which is open in the middle. So results will differ.
The fenders are far from ideal, but they still reduce the total drag on the wheels by about 30% vs. open wheels. Open wheel car shown in second image.
Of course it depends on the shape of the car, but the stock full width wheels are your largest source of air drag with a low profile car.
Digging out a very old simulation I did years back:
Just considering the body, it looks like the bottom experiences more drag than the top, especially in the nose portion. This first car in the top image has a 0.25" thickness, and the front nose is pointed in the vertical center (symmetric top to bottom). Maybe a better design would be to have the point lower, but the difference will be very small.
My track in the model has a 0.25" flat raised portion for a rail, which is different than the besttrack cross section. which is open in the middle. So results will differ.
The fenders are far from ideal, but they still reduce the total drag on the wheels by about 30% vs. open wheels. Open wheel car shown in second image.
Not sure if this is related, but with a wedge, does it make a difference if the front of the car is angled down or up? The very front where it is tapered I mean. Seems like tapering down would increase the weight of the car, and tapering up with create a little lift and decrease it. I'm probably totally wrong of course.
First time poster, but I thought I'd share a bit. A friend was doing a new OpenFOAM CFD install, and wanted some data to run, to test it, so I sent him some pinewood derby fender designs. The baseline was the fender set my nephew ran in 2018, which was his 3rd generation car. Wood chassis is 5/16" thick for reference.
Cd= .561
Then we ran a full length single fender, more blunted, like a current LMP, going for as low a Cd as possible, without wheel skirts.
Cd=.450
Caveat being the Cg was moved forward, due to more fender mass.
I know we ran a fenderless car, but I can't find the data on that. I just wanted to toss these out, as food for discussion.
Cd= .561
Then we ran a full length single fender, more blunted, like a current LMP, going for as low a Cd as possible, without wheel skirts.
Cd=.450
Caveat being the Cg was moved forward, due to more fender mass.
I know we ran a fenderless car, but I can't find the data on that. I just wanted to toss these out, as food for discussion.
If im interrupting the data correctly a full length flat fender has less drag than two tapered fenders is that a correct interpretation?
Last edited:
Correct, you're turning the air less, and not hitting a second frontal face (the second fender).
Also of note, those runs were done with the car straight. A rail rider is actually in yaw going down the track, thus, their is some side force on the car. In theory, the side force would help hold the car against the rail, so you could dial a bit of steer out of the car, and reduce the rolling friction, but do you lose more with the side force Cd than you gain in reduced rolling drag?
Best case, would probably be to skew the fenders, so they are straight down the track, when the car is in yaw.
Best case, would probably be to skew the fenders, so they are straight down the track, when the car is in yaw.
Hats off to you CAD_Jockey for taking the time and energy to help your nephew in his racing career. He is a very lucky kid the world needs more uncles and people like you! I hope your input helped him in his quest for a trophy or two!
Here are some images from a non-fendered car, that Corvid ran CFD on (http://www.corvidtechnologies.com ).
https://www.aeriusphoto.com/OurRaceCarsSpeedwaybikesKartin/First-race-car/
https://www.aeriusphoto.com/OurRaceCarsSpeedwaybikesKartin/First-race-car/
It would be better to run the CFD's with cars on either side to get a more real world analysis. Cars don't race alone (usually) 
Identical Cars on either side could easily be simulated by just using symmetry planes on the two sides on the plane between the cars. No need to actually model the other cars to get their effect. Symmetry plane has conditions: Vx = 0 (Normal Velocity to the plane = 0), and dS/dx = 0 (all scalar gradients normal to the plane are also = 0.)
I have a feeling the effect from other cars will be pretty small if they are side by side. (If the cars were streamlined) blunt bodies and larger cars will have more of an effect as they plow more air.
I have a feeling the effect from other cars will be pretty small if they are side by side. (If the cars were streamlined) blunt bodies and larger cars will have more of an effect as they plow more air.
Have you ever raced one of these cars? The cars next to you have a HUGE effect on your car. There must be a thread or threads on the forum that dives into the effects.
I think the key to CFD_Engineer's statement, was "streamlined". If all of the cars on track are aero efficient, there should be less wake.
When dealing with the wake from other cars, you would have to run cases with the opponent's car at various positions ahead, and behind your car, and vary from all 4 lane positions. Then you factor in the difference between running against non-fendered, 2 fender, 4 fender, leading and trailing front fenders with rear leading fenders.... at this point, you are leasing cpu time on Amazon's cloud, to get all of the runs done
When dealing with the wake from other cars, you would have to run cases with the opponent's car at various positions ahead, and behind your car, and vary from all 4 lane positions. Then you factor in the difference between running against non-fendered, 2 fender, 4 fender, leading and trailing front fenders with rear leading fenders.... at this point, you are leasing cpu time on Amazon's cloud, to get all of the runs done
How a car run will run in traffic is one of the most important and valuable pieces of information on how well a car will do.
Agreed on the difficulty of many simulations with cars at different positions, that would be a horrible task. But I'm guessing the more aerodynamic the car by itself, the more aerodynamic it will be with other cars as well. The trends will be the same, so just concentrate on a single car with symmetry planes.
I'd have to disagree with real world testing being better, too many variables there, and you could only pull "total drag" out at best. With a wind tunnel you would need a moving track on the bottom as well, not just a stationary base. Also you can't tell what part of the vehicle is making a difference and improve on that locally.
A single, carefully controlled, real world test would be great to validate the CFD model of course, but there is so much more detailed information you can get with CFD, and changes can be made quickly and reanalyzed for faster optimization.
Maybe I should add that the CFD has to be well done, mesh quality and sensitivity analyses, boundary layers resolved properly, correct turbulence models, and perhaps transitional turbulence models. A poorly done CFD isn't too much use, Garbage in Garbage out. So yes, real world testing is definately much better than a poorly done CFD
I'd have to disagree with real world testing being better, too many variables there, and you could only pull "total drag" out at best. With a wind tunnel you would need a moving track on the bottom as well, not just a stationary base. Also you can't tell what part of the vehicle is making a difference and improve on that locally.
A single, carefully controlled, real world test would be great to validate the CFD model of course, but there is so much more detailed information you can get with CFD, and changes can be made quickly and reanalyzed for faster optimization.
Maybe I should add that the CFD has to be well done, mesh quality and sensitivity analyses, boundary layers resolved properly, correct turbulence models, and perhaps transitional turbulence models. A poorly done CFD isn't too much use, Garbage in Garbage out. So yes, real world testing is definately much better than a poorly done CFD
Last edited:
Similar threads
