Help with track car desing

[AlvarP]

Hi everyone,

I'm a first time poster so sorry if this is in wrong thread but anyways (also not native english speaker so missplelling on the way).

I'm designing a track car in solidworks and i ran into a problem with the downforce produced by the underbody. currently my models under body producing about 2400 N of downforce at 55 m/s (200 km/h) and about 1800 N in total because the top surface of the car (that is colored blue in one of the pictures) is producing about 550 N of lift.

So now some measurements coming :
Ride height is 30 mm, The tunnel is 20 mm higher than the rest of the floor (50 mm from the road), Raised part of the splitter is 50 mm high and 400 mm wide, Diffuser exit is 1000 mm wide and 250 mm high, Tunnels length from inlet to outlet is ~2500 mm, flat part of the tunnel is 1100 mm long.

So firstly some explanation:
Those wings and front fender aren't ready jet as you can see. The engine that i'll be using is from hyabusa and turbocharged so max 400 hp "reliably". Those "side pot" inlet are for cooling i haven't wet modeled the cooling inlet but the radiator will either cover the whole tunnel and after the cooling the air will flow to the back of the car (and now that i think about it those cooling outlets at the back should be smaller to slow the air down so that it will have more time to absorb the heat). The wheels are covered to reduce drag like in the red bull x2010 and lastly i made that tunnel so that if the car hit a bump i won't lose all the downforce.

Now the actual question:
So to my understanding 2400 N is not bad for a road car but i'd like to see figures in 5000 N - 7000 N range. According to wikipedia f1 car produces about twice it's own weight in downforce at 200 km/h so about 15000 N so my goal shouldn't be impossible because f1 underbodies are highly regulated. So should i make the tunnel in a shape of an hourglass when looked from from the bottom. And do you think i could get more angle from the diffuser (currently 15 deg). and what about the inlet in the splitter is it too small? And could some kind of vortex generators be helpful? Also the diffuser is 900 mm long because that's where the engine stops if i want the lowest center of gravity possible. And lastly should I tunnel the air differently, so in or on top of the diffuser?

TLDR; Look at the pictures and tell me how to get more downforce from the underbody

Lastly about solidworks am i doing it correctly by making the road as a big thin rectangle and then putting the computing area 1 mm inside the 3d rectangle and then making the surface as a wall or is there a way to make the road as 2D plane?

Sorry for a long post and thanks in advance

[Tim.Wright]

I'm not a fluids guy but I'd guess the easiest way to increase downforce would be to extend the floor up past the front wheels. This will nearly double your surface area.

[PhillipM]

I'd also be checking the optics on that windscreen before committing to running all my models on it...

[flynfrog]

also not a fluids guy you can cut some lift by rethinking your wheel fairings probably drop a good amount of drag as well.

[AlvarP]

I'm not a fluids guy but I'd guess the easiest way to increase downforce would be to extend the floor up past the front wheels. This will nearly double your surface area.

I know that but I really like the f1 front desing even tho i know that isn't the ideal desing but other wise i'm open to changing the design.

I'd also be checking the optics on that windscreen before committing to running all my models on it...

I'm not that knowledge about winscreen optics but i made it so that your eyes are at the same hight that they would be in f1 car, but of course that curved screen will bend light a little bit. I don't know if you ment that "unrounded" part in the middle and that's because i wasn't able to round it completely with the variable fillet tool.

also not a fluids guy you can cut some lift by rethinking your wheel fairings probably drop a good amount of drag as well.

I'm still working on those but if i remember correctly (i'm in school so don't have the numbers with me) the drag wasn't that bad about 700 N but of course all the drag should be eliminated if possible.

[CBeck113]

You work for Ferrari, don't you?

[AlvarP]

You work for Ferrari, don't you?

I will neither confirm or deny that claim

[Vyssion]

So to my understanding 2400 N is not bad for a road car but i'd like to see figures in 5000 N - 7000 N range. According to wikipedia f1 car produces about twice it's own weight in downforce at 200 km/h so about 15000 N so my goal shouldn't be impossible because f1 underbodies are highly regulated.

I would just note that those approximations on performance are for the full car - not just the underbody. Given that the underbody is responsible from 30-40% of a cars total downforce, 5000-7000N as you say will be difficult to attain...

So should i make the tunnel in a shape of an hourglass when looked from from the bottom. And do you think i could get more angle from the diffuser (currently 15 deg). and what about the inlet in the splitter is it too small? And could some kind of vortex generators be helpful? Also the diffuser is 900 mm long because that's where the engine stops if i want the lowest center of gravity possible. And lastly should I tunnel the air differently, so in or on top of the diffuser?

Few things I can see just at a glance, on your streamlines plot for the top surface of the car, there is a strong recirculation region towards the back of the vehicle between the rear wheel shroud and the centre chassis part. My guess as to why this is occurring (without having the full post-processing in front of me) would be that the gurney flaps which you have placed at the rear of the vehicle are much too large and incorrectly used. Typically, a gurney flap needs to be large enough to extend just beyond the boundary layer which forms on the surface it is mounted on. Most gurney flaps I have implemented have been on the order of 5-15mm. They are typically used on wings though - the way in which you have yours mounted makes me think that you have gone for a spoiler effect (i.e. to reduce drag a little) which if so, would mean that they need to not be at a 90° angle to the chassis. Also, spoilers are usually employed on their own - not with a rear wing - so if you are looking for downforce, I would ditch them. The aim of a gurney flap is to create a small stagnation pressure on the top of a downforce generating wing in the form of a vortex which then sheds over the top of the flap alternating from one shed off the lower side of the aerofoil.
What I would say is occurring is that you are getting a large stagnation pressure at the gurney flaps at the back of the car which is causing a rather large vortex to be shed off the flap when the pressure builds up. Due to it's proximity to the rear wing, I would say that the vortex is interacting with the boundary layer flow on the underside of your rear wing and causing it to separate, reducing downforce and increasing drag.
The rear of your car is also very "square"... Have you ever looked at the flow around an Ahmed Body? You would find that the flow is again separating off the sides of your rear tyre shrouds which would cause a bit of a vacuum effect behind the car.
I also can see that you don't currently use any endplates on the rear wing, and just have it floating in space. In that set up, the high pressure on top of the wing will want to roll over (generating an edge vortex) to the lower pressure side of the aerofoil. This serves to decrease the performance of the wing - so if you employ even just plain old rectangle endplates to hold the rear wing, you should see an improvement in downforce.
Also, I am not sure how refined your mesh is... the stagnation pressure located on the symmetry planar contour you supplied at the splitter looks to be a little "too" neat of a square...
You also seem to have a lot of 90° steps all over the geometry. Modelling a car is tricky with solid modelling which it looks like you have used for this design, but each of those steps will induce flow separation and muck up the airflow.

Lastly about solidworks am i doing it correctly by making the road as a big thin rectangle and then putting the computing area 1 mm inside the 3d rectangle and then making the surface as a wall or is there a way to make the road as 2D plane?

You can do it that way yes. I have also seen people set the lower boundary to be a few mm above the bottom of the wheels to kind of cut them off a bit, whilst also simulating a bit of a contact patch as well. This would result in an infinitely receding geometry as you converge the round tyre to a flat ground though, so perhaps a fillet or a plinth could help with meshing and solving in that region.
Also, are you defining your wheels as a rotating wall during your simulations? Tyres generally shed up to 6 vortices which will interact with the flow. Yours are shrouded quite well, so you may find that you only have to deal with the two at the base of the wheel, but nonetheless, it would be good to include that detail.
Sorry for the long post - without looking more closely at the post processing data, there isn't much more I can help with. It also might be an idea to restrict the contour range to one which highlights areas of interest more than just a global range which tends to stay in teh light blue/green/yellow region.

[AlvarP]

So to my understanding 2400 N is not bad for a road car but i'd like to see figures in 5000 N - 7000 N range. According to wikipedia f1 car produces about twice it's own weight in downforce at 200 km/h so about 15000 N so my goal shouldn't be impossible because f1 underbodies are highly regulated.

I would just note that those approximations on performance are for the full car - not just the underbody. Given that the underbody is responsible from 30-40% of a cars total downforce, 5000-7000N as you say will be difficult to attain...

So should i make the tunnel in a shape of an hourglass when looked from from the bottom. And do you think i could get more angle from the diffuser (currently 15 deg). and what about the inlet in the splitter is it too small? And could some kind of vortex generators be helpful? Also the diffuser is 900 mm long because that's where the engine stops if i want the lowest center of gravity possible. And lastly should I tunnel the air differently, so in or on top of the diffuser?

Few things I can see just at a glance, on your streamlines plot for the top surface of the car, there is a strong recirculation region towards the back of the vehicle between the rear wheel shroud and the centre chassis part. My guess as to why this is occurring (without having the full post-processing in front of me) would be that the gurney flaps which you have placed at the rear of the vehicle are much too large and incorrectly used. Typically, a gurney flap needs to be large enough to extend just beyond the boundary layer which forms on the surface it is mounted on. Most gurney flaps I have implemented have been on the order of 5-15mm. They are typically used on wings though - the way in which you have yours mounted makes me think that you have gone for a spoiler effect (i.e. to reduce drag a little) which if so, would mean that they need to not be at a 90° angle to the chassis. Also, spoilers are usually employed on their own - not with a rear wing - so if you are looking for downforce, I would ditch them. The aim of a gurney flap is to create a small stagnation pressure on the top of a downforce generating wing in the form of a vortex which then sheds over the top of the flap alternating from one shed off the lower side of the aerofoil. http://i.stack.imgur.com/9CHVO.jpg
What I would say is occurring is that you are getting a large stagnation pressure at the gurney flaps at the back of the car which is causing a rather large vortex to be shed off the flap when the pressure builds up. Due to it's proximity to the rear wing, I would say that the vortex is interacting with the boundary layer flow on the underside of your rear wing and causing it to separate, reducing downforce and increasing drag.
The rear of your car is also very "square"... Have you ever looked at the flow around an Ahmed Body? You would find that the flow is again separating off the sides of your rear tyre shrouds which would cause a bit of a vacuum effect behind the car.
I also can see that you don't currently use any endplates on the rear wing, and just have it floating in space. In that set up, the high pressure on top of the wing will want to roll over (generating an edge vortex) to the lower pressure side of the aerofoil. This serves to decrease the performance of the wing - so if you employ even just plain old rectangle endplates to hold the rear wing, you should see an improvement in downforce.
Also, I am not sure how refined your mesh is... the stagnation pressure located on the symmetry planar contour you supplied at the splitter looks to be a little "too" neat of a square...
You also seem to have a lot of 90° steps all over the geometry. Modelling a car is tricky with solid modelling which it looks like you have used for this design, but each of those steps will induce flow separation and muck up the airflow.

Lastly about solidworks am i doing it correctly by making the road as a big thin rectangle and then putting the computing area 1 mm inside the 3d rectangle and then making the surface as a wall or is there a way to make the road as 2D plane?

You can do it that way yes. I have also seen people set the lower boundary to be a few mm above the bottom of the wheels to kind of cut them off a bit, whilst also simulating a bit of a contact patch as well. This would result in an infinitely receding geometry as you converge the round tyre to a flat ground though, so perhaps a fillet or a plinth could help with meshing and solving in that region.
Also, are you defining your wheels as a rotating wall during your simulations? Tyres generally shed up to 6 vortices which will interact with the flow. Yours are shrouded quite well, so you may find that you only have to deal with the two at the base of the wheel, but nonetheless, it would be good to include that detail.
Sorry for the long post - without looking more closely at the post processing data, there isn't much more I can help with. It also might be an idea to restrict the contour range to one which highlights areas of interest more than just a global range which tends to stay in teh light blue/green/yellow region.

Thanks that was helpful . The "complete" model is just to show what i'm aiming for. I didn't run those fenders and wings in the simulation because they aren't properly designed (no end plates, floating in the air etc.). As for the gurney flaps I thought that they would ctrate a high pressure zone on top of the car but now that I think about it there was a vortex behind the car pushing air to the diffuser. So I'm probably better of getting rid of them. I'll post more pictures with the "flow lines" and without gurney flaps when I get to my computer.

[Jersey Tom]

How do you know the numbers you're getting from this are at all accurate and not just nonsense?

[AlvarP]

How do you know the numbers you're getting from this are at all accurate and not just nonsense?

They use the same software in formula sae and there are couple of articles that tell that solid works is abot 95% accurate in aerodynamics.

[krisfx]

How do you know the numbers you're getting from this are at all accurate and not just nonsense?

They use the same software in formula sae and there are couple of articles that tell that solid works is abot 95% accurate in aerodynamics.

That doesn't make the numbers accurate.

[PhillipM]

It's only accurate if it's setup correctly and correlated to be accurate.

WRT the screen, I don't see what F1 eye level has to do with it, your optical distortion will be immense.

[Vyssion]

How do you know the numbers you're getting from this are at all accurate and not just nonsense?

They use the same software in formula sae and there are couple of articles that tell that solid works is abot 95% accurate in aerodynamics.

Having used SolidWorks in school for flow sim and then using proper commercial codes in uni and work, I would be very wary of trusting the actual numbers that SolidWorks spits out... Aside from the fact mentioned above that the numbers which are calculated depend on the setup being correct, it is my own personal opinion that SolidWorks Flow Sim should only be used as a method of light visualisation of what the air is doing as it flows over your object. Regarding numbers, I would be hesitant to assume that, for example, 2345N of downforce meant exactly that... I would be more inclined to believe that the thrust is around the 2300N mark +/- 10%... You could use it (assuming all setup variables are kept constant) to see whether the delta between two small tweaks (i.e. an angle of attack change or a lengthening of chord for an aerofoil) had a positive or negative impact on the results, but beyond that, I would be careful in quoting exact numbers.

[Jersey Tom]

How do you know the numbers you're getting from this are at all accurate and not just nonsense?

They use the same software in formula sae and there are couple of articles that tell that solid works is abot 95% accurate in aerodynamics.

Formula SAE students don't necessarily have the best grasp on what they're doing.

Source: Former student.

Beyond that, "95% accurate in aerodynamics" doesn't mean anything. How could you even quantify that? All simulation tools have strengths and weaknesses, things they're good for or not. You need to know whether what you're trying to achieve is in the "95%" of what a tool does well, or in the 5% that's junk (or whatever that proportion really is).