F1technical.net

Joe7218 wrote: ↑

23 Apr 2020, 13:17

thanks for the tip!
I'm pretty certain I'm not getting as much cooling flow as I need because one of my main bargeboard elements is stalled and preventing som air from entering some sidepods. Once I fix that I should be getting good cooling again.

Heres my sneak-peek

https://i.imgur.com/MRVYbTN.png

Looks nice. I've also found a traditional bargeboard struggles to keep flow attached. The volume is quite short compared to the f1 volume. You can slot it like a multi-element wing. Or as I've done move away to a downwash wing which helps clean up the suspension wake. I've found this approach helps cooling but hurt the front of my floor and sidepod undercut - downforce and drag.

Sneak peak. Trying to start fairly simple and get things working properly. Reading through the forum from last year I'm getting a bit nervous about cooling. Hopefully my simulations match the official results...

Looking pretty good!

As for cooling, have a look at the .stl of your monitoring surface and check if the mesh looks good. That's what tricked me last year.

spacehead3 wrote: ↑

23 Apr 2020, 14:25

Sneak peak. Trying to start fairly simple and get things working properly. Reading through the forum from last year I'm getting a bit nervous about cooling. Hopefully my simulations match the official results...

I can't really see from this angle, but just make sure your gurney flaps are inline with the specific regs about gurneys and minimum surface thickness

Also how tall are they?! MAAAAAHUUSIVE!

Nice to see a fellow Onshaper though.

LVDH wrote: ↑

23 Apr 2020, 08:00

Yesterday night, I uploaded a new MFlow version. It fixes issues involving the strange behavior you have seen in Windows. You should now be able to start the simulation through the GUI again and the post-processing images get put into the correct folder again.
The fix killed something else though, that I had to deactivate. The terminal output from the mesher and cfd solver does not get piped into the MFlow GUI anymore (for now). This means that you will have to take care and check for yourself if the simulation is still running.

I can confirm that the post-pro issue has been resolved. Pictures go in the right folder and report generated.

I also found some extra potential bugs:


At post processing the vehicle surface used seems to be the original instead of the raked one used in the mesh.

Also, are the Clf,Clr coefficients correctly calculated at run time?
This is the output from the gui:
Force Coefficients averaged Total - Cd: 1.254, Cl: -2.126, Cl/Cd: -1.695, Cl(f): -1.495, Cl(r): -0.631, CoP: 1.009
But looking at the per iteration trace it shows positive Cl(rr)!?


Has anyone experienced this issues?

AratzH wrote: ↑

24 Apr 2020, 23:00

I also found some extra potential bugs:
https://i.imgur.com/6QL6LkR.png
https://i.imgur.com/U7fxl68.png
At post processing the vehicle surface used seems to be the original instead of the raked one used in the mesh.

Also, are the Clf,Clr coefficients correctly calculated at run time?
This is the output from the gui:
Force Coefficients averaged Total - Cd: 1.254, Cl: -2.126, Cl/Cd: -1.695, Cl(f): -1.495, Cl(r): -0.631, CoP: 1.009
But looking at the per iteration trace it shows positive Cl(rr)!?
https://i.imgur.com/2QCfWNa.png

Has anyone experienced this issues?

Regarding pictures - same thing happens on slices with transparent model, actually, it would be nice if on slices transparent model could be turned off.

About coefficients - it has to do with lRef in settings. It is set to "1" during simulation, and then as far as I remember, but @LVDH can confirm, that balance is calculated in the report based on Cm

Ok! That makes total sense, I was getting a bit confused. Thanks for the info!

wb92 wrote: ↑

25 Apr 2020, 10:06

About coefficients - it has to do with lRef in settings. It is set to "1" during simulation, and then as far as I remember, but @LVDH can confirm, that balance is calculated in the report based on Cm

That is pretty much the correct answer. But obviously right now it is a bit confusing. I will try to clean up the plot as well. The hard numbers are correct though.

Also not having the iso surfaces and so on not being aligned with the car model is a bit annoying. The question is what should I rotate for the images, the car geometry or the iso surfaces? I assume you guys would prefer the first option, while the second one is probably the smarter choice. Let me know what you want and I will take care of it.

LVDH wrote: ↑

26 Apr 2020, 14:20

wb92 wrote: ↑

25 Apr 2020, 10:06

About coefficients - it has to do with lRef in settings. It is set to "1" during simulation, and then as far as I remember, but @LVDH can confirm, that balance is calculated in the report based on Cm

That is pretty much the correct answer. But obviously right now it is a bit confusing. I will try to clean up the plot as well. The hard numbers are correct though.

Also not having the iso surfaces and so on not being aligned with the car model is a bit annoying. The question is what should I rotate for the images, the car geometry or the iso surfaces? I assume you guys would prefer the first option, while the second one is probably the smarter choice. Let me know what you want and I will take care of it.

For ISO pictures - use already rotated geometry from the case if possible (but I don't care much about it)
For slices - I would be very happy when the transparent car geometry could be turned off

Trackworkers leaked our private test

Cool! I wish i could do that!
Nice design too, and interesting cooling system, very similar to what we used to see in F1 in the recent years. I was tempted to try something like that, but it takes too much modeling for my taste. Let us know if it's worth it!

How is development going, everybody?
It's taking me a lot of effort to make the smallest progress, and i'm running out of options. And the fact that drag reduction is so predominant over downforce is particularly frustrating. On the bright side, you're going to see a couple of interesting things on my car.

variante wrote: ↑

27 Apr 2020, 19:48

Cool! I wish i could do that!
Nice design too, and interesting cooling system, very similar to what we used to see in F1 in the recent years. I was tempted to try something like that, but it takes too much modeling for my taste. Let us know if it's worth it!

How is development going, everybody?
It's taking me a lot of effort to make the smallest progress, and i'm running out of options. And the fact that drag reduction is so predominant over downforce is particularly frustrating. On the bright side, you're going to see a couple of interesting things on my car.

I wanted to try to make such one as well, but it would consume too much time to get it to work (which I had less due to some completely missed results), and knowing my last issues with cooling I guess I will stay with what I have so far (I want to check some crazy stuff elsewhere).

About the 2nd question - I have similar issues like you, but also some new on top of that. The final decision which chassis I will use for first race is still open...
Regarding predominant drag reduction, that's indeed super annoying, because to add downforce I need to add some drag, which results in actually slower laptime if it results in the same efficiency, however to reduce the drag, I need to change completely the philosophy of the car (including the change of one-and-only 0,6deg rake). So far it's more tricky than I expected.

Hi I am a bit confused about the calendar, the dates are race days or submission days? When is the next submission scheduled?

I'm happy with the results right now. Cl/cd is around 1.80 (it was 1.30 at yas marina last year so good progress) and cooling is good. Bigger exit from merging the sidepods and engine cover as one and some louvres helped a lot on the cooling side.