[MVRC] Mantium Virtual Racecar Challenge 2019 (Grand Prix Cars)

[rjsa]

Isn’t there a service that will host images without an account any longer? Cubeupload and Imgurl asking for accounts now.

[machin]

Time was really tight to get things under controlled development, hopefully few big issues should be minimalized, and here we go, reveal of the rookie contender:

https://ibb.co/9tV0YdQ
https://ibb.co/27Q4pPB
https://ibb.co/1brpLF9


Enjoy!

Another great looking car!

[Ft5fTL]

Is late submit still a thing like we had on the prev. seasons?

Anything about this?

[jjn9128]

Is late submit still a thing like we had on the prev. seasons?

Anything about this?

In the rules

Extend the submission deadline by two days. This may only be done once per season.

[Ft5fTL]

Ah, my bad. Didn't read that part of the rulebook. I'm gonna submit it tuesday then.

[Thomas2019]

Submitted. Still much to test and try out. I only just begin to understand the inter-dependencies.


Good luck to everyone!

Thomas

[rjsa]

Sent. It wouldn’t qualify for a F1 race by a safe margin though.

[BlakjeKaas]

Ah, my bad. Didn't read that part of the rulebook. I'm gonna submit it tuesday then.

Gonna do the same, it's going to be a terrible car though, but at the very least I want to get a simulation in before I send it in.

[CAEdevice]

I am comparing the resusts of two different simulations: I am not so sure that the issue related to the component CoP computation does not affect the total results:

Car A

Force Coefficients averaged Total - Cd: x.xxxxxxxx, Cl: -x.xxxxxxxx, Cl/Cd: -x.xxxxxxxx, Cl(f): -x.xxxxxxxx, Cl(r): -x.xxxxxxxx, CoP: 0.926
... [all other numbers are almost identical between the two cars] ...
Force Coefficients averaged wing_front - Cd: 0.2662126972, Cl: -1.1066677259, Cl/Cd: -4.1570809260, Cl(f): -3.4186499550, Cl(r): 2.3119822290, CoP: -
Force Coefficients averaged wing_rear - Cd: 0.2789338788, Cl: -0.6844554475, Cl/Cd: -2.4538268728, Cl(f): 1.2724959805, Cl(r): -1.9569514277, CoP: -

Car B

Force Coefficients averaged Total - Cd: x.xxxxxxxx, Cl: -x.xxxxxxxx, Cl/Cd: -x.xxxxxxxx, Cl(f): -x.xxxxxxxx, Cl(r): -x.xxxxxxxx, CoP: 1.408
... [all other numbers are almost identical between the two cars] ...
Force Coefficients averaged wing_front - Cd: 0.2760665220, Cl: -1.1495621821, Cl/Cd: -4.1640767362, Cl(f): -1.4451989997, Cl(r): 0.2956368176, CoP: -
Force Coefficients averaged wing_rear - Cd: 0.2659228715, Cl: -0.6662020416, Cl/Cd: -2.5052453659, Cl(f): 0.1257940060, Cl(r): -0.7919960476, CoP: -

Can a wing, that has about the same downforce and drag, have such different impact on the CoP?

Notes:

• Car A and car B are almost identical, the only difference is +1° of the angle of the rear wing and-0,5° of the angle of the front wing: the computed forces of the wigns are congruent with this variation, but the contribution to Cl(f) and Cl(r), for both components and the whole assembly, are completely illogical. Any other part is unchanged (I mean: same STLs).
• Considering the the variation of the force is reasonable, I guess that there isn't a convergence or stability or turbolence model issue: this is confirmed buy the pressure distribution I see with Paraview (but it's the same with the default post processing), that is reasonable and looks as expected.

I think the problem is only in the way the final number are computed, not in the simulation itself.

EDIT: I hid the data about total df and drag in order to not reveal the results in advance, but I can show them if you think it is useful.

[LVDH]

Well, without seeing the whole simulations, I think it is very possible. The front and rear wings act with a large lever. Then, if they are the major downforce distributors, changing something with them, will make a big difference. Also you should not underestimate what one deg of AoA can do, esp on the front wing, it is very wide.

[CAEdevice]

Well, without seeing the whole simulations, I think it is very possible. The front and rear wings act with a large lever. Then, if they are the major downforce distributors, changing something with them, will make a big difference. Also you should not underestimate what one deg of AoA can do, esp on the front wing, it is very wide.

Yes, it would be better to show the whole data, but I prefer to not anticipate the official results (anyway I can send you privately the case folders).

I am convinced that is not possibile that a variation of 4% of drag or lift of a componet (rear wing) placed in the same place (same moment arm), without any other significant force on the other parts of the model, can cause 200% or 300% variation of the effect of that component on the general balance of forces.

The front and rear wings act with a large lever.

Large but still proportional... M = b * F

... and a 4% variation is comparable with numerical convergence "errors": does that mean that the usual numeric approximation (less then 5% I estimate) can lead to a variation of the resultant moment of the component of 200-300 (or more)? If so it dos not make sense to simulate the car...

Anyway: I decided to submit the car that has the better balance indipendently by the numbers produced by the software.

[LVDH]

Some polars, I found by using a search engine:

[CAEdevice]

Some polars, I found by using a search engine:
https://www.researchgate.net/profile/Ca ... or-all.png

We are not talking about the same thing.

I try to explain: provided that 1° of variation can make a signficant difference about the wing lift or drag, the effect of this differcence on the computed moments (not matter if it is 5%, like it happens in the simulations I shown or 1000% as you may find elsewhere), if any other force on the assembly remains the same, must be proportional

I would be ok if we had a 10000% downforce difference on a wing, after changing 1°: I am not an aerodynamicist, I must accept this, but the balance of forces produced by this variation (a pure static problem) should be congruent.


Force Coefficients averaged wing_rear (same numbers presented in a different way)

Cl: -0.6844554475 >>>> -0.6662020416 (delta 2,6%)
Cd: 0.2789338788 >>>> 0.2659228715 (delta 4,8%)

Cl(f): 1.2724959805 >>>> 0.1257940060 (delta >1000%)

(Moment arms and all other force not varied)

Navier and Stokes are peaceful but Galileo is suffering.

[Thomas2019]

I used a different OpenFOAM variant and dont have the CoP from the Mantium report.
I calculated the CoP from the output of Cm and Cl in the forceCoefficient log file. The reference point is defined at x=1.7m.
So the Cl acts around that point to create Cm. If you know Cl and Cm you can work out the position of CoP on the x-axis.
At least thats my thought process.

EDIT: And Cd plays a role as well, but the lever of Cd is quite small compared to the main lift producing parts.

Thomas

[rjsa]

Do you see any crazy going on on the 3d pressure plots? The red one for instance?