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

[wb92]

Come on! We're all late...I've made my first simulation today after putting together a car in 3 days!

In 3 days!? That's quick, is there any part based on your F1 2017 car thread?
I hope on Monday/Tuesday I could have my car ready for calculation.

Seems like Cl ~2 is minimum target so far...?

[rjsa]

now this:

FOAM Warning : 
    From function Foam::labelHashSet Foam::polyBoundaryMesh::patchSet(const Foam::UList<Foam::wordRe>&, bool, bool) const
    in file meshes/polyMesh/polyBoundaryMesh/polyBoundaryMesh.C at line 794
    Cannot find any patch or group names matching "Streak1_mk01_0001.*"

[variante]

In 3 days!? That's quick, is there any part based on your F1 2017 car thread?

I've made a simple car...no complex shapes or anything (and it's gonna be pretty ugly...). The car is 100% new; even the airfoils have not been carried over from other models.

Seems like Cl ~2 is minimum target so far...?

Yeah, something like that. But i see there's the potential to reach ClA 4 very easily (not that i have, unfortunately).

Also, let's not underestimate balance (which is all over the place for me ATM) and cooling.

[CAEdevice]

Yeah, something like that. But i see there's the potential to reach ClA 4 very easily (not that i have, unfortunately).

Cl4 is science fiction for my car... ClA2 is my target at the moment

[wb92]

Okay, thanks for some clues where to end up.
For me currently cooling duct is the biggest challenge to fit in general concept.

Meanwhile - what kind of mesh size is set up by default in calculation (approx.) ? I would like to know if the calculation will take half or whole day.

I think my car is also simple, but simplicity is relative.

[CAEdevice]

The official setup is "MVRC" (2500 iterations). It requires about 12 hours on a PC with 4 cores, 3.2 GHz and 16gb of RAM.

[LVDH]

It is running despite the wheels and bluecfd error messages.

BUT:

60803 wall                MVRC_bool_reduced_patch60794  
60804 wall                MVRC_bool_reduced_patch60795  
60805 wall                MVRC_bool_reduced_patch60796  
60806 wall                MVRC_bool_reduced_patch60797  

Seems to be on and endless loop.

No it's not caught on a loop, moved on to another part.

Ok, many questions to answer.

So merging the wheels with other parts was not a good idea. The wheels rotate and need a corresponding boundary condition.
The warning about the wheels is somewhat clear though. Somehow your wheels seem to be in the wrong position. You should have a wheelbase of 3.4m.
I really do not understand your issue with the 30 files, when using an MVRC template. I can image that you open the GUI, while having some other template selected and then opening your case. Then it will not load all parts into the GUI. When building the case after selecting the correct template it will include all parts. If you continue to have issues, please send an email.

@Andre, what causes this?

60803 wall                MVRC_bool_reduced_patch60794  
60804 wall                MVRC_bool_reduced_patch60795  
60805 wall                MVRC_bool_reduced_patch60796  
60806 wall                MVRC_bool_reduced_patch60797  

It happens with the provided parts I decimated using magics to reduce size.

This is a bit out of context, so I have no idea. I can only check if you send me the case.

now this:

FOAM Warning : 
    From function Foam::labelHashSet Foam::polyBoundaryMesh::patchSet(const Foam::UList<Foam::wordRe>&, bool, bool) const
    in file meshes/polyMesh/polyBoundaryMesh/polyBoundaryMesh.C at line 794
    Cannot find any patch or group names matching "Streak1_mk01_0001.*"

It is a warning and you can ignore it.

I finished my first couple of cfd's with the new software and my new cpu. There is some significant difference on the results that i got from different cfd options.

Time = 1200
forceCoeffs forceCoeffs write:
    Cm    = 0.27190957
    Cd    = 1.1457955
    Cl    = -1.63347
    Cl(f) = -0.5448254
    Cl(r) = -1.0886445

Time = 1200
forceCoeffs forceCoeffs write:
    Cm    = 1.366718
    Cd    = 1.1547571
    Cl    = -1.4156811
    Cl(f) = 0.65887749
    Cl(r) = -2.0745586

The first one is the standart MVRC option. Second one is the MVRC_Fast option with same car. I think this is a serious thing since i often use the fast option for quick results.

Why "Time = 1200"? Iterations in "fast" go up to 2000; in "normal" even more, probably. Given the different settings between "normal" and "fast", i don't think that residuals compare. Let them converge and then lets see what happens.

I've just completed my first simulation in "fast" and everything seems right.

BTW if anyone has similar info about fast and normal, please share because i probably won't have the time to do any comparison.

I quit that one around 1250. Thats why i posted the results at 1200. I completed 3 sims at fast settings, i ended up with these results:

Cd: 1.1227065555, Cl: -0.3831973513, Cl(f): 1.7312270195, Cl(r): -2.1144243705, CoP: 18.761

Cd: 1.1888227472, Cl: -1.7330100288, Cl/Cd: -1.4577530863, Cl(f): -1.7699407171, Cl(r): 0.0369306882, CoP: -0.072

Cd: 1.2144504931, Cl: -1.8959802335, Cl/Cd: -1.5611836335, Cl(f): -1.7225041266, Cl(r): -0.1734761067, CoP: 0.311

Look, you stopped both simulations pretty early. You should check if the plots started leveling out (converged). And due to a coarser mesh in the "fast" setting there will be a delta in the numbers. Right now your comparison just does not make much sense.

The official setup is "MVRC" (2500 iterations). It requires about 12 hours on a PC with 4 cores, 3.2 GHz and 16gb of RAM.

The races run in the "long" option. This one is just like the normal template but runs for 5000 iterations.

I am working on possibly sending out a new version of MFlow this weekend. I had some issues with my computer otherwise it would have already been out.

[jjn9128]

I've decided not to compete (various reasons for that) but I have a front wing/rear wing and floor (all untested) if anyone wants the STEP files to try with their model. I think it's all legal (maybe not the diffuser strakes..) but obviously that's at your own discretion if you chose to use it.

[variante]

Yeah, something like that. But i see there's the potential to reach ClA 4 very easily (not that i have, unfortunately).

Cl4 is science fiction for my car... ClA2 is my target at the moment

ClA 4 as long as the "fast" option is proven reliable. I have that as reference.

The official setup is "MVRC" (2500 iterations). It requires about 12 hours on a PC with 4 cores, 3.2 GHz and 16gb of RAM.

And the "fast" with 2000 iterations takes like 6 hours on a i7-7700HQ laptop.

[Ft5fTL]

I didnt stop both simulations. I just stopped the fast one because it didnt make any sense and i already completed 3 fast sims. There is just no way to get a result like Cl(f): 1.7312270195 with a 4 element fwing. (unless the wing is upside down) Nevertheless i will complete the fast one i linked before and share.

[LVDH]

I didnt stop both simulations. I just stopped the fast one because it didnt make any sense and i already completed 3 fast sims. There is just no way to get a result like Cl(f): 1.7312270195 with a 4 element fwing. (unless the wing is upside down) Nevertheless i will complete the fast one i linked before and share.

You can send me your car or preferably the report. Maybe something is going bad.

[LVDH]

I've decided not to compete (various reasons for that) but I have a front wing/rear wing and floor (all untested) if anyone wants the STEP files to try with their model. I think it's all legal (maybe not the diffuser strakes..) but obviously that's at your own discretion if you chose to use it.

You have no idea how it is to be sitting there on a Sunday, waiting for the results to come in. You should really try it.

[rjsa]

This is a bit out of context, so I have no idea. I can only check if you send me the case.

How can I send it?

[variante]

I've decided not to compete (various reasons for that) but I have a front wing/rear wing and floor (all untested) if anyone wants the STEP files to try with their model. I think it's all legal (maybe not the diffuser strakes..) but obviously that's at your own discretion if you chose to use it.

You have no idea how it is to be sitting there on a Sunday, waiting for the results to come in. You should really try it.

Yeah and, also because of the new regulations, i suspect you would have good chances of victory.

[wb92]

Is there any 'test case' to check if software works properly? Also where I have to place STLs for my own runs?

EDiT:
Nevermind - handled tutorials, so hopefully I can now manage it by myself.