[MVRC] Mantium Virtual Racecar Challenge 2017

[rjsa]

Nice to hear you are doing fine .

[Alonso Fan]

Nice to hear you are doing fine.

Agreed

[LVDH]

OK, reports are out.

It is always a good idea to have a blog or at least start a thread here in the forum to get help with your cars. Esp. IDADOX and SHM Racing would profit from that. In the past we have seen that some participants went that route and got good and valuable help.
The IDADOX car is tessellated very badly. And keep in mind that you create fantastic surfaces with your CAD tool to then have them butchered in the very last step. As the CFD mesher does not like these bad surfaces either you are loosing performance. Also with that car there were issues with the expected results vs. what I got. This is probably just a rookie CFD mistake. That car converged so slowly that I had to run many iterations on it. When you see plots that are still falling (like in that case the HX volume flow) you have to continue running the simulation until the plot either flattens out or ate least somewhat fluctuates around a mean value.

SHM needs general help. While his rendering skills are by far the best of the pack, the flow around his car is not so smooth.

I hope to get to sorting some of the support questions that have pilled up today and tomorrow. Please also keep in mind that you can use the Q&A section on the challenge page to help each other out. I know most of the time I have to fix something, but sometimes there are issues that are not new.

[rjsa]

I'll try deal with the meshing. Probably for the 3rd race though. One question, what's the minimum triangle size you guys use? Like n usually meshers have a biggest/smallest mesh size setting, what's the smallest?

I think I know why it went so bad. I had two sets of gurneys on the high res dir that shouldn't be there. Moving those to the body dir would make things go crazy like in the race report, with the RW stalling epically. That's been dealt with

On top of that my air intake'd been acting a lot, lots of buffeting. It's way better now. Anyone seen this?

[Ft5fTL]

Looks more like a seismograph reading.

[LVDH]

Unfortunately most cars show some fluctuations but yours are the most intense. Fixing the surface meshing issue is a very hard step but the results will be worth it.

[rjsa]

Today I have a very convoluted workflow. I can improve it but I need to move from Rhino to Fusion to make it simple. And I've been using Rhino since the beta in the 90s, you can see why I resist.

[rjsa]

How hard is the 80mb limit? I got the body under 60mb but my HX parts are pretty heavy with 1cm square meshing. Will a 2cm or 3cm mesh cut it?

[LVDH]

Well it is up to Chris but I am sure that you should not violate it twice. Some HXes are only the minimum amount of triangles you need to define a box. Just keep in mind to properly mesh the monitoring surfaces.
Some guys have their cars in the range of 10MB and they look good. Properly meshing the surfaces is the hardest thing you have to do as CFD engineer.

Maybe you can explain your meshing process in detail and somebody can help improve it.

[cdsavage]

How hard is the 80mb limit? I got the body under 60mb but my HX parts are pretty heavy with 1cm square meshing. Will a 2cm or 3cm mesh cut it?

The 80MB limit should be considered a regular rule like all of the others.

[rjsa]

Less wheels and suspension, right?

[cdsavage]

Less wheels and suspension, right?

It applies to all of the STLs included with the submission - the submission shouldn't include wheels or suspension.

[rjsa]

ok, tks.

[LVDH]

Hi guys,
sorry I was so quiet lately. I was simply working like crazy. I have made some small improvements to the software. You can now download version 2017.7. It seems to solve some issues you guys had with solver stability. I say "seems" because I did not have very much time to check it extensively. Maybe some people can test it and esp. rsja can possibly confirm that his simulations are now a bit more stable, we should go for using this version for the next race which seems to be next week already.
Also there is an update to the rule book. I did not have a look yet but also there the changes should be minimal.

[CAEdevice]

Thanks LDVH, may Chris summarize the rule update?

I am getting crazy with the first race model: I only changed two small things (minor legality issues) and now it stops after meshing (successfully) when solving the first iteration.
Tomorrow I will try with MFlow v2017.7.