[MVRC] Mantium Virtual Racecar Challenge 2023 (Grand Prix Cars)

[LVDH]

It is time for MVRC 2023!
Yes, it is time for a new season with new rules and new challengers.
You can sign up here: https://mantiumchallenge.com/mvrc-2023-registration/

A big thank you to the MVRC staff (Christian and Max) for all the hard work that goes into this behind the scenes. Also a big thank you to all others that have contributed (Matteo, Ferit and some that I may have forgotten right now).

For this season, we have improved rules, better CAD files, codified sporting regulations and some other small surprises that you will see later and hopefully make MVRC 2023 the best season so far.

In the registration form, there is a field where you can leave a comment about your team or yourself. In this championship, we see great engineering, please use the field to promote yourself. You can use words or a LinkedIn profile, whatever you want.

As the Wordpress page broke, after updating to the 2023 calendar, you can have a look here:

• Race 1 – Monaco GP, Circuit de Monaco – 21.05.2023

• Race 2 – British GP, Silverstone – 25.06.2023

• Race 3 – Italian GP, Monza – 20.08.2023

• Race 4 – United States GP, Circuit of the Americas – 08.10.2023

• Race 5 – Brazilian GP, Interlagos – 05.11.2023

The only thing missing right now, is an updated version of MFlow (official CFD software used in the competition) adapted to the new rules. This should be done in the next one or two weeks. In that time, everyone can familiarize with the new rules (point out mistakes) and design a first batch of designs.

I wish every one a lot of fun and good luck for MVRC 2023,
André

[yinlad]

Fantastic stuff. The rules and parts changes are all positive steps imo. Looking forward to competing again, it's a shame I ran out of spare time to keep developing the car last season

[BlueCheetah66]

Big fan of the new Rear Wing volumes to make F1 realistic wings easier

[CAEdevice]

Great job, the new rules writing is very professional.

Just a doubt: doesn't rule 14.3 risk being interpreted as an arbitrary lowering (associated with an infinitesimal rake variation)? Even adding a clarification regarding the interference with the ground, could the excessive proximity of the plane to the ground cause convergence problems?

[beschadigunc]

I would highly recommend meshing the floor-ground volume a little more finely. As the experiments I made with the old mantium solver with various settings have massively influenced the performance and reality of the airflow. Those cells I would spare from the engine complexity. Even though the geometry looks cool it is very pointless that its soo realistic with sharp edges etc. Just make it a square box with the same volume %1000 better for cfd

[jjn9128]

Mesh refinement volumes is an absolute must for ground vehicles. I'm not sure how it's handled with this competition?

[CAEdevice]

Mesh refinement volumes is an absolute must for ground vehicles. I'm not sure how it's handled with this competition?

When you subscribe, you can download the software to generate the simulation case including all the mesh settings.

Consider that the mesh around floor fences and planck (but I am not sure about the sceond one) is already well refined. Maybe only the diffuser actually needs a bit more accuracy, but always remebering that 24h for a complete simulation on a home PC is the limit if we don't want a competition only accessible to professionals/small companies or people who have access to computing power at universities.

Currently the mesh consists of 6 million cells per half machine (symmetrical model): it would not be reasonable/possible to manage more than 8-10 million cells.

[jjn9128]

Mesh refinement volumes is an absolute must for ground vehicles. I'm not sure how it's handled with this competition?

When you subscribe, you can download the software to generate the simulation case including all the mesh settings.

Consider that the mesh around floor fences and planck (but I am not sure about the sceond one) is already well refined. Maybe only the diffuser actually needs a bit more accuracy, but always remebering that 24h for a complete simulation on a home PC is the limit if we don't want a competition only accessible to professionals/small companies or people who have access to computing power at universities.

Currently the mesh consists of 6 million cells per half machine (symmetrical model): it would not be reasonable/possible to manage more than 8-10 million cells.

But is it just surface and growth rate? Or are there refinement boxes e.g. aroud the contact patches and the wings?

Also I had a look and the tyres don't appear to have a plinth around the cpatch. You probably want something at least 1n mesh cells tall around the base of the tyre.

[variante]

Mesh refinement volumes is an absolute must for ground vehicles. I'm not sure how it's handled with this competition?

When you subscribe, you can download the software to generate the simulation case including all the mesh settings.

Consider that the mesh around floor fences and planck (but I am not sure about the sceond one) is already well refined. Maybe only the diffuser actually needs a bit more accuracy, but always remebering that 24h for a complete simulation on a home PC is the limit if we don't want a competition only accessible to professionals/small companies or people who have access to computing power at universities.

Currently the mesh consists of 6 million cells per half machine (symmetrical model): it would not be reasonable/possible to manage more than 8-10 million cells.

But is it just surface and growth rate? Or are there refinement boxes e.g. aroud the contact patches and the wings?

Also I had a look and the tyres don't appear to have a plinth around the cpatch. You probably want something at least 1n mesh cells tall around the base of the tyre.

Correction: there are 3 standard mesh quality options. The roughest one produces around 6 million cells, for half a car. The medium one produces around 13 million cells, for half a car. The finest one, officially used for race results, produces more (no idea how many because i never dared to launch such a simulation on my laptop).

And yes, there are refinement boxes. They encompass wings, bargeboards, diffuser, etc. But they're also in the car's vicinity and wake.
Still, to correctly model lower ride heights, a finer mesh should be used under the floor. To compensate, the refinement box in front of the car can be surely shortened.

The tyres have a LARGE contact patch, if you were referring to this. That's achieved by intersecting wheels and ground.

[LVDH]

The issues with the rules around this topic will get addressed with the new version of MFlow. I have to run my own experiments on this. We have two restrictions that we have to live with. Many of the teams use Windows machines for CFD. In this case, it means, we are limited to OpenFOAM 5.0. Newer versions of OF support better meshing strategies that would help esp. in the floor area. I have ideas on how to getting newer versions of OF running for the Windows people but this will probably have to wait until late in this year. The other restriction is that many people actually run their cars on laptops. The competition shall be open to a good number of people so the amount of CPU power and RAM needed will always be a bit limited. These compromises mean not everything will be super realistic. When I am done with the new rules, there will be limits on rake vehicle height and the contact patches might look a bit better.

[Ft5fTL]

Many of the teams use Windows machines for CFD. In this case, it means, we are limited to OpenFOAM 5.0.

Actually bluecfd have a 2020 build for windows that uses openfoam 8.

http://bluecfd.github.io/Core/Downloads/

[LVDH]

Many of the teams use Windows machines for CFD. In this case, it means, we are limited to OpenFOAM 5.0.

Actually bluecfd have a 2020 build for windows that uses openfoam 8.

http://bluecfd.github.io/Core/Downloads/

That is very cool. I will have a look at that, no idea how I missed this.

[variante]

The new CAD parts look like a nice step forward. But here are a few criticisms:
- there are useless internal geometries (rims, exhaust parts)
- rim covers are still flat, rather than sculpted
- brake intakes oddly intersect the new wheel deflectors (might as well eliminate them)
- weird alignment between battery/fuel cell and chassis (prepare to see radiators intersecting the fuel cell)
- empty space between battery cell and fuel cell (might cause meshing problems)
- RV_Vehiclebody still too narrow around the rear compared to F1 (radiator exhaust area)
- RV_Vehiclebody still going lower that RV_Floor_Body. Not sure how the rules deal with this.

[CAEdevice]

The new CAD parts look like a nice step forward. But here are a few criticisms:
- there are useless internal geometries (rims, exhaust parts)
- rim covers are still flat, rather than sculpted
- brake intakes oddly intersect the new wheel deflectors (might as well eliminate them)
- weird alignment between battery/fuel cell and chassis (prepare to see radiators intersecting the fuel cell)
- empty space between battery cell and fuel cell (might cause meshing problems)
- RV_Vehiclebody still too narrow around the rear compared to F1 (radiator exhaust area)
- RV_Vehiclebody still going lower that RV_Floor_Body. Not sure how the rules deal with this.

I answer only about the rims, because I haven't downloaded the template yet.

Rims will be "filled" with cylindrical volumes (if you can't find them in the template, maybe the staff forgot to include them), no additional cells (no cells at all) there should be inside the internal rim volume.

[variante]

I answer only about the rims, because I haven't downloaded the template yet.

Rims will be "filled" with cylindrical volumes (if you can't find them in the template, maybe the staff forgot to include them), no additional cells (no cells at all) there should be inside the internal rim volume.

Yes, but they're still consuming CAD processing power.