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

[LVDH]

The data sheet doesn't show up properly in any browser. A screenshot of your excel sheet would do it.

This has been bothering me for years now. I finally figured out why this happens. However, I cannot fix it right now, so this is my pro solution:

Will the CoP issue (including the single components numbers) resolved in the next MFlow release (for Singapore)?

Yes, I will most likely release the new version today.

Also, the CoP calculation on my car is surely wrong (big time). There must still be a tiny evil bug in the software.

This what what I am somewhat afraid of. I have gone through the math again and again the last weeks. It should be correct.

[LVDH]

Looking at the numbers indeed suggests that OK Racing's closest followers have more low hanging fruit to improve the car.
Anyway, I have uploaded the new MFlow version.

[rjsa]

I had COP 1.88 on my last full run.

[jjn9128]

Okay so the proper sim gave me more drag and more downforce, the variance is 0.079 on CxS and -0.067 on CzS, but with a 3% difference in the balance (0.1m). Interesting that my cooling didn't seem to work as planned, I'd been really conservative with it. That said I have an inkling about an issue.

[LVDH]

I had COP 1.88 on my last full run.

Okay so the proper sim gave me more drag and more downforce, the variance is 0.079 on CxS and -0.067 on CzS, but with a 3% difference in the balance (0.1m). Interesting that my cooling didn't seem to work as planned, I'd been really conservative with it. That said I have an inkling about an issue.

Both of you used the long-option and are sure that we are talking about the geometry you submitted in the end?

[wb92]

Congrats for OKRacing, nice job.

For me main point was to 'not be the last', and that's accomplished. I've found out what's wrong with my cooling and some other points of my car.

Regarding overall numbers - it's very interesting that rear gives lift (!) instead of downforce. I kinda still don't believe what happened with rear when I saw official post-pro images.

For proper calculation - is it better to use "MVRC" or "MVRC-long" option?

[Thomas2019]

Congratulations to the winner!

I have worked on an evolution of my car the last days (Team Baltic) using a different OpenFoam version (since I cant get MantiumFlow to run) but mimicking the settings as closely as possible.
Seeing the official values now, they differ significantly from my results I used for "optimizing". Cl(f)&Cl(r) of -1.25 & -0.07 officially vs. my values of -0.6 & -1. So this variation is obviously way too large. Rather sobering. So I will take a break. Maybe go over the settings once more with a fine comb. But that difference is worrying.

Thomas

[machin]

I’ve updated Virtual Stopwatch for the Singapore Track...

This is a track like Monaco; the corners are slow, so downforce doesn’t have a big impact... however since the straights are all quite short there is also very little penalty of high drag... so the outcome is that you guys should aim for a high downforce set up.

Virtual Stopwatch MVRC Timing

[CAEdevice]

Thanks Richard.

I was looking for two information in the threads of previous editions, but I could not find them.

The first is the CoP correction: how much can it be corrected? The second is the relation between max available power and cooling flow. I remember that a part of the engine power is available even if the cooling flow is zero, but I don'remember how much.

[machin]

Yes, to stop anyone getting too disheartened we set the minimum power to 40% even with no cooling flow... however in reality if you’re down at 40% power you’re not going to be competitive, so it’s almost irrelevant....

Andre adds a COP correction, I think it is a maximum of 0.1m in the direction of the optimum COP... so that represents the small wing adjustments a team would make at the track.

[Continues]

Thanks everyone! I'm currently looking for an industrial placement, need to add this to my CV right away

After seeing the times I expected the deciding factor was engine power with my car being the only one in the 90% range, but some of you got the same or better cooling in much tighter package, really impressive.

I was rather worried, because everything I tried since then made my car slower. I might end up going back with only minor changes.

Good to see Singapore GP available in the Virtual Stopwatch.

[machin]

Thanks everyone! I'm currently looking for an industrial placement, need to add this to my CV right away

Yes the competition is great for that sort of thing; you can demonstrate your CAD skills, experience with CFD, knowledge of fluid dynamics, iterative design, and I’m sure many other things. I like to think the competition helped Nicky get his job with Force India... and we’re just waiting to see where Variante’s career will take him (no pressure )...

[rjsa]

I had COP 1.88 on my last full run.

Okay so the proper sim gave me more drag and more downforce, the variance is 0.079 on CxS and -0.067 on CzS, but with a 3% difference in the balance (0.1m). Interesting that my cooling didn't seem to work as planned, I'd been really conservative with it. That said I have an inkling about an issue.

Both of you used the long-option and are sure that we are talking about the geometry you submitted in the end?

The same geometry, exact same folder. The config I used was the MVRC. If I have to use the long to get reliable results it’s going to get very tough.

[CAEdevice]

Thanks everyone! I'm currently looking for an industrial placement, need to add this to my CV right away

You should write about your victory and developement on LinkedIn too.

Meanwhile... congratulations and let's what happens in Singapore

[LVDH]

For those who missed it, I did release the new version yesterday.

I had COP 1.88 on my last full run.

Okay so the proper sim gave me more drag and more downforce, the variance is 0.079 on CxS and -0.067 on CzS, but with a 3% difference in the balance (0.1m). Interesting that my cooling didn't seem to work as planned, I'd been really conservative with it. That said I have an inkling about an issue.

Both of you used the long-option and are sure that we are talking about the geometry you submitted in the end?

The same geometry, exact same folder. The config I used was the MVRC. If I have to use the long to get reliable results it’s going to get very tough.

You car only starts converging after 3000 iterations, so if you used the MVRC template you will get confusing results compared to the MVRC_Long option. You and others should always check the Cm plot. If it is not flat at the end, you need to run longer. Some cars converge nicely in way less iterations but some simply take longer.

Thanks everyone! I'm currently looking for an industrial placement, need to add this to my CV right away

Yes the competition is great for that sort of thing; you can demonstrate your CAD skills, experience with CFD, knowledge of fluid dynamics, iterative design, and I’m sure many other things. I like to think the competition helped Nicky get his job with Force India... and we’re just waiting to see where Variante’s career will take him (no pressure )...

Before the season started, I was working on a certificate for last years champ, Variante. I still have not finished it. The idea is for you guys to have something to show future employers, as winning this championship definitely proves great skills as aerodynamisist.

Yes, to stop anyone getting too disheartened we set the minimum power to 40% even with no cooling flow... however in reality if you’re down at 40% power you’re not going to be competitive, so it’s almost irrelevant....

Andre adds a COP correction, I think it is a maximum of 0.1m in the direction of the optimum COP... so that represents the small wing adjustments a team would make at the track.

The CoP correction is only 5cm in your advantage. It is carry-over from the old KVRC days. I just kept it in.


I know that many of you are questioning the CoP calculation in general. At the beginning of the season, the l_ref used to calculate it was wrong as it was using the wheel base (which should be the thing to do), however there is a semantic error in OF, my workaround is to use l_ref=1, as this cancels out their mistake. I think this dumb issue destroyed quite some trust for you guys. Also some of you are only running the MVRC_fast or MVRC template, while I use MVRC_long for the races. Also I know how it is to submit a car last minute. Mistakes get made. You might be expecting a certain result, but submitted untested or the wrong CAD data. In addition, I often have to fix some of your files, this can also affect the results. So there are reasons, why the official results can differ from what you expect.

But I will once more go through all the math, right from the beginning and make sure that everything is correct. If I do indeed find an error, I will fix it in MFlow and release a new version today. For the coming race then, I will keep the current and the potential new CoP calculation in MFlow and you guys will get to use the one in your advantage. For the following race, I will then hopefully figure out which one is correct.