My 'MVRC' run with the R02 car was very close to the official results, so it was more or less as I expected for me.
Yeah, that's pretty close.
But... both me and G-raph expected 0.2 CLA more (from our fast simulations). Also, G-raph has the same rear DF as in Race 1 despite a much bigger Rear Wing, which is unexpected even with a bigger Front Wing. Busor had that weird improvement i talked about. PurePower had underwhelming results, as himself reported. Same goes for CAEdevice. Koldskaal didn't even compute.
So, within the top 7-8, you're the only one who got it smooth. Maybe MaxTaylor too.
Yeah, I only run "MVRC template" for the last few years now. By results this year have been within 0.5% error CLA of the official. I'm impressed y'all do as well as you do with the fast!
I ran the "MVRC Template" today with the submitted car and i can confirm that "Fast" simulation results are rubbish. Race result match with the "MVRC Template" run.
Congratulations to GRaph and also Panthera!
Very cool to see how teams slowly can climb up the ladder in this championship.
Very big thank you to Max and Christian. I hope you guys liked the new feature in the video. I knew that it was coming and there might even be something a bit better in future, who knows...
I liked the "DNS". At least the wording.
Koldskaal's sim crashed. The main suspect is a part that was used in race one already. So I have to investigate this a bit further and also the other complaints, I see here.
It is interesting to see that some teams get the numbers as expected while others have larger discrepancies. Without having looked into it, I can suspect the mesh decomposition and maybe rogue settings files.
When you simulate on your local machines, you probably use something in the range of 8 CPU cores, while on our new cloud system, we run on 48 cores. OF can show different results based on how and into how many blocks the mesh gets decomposed. But this usually only very slightly changes the results.
In your submissions, I have noticed that many teams also submit a settings file. This usually annoys me, as I have to delete it, to make sure you are all running the same settings. So if you submit the file to me, maybe your are also copy-pasting it into your own runs and have even forgotten that you do this. So, please look into this.
Also, you can all send me the results, you expected, so I have better data to look into these questions.
Regarding Variante's protest:
First of all, protesting here in the forum is the way to go.
His monitoring surface does not match the front surface of the radiator volume as there is a radius in the volume, that is not present in the monitoring surface. My take is that adding a radius to a surface should not be an issue to anyone here that is part of this championship and has been warned about it. Also, while it is a super small violation that maybe did not even change anything, I just have to be a bit hard, as I do not like wiggle room between, it is kind of OK and this is too much. Handing out a penalty after a warning is clear to everyone - you and me. So it might sound hard, but fixing it would have been very easy. However, I am happy to hear arguments about this, defending either position.
I am sure that the violation by Variante does not affect performance, but if the penalty comes after a warning, I think this policy is strict but fair. The same happened to me with the double (overlapping) mirrors in 2023.
About the performance difference. I agree that mesh decomposition can cause differences (I sometimes use 2 cores and sometimes 12 cores and 1-2% downforce error may happen, but the difference between my 12 core simulation and the 48 core of the official one, with the same template, was almost 10%). I have the suspect of a missing component or a wrong parameter (the only parameter I use is ride heigth -20 mm).
His monitoring surface does not match the front surface of the radiator volume as there is a radius in the volume
Ah, I didn't even imagine this was the issue.
You're technically correct with that penalty, so I won't challenge it anymore, and give up my 2nd place.
To be clear with the curious ones: my cooling monitoring surface has some excess area which, however, is not in contact with any airflow, thus it gives me 0 (zero) additional performance.
His monitoring surface does not match the front surface of the radiator volume as there is a radius in the volume
Ah, I didn't even imagine this was the issue.
You're technically correct with that penalty, so I won't challenge it anymore, and give up my 2nd place.
To be clear with the curious ones: my cooling monitoring surface has some excess area which, however, is not in contact with any airflow, thus it gives me 0 (zero) additional performance.
Very much a technicality and a shame when they happen, but I think we've probably all fallen victim to a penalty like this over the years, I know I have
The thing is, besides the fact that this clear way of handling things removes a lot of subjective reasoning, since we have this strict rule, the cars have become much, much more rule complainant, which is great as it saves a lot of time on my side.
But, it was a hard decision, esp. as it might be taking out the only car that can challenge GRaph. I do hope though that more teams can catch up.
Andre, how easy is it to run these cases on the CFD cluster now? I'd be interested to give it a try with the free allowance they provide when signing up. After that I'd be curious to know how many MVRC runs you could fit into that time and how much each would cost thereafter. Maybe I'll drop them an email to find out
It is actually very easy, this time I ran all the simulations. And I think it is pretty much ready for you guys.
We will probably very soon organize a quick webinar for you guys to show you how it works. Then you can decide yourself if you want to try it out.
Let me know what you think.
The good news is that Variante's penalty should not affect the championship, as he needs to win the next 3 races in any case. So that will avoid a controversy.
It is actually very easy, this time I ran all the simulations. And I think it is pretty much ready for you guys.
We will probably very soon organize a quick webinar for you guys to show you how it works. Then you can decide yourself if you want to try it out.
Let me know what you think.
This sounds amazing and I'd be interested to try it.
On another note, do you know when we should expect the reports?
Sorry to be that quiet, I really struggle to make the time at the moment, which sadly also can be seen in my development.
For the discussion around the simulation results:
You are right, I didn't change much on the car apart from front and rear wing, thats also why my balance is still far off. I did however reduce my rake angle, as in the last race that didn't work at all. So that may explain some of my rear downforce gains.
I can't really promise much development going forward, let's hope for some more spare time.
I understood the reason for the difference between my results and the official ones. The parts of the suspension not included in the covers (therefore also the brake area) had been exported with a very rough stl mesh and this affected the results. My fault, but at least I fixed it.
I understood the reason for the difference between my results and the official ones. The parts of the suspension not included in the covers (therefore also the brake area) had been exported with a very rough stl mesh and this affected the results. My fault, but at least I fixed it.
Glad you could get to the bottom of the issue at least!
1) How flat the floor is
2) How limited the outer floor is in thickness and the plan suggests slots are going to be limited.
3) How raised bottom of the diffuser appears to be
4) No floor fences at all (other than the mandated in-washers)
5) No beam wing
6) Rear wing end plates are back
