[MVRC] Mantium Virtual Racecar Challenge 2016

[CAEdevice]

JJR's diffuser really looks stunning. That 3d expansion requires really great modeling skills and incredible CFD optimization to make it work. Again quite different concepts. I really look forward for the results.
It is a shame that Variante did not enter this race. I was counting on him being second to grasp the championship

I would have liked Variante on the podium too...

[rjsa]

Well you get numbers and plots for everything you monitor. So if you did not monitor something by not using special_bc or monitoring surfaces, nothing showed up. But are you sure you get CoP? And you can use as many monitoring surfaces as you want. If you are for example interested in the underbody flow, add one there.

I've always had the monitor there. It just started showing as calculated flow now. It only reported pressure before.

What I was getting:

Code: Select all

Integrated Force Coefficients
Cd 	Cl 	Cl_front 	Cl_rear 	Cd*A 	Cl*A 	Cl_front*A 	Cl_rear*A 	A
1.01 	-2.44 	-0.679 	-1.76 	1.01 	-2.44 	-0.679 	-1.76 	1 m^2


mSurf_h_monitor_right_2d
-0.533931 Pa*m^2 

This block appeared in the last report:

Code: Select all

Cd*A 	Cl*A 	Cl(frt)*A 	Cl(rr)*A 	CoP
1.01378 	-2.4371 	-0.67958 	-1.75752 	2.16345 m
mSurf_h_monitor_right_2d
-1.0663[m^3/s] 

It just appeared there. I can't think of anything that I did to cause this.

The upper numbers are from the post-processing tool that also creates the images. The lower numbers from the older version without images. The newer one has a bug as it outputs the wrong physical dimension. Also it does not automatically double the values any more when using a symmetry plane. I took it out as most people started using names like ..._left and ..._right. I thought when the surfaces is divided doubling does not make too much sense any more. But I see the confusion I created. I will think about something better.

But both are in the HTLM report. I did run the old one before I ran the new one. How did it get into the HTML?

Anyway, are the numbers on the old report correct?

I hadn't found the CoP position anywhere in the reports until I've seen this number, so I'm going on with it.

[LVDH]

The tool that generates the report searches for anything that could belong into it. So if you execute both post-processing tools it will find both results. Maybe you just copy pasted a finished case and had the results of it in the new case because of that?
The reason for it searching for useful stuff is that in the full version of MantiumWFlow (it will be renamed soon) is that it enables the users to create their own custom post-processing utilities if they like.
The numbers might not match 100% percent as I think I slightly changed the number of iterations used to calculate the averaged results. In a fully converged simulation it will not change anything. But even with slightly changed numbers I have never seen the lap time change significantly and never enough a affect the race position.

[Alonso Fan]

Well that's a record high 6th place confirmed for me... Shame not many are entering.

[CAEdevice]

Well that's a record high 6th place confirmed for me... Shame not many are entering.

Next year I would propose to give double points for LeMans race. There three reasons for that:

1) Prestige of this track
2) The best car in the last race is the best car of the efficiency tracks
3) It will be a way to avoid loss of interests

[RicME85]

Double points won't make a difference to loss of interest.
Things have got too difficult too quickly for some of us. Whilst improving the CFD tools and the way things are measured is good for realism, actually getting your head around stuff, understanding all the results and being able to model changes that work isnt easy and is time consuming.

I've given up because it's no fun trying to solve the engine flow issues whilst also trying to balance CoP...and that's with using the car you supplied Matteo. Just using the geometry you have produced results in massive engine power deficiencies in every test I have done and that is before adding meaningful to the car.
I am certain this is mostly a Sketchup STL export issue as the geometry of yours that I have didn't have the HX monitoring surface so I created it and have had zero luck with the flow.
I know now is the time for the inevitable 'learn a proper CAD tool' argument but Im not going, I have explained why enough times.

I may investigate the issue at some point in the hope of joining in again next season but as things stand I wouldn't hold my breath, I'm having too much fun playing video games with the wife, going to concerts and watching Derby County to want to spend any time on Sketchup/MVRC as rather than being something fun it has become a chore and more like a thankless job.

Good luck to everyone, I will at least be following things to see what crazy ideas everyone comes up with next.

[CAEdevice]

I am not sure that it was a CAD issue: despite the original open source car had good cooling performances (2.8-3.1m3/s ), I noticed that small changes in the front wing angles could cause very different results. Also consider (now I can reveal it) that the "wing shaped" rear suspension cover helped to extract air from the cooling exits.

I think that it is a philosophical decision: next year I could provide a complete car, ready to race, with only rear wing and front flaps to be tuned and a second incomplete version with more design freedom.

About CFD realism: I agree it makes the design more difficult, but it is the reason I like it. I am a professional in a very different field (hydraulic presses, steel structures, stairs, CNC machines, ...) , but I know that when design meets reality it becomes very complex.

[CAEdevice]

I am sorry about Variante surrender too. I think that MVRC could/should compete with Formula SAE in the future. It could be an interesting competition for departements that deals with CFD (aerospace), less focused on car dynamics and more on CFD (a more detailed CFD model and simulation could help), but not limited to the academic world (in my opinion a weak point of Formulae SAE).

[LVDH]

Hi, I have already also started thinking about what to do next year. We can soon start a thread about it.
I understand RicME's frustration about things getting difficult. What he explains is what I found so cool about aircraft engineering (although after university I started in the automotive industry (my first car was recently revealed)): Improving one aspect of something will affect others and usually not in a good way. This is what we now have in this competition. You have to get drag, down force, the balance, cooling flow correct while not overlooking some rule and getting a penalty. This championship is a very cool challenge. Getting good results is a matter of putting a lot of time into the design. Either at night thinking about some new ideas or then at the computer trying to tell him what you want.
And on one hand I agree with Matteo that this championship should at least reach an audience of (almost) engineers at maybe 10% of FormulaSAE. And I was at some points planning on more championships which would be simpler than MVRC. On the other hand I have a suspicion that MVRC and related championships will never be much bigger than what we have.
As many of you have noticed this is difficult, very difficult and very frustrating at times. Frustrating stuff does not appeal to many people (esp. nowadays). Look at what happened with racing simulation games. I remember the times playing GP2 and spending tremendous time with the telemetry and car setup. These kind of hard core games have disappeared for quiet a while and I am not sure about the new "super realistic" racing games of today.
Also when you look at GrabCAD, they also have challenges and I looked through them the other day. I would expect them to have a large audience and people interested to submit something. You get good prizes over there. But the winning submissions do not impress me and the number of submissions does not either. And this is just CAD, no CFD.
So I almost think that this here will be a somewhat small championship for a while and I have been convincing many people to join. I often visit engineering companies and they have the people who like something like this. Who is new? Some people wanted to join but did not get their first car done.

But, I do not give up and want to see this here grow. So what are my current plans? Next year will be similar to this one. Some new twists will be introduced so you guys have to design new and more exiting cars. More on that later. And we need the intro car back. Also (I hope somebody agrees to help me) we need something like a flyer for the championship. This could then be laid out in universities or wherever you guys hang out and maybe find us some new contenders. Other ideas on how to market our championship better are welcome.

[RicME85]

I am not sure that it was a CAD issue: despite the original open source car had good cooling performances (2.8-3.1m3/s ), I noticed that small changes in the front wing angles could cause very different results. Also consider (now I can reveal it) that the "wing shaped" rear suspension cover helped to extract air from the cooling exits.

I think that it is a philosophical decision: next year I could provide a complete car, ready to race, with only rear wing and front flaps to be tuned and a second incomplete version with more design freedom.

About CFD realism: I agree it makes the design more difficult, but it is the reason I like it. I am a professional in a very different field (hydraulic presses, steel structures, stairs, CNC machines, ...) , but I know that when design meets reality it becomes very complex.

I tested without making changes to the car at all other than adding the hx slice.

[RicME85]

Andre and Matteo, what you have to understand is that not all of us are professionals with a background in this subject nor are we all students in this field, we started doing this stuff for fun when it was originally but a competition on aesthetics (Formula Sketch) and then took it further when Julian and Nick sorted out KVRC and from there it's snowballed to where we are now with MVRC.

I applaud all the work Andre has done but we really need the intro class back.

[LVDH]

but we really need the intro class back.

That is true, I also want it back for next year.

[CAEdevice]

Andre and Matteo, what you have to understand is that not all of us are professionals with a background in this subject nor are we all students in this field, we started doing this stuff for fun when it was originally but a competition on aesthetics (Formula Sketch) and then took it further when Julian and Nick sorted out KVRC and from there it's snowballed to where we are now with MVRC.

I applaud all the work Andre has done but we really need the intro class back.

I can understand what are you saying. Next year I will release a complete car and (maybe, it depends on the free time I will have) a complete virtual machine with all you need to simulate the car.

The question is: would it be funny anyway?

About involving professionals: André is right, the problem is to invest time. Professional are interested if they can spend pictures, results and victories as advs and Linkedin contents. My case is slightly different (I am not a CFD specialist, I use MVRC to learn), only with a bit of passion I could insist after so many 4th places

At the moment I can say that the most important thing I learned is that CFD is an hugely extended field and not so easy to deal with.

[CAEdevice]

@André: yes, we need a kind of brochure, but also a PowerPoint and some contacts in the right departments (I have some in Milan Politecnico maybe Variante can help too).

I can provide a more "stylish" version of my car you can use for producing demos if you need it.

[RicME85]

Andre and Matteo, what you have to understand is that not all of us are professionals with a background in this subject nor are we all students in this field, we started doing this stuff for fun when it was originally but a competition on aesthetics (Formula Sketch) and then took it further when Julian and Nick sorted out KVRC and from there it's snowballed to where we are now with MVRC.

I applaud all the work Andre has done but we really need the intro class back.

I can understand what are you saying. Next year I will release a complete car and (maybe, it depends on the free time I will have) a complete virtual machine with all you need to simulate the car.

The question is: would it be funny anyway?

You mean would it be fun using a complete car? Yeah it would be more fun than starting with a blank slate. I still think the intro class should come with a couple of 'bolt on' parts that are interchangeable ie a high, medium and low downforce front wing and the same for the rear. This would allow easier start up for complete beginners and would be a good base for those who have got the basics down but want to advance up the grid. Even with these pre-supplied wings you would get enough differences between the cars due to the combination of the wings and AoA entrants may use. Could have different nose solutions too, a high one like AlonsoFan's or a low one like yours. A wide diffuser exit or a narrower one. A more aggressive diffuser profile or a gentler profile.