[KVRC] Khamsin Virtual Racecar Challenge 2016

[CAEdevice]

Hi guys,

last Friday Friendship Systems held a lecture at the university of Berlin.
Part of it contained my car which was partially optimized using their software CAESES.
They also encouraged the students to join the competition.
So with some luck we will have a few more entries next season.

Great! I'm going to publish an article about my 2015 car and Karalit, hope it can help too.

[CAEdevice]

I am entuhsiastic about the new rules! Thanks Chris.

About the questions:

- Should the rules on the visibility of the front suspension templates be relaxed or removed? >>> Relaxed (suspensions covered only when view from above and from a frontal point of view with a direction angle of 60°)
- Should the maximum height of the diffuser (320mm in 2015) be lowered >>> Lowered (280-290mm)

I would add another point: I am working on a parallel project (importing my 2015 car into a realistic driving simulator). I have the impression the the ride height (45mm or 40mm I don't rember exactly) is almost realistic if interpreted as "static", but at 100mph I would expect lower values (25-30mm).

I think we would have more realistic diffuser with a lower dynamic ride height (my latest version of the 2015 car, and some other opponents, was efficient, but quite far from real LMP1).

[CAEdevice]

PS: 2015 KVRC car could turn at Eau Rouge around 270 km/h or even more ...

[CAEdevice]

Some other things things I was thinking about while driving to work

It would be useful to provide a CFD simulation of a complete car (I can provide a complete one), to allow the partecipants to make a raw tuning of their CFD tools (es. OCCFD for Windows vs OCCFD for Linux, or other software).

About the race results: the post processing of the podium cars should be the same for every race. I would add some CFD pictures (pressure map, streamlines, surfaceLIC ...), but no sections.

Last idea: I would dedicate a page of the rulebook to the mechanic specifications of the car (not to be changed during the season or established for each race before the first race).

[CAEdevice]

Sorry for writing so much, but I have read the Appendix, in particolar:

Cooling
K4.1 option 1

KVRC 2016
Flow rate boundary condition of 3 m 3 /s will be applied
Difference in surface integral of pressure (inlets outlets)
must be >= 20 Pa.m 2 (whole car)

KVRC 2017 and so on
Flow rate: 5 m 3 /s
Difference in surface integral: 56 Pa.m 2

Is there a reason for the 2016 surface integral values? In my opinion they are too low, 2017 values are much more realistic (I would have guessed 80-100 Pa.m2.

[machin]

Agreed: the 2017 figures are indeed more realistic, however looking at the figures that were achieved by the teams in round 6 of the 2015 championship (The pressure integrals were measured for all cars to gather data but were not used to determine engine power) we realised there was a risk that it might be difficult for some teams to get close to the realistic figures... as a compromise, whilst people find their feet, we thought it best to introduce a lower figure for this year (whilst still maintaining some prescriptive limitations on minimum inlet/outlet sizes) and then increase that up to a more realistic figure for next year (with the inlet size restrictions being relaxed or dropped entirely).

Its a difficult balance: we want the championship to be realistic... but making it too realistic too quickly just makes it too hard... I've said it before: real F1 teams have hundreds of personnel designing cars full time and they don't get it right all the time.... we don't want to make the championship so realistic that you all need to employ a team of designers!

[LVDH]

Some thoughts on the new rules:

• Cooling ducts: Having the two rules sets could be a good compromise. But I suspect one of the two will dominate.

• The coordinate system is still wrong.

• A bit more space on top of the front fenders would be good. Modeling them was always a tight deal. Having louvers while respecting the 10mm rule was very difficult (for me).

• I would keep the front suspension cover rule. It seems realistic and was always a big challenge in the vehicle design.

[CAEdevice]

Agreed: the 2017 figures are indeed more realistic, however looking at the figures that were achieved by the teams in round 6 of the 2015 championship (The pressure integrals were measured for all cars to gather data but were not used to determine engine power) we realised there was a risk that it might be difficult for some teams to get close to the realistic figures... as a compromise, whilst people find their feet, we thought it best to introduce a lower figure for this year (whilst still maintaining some prescriptive limitations on minimum inlet/outlet sizes) and then increase that up to a more realistic figure for next year (with the inlet size restrictions being relaxed or dropped entirely).

Its a difficult balance: we want the championship to be realistic... but making it too realistic too quickly just makes it too hard... I've said it before: real F1 teams have hundreds of personnel designing cars full time and they don't get it right all the time.... we don't want to make the championship so realistic that you all need to employ a team of designers!

I also simulate what would have happened with 2015 cars: I obtained a NEGATIVE differential!

Code: Select all

The pressure at intake and exhaust are:

Engine intake, Area: 0.015m2 - Compliant
  Surface integral of pressure: 13.04 Pa.m2
Engine exhaust, Area: 0.010m2 - Compliant
  Surface integral of pressure: 3.98 Pa.m2
Cooling intake, Area: 0.129m2 - Compliant
Cooling exhaust, Area: 0.072m2 - Compliant
  Differential of surface integral of pressure: -27.06 Pa.m2

... but consider that the inlets were designed in order to be rule compliant, not to be realistic

Anyway: ok for the 2016 proposal!

[julien.decharentenay]

Coordinate system: Sorry for this. The directions are hardcoded in OCCFD and I am concern of messing them up somewhere. I could use a trick of rotating the geometry manually - but again there might be some errors...

OCCFD: There is a "linux" version. It is more a release of the raw ruby source code which can be downloaded at http://www.khamsinvirtualracecarchallen ... edirects=0. There is no documentation to make it work at this stage, but as a starting point, you need to install ruby (ideally a version 2) and do ruby mainOneClickCFDVehicle.rb -h to get the command line options. Yes it is all through command line....

Integral of pressure: I will do a write up of how we (more like Richard) came up with these figures. Essentially there are two changes compared to KVRC 2015: the cooling flow rate is increased from 1.5m3/s to 3m3/s (2016) and to the final 5m3/s (2017). And the template is replaced with a pressure constraint. The pressure constraint drives engine power in VirtualStopwatch... The low pressure integral is a result of having a look at the results we saw from the last race of KVRC 2015 (with a flow rate of 1.5m3/s).

[CAEdevice]

Hi Julien, I do not have the data folder with me, so I ask directly through the forum. Wich is the flow rate used in OCCFD 1.1 for Windows?

[CAEdevice]

When the rules will be defined, it would be useful to see the relation between engine power and cooling results, and to have the possibility to simulate the lap with Virtual Stopwatch using the actual power.

[CAEdevice]

ANNOUNCEMENT from the sponsor CAEdevice (myself )

CAEdevice will keep supporting the KVRC during the 2016 season, in two (I hope interesting) ways.

1) Final prize: the winner of the next KVRC edition will receive a 12 months subscription to the magazine Racecar Engineering



(If the staff could Tweet the news, maybe RE could dedicate a row to our races? )

2) Are you ever wondering how driving a KVRC car would be? Now you can : ) In the next weeks a model of the MP003D will be available for AssettoCorsa, one of driving simulators with the most accurate physics. The MP003D is derived from the car that CAEdevice used during the previous season (MP003B), with some improvements that make it an hybrid between 2015 and 2016 KVRC cars.

[machin]

Nice one CAEDevice! Good choice in magazine!

Some thoughts on the new rules:

• I would keep the front suspension cover rule. It seems realistic and was always a big challenge in the vehicle design.

Personally I would tend to agree: keep the cars looking like LMP's this year and then maybe switch to RedBull X1 style cars (narrower cockpit, visible suspension) next year...?

[CAEdevice]

Nice one CAEDevice! Good choice in magazine!

Thanks to you for the precious help in reproducing the physics of the KVRC cars.

[machin]

No worries.

When the rules will be defined, it would be useful to see the relation between engine power and cooling results, and to have the possibility to simulate the lap with Virtual Stopwatch using the actual power.

Actually... if you head over to the Test Track page you'll find those things already incorporated (I just hadn't announced it!):-

http://www.competition-car-engineering. ... KVtest.htm

Power (as a %) is shown in the bottom of the data window at the bottom of the page.

I've made some other changes from last year; particularly to the tyre model which now shows some deterioration over the lap depending on the forces going through it... to be honest it shouldn't make much difference except possibly at the Nurburgring which is a relatively long lap.

Speaking of tracks, at the moment the page shows the 5 tracks used in 2015, I'll definitely add one more track so we don't have to do any repeats, and it is likely that we will change one or two of the others.

One other thing I wanted to make clear: the Virtual Stopwatch page link above is really just a "cover page" which works as a sort of "results look-up portal"; so the full mathematical model isn't present online. There's a couple of reasons for this, firstly: you don't need to download the full program every time you load the page (i.e. it is much quicker). and secondly; it means nobody can, errrrr, "borrow" all the hard work that has gone in to the data behind Virtual Stopwatch!