[MVRC] Mantium Virtual Racecar Challenge 2017

[etsmc]

The front looks very cool. Are those cooling intakes? I wonder what the Cp surface image of that view looks like. I hope a lot of red at that splitter.

I will be presenting our championship at the CAESES user meeting in three weeks (the OF guys rejected me...). If you have cool images or other material I could use, please send it to me. It will help make your virtual race team more famous.

Not as read on the splitter as I was hoping for.
I think the intakes at the front are too close to the top surface of the splitter, will move them up slightly and see what happens. The issue with the intakes though is making them so you can't see the front suspension template.


Will test the new layout with the new version of the software and see how it does.

[CAEdevice]

Hi, I tested the new release of MantiumFlow: it works perfectly and the new post processing features are great.

As promised, later (I am on the train now) I will show a picture of the hx stream lines of my car.

Just some suggestions for LDVH

1) Streamlines seeds are too thick sometimes (*)
2) I would be great to have some template-managed options (as with slices), although many things can be easily managed by editing the "streams" file ("... WorkingDirectory \ system \ streams")
3) The surfaces from which the streamlines start are visible (ex. I added a monitoring surface in fornt of the rear wing) in all images, it would be better to hide them or make them semi-transparent
4) In the case of hx streamlines, it would be better to make the car body semi-transparent


(*) Is the number of seeds related to the number of triangles of the monitoring surface? In that case, except for hx monitoring surfaces, my advice is to use 25mm triangles instead of 10mm.

EDIT: here is the HX streamlines image of my car, I hope it can be useful for the intro class teams that use a similar concept. Maybe the other teams will be interested too

[LVDH]

1) Streamlines seeds are too thick sometimes (*)

Well, I want to have them much thicker, then they look nicer. Now they are only lines. At some point I might add the option to change to image resolution. That might help then.

2) I would be great to have some template-managed options (as with slices), although many things can be easily managed by editing the "streams" file ("... WorkingDirectory \ system \ streams")

Will come and is mostly available. I am just not sure if it will also be for the MVRC edition

3) The surfaces from which the streamlines start are visible (ex. I added a monitoring surface in fornt of the rear wing) in all images, it would be better to hide them or make them semi-transparent

Yes, I can add that if you like.

4) In the case of hx streamlines, it would be better to make the car body semi-transparent

Some cars look strange in the transparent view though. I think it is because there not all triangle or oriented the same. I agree, though.

(*) Is the number of seeds related to the number of triangles of the monitoring surface? In that case, except for hx monitoring surfaces, my advice is to use 25mm triangles instead of 10mm.

I am not sure myself as I did not have much time to test. If you can do that please let me know. You can add as many monitoring surfaces as you wish. They can also be on top of each other.

EDIT: here is the HX streamlines image of my car, I hope it can be useful for the intro class teams that use a similar concept. Maybe the other teams will be interested too

https://www.dropbox.com/s/2bt9qervp9evy ... s.jpg?dl=1

=D>

I remembered your question about the symmetry plane. Yes, it is always used. I have no idea how you can get forces on the left mirror. If something is reported, I can imaging it to be some random numbers. So how they almost match the right one is strange. If you did not make a mistake like having them both be on the right hand side overlapping each other, please send me the case.

[CAEdevice]

1) Streamlines seeds are too thick sometimes (*)

Well, I want to have them much thicker, then they look nicer. Now they are only lines. At some point I might add the option to change to image resolution. That might help then.

Yes, sorry for my poor english. I meant "too many lines" not "too thick lines".

About the number of seeds and stl triangles I will do a test today. The same for the right mirror.

[CAEdevice]

I confirm: the number of seeds (when creating streamlines from surfaces) is influenced by the tringales (STL) resolution (in the picture below, 50mm vs 10mm).

[rjsa]

Now can we turn of the extras, especially the streams? This last round of updates brought my computing time up 30%, it's now 3 times what it was with the 2016 software.

[RicME85]

Surely these new things are done in post processing so shouldn't have any effect in the speed of your simulations as that's done separately?

[CAEdevice]

Nothing changed in the solver, but you can try to use an alternative template (mvrc_fast.ini) for initial development and then the standard template.

[rjsa]

I have been using the standard template all the time. I found the fast option misleading sometimes and I gave up on it long ago.

Processing time has increased with the solver changes and the last post processing changes brought another 30% time increase.

I can tell you how much each run I did took and what happened when things changed. The 30% increase for the streams, every single run, is not worth it for me since I can get those out of paraview if needed. That's why I ask for a way to shut it down.

And BTW, I got HX reporting back but parts performance still won't make it to the report, it appears on solver.log only.

[CAEdevice]

the last post processing changes brought another 30% time increase.

I did not have problem or time increase with MFlow v8 or v9 compared to v7. Maybe your hardware has not enough RAM? I think 16gb or more is the best setup. I noticed that using 6cpu over 8 is also a good compromise.

[CAEdevice]

A question for LDVH: does the presence of monitoring surfaces influence the mesh?

[rjsa]

I have 16gb. I'll watch memory use, but there isn't much I can do about, it's a macbook pro and it's topped.

[CAEdevice]

By the way, thanks to MFlow improvements the challenge is becoming more realistic, do you note any analogy?

[rjsa]

LOL, perfect.

It did go processing until 16 hours.

I'd find it nice if we could get .ini file switches to turn off streams.

Getting this too:

Curent force name: BodyFull_se019
Plotting of forces failed.
Curent force name: Fins_R_se000
Plotting of forces failed.
Curent force name: Fins_se000
Plotting of forces failed.
Curent force name: HxBaffle_L_se000
Plotting of forces failed.
Curent force name: HxBaffle_R_se000
Plotting of forces failed.
Curent force name: RW_CCA_200deg_010a
Plotting of forces failed.
Curent force name: StreaksF_se000
Plotting of forces failed.
Curent force name: StreaksR_se000
Plotting of forces failed.
Curent force name: cfd_parts_suspension_100mm_032
Plotting of forces failed.
FSource folders: ['faceSource_engine_in_se015', 'faceSource_exhaust_R_se015']

[AratzH]

My turn!

I love the new streamline feature. In my case I didn´t notice any increase in run time...