[MVRC] Mantium Virtual Racecar Challenge 2016

[LVDH]

The more important question is who will take care of design and structural analysis of the uprights? And will they enable such a flexibility with the Z-position of the suspension geometry?

[CAEdevice]

The more important question is who will take care of design and structural analysis of the uprights? And will they enable such a flexibility with the Z-position of the suspension geometry?

Hi, I fear to not have understand your answer once again

I was talking about an alternative standard (given) geometry of the front suspension, not about the possibility to design it.

[LVDH]

Look, it is not really worth it. Yes, it would be cool to have various variable geometry and the position of the heat exchangers affects CoG....
Maybe we can implement some of these thing for next season. You might have also noticed that I created a challenge homepage. There will be more challenges.

If you go through all the trouble of creating a new CAD geometry with FEM simulations and some other cool stuff that helps promote our cool little championship, maybe I will add your new geometry to the challenge. In the end wesley123 could very well be right and everyone opts for the same geometry and no real value to our competition was generated.

I have all the cars and I can tell you they all look cool and very different. I also have most results already. The standard positions have been mixed up a bit so there is no need to worry about too little action.

[CAEdevice]

Ok! I will find another sokution to clean the flow on that region

The other challenges are interesting: could you say more about them?

I am working on a single event CFD contest, maybe in late December or January (or in any date fsr from MVRC races, in order to not overlap). I would like to involve you (or Chris if he is interested) as partecipants. I am also trying to involve an external sponsor. But I will think about it after two or three races of MVRC, after completing the new car.

[CAEdevice]

About providing geometry for the MVRC: I am working on a customer release of my car, it will be available in about a week.

[TalnoRacing]

I would have a request for Chris: would it be possible to have a "pull rod" version of the front suspension to be choosen as an alternative to the "push rod" version?
(http://images.automoto.it/images/598951 ... errari.jpg)

I think we all face the same challenge in finding a clean solution in this area. Changing the front suspension will still present a challenge to us all, it will not change the competition as such and will not be more of a challenge.

I would rather look at adding front and rear brake ducts with control surfaces similar to the heat exchangers, where a specified airflow is required on a fixed surface. Basically taking the principle for engine cooling and applying it to the brakes.

[etsmc]

could anyone recommend what I can do about an issue I am having with my stl files?
I received an email saying that the quality was not good enough, I have been using the same process as I did for KVRC.
steps I take are as follows.
design half a car in SketchUp
export from SketchUp to stl
open the stl in meshlab -> duplicate -> flip -> flatten -> export stl
open stl in netfabb -> repair tool -> close all holes -> export stl

when I have a look at the stl in meshlab after exporting it from SketchUp it looks ok with a nice smooth surface
when I look at the file that has been exported from meshlab after duplicating and flipping it still looks smooth
the file that is exported from netfabb after the holes are closed is where I have the issue, the surface is no longer smooth ("looks a bit like it was carved out of a piece of wood." - Andre)

I have looked at the files that i submitted for KVRC and they have this same rough surface.

What am i doing wrong or is there a better way to do what i am doing? is anyone else using SketchUp? what are your processes and do you have similar issues?

[CAEdevice]

I have the suspect (remembering KVRC cars) that sometimes the original geometry (before treating it with Meshlab) is critical. If you don't mind I see your design, I can take a look to you stls. I know it is difficult, but I would use other tools than SketchUp (FreeCAD, or a Student Edition of a commercial package). SketchUp gives the illusion to be easy to use, but, if you want to use it at an higher level (es. Variante) it is as difficult to learn as ant CAD.

[etsmc]

I have the suspect (remembering KVRC cars) that sometimes the original geometry (before treating it with Meshlab) is critical. If you don't mind I see your design, I can take a look to you stls. I know it is difficult, but I would use other tools than SketchUp (FreeCAD, or a Student Edition of a commercial package). SketchUp gives the illusion to be easy to use, but, if you want to use it at an higher level (es. Variante) it is as difficult to learn as ant CAD.

do you have somewhere i can upload them too? happy to send them and see what you think..

[CAEdevice]

Please send me a pm with your email, I will answer with a Dropbox file request (you don't need to have Dropbox). I can read both STL and Sketchup, but I prefer STL.

[etsmc]

Please send me a pm with your email, I will answer with a Dropbox file request (you don't need to have Dropbox). I can read both STL and Sketchup, but I prefer STL.

you have a pm

[CAEdevice]

Hi etsmc, I had a look to your STLs (I will send you an email privately with the pictures of your CAD model showing the critical areas). Here are my remarks:

1) The original geometry is critical: you used some strange variable curvature radius and there are many cusps if you cut the models into sections
2) It is ok to use MeshLab or Paraview to decimate the triangles
3) If the native software (SketchUp) is not able to generate a closed STL, there is no reason to think that NetFab can do it (btw: NetFabb has been acquired by Autodesk too?)

NetFabb actually generates a closed STL, but it "guesses" too much around the critical areas, this is an example:



What you have to do, is simply to design a correct and clean geometry with SketchUp or with other software (a more "honest" CAD tool than ShetchUp would have rejected that geometry) in order to not use NetFabb at all.
(You would not need to use NetFabb with a clean geoemtry, but 9 times over 10, it would be better to use NetFabb before decimating the mesh with MeshLab). More details and pictures by email

[LVDH]

Thanks Matteo, good work.
Later I will check his geometry before it entered Netfabb. It does not look too bad. Maybe it is good enough for OF.

[CAEdevice]

Thanks Matteo, good work.
Later I will check his geometry before it entered Netfabb. It does not look too bad. Maybe it is good enough for OF.

My CAD finds many gaps in the geometry before it entered Netfabb... I am not an expert about SnappyHexMesh, but none of my FEM mesher for structural analysis would accept it (except a cartesian mesher and other exotic ones)... but I have finger crossed

I am pessimistic, but you can try using NetFabb before MeshLab and then trying with a "smoothing" function before decimating ...

[CAEdevice]

Coming soon...



CAEdevice MP004C "open edition"

- carbody simplified to make the modeling easier (and to not reveal too much)
- cooling inlets/outlets, heat exchanger completely legal (the areas are bit bigger than the minimum reuired surfaces: it should be safer, but it depends on what you will put in front of the car... )
- theoretically customizable only with extruded cuts and protrusions
- MVRC (Wflow) STLs folder structure provided
- DF/DR ratio around 3.0 (out of the box)

I am waiting for the results of the legality check of the original car (MP004B) before making the download available.
My aim is to make the modeling easier, providing a good base for the cooling ducts and general body.

EDIT: does SketchUp accept STLs?