FreeCAD as pre-post processor for MBDyn

[josegegas]

Wonderful @josegegas. It works perfectely !

Did you manage to put together the whole simulation?

[SuperMarty]

Yes @josegegas ! Everything went well after this glitch at start.

Now I understand that the "cilindros" you added are just a way to place the structural nodes.

We already have several tools to create well placed and oriented Local Coordinate System (LCS).
Please, could you consider adding in your workflow the possibility to snap a structural node to an already created LCS ?

[RatonLaveur]

Ah yes, that sounds like a great plan. Why not use some of the Assembly4 tools directly ? afterall, if anyone masters the LCS, it's Zolko!

[barrti]

Jose

I'm using the latest workbench but I'm starting with your Crank-Rocker-Four-Bar linkage model.
I'm following your "FreeCAD + MBDyn crank rocker four bar linkage" from scratch tute on Youtube ( April 7)

Most of it so far is good except that I just can't get the draft WB tool to re-orient the axis for the joints ( see Time 4m.01s on the video https://www.youtube.com/watch?v=QRWepZ2JPEs&t=98s ).

I have selected the snap in the Draft WB to snap to centres of circles but it just never works.

I can manually move the position of z:joint 2 for example, by typing in co-ordinates for the x,y and z but your method is obviously easier and accurate

Any ideas?

[SuperMarty]

Hi @barrti,

I had the same kind of problem with the gyroscope. I succeeded by removing almost every snaps except : snap lock and snap center.

Hope this helps !

[josegegas]

Jose

I'm using the latest workbench but I'm starting with your Crank-Rocker-Four-Bar linkage model.
I'm following your "FreeCAD + MBDyn crank rocker four bar linkage" from scratch tute on Youtube ( April 7)

Most of it so far is good except that I just can't get the draft WB tool to re-orient the axis for the joints ( see Time 4m.01s on the video https://www.youtube.com/watch?v=QRWepZ2JPEs&t=98s ).

I have selected the snap in the Draft WB to snap to centres of circles but it just never works.

I can manually move the position of z:joint 2 for example, by typing in co-ordinates for the x,y and z but your method is obviously easier and accurate

Any ideas?

Not sure what may be happenind. Sometimes it may be tricky to select an entitie from the 3D view. Sometimes I switch the view mode to "wireframe" so that it may be easier to select edges and not faces.

[josegegas]

Ah yes, that sounds like a great plan. Why not use some of the Assembly4 tools directly ? afterall, if anyone masters the LCS, it's Zolko!

I´ve been thinking about using LCSs as nodes for a while. I can see the possible advantages of using the tools that already exist to place LCSs at different locations given some references in the 3D geometry. In adition to this, what could be other advantages?

So far, using the tools of the Draft workbebch to place nodes at different locations has worked for me. I guess the key question would be, what advantages could LCSs would provide over the placing tools one can find in the Draft workbench? One usually wants to place nodes:

*At the center of mass of the bodies.
*At the same location of other nodes.
*At the center of a circular face, such as in a cylinder.

These can be achieved by the "snap" tools of the Draft workbench.

Another possible advantahe of using LCS would be that it may be easier or more intuitive to place nodes this way...

What do people think? What could be the next steps to follow with this project? If I will use LCSs as nodes, I prefer to do so before I start to implement more functionalities, otherwise it would be more work to migrate to LCSs once I have written more code.

[barrti]

SuperMarty,

Thanks, that helped. If i have more trouble, I'll use the wireframe suggestion from Joseegas

cheers

[RatonLaveur]

ping

I think Zolko may see whether there is (or not) value in using LCS instead of parametric points/nodes.

My intuition is that it may enable cross-functionality between the Assembly4 module and MB-Dyn, therefore essentially allowing a full FreeCAD Design-Assembly-Simulation workflow. But there must be a level of cross-functionality with A2+ or Assembly3...or future assembly results. Therefore, one question might also be "is using LCS instead of simple nodes more difficult/challenging when considering other sources of assembly than simply Assembly4?"

Hence, really there are two aspects:
- Are the LCS placement tools better/more parametric than the draft nodes?
- Is using LCS instead of nodes more complex/less cross-compatible with all types of assemblies?

If the answer to these questions are Yes and No, then I'd say go for it, this would streamline the model setup phase of your workbench, and place more emphasis on the simulation itself

[Zolko]

I think Zolko may see whether there is (or not) value in using LCS instead of parametric points/nodes.

- Are the LCS placement tools better/more parametric than the draft nodes?

for me, Draft snap is not parametric at all: when I create a cylinder, and place a point in the center of the face with a snap, and later move the face/object, the point doesn't follow. As far as I understand, Draft is not parametric: you draw what you draw, and there it is, but objects are not constraint. You CAN draw a point in the centre of a circle, but the point is not constraint to be the centre of the circle: if you move the circle, the point doesn't follow.