Hypermesh: Unwanted reference system conversion from node based

Altair Forum User

Hi everyone

I am new to this forum. I am using HW 9.0.

The current issue that I have is regarding reference coordinate systems. I have an assembly (A) which is composed of several parts connected by rigid bodies which have well defined spatial locations.

The model A was set up with several coordinate systems and nodes correctly positioned. All components were individually modelled on node based reference coordinate systems. This meant that they could be really simply located by importing into A and using the 'POSITION' function in the TOOLS panel to transform the origin, x and xy nodes of the coordinate system and hence transform the entire component to the correct spatial location.

Model A in turn needs to be included in a number of similar but different models. So similarly I ensured that all of the systems' driving nodes in the components of A cascade off a master coordinate system (node based). This allows me to simply transform these 3 nodes and the whole assembly will follow.

This was working fine, however this brings me to my current problem:

When I review the systems in my model A now, the driving nodes have been deleted somehow and all the systems have reverted to fixed systems. Although all the references are still correct, this will cause me a problem when trying to use POSITION, as the master nodes no longer exist, and the option to transform a coordinate system directly is not available in the POSITION function.

It is not a realistic option to use translate or rotate, as it is a composite affine transformation consisting of several translations and rotations about different axes (which I would prefer not to calculate by hand).

My workaround at the moment is to use the master CSYS to define nodes, then create a coincident CSYS, then transfer the references accordingly and delete the original master CSYS.

I am curious as to how or why this could happen.

I would appreciate any comments that you may have.

Thanks and regards

Andrew Sims

UNSW

Australia

Altair Forum User

An update and workaround:

OK this is getting wierd.

I performed the following precise actions:

1. Translate CSYS 1 by x=10mm

2. Reject before accepting the translation. This moves the model and then puts it back where it started.

This restores the driving nodes of the coordinate systems. When I review them now, the origin, x and xy nodes are shown. However for some of the systems they are offset and/or rotated in no way that I can see rhyme or reason for.

Then I tried importing this into a new model and positioning the master CSYS nodes in the correct place, everything then moved and updated according to the references, however due to the offsets and rotations, some components moved with respect to the master CSYS and of course from the CSYS with rotated nodes, it is now rotated which will throw off the principal inertias of a nodal rigid body in this coordinate system.

Hmmm.

I have been able to get a workaround. This is done by translating the defining nodes of the CSYS to their correct locations. Associativity is then correct and transformations of the models are correct.

So I can move on, but it does not explain the issue. It was a subtle error which could have easily gone by un-noticed.

I guess sorry for the long posts etc, but it looks like I have found the resolution. Will be aware in the future. I am still interested if anyone has any comments about this.

Thanks

Andrew Sims

UNSW

Australia