Hi ACaputo,

 

To illustrate how I’d go about this problem, I made a Solidworks (SW) assembly of three geometrical shapes, this is what it looks like on an isometric view:

I then exported the assembly as a single *.STEP file and opened it in OpticStudio (OS):

The first (anecdotal) thing to notice is that the coordinage system is rotated about the Y axis by 180 degrees when using the isometric view. When importing a *.STEP file in OS, the origin in the STEP file is preserved. Meaning that in this particular case, it is still the barycentre of the sphere. We can confirm this by rotating the whole assembly using the Tilt About X/Y/Z columns in an X-Y projection for example:

I let you check that all the rotations are made about the sphere in the other projections (Y-Z and X-Z).

If you explode the assembly, and rotate only the pentagonal prism by 15 degree, this is what you would see in an X-Y projection (I kept the whole assembly in green and increased its transparency):

As you can see, the rotation is still done about the origin of the whole CAD assembly, or the sphere in this particular case. Let’s now see how you can rotate the pentagonal prism about its barycentre.

First, let’s look at a (poor) drawing of the assembly:

This shows the actual location of the pentagonal prism with respect to the assembly origin (the sphere). This is were it gets a little bit tricky, but bear with me. The idea is to insert a Null Object before the exploded pentagonal prism, and change the location of this dummy object (X/Y/Z Position) to the location of the pentagonal prism. In SW, the pentagonal prism is located in X = -1.0, Y = 2.0, and Z = -3.0 so we’ll use those exact same values in OS. Then, I’ll use the Ref Object colunn of the pentagonal prism and type in the Null Object number. That way, the pentagonal prism location is referenced to the Null Object. However, this moves the pentagonal prism from its original location. We bring it back by settings Pick-up solves with a -1.0 factor from the Null Object.

In the layout, everything will seem as if the pentagonal prism hasn’t changed. But we now have access to the location of the barycentre of the pentagonal prism through the Null Object. If you try rotating the Null Object about Z by 15 degrees this is what you’ll see:

I leave you to check the other rotations, and you can try to do the same for the cube. I’m attaching all these files to my answer.

I hope this helps. Take care,

 

David

Hi David,

Wow this is very thorough! Thank you so much for taking the time, I think this cleared up a lot. Playing with this file helped to put together what you were describing as well.

The big connection for me was the need to set the coordinates of the null object to be exactly at the coordinates of the desired object. I have this Zemax file provided by someone else, without my own access to the original solidworks file. So what this is telling me is there is no way to rotate this object about itself without knowing its real coordinates in the assembly, correct?

I just find it strange (perhaps due to being new to Non-Sequential Systems), that although the assemblies retain the positional data of each part by way of their placement, that it cannot be accessed or verified within Zemax. You actually need to go reference the Solidworks model for the numerical coordinates, is that correct?

This was greatly appreciated, thanks again.

Andrew 

Merci David

Hi Andrew,

Thanks for posting in the forum. As David has already covered the introduction and basics in great detail, let me pick up on the second question. 

Q: So what this is telling me is there is no way to rotate this object about itself without knowing its real coordinates in the assembly, correct?

Yes, there is no straight forward way of retrieving the local coordinates of a specific part of an exploded system within OpticStudio. The main reason for that is here is that the meshed CAD Parts have to be fitted into the modeling approach of OpticStudio, which in NSQ can be described as a hybrid between a “first-guess-meshed” and then equation driven surface. For more information allow me to reference to the follow articles:

How to import CAD objects – Knowledgebase (zemax.com)

How to explode a CAD assembly – Knowledgebase (zemax.com)

Impact of drawing resolution on ray tracing – Knowledgebase (zemax.com)

So, I can see two workaround approaches (WA) here:

WA1: David has already briefly mention that you could use rays and ray filter in order to see where objects and rays intersect and get the coordinates this way. But this is connected with quite a bit of effort.

WA2: I am not sure how complex your system is, but you won't need to go back to the actual design if you have the STEP assembly. Because other than OpticStudio can other CAD Platforms display the local coordinates of a part within an assembly quite easily. As an example, I took the STEP Assembly which David has created in SW and opened it up in Ansys SpaceClaim:

In red you can see the Global Coordinate Origin, which is in the middle of the sphere. And in green, you can see the two created local coordinate points for the other two objects.

Having said all of that, I agree with both of you that this information must be conveyed in some shape or form, otherwise OpticStudio would not be able to explode the CAD Assembly properly.

Let me run that past the Product Team quickly and see if they can help us a bit further here. We could potentially also file a feature request if necessary, to make this workflow a bit easier in the future.

Feel free to ask any follow-up question you might have,

Hi David,

This Check box reads in any optical properties that have previously been saved to the individual parts using the Save CAD Assembly/Part Properties tool. The CAD Assembly/Part Properties saves optical properties of an (dynamic!)-object back in that programs format.

So, this would only be useful with optical properties of the parts and subsequentially only be used with Creo or Inventor Parts (Parametric CAD Parts, Dynamic Link). Thus, I don't really see the point of having this within the STEP Explode dialog window. I`ll pass that on.

Best,

Hi Andrew,

I suspect this is the same problem but do you have an example to share for troubleshooting?

To centre a Source on a part of the assembly, you’d also have to know the location of that part with respect to the assembly origin.

Take care,

David