View Single Post
  #5  
Old 06-18-2008, 11:58 AM
Steve Steve is offline
Registered User
 
Join Date: Mar 2003
Location: Alabama
Posts: 309
Re: dynamic modeling vs. synchronous technology

Quote:
Direct editing on imported geometry is fairly straight forward, but what about direct editing of a native model with a full history tree? What happens to the history tree? I have used a few history-based systems that have direct editing, and with those systems the direct edits were actually recorded in the tree and have the typical parent/child relationship. After a while the tree became much more complex and reordering the tree was almost impossible. I am not sure how they do it in SolidEdge, but I bet it is not too different. The edits would have to be maintained in the tree, otherwise a regeneration of the model would overwrite the direct edits. In most cases that I tried, topology changes were not allowed.

If your processes and needs are better supported with direct editing, use CoCreate. If your processes require more control and management use ProEngineer

From a product design point-of-view, I am not sure that mixing the two is really going to work out in the long run. However, for those using a history-based system and have an occasional need for making minor edits to imported geometry, it certainly would be a nice-to-have feature.
From what they are telling us, right now it is an either-or situation. You can either continue to create models the "old" way, with a history tree, or you can create them (or convert them) the "new" way, with no history tree.

The big question on everyone's minds is how they preserved the parametric functionality while doing explicit modeling. The problem with explicit systems like CoCreate and SpaceClaim is that it is difficult if not impossible to parameterize some types of geometry. For example, in OSD you cannot control the base diameter of a tapered cone. In a history-based system, the system remembers the parameter associated with the cylinder, and then it remembers the parameter associated with the draft applied to it.

According to the Solid Edge folks, this problem is not present with Synchronous Technology, but it remains to be seen.

Another advantage to a history-based system, or rather, a system whereby features are derived from embedded, persistent sketches, is that the relevant parameters for a feature only become visible when you go to work with that specific feature. It remains to be seen how this will work in Synchronous Technology. In OSD, with what parameters you are able to apply to the model it is basically an all-or-nothing situation to displaying them. There is a dialog where you can turn on and off the display of individual parameters but without knowing what features they go to this is tedious to use. They say with ST that the system, while now history-free, is still feature-based, so the relevant parameters will still be accessible relative to the features they control. But all this remains to be seen.

It's worth noting that Solid Edge has had boolean functionality for some time. You can, for example, move faces, offset faces, delete faces, modify holes and rounds, all without regard to whether these items were controlled by parametric features or not.

Synchronous Technology claims to go beyond this by actually sensing intended geometric relationships on the model for you.

For example, if you move a face, and the software senses that the face you are attempting to move is coplanar with other faces, it will move them all when you move the one, as it infers a geometric coplanar constraint. They say you can override this behavior if desired.

Similarly, they say that if you move a face that has other faces that join tangent to it, it will preserve those tangencies. Likewise if you move a face that is sensed to be concentric with other faces, it will move them as a group.

Basically, they say that rather than constraining such "obvious" geometric relationships in a sketch, their solver senses them on the model and preserves them anyway, provided you do not override the system and tell it to ignore them.

All this sounds well and good - my biggest concern, as I said before, is how easily can I dimensionally constrain my model to embed and preserve design intent.

Another huge achille's heal that ST will have, they say, that OSD and Spaceclaim also have, is the inability to create assembly level features. This is a huge problem to anyone who creates machined parts based off of cast parts, or machined welded assemblies.

It is supposed to ship this summer. It will be interesting to see how it all works out.

Steve
Reply With Quote