cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
Showing results for 
Search instead for 
Did you mean: 

Edit definition view refit problem ?

Newbie

Edit definition view refit problem ?

Hi. (excuse my english 😄 )


I recently find weird thing... you zoom on a part to redefine it, when you click edit defenition it stupidly made a "pan" move and now the part you want to redifine is off-screen.



Someone Know how to fix that?



Thanks

7 REPLIES 7

Edit definition view refit problem ?

I think this sort of behaviour is just standard Pro/E.

If you have maintenance and a PTC.com account, you can vote up my 'idea' to get rid of it here:

Edit definition view refit problem ?

This behavior can be caused by have a feature out in space. Most likely from a datum coordinate. The model could have been defined in assembly mode and used and copied a feature from a different model that was not located at the screen center. Then the other features were built upon using the original feature as a reference.


 
David Brainsky
Certified PTC Instructor
Chula Vista, Ca. 91910
email: -
Cell PH:619-253-5137


----------

RE: Edit definition view refit problem ?

Attached is a picture of another example related to this refit problem...


It would be nice if Creo had an high level toggle or configoption-



  1. refit using geometry/datums/features existing on Screen.

  2. image on screen fixed relative to default coordinate system.

RE: Edit definition view refit problem ?

This happens because components later in the tree are removed from the model, so the overall size of the model extents change due to the decrease in geometry. Panning is based on an offset from the center of the model extents.


The way to prevent this from happening is to prevent the model extents from changing, which can be done by creating a sketched datum curve that is larger than the solid model components and earlier in the tree than the component that is to be redefined.


The behavior is an irritant. Given that every model has an implicit coordinate system, it would have been nice for pan/zoom/orient to be based only on the implicit coordinate system. An explicit refit-to-extents or center-on-extents command can still be used to return a model to the visible portion of the screen coordinates.

RE: Edit definition view refit problem ?

David,


I have to challenge your suggestion of adding a large datum curve out in space to keep the model from moving around. This is only safely advisable if the modelhasabsolute accuracy enabledand not relative accuracy. Relative will take into account the effect of that curve regarding the bounding box size and later on, if one decides to delete that curve, features that were created/computed and resolvedwith the curve there, have a possibility to failing due to accuracy issues.


My $.02

In Reply to David Schenken:



This happens because components later in the tree are removed from the model, so the overall size of the model extents change due to the decrease in geometry. Panning is based on an offset from the center of the model extents.


The way to prevent this from happening is to prevent the model extents from changing, which can be done by creating a sketched datum curve that is larger than the solid model components and earlier in the tree than the component that is to be redefined.


The behavior is an irritant. Given that every model has an implicit coordinate system, it would have been nice for pan/zoom/orient to be based only on the implicit coordinate system. An explicit refit-to-extents or center-on-extents command can still be used to return a model to the visible portion of the screen coordinates.


RE: Edit definition view refit problem ?

Interesting insight, but there's no reason to ever remove the curve. If I use one, I'd pack it into a group at the top of the model tree along with other default geometry.


It should not make a negative difference to accuracy either. If one were to re-order the model tree to put larger items earlier, that would do the same thing to accuracy, so it's not that the curve is unique in that regard. Putting a curve that is suitably sized would stabilize relative accuracy early, preventing relative accuracy from affecting the models due to any number of causes.


It would also only affect assembly features; the individual parts have their own accuracies and should be unaffected.


I've used such features before when making frame-grabbed animations that start with a small item and add others to keep the model from wandering around the screen with each added item.


x

RE: Edit definition view refit problem ?

David.


All true given the luxury of staying native within Creo. However, the models I work on like interior and exterior vehicle body panels, where the ratio of large to small relative values are significant,when exported to CAM software thru Step or IGES the surface quality and stitching(I.E accuracy)canbe compromised. The surface accuracy when"translated" out, if that artificial bounding box entity is there way outside the solid, tends to effect the data to such a degree that the gaps and slivers cannot be overcome on the far side. I need to make sure that I keep any over-build geometry as tight as possible to the final solid if I get into a situation where relative accuracyis the only option. I like absolute for big body panels and class A surface work, although regen times can get ugly.


Just thought I would throw that out there so others know there can be a hidden, "after model is done, oh crap I can't delete that curve because 800 features fail"effect down stream.


Another $.02

In Reply to David Schenken:



Interesting insight, but there's no reason to ever remove the curve. If I use one, I'd pack it into a group at the top of the model tree along with other default geometry.


It should not make a negative difference to accuracy either. If one were to re-order the model tree to put larger items earlier, that would do the same thing to accuracy, so it's not that the curve is unique in that regard. Putting a curve that is suitably sized would stabilize relative accuracy early, preventing relative accuracy from affecting the models due to any number of causes.


It would also only affect assembly features; the individual parts have their own accuracies and should be unaffected.


I've used such features before when making frame-grabbed animations that start with a small item and add others to keep the model from wandering around the screen with each added item.


x