model symmetry

I'm building a part with several pro/surface and Isdx functions.
It's somethink like a bike, so it's have a symmetry plane.
So, I'm modeling the half of the model and making symmetry function near to
the end of the function tree (some details aren't symmetrics).
Near the end the whole model is an open quilt so I've tryied 2 methods :
1/
- I make a planar surface in the middle plane
- I merge this one with my surface (so I got a closed volume)
- I solidify this volume
- I make a symmetry function on this closed and solid volume
2/
- I make a symmetric quilt (with wymmetry function)
- I merge the two quilts
- I solidify this volume
So, in both cases, one of the functions spend a lot of time (seems to
depending on tangency condition in the symmetry plane).
(In case 1 : symmetry - in case 2 : merge).
a lot of time : I mean that this particular function spent more time for
regen than all the other ones.
So I'm wondering what would be the best approach in this sort of problem.
I can't change precision because other functions.
So what is the best method ?
How handle this sort of symmetry/precision problem ?
Thanks,
GB
Reply to
g. bon
Loading thread data ...
I'm building a part with several pro/surface and Isdx functions. It's somethink like a bike, so it's have a symmetry plane. So, I'm modeling the half of the model and making symmetry function near to the end of the function tree (some details aren't symmetrics).
Near the end the whole model is an open quilt so I've tryied 2 methods :
1/ - I make a planar surface in the middle plane - I merge this one with my surface (so I got a closed volume) - I solidify this volume - I make a symmetry function on this closed and solid volume
2/ - I make a symmetric quilt (with wymmetry function) - I merge the two quilts - I solidify this volume
So, in both cases, one of the functions spend a lot of time (seems to depending on tangency condition in the symmetry plane). (In case 1 : symmetry - in case 2 : merge). a lot of time : I mean that this particular function spent more time for regen than all the other ones.
So I'm wondering what would be the best approach in this sort of problem. I can't change precision because other functions. So what is the best method ? How handle this sort of symmetry/precision problem ?
Thanks, GB
Any time I've spent a lot of time with merge or merge/cutout type operations, it's be the accuracy setting ('Edit>Setup>Accuracy'). Increase it by 10 or 100 fold (decimal left one or two places) and the problem was solved, no slowing, no hesitation, no agonizing over the results. The next time? seconds not minutes. Sometines it took enabling absolute accuracy (my favorite) and setting this to some accuracy of accuracies level (ratio of smallest part diagonal to assembly diagonal). This generally knocks some sense into Pro/e.
David Janes
David Janes
Reply to
David Janes
Recheck your basics.
Are your your midplane curves definitely planar curves? - sketches on the midplane, planar ISDX or curve thru points made planar.
When you mirror the model [at the point it becomes assymetric] are you choosing the model name at the top of the model tree and mirroring the whole?
Sean
Reply to
Sean Kerslake
How about case 1 but mirror the part (or just the closed quilt) before you make it solid. Then solidify your first closed quilt, then solidify your second. Perhaps the second solidify won't need to calculate tangency to the solid geometry for it to solidify. Maybe.
Reply to
graminator
How about case 1 but mirror the part (or just the closed quilt) before you make it solid. Then solidify your first closed quilt, then solidify your second. Perhaps the second solidify won't need to calculate tangency to the solid geometry for it to solidify. Maybe.
Has anyone ever tried the datum ribbon feature for ensuring tangency and good symmetrical merging? Another concern, echoing graminator's, is that surfaces be set normal to the mirroring plane (that's also where the datum ribbon helps). I guess I really haven't "gotten" what the problem is, beside slow regen speed, which was already answered. Are you looking for something that tricks Pro/e into not being so slow or that, when mirrored, doesn't EXAGERATE the accuracy problem? And it must as the part envelope just doubled with mirroring.
David Janes
David Janes
Reply to
David Janes
As I understand it the datum ribbon is for draft angles, setting normalcy should be robust for symmetry.
My first question on midplane issues is always are the curves 'actually' on the midplane rather than 'theoretically' on the midplane. Classic example being someone has a curve thru points which all lie on the midplane and assume this generates a planar curve when it is actually/mathematically still a free curve until you change its Attributes to Quilt/Surf.
Sean
the
near
:
for
problem.
How about case 1 but mirror the part (or just the closed quilt) before you make it solid. Then solidify your first closed quilt, then solidify your second. Perhaps the second solidify won't need to calculate tangency to the solid geometry for it to solidify. Maybe.
Another concern, echoing graminator's, is that surfaces be set normal to the mirroring plane (that's also where the datum ribbon helps). I guess I really haven't "gotten" what the problem is, beside slow regen speed, which was already answered. Are you looking for something that tricks Pro/e into not being so slow or that, when mirrored, doesn't EXAGERATE the accuracy problem? And it must as the part envelope just doubled with mirroring.
David Janes
David Janes
Reply to
Sean Kerslake
I wouldn't have though you would want the curve to be 'actually' planar. Thru points is best, as I learnt from Dave. The 1st point of the curve has to be normal to the plane, assuming you're starting in the middle, i.e. on the mirror plane. Same for all the curves. Then the boundary surface would have normalcy constraint to the mirror plane on the adjacent boundary.
It occurs to me I might be reading your post wrong. Perhaps by midplane curve you mean one that lies on the mirror datum. I was thinking of the ones going in the adjacent dir. Pity we can't put sketches up on this board.
messagenews: snipped-for-privacy@p47g2000hsd.googlegroups.com...
Reply to
graminator

Site Timeline

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.