Welcome to MCAD Central

Join our MCAD Central community forums, the largest resource for MCAD (Mechanical Computer-Aided Design) professionals, including files, forums, jobs, articles, calendar, and more.

Register Log in

small rounds on big parts

survey

New member
Hello,



I am tring to add a small round (R= .02) onto a comparitively large part (12 x 6 x 9), and pop-up window tells me that I can not enter a value under .04.



If I copy the surface geometery to another part (just for a test) and create a solid out of it, I can add small rounds >.01 just fine.



I have tried to adjust the accuracy (both ways) and it does not help.



Does anyone have any suggestions?



Thanks



Survey
 

gggggggggg

New member
12 x 6 x9 does not sound too big. I'm surprised you are getting that error message. Could you post the part so we can have a look?



These types of problems are hard to diagnose without actually being able to work on the part. Pro/E is funny about stuff like this.
 

survey

New member
Unfortunately I can not post the part ( do to confidentiality).





I do have some Geometry Checks that are creating tiny edges. I'm trying to fix them and I have had some success putting smaller rounds on.



Is it normal for features with geometry checks to cause these problems?



survey
 

donha

New member
Not sure if Setup>Accuracy will allow you to enter a smaller radius or not? First thing I would look at.
 

survey

New member
I changed to absolute accuracy and it looks like it's going to work. Thank you so much for the help!!!!!



Survey
 

dougr

New member
Absolute accuracy is actually based on a physical quantity in the model (smallest edge) and is the critical one.



When relative accuracy is held constant (default) then absolute accuracy is free to change as the model bounding box changes in size. This explains why models can start to fail for no seemingly apparent reason.



As edges are created that are smaller than absolute accuracy then small edge geom checks are generated which can be fixed by dialing down the accuracy.



I set all my parts to use absolute accuracy to make sure that Pro/E can never adjust this value and inadvertently trigger feature failures.



Think relative accuracy was created way back when just for computational convenience and was the only accuracy available until Rev 17.
 

Brian_Adkins

Moderator
I think relative accuracy was created for regeneration speed. It basically allows Pro/E to be only as accurate as it needs to be. A simple cube part could be mathematically evaluated using much less stringent accuracy than a complex part with lots of intricate features.



PTC used to tout this as a feature that enabled Pro/E to outpace other systems. (maybe they still do)



For complex parts, though, I like to dial in an exact accuracy (absolute). This is especially true when the ratio of smallest to largest edges on a part is apt to vary significantly throughout the design cycle (as dougr stated).



Absolute accuracy is also a great way to ensure that the master-model-merge technique does not cause accuracy problems. A large master model that gets dissected into many smaller models can lead to the individual models having different accuracies (because they could all be different sizes).



-Brian
 

dougr

New member
Absolute accuracy is also a great way to ensure that the master-model-merge technique does not cause accuracy problems.



For merge and cutout operations it's mandatory for part absolute accuracies to match..
 

Brian_Adkins

Moderator
>>For merge and cutout operations it's mandatory for part absolute accuracies to match



Not what I was referring to, but I'll take your word for it.



-Brian
 

Brian_Adkins

Moderator
Sort of...



I was really trying to get at what happens after the slave parts have been set up. Generally, all the slave parts are empty initially, and the master part is not... so they do not have the same accuracy (if relative is used). The problems can arise weeks later as the slave parts are trimmed down to represent smaller parts and the accuracy changes to reflect this smaller size. On complex styled parts, this can lead to situations where the slave parts have wildly different accuracies from one another.



In other words, if using relative accuracy, the slaves all start off with the same accuracy, but then quickly diverge as they are trimmed into smaller parts. Then, additional features (i.e. a small round) added to one part might regenerate fine, but a round of the same size might not regenerate on another of the slave parts.



Setting them all to absolute in the beginning would put all the slave parts on a level playing field so that the same feature size rules would apply regardless of the trimming you do to them.





-Brian



The regular MMM technique using the merge feature was around long before absolute accuracy was available. However, I now use the external merge, external copy-geom, or inheritance feature types instead.
 

Sponsor

Top