tolerance analysis

[momi]

hi,
can anybody tell me that whether PRO E be helpful for
tolerance analysis of any designed parts.
also any tutorial to understand upto which extent this can
be helpful...
best regards.
momi

 

[baseer]

hi, momi
yes, it is possible!
please see attachment.

2010-11-15_074253_4328_TolAnalysis_DS_EN.rar
regards

 

[momi]

hi,
nice to c this,
however its 1D analysis, how i can do 3D analysis, if its
not in PRO E, anyother software.
can u give me any tutorial to explain it please
regards and thanks

 

[baseer]

hi, momi.
unfortunately me too does not have that, it is appreciative if you find and can share with me.

regards,
ptctips@gmail.com

 

[Kaz Z06]

You can do a 3D tolerance analysis using "Behavioral Modeling Extension" (BMX). In fact, WF5 has some nice 6 sigma tools to carry out DPM studies similar to CETOL. You must have enough experience/knowledge on how to use BMX before jumping in - too much to get into it in this forum - please check manuals & tutorials.

Kaz

 

[Blackline]

Have alook here:

http://www.ptc.com/WCMS/files/75687/en/flash/index.htm?im_db key=75687&icg_dbkey=362

Best regards

 

[momi]

hi everybody,
1-thanks for everything.
2-i have seen the link given by blackline, its good but
still it gives 1D solution and not the 3D solution,
3-KAZ im getting to BMX now, can you suggest me any
specific material,just to save my time.. thanx.
take care.

 

[Kaz Z06]

There isn't any material I am aware of. You just need to understand how BMX works and create measurements that represent the dimension you need to control. Then vary all the dimensions that create the measured item of interest and run it through MODS (really a DOE). Apply your process capability and run a Monte Carlo. WF5 can run Monte Carlos.

Cheers,

Kaz

 

[Ray Ellender]

What you need is CETol from Sigmetrix (www.Sigmetrix.com). Tolerance Analysis Extension (TAX) in Pro/E is the cut-down 1D version, but CETol has full 3D statistical tolerance analysis capability. I think you would spend a lifetime trying to do the same thing in BMX unless it was a really small model, in which case you could do it by hand in Excel anyway!

 

[Kaz Z06]

CETol is a very good tool. I have used it and have had formal training for it. BMX can handle this problem and it would not take a lifetime to accomplish. Both approaches do require you follow formal methods to obtain good results and require a good amount of extra effort. Pro/E's implementation of CETol in one dimension (TAX) is more efficient than the way CETol's tool works, but again it is only 1-D so things can be simplified. CETol was very seldom applied at the companies I'm familiar with due to its level of complexity - only more highly experienced engineers successfully used this tool as it requires consistent use. Bottom line is, you need to be properly prepared to attack this type of problem. Adequate knowledge of how the parts might assemble together is most important to apply either method. When this type of study is done well, it is very valuable.