Tolerance Analysis books 4

[ntuedutw]

Hi All,

After we introlduced GD&T(Y14.5) into our design requirements, our traditional worst-case tolerance stack analysis can't work(Plus-Minus system is our original system). Could everyone suggest some tolerance analysis books for self-study?

Note: We are investigating to buy VSA for our design jobs. This software seems powerful but is hard for us to understand if we are lack of tolerance analysis ideas.

Thanks in advance,
Ntuedutw

 

[wes616]

"Advanced Tolerancing Techniques (Engineering Design and Automation)" By. Hong-Chao Zhang


Wes C.
------------------------------
No trees were killed in the sending of this message, but a large number of electrons were terribly inconvenienced.

 

[looslib]

CeTol is another program for tolerance analysis.

http://www.sigmetrix.com/

Al Neumann has a course and book or two on tolerance and stackup analysis.

http://www.geotol.com/

"Wildfires are dangerous, hard to control, and economically catastrophic."

Ben Loosli
Sr IS Technologist
L-3 Communications

 

[TomGDandT]

Both CeTol and Vis-VSA are great packages.
We have both packages where I work.
We have engineers that are experts for each.

I have taught the Basic CETOL class for Sigmetrix.

What products are you designing and with what CAD system.
CeTOL is connected to Pro/E or Catia.
Vis-VSA requires that CAD data be sent thru one of their translators to bring the data into the Vis-Mock.

Do you have any questions about either package?

Tom Rhodes, GDTP-S
QMC LLC;

http://www.qmc.com

Senior Dimensional Management Engineer.
CeTOL 6 Sigma

 

[McLeod]

The best book I've encountered for tolerance analysis is the Dimensioning and Tolerancing Handbook (Drake, et al, 1999).

http://www.amazon.com/Dimensioning-Tolerancing-Handbook-Paul-Drake/dp/0070181314

There are also a lot of good papers online at BYU's Association for the Development of Computer-Aided Tolerancing Systems (ADCATS)

http://adcats.et.byu.edu/home.php.

The software we have is MechTOL, which isn't graphical, but it handles geometric controls.

http://www.mechsigma.com/

(click on "Software").

 

[TomGDandT]

The Paul Drake book is great.
Has many contributions from authors and experts in the field. I enjoy my copy.

The adcats site has alot of good stuff.
Too bad the Adcats group is not current.
BYU has provides Sigmetrix with many excellent employees that make CeTOL the software I use.

The Mech Sigma package was shown to me by Paul Drake at one of the ASME classes he teaches. It looked like a good 1D or 2D package.


Tom Rhodes, GDTP-S
QMC LLC;

http://www.qmc.com

Senior Dimensional Management Engineer.
CeTOL 6 Sigma

 

[looslib]

The ADCATS program, I think, has accomplished what Dr. Chase wanted. Sigmetrix has taken over the development and further enhancements of the program.

We used CeTol where I used to work. The design engineers had a hard time with understanding the need for the program. The instructor from Sigmetrix, a fellow BYU grad, did a great job of explaining it, as he was a developer of the interface to Pro/E.


"Wildfires are dangerous, hard to control, and economically catastrophic."

Ben Loosli
Sr IS Technologist
L-3 Communications

 

[TomGDandT]

Has Gordon C. or Ted A.?


Tom Rhodes, GDTP-S
QMC LLC;

http://www.qmc.com

Senior Dimensional Management Engineer.
CeTOL 6 Sigma

 

[ragtop1955]

Ntuedutw,

Transitioning from a traditional ± tolerance system, to GD&T Y14.5 is not always easy. Many of the correct questions to ask are difficult, especially since you do not always know what to ask. This is especially true when choosing a tolerance analysis tool. Should you choose [!]VSA[/!] (full blown 3D tool), [!]CETOL[/!] (Vector based 2D-ish spreadsheet), or even an [!]Excel spread sheet[/!]. All can do a good job, but all have advantages and disadvantages.

In essence you have to choose between ease of use (training time, time to perform analysis) vs. accuracy (extra effects of a 3rd dimension, depth of software "built in smarts").

Other factors include what CAD system do you use? [!]CETOL[/!]works only with Catia v5 and Pro/e. [!]VSA[/!] works with all - UG, AutoCAD, Catia v4 and V5, Solidworks, etc. Some tools like [!]TSV[/!] is only in UG NX . And of course, an Excel spreadsheet works with all above, including manual drawings.

Another factor is do you use embedded 3D GD&T in your designs / CAD files, or nonparametric GD&T in a flat 2D drawing. Some tools take advantage of the embedded "intelligent" GD&T and uses it directly as input to the tolerance analysis.

For more information, here is a great article in MCAD Cafe on the above subject.

Today's Culture in Tolerance Analysis Blocks Real Benefits

http://www10.mcadcafe.com/nbc/articles/view_article.php?section=UserArticles&articleid=244778

We have a couple of tools to help make this decision. If you are interested, I could send these spreadsheets and Word documents to you. It helps you evaluate your current corporate environment, and helps direct you to a "class" of tools (not a specific tool). This may help you make your decision. Let me know if you are interested.



Curtis DeSaele
Geometric Solutions

http://www.GeometricSolutions.net

VP Engineering
Dimensional Management Specialists

 

[TomGDandT]

VSA operates inside Vis-Mockup.
Vis-Mockup handles the CAD data for viewing and picking.
The CAD data must be translated into a JT file for Mockup.

Tom Rhodes, GDTP-S
QMC LLC;

http://www.qmc.com

Senior Dimensional Management Engineer.
CeTOL 6 Sigma

 

[ragtop1955]

Technically, Vis VSA needs two software modules. TeamCenter's Vis Mockup and TeamCenter's Vis VSA. TeamCenter's Vis Mockup handles the base 3D CAD functions, then Vis VSA does the specific Tolerance Analysis "on top of" the Vis Mockup 3D CAD "engine".

All modern 3D tolerance analysis tools are "tied" to some form of 3D CAD engine. But to not be "loyal" to a specific CAD system (i.e., Catia V5, OR UG NX, OR Pro/e), Vis VSA uses TeamCenter's Vis Mockup. Vis Mockup is the [highlight]"CAD Neutral"[/highlight] solution used in TeamCenter PDM systems to view CAD files from multiple CAD systems.

Using Vis Mockup is advantageous for a few reasons:

1 - You do not need a dedicated CAD seat to perform tolerance analysis (Vis VSA runs on Windows based PC)
2 - You do not have to change the master CAD files, when performing [highlight]"what if"[/highlight] studies, to improve results.
3 - For companies that run multiple CAD systems, you only need [highlight]one tool[/highlight] to buy / train staff / schedule work flow through.
4 - Other seemingly "stand alone" 3D tolerance analysis tools have a basic 3D CAD system "built" in. Since many of these products are primary tolerances analysis tools, the 3D CAD portion is limited. TeamCenter's Vis Mockup is improved annually, and get more functions added every year.
5 - if you already have TeamCenter, you have access to all the CAD files from all CAD systems.
6 - If you do not have TeamCenter, then Vis Mockup definitely less expensive to buy than a seat of a major CAD system (see below).

To be "CAD Neutral", this does require some form of translation from the native CAD system into a format that Vis VSA and TeamCenter's Vis Mockup can understand. For the UGS products (i.e., UG NX, IDEAS, SDRC), no special translator is required (see below). For other CAD systems, there are IGES, STEP and ones developed for each specific platform.

Now the JT file format is a "tessellated" file, with NURB data. A "tessellated" file basically is the graphical representation of the CAD file you see on your CAD screen (this is why a hole sometimes looks like an Octagon, instead of a circle). The real "CAD" circle is within the Math on the hard drive, but the octagon is displayed graphically on your screen. Thus when you look at it, it looks incorrect -- but when you query the CAD file, it looks at he hard drive and states this is a circle of XXXXXXXX precision. All modern CAD systems use tessellation, to show graphics quickly on screen.

Most modern CAD files consists of many things - Parametric parameters, NURB data, PMI information, "tessellated" data, and all other types of important stuff. The "tessellated" data is the bulk of the JT file, with the NURB data added for accuracy. For UGS products, the same "tessellation" is used throughout. Thus UG NX a JT files have the same tessellated data. Thus a UG NX file opens directly in Teamcenter products. For all other CAD systems, they use different form of tessellated data, and need to be translated.

Typically, the JT file is 10-20% the size of the original CAD file.

Thus the bottom line -- if you are planning to do 3D tolerance analysis, do you want to "tie" it to a single CAD system, or one that is can use multiple CAD systems. Both require a CAD "engine" to run (i.e., Vis Mock, Catia V5, etc.)

Also beware, some "3D" tolerance analysis tools use the 3D CAD system to generate "input" into an analysis tool that generates results outside the 3D environment in essence, creates a spreadsheet like analysis). This does not allow for true 3D geometry "effects", and may / will give misleading results. These tools seem "3D", but are more 1D / 2D -ish


[!]Curtis DeSaele[/!]
[blue]Geometric Solutions[/blue]

http://www.GeometricSolutions.net

VP Engineering
Dimensional Management Specialists

 

[TomGDandT]

CETOL is not a "Vector based 2D-ish spreadsheet".
There is no speadsheet with calculations presented.
CETOL provides the variation results in the Analyzer.
Results can be viewed quickly in a column format or in a HTML format. There are standard HTML formats or you can create your own. I have asking Sigmetrix for years to provide the results in a spread sheet format.

Curtis, if you want any more info about CETOL you can reach me at QMC. I remember you and Bryn N. from your VSA days.

Tom Rhodes, GDTP-S
QMC LLC;

http://www.qmc.com

Senior Dimensional Management Engineer.
CeTOL 6 Sigma

 

[sballcoach]

You will most likely find a significant amount of books on the market for GD&T, lesser for 1D variation analysis / stack ups, and far less, if any, covering the topic of 3D variation analysis. They might provide you with some of the basics surrounding 3D variation analysis, but nothing to the extent seen in a 3D training class specific to a particular product.

In short a 3D variation analysis model consists of the following:
1.) CAD geometry (or points that represent the geometry when not available)
2.) Assembly process
3.) Tolerance part/assembly
4.) Measurements
5.) Analysis results (Monte Carlo, Vector Loop & Sensitivity)

CAD Geometry:
As previously stated there are different solutions on the market today that can work in either the integrated CAD or stand-alone (CAD independant) environment. The major players on the market today are 3DCS, VIS-VSA and CeTol.
The last two have been discussed already. The 3DCS offers both a CAD independent as well as a CATIA V5 integtated solution.


Assembly Process:
This is the definition of exactly how the parts form sub-assemblies and final assemblies. Similar to what many CAD solutions call 'constraints'. However the CAD software' traditionally do not provide the detail necessary to exactly define the assembly process. For example how a wheel is attaching to five lugs or hanging of an aircraft door to the fuselage. 3DCS provides an enormous amount of flexibility in defining the assembly process.


Tolerance part/assembly:
This is straight forward in that the variation model requires that the detail components have tolerancing assigned to features on the part. Different solutions go about this in different ways. The 3DCS software offers two methods for the tolerancing of components, either manually when limited or no GD&T exits and or through the use of the embedded GD&T information (as in CATIA V5, FTA). Additionally 3DCS offers many distribution types (Normal, Min-Max, Trapizoidal, Wiebull...) where other products offer only a "Normal" distribution type.

Measurements:
Measurements are simply the areas to be analyzed in the variation analyis model. The measurement would consist of the nominal information, ex. distance between to features, and the tolerance which is deemed acceptable. For example the gap between a door and fender is 4.5mm
+/- 0.75mm. This information provides the software the necessary information for comparing the 3D simulation results to determine the amount of mean shift in the process, the percentage outside the specification limits, and numerous other statistical results.

Analysis Results:
Analysis results consist of Monte Carlo or Vector Loop which predict the amount of variation within the assembly and Sensitivity determines the contributors or the "source" of variation. 3DCS offers all three types of analysis.

With respect to integrated versus non-integrated tools, there are benefits to both that can be discussed outside of this forum. Again, 3DCS is available both as a CATIA V5 integrated solution and as a STAND-ALONE CAD independent Windows based solution. Both versions work very effectively with preliminary CAD geometry and GD&T information in order to perform variation analysis early in the design phase. As the design matures, so does the 3DCS model with automatic updates when new geometry is available.

For a more details explanation visit:

http://www.machinedesign.com/asp/viewSelectedArticle.asp?strArticleId=60916&strSite=MDSite&catId=2

Rick Rots
Dimensional Control Systems

http://www.3dcs.com

 

[TomGDandT]

Hello Rick.
Now we just need someone from Sigmetrix and UG Solutions to tell us about CETOL and VIS-VSA.
What about the guys in Holland?

Tom Rhodes, GDTP-S
QMC LLC;

http://www.qmc.com

Senior Dimensional Management Engineer.
CeTOL 6 Sigma