Tolerance Stack-up Analysis or GD&T Design Notes

[MikeH]

Has anyone attempted to use Mathcad as a tool to analyize or help document tolerance stackup and accumulation. I would be interested in any template someone may have started. Both worst case and root sum of the squares method. I'm also interested in any GD&T related design notes.

[ptc-1368288]

Attach the work sheet.

What you are asking is still obscure !

jmG

[MikeH]

On 12/22/2004 1:12:36 PM, jmG wrote:
>Attach the work sheet.
>
>What you are asking is still
>obscure !
>
>jmG
--------------------

What worksheet are you referring to? I do not have any Mathcad file/template created at this point in time. I have never used Mathcad before or the user forum either. Engineers at my workplace are being asked to better document our design and analysis so I'm investigating a means of accomplishing this. I'm trying to set up a basic template or document which we can use as a starting point and tailor as required to complete tolerance stackups for various projects. I haven't figured out yet what capabilities MathCad has, or if using a spreadsheet would be easier.

[StuartBruff]

On 12/22/2004 2:26:14 PM, MikeH wrote:
==What worksheet are you referring to? I do not have any Mathcad file/template created at this point in time. I have never used Mathcad before or the user forum either. Engineers at my workplace are being asked to better document our design and analysis so I'm investigating a means of accomplishing this. I'm trying to set up a basic template or document which we can use as a starting point and tailor as required to complete tolerance stackups for various projects. I haven't figured out yet what capabilities MathCad has, or if using a spreadsheet would be easier.

Do you have a document giving the tolerance calculations that you could post to the Collab? This would give members some idea of the type and size of the calculations.

My initial feeling, from the para above, is that this might well be the sort of thing Mathcad is particularly good at.

Stuart

[#5]

I have done many worst-case and "Root-Sum Squares" mechanical tolerance studies in the past. Off hand, I don't see any particular advantage to MathCad in this regard. For most cases, I think Excel works fine since usually only arithmetic and trig functions are involved. It's an interesting idea, however.
Garth

[IRstuff]

For a relatively simple RSS, Excel works adequately, but for a more complicated scenario, Mathcad is superior due to the ability to use symbolic names and the ability to retain units.

We just went through this exercise with Excel and Mathcad for some analysis at work and Mathcad provided a clearer final document


TTFN,
Eden

[akahighlander-d]

Eden,

How do you do this in mathCAD in excel there is a function for it?

[IRstuff]

On 1/7/2009 11:39:32 AM, akahighlander wrote:
>Eden,
>
>How do you do this in mathCAD
>in excel there is a function
>for it?

Not sure what function you might refer to. The basic stuff are summations, squares, and square roots, all part and parcel of Mathcad.

TTFN,
Eden

[kmckee-disabled]

I too am very interested in finding how Mathcad could perform this analysis. It's easy in Excal but very slow. If you don't have more than 25 variables try a program called MITCalc Tolerance Analyzer. It does all you are looking for. However, with many variables and several intermediate results, the program will run for 30 to 45 minutes for an answer.

The problem with using Mathcad is my inability to find how to use tolerances. i.e. X=.250 +/- .001. To me this is the crux of the difficulty. I have a Mathcad worksheet started but it's very simplistic.
X=.250 Xtol=.001
Xp=X+Xtol
Xm=X-Xtol

You get the idea. It works but it's not elegant. If there's someone who has figured this out I'd be happy to hear from them.

Cheers,
Ken

[PhilipOakley]

The usual problem with "tolerances" in a design
scenario is the need for choosing the many different
methods of combination between the different parts.

In some cases it will be (alleged) worst case, in
other 6 sigma (which version?), others a root sum
squared, etc. Along with knowing if the reported
tolerance is a reasonable map to actual tolerances..

For example it was common practice in optical
components for them to be machine 'thick', so that a
second pass could be attempted if the curvatures
were wrong. This would require careful
considerations (sometime to the point of miss
setting the nominal size & tolerance!)

Yes it can be a problem...

Philip Oakley

[ptc-1214470]

The attached is rather primitive but may suggest an approach to your issue - I believe I picked this worksheet up from the forum a few years ago.
JIm S.

[iDZignit]

I'm looking into this right now.  This sheet doesn't seem to come thru...& open.

We have an elaborate Excel sheet floating around & I wanted to see how well this could be driven from MathCAD.

[Werner_E]

---

@iDZignit wrote:

I'm looking into this right now.  This sheet doesn't seem to come thru...& open.

---

Guess you are using some version of Prime, not Mathcad.

Mathcad 15 and below should be able to read that attached worksheet, but neither version of Prime can.

You must use the converter to convert the legacy format sheet to the format of your Prime version. For the conversion to succeed you must have a full version of a current version of Mathcad 15 installed. If you bought Prime you are eligible to use Mathcad 15 using the same license file you generated when you installed Prime.

But then .. if you have Mathcad 15 installed, why would you use Prime? MC15 still is more powerful and capable compared to Prime and there is no guarantee that the converted sheet will still work as expected in Prime.

P.S.: To give you an idea what the file is all about, I attach a pdf print of that file

[LucMeekes]

You're probably using Prime (as opposed to 'real' Mathcad). Note that the file is from 2009, written in Mathcad 11 (or below). If you have a Prime license, you can also install Mathcad 15 (using the very same license file that is used for licensing your Prime).

And Mathcad 15 will open the file. You should also be able to convert the file to your Prime format. (That does NOT mean that it 'll work in Prime as expected...)

Success!

Luc

[VladimirN]

What version of Mathcad are you using?

[iDZignit]

P4, we have P5 waiting to install.

[VladimirN]

Please find in attachment "Statistical Tolerancing" worksheet for Mathcad Prime 3.1.

[artika]

Hello Vladimir,

is it possible to have this file for Prime 3.0?

Many thanks!

[LucMeekes]

There is no way in Prime to save to an older version.

Attached is a PDF of the file. It's simple enought that you should be able to recreate it in Prime 3.0.

If you want, you can post your Prime 3 file in a reply to this thread for other users of Prime 3 to use.

Success!
Luc

[VladimirN]

U vergat de mogelijkheid om regio's tussen twee documenten te kopiëren 😉.

[VladimirN]

Here it is.

[nas0k]

Please see the P4 file attached. I hope this information is helpful.

Norm