Smart questions
Smart answers
Smart people
Join Eng-Tips Forums
INTELLIGENT WORK FORUMS
FOR ENGINEERING PROFESSIONALS

Member Login




Remember Me
Forgot Password?
Join Us!

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips now!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!

Join Eng-Tips
*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.
Jobs from Indeed

Link To This Forum!

Partner Button
Add Stickiness To Your Site By Linking To This Professionally Managed Technical Forum.
Just copy and paste the
code below into your site.

tdculbert (Mechanical) (OP)
8 Jan 08 16:30
Our company has historically dimensioned exclusively in english; our standard title blocks contain the typical Tolerances unless otherwise stated:
.XX+-.01
.XXX+-.005
.XXXX+-.0005

Lately we've had to produce more metric drawings or metric variants of existing drawings. Without using dual dimensioning, this raises questions regarding the title block standard tolerances.  If you were to do a direct conversion, your title block would contain fairly silly tolerances:
.XX+-0.25
.XXX+-0.127
.XXXX+-0.0127

In addition, in metric tolerancing trailing zeros are omitted. Thus, 0.240mm is written 0.24mm and then the decimal-place-specific standard dimensioning rubric becomes irrelevant.

What's the proper way to specify title block tolerances in metric? A single symmetric tolerance with every different tolerance specifically called out?
ctopher (Mechanical)
8 Jan 08 16:34
There is nothing with what you show. Add a note that they are metric.

Chris
SolidWorks/PDMWorks 08 1.1
AutoCAD 06
ctopher's home (updated 10-07-07)
ctopher's blog

MadMango (Mechanical)
8 Jan 08 16:47
Don't you normally drop one place when converting?

+/- .3mm
+/- .18mm
+/- .013mm

"Art without engineering is dreaming; Engineering without art is calculating."

Have you read FAQ731-376: Eng-Tips.com Forum Policies to make the best use of Eng-Tips Forums?

tdculbert (Mechanical) (OP)
8 Jan 08 17:01
Yes, you would drop a decimal place.  The overall question of the use of such a decimal-place-dependent tolerance scheme in a system that truncates decimal place zeros remains, however.
ctopher (Mechanical)
8 Jan 08 17:05
Yes.
Inches
.XX +/- .01
.XXX +/- .005
.XXXX +/- .0005

...Would be
Metric (mm)
.XX +/- .3
.XXX +/- .13
.XXXX +/- .013

Chris
SolidWorks/PDMWorks 08 1.1
AutoCAD 06
ctopher's home (updated 10-07-07)
ctopher's blog

drawoh (Mechanical)
8 Jan 08 17:32
tdculbert,

  I set up our metric title block as follows...

X = +/-0.5
X.X = +/-0.2
X.XX = +/-0.1

   As you correctly noted, the ASME Y14.5M-1994 standard requires that you drop the trailing zeros from your metric dimensions.  This messes up any scheme of controlling tolerances from the title block.

   Oh well.

   Now, I set each tolerance on each metric dimension.  This is not a bad thing, and it does not take that long.  The only people will be inconvenienced are the lazy slobs who apply default tolerances to everything.  

                          JHG
Helpful Member!  KENAT (Mechanical)
8 Jan 08 17:36
Our standard inch block where I work (US) is

1 PL = +-.03
1 PL = +-.01
1 PL = +-.005

When using mm we change this to

1 PL = +-0.75
2 PL = +-0.25
3 PL = +-0.1

Obviously these aren't directly equivalent so that has to be taken into account with tolerances etc.

However, you do have a point with the trailing zeros issue.  

Quote:

ASME Y14.5M-1994 1.6.1 (c)  Where the dimension exceeds a whole number by a decimal fraction of one millimeter, the last digit to the right of the decimal point is not followed by a zero. See Fig. 1-2

So you can’t simply have a 10mm rod with a block tolerance of +-0.1 by putting 10.000 which limits the usefulness of the block tol.

In the UK where I worked we had a single block tolerance that didn't vary by decimal places.  Typically this was +-.25 although we changed it to suit the drawing.  Any dimension that varied from this had it's tolerance specifically called out (we usually showed limits not +-).

Alternatively, there are horrible standards like ISO 2768 which you can invoke.  

For instance the last paragraph of the standard says:

 

Quote:

A.4 The tolerance the function allows is often greater than the general tolerance. The function of the part is, therefore, not always impaired when the general tolerance is (occasionally) exceeded at any feature of the workpiece. Exceeding the general tolerance should lead to a rejection of the workpiece only if the function is impaired.

That said, it may be that the implementation of ISO 2768 I’ve seen makes it seem worse than it is.

ASME Y14.5M-1994 doesn’t appear to directly say what to do about block tols for metric or inch but does reference ANS Y14.1 at para 2.1.1 (e).  I believe this is now ASME and for metric you’d want ASME Y14.1M Metric Drawing Sheet Size and Format, I don’t have this standard.

Sorry I couldn’t be more help.

KENAT, probably the least qualified checker you'll ever meet...

fcsuper (Mechanical)
9 Jan 08 12:40
tdculbert, just my opinions:

Of course, dimensioning in metric isn't really supposed to be done using decimal places, hince the rule to not include insignificant zeros after the decimal place.

The ISO and ASME standards utilize a common tol table for metric dims.  All you need to do is call out the class of tol you want on the drawing and reference the standard.  The idea behind the common tol table is that tolerances are based more on feature/part size and type, rather than a somewhat arbitary place decimal system.  

Barring this, you can just copy the method of using places.  It's safe, familar and good enough for most circumstances.  

If you are converting drawing to metric, then a direct (fully accurate) translation from inch to metric in your place tol scheme is best (even if the numbers are a bit crazy).  Otherwise, you might be introducing unintentional errors or contradictions to the design intent simply to show the part dim'ed with a different unit of measure.  

If you are drawing the part new, then it doesn't really matter which system you use, or what tols, as long as you use the choosen system properly within the design intent of the part.

Matt Lorono
CAD Engineer/ECN Analyst
Silicon Valley, CA
Lorono's SolidWorks Resources
Co-moderator of Solidworks Yahoo! Group
and Mechnical.Engineering Yahoo! Group

KENAT (Mechanical)
9 Jan 08 13:21
fcsuper, what ISO & ASME standards for common tol table are you referring to?  Are you just talking about shaft/hole fits or something like ISO 2768?

These threads show some of the trouble I had witn a vendor using 2768 & joebk had issues too.  thread1103-196260: Tolerance analysis ISO2768 thread1103-197786: tolerance analysis to ISO 8015 & 2768 part 1 (again!)

Done properly the decimal place system isn't arbitrary, if those tolerances aren't suitable for a specific feature you asign it its own tolerance.  They only become arbitrary when people don't consider tolerancing and just throw numbers down.

KENAT, probably the least qualified checker you'll ever meet...

WHITMIREGT (Aerospace)
9 Jan 08 21:51
The following came from the Canada dimensioning and tolerancing standard which is based on the ISO dimensioning system:"For drawings indicating dimensions with no specified tolerances, it is recommended that a general tolerance note applicable to such dimensions be added to specify the universal tolerance or tolerances most applicable to the drawing. Such a note may specify a single tolerance. A typical note "TOLERANCES FOR DIMENSIONS NOT OTHERWISE LIMITED SRE AS FOLLOWS: ANGLES ±2°, AND LINEA DIMENSIONS ±0.25".

I was also searching a number of ISO drawings, and none of them did not have a title block tolerance other than one tolerance.
Joshmo (Aerospace)
10 Jan 08 13:50
All,

My only comment regarding metric default tolerances is that going to 3 decimal places in metric is almost useless since .001 mm is .00004" !!!

JMo
ctopher (Mechanical)
10 Jan 08 13:55
If it's outside the realm of your tol block, call it out on the dim.

Chris
SolidWorks/PDMWorks 08 1.1
AutoCAD 06
ctopher's home (updated 10-07-07)
ctopher's blog

fcsuper (Mechanical)
10 Jan 08 15:32
KENAT, technically, both.  I didn't say they didn't have issues...only that these are how metric parts are intended to be toleranced (hince the removal of the insignificant zeros after teh decimal).

As far as the decimal place scheme, it can get arbitary if it isn't set for the type of part, which is an often occurance.  It is useless on molded part over a certain size since tolerances on those parts is best measured in per inch terms.  Sheet metal follows very different rules too, where similar features can have completely different tolerance ranges simply because one is across a bend from the other.  It seems to me, the decimal place scheme was developed for machined parts, where it is most useful.  

Of course, having said this, I rely on the decimal place scheme almost excusively.  I just pay attention to where tols need to be added to individual dims to account for processes.

Matt Lorono
CAD Engineer/ECN Analyst
Silicon Valley, CA
Lorono's SolidWorks Resources
Co-moderator of Solidworks Yahoo! Group
and Mechnical.Engineering Yahoo! Group

LONDONDERRY (Mechanical)
16 Jan 08 14:12
Technically if you are going to use Metric tol. system as I have done I would suggest going with the ISO tol schem ISO 2768-mKE

0.5 to 3        +/-0.1
over 3 up to 6  +/-0.1

CheckerRon (Mechanical)
16 Jan 08 15:01
Londonderry (Ireland?):
 I don't have ISO 2768-mKE, but are you sure you got that right? <3 and >3 are the same, ±0.1?
LONDONDERRY (Mechanical)
16 Jan 08 16:13
CheckerRon-
Londonderry, New Hampshire.
Yes It is correct, I'm looking at it right now and
0.5 up to 3 is +/-0.1

That is because I'm using ISO-mKE  m = medium tolerance class.  If I was using ISO-fKE then it would be
0.05 up to 3  is +/- 0.05mm


I prefer this tolerance system based on length as opposed to decimal for.  At least for DIN/ISO drafting.  

Frank
KENAT (Mechanical)
16 Jan 08 16:28
LONDONDERRY, have you read the last paragraph of part one of the standard?  To my mind it pretty much makes it meaningless.

I have real issues with 2768 but as I said before part of that may have been the implementation of it I've seen.

I'd also be concerned that overreliance on 2768 might make people forget about tolerancing properly (although that can happen with block tols too)

KENAT, probably the least qualified checker you'll ever meet...

LONDONDERRY (Mechanical)
17 Jan 08 7:05
KENAT-
For years I used ANSI decimal tolerance system, while its very good, I don't see it as a practical way of using it metric DIN/ ISO.  Mostly because very rarley do you see metric drawing go down to 3 decimal places.  The one thing I do like about using 2768 is the tolrance is based on the length of the part being machined.  It is harder for a machinist to hold tight tolerance on longer lengths that on smaller parts.
On the subject of proper tolerancing on drawings.  I blame this on the lack of college  and professional education I've seen most mechanical engineers have on drafting standards.  I've seen some ME's place very dimension to 3 decimal places because they lack the understanding.
 While I'm all so a ME I have my drafting degree and years of experience before I made my leap.  However, I still need to consult my drafting books all the time.


frank
KENAT (Mechanical)
17 Jan 08 12:14
Londonderry,

As a guide to what tolerances are typically achievable from a manufacturing point of view I agree 2768 looks pretty useful but as a tolerancing standard to reference from the drawing I have concerns, biggest of which is:

last section of 2768 part 1 - A.4.

Quote:

Unless otherwise stated, workpieces exceeding the general [geometric] tolerances shall not lead to automatic rejection provided that the ability of the workpiece to function is not impaired.

So it doesn't give an explicit pass/fail criteria which I thought was probably the most important thing for a drawing.

KENAT, probably the least qualified checker you'll ever meet...

fcsuper (Mechanical)
17 Jan 08 14:46
Actually, why does it give pass/fail criteria at all?  It only needs to state how tols are determined.  Pass/fail is determined by the system in use.

Matt Lorono
CAD Engineer/ECN Analyst
Silicon Valley, CA
Lorono's SolidWorks Resources
Co-moderator of Solidworks Yahoo! Group
and Mechnical.Engineering Yahoo! Group

KENAT (Mechanical)
17 Jan 08 14:52
Um, am I missing something.  Tolerances surely give the dimensional pass fail critera?  If the dimension is within tolerance it passes.  If out side it fails.

Tolerances shouuld come from the part requirements which the drawing defines...

Maybe this is an ISO/ASME issue but it seems pretty dumb to me that as a general rule a part can be made not to drawing but that that isn't reason for rejection.

In practice I'm sure many of us have been in situations where parts have come in out of tol and due to cost/timescale we've had to determine if they can be used anyway or come up with some scheme to rescue them but that should be very much the exception.  That wording in the standard makes it sound like the norm.

KENAT, probably the least qualified checker you'll ever meet...

JMarv (Mechanical)
17 Jan 08 15:14
I think you're misinterpreting that last paragraph and/or taking it out of context.  Pass/fail criteria may be the most important thing for a drawing, but isn't fit/function the most important thing on the finished product?  I think that part of 2768 is geared toward mass production.  As a tier-2 automotive parts supplier, there is a BIG difference between receiving a Corrective Action Request and receiving a Defective Material Notice (8D rejection).  If a product is found to be out of spec, but the fit/function is not impaired, a CAR is issued.  The manufacturing process gets fixed, it's documented for the customer, and the issue is normaly closed.  When the fit/funtion is impaired, then a rejection is issued.  This may involve sending someone to the customer to sort tens of thousands of pieces (expensive!), paying the customer for each hour their assembly process was shut down due to the parts being quarantined (more expensive!), and paying the auto manufacture for each hour their assembly plant was shut down (don't even ask how expensive!). Not to mention all the months worth of paperwork to follow up afterward.  Having ISO 2768 invoked keeps "reject happy" purchasing agents from making their customers jump through all these hoops when it's unnecessary.

In my situation, I'm refering metal stampings that are mass produced at 30-40 pcs per minute.  In the industry that you work in, things may be viewed differently.

Joe
SW Office 2006 SP5.1
P4 3.0Ghz 1GB
ATI FireGL X1

KENAT (Mechanical)
17 Jan 08 16:02
Jmarv, while there are some exceptions, I'd like to think most of the time that if the drawing is done properly (particulary tolerancing) it pretty well reflects the fit/function of the finished product.  In fact that's kind of the point isn't it, to give a definitive requirement for the part.  If it meets the requirement it's good if it doesn't then it's bad.  My boss frequently talks about the drawing being a legal document defining what you'll accept.

If parts are coming in that don't meet the drawing requirements but are fit/function OK shouldn't the drawing perhaps be changed to better reflect the real requirement?

Like I said I know there are situations where out of spec parts are accepted or some how salvaged.  In aerospace I got involved in this, we made some parts that were fairly high production and was once involved in scrapping several days worth of production which as you say is painful.

I hope you're right and I'm misinterpreting it, perhaps it is just meant to allow dispositioning of defective parts but I still don't like it.

Anyway this has got way off topic and I'm starting to rant, sorry.

KENAT, probably the least qualified checker you'll ever meet...

fcsuper (Mechanical)
17 Jan 08 17:10
Yes, drawings should detail form/fit/fuction accurately.  If a looser tol is ok, then it should be used in the first place, or when discovered; just as it would be made tighter if the converse is discovered.  That line doesn't make much sense, and I question why it is even in the standard at all.

Matt Lorono
CAD Engineer/ECN Analyst
Silicon Valley, CA
Lorono's SolidWorks Resources
Co-moderator of Solidworks Yahoo! Group
and Mechnical.Engineering Yahoo! Group

mwbaker (Mechanical)
5 Mar 08 11:56
Instead of being bound by the entire drafting standard, at my company we call out an exception in our notes as shown below:

1. APPLICABLE STANDARDS/SPECIFICATIONS:
ASME/ANSI Y14.5M-1994, DIMENSIONS AND TOLERANCES.
ASME/ANSI Y14.38-1999, ABBREVIATIONS.
EXCEPTION: TRAILING ZEROS DENOTE TOLERANCE.

I'm sure that everyone has written a requirement based on a specification and up front called out the exception to the standard. It should be OK to note your exception to the standard since you are declaring it up front.
KENAT (Mechanical)
5 Mar 08 12:03
I don't see a problem with that in principle.

In fact we do something similar with metric threads, as well as calling up B1.13M we say that threads without pitch specifed are coarse pitch.  This then brings our drawings in line with the rest of the worldwinky smile.

KENAT, probably the least qualified checker you'll ever meet...

CheckerRon (Mechanical)
5 Mar 08 12:55
mwbaker: Since this thread is addresssing metric tolerances, I hasten to add that in metric dimensioning, trailing zeros are dropped and do not define the tolerance. The last significant digit defines the tolerance.
KENAT (Mechanical)
5 Mar 08 13:02
Ron, I think that's his point.  By adding his note he's adjusting it so the same convention on number of decimal places invoking block tolerances can be used with metric & inch.  I'd probably clarify with something like "EXCEPTION: TRAILING ZEROS INVOKE BLOCK TOLERANCE ON METRIC DIMENSIONS UNLESS OTHERWISE STATED." but i don't see the problem in principle, it's a bit like what we did with metric threads as I said.

KENAT, probably the least qualified checker you'll ever meet...

CheckerRon (Mechanical)
5 Mar 08 16:42
OK KENAT, but seems like confusion for convenience, and not something I would want to do. Here we have a separate metric format and an english(not really),imperial(not either), American (yeh!)format, as do you. After all it is not EMSE, it's ASME.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!

Close Box

Join Eng-Tips® Today!

Join your peers on the Internet's largest technical engineering professional community.
It's easy to join and it's free.

Here's Why Members Love Eng-Tips Forums:

Register now while it's still free!

Already a member? Close this window and log in.

Join Us             Close