GD&T and manufacturing process 1

[Dhavalvp]

Hi all
I m working on a typical problem regarding identification of the relationship between the GD&T and different manufacturing processes.
I would like to know that if there is any formulas or guidelines available for assigning different feature tolerances like circularity,cylindricity,parallism for different machining processes like planning ,milling etc?

 

[ctopher]

A good book to have a copy at your desk is the "Machinery's Handbook". In general, you dimension your part and add GD&T for fit/form/function, not for process or machining. You generally don't know how the part will be made in the end.

Chris
SolidWorks/PDMWorks 08 3.1
AutoCAD 06/08
ctopher's home (updated Apr 30, 2008)

 

[drawoh]

Dhavalvp,

You need to chat with your fabricator. There is a chart in the Machinery's Handbook showing the ISO tolerance classes that can be achieved with a given manufacturing process. Search around the limits and fits pages.

You are not supposed to provide manufacturing instructions on your fabrication drawings, but I have rarely proceduced fabrication drawing where it was not obvious what fabrication process would be used. When you know this, you should know what tolerances can be reliably specified on your drawings.

JHG

 

[KENAT]

Agree with above, function should drive the tolerancing & you don't normally define the manufacturing process on drawing unless it materially affects performance.

However, you don't want to end up with tolerances that can't be met except by expensive processes unless you really have to. Sometimes a change in design can allow looser tolerances and this is good practice, to know when to do this you need some idea of the process capabilities.

If you have your own shop, or one/few external shop/s that does/do most of your work you may be best off talking to them.

The chart in Machineries is probably based on ISO2768. This is a DOG of a standard that should not be invoked in its entirety by any self respecting Engineer/Designer (see recent posts on the subject by myself and others). However using it as guidance on what are typically achievable tolerances may have merit. Extracts of this standard are available on the net if you do a google search. I may have even linked them in my earlier posts on this standard.

Also, while I'm not sure it has exactly what you need, Ctopher posted a few months ago a pretty good link to a machine shops website with information on what causes them problems/make it easier for them. Take some of it with a pinch of salt as it's biased heavily to making the machinists job easy without much consideration for function but it has some good stuff.

Finally I seem to recall Drawoh having a website with a little such info, or was it dingy, one of them.

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

 

[Dhavalvp]

thsnk you all
actually you all have make my problem easy,but while surfing about ISO2768 I found another standard i.e.ISO 2768-2:1989. it is something about "General tolerances -- Part 2: Geometrical tolerances for features without individual tolerance indications." but on some of the sites it is written that it is highly recomended that u should not use this standard! can anyone expalin about this standard and about the notification?

 

[KENAT]

ISO 2768 is a 2 part standard. Part one gives +- dims for various size features and various ‘qualities’ of machine shop. Part 2 gives information for geometric tolerances.

As to why this standard is a @#%%$^@#$% @#$%$@%!@# @#$%$# piece of #$%@#$5234; I’d have hoped that looking at my and others earlier posts on the subject would have answered that. Take a look at these links then feel free to draw your own conclusion.

thread1103-216401

thread1103-196260

thread1103-197786

thread182-208001

thread182-197787

http://www.tec-ease.com/the-new-gd&t-article.htm

In summary the last paragraph of part one says:

Unless otherwise stated, workpieces exceeding the general geometrical tolerance shall not lead to automatic rejection provided that the ability of the workpiece to function is not impaired

Most designers/engineers I know would argue that a major point of the drawing is to define what tolerances need to be met to maintain function, and that hence invoking this standard in its entirety may not be that smart.

Out of interest what drawing standards do you work to? Most of the people that really have issue with 2768 are I believe ASME or ex BS users. Perhaps to an ISO die hard it’s more palatable.


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

 

[JMarv]

Dhavalvp,

To expand on what Kenat stated, 2768-1 (when invoked) is a general tolerance for untoleranced dimensions. If you like to rely on a general "title block" tolerance to define an entire drawing, this standard may cause problems. (See Kenat's quote). Since the standard only applies to untoleranced dimensions, you should add a specific tolerance to any dimension that you don't want the standard invoked on.

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

 

[KENAT]

I suppose the other reason I don't like it is that different people seem to interpret it differently. The way the German supplier I worked with used it didn't match what I understood the spec said or what anyone else I could find thought it meant.

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

 

[Xplicator]

If you would really like to guide the vendor to a particular manufacturing process, you can do so with surface finish specifications.

Hope this helps.

 

[Dhavalvp]

thanks KENAT
thanks for the links that u have given.
I have another doubt regarding GD&T,while refering to the ANSI Y14.5 standard I came to know about the formulas like the fixed fastner and the floating fastner that are used to allocate the postional tolerances.
Is there any other such formulas avilable for allocation of different geometric feature tolerances ?

 

[KENAT]

Coaxial features when using position is in 14.5, just after fixed fasteners.

I can't think of anything else off the top of my head.

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

 

[GMIracing]

I agree with ctopher, tolerances should be based on fit, form, and function. Make the tolerances as large as possible, but no larger. Tolerances should be used to ensure the parts fit together, maintain clearance/distance from other parts of the assembly, and that everything functions.

 

[MechNorth]

ISO 2768 is based on average shop capabilities at the time the studies were undertaken. 2768 dates back to at least 1973 (that's the edition replaced by 2768-1 and 2768-2 in '89). Let's say that they took 20 years to collect, collate, validate and publish the data (if you work with standards groups, you know this isn't far off!). So that puts us back to the early/mid 1950's when they started collecting the information. So, with an average shop using let's say 20 year old machinery, that would take you potentially back to machinery from the '30s. While some of those manual machines were far better than the average machinery produced today, I tend to think that the "general shop capabilities" that the 2768 standards represent are not representative of the NC machine capabilities in use in most shops today. In addition to the questionable relevance to today's shop, 2768's tightest class still doesn't come close to what we would typically call a precision tolerance today, so the standard has limited relevance to today's industry as well.

If you are looking for appropriate tolerances for specific interfaces, then Machinery's Handbook, ISO 286, and other standards which specify fit classes are a great start.

Jim Sykes, P.Eng, GDTP-S
Profile Services
CAD-Documentation-GD&T-Product Development

http://www.profileservices.ca

 

[KENAT]

MechNorth, you make a good point on how current the information contained there in is.

Although I'm tempted to say that Germany where the standard originated probably re-tooled significantly in the 50s so the figures may not be quite based on 30s tooling - though it's not beyond belief.

The same goes for the charts you see showing cost increase with tighter tolerance.

I'm forever having people tell me they spoke to the machine shop and they say they can routinely hit +-.002 or tighter with no problem/cost increase.

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

 

[fcsuper]

.002? On CNC, that's huge!

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

 

[MechNorth]

Good point, KENAT. I used to have managers demanding that we open tolerances "to reduce costs". So, we'd re-design the components to shift the tighter tolerances to other pieces and magically the cost of the original piece didn't decrease, but the cost on the other piece often went up. What they (managers) weren't realizing was that if you aren't changing machines & processes, then your basic capability stays the same over a range of tolerances. Management by Magazine ... read an article ... believe the article ... try to implement the article ... fail to achieve the same results ... No, really?

Jim Sykes, P.Eng, GDTP-S
Profile Services
CAD-Documentation-GD&T-Product Development

http://www.profileservices.ca

 

[KENAT]

fcsuper, we routinely need that and tighter but when we don't I don't see the point in specifying it anyway.

I've worked places where sure the newer machines could do it day in day out but some of the older machines struggled a bit. Putting tighter tolerances meant that when there was a lot of thru-put they couldn't use the older machines so things got delayed.

While what constitutes a tight tolerance may have come down a lot, I think the principle of not specifying tighter than you need holds.

Otherwise I'll just change our block tol to +-.001 and never do another tol calc;-)

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