Major benefits of GD&T usage 6

[pmarc]

I would like to start this thread based on some thoughts that came to my mind during answering to Evan's post about surface interpretation of position tolerance:

http://www.eng-tips.com/viewthread.cfm?qid=286461&page=1

As I said there, I have a feeling that quite a lot of GD&T users (at least the ones I work with) would say that hard gaging is the only way for checking positional tolerance at MMC. But since they know it is not really possible to use gage simulators in verifying their design (let's say because there is a company strategy that prohibits hard gaging usage), they simply do not apply M modifier, even if they feel it would be reasonable or even required from functional point of view. By doing this they intentionally or unintentionally loose one of the biggest advantages of GD&T language over traditional coordinate dimensioning - tolerance zone is becoming unnecessarily fixed in size regardless of feature of size.

That's why I wanted to ask you Guys about your experience with any other important benefits of GD&T usage that you feel are not utilized properly or not used at all in your everyday's pratice.

Any remarks are really welcomed.

 

[powerhound]

321GO,
It looks like you've taken your ball and gone home but in case you're still lurking just to check out any further responses, I'm going to add my two cents.

JP was giving you tremendous amount of leeway when he conceded that any increased costs associated with the M modifier could be attributed to training. Thats WAAAAAY more than anyone else would have likely given you. The M modifier reduces cost...period. If you don't use it just because you're afraid someone won't understand it, you're still not making a less expensive part. This argument is more accurately directed at GD&T in general. You can't just have a training class in the use of the M modifier for bonus tolerance. That class would take all of 5 minutes. Explaining its use regarding datum shift is a totally different thing, but that's not what we're talking about. The bonus tolerance allowed by MMC is very easily understood once you sit down and think about it.

Your argument about the M modifier not being intuitive is true, but only in the sense that nothing about GD&T is intuitive. Intuitive means that you can just look at something with no prior knowledge and be able to tell how it's used.

No matter how offensive it sounded, drawoh was spot on with his comment. Shop floors are rife with screwed up prints that drafters think are perfect but the machinists have just learned to deal with. When good parts are produced, the drafters think the prints are good.

If you continue your current thought process that just because we are in the modern CNC era, you can apply unnecessarily tight tolerances with no added cost, you are going to cost your company more money than any MMC modifier ever could. If you really want to control costs, you should start with that...AND you should listen to what people like JP and others are saying. They've been on this board for years and have dealt with these issues on a daily basis. You are the only one on this thread that thinks the MMC modifier adds cost. You should consider that someone else might know something more about this than you.

Powerhound, GDTP T-0419
Engineering Technician
Inventor 2010
Mastercam X4
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[KentMc]

I have seen the good and the bad of GD&T. Some companies view it as a miracle cure that you slap some more information on a drawing and get good parts. The truth is poorly applied, GD&T will cost a company more. It can be like the new toy for engineers to play with and over specify parts. Even worse to have grossly contradicting or incorrectly specified elements.

At the same time, it provides a simple and common means to communicate design intent. Whenever a standard tolerance band applies, by all means use it. But by using GD&T, including the "bonus" tolerance, you can dramatically reduce costs. How many times are parts found to be defective only to then be treated as "use-as-is"? Why, because the default tolerance was easy to use and understand, but did not fully communicate what the design could tolerate. Or parts carefully and methodically CNC machined to ultra-tight tolerances (because they CAN), but at the cost of being less tolerant of how the part actually functions?

If your industry or design doesn't really need them then don't use them. But you will find in high performance or well optimized designs, they will have a place. The cost savings of understanding how to communicate the design through effective GD&T truly outweighs the investment in training. But if your company insists that CNC operators just turn the machines on and off and shouldn't need to understand tolerances, then GD&T isn't for you. Those are parts most likely to be made in China. Sloppy designs, either over-specified (expensive) or underspecified (expensive-prone to not function as needed) are just ridiculous. Robust design analysis should tell you when you need GD&T. Otherwise don't use it.

 

[ptruitt]

I like the fact that the trend is to reduce the number of GD&T symbols. Many folks are boiling it down to profile of a surface and position. I am not too knowledgeable, unfortunately, but I think there may be people reading this who could give examples of where it is possible to replace the position tolerance/M-modifier application with a simpler profile call-out.

 

[fcsuper]

As we move more into Model Based Definition (relying on the CAD's digital data set such as solid models and associated metadata) to spec parts, GD&T will become more important again. I'm sure organizations that are moving into MBD are requiring GD&T.

In many cases, it can save money and is quicker to draw on a drawing. In some cases (as when its being used improperly), it can cost money. However, there are legitimate uses that may be time consuming to figure out how to do them correctly if you aren't familar with that portion of the standard.

Matt Lorono
Lorono's SolidWorks Resources & SolidWorks Legion

http://groups.yahoo.com/group/solidworks & http://twitter.com/fcsuper

 

[KENAT]

The basic concept of M is actually simple for typical applications:

"As the clearance hole gets bigger, it's location can vary more and still allow the screw/bolt through"

Really simple, if the holes bigger, it doesn't need to line up as well. Now applying this properly is a bit trickier but not rocket science.

M allows you more tolerance, I struggle to see how this doesn't tend toward being cheaper.

You can choose to inspect ignoring M. If it passes great, if it fails, well if you take M into account it may now pass. Or you can decide to scrap the part. Your choice.

321GO, from a US (and to an extent as I experienced it UK) perspective, Function is the primary driver of dimensioning & tolerancing. The most important thing being that parts work.

With the Iso system, at least 2768, it seems that manufacturing is given priority in driving tolerance scheme. 2768 is entirely based around 'what typical shops can achieve' etc.

This perspective, and the predominance of US based posters on this site may explain some of the disagreement.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[321GO]

KENAT (and others),

yes, the German DIN-ISO-2768 is in essence indeed based on the "shop capabilities". The idea is to simplify the drawing(nothing new there).

To be precize, it has 2 parts(1,2). The first part covers tolerances for linear/angular and chamfer dimensions based on "shop capabilities".
The second part covers GD&T also again in different classifications based on shop capabilities.

Generaly it would be stated as below on a drawing:
Tolerances and GD&T according: DIN-ISO-2768 mK (the "m" is a indication for the toleance class and the "K" for the GD&T class)

To ensure a correct fit, one can specify overruling tolerances and GD&T on specific features when needed.

(KENET), is the US approach not nearly indentical in essence?

 

[dtmbiz]

From my 1st GDT class I have been taught was that the whole premise for its use is,
"The ability / mechanism to accept as many functional parts as possible using cost effective tolerancing".

My list of GDTadvantages:

Datums to more clearly define / convey part interface, feature function; inspection setup; mfg;

Cylindrical tolerance zone vs. Cartesian

Form controls

Orientation controls

Rule #1 controls perfect form at MMC (unless accepted by note)

Bonus tolerance relative to MMC / LMC

Composite flatness

Runout

I have been around long enough to have had to use various "verbiage"
on drawings prior to using the 1973 ANSI Y14 standard to "describe" the list above, which can be
open to more varied interpretation than the standard's symbols. concepts and definitions.
In short a good design and communication of that design needs to be
understood by a myriad of people. Whether or not it seems difficult to explain
the concepts, definitions, and application of the standard, just try to do it without the
symbol language of GDT. It is very possible to put out a simple drawing without GDT
but more than likely it will not consider nor communicate the design intent nearly
as well or completely compared to using GDT.

 

[KENAT]

No, the US approach is not nearly identical.

ASME Y14.5M-1994 - the Dimensioning & Tolerancing standard - puts the emphasis clearly on function.

2768 - which I am unfortunately moderately familiar with - has a bit at the end which (to paraphrase) essentially says a part doesn't have to meet the requirements of the standard, it should be accepted or rejected based on function.

Well if the drawing doesn't pretty much capture the required the functional requirements, then what does?

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[ptruitt]

I think Kenat makes a very good point about function. In my opinion, one attribute of a good DESIGN documentation package is that it is goal-oriented, not process-oriented. (There will always be exceptions. Try to make a design drawing of a packed parachute. So, sometimes, the design will need to include the process.) Of course the design must be process-capable and designers need to know what is reasonable. It sounds like DIN-ISO-2768 is going to stifle innovation and result in designers that don't collaborate with manufacturing engineers. Without that collaboration, expect suboptimal products. Perhaps ASME is better in that respect.

 

[CheckerHater]

KENAT,
Comparing 14.5 with 2768 is not even "apples to oranges". It's rather "apples to orange peels".
14.5 is EVERYTHING about dimensioning and tolerancing.
2768 is merely dealing with dimensions without explicit tolerances attached to them.
In that sence the closest thing to 2768 would be ANSI B4.3, which unfortunately now is "missing in action".
And about "non-rejection" the standart clearly sais "Unless otherwise stated", so it is your right (and responsibility) to state oterwise. You can even enter some legalese like "Clause A.4 does not apply" in your contract with manufacturer.
They probably don't ask for your advise when writing a contract (I know they don't care about mine), but that's different problem.

 

[KENAT]

I wasn't comparing 14.5 as a whole to 2768.

Just the concept that ASME accentuates function, while 2768 accentuates manufacturing without regard to function.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​