Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations TugboatEng on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
Law of diminishing returns and GD&T - open discussion- 7
- Thread starter aniiben
- Start date
An old adage that shapes my view on your question goes something like this: "We cannot simply make drawings that can be understood. We must make drawings that cannot be misunderstood."
Of course someone could put too much GD&T on a drawing. But that's not the GD&T's fault -- it only means they don't understand the proper use of GD&T.
John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems
Of course someone could put too much GD&T on a drawing. But that's not the GD&T's fault -- it only means they don't understand the proper use of GD&T.
John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems
I never use GD&T on drawings. Feature control frames and related datum feature symbols maybe.
The reality is that FCFs are a way to reliably define how a part is to be oriented in preparation for making a comparison to a well defined tolerance zone. Not that users always manage to accomplish that, but the alternative is to use a significant portion of assumed convention.
For example I saw an entire project** that had no controls on the angle relationship between almost all faces of parts. No default angle tolerance and no angularity tolerances for anything but non-90degree angles. Just left to the fabricators and inspectors to decide what they wanted to do. No one complained and parts seemed to fit.
I had a supplier strike off all the interface dimensions on an interface control drawing. We had created the drawing because the supplier had sent non-interchangeable parts for an application that basically required interchangeability. His reason? Because his company's inspection would not agree to inspect the parts. When I say "his" it's because he was the President of his company. Sigh. I had even discussed this before creating the drawing with the allowance that if the particular scheme of references wasn't what they worked to they could change it. The markup looked like a victim in a Saw movie.
If there's a desire to prove what you are asking for, use FCFs. If it doesn't matter what might happen, then don't. Maybe it will work out; maybe not.
**Hundreds of drawings, some were complex machinings, not a 90 degree angle tolerance on any of them.
The reality is that FCFs are a way to reliably define how a part is to be oriented in preparation for making a comparison to a well defined tolerance zone. Not that users always manage to accomplish that, but the alternative is to use a significant portion of assumed convention.
For example I saw an entire project** that had no controls on the angle relationship between almost all faces of parts. No default angle tolerance and no angularity tolerances for anything but non-90degree angles. Just left to the fabricators and inspectors to decide what they wanted to do. No one complained and parts seemed to fit.
I had a supplier strike off all the interface dimensions on an interface control drawing. We had created the drawing because the supplier had sent non-interchangeable parts for an application that basically required interchangeability. His reason? Because his company's inspection would not agree to inspect the parts. When I say "his" it's because he was the President of his company. Sigh. I had even discussed this before creating the drawing with the allowance that if the particular scheme of references wasn't what they worked to they could change it. The markup looked like a victim in a Saw movie.
If there's a desire to prove what you are asking for, use FCFs. If it doesn't matter what might happen, then don't. Maybe it will work out; maybe not.
**Hundreds of drawings, some were complex machinings, not a 90 degree angle tolerance on any of them.
-
1
- #4
anniben: This is how I respond to your question, which there is no real answer
At it's foundation, GDT is a philosophy - a way of thinking about the size, form, orientation and location characteristics of part features. If you are really trying to comprehend how your parts "function and fit-up", GDT "thinking" must be involved. Also, nothing is perfect, everything has error. Defining the limits of error is critical to knowing if the part will work as the design requires. All the symbols do is provide tools to communicate this information. The Y14.5 Standard is a dictionary, allowing the "words" (symbols) of a technical language to be interpreted the same by everyone - at least that's the goal.
The "amount" of GDT applied to a drawing depends of the level of risk you are willing to accept if the above is not communicated clearly. The risk directly translates into dollars spent dealing with scrap, rework, loss of customers, employee turnover, plus all of time wasted arguing if the parts are acceptable or not because no one defined how bad it can be. To me: more GDT - less risk - more profit.
One of our customers kept referring to "make it to the shop standards everyone uses" (paraphrased) when something on their drawings was not clear to us - no GDT to be seen anywhere. I asked them to get me a copy of these "shop standards" so we can use them to interpret their drawings. I never got a copy. It was just a cop-out. We ended up telling them what tolerances we could work to - based on our mfg capability and not their design intent - and got them to sign-off. The tolerances may have been much better or worse than required by the design. They may have paid for something they didn't need, making them less competitive, or they didn't get what the design needed and the part didn't work as intended. Choices will be made and consequences found!
Certified Sr. GD&T Professional
At it's foundation, GDT is a philosophy - a way of thinking about the size, form, orientation and location characteristics of part features. If you are really trying to comprehend how your parts "function and fit-up", GDT "thinking" must be involved. Also, nothing is perfect, everything has error. Defining the limits of error is critical to knowing if the part will work as the design requires. All the symbols do is provide tools to communicate this information. The Y14.5 Standard is a dictionary, allowing the "words" (symbols) of a technical language to be interpreted the same by everyone - at least that's the goal.
The "amount" of GDT applied to a drawing depends of the level of risk you are willing to accept if the above is not communicated clearly. The risk directly translates into dollars spent dealing with scrap, rework, loss of customers, employee turnover, plus all of time wasted arguing if the parts are acceptable or not because no one defined how bad it can be. To me: more GDT - less risk - more profit.
One of our customers kept referring to "make it to the shop standards everyone uses" (paraphrased) when something on their drawings was not clear to us - no GDT to be seen anywhere. I asked them to get me a copy of these "shop standards" so we can use them to interpret their drawings. I never got a copy. It was just a cop-out. We ended up telling them what tolerances we could work to - based on our mfg capability and not their design intent - and got them to sign-off. The tolerances may have been much better or worse than required by the design. They may have paid for something they didn't need, making them less competitive, or they didn't get what the design needed and the part didn't work as intended. Choices will be made and consequences found!
Certified Sr. GD&T Professional
aniiben,
I systematically use GD&T on my drawings. GD&T is a language that accurately and unambiguously describes what I want fabricated. If my drawing gets covered by GD&T symbols, it is because the part is complicated, and full of accurate features. Drawings like this are really nasty to do without GD&T.
Your drawing would consist of plain dimensions, no boxes.
--
JHG
I systematically use GD&T on my drawings. GD&T is a language that accurately and unambiguously describes what I want fabricated. If my drawing gets covered by GD&T symbols, it is because the part is complicated, and full of accurate features. Drawings like this are really nasty to do without GD&T.
A perfectly legal GD&T note said:
Your drawing would consist of plain dimensions, no boxes.
--
JHG
- Thread starter
- #6
Thank you very much for all the answers.
I just like to reinforce that I am questioning if there is such a point (of diminishing returns) or you think there is absolutely not.
I am not suggesting of not using this heavy technical language, but inquiring when it is too much.
I just like to reinforce that I am questioning if there is such a point (of diminishing returns) or you think there is absolutely not.
I am not suggesting of not using this heavy technical language, but inquiring when it is too much.
-
1
- #7
I expect that many here have a different experience. Based on the questions from many visitors on this forum there are a lot of people who have experience more similar to mine.
I worked in one place where the producer didn't even use the drawing and neither did inspection. The source had built to the original model and to smaller variations than were originally called for. Production problems had occurred and they asked for changes, changes they made to the producer manufacturing models, that were never incorporated into the company drawings. This became evident because there was a mirror assembly version that the changes were, partly, incorporated into. When I was brought in to fix a pile of other problems with the drawing package I noted the differences and then, comparing them with finished items on the factory floor, found a bunch more.
There's not a diminishing returns place - one might never use FCFs at all and get a high quality product. Like I said, the only reason for FCFs is to predict what you might get and confirm that's what you got, but if you can put a part into assembly production without inspecting it and it works, it doesn't matter at all what's on the drawing.
In the industry parts I've dealt with, both military** contracting and FDA controlled medical equipment, they could have eliminated all use of FCFs and seen no change at all. More time was spent arguing over minor details of FCFs while simultaneously making horrific datum reference choices than I care to recall. Those choices would allow completely useless parts to be accepted. And then forgetting to specify any tolerances at all in many cases. And it made no difference because the CNC equipment was far more reliable at producing near perfect form and orientation than the tolerances allowed.
What I would do if money was the biggest deal would be to just put plain dimensions and tolerances on sizes and locations; mostly the tolerances on dimensions would be for close/bearing fits, not on typical dimensions. Only if an assembly problem came up would I bother with an FCF. But it pays to put a few of them on a drawing so management thinks some actual work is being done.
The only time I would care about using FCFs is if the company was managing yield and using factory measured variations to drive the engineering design yield calculations. Note this is not process control charting; that just makes Quality managers happy. If the company is not closing the loop on factory floor quantitative results and their manufacturing costs with design engineering, then nope. Not worth it. I'd keep the Envelope principle and Continuous Feature and chuck the rest. If someone wanted something special, I would put a fat note on the drawing.
The dilemma, and I've had my career bashed in over it, is that if you 100% use FCFs to control all aspects of parts you get kicked for doing too much. And then if you take even one FCF off a different faction can use that to complain that you don't know what you are doing at all; the same faction that will select bad datum references and refuse to look at simulation results because "manufacturing would never use all the tolerance anyway." But make the tolerance smaller and a different faction demands the cost will go up and the tolerance can't be changed. And so the ignorance goes.
Like I said, I'd scrap it all unless the company is closing the loop and your job specifically involves using production data on process variability.
**Worked specifically for the US Air Force, the US Navy, TACOM, and the US Marine Corp, all with different procurement expectations for documentation. TACOM was the one accepting drawings without any angle tolerances between nominally perpendicular surfaces/features.
I worked in one place where the producer didn't even use the drawing and neither did inspection. The source had built to the original model and to smaller variations than were originally called for. Production problems had occurred and they asked for changes, changes they made to the producer manufacturing models, that were never incorporated into the company drawings. This became evident because there was a mirror assembly version that the changes were, partly, incorporated into. When I was brought in to fix a pile of other problems with the drawing package I noted the differences and then, comparing them with finished items on the factory floor, found a bunch more.
There's not a diminishing returns place - one might never use FCFs at all and get a high quality product. Like I said, the only reason for FCFs is to predict what you might get and confirm that's what you got, but if you can put a part into assembly production without inspecting it and it works, it doesn't matter at all what's on the drawing.
In the industry parts I've dealt with, both military** contracting and FDA controlled medical equipment, they could have eliminated all use of FCFs and seen no change at all. More time was spent arguing over minor details of FCFs while simultaneously making horrific datum reference choices than I care to recall. Those choices would allow completely useless parts to be accepted. And then forgetting to specify any tolerances at all in many cases. And it made no difference because the CNC equipment was far more reliable at producing near perfect form and orientation than the tolerances allowed.
What I would do if money was the biggest deal would be to just put plain dimensions and tolerances on sizes and locations; mostly the tolerances on dimensions would be for close/bearing fits, not on typical dimensions. Only if an assembly problem came up would I bother with an FCF. But it pays to put a few of them on a drawing so management thinks some actual work is being done.
The only time I would care about using FCFs is if the company was managing yield and using factory measured variations to drive the engineering design yield calculations. Note this is not process control charting; that just makes Quality managers happy. If the company is not closing the loop on factory floor quantitative results and their manufacturing costs with design engineering, then nope. Not worth it. I'd keep the Envelope principle and Continuous Feature and chuck the rest. If someone wanted something special, I would put a fat note on the drawing.
The dilemma, and I've had my career bashed in over it, is that if you 100% use FCFs to control all aspects of parts you get kicked for doing too much. And then if you take even one FCF off a different faction can use that to complain that you don't know what you are doing at all; the same faction that will select bad datum references and refuse to look at simulation results because "manufacturing would never use all the tolerance anyway." But make the tolerance smaller and a different faction demands the cost will go up and the tolerance can't be changed. And so the ignorance goes.
Like I said, I'd scrap it all unless the company is closing the loop and your job specifically involves using production data on process variability.
**Worked specifically for the US Air Force, the US Navy, TACOM, and the US Marine Corp, all with different procurement expectations for documentation. TACOM was the one accepting drawings without any angle tolerances between nominally perpendicular surfaces/features.
A special thanks to everyone for sharing their experiences with "how much GDT" to apply.
3DDave's above quote is profound as it can allow GDT to appear to be "irrelevant"....why put tolerances(GDT)on the drawing if we never have issues with tolerances. I worked at a shop where the machine tools, as 3DDave stated, made parts better than they needed to be. Inspection of part features was almost eliminated - especially hole locations - over the years because nothing was found out of spec. Over time this "attitude" migrated into engineering and drawings without complete tolerancing were released. Over the years there was no investment in the machine tools and their precision and accuracy gradually deteriorated. Over time more and more serious errors started showing up at the assembly stage and right in front of the customer witness rep. They started to investigate. Measuring everything in site at great cost! Because the drawings lacked complete tolerancing, no one knew how bad the part could be and still work. The precision required was "lost" in the precision of the machine tools. Of course this "realization" is all 20-20 hindsight and the time scale is years.
Certified Sr. GD&T Professional
3DDave said:
3DDave's above quote is profound as it can allow GDT to appear to be "irrelevant"....why put tolerances(GDT)on the drawing if we never have issues with tolerances. I worked at a shop where the machine tools, as 3DDave stated, made parts better than they needed to be. Inspection of part features was almost eliminated - especially hole locations - over the years because nothing was found out of spec. Over time this "attitude" migrated into engineering and drawings without complete tolerancing were released. Over the years there was no investment in the machine tools and their precision and accuracy gradually deteriorated. Over time more and more serious errors started showing up at the assembly stage and right in front of the customer witness rep. They started to investigate. Measuring everything in site at great cost! Because the drawings lacked complete tolerancing, no one knew how bad the part could be and still work. The precision required was "lost" in the precision of the machine tools. Of course this "realization" is all 20-20 hindsight and the time scale is years.
Certified Sr. GD&T Professional
TheTick:
My use of word "drawing" includes all aspects of the design organization - anyone who approves the drawing. In my example there were small glimpses of the problem over the years, but the whistle blowers were responded to with "yeah right".
Certified Sr. GD&T Professional
My use of word "drawing" includes all aspects of the design organization - anyone who approves the drawing. In my example there were small glimpses of the problem over the years, but the whistle blowers were responded to with "yeah right".
Certified Sr. GD&T Professional
-
3
- #14
Law of diminishing return states that even the most beneficial principle could be harmfull if it's carried far enough.
Also I read on the web an article from Alex K and among other of his opinions he stated:
Silly Thing #4: Sure this stuff is all well and good, but those !#@%**! (for "!#@%**!” substitute; "guys in the shop", or "vendors", or "inspectors", etc.) won't/don't understand it.
This is the "Those Other Dummies" argument. In using it, the speaker allows as how "we" understand alright (because we're so smart), but "Those Other Dummies" won't or don't, and because they won't or don't, "we" shouldn't use GD&T.
While it is true that not all people understand tolerancing (or, for that matter, any number of other technical subjects) as well as "we" might wish, it is also true that many people in many functions in many organizations understand it quite well. In those cases where ignorance actually exists, the answer is education, not capitulation.
If you assume that your audience is illiterate, why are you sending them written messages? It is self-defeating to assume incompetence on the part of "The Other Guy”. 'We" must assume that the people who receive our drawings are capable of reading them, or why prepare drawings at all?
Also I read on the web an article from Alex K and among other of his opinions he stated:
Silly Thing #4: Sure this stuff is all well and good, but those !#@%**! (for "!#@%**!” substitute; "guys in the shop", or "vendors", or "inspectors", etc.) won't/don't understand it.
This is the "Those Other Dummies" argument. In using it, the speaker allows as how "we" understand alright (because we're so smart), but "Those Other Dummies" won't or don't, and because they won't or don't, "we" shouldn't use GD&T.
While it is true that not all people understand tolerancing (or, for that matter, any number of other technical subjects) as well as "we" might wish, it is also true that many people in many functions in many organizations understand it quite well. In those cases where ignorance actually exists, the answer is education, not capitulation.
If you assume that your audience is illiterate, why are you sending them written messages? It is self-defeating to assume incompetence on the part of "The Other Guy”. 'We" must assume that the people who receive our drawings are capable of reading them, or why prepare drawings at all?
-
2
- #15
I am trying to provide everyone who has to work with my drawing (fabricators, vendors, MFG, QA, etc) a fully defined part. That means no feature on the part is ambiguous or open to more than one single interpretation. When there is ambiguity people are forced to make assumptions about what is required. Many times these assumptions will be incorrect.
When people tell me that there is "too much" and/or "unnecessary" GD&T on my drawings I always tell them that we can choose to require inspection on every feature or something less than 100% of features (if we are willing to take some risk). However, if we cannot agree what the part should be we are doomed from the start. How can we agree on what the part should be if it is not fully defined?
In short my philosophy is:
1) fully define the part.
2) inspect how ever many features "makes sense".
3) fall back on the full definition of the part if problems arise with step 2.
When people tell me that there is "too much" and/or "unnecessary" GD&T on my drawings I always tell them that we can choose to require inspection on every feature or something less than 100% of features (if we are willing to take some risk). However, if we cannot agree what the part should be we are doomed from the start. How can we agree on what the part should be if it is not fully defined?
In short my philosophy is:
1) fully define the part.
2) inspect how ever many features "makes sense".
3) fall back on the full definition of the part if problems arise with step 2.
- Status
Similar threads
- Locked
- Question
- Locked
- Question
- Question