Not using GD&T? 10

[ctopher]

My last company, and current, have argued with me about using GD&T.
Engineers, purchasing, and managers have argued that it makes the parts more $$.
I tell them no it doesn't.
Last company had parts machined in China, current in Thailand and in USA.
China and Thailand have said not to use it because they don't understand it.
Often I see parts made that don't meet print with GD&T, but are bought off anyway.
I'm at wits end, tired of arguing with everyone.
I also find more people here in USA that don't understand it.
Anyone here run into this? If so, what do you do?

Chris, CSWP
SolidWorks '20
ctophers home
SolidWorks Legion

 

[Burunduk]

I wasn't implying that you said that manufacturing comfort is preferable over functionality. My point was that doing it the way they did can't be explained away by manufacturing comfort or by "it doesn't matter" as they told to ctopher.
Manufacturing only needs the holes dimensioned from the edges. It doesn't need them toleranced with position relative to the edges to be able to fixture to the edges and produce them reliably.
If the designers chose edges from opposed corners on the two parts as datum features, that's a warning flag that they don't know what they are doing.
This is why ctopher brought that example.
A company should prevent that by motivating employees to learn proper tolerancing and communicating the importance of it to them.

If you don't know how to specify tolerances that are both functional and manufacturable by different common technological solutions but still expect to get quality parts that will always work, then you depend on that one supplier (maybe the Trumpf you mentioned?) who will make it for you very close to the CAD model. That supplier can charge you whatever he wants.
Or if you manufacture in-house you can't do without the expensive machines, tools, and processes that will make the entire part very accurate, even the features that don't need that accuracy. That's a waste. Having designers machinists and inspectors acquiring knowledge needed to do their work properly costs less.

 

[3DDave]

"If the designers chose edges from opposed corners on the two parts as datum features, that's a warning flag that they don't know what they are doing. "

Or it's a sign that it is so insignificant a decision that no one cares to go and rework what different teams did independently because the parts, as fabricated, function correctly.

"A company should prevent that by motivating employees to learn proper tolerancing "

Yeah, strike them with a whip over aesthetics so they hate the topic over a detail that doesn't matter. For sure "pay them double" isn't going to happen.

That's not part of proper tolerancing. That's part of a thoughtless devotion to drawing appearance and can be an extra expense with only one person who never uses the product seeing that.

Two factors limit tolerance - stress and deflection. If that's not part of your tolerance analysis then you have never done tolerance analysis. You've done "thoughtless devotion."

You don't care for Trumpf? Why have you got an opinion on sheet metal then? If you don't know the tools you cannot make decisions on what will be economical to make with them. Is your sheet metal done by starving children with hand files so that the quality of the work can be adjusted on a feature-by-feature basis?

 

[ctopher]

I agree about having employees trained. But, it's not that easy because a lot of companies are scaling back and cutting costs.
I recently interviewed at a couple companies. They want designers/engineers with GD&T knowledge already, not willing to train.
The only reason I didn't get the jobs is because they want to pay at low end.
These companies told me they use GD&T, but don't have enough knowledge to implement most of it. So they need someone that knows it.
They also told me that most interviewed didn't know it, or very little.
This is real world.

Chris, CSWP
SolidWorks '20
ctophers home
SolidWorks Legion

 

[Burunduk]

3DDave,
You don't know that parts as fabricated function correctly, or even that the edges are not functional for the alignment of the holes. That is your wishful thinking. We do know, as ctopher indicated, that the drawings are still being discussed. You still didn't provide an explanation on how referencing nonfunctional datum features in a geometrical control aids manufacturing.
If it was done without giving any thought to function or the manufacturing process, why do you try to make excuses to support it?

Datum selection has nothing to do with "drawing appearance", it has a lot to do with dimensioning and tolerancing for function of assemblies and parts.
Motivating the employees to learn could be providing the training. Usually people do not resist to increases of their qualification. Raises according to qualifications don't have to mean "pay them double" and can happen.

Maybe stress and deflection are the only two considerations you ever make for your tolerance analysis. There are many other factors out there for different applications. One example from the top of my head - positional tolerances on lens seat features to locate and orient lenses within a housing in optical equipment.

Sheet metal is even more unpredictable than rigid parts when the geometry is not controlled properly. The tolerancing principles that apply to any type of components apply to them too. Part of it is that if you don't need precision, apply loose tolerances, not ambiguous ones. I don't know what kind of thing you have with Trumpf, but my suggestion is - don't underestimate the potential of any manufacturer to supply you a crappy product. And if you don't have a robust documentation of your requirements, you will have no basis to make any claims.

ctopher,
I find it sad that companies choose to cut costs at the expense of making their employees better professionals that could increase the company's ability to deal with competition and increase profits in the long run. That's the opposite of a good business strategy.

 

[ctopher]

I agree. New management strategies these days are strictly to make profit, without any knowledge of what it takes to get there. They just want to see quarterly income so they look good. Then they move on to another company after a couple years. I have seen this too often.

Chris, CSWP
SolidWorks '20
ctophers home
SolidWorks Legion

 

[3DDave]

B - you don't know either and could have stopped right there.

I made no excuses supporting specific datum selection, in fact suggested that eliminating them was sometimes the right thing to do, however people pick edges without reasons all the time. They do it so it looks like the Y14.5 picture book examples.

OK - so you know I have designed electro optical elements, but sure, there are plenty of special cases - but that is far off topic and is another of your distractions. Optics are evaluated as a stack and typically don't requires use of Y14.5 to manage it.

"Usually people do not resist to increases of their qualification" Not in my experience. Most want a book to tell them what to do and point to so they avoid responsibility for thinking problems through. They love nothing better than to quote the contents as a defense. They match a picture in the book to a part of their design and copy and paste.

Increased qualification means - same pay, more work, not hiring help and resentful co-workers. Management sees it as a reason to be more critical as any error is a sign they knew better and aren't trying so hard.

Why are you so uncomfortable with this? Don't let your self-worth be tied to a book no matter how dedicated you feel to memorizing and regurgitating it. I'm not the one responsible for the poor rate of adoption and the major misunderstandings about it - that's the ASME's fault. Ask yourself why there are companies who don't use it. It cannot be because they don't want to as you demand that everyone would as an increase in their qualification.

You are just making people who want to do better feel bad as they are trapped with bad management. But if you can supply them with a clear, industry wide, thousand interview cost-cutting report with a few hundred Chinese companies, that would help them out.

 

[3DDave]

Ctopher,

"they want to pay at low end."

And that's it. Wanting to use it but not spending on training tells me they won't spend any on any other part of the process either.

It's not rocket science to learn the drawing decoration that is any more that slapping a +/- default in the title block is. It's not even as tough as high school geometry, but ASME piled on gobbledy-gook definitions and chewed up the layout time after time.

I have a copy of the 1963 "A Treatise on Geometric & Positional Dimensioning and Tolerancing" by Lowell Foster. Almost every concept is presented on one page and the diagram is on the facing page of a fold-flat book to make reading and understanding very simple. They even have diagrams of the sort of gauges that might be used. No multi-column text wrapping either - no splitting in mid-sentence to wrap back to the top of the page and limited hyphen breaks of words. It's for someone to learn the subject and reasonably effective despite being done by an amateur.


What is difficult isn't the decoration, but determining how to allocate available acceptable variations across an assembly in a way that fits well within production precision capacity that meets the cost targets. Sometimes that means choosing a really small tolerance so that expensive manual alignment can be avoided. Sometimes it is realizing that a feature can be shifted from one part to another and therefore eliminate tolerance concerns entirely.

The number of people interested in determining all the factors and the availability of accurate information about those factors is limited. It takes a tremendous amount of data gathering that no one wants to pay for and that suppliers will want to keep proprietary so you cannot go to a competitor and say "can you do this better?"

HR sees it as a check-box item because management does. They don't ask "How do you manage assemblies with differential thermal expansion during transient temperature swings?", which is an area that can genuinely cause unexpected and very expensive problems. They are more looking for "Can you pick a clearance hole size and tolerance from Machinery's Handbook if I tell what size bolt to use?"

 

[Burunduk]

"you don't know either and could have stopped right there."

The example was brought up in the context of a problem. I tried to tell you that there is no reason to deny the problem by giving alternate explanations based on guesses, especially since whatever explanation you might offer leads back to the same problem:

"people pick edges without reasons all the time"

This is the problem. Don't expect those who do this to describe a functional requirement properly in a drawing, when it is important that they do so.

"but sure, there are plenty of special cases - but that is far off topic and is another of your distractions. Optics are evaluated as a stack and typically don't requires use of Y14.5 to manage it."

Geometric tolerance controls driven by considerations other than stress or deflection are not uncommon and are not exceptions or special cases. If you didn't use Y14.5 in your optical designs, it doesn't mean that it is not useful for this purpose. Or if lens seat design wasn't good enough example for you or you believe optics don't require standardized tolerancing, a different example could be when it is required to minimize pressure drops at the flow of liquid from one component to another. In addition to sealing and surface roughness, the design would require appropriate control of the size of the holes through which the liquid passes in the two components, and the alignment of these holes. The alignment would be achieved through a positional tolerance that uses mating surfaces as datum references, with the datum precedence order that considers how the components are connected to each other in the application. This can't be achieved without standardized tolerancing.

Even if from some reason you are going to specify everything in a note, you still need to be familiar with the concepts. The concepts are not likely to be recreated by an individual making up a note. The concepts are described either in Y14.5, ISO GPS, any national standards based on them (JIS B series, BS 8888), or you could refer to Machinery's Handbook which covers the basics while addressing both systems.
And it has nothing to do with being "tied to a book".
This is why "not using GD&T" makes no sense. I's half-work. Would you recommend a hospital to instruct its doctors to ignore some portions of documented medical knowledge to cut treatment costs? Before you shout that it is another distraction, it is not, it is an analogy.

In the end, it comes down to whether we need unequivocal descriptions of geometrical limits that address functional considerations, or not?
It has to do with another point you keep talking around - do you wish to use the most expensive suppliers or manufacturing equipment that will make all features on all parts very close to nominal?
Do you want to trust them blindly without having a proper documentation of the geometric design requirements?

The common "non-GD&T" method of dealing with problems that occur due to interpretation ambiguity of such drawings, is tightening the +/- tolerance. Providing very little allowance for manufacturing. This requires expensive equipment and processes. This is the opposite of cutting costs.

 

[3DDave]

B. Of course you get notified on this - cannot let even one second go by. So here's another notification for you.

 

[Burunduk]

3DDave,
Nice responses.
You have nothing to answer, so you reverted to misbehaving. Well done.

 

[3DDave]

Seems to me like first you say that engineers all just want to learn. And then that management needs to motivate them to learn. And then that suppliers should all just cooperate with your vision of how the one thing you think you are good at should work.

And then you don't do any of the work to fix any of that and instead are falling back to "Do what I want you to do because I want you to do as I say."

Still waiting for proof of anything you have claimed.

While you are learning to write examine "straw man argument." You are using it all the time and should know it has a name.

 

[Burunduk]

Engineers like to learn and improve, just like any other group of professionals. But they are not likely to be very effective at it when the environment they work in does not encourage or support it. It goes without saying that management's strategy and actions have a lot to do with what that environment ends up being. It should be embarrassing for someone who claims to have long experience in the industry not to see that.

Your tales of engineers who are only concerned about pointing to a picture in a book are either rubbish or this is the kind of response you provoke with your attitude, not surprisingly. I've been a checker and approver for years, helped young and senior engineers improve at dimensioning and tolerancing, while getting across the message that functional datum selection is part of the design process and not just a drafting thing intended to make the drawing look a certain way. Not even once have I encountered someone who was only pointing at a picture in a book and not willing to listen. If you found yourself in such situations - it must be your fault.

I have little left to prove, as you've done a lot to prove my points yourself, while trying to do the opposite. First you brought an example of a bunch of engineers who did a poor job at design requirements documentation, the problems of whom you allegedly solved by fixing their drawings. When I asked you some leading questions about how their component drawings were made, you refused to give any answers. Most probably because you didn't want to admit that those engineers were lacking the kind of knowledge that is usually acquired by training and continuous practical application of tolerancing based standards such as ASME Y14.5 or ISO GPS. Then you pointed to another user in this forum, who told that his company is against training due to cost considerations, and doesn't even provide them access to the standard despite officially requiring to work by Y14.5-2018. While you think that this is a smart management policy, and that engineers don't want to learn, that same user called his company's policy "sad", made the effort to get access to a copy of the standard anyway, and invested his own time in learning from it and asking questions here.

You also failed to clarify how notes can be effective as an alternative to GD&T, or tell about an example of a product redefinition you were involved in, that cut costs and provided better control on quality when all datum references and geometric tolerances were removed from it and replaced by +/-, and not vice-versa. It makes no sense that you require proofs to what was already proven while providing no proofs of your own to your claims.

By the way, some of your writing is completely unintelligible and lacks any context or logic. So take your own advice about learning to write. Some reading comprehension lessons will definitely help you as well.

 

[3DDave]

If management is the problem why aren't you addressing their concerns? Cost is the main one. All you have to do to make everyone happy is show all management everywhere a detailed quantitative analysis of how much savings there will be.

Break down the additional cost for training, the additional cost for drawing review, the additional cost that those companies who spent a lot of their own money on training and skills development will charge for access to that ability, the additional cost for inspecting the additional requirements that you want to put on the drawing and then show, absolutely, a huge savings because of all the extra spending they will do.

Engineering is quantitative, not qualitative. Show the monetary savings in all that spending.

----

"doesn't even provide them access to the standard despite" It's book that is not only just $250, which any interested engineer can afford, there are other sources for it far cheaper than that. If that's the obstacle they aren't trying very much. With a 4 year engineering degree they will have spent far more money on books and should have developed the discipline to self-study, but most won't do that when there is no quantifiable benefit to do so.

Again - up to you to demonstrate that clear, quantifiable benefit.

-----

"You also failed to clarify how notes can be effective as an alternative to GD&T, or tell about an example of a product redefinition you were involved in, that cut costs and provided better control on quality when all datum references and geometric tolerances were removed from it and replaced by +/-, and not vice-versa"

And there's another straw man.

First, is the standard based on text to back up the symbols? Without that text the symbols aren't meaningful. That's why you insist on copying and pasting it. If you don't believe that notes work, stop copying and pasting notes. Every time you have copied from the standard and pasted it - you have done exactly what I said works.

Second I did not say the control would be better; that's a straw man argument. I said that not using Y14.5 controls can still result in a product that is suitable.

You are insisting it costs less to have more requirements, a quantifiable argument you have failed to quantify.

------

I don't need proof. The management of other companies do. Proof you cannot possibly provide. So they look at their costs very carefully and see no reason to proceed as you suggest. They see what I see.

Perhaps you think you are smarter than every manager in every company? Start your own company dedicated solely to Y14.5 and show them by making products that are far less costly to consumers.

 

[Burunduk]

The suggestion that converting +/- drawings to GD&T adds requirements for inspection to the drawing is ridiculous. You probably never converted specifications from one type to the other, but you could realize that anyway, if you can count.
One position tolerance applied to a hole replaces 2 coordinate dimension tolerance requirements and the general tolerance block "control" of right-angle orientation.
One profile tolerance "Between" A and B can replace 4 or 5 directly toleranced dimension requirements, or much more. A loose general profile tolerance such as given in a minimally dimensioned drawing provided in conjunction with a basic CAD model may replace tens of requirements, many of which are usually unnecessarily tight.

Luckily, the management in my company acknowledges the financial benefits of applying geometric tolerance controls and I no longer have to prove it. I am requested more and more often by management and by the manufacturing unit to replace old drawings that used only a small amount of geometric tolerance controls and a multitude of toleranced dimensions, by data sets made in accordance with the standards. The old drawings, which don't exclude "GD&T", but often have it applied wrong, are typical of designers who didn't get training at their time or were not taught the right tolerancing principles on the job. It's the same everywhere.

Training costs should be the last to cut. Investing in knowledge is what allows companies to survive or succeed in a competitive reality. Training or mentoring doesn't have to be external if the right people are hired.

Since you provided no other evidence to support your claims, show an example of a company that moved to applying tolerancing by the principles covered in ASME Y14.5 or ISO GPS, and then returned to having all drawings made the old way.

About notes, you repeat the same argument over and over while ignoring the explanations you were provided and the questions you were asked;

1. What makes you think that someone who never learned the concepts (because his company is "cost-effective" and doesn't want any training or standards) would be able to recreate the concepts he never learned, or something similarly effective - in a drawing note?

2. Why would an adequate supplier prefer to work by a drawing cluttered by someone's notes over a concise definition based on symbols with agreed upon meaning? It's clear that a supplier may only prefer the drawing with the multitude of notes if he knows he is going to ignore them, mistaking them for measurement instructions for specific metrological technology he doesn't have to apply (and it is very probable that this is what such notes may end up looking like).

More points you should think about:

3. Have you ever thought why they place traffic signs on roads and don't require drivers to read literal driving instructions such as "give way to oncoming vehicles"?

4. Why would an inspector using a CMM want to read someone's attempt of a description in a note about how a hole is to be located, and figure out how to program that into the measurement, when the drawing can clearly show the datum features and the Position symbol that he can click on in the software's user interface?

5. Why would someone who does manual inspection, and was taught the purpose of a sine bar and how to use it, want to read a designer's attempt to describe an angular tolerance zone rather than seeing the angularity tolerance symbol and the applicable datum features on the drawing?

"not using Y14.5 controls can still result in a product that is suitable"

Maybe, if it's a part that has no relationship with any other component and not used in any mechanism.

 

[3DDave]

"You probably never converted specifications from one type to the other, but you could realize that anyway, if you can count."

You are wrong. Did a full gun laying mount drawing package review for B-52s on that very topic. I got the mug and everything. Unlike your efforts, however, it required a full system investigation for the need, not just plugging numbers into a calculator.

"Luckily, the management in my company acknowledges"

That's not the topic. You need to prove to all the other people the cost advantages. Which is the only thing you refuse to do.

"you provided no other evidence to support your claims" You say that about yourself the best. No proof of cost/benefit.

Plenty of companies can waste money on things and in doing so feel like their success must depend on that waste. I find that those who feel their self-worth depends on such things get very defensive about such things. You have been very defensive. Let the ASME do the work.

1-5 are more straw man arguments and are off topic.

 

[Burunduk]

You don't know what type of investigations my efforts required, so your finger pointing is another of your defensive distractions. If you reviewed the type of drawings I talked about, you should have known that redrawing to replace ambiguous direct linear and angular tolerances for controlling form, location, and orientation of features by geometric controls does not increase the overall number of requirements - it reduces it significantly, as I detailed. You claimed the contrary, so you should have explained how "more requirements" happens. But you don't do that, because that's another claim that you make but can't support.

If you didn't experience the reduction of the number of requirements, the increase in allowances for manufacturing while decreasing uncertainties and risk by setting robust tolerance zones for features, then the B-52 drawings review that you did was not "on that very topic". It is more likely off-topic. As in the previous case, you tell a tale without making a point. You just say you did a B-52 drawings review "on that very topic" without saying anything about what the outcome was. Did your review increase the manufacturing and inspection costs? If the contrary, you add another proof to my points.

Sure, every question you can't answer is a "straw man argument". Don't let me interrupt you while you advocate bad policies and amateurism.

 

[3DDave]

No ROI? Then you have no answers. You are willing to spend money with no quantifiable benefit. That's not a business decision basis.

You have had no proof to add to.

 

[Burunduk]

My opinion is based on my experience. I told you what cuts costs when products are redefined per Y14.5, according to that experience. You didn't tell anything thought-provoking or provided any proofs or indication to support the contrary, not even by the stories you told. Your stories and the details you conceal about them indicate that I am correct.

I'm far not the first ever to indicate that:
1. Fewer requirements mean shorter inspection.
2. Increasing tolerance values by eliminating stacks and tight +/- means lower-priced process.
3. ROI? Think one-off investments versus continuous waste.

You on the other hand, define your own opinion as "nuanced". Looks more like a biased one. If you go against common logic, you are the one who needs to provide proofs - at the very least, before you require them from others. To finally provide a proof, show the figures that support sticking with mostly +/- and some poorly thought through geometric tolerancing, rather than training employees and adopting standards.

 

[3DDave]

" I told you what cuts costs "

What you don't tell anyone is how much that knowledge costs. And you did not tell by how much it cuts costs. And that is why you will always be wrong in this effort and why businesses won't do what you think they should do.

The most hilarious thing is - by arguing with me you are telling every business leader that you have no support for your position.

 

[Burunduk]

As I explained in my last response,
The burden of proof is on you. How much $ is saved by keeping the employees unqualified?

Why didn't you like that little plus sign in the tolerance compartment in the other thread? Looks nice and simple. Costs nothing to inspect the requirement. And they saved a fortune by not training the engineer who eventually did this.

"by arguing with me you are telling every business leader that you have no support for your position"

The same applies to you.