UF and collection planes in ISO GPS

[greenimi]

ISO GPS system:
Can UF and collection plane be applied together?
(UF above the tolerance indicator and collection plane as a supplemental indicator)

I don't have access to all ISO GPS standards, but in the ones I have I cannot find an applicable example where UF and Collection plane are used together.

Is it feasible? Is it legal? Any good examples?

Thank you so much and have a great day/ weekend

 

[pmarc]

greenimi,
Besides CZ and SIM, there is also CZR (combined zone rotation). All three group individual tolerance zones into a pattern of tolerance zones.

There is also UF which groups multiple features into a single feature.

And there is CT (common tolerance) which may be used in conjuction with a directly toleranced dimension to specify that a collection of features of size "shall be considered as one feature of size.

Regarding your second question, the standards say that these modifiers may be specified but don't have to be when all non-redundant DOFs are constrained. So I think it's a matter of preference and there are pros and cons of both options. I would definitely not recommend mixing the usage of both options within one drawing, though.

 

[Burunduk]

If the front surface is not parallel to F then it is also collected?

This question is a bit odd because the front surface is datum feature F unless you meant something else by "front".
Ok, the actual as-produced datum feature F will never be purely parallel to datum F in the strict theoretical sense (since it won't be perfectly flat and thus will not be fully coincident with the datum and thus will not be parallel to it). But that doesn't matter because the identification is based on the nominal relationship regardless of any specific physical outcome.

 

[3DDave]

Sorry - If the BACK surface is not parallel to F then it is also collected?

 

[Burunduk]

If it's nominally non-parallel - yes.

 

[3DDave]

Then if there was a hole in that face it would also be contiguous and the tolerance would apply to that as well.

I'll wait for an expert to chime in.

 

[Burunduk]

While waiting for an expert to chime in, let's try to interpret based on what's written in the srandard.

The definition is:

"3.6
collection plane
plane, established from a feature on the workpiece, defining a closed compound continuous feature.

Note 1 to entry: The collection plane is always used when the “all around” symbol is applied."

Further, there is:

"16.1 Role of collection planes
Collection planes shall be used when the “all around” symbol is applied. The collection plane identifies a family of parallel planes, which identifies the features covered by the “all around” indication."

And:

"16.4 Rules
A collection plane shall be indicated when the “all around” symbol is used to identify that a specification applies to a collection of features. A collection plane identifies a set of single features whose intersection with any plane parallel to the collection plane is a line or a point."

So, does that "collection" include the inclined back face and the hole from your scenario? What is your opinion based on the above texts?

 

[3DDave]

I'll wait for an expert to chime in.

 

[Burunduk]

That's part of the problem.
How useful is this standard for defining, manufacturing, and inspecting mechanical parts if it requires an accompanying expert to interpret it?

 

[3DDave]

I've asked the same of Y14.5. I'll work on pounding on Y14.5 for more clarity and you can take on ISO GPS. Deal?

 

[Burunduk]

There's no comparison at all. Y14.5 can be improved, but at least it doesn't have these ridiculous callouts:

 

[3DDave]

It's easily readable and more explicit than the implied methods in Y14.5, aside from the apparently weird "collection plane" behavior that doesn't appear to be problematic in this example.

 

[Burunduk]

I want to hear that from a machinist, saying that those callouts make it easier to understand the requirement.

 

[3DDave]

Such things normally go to QA first to review the drawing and then to the manufacturing engineer who specifies the machining operations, then some back and forth with the CNC programmer to determine any tooling and fixturing required and which machines are process capable. I'm not sure the machinist ever sees the drawing. For the most part the machinists I worked with got a process routing sheet with the details of what to do already worked out.

Even if it is a one-off shop with a guy and his Bridgeport, the initial conversations about dimensioning and tolerancing should happen before the drawing is handed off. If there are doubts or questions, they would have my phone number to call.

Your mileage may vary. If you are in a shop where the engineers refuse to deal with shop questions, that's an entirely different problem than how the information is depicted.

 

[Burunduk]

In every mass production plant I've been to, the part's drawing is hanging on the CNC machine. The operators conduct in-process inspections and need to know what they are manufacturing.

 

[3DDave]

So what? If they don't understand do they just guess? If so then that is a management problem, not an engineering documentation problem.

I cannot imagine a better way to get out-of-control production with every machinist winging it.

Also, mass production? On a piece part basis? Sure.

Process sheets tell the machinist what they are making and the steps in production they are to perform.

 

[Burunduk]

"So what? If they don't understand, do they just guess? If so, that's a management problem, not an engineering documentation problem."

If it's a management problem, how would you handle it? Hire ISO GPS experts to be available for every machinist at any given time?

Suppose a machinist needs to do in-process inspections every X parts to check if the all-around profile requirement is still met, or if the new tools the company started ordering wear out too quickly.

After seeing an unnecessarily complex callout that he doesn't understand, the machinist may ask the manufacturing engineer, who will likely contact QA, which will then call the design engineer who made the drawing but doesn’t necessarily fully understand the callout he copy-pasted.

Asking questions and getting answers takes time and can result in delays and additional costs. See? You are still waiting for an expert to chime in and answer your question about the collection plane.

"I cannot imagine a better way to get out-of-control production with every machinist winging it."

So what’s the better approach? Should machinists be ignorant of what the drawing requires?

"Also, mass production? On a piece-part basis? Sure."

I didn't say all those plants made parts only on a piece-part basis, but even in assembly-line production, there are still machines making individual parts, whether finished or as part of the process. The machinists responsible have to ensure their production meets requirements anyway. And what's wrong with mass production on a piece-part basis? Spare parts, replacement parts, etc., are often mass-produced on a piece-part basis.

"Process sheets tell the machinist what they are making and the steps in production they are to perform."

That is not always the case.

 

[greenimi]

After seeing an unnecessarily complex callout that he doesn't understand, the machinist may ask the manufacturing engineer, who will likely contact QA, which will then call the design engineer who made the drawing but doesn’t necessarily fully understand the callout he copy-pasted.

Burunduk,
I think we are working for the same corporation, don't we? Which department are you in?

 

[Burunduk]

greenimi, I bet there are several more people here that feel the same

 

[3DDave]

me: "If they don't understand do they just guess? If so then that is a management problem, not an engineering documentation problem."
B: "If it's a management problem, how would you handle it?"

Same as any other case where the machinists are making bad parts. It could be they don't understand the material callout, or the surface finish requirement, or the quantity. Whoever they report to asks why they f'd up and if they say they didn't know, then their supervisor tells them to ask and ensures that they understand. If the machinist continues to f' up because they don't ask, then they get fired.

If the supervisor isn't doing their job, plant manager fires them.

"Asking questions and getting answers takes time and can result in delays and additional costs." Yup. Like not asking what to do with the 4 bolts, time was saved, right? Boeing lost a few tens of billions of dollars in value because someone didn't ask that question. Maybe all the factory workers should all be trained to be aircraft design engineers?

"Suppose a machinist needs to do in-process inspections every X parts to check if the all-around profile requirement is still met, or if the new tools the company started ordering wear out too quickly."

That's what the built-in CMM software in the CNC machine that the CNC programmer set up for the machinist does. However, a couple of caliper or micrometer dimensional checks should tell if the process is drifting without needing to check the entire profile to ensure the tool isn't wearing too quickly, though getting dull and producing bad chips and a bad surface finish happens first. Even a temperature gun will tell if the cutting tool is wearing as the temp will go way up.

If the engineer doesn't know what he copy/pasted? Maybe the machinist should do an FEA on all the parts because maybe the engineer doesn't know how to do a strength analysis either. Maybe the machinist should also have a doctorate in materials engineering because the engineer just copy/pasted the material callout.

In any case, if the engineer doesn't understand what they are doing, then that's another management problem, hiring the uneducated.

"Piece part" isn't "making individual pieces," it is open setups by individual machinists. The way mass production works on a piece part basis is to have 10-1000 individual machinists, each with their own tools, doing what they want. Do you think mass production is having a single 5000 spindle CNC machine cutting 5000 parts simultaneously?

"Should machinists be ignorant of what the drawing requires?"

I don't think so, but there you are putting words in my mouth. Already said, if they don't know they should ask. Then they won't remain ignorant. Do machinists look up feeds and speeds or do they have those all memorized for all materials? What if the material isn't in their list? Do they remain ignorant or do they ask someone?

me: "Process sheets tell the machinist what they are making and the steps in production they are to perform."
B: "That is not always the case."

Then that is also a management problem. Workers should be given the necessary instructions to do their work. If those instructions are incomplete or following them produces bad parts the answer is to improve the instructions.

Is your factory a bunch of children on a dirt floor shed with hand files? It sounds more and more like it.

 

[Burunduk]

Taking things out of context doesn't help you present good arguments.

Asking questions when necessary is an important part of any job, but the whole purpose of a standard is to reduce the number of questions and enable more efficient communication through the drawing.

There is no added value in the new ISO GPS indicators when all they do is add complicated symbolism and strange use of datums, just to neutralize defaults that are intuitive to anyone who knows how to read a drawing.