Goal R.002 *Pulls hair out!* 5

[ModulusCT]

OK, so I was told to do the following by my supervisor today. I really hate having to tell him that he has no idea what he's doing and that he doesn't know the applicable drafting standard, so I thought that it would be better to simply offer a solution that does jive with the spec.

He wants to indicate R.002 +.003/-.000, but with an emphasis on trying to produce the radius at .002 rather than taking the total permissible error and shooting for the middle. His solution to this problem is to say .005 MAX; GOAL R.002, but I hate this personally because the word GOAL is not mentioned in the spec as an acceptable descriptor for tolerancing a part. In fact, I'm pretty sure there are parts of the spec that say not to do things like this.

Any suggestions? Is there a way that you can say, what my boss wants the drawing to say by using acceptable symbology and terminology?

 

[Walterke]

I'm just saying that a unilateral tolerance communicates something.

You're absolutely right!

It communicates that a given dimension can be bigger(or smaller) then the given dimension, but not smaller(or bigger)

NX 7.5
Teamcenter 8

 

[Belanger]

Agreed, Walter. But it seems that the prevailing opinion here is that the "given dimension" itself means nothing. (Other than a convenient way to discuss a nominal value in conversation.)


John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

http://www.gdtseminars.com

 

[Walterke]

I just reread my sentence and noticed it is quite confusing, though you seem to understand what I was trying to say.

I understand you (and other people) try to explain to us that there is a difference between a X+0/+0.5 and a X.25+/-0.25 dimension, but all throughout this (pretty long) thread, I have not seen anyone define the difference, without using words as 'common sense' or 'what the designer means'.
Technically there is no difference, so there is no reason why a machinist shouldn't strive for the X.25 dimension, regardless of the way it was noted on the drawing.

Just my opinion though, which, apart from my school days, unfortunately only has a whole 2 years of actual experience to back it up. But this is how I was thought to use tolerances.

NX 7.5
Teamcenter 8

 

[Belanger]

Here's a simple analogy to illustrate the difference. Suppose I want to redecorate my living room. My wife sends me to the paint store, telling me to get hunter green. "But if you can't find the exact shade," she says, "get a pale green or maybe even a light blue."

Now if the store has hunter green, but I come home with a can of light blue paint, that's within spec, right? So all is well, as far as the wife is concerned! But I had the capability to get closer to the "desired dimension." So the desired dimension can be thought of as a "datum" from which acceptable variation is measured. But I agree that in the end, there is no legal difference when it comes to accept/reject criteria.

Obviously, I'm not saying that manufacturers purposely try to go for the bare minimum to achieve a passable part. But my point is that there is still some meaning to the "desired dimension," and besides your reference to common sense, the closest thing in Y14.5 would be the latter part of paragraph 1.3.22.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

http://www.gdtseminars.com

 

[ajack1]

I would say that manufacturers do the bare minimum to achievable passable parts, well sort of.

I was always taught that in a production situation tools or cutting dies wear, i.e. shafts get bigger and holes get smaller over time, so you should start near the limit furthest away to allow the maximum run time whilst still staying in limit.

Also say a piece of stock material will conform to the limits you would not buy an oversize piece and machine it down just in order to get nearer to mid limit. The same applies to holes if you can get away with just centre drilling and then drilling you would not start adding extra processes like reaming, boring or honing just in order to get nearer to mid limit.

I would think most manufacturers try to produce acceptable parts in the most cost effective way, rather than aiming for mid limit however it is toleranced.

 

[ModulusCT]

I was always taught that in a production situation tools or cutting dies wear, i.e. shafts get bigger and holes get smaller over time, so you should start near the limit furthest away to allow the maximum run time whilst still staying in limit.

I wonder if it would be advantageous to design parts with this in mind? Maybe when designing parts that will be created as part of a mass production cycle.

I'm not a vegetarian because I dislike meat... I'm a vegetarian because I HATE plants!!

 

[ewh]

It depends on the design.

Say you get optimum part performance at .XXX and have an operating tolerance range of +.000/-.005.
.XXX +.001 will not work in this design.
.XXX -.005 is the other limit in the design and will work, but not at optimum.

If we are talking racing, for example, I think the machinist is going to do his best to achieve the requested dimension.

Mass production introduces the trade-offs of "median" dimensions.



Technically, the glass is always full.

 

[ptruitt]

Production does not need to be very massive before process capability of small features becomes unrealistic, especially when considering measurement uncertainty and bias. I must wonder what the process capability has been or will be achieved in the OP.

Try this: Ask ten people to do a simple tolerance stack analysis on three mating parts where non-symmetrical tolerances are specified and see how many people get the right answer. Then do a similar exercise adding a hole position with bonus tolerance. Any right answers, now? Now make all of the tolerances symmetrical and see how many wrong answers. How many migraine headaches resulted? I like to mitigate risk and using symmetrical tolerances is a significant part of that effort in my designs.


Peter Truitt
Minnesota

 

[ajack1]

Ewh I would not doubt for one minute that for things like race engines machinists would aim for the optimum size, especially high end racing, F1, Indy car and the like.

What I was trying to say is that limits will be used by machinists not only to aid production but also to dictate the process and keep the costs down.

For example if you could take a piece of boiler plate and laser or water jet a finished part that would meet all criteria it is highly unlikely that anyone would fully machine the same part and start polishing, honing etc just to get nearer to the mid limit or desired size. Or at least if they did it would be a good idea to use them now as they probably won’t be in business next month.

 

[ewh]

I am not disagreeing with any of these manufacturing or inspection points; I'm aware of manufacturing realities. I'm just trying to point out that the requested dimension has a purpose in the definition of the design. That's all. While manufacturing and quality may take the total tolerance allowance and aim for the median in fabricating and inspecting the part, the original design intent is still documented and preserved. If the design needs to be changed in the future, such toleranced dimensions should raise a flag to the revisor that there may have been a reason for such tolerancing, other than mere laziness. Why force mediocrity when [i[design[/i] excellence is within grasp?

Technically, the glass is always full.

 

[Belanger]

Peter, your example illustrates one big pitfall of always thinking in terms of symmetrical, bilateral tolerances. And that is the idea of "bonus" and "shift" tolerances when performing a stack analysis.

Bonus tolerance is usually not equal bilateral. Guess how most people handle it? They ignore it. Or they try to plug it in and just figure on splitting it down the middle around a nominal.

It is possible to make it bilateral for a stack done with the strict nominal/plus-minus format, but it gets confusing with the LMC being the nominal and the ± tolerance being the stated amount and the bonus put together. The method I prefer for stacks is simply a two-column method (one for max, one for min). It still requires some thought when it comes to bonus and shift, yet it's very straightforward and can easily accommodate RSS and other statistical factors.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

http://www.gdtseminars.com

 

[Walterke]

Why force mediocrity when design excellence is within grasp?

A tolerance on a dimension is usually a reference to the possibilities of the machine and thus, the price of the part.
If you laser out a circle out of a sheet of steel, the tolerance on the diameter will be around +/-0,2. This means that your laser is not capable of determining the dimension of your part any better then this tolerance.
you won't send a drawing to a laser shop with a +/-5mm tolerance, nor with a +0/+10mm tolerance. In stead, you'll send them with either a +/-0.2mm or a +0/+0.4mm tolerance (numbers as examples only), where one means "it has to be around diameter X" and the other means "it has to be close to diameter X, but DEFINITELY not smaller"
In the latter case, the diameter will be programmed on diameter X,2mm, to make sure most parts are within the requested tolerance zone.

Knowing the tolerance of where your product will be machined is part of the design progress. That, or, the machine used to create your part will be dependent on the applied tolerance on the drawing.

NX 7.5
Teamcenter 8

 

[ewh]

Again, I'm not questioning manufacturing/fabrication methods. I feel that is a different issue, to be aware of definitely, but it does not address the pure design aspect. You start with a problem, seek the best solution and then tolerance accordingly, ensuring that the original design intent is captured as well as that the part is manufacturable.

Technically, the glass is always full.

 

[KENAT]

Walterke, while process capability is very important to consider, per ASMEstds it's function that is the primary driver of assigned tolerances.

Posting guidelines faq731-376

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

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

 

[jctoce]

Sorry for off topic question, but where are you in CT ModulusCT? I work right in East Windsor.

 

[Walterke]

per ASMEstds it's function that is the primary driver of assigned tolerances

In that case the machine used to create your part will be dependent on the applied tolerance on the drawing, which is what I ended my previous post with.

@ewh: if I want to design an axis that slides into a hole, I'd go for a sliding fitting, which, according to ISO is describes as a H-h fitting (for example)
This would result in a axis with diameter X+0/-z and a hole diameter X-0/+z.
This results in a minimum 'passage' of 0 and a maximum passage of 2z.
The design intent here is NOT for the axis and the hole to be as close to X as possible, with a tolerance of z. Aiming for the middle is exactly what I want here.

NX 7.5
Teamcenter 8

 

[ewh]

Walterke,
This may be, and that is fine; you know what you want (though I question the minimun "passage" of 0 - there is such a thing as line-to-line interference). It depends on design intent, and that scheme obviously meets your requirements.

Design intent - in the design phase, as on a drawing, you really don't usually care how your requirements are achieved. I don't care what machine you use to create my parts, as long as the parts meet the requirements. You could whittle it out with a pocket knife for all I care.
Realistically, I do have to be aware of manufacturing capabilities and adjust my tolerances to allow for them, but again, I'm not questioning manufacturing/fabrication methods. I am addressing the definition of the "ideal" design, which will have to accommodate those issues before the drawings are released, but not when the design is being conceptualized. When I am starting a new design, manufacturing is not my first consideration. It is an important consideration that must be addressed, but that is a different issue. My first concern is how to make this design as good as possible within reason. When that is accomplished, then I can adjust the tolerancing to accommodate down stream functions, while keeping the "ideal" definition intact. I do not want to lose the definition of the "ideal" design to accommodate manufacturing/quality, when sufficient tolerances are already provided. Unilateral tolerancing is one tool to accomplish that, and a competent machinist should be able to take that data, adjust it within the parameters of his tools and continue making good parts, while being aware of the intended target as defined on the drawing.
Many of the participants in this forum seem to be most concerned with manufacturing/inspection, and that is good, as it is a reality we must deal with. I don't feel that it needs to be given such weight as to let manufacturing/quality drive the definition/documentation of the engineering design. Putting other functions before the design function is putting the cart in front of the horse.



Technically, the glass is always full.

 

[ModulusCT]

Sorry for off topic question, but where are you in CT ModulusCT? I work right in East Windsor.

Is there no way to private message someone on these forums?

Anyway, yeah, I'm very familiar with East Windsor. I work in Simsbury... Where do you work in EW? What's the name of the place?

I'm not a vegetarian because I dislike meat... I'm a vegetarian because I HATE plants!!

 

[ewh]

That should probably be directed to jctoce; I'm in TX.

Technically, the glass is always full.

 

[powerhound]

I think he meant East Windsor.

What part of TX ewh? I thought I was alone here.

Powerhound, GDTP T-0419
Engineering Technician
Inventor 2010
Mastercam X5
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II