What's new
What's new

Customer expectations on tolerance?

FredC

Diamond
Joined
Oct 29, 2010
Location
Dewees Texas
Customer sent me several prints with difficult tolerances. I emailed back and expressed reservations about holding the tolerance in the long part. This is on a 5 inch long aluminum tube with a 1.000 OD and an .8700 bore plus 0 minus .0004 on the bore.
The engineers reply "I just found I had left over McMaster-Carr 89965K681 General Purpose Aluminum Tubing from last time we made this and the ID is undersized. So I only need it Cut to Length, Reamed, and Plated." Plating will be clear irridite. The best I can tell is this will add nothing to the material.

Other close tolerance items in this package:
1/2 long bore .7495 same tolerance plus 0 minus .0004 on the bore (6061)
1 inch long bore 1.1248 plus 0 minus .0004 on the bore (6061)
1.0000 bore 1/4 long in Nylon plus or minus .001 Would have been comfortable in PEEK with this tolerance but nylon?

Thinking about asking the customer for the parts these fit and asking what kind of fit they expect. I do not think they can even measure this stuff, just put numbers on the drawing with the expectation they will be close.

Any other suggestions?
We hold closer tolerances down to plus .0001 minus 0 in small parts all day long. We use Deltronic pin gauges, super mikes laser mikes, and such in production parts. No way I can justify the gauges for 1 part of each on these things.
 
You may be dealing with a young engineer with little "real world" experience. It's up to you if you want to take the time to educate, but in this instance I think I'd no-quote.
 
or, point out to them that +/-.0002 will be 4x the price of +/-.0005

Maybe 10 times on that 5 inch long bore.

Milland,
I think the engineer is older and been with the company a long time. Drawings have at least one missing dimension and probably have not been reviewed in years.
 
Well...the first question is "do you really need those tolerances?" Never underestimate the value in having intercourse with a customer.
 
More importantly, the paperwork needs to match what you were told.
IE, if the print still has the tight tolerance, and your parts get stopped
at incoming inspection, you don't have a leg to stand on.
 
Well...the first question is "do you really need those tolerances?" Never underestimate the value in having intercourse with a customer.

I'd put it a different way. I'd say: "One of the ways I help customers succeed is to share insights from my experience. This way, your design ideas succeed more cost effectively. Everybody wins. So on this part, the plus 0/minus 0.0005 is about $3000 more than a similar part with standard tolerance. Is the spec needed? I mean, we can do it, we do it all the time, but guaranteeing those specs means I have to put in some really involved setup and metrology steps. And for a one-off, all that setup has to be charged to the one part."

The cost delta may not be of importance and you get to go ahead, but in a way that's well compensated.

BTW, once you get the tooling set up I'd probably make 6 more of the parts. I little bit of cost, but if they call back for more you'll have some on the shelf.
 
That opening line "One of the ways I help customers succeed is to share insights from my experience" is just the sort of thing that, for me, makes intercourse unappealing. When someone dollops on that much lubrication from the get-go, or even mid-sentence, I know they've been to one too many semenars. And if the person I'm performing intercourse with needs that sort of glop, I know I'm talking to the wrong person anyway. So here's how I would start: "These tolerances are wicked-ass tight. Unless you're planning to fly to the moon, it'd be a good idea to loosen them up or you'll get screwed on the pricing."
 
That opening line "One of the ways I help customers succeed is to share insights from my experience" is just the sort of thing that, for me, makes intercourse unappealing. When someone dollops on that much lubrication from the get-go, or even mid-sentence, I know they've been to one too many semenars. And if the person I'm performing intercourse with needs that sort of glop, I know I'm talking to the wrong person anyway. So here's how I would start: "These tolerances are wicked-ass tight. Unless you're planning to fly to the moon, it'd be a good idea to loosen them up or you'll get screwed on the pricing."


Yeah, that's just what I said... :D
 
I saw this all the time as a project engineer reviewing designers work. Engineers, young or old, will dig in deep if they sense criticism.

If it is a big company, there will be internal friction to changing specs after it is released for quote. Purchasing and engineering do not typically get along. Cost will not be the only factor.

If it is a small operation, they will be most interested in cost.

The best way is to ask for a call. Tell the engineer you reviewed the part and can do it but the cost will be high. Tell them why without judging their experience. Don’t bother putting a number on it. Tell them what you can do for a fair price and let them make the choice on how they want it quoted.
 
Going from a job shop to an inhouse supplier has taught me one glaring thing. Parts are usually detailed with too tight of a tolerance. Fortunately I can look at the model and then talk to a machine builder and come up with a real world tolerance.
 
Stacked tolerances are the headache for assemblers who outsource parts.

It is easier when part suppliers run with SPC and keep processes trending properly. That involves some work to qualify capable suppliers which defeats the bid it out to the cheapest source mindset.

Unfortunately, most view tolerances like goal posts and will correctly argue if they are in tolerance by the thinnest margin. The engineer is stuck with silly constraints or they own the mess when things don’t fit.

This case is a one-off so a different situation.
 
Email from the engineer says all these close tolerances are for a light press fit onto a ball bearing. He said he can supply the bearings these items fitting. Waiting for a response on that tube to see if it only has one bearing on each end, if so this has turned to easy peasy just bore each end 1/2 inch deep.

Curious thing about that .8700 dimension, seems like it ought to be .875 or 22mm to fit a standard bearing, almost seems that number is a typo.
 
1000 tons might be a light press fit for a 100 inch ID ball bearing, think it would not qualify for a light press fit on 7/8. I will ask though. Light press fit could be with 2 fingers or a small arbor press.
 
Email from the engineer says all these close tolerances are for a light press fit onto a ball bearing. He said he can supply the bearings these items fitting. Waiting for a response on that tube to see if it only has one bearing on each end, if so this has turned to easy peasy just bore each end 1/2 inch deep.

.

Sounds like a reasonable guy and the lines of communication are open. Work with him and you both gain.
 
What it for? Oh it's part of a handle:willy_nilly:

I kid the not I have seen +/-.001 on a handle that had indentations for your fingers. Medical part, customer would not open the tolerances up and paid through the nose on them.

As for the OP I would talk with the engineer as others have said.
 
Been talking today.

Many years ago this company had its own machine shop and only farmed out production work. This batch of parts is made for an instrument they are building. One each of several parts. Whenever they build one of these gadgets they knew what these drawings meant. I had no clue that some dimensions not needed were because parts from different sources were already partially finished. As an example, the take up pulley is shown with a smooth OD with no size indicated, it is actually turned from a gear blank obtained from Stock Drive. Need a program to understand these drawings.

Flanged bearings go in each end of the tube, boring 1/2 inch deep is perfectly acceptable. Too easy.

Asked about the press fit and 2 fingers on a 1/4 ton arbor press is OK.
 








 
Back
Top