+/- ,13mm (.005") on length tolerance not possible

[duk748]

hello - I have a small question - an outside company has designed a shaft for us which is approximately 3 meters in length - it is approx. 100mm in dia. - the length tolerance specified on the drawing is +/-.13 mm (.005") - someone in our group said that this tolerance is impossible to hold on that length that .4mm (.015") was the best that was possible - I was wondering if that statement is true - I would think that w/ laser measurement devices & cnc machining it would be possible in this day in age - the shaft length does need to be held & I would assume that w/ some redesign it could be changed to fit this tolerance - any info would be greatly appreciated just for a better understanding for myself - thank you - i am re-posting this from the drafting board on the advice of the people there - material is 1045 steel w/ 125 finish at each end - thank you again

 

[JNieman]

Of course it is possible - it's just a matter of whether or not you want to pay for the processes that make it so.

There doesn't have to be any CNC or lasers involved, to be honest. Just a calibrated measuring tool that can reach the 3 meters. That might be a CMM, it might be a very large caliper or micrometer.

I can't really say what a good process would be to manufacture the part without knowing the quantities and the machines available. It wouldn't be a problem for some lathes. Wouldn't be a problem for some mills. There are many ways to skin that cat.

 

[tygerdawg]

It may be necessary to specify the temperature conditions under which the measurements are taken. On 3m length, thermal expansion may be an issue.

TygerDawg
Blue Technik LLC
Virtuoso Robotics Engineering

http://www.bluetechnik.com

 

[KENAT]

See thread1103-406749 in GD&T forum for my response - along the same lines as JNieman.

tygerdawg FYI on the issue of specifying temperature, if drawings are to ASME Y14.5M-1994 it's specified at 1.4(k) that UOS all dims apply at 20°C. I believe ISO has similar but couldn't tell you what spec it's in.

However, the implication that you need to worry about being at this temperature during inspection and potentially machining is very valid one I made in my original response.

Posting guidelines faq731-376

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

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

 

[JNieman]

I agree with the temperature requirement. General shot from the hip tells me you'll grow .005" in length with an increase of just 6 degrees fahrenheit.

When I'm confronted with such situations, I typically will place a note reaffirming the boilerplate requirement. "Part to be verified/inspected to dimensions and tolerances shown at 68F/20C as per <insert standard document number>" just to say "Hey, I really mean it this time" because let's face it - it counts all the time, but who knows how many suppliers actually realize it, and how many control their quality room that well. Maybe throw in the fact that they need to let it normalize, and not just measure it 2 minutes after bringing it in from their 90F or 45F shop floor.

 

[CheckerHater]

General shot from the hip tells me you'll grow .005" in length with an increase of just 6 degrees Fahrenheit.

That means, not only shaft must be made / measured in controlled temperature environment, it also has to work in controlled temperature environment.
One can imagine that temperature of working machine may change for tens if not hundreds of degrees.
From that point the following OP statement looks interesting:

the shaft length does need to be held & I would assume that w/ some redesign it could be changed to fit this tolerance

Could "redesign" include compensation for thermal extension and/or manufacturing accuracy?
Way too often tight tolerances are the result of over-engineering, so I would be curious to have at least a hint on the shaft working conditions and how the tolerance was calculated.

"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[JNieman]

@CheckerHater, that first point gets mostly into the realm of design requirements which is hopefully already considered. I used a rule-of-thumb coefficient for AISI 4140 alloy steel to do my quick math, but I have no idea what's being used. If the components are of comparable thermal properties, and the mating pieces shrink/grow comparably, things work. I did not assume that the bar was being 'trapped' on both ends where thermal expansion may induce binding or undue stresses. That's a bit outside the scope of this discussion.

But yes, I agree that redesign should always be considered when faced with extraordinary manufacturing requirements. "Value engineering" right

 

[duk748]

hello again & thank you for the info - yes thermal expansion was taken into consideration as the working environment will be only ambient air temp w/ the mating pieces of simular properties - yes i too believe that some over-engineering was a part of the design but i also believe that this was to compensate for a very low end manufacturing facility off shore - some parts manufactured & evaluated by this same outfit have caused considerable time to be spent tailoring parts & designs to their manufacturing process not due to what is normally accepted practice or what can be accomplished - i wanted to know if & what was possible in a modern manufacturing facility - thank you again for your help

 

[JNieman]

Your assumed context of a "modern manufacturing facility" does not say anything, unfortunately. The amount of variance from one shop to another is vast as the ocean.

A "CNC" machine shop, for example, could vary from a tin building lacking climate control, with a bunch of clapped out Bridgeport clone knee mills with retrofitted programmable controls to a modern building with calibrated climate control and brand new top of the line well-maintained machining centers.

If you are dealing directly with a supplier and they tell you what they can and cannot do, they are probably speaking based upon the capabilities of their machinery and personnel. Without knowing what they have, it's hard to say what they can or cannot do. A large portion of their communication will depend on the personality/experience of the correspondent and how much they will push you back on troublesome requirements. It may be truth, laziness, distrust of engineering, or someone merely passing on what a shop floor employee said is 'impossible' which, often times, isn't.

 

[CheckerHater]

@JNieman: I get your point. This discussion originated on GD&T forum and is still about the tolerance, hence the natural curiosity for "functional requirements"
Tight tolerance usually implies "fit" condition, otherwise why bother? So the idea of 3 meter long shaft only 100mm in diameter being "fit" (or "trapped") from the ends rises questions.

"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[TheTick]

I think your drafter is trying to tell you he needs to do something different with his life.

 

[3DDave]

It's only money.

For reference, .13mm/3000mm is 43 ppm; steel is 30psi/ppm, so this is about 1400 psi worth of stress. Over the 10mm diameter this is 15600 lbf. As mentioned steel thermal expansion coefficent is around 6.4 ppm/F, so only 6-7 degrees F is required to move from nominal to out-of-tolerance; 12F will move it completely out.

What's interesting is that this is easily the temp range from floor to ceiling in residential homes. Commercial use is less well regulated unless it is specifically temperature controlled.

 

[racookpe1978]

However they designed it, they came up with a length requirement and a tolerance.

The temperature of that big a rod WILL NOT change quickly: day to night, machining heat or static cooldown or sunlit and shadows or truck to building to measurement lab. But! it WILL change. The temperature must be stablized BEFORE measurement starts each time. The temperature of the device (caliper or slide or fixture or laser) must also be static and be within its specification. For example, the shaft is at 120 deg F, the length is "adjusted" or corrected for 120 deg growth from 70 nominal degrees, but the caliper itself heats up from measurement 1 to measurement 2 to measurement 3.

 

[TheTick]

Machining tolerances aside, there is something very wrong with a design that requires such tight tolerance on such a large shaft length.

 

[Tmoose]

Any way to post an assembly drawing to see the function?