Shaft Machining - How do I convince Vendors that the Quality is No Good? Standards?

markz528 · May 4, 2019

Did a qualification inspection on a shaft for a rotating piece of equipment. Material is 4140 and has a shaft extension (where coupling will be mounted) diameter tolerance of 6.499 to 6.500 inches. Surface finish limit is 63 microinches RA.

Shaft extension measurements and surface finish measure in spec. But............

Around 2 inches from the end of the shaft, there is a step in the shaft of around 0.0004 inches. The end of the shaft is larger than the rest of the shaft. The surface finish at the end of the shaft is around 20 RA and elsewhere is around 40 RA. Just terrible quality overall - not sure how they messed it up.

The vendor argued that the surface finish is good and measurement is good so the shaft is good. My answer to that is no - you will not have a step in the machining under an interference fit coupling.

My questions are the following:

- Machining defects/steps are not quantifiable with a profilometer to the best of my knowledge. How do I qualify what machining defects/steps are acceptable? Is there a standard that I can use? I have had cases in the past where the machining was terrible but yet the profilometer surface finish readings were fine because the machining marks were too coarse for the profilometer to pick up. In my opinion, proper machining does not allow for steps but I don't know how to quantify/qualify them.

- How would you have responded to the vendor in my situation?

BTW - they say they have corrected the shaft and that I will be happy with it during the re-inspection next week.

HuFlungDung · May 4, 2019

Sounds as though the cutting edge blew up, and then the resulting built up edge on the tool resulted in a pile of crap on the end of the tool that rubbed its way undersize.

You shouldn't have to spec 'no steps', IMO. What is this, kindergarten? For that sounds like a fairly beefy and expensive piece, so for that you'd maybe have to spec the finish with "done by an old guy who has already done a few miles of turning without steps"

markz528 · May 4, 2019

HuFlungDung said:
You shouldn't have to spec 'no steps', IMO. What is this, kindergarten?

AMEN! Thank you!

But reality is proving me otherwise............

Joe Miranda · May 4, 2019

"In spec" is "in spec". Too bad that there is a .0004 step. Is it in spec? The surface finish is in spec. If they supplied you a shaft that can in any way be interpreted as being within tolerance - it's YOUR bad - not there's. If it is not what you wanted or needed then change the drawing!

Conrad Hoffman · May 4, 2019

Me thinks you have insufficient callouts on your print. As far as profilometers, there are all different types. I can measure a .5 mm step, but just not on a large part. There are standards for how far you scan to pick up various features at different surface finishes. Plus a bunch of other parameters. If you just call out the Ra, you get the defaults called out in the standards. Of course you don't have to. A surface finish callout can be quite complex, right down to the filters used for the measurement. Most people won't (and shouldn't) bother with that, because the shop won't have a clue. The defaults plus the right GD&T should do the trick.

gbent · May 4, 2019

If the small diameter and the large diameter are both within your 6.499/6.500 diameter your diameter is within specification. A .0004 step in a 6.5" diameter part and both sides of the step are within specification? You, my friend, are trying to make someone else pick fly shit out of pepper.

Milland · May 4, 2019

If there's a step in a surface meant for clamping a shaft coupling to, then I'd be concerned about it too. If nothing else, it's bad craftsmanship to leave it, better to clean the surface to minimum if possible.

I hope for the OP's sake that what happened, per his last sentence.

Rob F. · May 4, 2019

Can you put a parallel call out on the area in question? Something like Dia to be parallel within .00001. Even .0001 parallel would be out of spec from what they delivered.

Delw · May 4, 2019

maybe you should have them ground next time.

used to see this all the time with some customers. they want to turn a part due to the expense of grinding cost.
me knowing what the part is for I would tell my customer that should be ground not turned. there answer is just make it close it will be fine. my answer ok but I aint going to guarantee it will not fail. I write it right on my invoice.

we see it alot on bearing journals for repair shafts. everyone wants it fast and cheap but there are some things that need to be ground. and dont even get me started on oh dont worry about heat treating just used stainless...

Tyrone Shoelaces · May 5, 2019

If the " step " is still within whatever tolerances and specs on your drawing In my opinion it's your problem. Obviously machining the shaft with a step on it is pretty poor practice by the shop you sub contacted the work to and I wouldn't be too happy shrink fitting a component onto it. I've worked on similar equipment in the past and the shafts were always finish ground if couplings etc were to be shrunk onto the shaft ends.

Sounds like a grind down, chrome plate and re grind to size in my opinion. See if you can get the guys who made the shaft to pay for this procedure.

Maybe they've already done this.

Regards Tyrone.

gustafson · May 5, 2019

would not the step fail surface finish? if you had a 4 tenths step every 50 thou it would. 4

CarbideBob · May 5, 2019

A .0004 step or jump on a diameter is .0002 per side.
A 63 Ra would allow a whole bunch of these steps on your part and you'd probably not be so concerned if there where many and no taper end to end.
I do wonder if the original is indeed functional for mounting a shaft coupling since they are compliant and will "mush" the step.
If the shaft is 4 Ra and tapered .001 end to end is that okay? Will that lock up for the torque required?
In the computer this connection is a line contact with no ability to transmit force. In the real world different.
Are you an engineer understanding this or a inspector who thinks it not nice and pretty?
You may be beating your vendor for no good reason.
Bob

Guest · May 5, 2019

I agree with the picking fly crap out of pepper comment. The parts sounds in print to me. Does the step effect fit, function or durability? You sound like a QC guy that will do whatever he can to reject a part. If 20-40 RA with a .0004 step is considered poor on a 6.5 inch diameter part I guess you better pay to have them ground.

Rob F. · May 5, 2019

OP did not really specify how the step is being measured. I was interpreting what he said as a .0004 step, not a .0004 dia increase, which would be a .0002 step. I was thinking the .0004 step gives a .0008 dia increase, still within 6.499 to 6.500 inches.
If he has the print changed for the future to have a parallel call out, which "should" be no problem on a lathe or grinder, this issue wont happen in the future.

CarbideBob · May 5, 2019

Rob F. said:
OP did not really specify how the step is being measured. I was interpreting what he said as a .0004 step,....

Yes I could be wrong about it but a .0004 step on a short scan would trip the Ra limit making the part out of print.
Bob

adama · May 5, 2019

Honestly, best thing is do what ever you can then shit can that supplier. Trying to make drawings moron proof is the truly futile bit.

Nothing beats picking up a phone call in 3 months time when they come looking for more work, then politely explaining that sorry but you can't use them do to the shit they provided you previously. Realy takes the wind out of there sales.

If the surface roughness fails over that step, you got your part out of print, fail it at qc and its there problem! If its in roughness wise over that step, you lose they win. Simples as that IMHO. But yeah, if your having to argue over this kinda stuff, the only long term fix is find a different supplier that simply gets it!

Mike1974 · May 5, 2019

Dualkit said:
I agree with the picking fly crap out of pepper comment. The parts sounds in print to me. Does the step effect fit, function or durability? You sound like a QC guy that will do whatever he can to reject a part. If 20-40 RA with a .0004 step is considered poor on a 6.5 inch diameter part I guess you better pay to have them ground.

Have I told the story about the QC guy checking +/-.005 with a tenths indicator?

I know, everyone kind of agreed that -.0051" is out of spec, but it's the point I guess.... BUT a quick glance at ass'y would have shown that to be BS callout... I know I know print is print, blah blah blah :rolleyes5:

Clive603 · May 5, 2019

adama said:
if your having to argue over this kinda stuff, the only long term fix is find a different supplier that simply gets it!

Indeed so. Arguing that the step falls within tolerance is Monday morning quarterback or barrack room lawyer territory. Whether the shaft is turned, ground or nibbled out by beavers a correctly implemented process will not leave visible or feelable steps. A step says something is wrong in the process. Whether systematic or individual. There is no need to call out anything. If you really want to I guess the old "Free from steps, gouges and visible defects" note or one of the countless individual variants should cover everyones ass.

But shit happens. On a big, expensive, part the vendor should be picking up the phone to say that "something" went wrong and that, although the part is technically within specifications, it might not quite be what you were expecting.

Ultimately its impossible to specify everything and cramming too much onto the drawing increases the likelihood that something will be missed. It also increases the likelihood of producing nonsense by mixing incompatible call outs. The more you call out the harder it is to check properly. Rolling with the natural properties of the process is usually both easier and more reliable. Thats why finish grinding became the norm despite lathes being capable of matching all but the most stringent surface requirements. Grinding being reliable and predictable. Turning less so.

Clive

eKretz · May 6, 2019

Is this a pilot diameter that aligns the coupling or what? The description is a little confusing. I would say that if the diameter and the surface finish are both within spec, it's fine. It is poor craftsmanship but it won't affect the function of the part. A shop with good skilled craftsmen would never have let it get out the door that way, but the part will still work just fine that way. A 4 tenths step at a 6½" diameter is inconsequential on either a pilot fit or an interference fit. A running fit might be a different story. My guess is that they missed the size and tried to polish the diameter - which would explain the larger part having a finer finish.

metlcutr55 · May 6, 2019

320 grit cloth. gently with oil.

Shaft Machining - How do I convince Vendors that the Quality is No Good? Standards?

Hot Rolled

Diamond

Hot Rolled

Titanium

Diamond

Diamond

Diamond

Diamond

Stainless

Diamond

Diamond

Diamond

Guest

Guest

Diamond

Diamond

Diamond

Diamond

Titanium

Diamond; Mod Squad

Cast Iron

Similar threads