Inspection of keyway in bore

[PetkovStoyan]

Hello,

Does anyone know how symmetry control of internal keyway can be performed with portable measuring device, and if there is such device at all in existence. I am not interested in the measurement of the width of the keyway, but rather its position with respect to the bore axis, so that this measurement can quantify the eccentricity between the keyway center plane and the bore center plane.

Thank you.

 

[3DDave]

How large is the bore and keyway?

 

[PetkovStoyan]

Bore is 110mm, and the keyway is 28mm wide. Lenght is 120mm. I can make and attach a sketch if this explanation is not clear enough

 

[Belanger]

It depends on whether the callout is the symmetry symbol or the position symbol (you used both terms). Also, it might depend on which tolerancing standard the drawing subscribes to (ASME or ISO).

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

 

[PetkovStoyan]

@Belanger,

either is fine, as long as the offset with respect to bore center plane can be measured, or calculated. Or any other measurement which will prove that the hub-key-shaft assembly is not possible due to wrong manufacturing of the keyway in the hub.
The keyway looks off-center, and I can tell that just with a naked eye. I am pretty sure that it will not fit on the shaft, but I don't have the possibility to assemble/disassemble the parts. So I need to prove that with measurements, since I have to reject the parts and claim non-conformance to external shop.

EDIT: If you ask me what is the callout on the drawing, there is none for the keyway. I know that this is a miss in the design, but I don't want to discuss the reasons behind this.
For now I just need to know how to measure the deviation.

 

[Burunduk]

If you notice a large offset with the naked eye, a simple optical comparator measurement can aid you in getting a rough estimate of how displaced the keyway is relative to a plane that passes through the bore's center.

Such measurement would be 2-dimensional, but as a rough evaluation of how bad it is, it may suffice.

 

[PetkovStoyan]

@Burunduk,

the part is quite heavy, and its diameter is above 500mm. This is not an option, unless some portable device is what you have in mind.
I found a paper online here:

https://www.cmmi.tuiasi.ro/wp-content/uploads/buletin/2015 fasc 4/L4 CM 4_2015.pdf,

but I could not find such device for sale, maybe it is just a design concept for a prototype. This is what I had in mind

 

[3DDave]

If the supplier relationship is such that a visibly off-center keyway cannot be rejected and you have no other requirement for how centered it should be, I think making a measurement is of little use.

For a one-off measurement - cut a metal strip to have a close fit into the keyway, a strip that extends farther than the bore; lay the end of the strip into the keyway and you can measure from each side to the maximum exposed width of the bore. Nominally that should be (110-28)/2 = 41mm. To remove errors in the strip, flip it across the width and measure again to removed effects of lack of straightness in the strip.

 

[PetkovStoyan]

@3DDave,

I agree on the first part, but still I consider having some quantitative value is a stronger argument than just saying "it looks off-center to me".

Thank you for the proposal for measurement method, this also came to my mind as an option.

By the way, did you check the link I provided in the previous comment, have you ever seen such measuring device?

 

[3DDave]

I was surprised to see a paper that covered your exact problem range. If this was an ongoing industry problem there would be a number of suppliers of tools for this job. Errors of this type should be in the 0.0001 inch range, not visible to eyeball estimation and are of a sort to find a replacement machinist than to find parts to scrap.

 

[PetkovStoyan]

@3DDave,

in my opinion the machinist used smaller size tool, and widened the keyway left and right until the desired size was achieved. Other than that, you are correct. Are you sure that you didn't add one extra zero on the error? 0,0001 inch equals to 2,5 micrometers, seems quite tight

 

[3DDave]

https://youtu.be/Pifjw29ss7A?t=1044

is cutting a similar sized keyway, 25.4 mm. Per the description:

This particular job is for a 1.000" wide (+/- .001) by 10.5" Long with a center tolerance of +/- 0.004". The shop who was doing these has been struggling and causing all kinds of delays and scrap parts. Now, I get the chance to prove myself again.

The final outcome of the inspection of my part was nearly perfect. The width of the keyway was 1.0005, and the center was only off by .0017". All within tolerance, but I intend to do better on the next one. The keyway cut by the former shop was 1.0055" wide, and 0.045" off center.

0.0017 vs 0.0450

Not bad for a first try.

 

[PetkovStoyan]

I wonder how he measured the center offset, and can he do it directly on the machine. Because if he has to take down the part to measure it elsewhere, it would be a nightmare to adjust it again in the same orientation, if required.
I wonder also how does he deal with tool wear, and does he widen the keyway left-right if the tool wears out and produce undersize dimension.


Yes, that's normal (or should be). Accuracy in the range of thousandths of an inch (or hundredths of millimeter). But not ten-thousandths of an inch (or microns) as you wrote in the previous comment. I think you mistyped an extra zero there after the decimal:

 

[geesaman.d]

If the keyslot is the correct width, perhaps cut a short piece of keystock (say 60mm long), fit it in the keyslot tilted in toward the bore centerline, and measure from the side of the key to the other side. The error will be a few .00x inches due to the fit of the key to keyslot and parallelism of the keyslot faces, but it should very easily measure a visible asymmetry.