What's new
What's new

Cross Hole - True position measurement?

colton_m

Aluminum
Joined
Apr 16, 2015
Hi Everyone,

I'm looking for suggestions on how to measure the true position of a cross-hole on a cylindrical part?
Pictured below there is a 0.275" +/- 0.0003" Dia. cross hole that needs to be centered in the tube O.D. within .0005"
Any ideas? Thanks!

Cross Hole Position Measurement.jpg

Tube OD: 0.875" +/- 0.001"
Tube ID: 0.625" +/- 0.001"
Overall length: 1.0000" +/- 0.0002" < yep 4 decimal places
Large cross hole: 0.400" Dia. +/- 0.002"
Small cross hole: 0.275" Dia. +/- 0.0003"

Material: 4140 at 38-44Rc
Quantity: 1000pcs.
 
Last edited:
Set the part in a vee block with a gage pin thru the hole. Level the pin and zero your indicator.
Rotate the part 180 and level the pin again. You'll need to be within .0005 TIR.

Any MMC on the callout?
 
Probably obscure in this case, but is the pin to be centered in relation to the ID or OD? Is there a concentricity spec (yeah, worst thing to use with GD&T, but you know what I mean.)
 
Set the part in a vee block with a gage pin thru the hole. Level the pin and zero your indicator.
Rotate the part 180 and level the pin again. You'll need to be within .0005 TIR.

Any MMC on the callout?

There is no MMC callout.
I did try this method, however since there's such a small cross section where the pin is installed I get some noticeable wobble / droop on the pin. I even ordered some custom sized, tighter tolerance pins but it needs to be nearly size for size to eliminate the movement making it nearly impossible to install repeatably. Some sort of precision expanding mandrel could help, but I'm not sure such an item exists?

Probably obscure in this case, but is the pin to be centered in relation to the ID or OD? Is there a concentricity spec (yeah, worst thing to use with GD&T, but you know what I mean.)

Sorry, I am unable to share the print. It is called out as centered in relation to the OD. There is no mention of concentricity on the print. I am turning the ID and OD in a single lathe setup, parting off then surface grinding to overall length. This milling OP is the final challenge.
 
What is the material made of? Maybe you can heat it up, install a larger cross pin and when it cools it's tight.
 
What is the material made of? Maybe you can heat it up, install a larger cross pin and when it cools it's tight.

4140 at 38-44Rc.

That may work for an initial setup but wouldn't be practical to check parts over the entire run (1000pcs).
 
There's the missing link in a thought that I had....
... well ... maybe....

How 'bout you make up a stepped pin to fit through both bores, AND have an equal step as the smaller bore on the end of the big end.
Concentricity is your friend here!

Then you could indicate the too smaller and equal sized steps level, and then compare to touching off of the top of the OD?

Getting the slop out of it was my concern.
Not sure that you are going to heat it up enough to help w/o warping the part tho?



I still stand by my first answer tho.



edit:
Well, you would end up tempering the P/H, so that dog don't hunt.


---------------

Think Snow Eh!
Ox
 
Sounds like a job for a cmm...

Yes, our CMM does work, unfortunately we don't have one on the shop floor.
I'm hoping to come up with a measurement method we can set up at the machine.
We have to chase the position throughout the day and our CMM is about 500' away and is often booked up.

Thanks for the suggestion though.
 
Well, a hard tooled approach that we use on some parts is by puting a cross pin through the hole(s) and then you can use a gauge pin(s) between that and the inside bore, but with your tollerance, IDK any "shop floor" method at all.

I don't even see one of these vision units working for that.

Now - if you want to $pend real bucks at this app, you could git one of those shop floor CMM's from Keyence.
They seem to be quite the little units!
I think that they are from $50K to $60K. (in 2018 dollars, so _ like $100K now?)

CMM | Portable, shop floor CMM for easy 3D and GD&T measurements | KEYENCE America


1605654249_0_edit.png



-------------------

Think Snow Eh!
Ox
 
With only .125 depth of hole to make a centerline a bit surprised that a CMM gives decent numbers.
Is this axis also defined off the second hole in the cmm? That may be questionable or correct.
 
1. I’m going to pretend the outside is so round that the axis of the actual mating envelope is equivalent to the axis of however you might be touching it. I’m also going to assume the cross drilled holes have perfect concentricity to themselves, Ie they are not eggs or have an odd number of lobes.
2. I’m assuming the two cross drilled holes are coaxial. Given their short lengths, good luck proving they aren’t.

Get a fixture (v blocks, bearing rail, whatever) that holds a 0.3980” pin and 0.2747” pin coaxially. You could also make a stepped pin, but Detronic pins are cheap and accurate if you can hold them well. Put the part on the two pins. Put a dial indicator/drop gauge on the high point of the tube over the pins. Flip the tube, and check again. Difference between the results is your Position error. Zero the indicator on the first measurement and you don’t have to do any math. This also covers the GO pin diameter check.


More thoughts:
3. How will the customer validate this? How would they prove it’s incorrect?
4. I wouldn’t trust a Keyence CMM to that tolerance, never mind the cost.
5. There are ways with expanding tools, or using a height stand to sweep the high/low points and letting a spreadsheet do some basic math, but I’d start with what I described above.
 
With only .125 depth of hole to make a centerline a bit surprised that a CMM gives decent numbers.
Is this axis also defined off the second hole in the cmm? That may be questionable or correct.

The hole axis is only being defined off the small cross hole. I agree, there is a lot of trust being placed on the CMM at this point.



How are you holding the part for drilling?
V fixture with the OD sat in it once set, won't move the centreline. So "should" be very repeatable?

As for checking the position while running parts, if you ream both holes (to tie the size down) and use a stepped pin, that's my thoughts for functional gauging.
Perhaps have the pin have a slow taper lead on the small hole, so it tightens going into the hole?
Then check with a v block and dti as Booze D says.

Although you will still get theoretical slop/droop in the larger hole as the pin has to slide through it....These types of things are a challenge because as Bob hints, CMM readings can be an utter waste of time on things like this!

We are milling the holes with a 3/16 endmill, I had trouble using reamers because of the semi interrupted cut when it first contacts the tube.

The current fixture is actually not a V-block style, it's pictured below. I inherited this from a previous family of parts (where the tolerance wasn't so tight). Might be time to ditch this thing and make something using a V-block. It was initially done on the 4-th axis because we had some parts where the cross holes were 90deg apart. This also allows for a chamfer on all 4 hole edges with a lolipop cutter.

Fixture Example.jpg
Fixture Example 2.jpg




1. I’m going to pretend the outside is so round that the axis of the actual mating envelope is equivalent to the axis of however you might be touching it. I’m also going to assume the cross drilled holes have perfect concentricity to themselves, Ie they are not eggs or have an odd number of lobes.
2. I’m assuming the two cross drilled holes are coaxial. Given their short lengths, good luck proving they aren’t.

Get a fixture (v blocks, bearing rail, whatever) that holds a 0.3980” pin and 0.2747” pin coaxially. You could also make a stepped pin, but Detronic pins are cheap and accurate if you can hold them well. Put the part on the two pins. Put a dial indicator/drop gauge on the high point of the tube over the pins. Flip the tube, and check again. Difference between the results is your Position error. Zero the indicator on the first measurement and you don’t have to do any math. This also covers the GO pin diameter check.


More thoughts:
3. How will the customer validate this? How would they prove it’s incorrect?
4. I wouldn’t trust a Keyence CMM to that tolerance, never mind the cost.
5. There are ways with expanding tools, or using a height stand to sweep the high/low points and letting a spreadsheet do some basic math, but I’d start with what I described above.

Thanks for the idea, if I understand correctly your setup would look like this?
Setup Example.jpg
 
The hole axis is only being defined off the small cross hole. I agree, there is a lot of trust being placed on the CMM at this point.
That isn't trust, that's irrational exuberance.
We are milling the holes with a 3/16 endmill, I had trouble using reamers because of the semi interrupted cut when it first contacts the tube.
Makes sense.
<snip>
Thanks for the idea, if I understand correctly your setup would look like this?
View attachment 332477
Correct. It may prove easier with a stepped pin (no risk of misalignment between the two pins that way), but a drop gauge top center should do the trick regardless. I'd put the tip in a fixed position with a ground flat end rather than trying to sweep a rounded one. This way you can rock the part forward/backward until you find the low reading, which will correspond to the tube being perpendicular to the flat on the end of the indicator. If it rotates about the cross drill axis you'll get an artificially high reading. Bell flat tips tend to have a tenth or so of slope to them (or maybe the indicator tip has a half tenth error where it seats?), but if it's in the same spot each time it won't matter.
 
With .125 wall, then I could get a probe to measure it. Now, as for my practice and possible understanding of what you are dealing with: It depends on how the GD&T call out is constructed. Let's say one of the faces is A, the ID or OD is B, and the other hole is C. C would be tertiary and needed to clock to.
So, say TP, Dia .005 /A/B. Then my interpretation is that they want a length from one of the ends. You might get 2-3 very close depth measurements, and see the perpendicularity of the hole. That might mean taking a measurement circle of the hole and make a clockline.
Id A,B,C, C the opposite hole, then that is clocked to it. I would set it vertically on a magnet. Then need a Z probe and two X probes or two Y probes.
 








 
Back
Top