Rotating parts in a V-block

Not following you.
There is backlash in a dividing head meaning that each individual part would need to be aligned horizontally, And then when indexing the dividing head, you would have to account for the error in the head, (that is backlash, imperfections in the head), which may not be good enough for the part. If you were to locate one of the upper internal flats of the internal square feature on a known height, known horizontally parallel to the surface table, all that you would have to do is calibrate your height gage mounted DTI and the check the lower internal flat for height and parallei, rotate, rinse and repeat.
 
There is backlash in a dividing head meaning that each individual part would need to be aligned horizontally, And then when indexing the dividing head, you would have to account for the error in the head, (that is backlash, imperfections in the head), which may not be good enough for the part. If you were to locate one of the upper internal flats of the internal square feature on a known height, known horizontally parallel to the surface table, all that you would have to do is calibrate your height gage mounted DTI and the check the lower internal flat for height and parallei, rotate, rinse and repeat.
A dedicated fixture to "hang" the socket from is a good idea!
 
You could use something like a Haimer in a mill setup with DRO. Just run the axis from 0 on the Haimer to 0 again and the DRO would read the width of the socket. You would have to align the socket closely to the mill axis but the cosine error would be pretty small if you were off a little. Perhaps a primitive jig to line it up and clamp it? I am not sure how you would record the DRO result but I'm sure it could be done.

Actually the cosine error is a little worse than I thought. If you need .001" accuracy, assuming the socket is about 1" ID you can only have rotational error of about .06 deg if I am doing my math right? But then this type of error would apply to a lot of solutions.
 
Last edited:
You could use something like a Haimer in a mill setup with DRO. Just run the axis from 0 on the Haimer to 0 again and the DRO would read the width of the socket. You would have to align the socket closely to the mill axis but the cosine error would be pretty small if you were off a little. Perhaps a primitive jig to line it up and clamp it? I am not sure how you would record the DRO result but I'm sure it could be done.

Actually the cosine error is a little worse than I thought. If you need .001" accuracy, assuming the socket is about 1" ID you can only have rotational error of about .06 deg if I am doing my math right? But then this type of error would apply to a lot of solutions.

Yes, it's quite sensitive to rotation. I calculated that 1 degree of error results in +/-0.006 as you sweep across.

I like the idea of inspecting in-place on the mill. That's one of the ideas I took away from the Moore books---their machines were precise and accurate enough to that it was quicker for the inspector to come to borrow the operator's seat than for the oeprator to bring the work to the inspection department. The only argument against in-place inspection is that I don't have a way to get output from the DRO into a computer. Also, I'm trying to get 100% coverage of dimensioned features, which requires multiple setups.
 
Back
Top