Thread: Locating a bore

Thanks Michael for showing this, learning something every time I look here.

Thanks for sharing Michael.

There have been times where I have had to position a bored hole in the mill and used a DTI and a gooseneck fitted in the collet. One would expect that it would be child play zeroing the spindle axis in the centre of the hole but there have been times when the readings gave an impression the hole was more an ellipse than a circle! Looking back, I can only assume it's a consequence of a less than perfect geometry between the dial stylus and the hole?

Simon

Hi Michael,

If I'm understanding your first test correctly I think its invalid(though granted it its only a 0.1mm offset I'm not sure the error is worth talking about by I havent done the maths, it depends on the hole size). You centered each axis separately?(I think so) By finding the local minima you've aligned the center of the DTI with the center of the bore(not the center of the spindle) when you turn the spindle 180 the center of the DTI is now offset from the center of the bore by twice the offset between the spindle and the center of the DTI. Clear as mud?

Originally Posted by Michael G

From memory you can run up to 500rpm, but the slower the better if you don't want the needle to be a blur.

800rpm for what its worth(its not going to be less of a blur going 300 rpm faster )

Originally Posted by Michael G

The dial is graduated in 0.0005" (5/10th) but the fine print in the instructions say that is for a 4" outside sweep tool.

Sort of, I've not seen one for outside sweeps, but you aren't doing that so it doesn't matter ATM.

With feeler length and bore diameter you can apply a correction

http://www.blakemanufacturing.com/pages/coaxvalues.html

1" bore and 2" feeler? each grad is 0.00054"


Stuart

Originally Posted by simonl

There have been times when the readings gave an impression the hole was more an ellipse than a circle! Looking back, I can only assume it's a consequence of a less than perfect geometry between the dial stylus and the hole?

There's a big can of worms here. I was getting readings that looked elliptical too with differences of up to a thou. Could be less than a perfectly round probe, could be the hole really is elliptical (bearings have clearance etc) may be misalignment between the axis of the measuring device and the machine. I figure if I can get readings on opposite sides the same and similarly at 90 degrees even if there are roundness errors I probably have a concentric axis or close to. Thermal effects are another one that I have not got a clue about. I'd like better than the numbers I've quoted but at the same time realise that it could take some time to work out the why of it all.

Michael

Originally Posted by simonl

Looking back, I can only assume it's a consequence of a less than perfect geometry between the dial stylus and the hole?

Shouldnt be, IF the hole is round and the bore is centered, the contact point between the bore and the styles will be the same through 360

Stuart

p.s. Michael how close did you tram your mill?

Originally Posted by Stustoys

If I'm understanding your first test correctly I think its invalid(though granted it its only a 0.1mm offset I'm not sure the error is worth talking about by I havent done the maths, it depends on the hole size). You centered each axis separately?(I think so) By finding the local minima you've aligned the center of the DTI with the center of the bore(not the center of the spindle) when you turn the spindle 180 the center of the DTI is now offset from the center of the bore by twice the offset between the spindle and the center of the DTI.

This was my reasoning -
Set up.jpg
Michael

Ok I didn't understand your test correctly, sorry about that.
How close was the repeatability when measuring Y1?

Stuart

Originally Posted by Stustoys

How close was the repeatability when measuring Y1?

Positional measurements came off a DRO with 5 micron scales. To make sure I had the minima I would wind back say 2 divisions (where the movement was clear and positive), note the reading, then wind forward until I'd gone through the minima and back to the division I noted the first measurement from. The average of those two readings was then Y₁ or what ever...
I think my method for getting the true centre is reasonable. I just wish that the deviations from true position were smaller as at the moment the size of them makes me think there is a significant error in there somewhere rather than a random scatter - of course it could be that I'm just measuring beyond the repeatability of these devices or my mill.

Michael

Anyone used an Indicol?

Originally Posted by .RC.

Anyone used an Indicol?

I just use the TLAR method myself....

PDW

If its good enough for Moore it good enough for me.


. moore jig grinder.jpg

Like all precision measurement there are inaccuracies built into every layer of the system. The more accurate you need to be the more you need to try and minimise these. So the more accurate the reference circle you are measuring the better (ie drilled hole vs bored/reamed/ground ) and the more accurate measuring device (ie mill with indicator vs say a jig borer or grinder) the more precise you will be. So low speed pulley measured in mill is more than adequate whereas locating a hole for a high speed jet engine shaft would require moore ( machining pun!!) accuracy.

Having said that the exercise of trying to be as accurate as possible, even when it is not essential, is important practice for when it is needed.
Thanks Michael for your efforts in this regard.

Mark

Very valid comment Mark.

The fundamental thing about accurate hole location, no matter which method is used is to have the bed-to -spindle relationship absolutely square in all planes. With a jig borer this should be a given but with a mill there is scope for misalignment. Any such error will show the hole indicated as being elliptical when using a co-axial indicator.

Moore also say, if you're going to stuff about with 1/10th indicators "an extremely rigid mounting must be provided"

Now while my testing did show errors up to 5 thou, that was with mounts that look far less rigid than the set ups Michael is using so I doubt its the full story.

I did do some deflection testing on DTI's(well its more the mounting really) but not on my Blake knock off (I'll try after lunch if I can get to it).

(In theory, I've not tested this as yet) There is a way to check this. 0 your hole in with your DTI, then Z the DTI out of the hole and back in taking a reading at all 4 points, this should mean that the dti only reads "on the way up" this should remove errors due to the difference in deflections between the "up reading" and the down reading"

Stuart

Originally Posted by Stustoys

p.s. Michael how close did you tram your mill?

Originally Posted by Peter Fou

The fundamental thing about accurate hole location, no matter which method is used is to have the bed-to -spindle relationship absolutely square in all planes. Any such error will show the hole indicated as being elliptical when using a co-axial indicator.

My mill head doesn't nod but will swivel side to side. Typically I tram it in to be less than a thou difference over a 12" span. (The target is spot on but the adjustment method is such that less than that becomes a bit hit and miss).
Understand about a non square hole showing up as elliptical but in this example the tram of head had not been moved since boring the holes, so I think that cause can be ruled out in this case.

I agree with Mark that practice is important for when it is important - in this case it's not but at the same time I am puzzled why I have these differences

Michael