Not enough discrimination? 4-digit Micrometer for .0015 Tol :burr ht of slit coils

I am attemting to do a control chart X-R on the Burr height of slit coils. The customer tolerance is .0015. We use a 4 digit micrometer, as specified by our customer, to do the measurement. My problem is that their is not enough discrimination in the micrometer to establish a good range chart. Can anyone help me out? If anyone is currently monitoring burr on slit steel coils
please drop a line?

[Kevin Mader]

How large is the coil? Could you use a Drop Indicator or a Dial Indicator that is sensitive enough? I would think a Super Mic would be out of consideration unless your organization already has one.

The coils can average anywhere from 6000lb.to 50,000lb. The measurement would be taken after the coil is run through the Slitter knives. The thickness of the coil sheet is measured just inside of the edge and then the edge will be measured where the burr is located. The reading of burr height is determined by taking the difference between the burr height thickness and the thickness just inside the edge.

[Kevin Mader]

I take it that you are not in a position to have a sample of each role at your disposal and that measurements must be made at the location from the coil itself. Hmmmmm.....good question. I'm afraid that I haven't an answer for you (as to measuring device). Sorry. Is there a way that you could get samples during the process (beginning or end of each coil, a sample is cut)? I am not aware of how the process runs or if this is possible. I do know that if one of these coils should happen to become unsprung, it could be a deadly outcome (even at a far lower tonage).

Regards,

Kevin

Hi J Verdi;
Are you measuring thickness or height? By insudtry standards, a .0001" micrometer is all you need and is certainly available, if you are measuring thickness, as from your description. Federal and others make measurement instruments that read to .00001."

If you are measuring actual height, this is of course a little more difficult, and an indicator is certainly a good choice.

But if I understand the production, these coils of steel are continuously moving? Difficult to measure in any way.

In-line laser micrometers can be also used to advantage, although in this case it would only give the thickest cross-section. If you know the amount of the slit thickness and its variation you could calculate the burr.

[Don Winton]

J. Verdi,

I am attemting to do a control chart X-R on the Burr height of slit coils.

I assume you mean an Xbar-R chart.

It appears that you measure the thickness of the slit coil towards the center of the cut, measure the thickness at the edge of the slit, and use the delta of the measurements to determine the burr height. Is this correct? If so, this may present some MSA issues, but not serious ones.

You also indicate the customer 'specifies' a four digit micrometer. Is this also correct? I would think that with a 0.0015 tolerance, the customer would want more resolution, but I may be wrong.

If so, it appears there may be two areas that need to be addressed. First, if you are going to try to determine range for an Xbar-R chart, you should be using at least a five-digit micrometer. But, this may conflict with customer requirement. Therefore, clarification from your customer may be in order.

Second, Kevin indicated a dial micrometer could be an alternative. This would require the micrometer be 'zeroed' at the area next to the burr, then moved to the maximum burr to give a reading of actual burr height. This would also give the resolution you need. But, this alternative may not be practical in your operation and more details would be needed.

There are also other things not covered in your post that may help. Under the assumption the coil thickness is 0.020, what is the variance over the length of the coil? How many subgroups are taken for a given length of coil? What is the purpose of the Xbar-R chart? You indicate that you are having trouble getting resolution for the range, which would indicate the process is running rather consistently. Does the chart add value to the process? Things to consider when implementing control charts in a process.

Regards,
Don

------------------
Just the ramblings of an Old Wizard Warrior.