Tool Wear Control Charts to Help Control Machine Tool Offsets

Can someone give me some guidance on which type of control charts can we use to control a machining process in which tool wear is prominent.

We need a control chart that keeps the process in control while it also informs when an offset is required.

Using a normal X-R chart only detects the process variability and does not take into account the tool wear. Our operators have to constantly do tool offsets in order to stay in control. The machine capability is over 3 showing that the machine is very capable and that we can ride the tool wear but that takes us off the control limits.

We need something easy for our operators to follow and control the offsets without processing any out of spec parts.



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

[Don Winton]

I am not sure of the details, so I must speculate.

I would assume that the tool wear is directly related to the amount of parts produced. A simple DoE test would determine the relationship. Thus, a simple linear regression chart may be appropriate.

Use the control charts to monitor the parts, use the linear regression chart to monitor tool wear. Used correctly, the operators would know when to adjust before the tool wear caused problems with the control charts.

Regards,

Don

[Laura M]

I've seen the regression line as Don suggests built into the control chart. The control limits also follow the regression line slope +/- 3sigma on either side of the line. If tool wear is fairly consistent between tools, then the slope can be determined, applied to a starting point of a new tool, and control limits "drawn" from there. Kind of like confidence intervals around the regression line. A point over(or under) the control limit "early" in the life of the tool would indicate a special cause. Later in the life of the tool, the same sample mean may occur, but be within the limits... not a special cause. Hopefully this makes sense...easier with a chalk board...May have the formula's in some old notes somewhere. If someone wants me to dig them out, I'll try.

[Dawn]

We SHALL monitor on-going capability monthly for a customer requirement. We just started doing this and the customer hasn't seen them yet, but will soon. Almost all of our cpk's show downward trend. I think this is due to toolwear-but is only my assumption. Your thoughts? If so, could we not use these cpk's to monitor tool wear?

Thanks Laura:

It makes a lot of sense what you replied. The tool wear is due to the number of parts machined and according to a simple DOE performed by turning 300 parts without any offsets I have a clear idea of how the measurements are affected. I will therefore construct a chart in which every so many parts there will be a different target XBar and post the control limits allowed for that target Xbar.

Thank again for your valuable input to everyone.

Q. Balderas

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

[Don Winton]

Dawn,

You can chart the process capability on a control chart, just as any other variable. As for the assumption regarding tool wear, it is risky to assume.

Regards,

Don

[bobdoering]

Originally Posted by Quirino A. Balderas

Can someone give me some guidance on which type of control charts can we use to control a machining process in which tool wear is prominent.

We need a control chart that keeps the process in control while it also informs when an offset is required.

Using a normal X-R chart only detects the process variability and does not take into account the tool wear. Our operators have to constantly do tool offsets in order to stay in control. The machine capability is over 3 showing that the machine is very capable and that we can ride the tool wear but that takes us off the control limits.

We need something easy for our operators to follow and control the offsets without processing any out of spec parts.

This is a great explanation of the classic problem! The correct chart is the X hi/lo-R chart. For more information on preparing this chart, see: Statistical process control for precision machining