[pwilson101]

There is an on-going debate at our shop on whether or not gage cuts SHOULD be used when finish boring on a CNC horizontal machining center on tolerances of about +/- .002 or less.

The shop floor personnel say "yes" because we have problems of getting chips in the spindle which will help to cause oversize bores. However, management disagrees. The main reason is because we have a sister company and the management there says they don't do it and don't have any over-sized bore problems. We both run the same parts on similar machines.

(We have about 15 horizontal CNC machining centers and boring mills. We machine part sizes that would fit on a 500mm x 500mm table to 2500mm x 1500mm table. Material is usually ductile iron castings)

I know we get chips in the spindle, in fact if you look at our pockets in the tool magazines there are impressions of where chips were, therefore I think that's proof that there would be this problem. What I don't understand is how the sister company "doesn't have any problems" when it comes to this.

Any feedback on this issue? It's driving me nuts hearing from management that we should take the gage cuts out when I know (at least I believe) it will cause over size bore problems. And don't want to scrap parts out to prove a point...

And input is appreciated...

[miljnor]

It's easy to tell someone how to do something when they aren't involved with the solution or the consequences of their actions.

I can always tell you I want something perfect but you have to do all the machining with a hand file. Sure it may be possible but the guy asking isn't the one doing the work or troubleshooting the problems.

Management can say anything they like, but the fact is, 2 identical machines running identical parts and programs could have 2 very different results.

It's not the norm, but it happens, and the tighter the tolerances the more common these kinds of things become.

I had a partner that always ridiculed me for saying it wasn't possible, because he knew it was possible. So I had to change my wording to : its not possible for the amount of cash we want to spend on our parts. $3.00 parts that cost $10.00 is not a good trade for an insignificant Prima Facie feature.

[mfgbydesign]

I am inclined to agree with you, but have another approach for you to try:
First inform them you are not comfortable removing the gage cut without quantifying actual error.
Then, chart actual deviation from nominal you find on every first try before adjustment. Get at least 30 observations and run a standard deviation calculation on the data. If 8XS (eight times the standard deviation) is greater than your smallest tolerance you need to keep the "gage cut." This exercise may help you quantify the tolerance level at which you can safely remove the check. It's not a full process capability study using 1.33Cpk, but a valid use of statistics to quantify the error, and your manager should appreciate the objectivity.

[miljnor]

Then to add what MFG said:
Take the amount of scrap taking the cut out will generate and calculate the total time saved on the removal of the gauge cut. Give this data, as well as the statistical data generated by MFG's methods, to the management and let them make the decision.

After all, its their job to "manage".

good luck

[pwilson101]

Thanks! I will do just what you both recommend, and take it from there.

Much appreciated!

I just wish I had a root cause. I find it hard to believe we are the only shop in the universe with this problem. Anyone have any ideas/comments as to how chips get into spindles on horizontal CNC machines? When it tool changes, there is a strong air blast to keep chips out.

And what I mean by gage cuts is we do not back the diameter off, instead we just bore just deep enough to allow a measurement (in most cases this is 5% or less of the bore length) and we adjust the diameter (if necessary) from there.

The problem is across the board on the horizontals in the shop. From the old to the new to the small and the large. Obviously on our 10K+ castings we are taking preventative measures no matter what. On the mid to smaller stuff, it would be "nice" not to take the gage cuts but I, in good conscience, can't just eliminate them across the board because of the potential scrap that will incur. Regardless of CYA. But now I have some ideas on how to determine the capability and also cost of scrap.

Regarding the CYA, I know how to go about that with the emails/documentation and etc. And I will do this when the time comes to start eliminating the gage cuts. I want to have an idea of what is causing this oversize bore problem and make a change of some sort before removing the gage cuts, however. Which brings me to the root cause analysis...

...I'm at a loss as to the contamination and here's why: Our sister company uses the same equipment in a lot of cases, some are different, but similar across the board. When I ask the question (how do you keep chips from entering the spindle, etc) the reply is they don't do anything "special". So I ask specific questions hoping what they don't think is special is something we would think is special. Some ideas are that the chips are accumulating on the spindle housing/face and dropping on the tool taper when tool changing. So do you (my sister company) keep an eye on this or clean it every so often, etc. The answer is, "not particularly" & of course on special occasions. And to be honest, these are CNC automatic machines. I cannot imagine that shops that don't have this problem do (or can possibly) keep their spindle housings/faces clean of chips at all times. So that being said, at some point they would have to experience the same problem I'm having, IF that is the root cause. But I'm not hearing this. Another idea is the chips are coming from the tool magazine pots (but HOW do they get there to begin with?) and transferring to the spindle, etc. etc. Do they have PM that cleans these pots out regularly? Again though, how would the chips get there to begin with? And I'm talking when I look at these pots (plastic liners), they are embedded with indentations from where chips were. If they are getting there, they are getting in the spindle. HOW??

So, I would like to find the root cause if at all possible. I need other ideas cause I'm out. I'm hoping we are doing something different than those that say they don't have this problem, and I can't think of anything. And btw - regarding the frequence: We can run 10 - 20 - 30 parts in a row with nothing but normal tool wear causing the size variance, but we will eventually get an oversize bore.

Thanks for taking the time to read this and giving me ideas...

[pwilson101]

Question for mfgbydesign:

You say to chart the actual deviation from nominal I find on every first try before adjustment. What if I make a change (adjustment)? What is it going to tell me if I find the distance from nominal again (after the adjustment)? Should I instead be recording what the difference is from part-to-part? I don't know statistics well, so just asking...

Thanks!

[beekeeper]

I think we are looking in the wrong direction. I would be trying to find out how the sister company is avoiding getting chips on there arbors. The chips getting on the arbor will shorten tool life and cause all kinds of problems if it is a high rpm spindle. Once you fix the chip problem the gauge cut problem is solved

[mfgbydesign]

by evaluating whether your process is in control. By "in control" I mean in statistical terms. The best way to do this is take your data (deviation from nominal before adjustment) and plot a histogram of the data. For an example of this please visit:
1.3.3.14. Histogram
The X axis is nominal deviation. Y axis is frequency. Pick "buckets" of .0002, .0005, whatever gives you a nice looking graph.
Then observe whether you have a "normal distribution" (bell-shaped curve). You can click on the examples in the web page link provided.
If your distribution is normal, then we can move to the next step.
If it is anything but normal, we need to identify and remove the "special cause" variation. It is one of the "6M's" Man, Machine, Material, Method, Mother Nature (environment), or Measurement.
The prefered method for brainstorming potiential causes is the Ishikawa (fish-bone, cause & effect) diagram.
Another technique I might encourage you to try is the 5W's of root cause problem solving. Why, Why, Why, Why, Why.
I might have some suggestions for you but you are on the right track, visiting the problem, asking questions, looking for physical differences. You will make progress. First measure where you are at so you can tell when a change has helped.
One more comment for now. Only measure the deviation from nominal before adjustment. When you factor in the adjustments of an out of control process, it gets complicated to control (account) for that in the analysis.
Please, start with some measurements and a histogram (frequency diagram). As Demming said, it is much easier to fight an enemy you can see.

[miljnor]

sometimes the chips get on the tool not the bore and the resulting issue is the same (like was stated above).

I had 2 machines with the same tooling running the same problem and one of them had a chip issue. Chips building up on the tool changer cause it to get stuck. We made a shield to prevent the problem on the 1 machine and it took forever to figure out why one machine was different than the other.

Ended up being the coolant pressure on one was vastly different than the other. The higher pressure and the direction the nozzle was shooting it was just enough to push the chips up onto the tool changer.

So in conclusion it might be something as minor as a coolant direction being off by as much as a millimeter and the pressure being higher/lower than the sister companies machine.

good luck.