Skip navigation
AM Forum

Know the problem before applying technology

After reading the article on Spindle Probing Benefits (page 56, 6/11/2008), I am compelled to respond for fear that the article may lead some less experienced individuals into a tail chasing episode, and it is my opinion that Moise Cunnings misstated the best and proper use of “secondary” fixture offsets and further, spindle probing.

If a part has tolerances such that the machine metrology does not provide a six sigma level of assurance to be able to hold those tolerances, then the part should not be run on that machine in a production capacity.

Newer and higher quality machine tools have laser qualified pitch error compensation which is sometimes enhanced with thermal compensation features. In addition, one of the primary reasons for the use of datums and geometric dimension and tolerancing (GD&T) is specifically to prevent the very scenario presented.

Rough stock?

Engineers designing precision parts that can be manufactured from sawed or, worse yet, sheared bar stock are applauded for their cost saving foresight. Forged and cast surfaces or features that are not machined are an extension of that same thought process.

The Manufacturing Engineer must be mindful of all upstream and downstream processes that will affect the current process.

If it is given that:

The machine tool is in good condition.

The fixture is correctly designed and constructed.

The cutting tools are sharp and used correctly.

The part is held firmly without distortion.

Proper machining methodology is followed.

The CNC program is correct.

It should follow that all features created in a single clamping will be in relation to each other within required tolerances.

In the cast part scenario of the article, if the part is loaded incorrectly and the program puts the first hole at (X250, Y200) and the second hole at (X250, Y200); the two holes are concentric because they were put in at the same coordinates.

The second hole is correct to location because it is concentric to the first hole.

The part, however, is scrap due to the fact that the first hole is off location from the edge.

Now let’s look at the same scenario with spindle probing and a secondary offset.

The part is loaded incorrectly and the program puts the first hole at (X250, Y200), the probe finds the center of the hole and sets a second offset (G55) at (X250, Y200) from the first offset (G54). The second hole is then placed at (X0, Y0). The two holes are concentric because, like it or not, they were put in at the same coordinates.

The second hole is correct to location because it is concentric to the first hole.

The part, however, is scrap due to the fact that the first hole is off location from the edge.

The difference is that, this time, it took longer to create that piece of scrap.

As with any problem of this nature, about 95 percent of the fix will be diagnosis of the true root cause of the missed location. In other words, when you know what the problem really is, the fix is generally pretty clear. This time the issue is that evil snake we all know as human error.

Most machine tools today that offer spindle probing come with a variety of canned cycles or probing routines that can edge find, center find and even rotate a coordinate system.

The fix then, is to insert the probe routine to edge find, rotate the coordinate system and adjust the fixture offset after the human contact and before any material is removed. However, this particular solution is not quite complete.

Knowing that we are now going to probe the side and end of this casting to set the coordinate system, we should give ourselves the best shot at getting it right every time.

In the classic 3-2-1 nest of the fixture, I would suggest that the fixture stop points for the secondary (2) and tertiary (1) locators be moved off to one side or the other far enough to allow the probe to access those points under program control.

Finally, if the part is mislocated by a significant distance, the first probe cycle will not hit the actual tooling points. A second or iterative probe cycle will ensure that the coordinate system is set correctly and the part will be right.

We have now created a good part without the added confusion of two or more offsets but, rather, by utilizing one correct offset and a less confusing program with a single datum.

Our benefit here is not any kind of change in throughput.

The benefit actually comes in the form of more saleable parts as a result of the elimination of scrap.

By the way, in the unlikely event of employee turnover, we also now have a process and program that is easier to understand, read and therefore to diagnose should problems arise in the future and a different manufacturing engineer or CNC programmer must provide a resolution.

Obviously, this is a quick overview and there are many other ways to solve these kinds of issues.

We have not discussed machines with marginal metrology. Before I had a G code for positive approach, I used to program this long hand on older machines making closer tolerance parts because, as an apprentice, I was taught to always take the slack out of the lead screws in the same direction on a Bridgeport.

Nor have we talked about the human factor. Highly skilled, real machinists vs. “I run 1 operation on 1 part on 1 machine so I am a” machinists. That would be a separate and much longer letter.

My bottom line... My primary point here is that if you wish to utilize technology to fix a problem, try to make sure that you know what the problem really is and how to properly apply that technology.

Your bottom line will improve.


David Jones
Senior Manufacturing. Engineer
Pratt & Whitney DARO
1177 N. Great Southwest Pkwy.
Grand Prairie, TX 75050

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.