A collet application that is unique to optics manufacturing

I recently was asked about gage pin vs. arbor precision. Clearly, certified gage pins are more precise than an arbor. The question was prompted by differing collet runout readings with the two. Most notable was that the gage pin had more runout than the arbors. And, the gage pin runout was different depending on how the pin was held in the collet while closing it. How could that be? One explanation could be that plastic arbors are more “forgiving” and do not respond to minor defects in the collet. But a new collet should give the same readings from a gage pin to an arbor. 

Many of you know that we give away certified gage pins at the lens manufacturer meetings. The 12.7 mm diameter gage pins are used to validate the lathe collets. This has been a nice promotion that has real value. It fits in with our mission to apply measurement to the CL manufacturing process.

We looked at the certification of these pins and found that it has to do only with the diameter and straightness, not with the ends. So, with a gage pin inserted onto a collet with a bottom ground face, the pin will either fit against the face or the diameter of the collet. If it is against the face it may not be straight in the collet.This problem is partly due to traditional use of collets from the machine tool industry. That is, metal turning is usually done with long (meters) bar stock where the collet is straight through. One part is made, and then the rod is pushed through to a stop for the next part until the rod is used up. Step collets are rarely used and usually only for second operations. If a machinist wants to validate a collet, a gage pin is inserted fully engaging the collet and a reading is taken. There is no face or shoulder to worry about. The CL industry uses step collets or bottom stop collets to hold the base curve or arbor. So, this requires an end of the arbor that is perpendicular to the shank of the arbor. This squareness is a result of how the arbors are made and is usually very good. This explains the better readings with a good arbor vs. a gage pin. 

What to do with the gage pin? Since the pins are about 50 mm long, the pin can be cut one in ½ and grind the ends perpendicular to the diameter. This is a bit of work, but this is what is required for a validation tool.One creative lab tech asked about locating off of the nose of the collet, thus avoiding the gage pin end problem. To my knowledge, collet noses are not precision ground (unless specified) and should not be used for precision location. A better approach is to use the face of the spindle for a flanged arbor location which is about as precise as it gets. This arrangement also addresses the dead length approach to holding the arbor. With collet pulling the flanged arbor against the nose of the lathe spindle provides a very stiff and secure chucking arrangement.