The Uniquity of the Precision Ball Manufacturing Process

The entire precision ball manufacturing process is totally unique! Quality control techniques used on conventional products, such as first article inspection, or statistical process control and many of the other accepted techniques of normal quality control simply cannot be applied to the ball making process. What I really love to see, is someone trying to apply three sigma or even better six sigma to the ball making process. We produced 64,000 balls in a 3/16 inch ( .1875", 4.76 mm ) diameter for a major university. They wanted all of the 64,000 balls to be within three micro inches (75 nanometers) of the same exact diameter. The customer spent 4 1/2 months measuring every single ball. There was not one single ball out of the entire 64,000 balls that was outside of this very stringent tolerance.

What are the big differences in Ball Manufacturing?

All precision balls are produced in a gross-batch process. In the previous example, we only had to check 150 balls during the manufacturing process, but every single one of the 64,000 other balls was perfect, and this is by no means the ultimate quality that can be produced. Every single ball in a give batch of balls is in the same machine at the same time. None of the balls are finished until all of the balls are finished. If a few of the balls are out of round, all of the balls will be out of round and if a few of the balls are oversize or undersize, the entire lot will be also.

If the balls in a given batch are oversize, the machine is simply turned back on until they arrive at the desired smaller size. We always choose the running time conservatively, so running a machine load of balls undersize almost never happens. If one of the variables that is outside of the totally computer controlled machines program is misjudged by the machine operator, and the balls do end up below the planned tolerance, the balls are never scrapped. They are simply stocked and sold as the next smaller marking increment. See the AFBMA standard, Book #10. Making this statement doesn’t imply that lapping balls undersize is a common occurrence. It is rare indeed when a lot of balls are lapped undersize, even if this does happen this does not reduce the product to scrap. They are simply measured to a smaller marking increment that is added to our more than thirty thousand lots of ball inventory.

Another thorn in the side of auditors is our solution to final inspection. The environment inside the entire ball departments is held at the international stand temperature of 20 degrees Celsius (68 degrees Fahrenheit). This totally eliminates the necessity of moving the balls to a temperature controlled lab for final inspection. Every temperature controlled department has its own laser scale measuring machine. These devices are considered absolute measuring machines. The error budget or uncertainty of these devices is out in the sevenths decimal place. All of the mechanisms in these devices are made of Invar®, a metal that simply does not expand or contract with temperature changes.

Now we come to the final determination of the ball diameter. The very act of making a measurement will alter it! In the case of ball diameter measurement, the contact force of the measuring machine will seriously reduce the measured diameter of the ball. The AFBMA standard requires that the ball be measured between flat parallel measuring surfaces. The force applied by the measuring surfaces, will flatten both sides of the ball and at the same time the ball will indent both of the measuring surfaces. We have absolute ball diameter data which we use to correct all of these elastic deformations. This reduction in diameter is even more complicated than meets the eye. Every different ball diameter and each one the various ball materials have a different amount of elastic deformation. We have the laser scale measuring machines coupled directly into our central computer system by our own custom written software. This software arithmetically averages a number of individual measurements of each ball diameter. It then compares this information with the bar coded specifications derived from the customers’ original purchase order. If these two data are not congruent, the computer will not print a shipping label and it will not print a shipping document, so there is no way of shipping incorrect diameter balls.

This same software system is used to count the balls for shipping using electronic counting scales, so any over or under shipments simply can’t be made. In addition to the diameter measuring capability, each department has its own Talyrond sphericity measuring machine. The uncertainty of these machines is plus or minus one micro inch ( 25 nanometer ) . In addition, both pieces of this inspection equipment are calibrated each and every day, not once or twice per year. Everyone wants to take advantage of the new age of computer technology; but when someone actually does, there are many arcane thinking people who thump their chests and cite some historical precedent, like first article or final inspection to muddy the water of true progress in the quality arena.

I will admit to a bit of frustration in having to deal with auditors who cannot or will not even try to understand the very fundamental uniqueness of the ball manufacturing and measuring process. We manufacture the master balls for one branch of the U.S. military. These master balls are calibrated absolutely by NIST. With this zero uncertainty data, we operate on an entirely different level of accuracy, with more certainty of the absolute ball diameter and of all elasticity correction, than can be made in any other way.