(Edited by Kyle Sawyer)
That Prism Thing
By Erik Larsen
One important goal in lens manufacturing is to produce prism free lenses. Over the years, I have heard various comments about prism and how to control it. Most of them have to do with the problem being “corrected” at the front cut lathe with the “finger (vibration) method”. This uses one’s finger or nail against the side of the lens blank while it is spinning in the lathe collet. People use this method to determine if the lens is running true enough. The problem with this is, people’s sensitivity varies and an individual’s sensitivity varies throughout the day.Another problem with checking Runout at this step in the process is the time and disruption it causes in the manufacturing process.I’ve actually had a technician tell me that anything less that a ¼ diopter isn’t worth trying to correct! He was not lathe cutting the edges.Excluding the manufacturers who are lathe cutting edges, is it worth trying to control prism better than 1/4 diopter? What if there was a systematic approach to reduce prism?Let’s look at what factors influence prism.
Base curve lathe collet condition.Front curve arbor (chuck) condition.Blocker alignment.Blocking compound temperature.Front lathe collet condition.
The base curve collet is often overlooked as a source of prism because only the outside of the blank tested for runout in the “finger” method. If the base curve collet has runout, the base curve and lens diameter will not be concentric with the outside of the blank. If the base curve is not blocked concentric to the O.D., then there will be prism in the finished lens. This prism will be a repeated amount, but the axis will be random.The front curve arbors must be in good condition on the12.7 mm diameter so that they fit well into the lathe collet. An outsized or damaged shank will result in random prism and axis. A recent trend is to use solid Delrin arbors because they are so forgiving with older collets.The alignment of the blocker must be confirmed. If it is out in either the side (radial) or the face of the blank, repeated prism and axis will result.It should be obvious, but the compound temperature must be cooled so that the lens will not move before the lens/arbor is removed from the blocker. I have seen them removed early and the operator told me to “hold the arbor upright so the lens doesn’t sag”. If the lens moves, the result is random prism and axis. I know of a few manufacturing managers who let the lens/arbor cool for the full 30 seconds to assure that the compound is stable.The front lathe collet must be validated to yield a minimum runout. Front collet runout usually results in constant prism and random axis.
So, the finger method to check runout may or may not find the entire prism, and certainly not to the precision necessary.
Here is where I’m going with this: To make prism free lenses, all of the above sources of prism must be dealt with.
I have received comments that blockers are hard to align, and should be made easier to calibrate. It is possible to calibrate the blocker “to itself”, that is, validate that it is blocking true. But, as shown above, this step is only about 20% of the prism control process (one of five steps). “Self aligning” the blocker misses the four other sources of prism.The goal is to make prism free lenses. I have seen small labs make 70 lenses per day and have to rework 20 of them in the afternoon to get them all out the door. Getting the process setup correctly and making the lenses right the first time is much more productive.The procedure is to periodically check the condition of the elements in the process. Early on, this should be done every 1000 lenses of so. This means that a company making 100 lenses in a day should do a check every two weeks (assuming good housekeeping). A company making 1000 lenses per day should check the process elements every morning. This is actually happening. You can make a lot of bad lenses at a 1000 per day production rate, so you better know that the process is working correctly. After a track record is established, longer periods might make sense.
CLEAN ALL OF THE COLLETS AND SPINDLES
Measure the diameter of a blank to find one that is 12.7 mm and is round. A gage pin may be more useful at this step. Put it into the base curve collet, and check it’s runout with a .002mm dial test indicator. Clean the collet to minimize the runout.Measure the arbor shanks and inspect them with a loupe. Make sure that the arbors are within the prescribed length tolerance. If you use 1” (25.4 mm) arbors they should be within .010” (.025mm). DISCARD any out of size or damaged arbors.
Pay particular attention to brass and stainless steel arbors.Place the gage pin in the front lathe collet and take a reading of the runout. Clean the collet to minimize the runout.Calibrate the blocker using the factory recommended procedure of blocking a known round blank and arbor and checking the runout in the front lathe.Make sure that the compound is cool before removing the arbor from the blocker.Place the lens/arbor in the collet for final validation. If there is repeating prism, use this information to make the blocker adjustment. If the prism is random, check the other elements: arbors and collets.The indications are: Random prism; check collets and arbors.
Repeating prism; Check the blocker alignment.After a few weeks of getting used to this routine, what usually happens is the collets get cleaned and the prism is under control. The blocker seldom requires adjusting. This systematic approach takes into account the interaction of fall of thee elements and shows where the sources of error are.Those of you who have ISO certification know about the improvement part of the protocol. This validation procedure fits right in with that section.So, what about the value of knowing you are making lenses with a good start? The first improvement is eliminating the judgment call and time to check lenses with a “finger test”.
That is good for at least 5 seconds of cycle time plus the benefit of the enhanced confidence in the process. The control that you gain on your process will allow you to start looking at precision edge profiling which will reduce the time at the final inspection station. Look into the reasons that the doctors are giving for a returned lens. You may find that the better process control will reduce the number of returns.
Many of you are familiar with my concerns about getting your process under control. This method of systematically monitoring the lens holding elements is a step in the right direction.