Recently we have seen two cases where the lab managers were experiencing prism problems to a serious degree. In one case we were able to correct it by phone, in the other case, I paid a visit to the lab to address this concern, and to correct a few other problems. In both cases adequate air supply was a problem, which lead to the blocking machines not operating properly due to a lack of air pressure and flow.
Case 1: For a few months, the lab manager complained that their new autoblocker was not performing as well as the five-year-old unit in their lab. They were experiencing 20-30% reject rates from this blocker. Over the phone and through email, we discussed the factors that influence prism; blocker alignment, collet operation, air supply, lathe collet condition, and blocking tool condition. When the above factors were tested, the lab manager found that the blocked base curve had as much as .5 mm of face run out from one side to the other. This tilt was inconsistent in magnitude and location. This is an extreme amount of tilt that exceeds the physical limitations of the blocking machine's movement. Since the tilt was random, blocker alignment was not the cause. He also found that the "System Pressure" was reading 80 PSI (5.5bar) when it should have been 90. This pressure value raised concerns about the lathes, since I understand that they require 80 PSI to operate before a safety low pressure switch shuts them down.
The lab manager said that since they had re-arranged the lab, the compressor had been putting out about 90 (PSI) and it was about 80 at the blockers. I asked him how far the blockers were from the compressor as compared to before the lab was re-arranged, and about the hose diameter. He said that the blockers were about 10 meters (33') farther from the compressor than before and the hose was 6 mm I.D. This created a restriction of air flow. I asked the manager to watch the gages as the blocker progressed through its cycle. During cooling, the "System Pressure" dropped to 50 PSI (3.5 bar). This was insufficient pressure to hold the base curve in the lower collet. When the air cooling came on, it blew out of each collet slit and lifted the base curve up in the collet, resulting in tilting the lens. The real cause of the prism error in this lab's lenses was an insufficient air supply for the new lab layout.
A temporary fix was to raise the air pressure at the compressor to 120 PSI (8 bar), and to buy an air tank and place it under the table at the blockers. The air tank acts as a reservoir that provides a volume of compressed air near the blockers, which neutralizes the restriction of the long, small hose. This took care of the low air flow and ensured that all the equipment had sufficient operating pressure. This case was handled remotely over the course of a few months. It was a very difficult case to diagnose because all of the information needed to be requested and then delivered. Getting the information necessary to solve problems takes time, patience and understanding.
I followed up with the lab manager the next week about his yield numbers. He was running 10-15% rejects on torics. He was happy with the improvement and is expecting it to get better as he tracks the rejects to identify the cause. A word about yield: Reject numbers have risen a little industry wide, I think, because all lens designs are being manufactured on the same production line. That is to say, specialty lenses have increased in volume relative to spheres compared to the years past, so much so that generally these lenses are no longer manufactured on a separate line from the "main production." Since specialty lenses have more parameters to hit, it's reasonable to assume that there will be more rejects. For example, on toric lenses, getting the prism axis correct can be difficult. This is especially true on low prism lenses. If there is an unknown small amount of prism at say 90 degrees to the desired axis, it will show up as an axis error. Prism is inversely proportional to axis error. This precision is dependent on the tooling, blocker, and lathe precision.
Case 2: Prism symptoms and yields similar to Case 1 were creating havoc in a lab. Again, it was proving difficult to diagnose the problem over the phone. We communicated via Skype, but still could not see the process as well as being there. The fact that this lab runs several production lines was helpful in monitoring the yields relative to each line. I visited the facility. We looked at the various lines and then concentrated on the one with the worst yield. They were not experiencing the severe tilt as in Case 1; prism axis on toric lenses was the problem.
The worst reject problem occurred on torics with the prism axis being 20-40 degrees off either side of the desired axis, especially on lower prism lenses. This indicated an accurate process, but with a lot of variation (poor precision). I knew, from when we introduced the original static blocker in the mid '90's, that the blocker was great for dialing in prism, but unless you had a very precise cylinder crimper, you would see off axis prism. With the new free-form lathes, turning prism is a matter of entering the parameter values. As long as the lens is blocked properly and the lathe collet is running true, the lathe will turn what you want.
The poor precision lead me to concentrate on the "blocking system." The "blocking system" consists of the lathe collet, blocking tools, and blocking machine as a unit. By treating these elements as a system, we are more easily able to correct problems. We checked the lathe collet run-out, picked several good condition blocking tools, and ran a test. We needed to find the cause of the random error in the process. One indirect problem was that the autoblocker was fastened to a table making it difficult to view the back panel. From my experience with Case 1, I insisted on viewing the back panel and sure enough, the lower collet pressure was low. We turned up the pressure to the required setting. We saw an immediate improvement in axis precision. Increasing the lower collet pressure caused the base blocking tool to be held more firmly and precisely, improving the repeatability of the system. We checked the blocker calibration, and it was within .015 mm TIR (.0075 mm prism). No calibration was needed.
After checking a few other aspects of the system; blocking compound drop size, over-all blocking tools condition, and the autoblocker thickness gauging device. All of these elements checked out, so we proceeded to make good lenses.
It is important to ensure that the basic facilities are in good working order and that machines are at the correct operating settings. A simple re-arrangement of a lab needs to be analyzed relative to all the systems: air, electrical, vacuum, floor support for lathes, etc. We were happy to be included in the solution.
