Progress Reports
Fall 2014
Spring 2015
Prior to the spring of 2015, I had, wearing lab gloves, cut light-guide fibers with a dremel and stored them in the fiber dark box.
At the beginning of the spring semester, I worked on taking apart a wood frame used for the bending of fibers. This took some time, as salvaging the wood and screws proved difficult without power tools. However, once I had the revelation that I actually had access to power tools, the progress was speedy and efficient. The only issues I had involved some very tenacious screws that would have been much easier taken care of using some bolt cutters. I also worked on reorganizing the supplies we had stockpiled in Schweitzer Lab, as well as restacking them for increased lab safety. They are currently on a shelving that canopies the computer, monitor and bending box.
For the most of the spring semester, however, I worked mostly on production of suitable-for-testing optical fibers.
In February, Liana taught me the polishing method for Scintillating Fibers (SciFis), which involves using a collar that is slightly shorter than the length of the fibers to rub the ends of the fibers on high grade sandpaper and then printer paper for a mirror finish. I found the process to be rather efficient. However, a grey grit would be formed when accidental rubbing of the collar on sandpaper occurred, and this could sometimes scratch the fiber ends or lodge the particles in the claddings of the fibers. This was unacceptable, but the use of Kim wipes on the polishing end greatly reduced the scratches and lodging. I highly recommend some quick swipes of a Kim wipe on the end every 45 seconds or so on the sandpaper, and certainly before using the printer paper to mirror the ends.
Because the SciFis were not produced in marked bundles, it was easy to bundle unfinished SciFis from a bundle in with a new set of fibers. That is to say, if 14 out of 36 fibers did not end up getting adequately polished, one could separate out the 22 fibers that were finished and bundle the 14 unpolished fibers with 22 new fibers that needed to be polished. This increased speed, as we did not need to resand the fibers that had been finished in order to finish off the fibers that were not hitting the paper and were thusly unpolished. My best yield from a group was 30 out of 36 polished, and this was achieved by giving the fibers more slack (pushing them out further from the collar) and giving them a bit more sanding time, taking care to keep the collar as level with the paper as possible.
The Light Guide (LG) fibers were a different story altogether. As the LG fibers were over a meter in length (as opposed to the SciFis, which came in at about 2 cm), there was some trouble bundling the fibers so that all 30 would be at the same distance from the collar. Bundling could take almost an hour without the bundling tool, and with it could take about 20 minutes. The fibers would end up not being all polished, and to avoid doing many of them singly we had to rebundle them over and over again. Thus, the main issue was bundling the fibers appropriately. There were also issues holding the fibers upright while polishing with one hand, which would lead to the collar being less uniformly level and the fibers taking longer to be polished.
Finally, another issue with the LGs was the fact that they were bundled and marked 1-15 and 18-32. The issue here was that the rollover with the SciFis could not be done here, since mixing bundles would result in mixed fibers, and we could end up losing track of what fiber was from what bundle. This could be alleviated with an additional numbering rubber band being added to indicate the bundle it came from, but I did not think of that at the time and it slowed production.
In the later part of the semester, I learned fusing. Fusing seemed to be down to an efficient science, and the fusing unit had no problems at all. The only issue is ensuring the pressures remain where they should be, and this is easily adjusted. Production is slow but consistent, with little confounding factors. I had one bad fuse, and I’m not quite sure yet as to what occurred. It may have been that I let a fiber slip by without inspecting the end (perhaps it was chipped) or perhaps some sort of air bubble formed during fusing. The only other issue is the presence of many SciFis that are about .1 millimeters thicker than the acceptable range. Seemingly, there was a good portion of the spool that was much thicker than expected. The other fibers are at an acceptable width of about 1.95 mm.
Summer 2015
Over the summer I worked significantly longer hours in order to expedite the fusing process. In addition to completing polishing for SciFis and LGs, I completed about 280 fuses. Fiber measurements were marked on the "Fibers - 2015" sheet before and after each fuse.
For one reason or another the lateral and vertical pressures we were using for fusing proved to be too high, creating large wings on each fuse site. Re-calibration of pressures through fusing of test fibers resulted in lateral and vertical pressures of 26 and 82, respectively, being preferred.
Unfortunately, however, the fusing seems to have become more art than science; a variety of factors influence each fuse, requiring intuitive and immediate reaction from the researcher. Two examples follow, though it seems more appropriate to leave the adjustments to an experienced researcher's discretion. First, there seems to be considerable play in the pressure gauges, as the numbers presented sometimes seem to have little bearing on how the fuse turns out. Thus, some fuses are adjusted based on the results of the previous fuse, which seemingly has produced many fuses of better quality than solely following the pressures recorded. The adjustments may require more or less pressure, as fuse quality seems to drift in either direction. Second, if multiple consecutive fuses occur, lateral and sometimes vertical pressure may need to be lowered slightly for each fuse (possibly due to the lamp heating up and causing the fibers to melt faster).
Sanding
When learning to fuse, I was taught that the wings created on the fibers are best reduced by sanding them down with 1000 grit sandpaper.
My schedule also allowed me to attend weekly meetings and give progress reports.
Fall 2015
During this semester I continued fusing the scintillating and light-guide fibers. I had found that the pressures Liana suggested to me perform well on strength tests--they held 1100g on average. However, over the summer I was a bit concerned with the excessive flaring that resulted from the cladding moving out of place. Some fuses resulted in lateral wings forming on the fuse site, the sizes of which could be up to 1mm. These wings had to be sanded down for the fibers to fit in their final housing at Jefferson Lab, and it is of some concern that sanding removes the cladding about the fuse site and promotes light loss. A razor blade could be used to slice the wings off closer to the fuse site and minimize sanding, but my concern persisted. After consulting with Jim, I fused a few spare light guide lengths to each other and re-calibrated the fusing pressures. High lateral or vertical pressure both produced wings, but too low of a pressure would result in the fiber cores not fully fusing. Fuses with 86psi vertical pressure and 24psi lateral pressure resulted in the smallest wings possible while performing relatively well on strength tests. On average, a fuse at these pressures would break at about 1000g. After this adjustment, most fuses were right where we wanted them to be--fully fused with an unavoidable slight flaring at the corners.
The histogram below displays the distribution of widths in LGs and SciFis before and after fusing.
Strength test breaking weights are reported in the table below.
| Bundle 2 breaking test weights | Bundle 3 breaking test weights | Bundle 4 breaking test weights | Bundle 6 breaking test weights | Bundle 7 breaking test weights | Bundle 8 breaking test weights | Bundle 10 breaking test weights | Bundle 11 breaking test weights | Bundle 12 breaking test weights |
|---|---|---|---|---|---|---|---|---|
| 1150g | 1050g | 950g | 1000g | 900g | 1000g | 1200g | 1050g | 1050g |
| 1050g | 1000g | 850g | 1050g | 900g | 950g | 900g | 1000g | 1000g |
| 1100g | 950g | 950g | 1000g | 950g | 1000g | 900g | 950g | 1000g |
| 1000g | 1000g | 900g | 1000g | 800g | 1050g | 1000g | 1000g | 1050g |
| 1000g | 1000g | 900g | 1000g | 950g | 1050g | 850g | 900g | 1000g |
At least 3 bundles were fused by the end of my semester. It was slow work compared to the summer where I had 20 hours per week to dedicate to fusing. Including polishing LGs and SciFis, setup of the fusing unit, strength testing, and sanding, I was able to average 10 fuses per 6 hour session at the beginning of the semester. With experience (and once all the polishing was finished), I was able to achieve 8 fuses per hour by the end of the semester, including sanding and strength testing time.
At one point, Christina and I took some of the older fibers to a dark room and shined a laser pointer through them to examine the light loss. Inspired by this, I looked at some of the heavily sanded fibers the same way and compared to the fuses at the reduced pressures. I noticed that the sanded fibers lost a lot of light at the fuse site while the new fuses lost almost none in this limited experiment.