Wrinkleless Mylar Shell and GEM Foil Prototypes for The BONuS Detector at Jefferson Lab

Transcription

1 Wrinkleless Mylar Shell and GEM Foil Prototypes for The BONuS Detector at Jefferson Lab A thesis submitted in partial fulfillment of the requirement for the degree of Bachelor of Science in Physics from the College of William and Mary in Virginia, by Wentao Xu Advisor: Prof. Keith Griffioen Prof. Irina Novikova Williamsburg, Virginia May 2018

2 Contents Acknowledgments ii Abstract v 1 Introduction 1 2 Design, Procedure, and Attempts 3 3 Conclusion and Next Step 18 i

3 Acknowledgments I would like to thank my adviser Professor Keith Griffioen for always being kind, patient, and supportive not just to my research and study but also to my life. ii

4 Abstract Jefferson Lab is preparing for another Barely Off-Shell Nucleon Structure experiment (BONuS) with four times the luminosity of the original experiment. Therefore, a longer Radial Time Projection Chamber (RTPC) is being built that can also handle twice the rate. The goal of this project for the first semester was to find an effective way to produce a cylinder made of thin conducting mylar, which is wrinkleless and so during the experiment. For the second semester, the goal was to conduct a similar research on GEM (Gas Electron Multiplier) foil in order to make a perfect cylinder out of them. These two pieces of apparatus are crucial for the BONuS experiment which will measure electron-quark scattering from a neutron target.

5 Chapter 1 Introduction The barely Off-Shell Neutron Scattering Experiment (BONuS) at Jefferson Lab [1] seeks to measure the momentum distribution of quarks. The RTPC is used to detect the low energy recoiling spectator protons released when an electron scatters from the neutron. The detector contains a cylindrical cavity filled deuterium target gas. The outer section is filled with a gas mixture. The cavity is placed in the constant magnetic field of 5 Tesla. When electrons scatter from the neutron, the spectator proton released will move in a curved path under the influence of the magnetic field. The mylar shell is one of several layers inside the detector. It is used to apply a radial electric field to the chamber. This cathode must be as thin as possible because the recoiling protons are easily stopped by too much material. It must be wrinkleless to provide a uniform field. Ionization trails in the detector are amplified by three layers of GEM foils, which also must be precisely cylindrical shells. The GEM foils creates an amplified electric signal. That can be further amplified by micro-electronics. The pattern of hits on the foil corresponds to various proton momenta which can be determined via tracking software. The structure of the detector is shown in figure 1. Figure 1. The cross-section diagram of the RTPC detector. Former William and Mary student Wenqing Zhao attempted to solve this problem and wrote an honor thesis on the subject [2]. By the end of her research, she 1

6 was able to make a 5 μm cylindrical mylar shell with a wrinkleless left side. But getting both sides simultaneously to be wrinkeless was problematic. She ascribed the wrinkles to the nonuniform tension caused by the glue and concluded that it is possible to make perfect wrinkleless 5 μm mylar cylinder. This is where I picked up with what she had left off. My job was to modify and improve the design to get better wrinkleless mylar shells and find out an effective method to make the GEM foil cylinders. 2

7 Chapter 2 Design, Procedure, and Attempts 1. Mylar Shell The basic production of mylar shell remained the same as Wenqing s process. I used a PVC pipe fixed on an apparatus that consists of a base and four adjustable vertical stands. The following is Figure 2 I took from Wenqing s thesis [3]. It illustrates the general process of making the mylar shell: Wrap the pre-cut mylar sheet tightly around the pipe and glue the sides together; slide the rings which just fit the size of the pipe onto both ends of the cylinder; glue the rings with the mylar. Figure 2. Diagram of the PVC pipe arrangement for producing mylar cylinders. The rinds, pictured as dotted circles, are able to slide onto the PVC pipe once it is wrapped with mylar To get familiar with the Wenqing s research, I started with mylar that has thickness of 100 μm. The first problem I encountered is how to cut a perfect rectangle of mylar. That means a rectangle with straight edges and right corners. The very first approach I tried was to use a scissor. That yielded relatively satisfying result, as the edges are easy to cut straight. However, the angle of corners sometimes was not ninety degrees. Therefore, I tried to use a trimmer instead, which is shown in Figure 3 below. 3

8 Figure 3. The ruler trimmer in the lab By aligning an adjacent side of the mylar sheet along the side of the trimmer, I could cut the square with right corners. The only problem with this trimmer was that it only fitted a sheet with maximum length of 30 cm, but the cylinder needed for the experiment at Jefferson Lab is 40 cm long. That means any sheet cut by this trimmer can only be used for trying ideas of cylinder production. After I was able to get an acceptable mylar rectangle, I started my attempts to make the cylinder. The problems I encountered here are, first how to make the length or height of the square just equal to the perimeter of the pipe, and second how to put the two sides of the rectangle together. To solve the first problem, I first wrapped a longer sheet tightly around the pipe and marked where one side met the other. Then I cut along the mark, but the result is that two sides cannot meet each other and this left a gap about 1 mm wide as shown in the picture below. 4

9 Figure 4. There is a small gap between the edges As you can see, the yellow part is tape. The two sides could not reach each other. Therefore, when I tried again, I cut a few mm far from the mark so that the two sides can just reach each other. Another method was to do calculation by measuring the diameter of the pipe, but again I needed to leave a few mm when cutting. Then for the second problem regarding putting together of the sides, I gave up the idea of using glue and used tape instead because Wenqing mentioned in her thesis that applying glues can easily cause nonuniform tension which gives rise to wrinkles on the cylinder. The way I taped the sides is that first taping half width of the tape to one side. Then wrap the sheet tightly around the pipe and put the other side on the other half of the tape. This is very hard to operate, but after several attempts, I was able to get a relatively good cylinder, although one end of the cylinder is slightly bigger than the other end due to bad operation. However, the tape can offer more consistent tension than the tape and add very little thickness to the surface. The product is shown in Figure 5. The experimenter just needs to pay extra attention when 5

10 taping the mylar and it may take a few tries. Figure 5. A sample mylar shell made with tape. After I got familiar with the procedures and techniques, I moved on to the 5 μm mylar as shown in Figure 6. The mylar sheet with a thickness of 100 μm has very few wrinkles and it is self-sustaining, but the situation is completely different for 5 μm mylar. Figure 6. The 5 μm mylar is very thin. 6

11 Notice that the 5 μm mylar is extremely sensitive to tension and can easily get wrinkled. I started with what I did to the 100 μm mylar, but the same technique did not work in this case. Cutting became a problem again. The trimmer I use to cut 100 μm mylar could not cut 5 μm mylar because it is so thin and fragile. The edge after cut had nonuniform sharp edges and sometimes the blade could not even split the mylar successfully. Thus, I tried to find a better trimmer and ended up finding one similar to what I was going to buy as shown in Figure 7. Figure 7. Another trimmer I used. However, this trimmer did not work as well as I expected, so I had to use a scissor again. In fact, the scissor works better than the trimmer in cutting 5 μm mylar because it can easily slice through the mylar, but cutting in straight line became even more difficult because the 5 μm mylar is so light and fragile that when it s cut, the edges wraps a little bit. The ideal way to solve this issue is to stretch the mylar and exert uniform tension (as much as possible). I achieved this by putting two heavy objects on two sides of the stretched mylar to make sure the surface is flat and then cut it. Then for taping, I used the same technique, but I was more careful due to the fragile property of the material. After a few attempts, I get a sample shown in Figure 8. 7

12 Figure 8. A sample of 5 μm mylar shell The sample is relatively good when a slight stretching force is applied (It was wrinkled after I finished. (It s much nicer when it s just done). Another potential solution to this problem is using a laser cutter to cut the mylar. This idea was not tested because I was afraid of unexpected consequences. The exact composition of the material is unknown. An intuitive guess for this experiment is that the mylar would melt under high temperature or the heat would make it wrap, therefore leaving uneven edges. By this point, the production of mylar cylinder was not a big problem anymore. What needed to be solved next was to make the cylinder maintain its shape during the experiment. This could be difficult due to its wrinkle-prone nature. The first idea was to apply an external force, like stretch, which can be achieved by fixing the rings on the cylinder in the RTPC, but this approach will produce minor horizontal wrinkles. An idea I thought about to solve this issue was to add supporting rings every a 8

13 few cm inside the cylinder. However, this method was soon proved unworkable by simple test because it was difficult to operate. Adding more rings would make the cylinder heavy and thus add another force going down to cause more wrinkles. It could also possibly cause more nonuniform tension if operation is inappropriate. Another way is to apply super pressure inside the cylinder by pumping air through it during the experiment. To simulate this environment, I made a 40 cm long mylar cylinder and pulled the majority of the body out of the pipe and inserted a hair dryer into the cylinder on the other end as shown in Figure 9. Then, I turned the hair dryer on. Figure 9. Experiment using a hair dryer. The result was surprising. Even with the lowest power of the hair dryer, the mylar surface suddenly became smooth and could stay that way as long as the air is flowing through it. Some wrinkles near the left end were caused by improper alignment. This method is the most promising for all the cases tried. 9

14 2. GEM foil Next, I needed to make a shell out of the GEM foil as shown in Figure 10. The GEM foil that I was using mainly contains cooper and Kapton. The same technique that I had used for mylar would not work on the GEM foil due to its special texture. The surface of the foil has countless small holes that allow electrons to pass through. Any scratching between the foil and the pipe might fill the holes with tiny bits of dirt, thus potentially making it lose its permeability. Figure 10. A small piece of GEM foil. This problem can be solved by using a pipe that has a slippery surface. For example, a surface covered with thin gold would be a candidate. However, gold is not an affordable material for this project. Therefore, that proposal was rejected. An alternative method is to reduce the contact between the foil and any surface as much as possible during the process. The new idea is making a few circular rings which serve as the outer support to constrain the shape of the foil. Figure 11 below is a 2D view of the method. 10

15 Figure 11. A conceptual diagram for the GEM production. The rectangular-shaped foil is released inside the rings and naturally extends until its surface touches the ring. Glue may be applied beforehand on one edge so that two edges are glued together as the foil extends. Alternatively, tape can be used after the foil extends inside the rings. Then the rings can be opened, and a GEM shell is made without sliding. Due to the scarcity and expense of the GEM foils, I used another plastic material as shown in Figure 12 for these experiments. The plastic has similar malleability to the copper GEM foil but is slightly stiffer. Therefore, techniques that work for the plastic should also work for the GEM foil. Figure 12. The plastic material I used for the experiment. To produce a cylinder with certain volume, I needed to cut the plastic into certain size first. Here I used both a blade trimmer and a laser cutter as shown in Figure 12 to do the job. Both yielded good results. The edge cut by the laser cutter had a bit unevenness due to melting of plastic under high temperature while the edge cut by the blade was clear (such result is not guaranteed with real GEM foil and corresponding experiment needs to be done). However, the blade trimmer can only cut material of 11

16 small size, while the laser cutter has less constraint in terms of size. Figure 12. The laser cutter I used for cutting the plastic. Figure 13. The cut comparison. The upper edge is cut by the laser cutter. The lower edge is cut by a trimmer. Then I needed to make the ring. An effective way to create such special tool is using 3D printing with the Ultimaker 3D printer in Maker Space as shown in Figure 14. I created the model on Maya, which is made of a semi-circle and two handles, and then used the 3D printer in the maker space. Two pieces of this model can be held together by taping the handles to form a complete circle. When the cylindrical shell is made, I can remove the tape and get the shell out from the rings. However, during printing, there was a big problem. The model was made with two handles and one 12

17 semi-circle. The overlapped part of the semi-circle and the handles could not be correctly printed by the printer, because the printer could not correctly read that part as shown in Figure 15. Figure 14. The 3D printer I used in the maker space. Figure 15. A failed model of the half-ring. As a result, I had to change the model. The new product looks as Figure

18 Figure 16. The new model of the half-ring. With a pair of sample rings, I tested the method mentioned above on both GEM foil and plastic. One thing noticeable is that the natural extension of circular material does not apply enough force for the two edges, when closed, to closely touch each other. Therefore, during the process, I had to manually press the edges to make them tightly stick together. The products are shown in Figure 17, 18 and 19. Figure 17. A cylinder of GEM foil made with the ring. 14

19 Figure 18. A cylinder of plastic foil made with the ring. Figure 20. After the ring is removed, the shape still maintains perfectly. By comparison, the rigidness of the plastic made it maintain in a relatively good shape. On the other hand, the soft GEM foil potentially does not have this property, of which people need to care during the actual production GEM foil cylinder. Next, I wanted to test whether this method works for making longer cylinders. Due to the malfunction of both 3D printers in the maker space during the last few weeks of the experiment, I was unable to make more rings. Therefore, I found alternative tools. I used the apparatus that was used to fix the PVC pipe during my previous experiment, and I moved the stands closer to each other. The inner surface of 15

20 each stand is circular so that it works just like the ring. Figure 21. The experiment with longer cylinder. Figure 22. Another view. 16

21 . Figure 23. The long cylinder after rings are removed. The result was satisfying. The long cylinder made with this method was able to maintain a cylindrical shape. and the whole process of production does not involve any sliding of the material. The only problem was that the middle part of the cylinder was not tightly glued because I could not reach the middle section from the inside to press it down. However, it is imaginable that such minor problem can be easily fixed. One can use long rod and enter from either end to provide an outward force on the edges. The corresponding experiment for larger GEM foil was not done due to lack of material, but the success of experiment on plastic suggested that the desired consequence could be achieved on the GEM foil. One only needs to deal with their slightly different texture while making the cylinder. 17

22 Chapter 3 Conclusion and next step By the effort throughout the year, I have found a method to make relatively perfect 5 μm mylar cylinder and most importantly I have proved that by applying super pressure inside the mylar, the cylinder can keep relatively smooth. On the other hand, the cylinder of GEM foil requires a different method of production. My experiment suggests that by applying several rings around the foil while gluing it, one could get a perfect cylinder. Then to reduce scratch on the surface, the ring can be removed by open it into two. The experimental success on plastic potentially means that such method also works for GEM foil. The potential improvement for the research is to test this method with larger piece of GEM foil to see the actual product, which could be done when the material is enough in stock. 18

23 Bibliography [1] Zhao, Wenqing Wrinkleless Mylar Shell Prototype for BONuS Detector: Design and Procedures [2] Howard C. Fenker, Wrinkleless Mylar Shell Prototype for BONuS Detector: Design and Procedures. Nucl.Instrum.Meth. A592 (2008) [3] CLAS Collaboration (N. Baillie et al.) Measurement of the neutron F2 structure function via spectator tagging with CLAS Phys.Rev.Lett. 108 (2012)