Summary of Glue Tests 1993

Abstract

I have reported most of the results of my adhesive testing to members of the VLPC design team at one time or another, usually verbally, but I am wnnng this summary as an easy reference to the results I obtained. The adhesives I tested were for two primary purposes. The first was adhering optical fibers to Torlon 7130; the other was for securing an aluminum nitride substrate to the same material. I have not had access to a scanning electron microscope and someone with the knowledge to determine actual failure mechanisms, so the deductions I have made about why some adhesives have worked well at low temperatures for some purposes and not for other applications while a different material never worked and another always worked are partially speculation. They should be taken merely at face value with no particular results 'carved in stone' so to speak. The first aspect of my testing was adhesion of optical fiber to torlon. Knowing that this is a very important joint, I tested a variety of glues of two primary types: acrylic and W cure. W cure adhesives are known to possess reasonably good properties at low temperatures and are quite convenient to use as long as a W source is available. The W cure adhesives I tested were: Loctite Utak 376 and also 7EN484(?), Master Bond 1 Component W 15-7, and Norland optical adhesive 61. I found them quite easy to use, and they were packaged in a way in which they were not likely to cause a mess. Lab 6 e Perimenters generally used the Loctite 376 optical cure adhesive in their research into connecting scintillating fibers to the standard type. The acrylics I tested were Loctite Speed Bonder 324 and Permabond Quick Bond 610. These worked reasonably well, but they require a considerably longer set time than the W cure adhesives and are more complicated to use. (5 minutes set time or so for the acrylics versus about 30 seconds for the W. The Loctite must have the activator applied about 5 minutes prior to the adhesive application and the Permabond must be mixed adequately.) I also used a cyanoacrylate ester (superglue type) adhesive which appeared to function adequately in this test, but I would not recommend it for extended use, and I am certain neither would anyone else. I would highly recommend using a W curing adhesive for this purpose if the adhesives and the W treatment can be determined to cause no damage to the fibers. There is no apparent physical damage, but transmission could potentially be damaged. The final optical fiber to torlon test that I did involved testing to see if individual fibers could hold the weight of the entire VlPC copper isotherm in the event that a small number of fibers shrink more than the others as the cryostat is cooled down. While this test was primarily for the purpose of testing the fibers themselves, I constructed a new sample to avoid breaking the others that had already been finished. The adhesive I used for this test was 3M 3535 BIA two part urethane adhesive. I had no problems whatsoever with this product, but like the other two part adhesives, it is considerably less convenient and more messy than the W curing adhesives. The short pot life of this adhesive was also a reason to avoid urethane, since mixing would be required frequently. The other portion of the adhesive testing for the VLPC that I performed was the adhesion of the substrate to the torlon used as a carrier. This bond is extremely small in practice, and I could not completely simulate the size and likely construction methods. I used larger pieces than those that will be bonded, but the primary goal of these tests was to test the performance of the adhesives. These tests contained harsher conditions of temperature changes and loadings than the substrates are likely to meet. By lightly loading the substrates before and after the cooling, it is possible to see if the cold temperatures had any effect on the adhesive-torlon interface, the adhesivesubstrate interface, or the adhesive itself. I tested about 10 adhesives intensely. In addition to these tests I also talked to people with about 20 different adhesive manufactures (maybe more) and Jay Hoffman, an adhesive expert here at the Lab. The general consensus from the. people that I spoke with is that the temperature range near liquid helium is very harsh and many adhesives that set quickly are not able to withstand these temperatures. Only one of the companies that I spoke with had any experience with temperatures that low (they made adhesives for space applications). I will go over each adhesive used in my testing. In addition, I will describe and speculate on the cause of any failures.