Page 20 - Demo

20Continued from page 18 Thoughts from Newmansuggested I write about seals. I thought, yeah, I can write all day about seal design. The problem is that it%u2019s impossible to do the subject any justice in a two-page rambling from a madman. But, now it%u2019s too late to change, as deadlines must be met, and I don%u2019t have another subject to share, so I will muscle through with some of the most basic stuff. Some important things about seals. Rubber compression is measured in pounds per square inch (psi) per a percentage of the compression. I think the technical term is the Compression Force Deflection (CFD) rating for the rubber. This basically means that the resulting gasket compression amount is the percent change in the gasket thickness caused by the pound-force applied to the total area of the seal. Don%u2019t worry, you probably won%u2019t be calculating this on a regular basis as the CFD data for a particular rubber is difficult to obtain. This number usually comes from empirically generated curves based on testing a type of rubber with a specific hardness. The only reason I bring all this up is to keep in mind the relationship between the contact area of the seal and the clamping force required to compress the seal. A wide flange and seal with a large seal area will require more clamping force than a smaller seal with a narrow flange and seal with less seal area to compress the same amount. Got it? That is why O-rings work so well.In the glovebox world, most applications are for flat seals, i.e., flanges, removable panels, doors, and windows. Over the years, we have come up with a few general %u201crules of thumb%u201d that most of us follow for creating leak-free joints. Things like the rubber durometer should be in the 30-40 durometer shore %u201cA%u201d range for a good seal. The bolt spacing for a flange should be on 3%u201d centers to provide a good helium leak-tight seal. If at all possible, cut gaskets from one piece of rubber and avoid bonded corners. Rigid thick flanges will require a flat machined face for a good seal, whereas formed thin flanges will bend slightly and conform to each other when bolted together, eliminating the requirement for a flat machined surface. You can find all of that buried somewhere in the AGS Guidelines and Standards. There are also very complicated seal applications. We had a project a while back that required a high velocity loop of helium cooling gas at 350 psi, to cool a target in an electron beam accelerator. The 3%u201d dia. stainless steel pipe had to pass through a high vacuum space maintained at 10-7 Torr vacuum, also in a high gamma radiation field with a 150%u00b0 C ambient temperature. The pipe loop was all welded except for a flanged joint that was required for component assembly. This flanged joint required a seal. Trust me, this was a challenging seal application. An elastomer O-ring could survive the environment but couldn%u2019t be used because it was in a high vacuum space, and our gas was helium. Why is that a problem? The helium gas molecules are small enough to actually permeate through the rubber at a rate high enough to contaminate the vacuum space. We were driven to metal seals for the solution. Conflat flanges would be normally used for an application like this with copper metal seals, but space was not available. We ended up choosing a silver-plated Inconel C-shape ring seal. It seemed to be the only option available. And, I%u2019m not going to tell you what a Conflat flange is either, ha ha, look it up. At the end of the day, it really is the exact same concept as our bottle seal, it%u2019s just the seal was made from metal. It was softer than the joining flange material and functioned to fill the gaps caused by the irregular surface of the flanges. Other considerations required to make this seal successful were flange thickness to prevent bending, flange material hardness, flange flatness, proper bolting to provide the required clamping force, very precise polishing, and a lot of luck. As always, the devil is in the details. Our seal design eventually worked, and with a few struggles, it passed the leak test. There was quite a learning curve on this one.The goal of the seal is always to pass the leak test required for the project, and depending on the test, it will drive the seal design. A helium leak-tight joint will require a much tighter seal because those pesky little helium molecules are so small. They can pass through much smaller holes than air or water molecules.Sometimes, though, it%u2019s not the seal that leaks. Many years ago, I had a project with some big vertical-acting doors on a large rectangular airlock for a beryllium powder sintering process in a large glovebox. When the glovebox was completed, the airlock wouldn%u2019t pass the helium leak test. Everyone was convinced that the doors were leaking. Finally, after days of working on the doors and seals, we discovered it wasn%u2019t the door at all. The polishers had gotten a bit overzealous and ground all the weld off of a weld joint right next to the door, causing a leak to be confused with a door seal leak. Things are not always as they seem.Sealing technology is a vast subject, with many variables to consider for a good seal design. Fortunately, there is a ton of information out there to help with any seal application you may come across. In a 3-minute web search, I was able to land on a free pdf, titled %u201cSeals and Sealing Handbook%u201d by Robert Flitney, which looked really handy. It only has 637 pages. It has always amazed me how a simple single concept can be used for so many different applications. Kind of like the wheel, right? So, I%u2019m thinking - Good luck and have fun with the %u201cArt of the Seal.%u201d And what did we really learn from all my rambling here? Put your finger on the damn cap before you shake the bottle of sauce. v