A Better Self Centering Technique for Sergent Couplers

I have proposed several solutions for self centering but I am always striving to improve. I admire the Kadee "whisker coupler" technique and wondered if it could be adapted. I grabbed my last #158, took a pair of pliers and yanked out the whiskers. They are crimped to the shank (more on that later). I glued the whiskers onto a Sergent narrow shank using CA in the same orientation as on the Kadees. I installed the result into an Accumate box and ... Viola! It works! Perfect repeatable centering, unrestricted lateral movement and importantly, very little force required to move the shank from side to side. The freedom of movement is important because that can sometimes be the difference between a successful coupling or not. If there is slight misalignment such that an open knuckle touches a closed knuckle the two couplers can sometimes slide apart to allow the join rather than pushing the open knuckle closed. This is more likely to occur if there is little resistance to sideways motion.

But that was my last #158. What to use for more whisker material? I got suggestions from another forum that the material was either carbon fiber or stainless steel and that it was .004" diameter. I searched fruitlessly for those materials. But the other day when rooting through my stash of railroad things I ran across a long tube that contained a (nearly) invisible strand of something. The label said "superelastic Nitinol .008". A lightbulb appeared above my head. This Nitinol wire is straight, stays straight, is super springy and is almost indestructable. 

I took out the (nearly) invisible wire and cut two pieces from it each 8mm long. I glued them on like the whiskers before, Tried it again in a box and... Huzzah! It works! These whiskers are a little stiffer than the Kadee ones because these are twice the diameter. They work but I'm going to acquire some .004 or .005 going forward. I saw a price of $8 for a 36" piece. You can get about 30 cars out of such a length. The technique works equally well using a standard shank in a wide box. The whiskers should be a little longer for the standard shank.

There is a short learning curve involved. I apply medium Starbond CA using a toothpick to the vise held coupler. Then using a fine tweezers I lay the wire in place with one end flush with the end of the shank. I carefully center and align it on the shank. Then I hold it down with a knife blade while detaching the tweezers. It could be squirted with accelerant at that point. Or just withdraw the blade and leave it alone for a while. The wire needs to be in solid contact with the shank for the entire contact length to assure that the angle of the wire is correct i.e. so that it matches the shank angle exactly. It can be pushed down with the knife blade without disturbing the positioning. This technique is non-destructive and reversible. See the pictures below.

I'm thinking that it might be advisable to bias the shanks slightly off center in the direction of the knuckle. It seems logical that this might decrease the chances of fouling the knuckle but further testing is required. If the conclusion is correct it should be possible to adjust the neutral position using a styrene shim between one whisker and the box wall.

Not that it matters but I'm pretty sure that Kadee is using this material. It looks like it and acts like it. Further the Kadee whiskers are crimped in place. That would be consistent with this conclusion since you would not want to apply heat to Nitinol in the manufacturing process.

That was using some .008 Nitinol super-elastic wire that I had on hand. Later I ordered some .005 wire because I think that is what Kadee is using. I converted several narrow shank couplers using that material. They work exactly like the Kadee 158 that I had previous experience with. I got reliable centering with very light resistance to sideways movement. But...there's always a but.

I noticed that the resting position was consistently not quite centered but rather very slightly to the left. If anything the shank should be slightly to the right in my opinion since I believe that will reduce the chance for a clash between an open knuckle and a closed one on the opposite car. I had proposed previously that a thin piece of styrene could be glued to the box wall to provide a little nudge in the opposite direction. However I remembered that I had all of that .008 lying there, now with no future for it. So I reasoned what if I put .008 on the weak side and .005 on the other side. The .008, being slightly stiffer should tend to move the shank away from its side. I tried it. It worked. In the picture below you can see that the shank on the coupler on the right is just slightly moved farther away. It is probably at center or slightly to the right of center. 

So this is one way to precisely manipulate the resting position of the shank. I'm thinking that a little stronger wire could be used here because I think the resting position should not be exactly centered but a little bit to the right. So maybe .010 or .012 wire might work. I don't have either on hand. 

Subsequently I have found that a lateral bias equivalent to substituting .008 wire for .005 can be obtained by installing two .005 wires on one side. Turns out it is just as effective and results in the maximum amount of displacement obtainable using this method. It's probably the maximum amount of displacement that you should strive for as well.