Moderator for small neutron detectors

[Doug Coulter]

Some of us have gotten some of the neutron detector tubes discussed here. I'm playing guinea pig and working on getting them going, but all my moderators are "busy" right now, so I needed another. The roughly correct dimensions for a moderator to get a high percentage of neutrons to thermal without losing them is about 7", which is the size of the Eberline Bonner Ball in that case used with bf3 tubes. But close is close, and I had some 6.5" diameter HDPE, which unfortunately was also UMHW, making it rather a task to do much with. The only way to cut that stuff is to melt it, more or less, with speeds and feeds beyond what many tools can manage. Twist drills screw into it instantly like a tap would, and you then lose both the drill and the piece. When drilling, you have to pull the drill out very often, or the molten plastic will stick it in there -- same result. And it's just a little too large to fit in my lathe without a chuck change I've been avoiding.

I discovered a modern Stanley spade bit goes through this like the wind -- as long as you pull the drill out every half inch, clear the chips, cool it off, repeat. Drilling and tapping for 4-40 screws to hold things on there was actually much harder as all I had in that size was twist drills -- and they tend to self-feed, cool, and stick. Tapping was easy, though. Due to experience with our 3He tube picking up EMI, apparently magnetic induction into the tube somehow, and because I had to cut the piece off with a chainsaw and got a sloppy cut -- I decided to go ahead and build this with a ground-able very low resistance shield with more or less full coverage. I used copper flashing for the top and bottom (so I can solder to it and mount a preamp box right over the tube), and Al flashing for the sides, as it's much cheaper and easier to work with. Neither will have much effect on neutrons.

Partially finished, with drill bit that really works shown.

Here's the parts, partway done, with that drill that really does work on this stuff -- at least if brand new and very sharp.

Almost finished

Here it is almost done. I'll bend up some Cu flashing for a preamp box to screw to the top and hold the preamp and HV input stuff. Since this pic was taken, I've niced it up a little by bashing the copper top and bottom to round them over where they stuck out a little bit. Nicely hides my lousy chain saw work.

Hopefully, I can get a decent test in before the end of the weekend -- I have one other major project I'm working on too -- an 8 core 3 ghz i7 computer, 12 gigs ram, 1 tb mag disk, 40gb SSD, 12 gigs ram -- doing the "build the computer" thing too. And oh -- it has a telsa card (NVidia) for doing very fast double precision floating point -- for simulations and neural net training tasks. I guess I just like hot rods.

[chrismb]

I discovered a modern Stanley spade bit goes through this like the wind

I should've mentioned this before. Having worked with 'hard' plastics for a while (because I have been using a lot of teflon and peek in my device) this is what I figured out too. Basically, use tools designed for 'wood', not metal. Chisel drills and the like. In my humble experience, you can only get up to about an M5 drill with hard plastics before having to move on to wood tools.

Threading hard plastic cores is also an a nightmare of amusing dimensions. The tap tends to undergo an oscillatory radial displacement, first cutting too deep on one side of the piece, then moving around to the other, developing a spiral of variable depth down the piece. The only way I have found to get around this is to always use a die thread a few sizes too big for the core. OK, so you end up with a trapezoidal thread but it's better than what you get otherwise - which is usually that the piece tends to plascially deform and twist once you've got just enough in the die to jam in there solid and render it unrecoverable!! I guess the inverse is true with taps as well - try a slightly smaller tap (or make a bigger hole) than is intented, but as Doug says, this is usually a breeze.

I'll leave you to establish what are or are not useful comments in the above by your own experimentation!!!... I should also mention - anticipate that such work will damage your wood tools, so use cheap ones!!!

[chrismb]

[[Incidentally, one other use for chisel drills is to make holes in aluminium enclosures (<3mm thick). Sometimes you want to make a big hole for a gland, or something - and you just don't have a big enough machine drill. Try a wood chisel drill instead! Make a 5mm centre hole with a machine drill to centre the spike on the chisel drill first. Then it goes through like butter after that, but make sure you don't let the drill get too hot. Also, it makes a hole a mm or two bigger than the actual chisel drill size.]]

[Doug Coulter]

I was only tapping this one 4-40 (English measurement the English no longer use, but think tiny) so it was easy. I used the tap fixture described in shop tips, works great. The real difficulty there was drilling the hole, for which I used the standard size for that tap (#42 drill). A twist drill tends to just bite and screw right in in 3-4 turns in this stuff, then slip on the chuck, which lets it cool and get hardened in there -- you have to be real fast on the reverse to get it back out of there or lose it. It took several "pecks" on each hole to get an open hole without plastic shavings still in there so I could tap it. In this case, tapping probably wasn't really needed except to start the screws more easily -- as Jerry has pointed out, simple displacement of this gooey stuff works fine and the screw will do that by itself. These screws are not really stressed much, just a place to put a ground lug under and somewhat nicer than using duct tape.

This one was fairly easy to bore as these things go. That's because it was only a 3/4" hole, and not so deep. "Gun drilling" anything -- say a 2 foot deep hole, brings other issues, mostly with chip removal and in this case, not having things cool and harden while in there. To drill a big hole, like twice that size as I had to do for the big B10 tube 20" deep, no hand drill had enough power to do it, even my 1/2" Makita that will break your wrists and lift you off the ground if you stick it in something.
The drill press has a limited quill travel, so it was frantic running of the table up and down to get the chips out in time. For that I used a wood type forstner bit....not recommended, but it did work "at all". After that my table cranking arm hurt for days.

I've had little trouble with PEEK or teflon here. Peek is stiff enough to machine like a free machining metal. The main issue with teflon is it's so soft that flex in the lathe is a big issue. But both cut nicely clean compared to UHMW HDPE, which machines like fast hardening glue....I suppose if you went slow enough it wouldn't melt, but that would be in the single digit RPM range for that stuff. Harder to do non-melting than plexiglass by far.

Next time I need to make a big one, I'll use another plan...but I have this piece of the stuff that cost me $300 for 6.5" by 36" to use up first.

You evidently really do want HDPE -- the issue is how many H's per cubic inch you can cram in there to make the thing small so there's less neutrons escaping, but you still are slowing them down. Plastics with halogens in them have their own cross section....there's just not much else to choose from here -- CarlW was right.

I've actually found that this stuff doesn't dull tools as much as wood. Other plastics, especially if they have "fillers" will, though. PVC is bad that way.

[chrismb]

I think you might find that 500 HDPE has the highest H per cc of any material. (...Slightly more than UHMWPE, in fact.) - well, certainly any 'common materials' - maybe perhaps solid H or some wierd boron hydride might beat it?

[Doug Coulter]

Yes, it seems to be the current best, and with few prospects in chemistry to do better. To pack in more H (or other light thing), you'd need high valence other stuff -- higher than carbon, and most of that has an absorption cross section of its own. Alternatively, something that makes smaller molecules (binds tighter) but again -- that would be halogens, which all eat neutrons. This is one of the more frustrating games, though I won't go to "you can't win, you can't even play" quite yet. I don't know the density of frozen D2 (or better, maybe, HD), but that's just too much hassle anyway.

I do note as a philosophical aside, that many people out there spouting "there's no shortages in the world that technology can't fix for us, it always has" are wrong. The periodic table is fully explored, secondary batteries already have energy densities approaching high explosives, we've run out of elements that can do more, the rules are what they are. At the limit, you couldn't genetically engineer crops to be more than 100% solar efficient, or need less water than is delivered in the product -- far past where we are now, but definitely a hard limit we will hit.

I've been frustrated that though you start with more or less unidirectional, narrow energy bandwidth neutrons, there appears to be no way to "downshift" like in a superhetrodyne (conceptually speaking here). Heck, you can't even decently reflect or focus them till the energy gets way low -- a few ev at most, where you can do the diffraction off a crystal thing. JohnF posted a neutron focus device for low energies that seems cool, but it won't work for this. I've seen some tricks with magnetic lenses (check the library) for imaging "at all" but....Still I think not worth the hassle.

It sure would advance the state of several arts to be able to make a neutron beam by means other than just stopping all the ones not already going the right way. It's something I think about, hoping to stumble on some slick trick that would help us, but so far, nope.

I do wonder, in re neutron detection in general, if a big blob of moderator would produce enough capture gammas to make it possible to do it that way -- just set a high threshold for their energy in an embedded gamma detector, and shield the whole thing in lead....Some things make pretty hot gammas when they eat a neutron, including H - 2.22 MeV.