Shields up!

For Farnsworth type designs.

Re: Shields up!

Postby Doug Coulter » Sun Jul 13, 2014 8:16 pm

Just a little note and a word of warning here - I'll have more cool results to show soon. I had been using copious amounts of cerro-safe here for X ray shielding. It was a VERY BAD MOVE. Didn't realize it at the time, since there are a variety of alloys that go by that name, but our version had 12% cadmium, which emits capture gammas when hit by neutrons that even several inch thick layers of it don't stop. So that was one issue - our neutron output, as it went up, was making these relatively huge capture gammas that go through inches of lead like it wasn't there...so, DON'T DO THAT. It should be lots better now that a few hundred pounds of that stuff have "left the building", but we won't test again till the rest is finished too. 100 msv in 30 seconds will make a believer out of one...in the weeks following when you don't feel quite right for awhile.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Shields up!

Postby Donovan Ready » Sun Jul 13, 2014 10:13 pm

At this point I might think of digging a trench. Let mother earth cradle you..
Donovan Ready
 
Posts: 239
Joined: Thu Apr 17, 2014 1:22 pm
Location: Austin, Texas

Re: Shields up!

Postby Doug Coulter » Sun Jul 13, 2014 10:41 pm

It's actually been a topic around here for awhile, Donovan. I have these really neato rock outcroppings on my land nearby that only need a wall at the "cave mouth" to have a little room with 30 feet of rock between me and it, in other words, the trench is dug, but it needs a roof. But at this point, even my best optics aren't as good as "being there" - and putting them closer in so as to get them good defeats itself - the X rays make digital optics just "white out" - and it seems I've made most progress by having mark 1 eyeball feedback so far. Not to mention, these outcroppings are on the side of what amounts to a cliff - it would not be fun to move heavy things down there to build the cave-mouth wall to keep the elements off the gear well enough to let it live (and not get full of not only condenstation but also, insects and stuff), not to mention the wiring required - latency and RFI issues mean, no internet or wifi is going to cut it for controls by itself.

Check the pic here - and blow it up. viewtopic.php?f=40&t=817#p4979 - that's not what my eyes saw or even close. I saw all 8 beams, one white dot in the center end on...and if you look close, though that's behind a lot of shielding (3/4" boro and 1/2" lead glass, a screen and so forth) - note the white pixels that were set by stray X rays and gammas. Cameras just don't get it. I need perhaps a telescope/camera combo, with more dynamic range than the usual several hundred buck digital camera.

It's definitely something to consider, and after 1-3 more advances, if they're not too major (otherwise it's next if they are) we make one we can move (this one's too heavy and has way more features than it turns out we needed anyway, but did make it easy to just try stuff at random) - we go to the cave, fer sure. No point if I don't live long enough to see this work! I am already working with writing software and working uP hardware to remote most if not all of this, but it's taking time to get done along with just you know, living and stuff. Recently it's been all data acq, and I've still got a little to go there - the scope has to be on the "hot side" of the shield, and me on the other, so how to adjust it? For example. Some other minor/trivial things, but they have to prove out 100% or close before we make that move to a simpler, more portable system that'll have to be tested at least a little "up here" before they go down in the pit. It's a case of limited human resources here - we otherwise have just about everything I could wish for.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Shields up!

Postby Doug Coulter » Thu Jul 24, 2014 11:04 am

Ok, still a lot to do. I used to labor under the mis-impression that X or gamma ray photons worked similar to light (since after all, they're photons). Well, that's not quite the case, due to the sheer energy being far higher per photon than chemical bonding kinds of energies (thousands or 100 thousands volts versus 1-3 volts). When you put out an X ray "beam" well, it's not like a flashlight - it's more like a flashlight in fog. Light goes "everywhere", due to compton scattering (and other things). Unlike visible light, this effect does lower the energy of the photons some - knocking out electrons from air and other atoms does that due to the higher energy photon's ability to do it vs visible light - but not enough to make something really unsafe into something safe.

Not knowing that is my excuse for having done a relatively sloppy job on parts of this, and I'm sticking by it. But now I know, so there's no excuse.

Here's one side of the fusor. I got that fairly well - there are leaks that aren't obvious in the picture (like out through the end of some feedthrough) but in general, this side is pretty tight (I'll do more anyway).
FusorLeft.JPG
Left side of fusor when facing the front from the op position.

I should note that most of these SS flanges are so thick they might as well be shielded with lead. SS isn't as good per cm, but I'm not using inch thick lead, either.
One obvious leak here is that little piece of coax with the probe on it. Yes, that's enough to make a difference. But this side is no longer the main offender.
You can see the recently almost-done neutron oven in this pic. I still need to attach the bottom part more firmly to the tank, so it doesn't lift off when I pull the top off. It's painted black because that's graphite paint to lube things a bit, but it's still on the sticky side. I'll probably solder on some tabs to the copper part and put baling wire (yes!) under the thing to hold it down.

But the main priority now is this side - I cheesed out and left major areas of the thin tank metal completely alone, as it's real work to handle all the warts and bumps well. Further, the tubes that stick out tend to leak pretty badly themselves. So it's not just the sides of the tank, it's those pipes too. Lots of fiddly work required here - and to make matters more fun, it's quite a tight fit between this side and my big lathe and mill.
FusorRight.JPG
The bad side.

The partly shielded HV FT you can see on the upper left of the tank is one of our faraday probes. This FT failed for HV - internal arc tracks, but is fine for this use, other than yes, ceramic slows down X rays pretty close to "none at all". And the X rays come right out of the end as well. The yellow box on the upper right is our primary neutron detector - while I DID use the box from an old CD ionization chamber/detector (you know, the kind that only read off zero after you're dead), it's just a box. The real detector is a hornyak button from Eljen, and a photomultiplier with our fine-wine preamp, running on its own wall wart and integrated, regulated power supply. It's by far the most stable detector in our, well, stable of various sorts of neutron detectors. We have a 3He one, a B10 one, some BF3's...none are as good for this. We don't need the sensitivity of the others at this point, we need accuracy and reliability, and the hornyak gives us that. Along with count rates that don't swamp preamps and counters...
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Shields up!

Postby Doug Coulter » Sat Jul 26, 2014 3:01 pm

Got some more done, but I also found out a depressing fact - one tiny leak is enough to scatter all over the room and raise the counts back at the op position. I say tiny in physical size, not in amount leaked, as a 1/4" inch hole in the "money zone" will hard-peg my survey meter at 3 feet set on it's numb-est setting (x100). It normally reads around half-scale on background at x1 for reference. This is with the fusor running "at idle" and about 3million neuts/second. I didn't stand around long with the geiger counter when I saw that. More fiddly bits to do, and this is getting really truly fiddly. But the lead works, you just gotta get it everywhere and with a bit of overlap. The fact that things are odd shaped and hard to get at doesn't help, but I just gotta do it. It seems that curing a few leaks but not all really doesn't do a lot for the stuff at the operator position - any single big one dominates, so it's gotta be perfect (no partly pregnant). I've been making some other changes, documented on other threads like "shields up" as well. I've just replace the old core duo quad core machine with my older of 2 intel NUCs. Wow, the new computer is blindingly faster despite worse official numbers, and nicely tiny and silent - as well as, I'll be real glad to get a couple sq feet of floor back.
NewDetectorAndFanLocation.JPG
Had to move some stuff...a new modified APAP machine now cools the bottom half, and the new neutron oven helps the top half, but there are leaks around it. Not trivial to fix.


Looks like I'll be pounding lead into crazy shapes for awhile longer. Leaks around the neutron oven are a special problem, since it has to move to be useful (you gotta take off the top to get to the sample).
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Shields up!

Postby solar_dave » Sun Jul 27, 2014 6:39 pm

Have you considered doing stuff like the old timey body guys did and just hot lead stuff into place?
Dave Shiels

My TED 5000 power monitoring
http://phx-solar.no-ip.info:8081/Footprints.html
solar_dave
 
Posts: 108
Joined: Tue Aug 21, 2012 12:33 pm
Location: Phoenix, AZ

Re: Shields up!

Postby Donovan Ready » Mon Jul 28, 2014 11:31 am

Hell yeah! Bondo Bob meets the flux capacitor! Image
Donovan Ready
 
Posts: 239
Joined: Thu Apr 17, 2014 1:22 pm
Location: Austin, Texas

Re: Shields up!

Postby Doug Coulter » Fri Aug 08, 2014 3:36 pm

Not far off...this is pretty kludgy work, but it needs to work right, even if I leave it on the ugly side. You'd think (and I wish) I'd be done by now, but not even close, way too many fiddly bits to do in an afternoon, and they all matter if I want to pursue that breakthrough I had - 2800x increase in output. We were ready for around 5x...it's a big deal, and I hate waiting, but I don't want to die either. A factor of 560 improvement is needed - on something that was already pretty decent...not trivial. That said, here's some of the latest progress. Bill came down and ran the fusor for me (at what is now our "idling range" of 3-6 million neutrons/second) while I ran around with a survey geiger counter to find the little leaks that are stil big enough to be problematic. EG >200 times the background at a foot from them for a few of them. It's like noise - you fix one, now you can hear another. Great news capitan - we've fixed the foot diameter hole in the boat. Now there's only 1000 1/4" holes to go...
Pix:
RightSide.JPG
Where I'm working now
I put some arrows on this one where leaks still exist. This stuff goes right through tiny cracks, ceramic insulators (lower right), and any place I didn't cover the stainless with at least 1/2" overlap. I had been under the (duh) misimpression that X rays worked kinda like light (it's all photons). But it's more like shining a flashlight into smoke. Sure, everytime they scatter off an air or other molecule, they lose 70ev or so - meaningless when they are at 50 kV or 2 MeV. They go around corners all too well. In this pic, I've made thick lead cans to cover the pressure gage (lower right) and angled feedthrough (uppright with blue wire) which is a faraday probe. Both were surprisingly bad leaks, but there are still little gaps here and there - so I go to cutting up bits of lead and sticking them on. There are related issues. Now that I have the op position in a relative "cone of silence", I can't read my gage for the system pressure (and related vacuum pump stuff). So I need a light there.
InvisibleGage.JPG
I can't see this in real life any better than the camera (look just above the mouse).

More fiddly stuff and wiring for that.

The backside is currently performing better than it looks like it should. I did a lot of work around the new neutron oven (top lifts off via rope and pulley arrangement), and I've added some lead and other stuff to make it stay cool better. I've not yet shielded the NUC which is on the back of the monitor (and from which I'm posting) and am trying to figure out how to do it without wiping out the wifi signal I usually use so as to prevent the odd lightning bolt from propagating throughout my network. I have enough blown switch ports as is, and have let the magic smoke out of a few mobos.
Back.JPG
The back side. The long FT tube is normally covered by an 8" PVC pipe with outer copper screening for EMI reduction. I'm adding lead foil to that too.


It stinks that it's taking this long to get safe to pursue a huge breakthrough first tested on May 1,2014, which gave 2800x (estimated) with no tuning whatever. That's getting us real close to actual net power gain after conversion via steam - not that far to go, maybe just tuning. But...the way it was, running for a minute would give me half a Sievert of radiation...and I'd not survive that. Sooo....we do this boring crap awhile. Dieing from rad poisioning is even worse than boredom.

And, Dave, yes, I'm doing a version of leading. But those guys had steel as a substrate and didn't have to worry about melting the other lead. If I work fast and precisely enough, I can use low-melt solder to stick the lead on (before the tin migrates and it all melts and slumps). It's a bit more challenging than that job. Bill brought us some expoxy paste and shot shell shot I'm going to mix up and stuff into the "can't get there" kind of spots as kind of the same idea - and Donovan, there's your bondo.
.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Re: Shields up!

Postby solar_dave » Sat Aug 09, 2014 1:03 pm

Yeah the body shop guys used a linotype alloy if I remember correctly.
Dave Shiels

My TED 5000 power monitoring
http://phx-solar.no-ip.info:8081/Footprints.html
solar_dave
 
Posts: 108
Joined: Tue Aug 21, 2012 12:33 pm
Location: Phoenix, AZ

Re: Shields up!

Postby Doug Coulter » Sat Aug 09, 2014 1:48 pm

Not sure what they used - it was real "white" like lino (which I use here for rifle bullets in .308 and 6.5mm Swede), but at the very least it was some non-eutectic alloy, which at application temps was about like a slushie - a bunch of crystals of whatever (lead most likely but could have been something else like antimony) that had crystalized out of the mix leaving a eutectic liquid behind. So it was like working mud, sort of.

Here there are some concerns about things getting activated (eg made radioactive) or having huge capture gamma ray energies (cadmium for example). Hydrogen actually has one of the biggest gammas (about 2.2 MeV) but it has a low cross section and doesn't capture many neutrons, just slows 'em down to the point other things get them. I found this out the hard way (it's in the books, but I wasn't reading the right ones). Things I thought should help did indeed stop the gammas from the fusor, but in turn emitted much higher energy ones...which they wouldn't stop.
630px-Pb-gamma-xs.svg.png
Lead effectiveness vs energy - not quite what you'd expect


The upshot is that while even a really thin layer (say 1-2mm) of lead easily stops almost all the 50keV X rays from the fusor, its absorbtion drops off rapidly with the energy of the photon. So if you double (or more) the energy, you need more than a proportional increase in lead thickness to get the same attenuation. Note the minimum right around 2 MeV...easily a factor of 10 worse than what it is at some tens of kv. That kind of makes this more engineering than brute force, because I can't build foot thick lead walls and expect the floor to hold them (already having issues there). So a big focus is to prevent those really high energy ones being generated in the first place. "An ounce of prevention". We might wind up coating the tank innards with a light element such as Al just to make sure that any fast protons etc can't make super high energy gammas in the first place. Neutrons are another whole story. If there's nothing around that moderates them well - they make it about 30 feet before reaching thermal energies where they get captured and make capture gammas. In that case, I have the square law working in my favor, at least some. It's a lot of things to trade off, not a simple issue, though friend and board member John Futter mentions that half a meter of concrete stops all the junk from his 2 MeV accelerator just fine...not that my floor will hold that either. We may, in the end, just build another building (storage shed) someplace a good bit away, and run remotely if this new breakthrough pans out - and watch via telescope and normal data acquisition.
We'll just have to see. With this batch of improvements (which include faster and better data acq), I can go to shorter runs as well, and keep the integrated dose in reason for a bit, or so goes the hope.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

PreviousNext

Return to Fusors, Farnsworth type

Who is online

Users browsing this forum: No registered users and 1 guest

cron