Fusor Forums

Here is the probe I'm using: http://www.flukebiomedical.com/biomedic ... ?pid=54801 (which does have a mica window)

New Run - Changed Measurement Equipment

5 minute run, 218k n/sec average
~0.5 gram Dysprosium, 15 cm distance to cathode
Data taken with RDA-3A beta scintillation detector and Ortec 4890 PreAmp/Amp/SCA

Peak CPM ~2x background average and 6+ standard deviations

It's unclear why the peak activity is delayed by 5 minutes. If this was a transition from Dy165m to Dy165 then the activity wouldn't have decayed as quickly after the peak (?). Potentially some other interaction with the setup is causing erroneous results.

Still getting odd decay numbers with the beta scintillation detector. The counts are clearly higher with neutron activated Dy than without, but the data is not intuitive. Does anyone have thoughts on this?

Related to my other thread on PoBe, I was able to activate dysprosium-164 to Dy-165 using 10 mCi of PoBe sources. The Dy was a baggie of 25g of Dy2O3, which I activated for about a day in my neutron oven and then taped around the GM tube of a GMC-320.

Since I'm working under different conditions than fusioneers are (much lower flux, but much longer activation and counting times letting me test higher half-life nuclides), I look for smaller deviations above average background counts over longer periods of time. I haven't tried finding Dy-165m and failed at finding anything statistically significant with silver, but I definitely got a detectable amount of Dy-165.

A t-test comparing the average of the GM counts from the first half-life against the background average (3.01 cpm above background) found an extremely significant increase at p < 10^-12, although the second and third half-lives weren't significantly above background (p=0.06 and 0.07 respectively). I defined the background as the average of everything from 5 half-lives until I stopped the test at about 12 half-lives of Dy-165 (total of 28 hours of counting). Only the first 7 are shown:



So, given my puny flux of something like 10k n/s, I think that the fusioneers here who are getting >250k n/s and who can run for more than a few minutes should easily be able to find Dy-165 in addition to Dy-165m. Then again, I'm not totally sure how far away the activated material can be in a fusor neutron oven - obviously the oven would have to be outside the fusor itself, and then there would be at least 10 cm or so of plastic on all sides. I get to place my target material directly on top of the PoBe source, which is a huge advantage given the inverse square law. How far away do you guys have to place your target material from the center of the fusor?

10-12CM is the norm from the central grid to the center of the neutron oven and material to be activated. The central grid is not the only source of neutrons in a fusor as the source is more difuse and the oven, itself, only incepts a fraction of the neutrons. A very complex situation. A neutron oven to intercept 100% of the fusor neutrons would have to enclose the entire device. ( A virtual impossibility due to expense and numerous mechanical issues.)

Richard Hull

Yeah, I realized that neutron emission from fusors is pretty much isotropic and that the oven can't cover the entire fusor or anything, but not that some of the fusion happens outside the grid. I hadn't thought about just how badly the inverse-square law hinders neutron activation with a fusor though.

Taking a fusor running at 250k n/s, and an activation target 10cm from the center of the grid (which I'll pretend is where all the fusion takes place), you would have 250000/(4*pi*(10 cm)^2) = 199 n/s/cm^2, whereas my targets are directly on top of the PoBe source, so let's say 10k n/s at an average distance of 1 cm would yield 796 n/cm^2/s.

Still, though, I'm kind of uncertain about how it all works out within the moderator given that both fusor and alpha-Be sources require something like 18 collisions with H atoms to get down into the thermal range, which presumably means that there are neutrons wandering around throughout the moderator and only some of them find their way to the target. So I'm not sure how a given fast neutron flux converts to a thermal neutron flux. I posted a related thread in New User Chat.

I have noted that in the average 6" spherical fusor you need to be producing about 500k n/s iso to do a good amount of easy activation. One million n/s iso will be necessary for advanced activation. This assusmes an eight inch cube of HDPE as a neutron oven placed in virtual contact with the fusor shell. Naturally, an assiduous, dedicated researcher with a fusor can get by with lower numbers than noted above.

Richard Hull

It seems short term activation produces erratic counts (but statistically significant compared to background to a high confidence factor).

I decided to target Dy-165 activation by repeating the cycle of running for 4 minutes at ~300k n/s and cooling down for 10 minutes a total of 5 times (60 mins total time). This strategy has produces intuitive results with persistent activity.

The size of Garrett's fusor is rather small compared to most fusors and he is intercepting far more neutrons in the moderator than others at 300k n/s iso. The few extra cm, closer to a larger oven does make a difference. In fusor activation, it is all about how many fast neutrons you can put into the largest moderator possible.

Richard Hull