Page 3 of 4

Re: Neutron detection moderator embedded activation!

Posted: Fri Nov 27, 2020 12:15 am
by Mark Rowley
Excellent results Richard. Now you have me wondering if there’s an inexpensive and easily obtainable rhodium alloy that could be used for activation.

Mark Rowley

Re: Neutron detection moderator embedded activation!

Posted: Fri Nov 27, 2020 12:44 am
by Richard Hull
Rhodium plating was often used in the past on cheap junk jewelry as it provided a had and bright overcoat that would not leave a green of colored remnant on the wearer's skin. Rhodium was often sold in solution by atomic weight for plating. Needless to say any rhodium alloy or chemical compound would be priced based on the market spot of the current $14,700 per troy ounce by its atomic percent weight.

Technical stuff

I will, in future, retain the 5 second count data point grabs recently replacing the original 10 second data points in the silver Xcel graphs. In future Rhodium graphs, only the first 300 seconds of decay will be grabbed at end of run counting. This means only 60 - 5 second count periods, instead of the old 120 - 5 second count counts with silver. This is due to the single Rh104 half-life of 44 seconds. (6 X 44) = 264 seconds. The extra time is to assure total decay. I will only have to change two lines of code in the Arduino Uno controller in my "count box" at fusor V. A real snap to change and blow into the Arduino via the USB port

Richard Hull

Re: Neutron detection moderator embedded activation!

Posted: Sun Nov 29, 2020 9:29 pm
by Richard Hull
Keeping fusor V busy and blasting rhodium up a bit. I attach the graph of today's run. Got over 100kcpm neutron counts.

Richard Hull

Re: Neutron detection moderator embedded activation!

Posted: Mon Nov 30, 2020 6:06 pm
by Richard Hull
I feel I must explain the red line in the graphs fully to those not so well heeled in statistics.

1. The Red line is the arithmetic mean, "average", of a full 10 minute long series of 5 sec readings from the un-activated metal probe. A total of 120 individual readings. This is often called "background".
2. To a statistics person, they want to know the standard deviation about the average or "mean" This is the square root of the mean or average value. It is represented by the Greek letter Sigma
3. A one sigma confidence level in the average of a data set. Tells the statistician that the range of values above or below the mean will all fall within plus or minus 67% of the mean or average value of the data set.

I have attached an excel graph of my latest background test, let us look at it. I had the Xcel spread sheet compute the mean by adding all the 120 data point counts to a large number and then divide it by 120. I got the result of the mean or average to be 6.5 counts every 5 second count. However the Sigma is the square root of the mean which is 2.4. Thus 67% of all background measurements will lie between about ~4 and ~9 counts for every 5 second count result. these limits are shown to lie between the two yellow lines.

You can see the span above and below the mean allows for a lot of what was actually measured to one Sigma confidence level. This is reflected in the jumpiness of the data over the entire graph.
This indicates 5 second counts in an already high background area is not really all that great, but does allow a highly visible, if jumpy, indication of the natural exponential decay of the activated Rhodium.

As an example, scroll to the post just above this post. I have added the 1 sigma yellow bars to the result of my 11/29/20 Rhodium activation results This give a better view of when the activation counting ended at about (60 on the graph), 5 second intervals totaling 300 seconds or about after 6 full half lives of the activated rhodium. I will endeavor to continue the one sigma bars in future.

Richard Hull

Re: Neutron detection moderator embedded activation!

Posted: Mon Nov 30, 2020 8:04 pm
by Richard Hull
Wow! This is a rolling post....I think I think too much after thinking and posting. (if you can decipher that sentence, you may need help)

Bill Kolb and I have been musing about my charts and graphs. I first noticed, during our discussions the intrinsic wild high count gyrations always seen during the first half life data grab be it Silver or Rhodium. These gyrations were way outside of the 5-6 six count wobble due to the one sigma background range. This is optically visible to the meanest intelligence. Bill mused over it as well. To put off any effort to deal with this to distraction, we left it to "it's what radiation does". It does random real well. It left me a bit uneasy, but we moved on to more pressing issues.

Alas, I was getting a soda after the above posting and a flash came to me. Beta recoil double counting!! I will explain below and in a diagram.

Beta recoil???

I have always loved this and include it in many radiation tutelages which I often I give. See diagram #1.... You can do this yourself. Grab a mica windowed GM counter and hopefully a beta only emitter Tl204 or Sr90 set up the scenario in the diagram and set two lead plates, ( I use tungsten), as shown. Remove plate #1 take a count, then put the plate back in place. Count again. Wow! cool, huh? Given betas of sufficient energy, they will recoil and create soft x-rays that will shine into the detector after impacting a super dense metal. This is beta recoil and bremsstrahlung.

Diagram #2 is the manner in which I think the huge peak variations occur during the first and part of the second half-life of these hot, short-lived, beta isotopes work to do this double counting of the same beta particle! It is partly due to the method of setup in my case.

The reasoning

1. Both silver110 and Rh104 have powerful peak beta energies, (2.87 meV and 2.44 meV respectively). As we know from the beta energy curves, they are not the bulk of the emitted betas but only a small fraction. Most beta energies gather around 0.3 to 0.6 the peak value.
2. Betas can easily travel through thin aluminum, like the shell of my Russian STS-5 tubes. Only the very weakest are lost to counting. Regardless, energy is lost as a beta electron moves through any solid matter
3. Acute angles of impact boost beta recoil as head on impacts with dense matter tend to just stop the beta particle....unless it penetrates the target at reduced energy.
4. From a materials science standpoint, both silver and rhodium are dense metals. These can both readily absorb and reflect beta particles.

I have tried to make the diagram and some of the ray trace keys fairly obvious. I have not covered every possible path, but more to show how double counting might take place.

The upshot is that the blisteringly hot betas that stand a chance of beta recoil, (few), are more likely to be numerous at the T zero in the decay scheme. As the activated material decays the number of hot betas naturally decrease. As the isotope "cools", the normal Sigma of background starts to become more prevalent. For double counting in the silver scenario the hot betas must leave the cylinder penetrate the GM aluminum make a count then re-penetrate the aluminum tube then recoil off the silver again and get through the aluminum again to trigger the second count.

The mass of active silver110 is only 1/2 the activity of the silver by mass. Not so for the Rhodium as 100% of it stands to be activated at its much higher cross section, so double counting is boosted in spite of the two possible unguarded escapes on the sides of the tube without a metal reflector of Rhodium.

I hope Bill likes this idea. Two heads are always better than one and that is a good thing. Anyone else's head have a better explanation?? Debate? Experimenting with radiation and numerous test instruments coupled with reading up on it a lot and often expanding on what you already possess is always to advantage.

Oh, as always, click on the diagram to enlarge it.

Richard Hull

Re: Neutron detection moderator embedded activation!

Posted: Tue Dec 01, 2020 4:15 am
by John Futter
Change the "beat gun " source to "Beta gun" source in your diagram

but well done

Re: Neutron detection moderator embedded activation!

Posted: Tue Dec 01, 2020 1:37 pm
by Richard Hull
Thanks, I often correct later in a clear moment of re-read. I changed "to" in the first sentence to "too" just now. I tend to type faster than I should. Sometimes I re-read posts I've done 3-10 years ago and find faux pas that need to be corrected. Mostly, it is the FAQs.

Richard Hull

Re: Neutron detection moderator embedded activation!

Posted: Wed Dec 02, 2020 10:32 pm
by Richard Hull
Attached is Bill Kolb's best fit reduction of my 11-27 Rhodium activation run. He noted the curve is virtually perfect based on the expected half-life decay. I ran two runs on the 27th and attach them here as well

Richard Hull

Re: Neutron detection moderator embedded activation!

Posted: Fri Dec 04, 2020 4:36 pm
by Richard Hull
Bill Kolb continues his work with me on his now famous best fit analysis of my Rhodium activation. I sent him some of my one sigma results and he has just started to look at this aspect of my activation results In this case, my 11/29/20 best run to date. I attach his results here.

Bill's long term complaint is that due to my high laboratory background average and its one sigma range, the data appears extremely granular to the eye. In his letter to me based on his reduction and one sigma limits, he is forced to admit that in spite of this visual granular nature he finds my results "tight" and accurate as every data point lies within the one sigma limits!! This is rarely seen in such reductions, making my setup capable of securing very reliable data. In Bill's own words In all cases, more than 98 percent of the data points fall within the 1 sigma dotted lines (1 std dev). As you have pointed out before, one standard deviation should contain 68% of the population. This suggests your measurement errors are small compared to the variance in background and Rh104 decay.

All of my Rhodium results are based on runs that never exceed 500- 600k n/s TIER over the last 3 minutes of a 15- 20 minute activation run! Thank you Rhodium!

For the sake of clarity I will now expose the more of less normal run on my uncooled fusor V on the specific run on this best one to date 11/29/20

This long elaboration is for newbies and older fusioneers who have never done activation to show the rather complete process needed to secure all of the activation information preceding this posting over this long thread of Silver and then Rhodium activation. My thanks go out to my long time friend Bill Kolb for his great effort on my behalf.

I get all the gear and pumps up to spec and the mostly "conditioned fusor V" is pinched off from 00 microns on the baratron gauge to 0.1 microns and pure D2 is sent in to a starting pressure of 12.3 microns, differentially pumped, (flowing D2 against the turbo pump working at the pinched or cracked valve.)

What follows are the data taken of over a full 20 minute run on 11/29/20... the numbers left to right...Pressure....KV applied.....mA of current....1 minute timed neutrons count, (CPM), at this level.

12.3 microns......33kv.......9ma......36,467cpm

There are thirteen one minute runs above. the other seven minutes are expended making adjustment to push the next timed count run to new heights. Pressure-voltage-current and new neutron count
In all those adjustment minutes the fusor continues to operate at some level between the last and next timed count. As the fusor is heating all the time, as more power is poured in, and rhodium needs about 3 half-lives at a stable level, we see that after the third or fourth run above the rhodium is decaying as fast as we are pouring on more fusion neutrons. As we keep adding more activation neutrons, The decaying rhodium is still kicked to ever high levels with each run over time. The effort is to grab the best 4 runs near the end when the fusor is at about 110 deg. C.

Power is then cut, and within 1 second, the series of 5 second counts begin. One can see that it is only the last 4 or 5, 1-minute runs, but especially the last one, that will determine the real best activation point of the entire run! Those early runs are effectively lost via decay over time. If you spend a little time noodling on this all will be made manifest. Nothing is to be gained by a long activation run only the last 4 minutes of activation will count. The long run is needed to hit the fusor's peak output, hopefully, more or less uniformly activating over the last 3 minutes.

I attach my personally generated Xcel file created right after the run. I dump the Arduino's 120, 5-second count periods into a text file, then blow this into Xcel on my little laptop to generate the graph. I submit this file should you wish to see the hard count columnar data and graph. (you will need a copy of Xcel to view this one.)

I also attach the jpeg image of Bill's best fit as mentioned above.

Richard Hull


Re: Neutron detection moderator embedded activation!

Posted: Mon Dec 07, 2020 8:45 pm
by Richard Hull
The beat goes on and I just did a super activation of rhodium. See attached.

The lab and all its components were at 32 deg F just now and I have been running the fusor regularly as you know. Coupled with the daily wall loading and a frozen SS fusor, I nursed the system to new heights. I have never broken 100 counts in 5 seconds on rhodium until today. I attach a photo of my "during run" data sheets. I use folded paper bits to record by-the-minute neutron counts. I refuse to constantly type these in as I did last posting. I get two days out of one folded sheet. All might do well to study 12/6 versus 12/7 on these sheets. The pressure at start and the counts at start are noted as they proceed. The ice cold 12/7 results speak for themselves. 12/6 the end run temp on the fusor was 121 deg C. 12/7 super run the end temp was 89 deg C. The results came up fast on the cold fusor making use of the wall loading. COOL your fusors! This is somewhat long known, (some years), but such a magnum impression required Jon and Mark to push the knife into the heart of the matter in their recent cooling efforts. Keep a cool head and a cool fusor and you will do good things and more fusion per unit watt expended.

Click on images to enlarge for old eyes.

Richard Hull