Neutron Attenuation Discrepency

[Joshua Guertler]

Greetings,

I have recently been testing a series of potential shields under conditions comparable to what may be expected for a fusor. I am relatively new to the various intricacies surrounding nuclear physics, thus, there may be a few misconceptions that will hopefully surface soon. Hopefully, by detailing the steps that I gone through, I will be able to adequately explain the source of my error.

PART 1:

As some background, I placed the various shields (comprised of a different moderator and neutron absorbing material, such as concrete) in front of a beam port opening in the side of a small experimental reactor as a means of logging data surrounding their ability to shield neutrons. results of the detection incidents had been recorded in CPS. This rate in CPS was converted to neutrons per second via dividing the CPS recorded by the detector's efficiency (roughly 1% for "cold" to 4 MeV).

PART 2:

Since the reactor's output of neutrons per second was relatively inconsistent, the background radiation (or initial intensity) fluctuated, thus, making an "apples-to-apples" comparison for detection results quite difficult. In order to compensate for this, I utilized linear neutron attenuation as a means of simplifying the matter. I used the equation below to solve for the neutron attenuation coefficient:

I = I0^(-μx)

Where:

I = CPS recorded by detector behind shield
I0 = CPS recorded as background (from reactor, no shield in place)
μ = Linear neutron attenuation coefficient
x = Thickness of target shield (cm)

In this case, the "I" vale was just the radiation that had seemed to pass through the shield where as "I0" represents conditions where the shield was not in place altogether. The result of this had been attenuation coefficients in the area of 0.0876... and 0.079734. Do these quantities seem correct, or are they result of error on my part?

PART 3:

As a means of validating my results against known information, I have been looking to go about similar means of proving that certain materials shall perform better than others using pre-existing cross sections. As a means of doing this, I found an equation here (https://www.psi.ch/niag/neutron-interaction-with-matter) which seems to suggest that a potential way of calculating the neutrons counts on the opposing side of the shield could be done in the following fashion:

I = I0 * 10^(-1*(atomic number density)*(total microscopic cross section)*(thickness in cm))

Where:

I = CPS on the other side of the shield
I0 = CPS before the neutrons hit the shield

However, when I solve for this equation, I am not met with results that are even in the neighborhood of the experimental results that I received (usually about 1000 times smaller than the experimental results). Thus, I presume that I am wrong in my work.

Any help or pointers would be greatly appreciated, as I do not have a great deal of experience under my belt yet.

Josh

[Dennis P Brown]

While this is a complex subject, a few quick points.

Neutrons do not attenuate like simple x-rays or even alpha particles. They are not absorbed but only scattered (unless you have added an neutron absorber to the shielding). As such, there are many neutrons 'floating around" so to speak as thermalized neutrons. Also, unlike other types of radiation, neutrons (free) have a life time that impacts their surroundings (ten to fifteen minutes.) One cannot easily shield a neutron detector from these effects.

Concrete is a very poor neutron attenuation/shield material.

Your detector 'efficiency" may or may not still be valid depending on its last calibration. An old, uncalibrated unit can be wildly off due to aging of the detector or issues with the electronics or both.

Further, I would not expect that detector to treat thermalized neutrons with the same capture rate as 4 MeV neutrons. That assumption would be my first issue followed by calibration of the unit.

Finally, a simple linear equation requires exactly that - all aspects of the neutrons interactions to a "shield" must be linear; a beam of neutrons with energies from thermal to 4 MeV is not in any manner linear. Also, what is the flux to energy ratio's? If 75% is 4MeV but 25% is thermal it is possible one type of neutron energy type could greatly change the results in a non-linear manner. The article you referenced pointed out the interaction rates for thermal vs. non-thermal neutrons run over an order of magnitude different - so you need a very single neutron energy beam to really expect linear results with shielding materials.

Research the subject in say a radiation physics book written for health physics personnel. I've found great insight using those when I had questions.

[Joshua Guertler]

Greetings,

Thank you for the response.

Based on my discussion with the people who worked at the reactor, I was able to determine that they utilized a computer program that registered counts from the ionization incidents in the detector (which had been a proportional BF3 tube). Thus, they were able to tell me that all neutrons within this range would be read and registered mostly the same (the 4 MeV and 0.02 eV was merely the maximum range with most neutrons falling around 2 MeV initially). Would it then be okay to use the linear attenuation coefficient for a shield comprised of a separate moderating and absorbing layer? The overall linear attenuation coefficient would consider the overall shield as if it were homogeneous.

Based on a few corrections of my work, I was able to discern that the equation for linear attenuation (as most of the neutron energies are around 2 MeV with a small amount being notable outliers) should have been more like:

I = I0 * e^-(attenuation coefficient*thickness)

with "e" being the number e rather than a filler for an exponential value. Am I correct in this realization?

As a final question, I was curious whether or not my assessment of converting counts per second to neutrons per second would function or if there is an effect that I should compensate for.

Thank you.

Sincerely,
Joshua Guertler

[Richard Hull]

Attenuation and Shielding..........This tells me you are looking for protection from neutrons.
It is true that concrete is not a good shield as a true shield should absorb and stop neutrons.
A good absorber typically demands thermalized neutrons.
Many neutrons are not thermalized all that well in most absorbers.
There are many materials that can thermalize neutrons very well.
Within the thermalizer, another extremely high cross section thermal neutron absorber should be used.
The thermalizer and absorber are best intimately intermixed within any good "neutron shielding".
A normal absorber material of thermalized neutrons would be boron, (borated compound). The best is cadmium.
Absorption and, thus, stopping of neutrons proceeds rapidly the greater the thermalization in and about the absorber.
Heavily borated concrete and borated cinder-block is commonly used as serious neutron shielding. For small fluxes borated paraffin is good.

I guess outside of a paper written on the subject, why would you be so worried about some precision in neutron attenuation from a fusor's output?
Neutrons from a fusor are pretty much of a non-issue, based on common outputs and run times, coupled with a moderately removed operator's station.

Richard Hull

[Dennis P Brown]

No one here can tell you about an unknown computer program used to correct for varying cross-section of neutrons depending on energy. Since your own people did, please consult them about conversion factors used.

We have just discussed this very point/problem about counts from a detector and trying to determine actual neutron flux. Read that post just written in the neutron & radiation forum section.

Frankly, your own in-house safety people will provide you all the details you need (how could they not?) and also, should have calibrated detectors for neutrons - I'm lost why you would use this forum considering the resources available at an actual nuclear fission reactor sight. Exactly where/what reactor place are you working at?

A few words on neutron shielding for forum members/readers: most people view 'concrete' as the cinder block and mortar mix they often see in construction and that is what I was referencing. I need to clarify my misleading statement about concrete and get into the details a bit; first, most articles reference cinder block as shielding material for many radiation threats- neither these cinder blocks held together with concrete nor simple cinder blocks are useful as neutron shielding due to these "shield" methodologies having near zero water content and it is water that is what offers neutron "shielding" effects in pure concrete.

As for concrete as a neutron 'shield", it is only 15% to 20 % water (by weight) so using it for neutron shielding is not a good method at all compared to other materials that Richard mention - as for pure concrete (if no sand or gravel mix; this then allows concrete to be more like 50% water by weight) one would need to use a lot of it to get enough water to be effective at scattering and thermalizing neutrons (it absorbs little compared to most typical neutron shield materials. Also, against fast neutrons, concrete can make matters worse for too thin of sections.) What one needs is the ability to absorb neutrons and Richard handles that question.

Now for neutrons and radiation issue - very short and incomplete comment: neutrons are far more damaging to human tissue than many other radiations of similar threat profiles. The weighing factor is 20 if I recall compared to x-rays for rough calculations. Besides their significant mass, and lack of charge, these particles can cause other atomic nuclei to become unstable/excited. Not good for biological creatures like us (we are about 70% water so moderate/low energy neutrons tend to interact well). Again, reference a health physics book for accurate details if one is interested in this subject.