Re: [fluka-discuss]: Re: Trying to understand an observation in photon dose rates (or photon fluence)

From: Alberto Fasso' <fasso_at_SLAC.Stanford.EDU>
Date: Fri, 27 Sep 2013 01:18:19 -0700 (PDT)

Hi Mina and Mario,

please let me give a little contribution to your discussion.
In my experience, the best shielding for high energy neutrons is iron, followed
by a hydrogenated material (plastics - not necessarily polyethylene - if
radiation damage is not an issue, otherwise simple concrete or even water if
it is technically possible). Via inelastic reactions, iron quickly slows down
high energy neutrons to energies where scattering by hydrogen becomes effective.
Heavy concrete is not as good, for the reasons explained by Mario (cost,
non-uniformity, unknown composition) but especially because the two moderation
effects (inelastic reactions and scattering on hydrogen) take place in
the same material, while separate iron/concrete shields perform the slowing
down in sequence: first the high energy neutrons via inelastic reactions, and
then the resulting low energy neutrons via scattering on hydrogen.
This is easily checked with FLUKA.
The widespread, but undeserved, popularity of heavy concrete is due mainly to
its use in pure photon fields without neutrons (i.e. photons of less than 7
MeV, that cannot produce neutrons). In these cases, all that matters are
average Z and density which are well provided by heavy concrete.
I also confirm that lead is bad for shielding, for the reasons given by Mario,
but also: for photons because each photon absorption is followed by
fluorescence, which must be shielded against too; and for neutrons, because
(n,xn) reactions increase the number of neutrons instead of decreasing it.
This latter effect is much less important in iron.

Alberto

On Thu, 26 Sep 2013, Santana, Mario wrote:

> Hi Mina,
>
> Just some thoughts?
>
> You know that lead is not the best shielding for high energy neutrons.
> They will bounce through the structure and make it to the other end of
> your shielding. A light material will reduce the energy of the neutrons
> very effectively at each collision. Moreover, it has a fairly large photo
> neutron cross-section. Also lead is toxic, and when exposed to radiation
> it will become 'mixed waste', so you may want to minimize its use. This is
> why it is often used close to the source, to absorb most of the
> electromagnetic cascades, and then concrete is used in the accelerator
> enclosure to 'take care of the neutrons'.
>
> Polyethylene is a good option for neutrons (specially if it is borated),
> but as you indicate it may degrade (and also catch fire). Also it is less
> dense than concrete so it does not absorb photons as well as concrete. An
> advantage over concrete is that its residual activity is mainly
> short-lived.
>
> About heavy concrete, it is a hybrid solution that seems to work
> successfully when shielding space is scarce, but it is expensive and you
> have to make sure that your heavy concrete is sufficiently homogeneous.
> Also some suppliers don't provide the composition, so you may have a hard
> time to simulate its effects.
>
> Mario
>
> On 9/26/13 6:21 PM, "Mina Nozar" <nozarm_at_triumf.ca> wrote:
>
>> Just wanted to update everyone with my finding on this. Because the
>> beam energy (75 MeV) is way above the neutron production (gamma, n)
>> threshold in lead (6.73 MeV), there are lots of neutrons generated in
>> the lead. These neutrons in turn produce other gammas through (n,
>> gamma) reactions. So the effectiveness of lead in shielding gammas stop
>> at some thickness (for energies higher than the gamma, n threshold).
>> The way I got around this was to layer the shielding design (Pb + PE +
>> Pb + PE ...). This design works in attenuating photons and neutrons so
>> that the total dose rate is below the required limit.
>>
>> However, I am wondering about the radiation resistance properties of
>> Polyethylene. The first (innermost) layer of PE is seeing Dose rates on
>> the order of 1E7 mSv/h or 10 kSv/h, so very high. Looking through
>> literature, I am seeing Total dose limits of 1E6 Gy (or Sv) - from CERN
>> 98-01. This gives 100 hours of operation before PE is damaged.
>>
>> So I am thinking the Pb/PE layering is not a viable option. Does anyone
>> know of PE being used in high radiation fields? Does anyone have
>> experience in this area?
>>
>> Another option I have investigated is to use high density concrete which
>> seems to do the trick. So we might go with that. I just want to make
>> sure I am not missing something before writing off the Pb/PE layering
>> option.
>>
>> Any help/advice is appreciated.
>> Best wishes,
>> Mina
>>
>> On 13-07-19 06:19 PM, Mina Nozar wrote:
>>> Hello everyone,
>>>
>>> We are working on optimizing the shielding for a 100 kW (75 MeV, 1.3 mA)
>>> electron beam dump. BD is basically a chunk of Al with a slanted plane
>>> in the middle to distribute the incident beam (square in shape 4x4 cm).
>>> The BD is surrounded with lead and concrete, lead and polyethylene (PE),
>>> or lead and borated PE(BPE) depending on the study. The BD is 111 cm
>>> long, 20 cm wide, and 12.5 cm high.
>>>
>>> In order to compare thickness of PE, BPE, or concrete required in
>>> shielding neutrons and finding the minimum thickness of lead required to
>>> bring down the dose rates from photons to meet different limits, I have
>>> been looking at dose rate distributions from neutrons and photons
>>> separately.
>>>
>>> I have set the production and transport thresholds for e+/e- to 100 keV
>>> and 10 keV for photons.
>>>
>>> I originally thought if I extracted TVL from the attenuation of photons
>>> in lead and neutrons in concrete (PE/BPE), I could find the required
>>> thickness of each material to reduce the dose rates down to the limits.
>>> However, I have observed a 'build-up' behaviour of photons in lead that
>>> I don't understand (and which is affecting my conclusions). I would
>>> like to find out whether this is a by product of geometry effects,
>>> thresholds set, etc. or if this is a real build-up effect I am
>>> observing, suggesting that past some thickness, the effectiveness of
>>> lead to shield photons goes down.
>>>
>>> I am attaching plan and side views of the geometry around the BD and two
>>> figures showing 1-D distributions of the dose rates (from neutrons,
>>> photons, and total), as a function of width and height of the BD. The
>>> slices I have made are 10 cm wide around the beam line and center of the
>>> beam dump. The arrows point to the areas showing the 'build up'
>>> feature which continues through the lead/concrete boundary.
>>>
>>> I am showing results of the lead and concrete study only but I observe
>>> the same behaviour for lead and PE.
>>>
>>> Thank you very much for any insight you might have on this. I
>>>
>>> Best wishes,
>>> Mina
>>
>
>

-- 
Alberto Fasso`
SLAC-RP, MS 48, 2575 Sand Hill Road, Menlo Park CA 94025
Phone: (1 650) 926 4762   Fax: (1 650) 926 3569
fasso_at_slac.stanford.edu
Received on Fri Sep 27 2013 - 11:14:09 CEST

This archive was generated by hypermail 2.3.0 : Fri Sep 27 2013 - 11:14:10 CEST