Simulation Results for SNAP• Galactic Protons

– Flux vs Energy and Particle Species– Dose vs. Shield Composition– Dose vs. Shield Thickness

• Solar Flares (worst day)– Dose vs. Shield Thickness– Flux vs Energy and Particle Species

• ¾ cm Aluminum shield results

Tom Diehl05/20/2004

The Rakhno Model• In “V1.0” the total mass included is 276 kg in the

region directly around the focal plane.– Shield is 2 cm thick, 35 to 41 kg aluminum

– Cold Plate was 99 kg molybdenum -> 49 kg.

– Silicon substrate is 5 kg mixture.

– Silicon itself is 200 thick amounting to 107 g.

– Radiator was 69 kg aluminum.

– The rest is deck, optical bench, supports, etc … amounting to 61 kg.

• New version “V1.1” has mass 226 kg.– I use this as “the standard” as it is closer to the present

design in Solidworks.

MARS 14 Monte Carlo• http://www-ap.fnal.gov/MARS/ provides a link to

documentation. • MARS 14 is a Monte Carlo code for inclusive and

exclusive simulation of three-dimensional hadronic and electromagnetic cascades, muon and low-energy neutron transport in shielding and in accelerator and detector components in the energy range from a fraction of an electronvolt up to 100 TeV. (straight off the web page).

The “Rakhno” Model

• See FNAL Technical Memo 2221 by Mokhov, Rakhno, Striganov, Peterson “Radiation Load to the SNAP CCD” 09/15/03 for more detail about the original version (V1.0).

Shielding and Detector Model (Rakhno)

Galactic Cosmic Rays @ L2• Creme96

provides the flux of nucleons from H to Fe

• Protons are the most abundant. Helium is second most.

• Flux of protons is 4.7/cm2/s.

No geomagnetic shielding.

No trapped particles.

Solar cycle minimum.

Galactic Protons @ L2• 2 cm aluminum shield

– Dose is 6.90.4 Rad/Yr

• Table of # cm/s vs. KE (in Si) converted to flux (divide by volume)*

• Integrals (#/cm2/s):

– p: 3.410.06

– n: 4.780.08

12.80.9

– e: 0.780.08 /k: 0.350.02

*See next slide

Not exactly a flux• MARS gives me a table of particles #*cm vs energy

– Number times Path length– Energy is that which it has when it starts in the silicon.

• For instance:

Effect of Changing the Material• Change the material that

makes up the shield (maintain thickness = 2 cm).

• Lower Z is a little better than higher Z & more weak evidence for “more shield is worse”.

• Carbon fiber mix was a little better than everything but not very different from Aluminum.

Changing the Material II

• Changed the material but kept the shield mass at 25 kg.

• No strong evidence for an effect

• Little difference in secondary fluences.

Aluminum is good choice from

Shielding POV.

Dose vs. Shield Thickness (AL)• The shield in the nominal geometry is 2 cm thick

aluminum cone. Is there an optimal thickness?

• Tested the dose in the silicon as change thickness (AL density) of the shield.

• No apparent reason to make the shield 35 kg to reduce the radiation in the silicon (so long as we have thick cold plate and other material around.

Material vs. No Material• What is the effect of ALL of the material?

• Set all the material to vacuum, except the silicon. Run the simulation using galactic protons & galactic electrons.– Dose (silicon) = 5.30.4 R/y, less than with the

material around the detector.

– Galactic Protons: Ratio Doses (satellite vs. no satellite) Rprotons = (6.9+0.4/5.3+0.4) = 1.3 + 0.1.

– Galactic (Jovian?) Electrons: Relec. = 2.0+ 0.2. But the electrons have ~100x lower flux than the protons and are therefore rather unimportant.

Galactic Cosmic Rays @ L2• Table of # cm/s vs. KE

(in Si) converted to flux (divide by volume)*

• Integrals (#/cm2/s): – p: 3.450.02– n: ~0

0.260.26– e: ~0 /k: ~0

• Ratio of total charged particle flux compared to 2 cm AL is 1.320.03 • Recall, ratio of Dose

compared to 2 cm AL is 1.30.1

Turn off all material except CCD’s

Material vs. No Material• What is the effect of ALL of the material?

– A fast cosmic ray simulation using 4.7 charged particles/cm2/s should be scaled up by 1.3.

– Therefore, 6.1 charged particles/cm2/s seems more realistic for the a fast simulation of protons at solar minimum.

Galactic Cosmic Rays: Solar Max. vs. Solar Min.

• The solar wind impedes the GCR– GCR Flux solar min: 4.7

pr/cm2/s– GCR Flux solar max: 1.7

pr/cm2/s– The benchmark for GCR

calculations is the flux at solar cycle minimum. (shown just prior).

– Didn’t take this any further.Tylka et al.

IEEE Trans. Nucl. Sci.

V44 #6, 2150 (1997).

Worst Day Solar Flares @ L2• Creme96

provides the flux of nucleons from H to Fe.

• It has other options: solar min./max., worst week, worst day, worst 5 minutes.

• I chose the worst day to start.

No geomagnetic shielding.

No trapped particles.

Solar Flare of Oct 20, 1989 averaged over 18 hours.

Worst Day Solar Flares @ L2

• 2 cm Aluminum shield resulted in dose of 23242 Rads/day

Integrals (#/cm2/s):

– p: 2900180

– n: 90070

2900450

– e: 6.53.3 /k: 0.320.03

Worst Day Solar Flares @ L2

• Shield finally does some good.• Dose decreases until as we can’t shield the higher energy protons.• ¾ cm AL is as good as 2 cm AL.

Dose vs. AL Thickness

¾ CM Aluminum Shield• Dose from Galactic Protons

is 6.78+-0.45 Rads/Yr.

• Dose from Worst Day Protons is 179+-34 Rads/day.

Integrals (#/cm2/s):

– p: 3.420.06

– n: 4.620.09

11.31.0

– e: 0.710.06 /k: 0.350.03

¾ CM Aluminum Shield• “Ratio” is particle flux

for ¾ cm AL over particle flux for 2 cm AL.

• Ratio of integrals is – p: 1.01+0.02– n: 0.970.02.10– e: 0.910.12 /k: 0.990.08

- Total charged flux is 0.988+-0.029 “ths” of 2 cm case.

Summary• Galactic Proton Studies

– The shield doesn’t stop most cosmic ray protons.– Aluminum seems like a pretty good choice of material from

shielding considerations. – Dose from galactic protons is ~ 6.9+0.4 R/yr with a 2 cm

aluminum shield.– There is a scaling for simple simulations that accounts for the

secondaries.

• Solar Flare “worst day” study is done.– An aluminum shield has an important effect.– Dose is 232+42 Rads/day on worst day w/ a 2 cm shield.

• ¾ cm vs 2 cm of aluminum– The thinner shield is as effective as the thicker.

Plan• Along the lines of this talk

– Continue Solar Flare studies. I can find information about the distribution of intensity of solar flares.

– There is some correspondence between #*cm and SEU & dose calculations and I’d like to learn how that works.

– Heavy ions. – Tests for electronics in different places (I don’t expect the

results to be very different elsewhere in the vicinity of the cold plate).

– Write-up these results.

• Use “Nibble” file to improve geometry of the MARS model. I expect that will be suitable as a starting point for the Geant simulation, too.

Buffer slide

• Stuff here on isn’t part of this talk.

Shielding and Detector Model (1)

• The trick with MARS is to select the appropriate level of detail.– Include the most important shielding

components. Consideration is given to proximity to detector and components mass.

– Put a lot of effort into detailing the detector.

Literature Search (partial list)• Some satellites planned-for L2 include NGST,

Herschel/Planck, GAIA …• References: http://astro.estec.esa.nl/GAIA/Assets/Papers/IN_CCD_operations.pdf.Herchel/Planck Project Document, JPL D-19155 (2003).http://www.ngst.nasa.gov/public/unconfigured/doc_0819/rev_02/Transient_Document_Section_1_and_2_V5.

pdf.

• These give me some confidence that I’m on the right track.

• The NGST suggests an additional concern:– that we may activate the shield/cold-plate. This would

produce additional particles on the focal plane.

• I talked to Nikolai Mokhov yesterday. He is not surprised the shielding is making Si dose worse.

Unshielded / 2 cm aluminum

• unshielded ccd plane over 2 cm AL (only proton spectrum in ccd’s).

Range vs. Energy in Aluminum• From a talk by

Mike L. in April 2001.

• The 2 cm of AL shield itself is 5.4 g/cm^2.

• log(5.4)=0.73