sPHENIX discussion of the superconducting solenoid

["Yip, Kin" <kinyip AT bnl.gov>]

Hi Roberto,

Like what I told you last week, this time, I have something to show you in pictures.  Over the weekend, I did a bit more to make sure that I didn’t

make a stupid mistake.  So, in the two .bmp files, you can see the energy depositions of a 10MeV proton and a 100 GeV proton (the difference

of 1 million times in energy J ) on my simple cable of 2.0 cm x 0.84 cm with a narrow width (0.04 cm) in the 3

^rd dimension.  In which

the copper portion (representing Superconductor) is 0.64 cm x 0.14 cm.

You can see that 10 MeV protons were ~kind of stopped by Al (before it gets to the middle part, copper); whereas the 100 GeV protons just penetrate

the cable with some energy deposited.

The nos. in the plots are in MeV/cm^3 (energy per unit volume) in these plots (not MeV/g).   Each volume (of the grid) is 0.04*0.035*0.04 = 5.6E-5 cm^3.

[  If you really want to get power, you may consider a total of 2.5E13 protons distributed (ie. divided) by 111 bunches.   Each bunch is separated by ~107 ns. ]  sphenix_cable5.m  contains the no. of the 2D plot, for 50*24 set/pair of  energy/volume plus its fraction statistical uncertainty.

I do this to make sure that I was not too crazy/wrong.   [ I’ve used a big beam size which didn’t affect my average energy deposition for the

entire part  but  it’d affect the maximum in these plots.  But here, it’s just to show … ] 

I’ll be just more confident in what I’d say, in case people ask.

Kin

================

Hi Roberto,

I’ve used the worst possible (or impossible) case that I dump all the bunches of the beams directly on the coil   (and that proton still has 100 GeV after going through

all the material before them).   Even so, it’d not melt anything.

I did this kind of simulation/calculation often (beam pipe, windows  etc.)  Only when it’s marginal, then we try to use ANSYS with the engineer.

In this case, it seems not.

Kin

From: Than, Yatming (Roberto)

Sent: Tuesday, February 27, 2018 4:47 PM

To: Yip, Kin <kinyip at bnl.gov>; Haggerty, John <haggerty at bnl.gov>; Franz, Achim <afranz at bnl.gov>; sphenix-magnet-l at lists.bnl.gov

Subject: RE: [Sphenix-magnet-l] energy deposited on the copper (superconducting wire)/aluminium-stabilizer

Kin,

We need the beam bunch average diameter, is it 2.8 mm? So figure out impingement Area

How many bunches will hit the magnet before the rest is dumped in beam dump.

Convert this to Joules and divide by time to get unit of power:  Watts

Heat flux: Power/ Area  gives W/m^2

Basically we need power and duration of beam hit.

The beam will hit the inner cryostat wall first, which is the aluminum wall.

With the heat flux number we can calculate a surface temperature and penetration depth vs time.

Paul Orfin has an Ansys model where we can change material to aluminum and we can predict the peak temperature and see whether it melts.

[ 1-D ANSYS transient model for high heat flux, (1x10^8 W/m^2) onto a surface.]

The model [ mesh size and time interval] is set up to handle fast transient, of order 1-100 microseconds.

Roberto Than

Cryogenic Systems, Collider-Accelerator Department

Brookhaven National Laboratory

Building 1005S, MS 1005S, Room 212

Upton, NY 11973-5000

Telephone: (631) 344-7165

Cell 631-487-6842

E-mail: ythan at bnl.gov<mailto:ythan at bnl.gov>

From: sPHENIX-magnet-l [mailto:sphenix-magnet-l-bounces at lists.bnl.gov] On Behalf Of Kin Yip

Sent: Tuesday, February 27, 2018 4:24 PM

To: Haggerty, John <haggerty at bnl.gov<mailto:haggerty at bnl.gov>>; Franz, Achim <afranz at bnl.gov<mailto:afranz at bnl.gov>>; sphenix-magnet-l at lists.bnl.gov<mailto:sphenix-magnet-l at lists.bnl.gov>

Subject: [Sphenix-magnet-l] energy deposited on the copper (superconducting wire)/aluminium-stabilizer

Hi John and Achim,

I used my usual MCNPX energy deposit simulation to do something quick (worst scenario).

From the BaBar paper （like this one) :

https://collab.external.bnl.gov/sites/sPHENIX-Magnet/Shared%20Documents/BaBar%20Documents/The%20BaBar%20Superconducting%20Coil;%20Design,%20Construction%20and%20Test.pdf

I made up a rectangle of 1.4 mm x 6.4 mm of copper (I didn't bother with NbTi) surrounded by the Al   8.4 mm x 20 mm.

Then, I hit proton directly on it (with some beam sigma if 0.28 cm  --- something that I used for the calculation in RHIC).   Quickly, I get the total energy deposit as such :

2.249 MeV/g per  proton  (hitting it) on the copper   and

1.18   MeV/g per proton                    on the Al stabilizer,

which are

3.6E-13 J/g per proton   and

1.9E-13 J/g per proton.

Our proton beam is said to have a max. of 2.5E13 (totaled from all 110 bunches).   I'm not sure who can make the beam turn 90 deg. and hit the Magnet.   But let's say that it's possible

that all the beams were dumped in one spot of the Magnet coil, the total energies are

~9.0 J/g  (Cu)  and  4.7 J/g (Al-stabilizer) .

From the Enthalpy curve (eg. Fig 3 on p.5) :

https://collab.external.bnl.gov/sites/sPHENIX-Magnet/Shared%20Documents/Miscellaneous%20(General%20Knowledge)/Material%20Properties%20at%20Low%20Temperature.pdf

or some others that Bob Lambiase showed me,

the temperature rise would be < 100K.    So, it'd probably quench the magnet but it won't burn a hole.

If George or somebody asked this kind of question again in future review, I would answer with the above.   Please let me know if you somebody find something critically wrong.

Kin

 

Attachment: 10MeV.bmp
Description: 10MeV.bmp

Attachment: 100GeV.bmp
Description: 100GeV.bmp

Attachment: sphenix_cable5.m
Description: sphenix_cable5.m