Progress Report from the Mechanical/Shield Group Feb 19, 2003


* The next meeting is Feb. 26th at 9:30 AM in the Fermilab Cafeteria.


* Shield and Cold Plate Engineering and Integration  
  Clark Reid continues to work on importing the SNAP satellite 
  integration model into Ideas. He has produced more 2D drawings 
  of some shielding components. They are on our web page.

  Sandor Fehers has imported some of the integration model with SolidWorks.
  We agree we need to have at least one SoldiWorks liscense within our 
  group, as that is the program they are using at LBL.
  
* Shielding Simulation

  Igor Rakhno and Sergei Striganov showed their progress in importing 
  heavy ion interactions into MARS. They showed the dE/dx(MeV/cm) and 
  range (cm) in 27Al for protons, 4He, 12C, 27Al, and 56Fe as a function 
  of projectile energy. The energy loss calculation is that by EM 
  interaction.

  Fritz Dejongh is working on the Spenvis simulation.  He noted that 
  some satellite shielding simulations use simple "line-of-sight" 
  traces of the material in the satellite in a given direction. He 
  asked if the 3d models can produce these sorts of summaries.

  TD showed comparisons of relative EM and nuclear-interaction shielding 
  thickness of aluminum, molybdenum, and tungsten. 
  If a cone is 180 kg and is 1 m tall with radius 1/3 m,
  then the thickness is T. A is atomic number. Z is charge.
  I is nuclear interaction length. I(cm) ~ 35 (g/cm^2) A^1/3. 
  Material    Density (kg/m^3)     A   Z  I(cm)  T (cm)   ZZT   T/I
  ------------------------------------------------------------------
  Aluminum    2740                27  13  39     8        13.5  0.21  
  Tungsten    1.93e4             184  74  9.6    0.9      49    0.094
  Molybdenum  1.03e4              96  42  15     1.7      30    0.11 
  ZZT (Charge^2 * T) is a crude measure of the effectiveness at shielding 
  from electromagnetic radiation. More is better.  T/I is the number of 
  interaction lengths.  With this few, it's not clear more interaction 
  lengths is better.