How to optimize x-ray tube simulation in egs++

[rtownson]

This question was originally posted here, and I have copied it over to preserve the discussion: https://www.reddit.com/r/EGSnrc/comments/sh1kwb/x_ray_tube_simulation_in_egs_through_toutor7pp/

I created x ray tube and ran. But its took too much of cpu time even though I used UBS and parallel run in 8 core. its took 30 hours for deposit 3k particle in phasespace file. below i mentioned the trasnport parameter which is used in this run. How can i reduce time for 50kvp to 150kvp simulation

:start MC transport parameter:
    Global ECUT                     = 0.512 # Global electron transport cutoff
    Global PCUT                     = 0.001 # Global photon transport cutoff
    Global SMAX                     = 1e10 # Global maximum step-size restriction for e- transport
    ESTEPE                          = 0.25 # Default is 0.25
    XIMAX                           = 0.5 # Default is 0.5, max. value is 1.
    Boundary crossing algorithm     = exact # exact (default), PRESTA-I
    Skin depth for BCA              = 3 # Default value is 3 for exact boundary crossing
    Electron-step algorithm         = PRESTA-II # PRESTA-II (default),PRESTA-I
    Spin effects                    = On # On (default),Off
    Brems angular sampling          = KM # Simple,KM (default)
    Brems cross sections            = NIST # BH (default),NIST
    Pair angular sampling           = Simple # Off, Simple (default),KM
    Triplet production              = Off # On or Off (default)
    Electron Impact Ionization      = On # On, Off (default), casnati, kolbenstvedt, gryzinski
    Bound Compton scattering        = Simple # On, Off, Simple, norej (default)
    Radiative Compton corrections   = On # On, Off (default)
    Rayleigh scattering             = On # On ,Off (default), custom
    Photoelectron angular sampling  = On # On (default),Off
    Atomic relaxations              = On # On (default),Off
    Photon cross sections           = xcom # si, epdl, xcom (default) or user-supplied
    Photon cross-sections output    = Off # Off (default) or On
    Compton cross sections          = default # user-supplied
    Photonuclear attenuation        = On # Off (default) or On
    Photonuclear cross sections     = default # default (default) or user-supplied
:stop MC transport parameter:

The user also asks if it is OK to use ECUT=AE=0.521

[Jaap1957]

I would not use MC always to get xray spectra since they were measured already a long time ago.
So here a non EGS suggestion, although I very much appreciate EGS:

For Xray tube "simulation" one can refer to the CD-ROM published with this 1997 report, part of a publication of the British NHS:

Catalogue of Diagnostic X-ray Spectra and Other Data, IPEM Report 78, K. Cranley, B. J. Gilmore, G. W. A. Fogarty and L. Desponds (Electronic version prepared by D. Sutton). The Institution of Physics and Engineering in Medicine and Biology. 1997. ISBN 090418188X

This program contains a large set of measured and simulated Xray tube spectra in the medical diagnostic range. See Simon Evan's 1998 review for additional information.

Yes, this is a very old program. But still useful. It can be executed in windows 3.11 (!!) See
https://www.lifehacker.com.au/2019/03/how-to-install-ms-dos-6-22-and-windows-3-11-in-a-virtual-machine/
for how to install an old version of Windows on a virtual (time) machine.

[ftessier]

@Jaap1957 Interesting, thanks! But we cannot distribute the CD-ROM files here, so I have updated your post with a reference instead, and a reference to a review on it. Now this is quite old indeed, and it would be an interesting exercise to compare with modern calculations.