
================================   xtubes   ====================================


 xtubes is an widget based graphical interface to calculate the spectum 
 of an X-ray tube with Mo, Rh or W target in the energy range 18-40 keV.

 It uses the method descibed in the manuscript:
 "Molybdenum, rhodium, and tungsten anode spectral models using interpolating
 polynomials with application to mammography"
 J.M. Boone T.H. Fewell and R. J. Jennings 
 Medical Physics 24(12) pp. 1863-1874, 1997. 

 The program interfaced by XTubeW has been downloaded from:
 ftp://ftp.aip.org/epaps/medical-phys/E-MPHYA-24-1863

 DESCRIPTION OF THE CONTROLS IN THE MAIN WINDOW:

  File:
    xtubes input parameters: This option allows to save the current
		parameters to a file for later loading. It also allows
		to save the current parameters as defaults for being
		used when the application is initialized. In the last
		case, the file is named "application".xop (where 
		"application " is the name of the current XOP
		application) and is written in the directory pointed
		by the XOP_DEFAULTS_DIR environment variable (which
		must be set). The parameter file is ASCII and can be
		read and edited with care.
    Write Files for XOP/Optics: Saves the current undulator
		spectrum in the SRCOMPE (flux) and SRCOMPW (Power) files
		to be used for the other XOP application from Optics
		menu.
  Quit: to exit from the program

 Set_Parameters:
  Set Parameters: to define the parameters for the calculation.
		The same result is obtained pressing the "Set Parameters"
		button in the main xtubes window. 
               Please refer to the information under the HELP
		button for a complete description of the parameters. After 
		pressing the ACCEPT button, xtubes starts running.
  Set Defaults: Sets the default parameters.

  Show: Display results
    Plot Results: Plots the xtubes results (see note on units).

  Help:   Shows the xtubes help (this text).

 NOTE ON RESULT UNITS:
     The units of the fluence in the results are not very clear 
     in the Boone's paper. This has creates some problems for users,
     and originated a discussion between Sutter (sutter@ntmail.desy.de) and 
     Boone. You can find below a short version of the discussion:

     In the paper for the W (only) [XOP program xtube_w] Boone's paper clearly 
     states that the flux per sec per 1keV bandwidth and per mm2 collected at 
     1m from the target. The result looks very reasonable, and I (srio@esrf.fr)
     have checked one result against the example in the doc of "Catalogue of 
     diagnostic x-ray spectra and other data", Cranley et al., Report 78, The
     Institute of Physics and Engineering in Medicine, with quite reasonable
     agreement.  However, in the paper on Mo, W and Rh [code Xtubes] the units 
     are not clearly stated. Moreover, both programs produced different 
     results for W @ > 41 keV (one case that can be calculated with both codes).

     I (sutter@ntmail.desy.de) have had several interesting exchanges with 
     Dr. Boone. 
       The units: The paper on Mo, W, and Rh states that the spectra
       collected on these tubes were binned into 500 eV intervals. (See Fig. 2
       for an example of these spectra for the Mo anode.) In addition, the 
       caption of Fig. 2 states that all the spectra were normalized to a 
       constant value of mA x sec, but does not reveal what this value was. 
       When I asked Boone about this, he said that his reason for not revealing
       it was the following: 
       "You are correct, I intentionally did not "reveal" the absolute output
       efficiency of the x-ray sources because x-ray tubes are so variable -
       the calibration of the mA (and the time) is often imperfect (and hence 
       the mR/mAs term would be in error) and the amount of inherent x-ray 
       tube filtration (we deal with tubes in the clinical environment, which 
       have collimators and other structures over the x-ray tube port which 
       make it difficult and impractical to visually check exactly what 
       filters are in the beam.  Thus, the total inherent (and added) 
       filtration is often not known.)"
       When I pressed him for at least an approximate value for the normalizing
       constant, he stated that it was 1 mA x sec. However, the above warning 
       about the variability of X-ray tubes should be kept in mind.

       In short, the units for the fluence, strictly speaking, should be:
       photons/mm2/500 eV/(mA.s). 

       One final note: Boone stated in another message to me that the fluence 
       from a point source drops off with the inverse square law. This 
       indicates that the area unit (mm2) in the fluence refers to the number 
       of photons per unit surface area on a sphere of given radius centered 
       on the anode, rather than to the number of photons produced per unit 
       illuminated area on the anode.

       Another quantity that isn't given with certainty in Boone's first 
       paper is the distance between the Mo, Rh, and W tubes and the detector. 
       When I asked Boone for this quantity, he said it was "probably 50 cm," 
       this being a common distance for mammography measurements. 
       The measurements for the W tube in Boone's second paper, corresponding 
       to the XOP program "xtubes_w", however, are clearly stated to have been 
       made at 1 m from the tube. 
       Perhaps this explains the discrepancy between the results from
       the two calculation routines xtube_w and Xtubes . Does this make sense?
       (The fact that 500-eV bins are used in the first paper, while the
       fluence in the second is given per 1 keV bandwidth, will also influence 
       the results.)

       Added srio@esrf.fr: Following this arguments, in order to
       transform the units of xtubes (fluence @0.5m with 0.5keV bandwidth) 
       into xtube_w units (fluence @1m with 1keV bandwidth) one should 
       multiply by (0.5/1.0)^2 * 2 = 1/2 (in other words, at 1m we have
       1/4 times the flux of at 0.5m multiplied again by 2 because we use the
       double of bandwidth). However, a strage factor of about 1000 is found: 

       W @ 41 kV
       Photon_Energy[keV]   Fluence[xtube_w]    Fluence[xtubes]
       25                   1.24e4              1.23e7
       35                   7.02e3              4.39e6
      
      I have no explanation for this discrepancy. 
      
 COPYRIGHT:
	XTUBES belongs to XOP package and it is distributed within XOP.
	PLEASE REFER TO THE XOP COPYRIGHT NOTICE

 REFERENCE:
	Published calculations made with XOP should refer:

	M. Sanchez del Rio and R. J. Dejus 
        "Status of XOP: an x-ray optics software toolkit"
        SPIE Proceedings Vol. 5536 (2004) pp.171-174

        http://dx.doi.org/10.1117/12.560903

	In addition, published calculations using the xtubes aplication 
	should also cite:

	J.M. Boone T.H. Fewell and R. J. Jennings "Molybdenum, rhodium, and 
	tungsten anode spectral models using interpolating polynomials with 
	application to mammography"
	Medical Physics 24(12) pp. 1863-1874, 1997. 


 LAST MODIFICATION: srio@esrf.eu 2008-01-28
