Home

RADIOACTIVE ION BEAM OPTIMISER, RIBO. USER MANUAL.

1. RRRR EEEEE DDDD FFFFF L DDDD R R E A A D F I E L D D RRRR EEE AAAAA D D EEEE FFFF I L D D R R E A AD DE F I E L D D R R A DDDD EEEEE I LLLLL DDDD THIS ROUTINE READS ELECTRIC FIELD DATA SI FROM A FILE and or from constants DEPENDENCIES checkCELL MERE readEfield El x 8 lt computedt Bn gt EMfield main x B electrans backupstep VARIABLES x y z absolute coordiantes x y z m E 3 electric field vector SI Defining Electic Fields 77 78 79 80 81 82 83 50 K K K K
2. CA EXAMPLE heat3D t The solid W sphere sees the inner side of the outer Ta spheric shell C 10 90 90 1 corresponds to the temperature in a point 38 39 40 41 42 43 44 45 2 3 USER DEFINED FUNCTIONS AND SUBROUTINES 73 of the Ta inner side The spheric symmetry allows to leave that same reference point for whatever position The Ta spheric shell celln 2 sees in the inside the W sphere C 31 90 90 1 and in the outside the open air T 298 AICI IF celln eq 1 THEN W shpere TO C 10 90 90 1 Temperature of inner face of Ta shell ELSE IF celln eq 2 THEN Ta shell IF norm R 1e 4 40 THEN inner face TO C 31 90 90 1 gt Surface temperature of the W spher ELSE outer face TO 298 0 gt Sky temperature END IF END IF radia eps 0 4 4 dA i CpRhoxdV EXAMPLE 2 CONVECTION Note that you can also include convection First define the variable conv and the parameter constant Uncomment these lines conv constant TO T radia eps 4 4
3. DRF T 2 1 5 of Diffusion and Effusion RIBO computes DRF 7p geometry and ERF t t2 ts 2 from the be decoupled analytic functions and it prints out the results in tables For example tao d s 0 10 0 0 0 0 0 1 8 0 1 7 0 1 6 0 1 5 0 1 4 T 1 2 s DRF RF tsl RF ts2 RF ts3 RF ts4 RF ts5 RF ts6 0 1E 2 7 727 0 000 0 000 000 0 000 0 000 0 000 0 0 3 2 13 120 0 007 0 007 0 007 0 006 0 003 0 000 0 0 0 6E 2 18 198 0 036 0 036 035 0 031 0 014 0 001 0 1E 1 23 022 0 116 0 116 114 0 102 0 046 0 003 0 3E 1 37 213 1 214 1 212 1 201 1 096 0 552 0 046 0 6E 1 48 936 4 329 4 326 4 294 3 995 2 247 0 225 0 1 0 58 506 9 598 9 592 9 536 9 005 5 584 0 678 0 3E 0 78 059 33 413 33 402 33 302 32 325 24 677 54327 0 6 0 87 058 53 647 53 637 53 541 52 597 44 483 14 564 0 1 1 91 622 67 231 67 222 67 143 66 364 59 327 25 862 0 3 1 96 960 86 780 86 776 86 738 86 364 82 766 57 048 0 6 1 98 441 93 026 93 024 93 003 92 797 90 770 73 942 0 1 2 99 051 95 719 95 718 95 705 95 575 94 298 82 933 0 3 2 99 673 98 533 98 533 98 528 98 483 298 036 93 739 0 6 2 99 830 99 256 99 256 99 254 99 231 99 004 96 784 0 1 3 99 893 99 548 99 548 99 546 99 533 99 396 98 046 3 D Diffusion module 2 1 Brief physical introduction The diffusion of atoms within a solid and the conduction of heat both follow the Fick s Law Ja a Vall 2 1 Where U is the concentration 2
4. KT T T 2 Cp T J Kem 3 0 CpTT T 2 t desorption s eps emissivity 0 1 p i DO KO DC KT DCC KTT t desleps CpO CpT CpTT 1 2 70 0 0 0 0 0 25 170 0 0 0 0 0 2 2 49 0 0 0 0 0 19 58 0 0 0 0 0 2 2 2 Activating 3 D diffusion calculations The user should give the following answers to activate 3D diffusion calculations see example in 3 1 4 1 inputfilename 2 outputfilename 66 10 2 3 CHAPTER 2 THREE DIMENSIONAL DIFFUSION MODULE cello cell to which generation is limited Type 0 if you don t need any restriction here inmode 5 for atomic diffusion or 6 for heat transfer output printing mode 8 average numbers 9 print full grid C custom Xmin Ymin Zmin Low edge coordinates of the 3D window Vacuum will occupy the outer space DX DY DZ Lengths of the sides of the 3D window box NX NY NZ Number of bins in each direction Termination condition and value User defined functions and subroutines The file sources custom3D f contains a set of functions and subroutines that can be edited to fit specific requirements in 3D diffusion problems For the time being the included objects are l CSTART f to specify the starting profile of concentration or temperature generf to define outer sources or sinks of atoms or heat D
5. DIAMETER pm Some indicative values pm He 62 Ne 76 r 142 Kr 176 216 Rn 240 N2 374 Consider also diameter 2 x 1 4 x A 1 3 lt 0 gt MEAN FREE PATH cm in vacuum n 0 gt RIBO will estimate the MEAN FREE PATH In case of ionic transport the emittance is computed with the program emittance f The central axis has to be modified to match the particularities of each case The output is written in M II AAA EEEEE II T T A A NN N C E MMM II T T AAAAA NNN C EEEE M II T T A A N NN C E EEEEE M M II T T A A N N CCCC EEEEE FCCC CCC ICCC The RIBO project MARIO SANTANA LEITNER 2000 2006 SUBROUTINE emittance epsilon X3 U3 1 8 X3 3 U3 3 li real 8 epsilon 2 o real 8 R 2 Pi FC CCC CC ICC CC IC IC IC CUSTOMIZE THIS FUNCTION TO FIT YOUR PROBLEM USAGE print emittance maps in given cross section f 21 22 23 34 CHAPTER 1 MONT
6. 4 Recompile the RIBO distribution with the script provided in tools make sh 5 Run RIBO and provide the right run time options discussed in 3 1 4 2 2 1 file The file data Cond dat contains the information about the diffusion parameter desorption time emissivity and volumetric specific capacity at constant pres sure The first 6 lines of the file are explanatory and should not be removed The code should find the valid information starting exactly in row 7 The first column 2 2 INSTRUCTIONS USE 65 1 contains the cell number celln The columns 2 3 4 have the information cm 8 on the diffusion parameter D 2 either as the diffusion coefficient D or the conductivity normalized to the specific capacity and to the density In any case the diffusion parameter can depend on the variable U Concentration or Temper ature In every boxel see eq 2 2 the code will compute the diffusion parameter as D D0 DU U Column 5 has information on the average desorption time s for atomic diffusion or the emissivity for photon radiation Special temperature depen dence laws 7 should be introduced in sources radia f See 3 1 4 Columns 6 7 and 8 Contain the specific heat capacity in 25 Three coefficients can be provided like for D 3 D DIFFUSION MASS HEAT DATA RESPECT FORMAT D T cm2 s DO DC C DCC C 2 K T W cmK
7. 1 4 0 0001 88 CHAPTER 3 EXAMPLES RIBO USE RRRR 5555 0006 5555 R 5 A AS RRRR SSS GG SSS RR SG GA A s EEEEE 5555 GGG 5555 INTRODUCE A NUMBER n IF n gt 0 gt n ATOM DIAMETER pm Some indicative values pm He 62 Ne 76 Ar 142 Kr 176 Xe 216 Rn 240 N2 374 Consider also diameter 2 x 1 4 x A 1 3 n lt 0 gt n MEAN FREEPATH cm in vacuum n 0 RIBO will estimate the MEAN FREE PATH 0 Free mean path between collisions with gas cm 1 T d P 3587 8078 298 44 0 0001 DONE DONE READING data valves dat file for moving walls no valves active in the system Post processing geometry DONE END OF INITIALIZATION SIMULATION STARTS RUNNING N Nion Nabs lt dist gt m 50 8 0 1 15077824 N Nion Nabs lt dist gt m 100 16 0 1 30160217 N Nion Nabs lt dist gt m 150 220 1 42556459 N Nion Nabs lt dist gt m 200 32 0 1 37736645 N Nion Nabs lt dist gt 250 36 0 1 34756701 N Nion Nabs lt dist gt 300 41 0 1 37057745 closing fitting release function to statistical momenta DONE Results are stored in test out Finally RESGAS asks us for the residual pressure Normally we would say 0 Here however we say it is 1E 4 torr We ask the code to compute the free mean path from the pressure Then RIBO ends up the prepa
8. SPECIAL EVENTS TERMINATION Please choose among these options Crossing of end surface lst card Tally No ionizations in the system Atoms can be ionized in a PLASMA ion source Histories of atoms and ions end when they cross the end surface detector Atoms can be ionized in a SURFACE ionizer Histories of atoms and ions end when they cross the end surface detector Atoms can be absorbed in the walls Trajec tories end at absorption or when crossing the end surface what do you want 1 2 3 or 4 As a function of the previous choice four options are then possible 18A fourth option including laser ionization could be available in the future The fifth option is not really an ionization mechanism it is used to map the distribution of radioactive atoms stuck in the walls as a consequence of prolonged sticking Configuring the ion source Activating absorp tion in the walls 26 IfR CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL IBO is run without ionization then no more data will be required at this point and the program will continue to collect the runtime parameters If ionization is activated then the code will ask into how many cells the ion source expands Please consult the example at 3 1 3 to learn more e If Plasma ionization has been chosen then the RIBO code will start by asking which are the indexes of all the ionizing cells then it will
9. dz cm 3 Useful to normalize generation PAR 2 concentration temperature at point X cm 3 T PAR 3 celin PAR 4 t desorption emissivity s W cm 2 K 4 PAR 5 dt s Dbq 1 DO cm 2 K 0 s k0 W cm 71 Dbq 2 DC 2 1 s kT W cm K 2 Dbq 3 DCC cm 2 k 2 s kTT W cm K 3 Dbq 4 t s s Eps W K 4 Dbq 5 W s cm3 1 Dbq 6 CpT rho s cm3 K 2 Dbq 7 W s cm3 K 3 rcr UR dA 3 dA i is the perpendicular area for normal i cm 2 CpRho CpxRho s K cm 3 TO ambient temperature sigSB stephan Boltzmann constant 5 6703E 12 W cm 2 K 4 CUSTOMIZATION replace eps by a function of temperature eps T replace TO by the ambient temperature or by the temperature of enfolding cells e g 70 10 4 9 1 ICICI CCCI AACA celln NINT PAR 3 IF M i 1t 1E 14 THEN CpRho Dbq 5 celln Dbq 6 celln T Dbq 7 celln T dV PAR 1 cm3 sigSB 5 6703 12 W K 4 cm 2 emissivity PAR 4 eps sigSB emissivity W K 4 cm 2 TO 298 0 K ambient temperature
10. 104 1 8 USER DEFINED SUBROUTINES CUSTOMIZE THIS FUNCTION TO FIT YOUR PROBLEM USAGE print values after each history eg x y z teff VARIABLES INPUT ofile Output file ocell Cell where particle was born Cell where particle trajectory ends ion Ionic state 0 neutral 1 ionized ads Adsordtion 0 none 1 particle was adsorbed endvec 11 Tells which of the paths you can put up to 11 has been first completed In normal cases a single end surface this is not relevant survec 11 Tells which is the end surface X03 3 Absolute position at starting point x0 y0 z0 cm X3 3 Absolute position at end point 2 1 U3 3 End velocity ux uy uz of the atom m s COL Total number of collisions outside powder particle COLP Total number of collisions inside powder particle COLP6 6 Collisions of type i per particle td Diffusion release time s ts Sticking time per collision s coll tde Particle elapsed desorption time spent in surfaces other than those of the powder s tdep Particle elapsed desorption time spent in powder cells s tPo Flight time in the powder s teff Flight time outside the powder s tp Total effusion time tp teff tPo ts COL COLP Total release time per particle tT tp td AUXILIARY eve Event index eve ion 2 ads EXTERN
11. 52 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL K K K K KK K K K K K K K K K K K FK FK K K K K K K K K K K K K K K K K FK FK K K K K K FK KK K K K K K K K K K K K K K K K K K K K K K K K K K K K FUNCTION Bn X0 n real 8 X0 4 Bn integer 4 n Bn 0 0 IF n eq 1 THEN Bnz0 0 ELSE IF n eq 2 THEN Bn 0 0 ELSE IF n eq 3 THEN Bn 0 0 Bn 1 0 X0 3 10 END IF END To obtain emittance plots modify the emittance plane and settings in the routine sources emittance f This function is presented in section 1 5 3 1 8 5 User defined output printing files userprint f This routine is called at the end of each particle history It allows the user to print in whatever relevant information in units 21 25102 SUBROUTINE userprint ofile td tp ts tPo tde tdep COL COLP COL6 U3 X3 X03 0cell celln ion ads survec endvec character 20 ofile integer 4 ocell celln ion ads survec 11 endvec 11 li real 8 X03 3 X3 3 U3 3 11 real 8 COL COLP COL6 6 td 5 tde tdep teff tp tT ld real 8 V3scaV3 norm eve a external rand e WRITTEN BY MARIO SANTANA LEITNER 2006 102The data will be stored in the files 2 out to 25 out in the directory from which RIBO is run 99 100 101 103
12. SOURCE CUSTOMIZATION You can edit the file sources customsource f and then recompile with tools make sh Additionally you can use the data file init dat by uncommenting read in customsource f And you can restrict generation to a cell source limited to a cell give cell number celln gt 0 gt generation limited to volume defined by celln celln 0 gt do not constrain to a cell celln lt 0 gt just generate a geometry plot SPECIAL EVENTS TERMINATION Please choose among these options 1 Crossing of end surface lst card Tally No ionizations in the system 2 Atoms can be ionized in a PLASMA ion source Histories of atoms and ions end when they cross the end surface detector 3 Atoms can be ionized in a SURFACE ioniser Histories of atoms and ions end when they cross the end surface detector 4 Atoms can be absorbed in the walls Trajec tories end at absorption or when crossing the end surface So what do you want 1 2 3 or 4 3 il 3 1 how many ionising cells max is 3 cell number of ion cell number 1 how many ionising surfaces in cell 3 Ionizer surface number 1 in ion cell 1 3 1 After giving the name of the input and output file we asked to restrict the source to cell number 1 the ellipsoid and activated simulations with a surface ioniser The ionising surface is the cylinder 9 belonging to the cell
13. defined in math f COL6 6 COL6 6 1 0 END IF EXAMPLE 3 Poth or Phong model The particle is reflected with a cosine n law around the mirror like reflected direction n depends on the roughness zero 20 0 n 5 887 gt 0 gt very rough cosphi V3scaV3 U3 grad3 norm U3 norm grad3 cos phi a b 0 Sample the alpha from by REJECTION alpha asin rand zero cosphi alpha uniform between 0 and 90 phi DO WHILE rand zero ge alpha n 1 0 48 49 39 51 52 93 54 1 8 USER DEFINED SUBROUTINES 45 alpha END DO asin rand zero cosphi Now sample the azimuthal angle beta rand zero 2 3 14159265 Now build local base and get CALL surfaceBase SU3 SV3 grad3 al SIN alpha COS beta bl SIN alpha SIN beta cl COS alpha U3 1 U3 2 U3 3 al SU3 1 bl SV3 1 al SU3 2 bl SV3 2 al SU3 3 bl SV3 3 CALL renorm U3 COL6 5 COL6 5 1 0 cl cl cl beta by INVERSE transformation exit vector U3 grad3 1 grad3 2 grad3 3 EXAMPLE 4 if incidence angle gt 60 Otherwise cosine law reflection zation sticking time exp t t0 mirror like maxwell Boltzmann no sticking therm
14. t s s Eps W K 4 Dbq 5 CpO rho W s cm3 71 Dbq 6 CpT rho s cm3 K 2 Dbq 7 CpTTxrho W s cm3 K 3 ncell number of cells of the geometry A CpRho CpxRho s K cm 3 UNITS of gener gener For MASS TRANSFER dV need par cm 3 s par s For HEAT TRANSFER dV CpRho 2 3 USER DEFINED FUNCTIONS AND SUBROUTINES need J cm 3 s EXAMPLES 1 constant generation of particles 1 5 par cm3 s gener par s 1 5 dV 2 constant heat generation 1kW cm 3 gt gener K s 1000 dV CpRho s sk s sek NOTE The generation may depend on time and on X dV PAR 1 cm 3 gener 0 0 IF inmode eq 5 THEN 3D mass diffusion gener 1 4 dV par s ELSE IF inmode eq 6 THEN 3D heat transfer CpRho Dbq 5 celln A Dbq 6 celln T
15. S 7 Tally S max 3 300 Figure 3 4 The target and ion source of example 3 298 0 25 1 1 0 0 0 0 0 2000 0 1 1 0 0 0 0 0 52 0 0 0 0 0 0 1 2 4 5 6 7 3 9 0 0 0 0 Alfa nx ny nz x y 7 R 298 180 1 0 0 0 0 0 0 999 Tmax Tpmax 750 10 0 0 25 0 0 249 The primary source that is used here is a sphere the radius of which equals the maximum radius of the ellipsoid The cell limited option will be chosen at runtime Runtime options The sequence of interactions with the program at runtime is screen printed below Some comments are interpolated 3 1 RRR II BBB OO RR II B B adio on eam ptimi R active ser R R II BBB 2000 2006 MSL WELCOME FEC CC AK aK aK aK Program Licensed exclusively to POE READER FE CCC CCC A ACI CK A KK version 30 06 06 full version Name of the input file test inp ifile test inp Name of the output file Beware it will overwrite this file test out READING INPUT FILE test inp 1 reading surfaces 2 reading cells number of cells 3 highest number of walls in a cell 7 number of bodies 9 reading tally 3 300 750 10 Nmax tmax tpmax 300 750 10 reading geometry storing surfaces bodies storing cells regions INPUT FILE HAS BEEN READ SUCCESSFULLY 82 CHAPTER 3 EXAMPLES BO USE
16. The output file includes a post analysis with several hypothetic ts Use 0 for noble gases 1 5 EXECUTING 3l lt 0 gt sample from law P t exp t ts ts It slows down calculations Only necessary if the number of collisions is low For ts ts X surface see examples 4 5 in the user routine customcollision f The following step decides if the module for continuous media shall be used If there is no continuous porous material only slabs or empty system etc then the Selecting target tell MESH that you have 0 regions filled with powder or fiber Otherwise input filling the number of regions were powder or fiber is present Reading the source settings M M EEEEE SSSS H H MMMM E S H H SSS HAHAAH M M E S H H M M 5555 H H STARTING MODULE FOR e diFFUSION IN POWDER How many cells contain powder e g 2 2 If a number bigger than zero is given in the example 2 then the program will pose some additional questions First it will ask which are the region numbers of the cells filled with target material what is the cell number of the powder cell 1 1 what is the cell number of the powder cell 2 2 32 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL In the example the first region is number 1 and the second is number 2 Then more information is requested In particular MESH will ask what is the average flight
17. The recommended methodology consists in sketching the geometry and number ing the surfaces and cells onto this drawing In some cases the concept under or over is not clear like with oblique planes then the guiding concept is the normal unitary vector at the surface if the cell is to the side of the normal of the surface then the sign is positive otherwise it is negative This may again seem to lead to dubious cases It should then be reminded that the gradient of the surface at a given point unambiguously defines the normal vector Remarks 1 The MC code does NOT impose any restriction to the maximum number of cells of a given problem nor on the number of limiting surfaces of a given cell Boolean logics based on intersec tions Sign defined by the gradient 15 tries for the source 12 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL The cells definitions have to be completed with zeros up to the maximum cell degree The zeros in the cells definitions have to be put at the end 1 Cells are convex elements They are exclusively defined with the boolean intersection operator and not with the union one The effort done with convgeom is directed to achieve the importation of geometry files whose format is compliant to other MC codes e g MCNPX 3 This shall enable to write the input files under those formats and to benefit from the plotting options offered by them 1 3 3 Birth
18. and therefore fitting the histogram to a smooth function may then become problematic Indeed very often the ultimate goal is to obtain an analytic function that describes the release distribution probability intrinsic release function R t Thus it may be useful to deduce this function without having to go through the cumbersome procedure of producing a histogram of the data and fitting a curve to it RIBO contains a release function fitting functionality that is based on the so called statistical momenta of the distribution The method is the following For every history the code updates a variable that contains the e g average flight time f For every new particle n this figure can be updated in this way f efus 14 n Thus the code can compute tf without need of storing t fi tf2 t fn The same can be done for the higher order momenta tf and for other magnitudes like t Po tf tPo Now if we believe that the release function behaves simply as a decaying exponential fi t C exp t t we can find the value of C and t easily by computing the momenta of f t and equalizing to those of the real distribution Statistical momenta 58 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL Moreover we impose that the distribution is normalized The parameters there after obtained are printed in the output file for example for the flight time a poss
19. dA i CpRhoxdV conv FGI CII I K K K K K K K K ELSE radia 0 0 END IF END 2 3 5 Customized the printout values userPRINT3D f SUBROUTINE userPRINT3D t N C C0 nbin Cmax Imax Rmax cellmax Cmin Imin Rmin cellmin sum sum0 integer 4 N 3 Imax 3 Imin 3 cellmax cellmin li real 8 C 300 300 300 2 C0 300 300 300 2 Cmin nbin i realx8 t sum sumO li real 8 Rmax 3 Rmin 3 fmax f min fav 4 i 46 47 48 49 50 51 74 CHAPTER 2 THREE DIMENSIONAL DIFFUSION MODULE logical fileNOTyetOPEN K K K K KK K K K K K K K K K K K FK FK K K K K K K K K K K K K K K K K K K FK K K K K K K KK K K K K K K K K K K K K K K K K K K K K K K K K K K K K CUSTOMIZE THIS FUNCTION TO FIT YOUR PROBLEM K K K K KK K ok ok K K K K K K K ck ck FK K K K K ok K K ck K K K K K K K K K ck ck K ck K K K K KK K K K K K K K K ck K K K K ck ck ck K K K K K K K K K OK K K written by MSL 2006 VARIABLES 1 Elapsed time s N 3 Number of bins in each direction x y z C 300 300 300 2 two values for each voxel 300 300 300 2 material index region number 300 300 300 1 Concentration Temperature C0 300 300 300 2 st
20. 16 1 8 4 Setting up the electro magnetic fields readEfield and Bn f These routines are called when ions are created and explicit transport is requested at runtime ELECTROMAGNETIC ION TRANSPORT You are dealing with ions which see the EM fields external or internal plasma fields IF you DO NOT know these fields but you have a rough idea about the extraction efficiency of the ions from the source you can give this number as well as the estimated extraction time ms e g 0 3 1 3 would mean that 30 of the initially ionised atoms make it to the outlet as ions The rest recombine The process takes about 1 microsecond in all IF you know these fields you can edit explicitly transport the ions by editing the functions readEfield f Bn f functions emittance f AND answering 1 0 now Ion extract efficiency 0 1 extract time ms 10 76 1 8 USER DEFINED SUBROUTINES 49 Thus by answering 1 0 this mode is activated first routine sources readE field f serves to determine the electric field vector as a func tion of time space either analitically or by reading it from an external binned file AC readEfield is subroutine of the RIBO code
21. 6 3 data are needed 1 6 4 for spheres and 1 6 8 for the rest An example of Source card In this example a Target source is used to simulate the generation of A 95 at energy 2000 with isotropic birth direction 180 in a cylinder centered at 0 8 5 0 of radius R 3 and full length L 13 cm Isotopes are born according to a Gaussian law centered at the axis with sigma 0 5 cm The axis of the cylinder is oriented parallel to 7 Source type Mass T K Alpha nx ny nz x y z R L sigma thita phi T o5 2000 180 0 1 0 0 8 50 0 3 13 0 5 90 90 A simulation for a given source gives way to various sets of events collisions flight paths If these events are sorted into histograms the resulting distribu tions can be added and subtracted to those of other simulations This permits to reproduce almost any source by performing additions and subtractions from the elementary sources Thus e g the flight path of atoms born in a cylindric ring can be obtained by subtracting the pondered flight path distribution of a small cylin dric source to that of a bigger one Normalization has to take into account the volume of the respective sources and the number of histories As a first step all In Monte Carlo jargon each individual repetition is called history and the whole group of histories is called simulation whose size will be the number of histories 1 3 INPUT FILE 15 histograms could be normalized to unity and then the
22. Dbq 7 celln T T IF celln eq 1 THEN EXAMPLE 10 kW cm 3 deposited in cell I gener 10000 0 CpRho W cm 3 K s END IF END IF END 69 2 3 3 User defined mass heat diffusion coefficients Dijkt f FUNCTION Dijkt I X t N C Dbq ncell inmode integer 4 N 3 1 3 ncell character 12 inmode real 8 C 300 300 300 2 X 3 Dbq 7 ncell t integer 4 celln real 8 CorT CpRho k real 8 Dijkt PK ck ck ok ok ok ck ok ok ok 2K K K FK DK DK K CE K K ck Dk 0k ck ck ck ok Dk K ck ck 2K CK CK ck ck Dk K ck ck ck ck ok ck ck ck K K K K 2K ck ck ck ck K K K ck ck ck Dk K K K K K K K ck K 2K K CUSTOMIZE THIS FUNCTION TO FIT YOUR PROBLEM 20 21 22 23 70 CHAPTER 2 THREE DIMENSIONAL DIFFUSION MODULE written by MSL 2006 This function computes the DIFFUSION coefficient CONDUCTIVITY for a given boxel that corresponds to position X and depending on the CONCENTRATION TEMPERATURE parsed constants Dbq and cell number VARIABLES MASS transfer HEAT transfer inmode 5 6 CpRho Cp xRho W s
23. Frenkel because here we check after a bunch of collisions actually when the particle exits the powder cell and NOT after each collision This will not imply any big error central limit theorem HIGHLIGHTS simulation of COLD spots THIS ROUTINE IS ALWAYS READ THROUGH VARIABLES INPUT celln Cell region from input file COLLPOW Number of collisions in the powder cell in the last passage through ts average sticking time per collision given at runtime 1 00 tdep Particle elapsed desorption time spent in powder domains s EXTERNAL rand Seed for random generation EXAMPLE 1 cells 1 3 5 contain powder the first one is warmer than the second and that one is warmer than the third 47 68 69 70 71 73 74 75 48 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL IF celln eq 1 THEN tdep tdep 1 0 15 ELSE IF celin eq 3 THEN tdep tdep 2 0 15 ELSE IF celln eq 5 THEN tdep tdep 3 0 15 END IF END 1 8 3 User defined source distributions customsource f This routine called at particle generation is explained in 1 3 3 on page
24. a exp t 100 0 999 Dz 1 r 2 r x a Dc IF t le 1 THEN Dz7 7778 Dc END IF END For space and time variations in D 109 b Options for 1 D calculations Figure 4 1 Screen captures of the interactive DIFFUSE program 110 Reference List 1 2 3 4 5 6 7 8 L Maunoury O Bajeat R Lichtenth ler and A Villari Temperature sim ulations for the SPIRAL ISOL target Nucl Inst and Meth B 2001 sub mitted Lichtenth ler et al A simulation of the temperature distribution in the SPIRAL target Report GANIL Caen August 1997 J S Hendricks et al MCNPX version 2 5 e Manual Los Alamos National Laboratory New Mexico 2004 mcnpx lanl gov Mokhov The MARS Code System User s Guide Version 13 95 Ref erence Manual Fermilab 1995 N V Mokhov Status of MARS Code Fermilab fermilab conf 03 053 edi tion 2002 www ap fnal gov MARS Fasso A Ferrari and Sala Electron photon transport in FLUKA status In F Barao M Nakagawa L Tavora and P Vaz editors Proceedings of the MonteCarlo 2000 Conference Lisbon October 23 26 2000 pages 159 164 Springer Verlag Berlin October 2001 Fass A Ferrari J Ranft and Sala FLUKA Status and prospective for hadronic applications In A Kling F Barao M Nakagawa L Tavora and P Vaz editors Proceedings of the MonteCarlo 20
25. and cell 3 and then open up there and close between cell 1 and cell 2 t 0 10 s 22 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL l 1 2 gt 1 2 QN 3 1 3 1 The corresponding definition would be start time t ramp time t ramp surface 111 112 0 10 0 1 1 3 10 0 0 0 1 1 2 REMARKS 1 If end time is smaller than start_time then the program assumes that the wall will no longer open after start time 1 The surface that acts as a valve needs to be defined in the input file and it must divide two existing regions 1 The user must specify between which cells the wall will be acting the order 15 irrelevant 1 The user can define up to 10 moving wall events 1 A wall can appear in several lines if there are several events concerning that wall 1 If the end time is smaller than the start time then the program assumes that in fact the end time is infinite 1 5 EXECUTING 23 1 5 Executing RIBO 1 5 1 The Isotope RElease Simulator IRES RIBO can be run on line from a server at www targisol csic es for simplified geometries and through a GUI front end that permits easy operation of the code The server contains a wide compendium of release data diffusion coefficients and sticking times which is used in combination with this on line version The server administrator requires registration 1 5 2 Expre
26. cylinder of the same radius and then by using the cell delimiter command The efficiency of the method corresponds to the fraction of the cylinder sector here it is 25 96 Surfaces n RC T 2 Y2 72 XZ YZ X Y Z C 1 0 5 298 0 0 0 0 0 0 1 0 0 2 0 5 298 0 0 0 0 0 0 1 0 0 10 3 0 5 298 0 1 1 0 0 0 0 0 0 1 16 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL 4 0 5 298 0 0 0 0 0 0 0 1 0 0 5 0 5 298 0 0 0 0 0 0 0 0 1 0 Cells n S1 S2 1 1 2 3 4 5 2 1 2 3 4 5 3 1 2 3 4 5 4 1 2 3 4 5 Source Type M T Alpha nx ny nz x y 7 L 5 th phi 40 298 18 1 0 0 5 0 0 10 0 0 0 5 Nmax Tmax Tpmax 2 1000 110 101 The cell delimited source command is more powerful than the composition additions or subtractions of simulations and can simulate more complicated dis tributions e g a ring can be simulated by both ways but a cylinder sector is only produced by the cell restriction as just shown Customized sources programmable option The instructions given up to this point already offer a a fairly high control of the geometry distribution of the starting atoms and also give the possibility to choose the initial speed the direction and the semi aperture angle of an isotropic emis sion cone However the user may want to define fancier distributions correla tion of velocity and position coordinates speed distributions position dependent weights the possibilities are
27. equivalently the two angles of the new 2 axis and to make a translation Finally the temperature of the surface is introduced and then the user can continue to define the second surface and so on The program writes into the input file the Surfaces card including the headers For example gt convgeom what is the input file name INPUT surface number 1 P plane S Sphere C Cylinder K cone SQ ellipse GQ gen YZ PX XZ PY XY PZ AX BY CZ D PX position 1 O 5 Ws 5 02 12 5 i rotation give polar angles of new x vector polar azimuthal 90 0 gt no rotation 90 45 rotation 90 45 alpha beta 0 0 0 0 0 0 0 707107 0 707107 0 1 translation give 0 0 20 110 0 0 20 1 32 0 0 0 20 0 0 0 0 0 0 0 707107 0 707107 0 2 414214 what is the temperature of the surface K 2000 add surface Y N Y surface number 2 P plane S Sphere Cylinder K cone SQ ellipse GQ gen 10 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL Radius eq 0 1 1 0 0 0 0 0 0 4 rotation give polar angles of new x vector polar azimuthal 90 0 gt no rotation 90 30 rotation 90 30 X alpha beta 0 250000 0 750000 1 0 866025 0 0 0 0 0 4 translation give 0 0 20 000 0 250000 0 750000 1 0 866025 0 0 0 0 0 4 0 0 20 0 0 0 X x0 y0 z0 0 250000 0 750000 1 0 866025 0 0 0 0 0 4 wha
28. infinite Responding to these potential needs the open source routine customsource f found in the directory SOURCES allows to create user define source distributions This subroutine deals with the follow ing I O variables x y z ux uy uz tp SOURCE where SOURCE is the vec tor that stores the parameters introduced in the input file in the card Source Thus the user can specify whichever dependency g x y z ux uy uz tp SOURCE For instance for a discrete generation profile over where the probabilities are pill position_y production probability Probability 1 2 08 2 85 0 19 0 19 1 3 INPUT FILE 17 2 4 22 2 63 0 17 0 36 3 6 36 2 41 0 16 0 51 4 8 49 2 20 0 14 0 66 5 10 61 1 98 0 13 0 78 6 12 73 1 76 0 11 0 90 7 14 83 1 55 0 10 1 00 Then the user should write the following instructions in customsource f a rand zero IF a gt 0 9 THEN y 14 83 ELSE IF a gt 0 78 THEN y 12 73 ELSE IF a gt 0 66 THEN y 10 61 ELSE IF a gt 0 51 THEN y 8 49 ELSE IF a gt 0 36 THEN y 6 36 ELSE IF a gt 0 19 THEN y 4 22 y 2 08 WARNING Remember that any changes in the source files will take place only after having recompiled The compilation instruction appears in the README file on the RIBO distribution and can be performed with the script tools make sh Reading source events from a file The
29. mathematical functions in math f Surface ionization is done separately EXAMPLE 1 snell reflection if gt 0 5 otherwise perpendicular 43 zero 0 0 IF rc gt 0 5 THEN k V3scaV3 grad3 U3 k grad3 1 U3 1 grad3 2 3 2 grad3 3 U3 1 U3 1 2 grad3 1 U3 2 2 U3 2 2 k grad3 2 U3 3 U3 3 2 k grad3 3 COL6 5 COL6 5 1 0 ELSE U3 1 grad3 1 U3 2 grad3 2 40 41 42 43 44 45 39 0 47 44 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL U3 3 grad3 3 COL6 6 COL6 6 1 0 END IF EXAMPLE 2 If the incidence angle is within a cone of 60 deg around the surface normal then reflect isotropically within that cone otherwise use mirror like reflections cosphi V3scaV3 U3 grad3 norm U3 norm grad3 b a b lt 0 IF cosphi 1t 0 5 THEN phi gt 60 deg k V3scaV3 grad3 U3 k grad3 1 U3 1 grad3 2 U3 2 grad3 3 U3 1 U3 1 2 x k grad3 1 U3 2 2 U3 2 2 k grad3 2 U3 3 U3 3 2 x k grad3 3 COL6 5 COL6 5 1 0 ELSE phi 60Ndeg CALL isotropic grad3 U3 60
30. molecular flow and the intermediate regime Cosine law e Effusion through porous media 1 Features of RIBO 2 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL e Effusion under multiple path passage conditions and surface crossing sense detection NEW FEATURE e Absorption condensation to the walls Definition of cold spots e Adsorption desorption on the walls temporary retention Frenkel sam pling NEW FEATURE e Specular effusion e Custom collisions NEW FEATURE e Effusion through crystal systems vibrations e Effusion through systems with moving walls valves e Surface ionization e Plasma ionization e Ion recombination e transport in electric and magnetic fields Emittance plots e Enhanced compilation simulation speed increased up to 40 NEW FEA TURE Other models are in consideration and could be implemented as a result of the common effort of a community of prospective users Developments should be centralized through RIBO s web page www cern ch ribo Under development 1 2 SETUP 3 Moreover users are invited to customize their RIBO distribution by modifying the provided routines in directory sources explained in page 1 8 e Source of Radioisotopes customsource f Customized collision law customcollision f Special desorption in powder or fiber powderdesorp f Magnetic field map Bn f Electric field map readEfield f User defined on line printing userprint f User de
31. of particles the Source card Predefined sources Source card input file card The Source card has to 12 14 entries that describe the source atoms their start ing position and the velocity distribution The first entry is a character that encodes the geometric distribution shape of the generated atoms The following vector of 14 numbers gives details of the source mass number A S 1 tempera ture S 2 K semi angle a of aperture of the luminous cone 5 3 with respect to the central direction of emission S 4 S 5 S 6 the birth coor dinates centroid 5 7 S 8 5 9 and details regarding the shape of the distribution Coordinates from S 7 to S 14 are explained below 1 Point source Only three geometric parameters are needed 5 7 x0 S 8 yo S 9 zo 2 Spherical source Like the point source with a radius zo yo 20 S 10 1 3 INPUT FILE 13 3 Box source Particles are sampled within a parallelepiped centered at xo yo 20 with sides of full lengths cm S 10 La S 11 Ly 5 12 L and oriented in space with the angles 5 13 0 S 14 y 4 Target generation The starting position is sampled inside a cylinder cen tered at zo zo zo of radius R full length S 11 L Gaussian radial dispersion cm S 12 and orientation angles of the target cylinder with respect to 2 of 13 0 S 14 Some remarks have to be
32. the sources custom3D f file Custom routines kW 43 radiation function starting temperatures The specific options 10 are already implemented in the open source functions and subroutines in sources custom3D f You are invited to look through them and identify the different choices Runtime options NOTE The input values are stored in batch file batchinputs heat3D batch RRR II BBB 00 RR II B B RRR on eam ptimi R active ser RR II BBB OO 2000 2006 MSL WELCOME Program Licensed exclusively to MANUAL READER FEC CCCI CK A OK aK a6 2K version 30 06 06 full version Name of the input file heat3D t 100 CHAPTER 3 EXAMPLES RIBO USE ifile heat3D t Name of the output file Beware it will overwrite this file heat3D out READING INPUT FILE heat3D t 1 reading surfaces 2 reading cells number of cells 2 highest number of walls in a cell 2 number of bodies 3 reading tally Nmax tmax tpmax 50000 750 10 reading geometry storing surfaces bodies Storing cells regions INPUT FILE HAS BEEN READ SUCCESSFULLY SOURCE CUSTOMIZATION You can edit the file sources customsource f and then recompile with tools make sh Additionally you can use the data file init dat by uncomment
33. the user during execution but in general terms these elements can be out marked Heading It includes authoring information and specifies the input file name and path Individual scores This section may contain the global and fractioned release time and total number of collisions for every simulated atom This usually constitutes the core goal of the simulations since it permits to recon struct the intrinsic release functions Preconfigured histogram This histogram unnormalized to bin size appears if the card Histogram is included in the input file Average figures of the release speed This group of data includes the aver age release time the relative time consumption in the diffusion and effusion phases the amount of in grain versus inter grain diffusion time In pres ence of ionic fluxes this output is divided in two groups for ions and for neutral atoms This allows assessing the impact of ionization in the effusion path and extraction efficiency Fitting of events to a release function see section 1 9 1 Estimated release fractions for given parameters and for conditions other that the ones specified i e different sticking times and diffusion time con stants explained in section 1 9 2 56 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL Average effusion time in each cell It permits to see which elements are slowing down effusion Average number of collisions to each surface This array pro
34. user can also read values from the file init dat in the DATA directory of the RIBO distribution This file must have been previously generated by another run with RIBO or by any other program e g FLUKA The file init dat can have whatever format you wish as long as the reading instructions in customsource f are coherent with the data in the file The variables to play with are the 6 space 18 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL coordinates starting time For example to read x y x ux uy uz tp from init dat just uncomment the following line in customsource f and recompile read 9 x y Z ux uy uz tp 1 3 4 Tally end card The end conditions are preceded by a line with the word Tally There are four inputs the first and last are relative to the individual histories and the two inter mediate fix the end conditions for the global simulation Up to now an end Tally would look like this still valid Simple Tally card Tally 5 Nmax Tmax Tpmax 3 6000 750 10 The text and the case of the first line are unmodifiable the second line is indicative and should not be omitted although it can be edited at the user taste The first number in the last line indicates the surface used as detector for the atoms When atoms reach this surface the history is completed and a new atom is simulated The fourth number alternatively terminates the current history if the individual fl
35. y z cm grad 3 Surface gradient normal at collision point always pointing inwards and normalized T Starting energy of the atom source card A Atomic mass uam source card of input file surn Surface number from input file rc First parameter of the surface after surface index number e g roughness coefficient ts Average sticking time per collision if it is constant all over it is normally define as 0 in the runtime options The output file then includes a sensitivity analysis for values other than 0 1 00 03 3 Incoming Outgoing velocity ux uy uz of the atom m s COL6 6 Counter for the type of collision for each history x 1 Maxwell type 1 2 Maxwell type 2 32 33 34 35 36 38 39 1 8 USER DEFINED SUBROUTINES 3 Mirror like 4 Debye semi classic 5 Custom 1 6 Custom 1 tde Particle desorption time s Normally desorption time in the collisions is taken into account at the end of each history as ts COLL If you use a special desorption time distri bution examples 4 5 update the value of tde as shown in those examples AUXILIARY U3 3 Base of auxiliary vectors to build up reflection phi alpha Auxiliary angles to build reflected direction EXTERNAL rand Seed for random generation NOTES Check available
36. 0 0 ELSE CorT C I 1 I 2 I 3 1 Temperature k Dbq l celln Dbq 2 celln CorT Dbq 3 celln CorT CorT CpRho Dbq 5 celln Dbq 6 celln CorT Dbq 7 celln CorT CorT Dijkt k CpRho END IF END IF END 2 3 4 Customized heat radiation and convection function radia f FUNCTION radia i M T C N R PAR Dbq dA ncell integer 4 i ncell N 3 li real 8 M 3 T PAR 5 dA 3 Dbq 7 ncell C 300 300 300 2 R 3 real 8 emissivity sigSB eps T0 dV radia CpRho norm a integer 4 celln a K K K SK K SK K 2K K K CE SK K K 2K FK SKK 2K CER CE 2K CE SK E 2K FK SR CE K CE K 2K FK K CE CK K 2K K 2K SE 2K 2K K 2K K 2 K CE K 2K K FK CE SK 2K 2K K 2K 2K CK E K 2K K 2K E K K K THIS SUBROUTINE checks if we are in a heat radiation problem inmode 6 and if so it computes the radiated power Qr W as a function of the emittance eps the exposed area dA i and the temperature difference 0 TO0zexternal VARIABLES i exposed area normal vector i j k M 3 material of neighbouring cell should be vacuum 0 T Temperature of the boxel Temperature i j k celln N 3 number of divisions in x y z 31 33 34 35 36 37 12 CHAPTER 2 THREE DIMENSIONAL DIFFUSION MODULE X 3 position of center of boxel x y z 1 dV dx dy
37. 00 0 00001 803 0 117 0 0 00211 0 00213 0 4 193 1 000 0 402 0 00000 0 00357 225838 4 36 0 0 00290 0 00067 0 3 952 1 000 0 951 0 00000 0 00008 4792 8 118 0 0 00198 0 00206 1 3 891 0 548 1 000 0 00000 0 00043 27154 5 132 0 0 00191 0 00234 1 4 316 0 185 1 000 0 00000 0 00420 265829 4 124 0 0 00222 0 00198 The first part contains the table with the individual events Outgoing particles are labeled with 1 ions or 0 neutral atoms The caption explains the different columns e g by multiplying COL and COLP with an average sticking time one can get the total desorption time 90 CHAPTER 3 EXAMPLES OF RIBO USE GENERAL FIGURES FOR NEUTRAL ATOMS Average intrinsic delay time s 0 0050 Slowest particle took 0 0338 s Average particle time consumption 0 0000 Diffusion 100 0000 Effusion AVERAGE DELAY in POWDER s 0 0035 DIFFUSION DISTRIBUTED AS 0 0000 grain diffusion 100 0000 intergrain GENERAL FIGURES FOR IONISED ATOMS Average intrinsic delay time s 0 0049 Slowest particle took 0 0292 s Average particle time consumption 0 0000 Diffusion 100 0000 Effusion AVERAGE DELAY in POWDER s 0 0035 DIFFUSION DISTRIBUTED AS 0 0000 grain diffusion 100 0000 intergrain This part of the output gives the effusion to diffusion and vacuum to powder time share Since effusion was not explicitly activated mode I or mode 3 100 of the release time is due to effusion Obviously th
38. 00 Conference Lis bon October 23 26 2000 pages 955 960 Springer Verlag Berlin October 2001 CERN Physics Analysis Workstation CERN paw support cern ch wwwasd web cern ch wwwasd paw 111 9 Persistence Of Vision Pty Ltd Persistence Of Vision Pty Ltd Williamstown Victoria Australia 2004 Software 10 B T Phong Illumination for computer generated images ACM 18 6 311 317 1975 11 M Santana Leitner A Monte Carlo code to optimize the production of Radioactive Ion Beams by the ISOL technique PhD thesis Technical Uni versity of Catalonia 2005 12 S Agostinelli et al Geant4 a simulation toolkit Nucl Inst and Meth A 506 250 303 2003 wwwasd web cern ch wwwasd geant4 geant4 html 13 H M Kalos and A P Whitlock Monte Carlo methods New York Wiley amp Sons 1986 14 M Fujioka and Y Arai Diffusion of radioisotopes from solids in the form of foils fibers and particles Nucl Inst and Meth 186 409 412 1981 15 Fick Ann Phys 94 59 1855 112
39. 0000 500 0000 1 2 9750 0 3750 0 0750 298 0000 2 4 4750 0 4250 0 1750 1 0000 0 136000 508 0000 1 222 025 0 3750 0 0750 298 0000 2 4 4750 0 4250 0 1750 1 0100 Be OBR EE 1 052380 561 9047 1 1 9250 0 0750 0 0250 297 4828 2 2 5250 2 3750 2 3750 1 0774 1 097014 564 5302 1 1 8750 0 0750 0 0250 293 7514 2 2 5250 2 4250 2 3750 1 0807 CALCULATION FINISHED DONE Results are stored in heat3D out Ribo initializes the 3D grid It then asks for the termination condition In this case the simulation will finish when the maximum temperature reaches 700 K or when the computational time reaches Tmax 750 s Output file interpretation The resulting output file displays the following information results after processing input file heat3D t preprocessing cpu time s 15 time elapsed time s Cmax Maximum value it can be multiple IM cell which Cmax belongs XM YM ZM coordinates of one of the Cmax Cmin Minimum value it can be multiple Im cell to which Cmin belongs Xm Ym Zm coordinates of one of the Cmin lt C t gt Average value of C t CO Relative increase or decrease lt C gt 3 1 EXAMPLES 105 time s Cmax IM XM YM ZM Cmin Im Xm Ym Zm lt C t CO gt lt C t gt 0 000000 500 0000 1 2 9750 0 3750 0 0750 298 0000 4 4750 0 4250 0 1750 1 0000 398 752 0 136000 508 0000 1 2 9
40. 2 s ze S The units are specified Fractional concentration Insert the remaining concentration e g 0 5 would mean that 50 96 has diffused out Insert the lapse length s It refers to the final time or the time at which the fractional concentration is measured Time definition Insert the printing time grid number intervals The time steps for the diffusion calculations are adjusted internally to keep the system stable but the time steps are often much too small to have them all printed Thus the user specifies what should be the time binning For the present status of the program the user needs to introduce the wished time dependencies directly in the corresponding subroutines SUBROUTINE timeN N i m t dt a real 8 N m 2 t dt tao pulse a x integer 4 i m decay law tao 20 097 tao 227 0 N i l zN i l exp log 2 0 dt tao pulsing source x a m i 1 0 5 IF mod t pulse 1t 1 0001 dt and t gt pulse THEN N i 1 N i 1 functC x a x 1 END IF END For time dependencies or FUNCTION functC x a 8 x a 1 10 0 1 1 0 1 1 3 1415927 2 functC 1 END For space variations of the concentration or finally FUNCTION D i m a t Dc realx8 x a t Dc r integer 4 i m x da i 0 5 D Dex sin x 3 1415927
41. 250 0 3750 0 0750 298 0000 4 4750 0 4250 0 1750 1 0100 402 743 0 274078 516 1242 1 2 2750 1 5250 1 1750 297 9584 3 0750 1 9750 1 6750 1 0202 406 794 1 046866 561 5804 1 1 9250 0 0750 0 0250 293 8229 2 5750 2 3750 2 3750 1 0770 429 468 1 052380 561 9047 1 1 9250 0 0750 0 0250 297 4828 2 5250 2 3750 2 3750 1 0774 429 630 1 097014 564 5302 1 1 8750 0 0750 0 0250 293 7514 2 5250 2 4250 2 3750 1 0807 430 940 Number of steps 50 CPU time s 765 CPU time step 15 3 s step CPU time sim time 693 922836 s CPU s diffusion Mario Santana Leitner ISOLDE CERN 2001 2006 2 The output file for these type of calculations is rather simple to interpret In output mode 8 the elapsed diffusion time is printed in the first column The highest found temperature the cell and the coordinates of a boxel with such temperature there may be many with the same value are printed in columns 2 6 The same information is displayed in columns 7 11 for the minimum value Finally the forlast column tells about the ratio of the average Temperature at time t and that 106 CHAPTER 3 EXAMPLES RIBO USE at time 0 and the leftmost column shows the evolution of the average temperature with time The file ends with some numbers related to the computational speed The primary source that is used here is a sphere the radius of which equals the maximum radius of the ellipsoid Th
42. 3 the extraction tube 7 0 2 5555 U U II 5 U U R R II OO OO NN N SSS U U RRRP FFF O O NNN S UUUU RR F II OO OO N NN SSSS UUU R R F II N N This program computes the probability of sur is used Parameters are the ionization work function and ionic statistical weights Wi g0 g g for an atom Z They are stored in the outer database sion dat The substrate work function is stored in workf dat written by Mario Santana Leitner CERN 2003 Nmaxl Nmax2 102 27 atom face ionization after each single atom surface collision Theory summarized in KOE2000 p 223 gt gt element 7 0 if you want to define parameters substrate element 7 0 if you want to define parameters 4 0 element 1 Boride 2 Carbide 3 Oxide 4 CeCompound gplus gzero gminus Wf Wi Ae 1 2 1 4 54 5 39 T K 000 positive surface ionization alphaS 0 00360563289 0 62 84 CHAPTER 3 EXAMPLES RIBO USE negative surface ionization alphaS 6 61641972 11 betaS 0 003592679 ELECTROMAGNETIC ION TRANSPORT You are dealing with ions which see the EM fields external or internal plasma fields IF you DO NOT know these fields but you have a rough idea about the extraction efficiency of the ions from the source you can give this number as well as the estimated extraction time ms e g 0 3 1 3 would mean that 30 of the initially io
43. 4 Setting up the electro magnetic fields readEfield and Bn f 1 8 5 User defined output printing files userprint f 1 86 Other user definedroutines L9 s e bee eo amp od bee RSE oh a x 4 1 9 1 Fitting of events to a release function 1 9 2 Computation of release fractions Three dimensional diffusion module Dif fuse3D 2 Brief physical introduction 2 2 221 data Cond datfile 2 42 244224 424244405 2 2 20 Activating 3 D diffusion calculations 2 3 User defined functions and subroutines 2 3 1 Customizing the starting distribution of Concentration or Temperature CSTARI 2 3 2 Defining source sink terms generf 2 3 3 User defined mass heat diffusion coefficients Dijkt f 46 48 48 22 68 69 2 3 4 Customized heat radiation and convection function radia f 71 2 3 5 Customized the printout values userPRINT3D f 24 Examples of RIBO use S L Bxamples oo uoo o dk Boe Ro 3 1 1 First example Geometry issues 3 1 2 Second example Bigger files 3 1 3 Third example test inp 3 1 4 Example of use of the 3D Diffusion 3D GRID heat3D t Diffuse A diffusion emulator Reference List i
44. 451 or the temperature K and J is the cm3 par cm s flux of particles 25 or of heat power 2 in the direction determined by f The constant a corresponds either to the diffusion coefficient D or to the thermal conductivity k Z gt RIBO divides the space in finite units of volume dV boxels the dimensions dV dx of which are determined by the user Each boxel will be labeled with its initial concentration or temperature and the cell number to which it belongs if any Time is discretized in time steps the length is adjusted dynamically Each boxel exchanges atoms or heat in the direction i through the interface area dA 1 Fick Law dx that communicates to the next boxel This is done by mass diffusion solved evenly meshed space 61 62 CHAPTER 2 THREE DIMENSIONAL DIFFUSION MODULE head conduction flows or by desorption or radiation if the communicating boxel belongs to vacuum For a given boxel i 7 k the balance of U in the interval dt expressed Carte sian coordinates is dU g d s dt 2 2 Where e gis the source term 23 for atomic diffusion J 2 3 J for heat calculations cm s e dis the diffusion or conduction flow term SIS 2U Ua dt 3 a dV 2 4 2 The constant D is mass or heat diffusion coefficient 91 Note that 5 the thermal diffusion coefficient is ob
45. 8 987272 97 813 93645 75 731 1 2 99 051 98 980 98 977 98 949 98 672 96 048 80 286 99 673 99 649 99 648 99 630 99 545 96 627 90 955 6E 2 99 830 QO 618 OO ris OO SOS 1 3 99 886 99 886 99 883 99 855 99 574 96 924 D wc GOD GO OQ Qo Doc oO ts s tao_d s 0 10E 1 0 0 0 0 1 8 0 1 7 0 1E 6 0 1 5 0 1 4 T 1 2 s DRF RF tsl RF ts2 RF ts3 RF ts4 RF ts5 RF ts6 0 1 2 2 481 0 226 0 246 0 000 0 000 0 000 0 000 0 3E 2 4 279 1 164 1 648 0 000 0 000 0 000 0 6 2 6 020 2 646 2 624 2 665 0 000 0 000 0 000 0 1E 1 4 425 4 377 4 141 0 000 0 000 0 000 0 3E 1 13 126 10 564 10 487 9 902 10 362 0 000 0 000 0 6 1 18 198 16 241 16 170 15 588 13 111 0 000 0 000 0 1E 0 23 022 21 473 21 413 20 900 17 793 0 000 0 000 94 CHAPTER 3 EXAMPLES RIBO USE 3E 0 37 213 36 342 36305 35 982 33 275 32 037 0 000 6 0 48 936 48 357 48 332 48 111 46 092 37 418 0 000 1 1 58 506 58 088 58 070 57 909 56 384 47 213 0 000 3E 1 78 059 TIRTS 772664 TETTIE F099 67 799 87 058 86 954 86 949 86 909 86 504 82 817 66 975 1E 2 91 622 91 557 91 554 91 528 91 271 88 844 74 264 2 96 960 96 937 06 936 96 927 96 836 95 943 88 480 6E 2 98 441 98 429 98 428 98 424 98 377 97 919 93 738 2 D O an m L 1 3 99 051 99 044 99 044 99 041 99 013 98 735 96 107 The exponential fitting functions
46. AL 33 105 106 107 108 109 110 111 112 54 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL rand Seed for random generation OUTPUT UNITS UNITS 21 25 are available for the user EXAMPLE 1 write to unit 21 the final and initial radius in a cylinder oriented with axis 2 write 21 FMT IX 2F9 6 sqrt X3 1 2 X3 2 2 sqrt X03 1 2 X03 2 2 EXAMPLE 2 write to unit 22 the number of collisions of each type write 21 FMT 1X 8F9 2 COL COLP COL6 sqrt X03 1 2 X03 2 2 END 1 NOTE A special routine exists to individualize the printing for 3D diffusion problems Please consult page 73 1 8 6 Other user defined routines RIBO distribution includes some other user customizable FORTRAN func tions and subroutines Indeed the directory sources contains also the open rou tines for the 3 D diffusion calculations of 3D GRID explained in chapter 2 and printed in 2 3 113 114 115 116 1 9 OUTPUT FILE 55 1 9 Output file The output file contains information about the run average numbers and variables linked with every history Its shape depends on the requests made by
47. ANUAL e All groups should at least have two elements so that no cell is isolated from the rest Every row of the connectivity matrix must have at least two numbers different from zero If the source of particles were entirely located outside the system or if the cell source contained some errors then a message on the screen would clearly warn the user Error in source card source out of cell domain This may happen due to an incorrect implementation of the source or to some flaw in the definition of the geometry of the system In some cases the errors of geometry are detected by a routine of RIBO that casts a message on the screen like GEOMETRY ERROR check cell 2 or surface 4 However many errors are not traced by the code a vast geometry case compiler is still a pending task for future upgrades of the program For the time being one more tool is available for debugging purposes At run time option 7 gives the opportunity to track a particle from birth to termination This tracking can be made effective in three subsequent choices Choose one option l coordinates x y z 2 cell history 3 surface history Normally this is enough to detect persistent geometry mistakes f that were the case an error message would show on screen and in the CCONM file 1 6 DEBUGGING RUNNING ERRORS 39 In addition to all this the user can change the end surface and choose closer surfaces e g instead of
48. E CARLO CODE RIBO USER MANUAL VARIABLES INPUT X3 3 Absolute position x y z cm U3 3 Velocity ux uy uz of the ion m s AUXILIARY R 2 Pi 3 141592 OUTPUT epsilon 2 Emittance in perpendicular directions R 2 Transverse position of the beam axis cm 3 14159257 Modify this according to the exit axis and position R 1 X3 1 0 0 R 2 Z3 3 1 0 Safeguard condition IF U3 3 eq 0 0 THEN write 6 Warning extraction axis perpendicular to velocity U3 3 z1E 12 END IF write 4 IX F8 4 F9 5 F8 4 F9 5 R 1 U3 1 U3 3 R 2 U3 2 U3 3 epsilon 1 epsilon 1 10 R 1 1000 atan U3 1 U3 3 Pi epsilon 2 2 epsilon 2 10 R 2 1000 atan U3 2 U3 3 Pi END At this point the program has all necessary elements to pursue simulations even tually including in grain diffusion in slabs particles or fibers inter grain diffusion through powders or fibers effusion in molecular or intermediately pressurized sys tems with mirror like walls diffusive walls or crystals and ionization in plasma chambers or surface ionizers After preprocessing some messages will appear in 24 25 26 27 28 1 6 DEBUGGING RUNNING ERRORS 35 the screen and simulations
49. Ez Erxz abs r lE 10 E 1 Ex E 2 Ey 84 85 86 87 88 89 90 1 8 USER DEFINED SUBROUTINES 51 3 Ez 91 KK FK FK K K K K K FK KK K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K 2K OK K 1 0 0 2 0 0 E 3 20 0 92 END As for the Magnetic field PK ck ck ok ok ok K K ok ok ok ck K CK FK FK K K CE K K ck Dk 0k ok ck ck ok K K ck ck ck ck ck ck ck K K ck ck ck ok ok ck K K K K K K ck ck ck ck ck K K K ck ck ck Dk K K K K K k K ck OK K K Defining Electic Bn is a function of the RIBO code CUSTOMIZE THIS FUNCTION TO FIT YOUR CONDITIONS Pk K ck ok ok ok ck ok ok ok ok ck ck ck ck ck ck K K K K K K ok ok ck ck ok ok K ck ck K ck ck ck ck K K ck ck ck ck ck ck ck ck K ck ck K ck ck ck ck K K K ck ck ck ck K K K K K K 2K K K K K 93 BBBB B B BBBB B B BBBB 4 FK K K K K K K K FK FK K K K K FK FK KK K K K K K K K K K K FK K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K DEPENDENCIES m Bn 95 l B S explicitEM S main f S 96 MEANING Bn is the nth component of the magnetic field Thus B B1 2 21 VARIABLES 0 4 space time four vector X0 m m m s 98
50. K cm 3 k conductivity W K cm Dijkt Diffusion coeff D cm 2 s k CpxRho cm 2 s Ce sae concentration celln Temperature celln Dbq diffusion constants conductance emiss Cond dat t elapsed time s 3 number of divisions in x y z I 3 boxel identifier i j k X 3 position of center of boxel x y z ncell total number of regions in geometry NOTES A quadratic law D C and k T are assumed but other laws can be implemented in this function The user may introduce fancy features now for example 1 Time dependency of diffusion coefficient Dijkt sin t 2 X dependency of diffusion coefficient PK ck ck ok ok ck K ok ok ok ok ck K K K K ck ck ck K ck K ok ok ok ck ck ok ok ok ok K K K K ck ck K ck ck ck ok ck FKK ok ok K ck K K CK K ck ck ck ck ck ck ck ck K ck K K K K K K OK OK K OK celln 1 1 1 2 1 3 2 IF inmode eq 5 THEN IF celln eq 0 THEN Dijkt 0 0 ELSE CorT C I 1 I 2 I 3 1 Concentration Dijkt Dbq l celln Dbq 2 celln CorT 24 25 26 27 28 29 2 3 USER DEFINED FUNCTIONS AND SUBROUTINES 71 Dbq 3 celln CorT END IF ELSE IF inmode eq 6 THEN IF celln eq 0 THEN Dijkt
51. Laplace Transformed and folded with the dif fusion release efficiency for several diffusion parameters The global release effi ciency factor as a function of the half life and sticking coefficient s are printed in tables Every table corresponds to a different diffusion time constant average effusion time in each cell 96 time share 1 10 344 2 88 634 3 1 022 average number of events each surface surface collisions lt absorptions gt lt eff gt 1 6 948 0 000 0 000 2 4 101 0 000 0 000 3 0 000 0 000 0 000 4 19 103 0 000 0 000 5 19 083 0 000 0 000 6 19 267 0 000 0 000 19 514 0 000 0 000 8 0 000 0 000 0 000 9 0 000 0 000 0 000 je meee eee eee CONNECTIVITY gt share 1 3 1 0010 These 3 tables tell about the effusion time share in the different regions 1 the number of collisions and absorptions in each surface 2 no particle collides with 3 1 95 last surface because they are forced to cross it and the connectivity between the starting regions and the end region 3 In this case they are all born in cell 1 because we forced that and they all die when crossing surface 9 which belongs to cell 3 SURFACE IONIZATION surface collision ionization probability 0 003592 ionization efficiency 19 40000 relative error 6 44565 IMPLICIT ION TRANSPORT Ion extraction efficiency 0 2 this factor is in
52. Moreover the output file includes a set of tables with the estimated release fraction for the input parameters diffusion coefficient sticking time for different half lives as well as for a variety of other diffusion and sticking time constants Briefly the way to obtain these values consisted in recording on line a covari ance matrix with the statistical momenta of the effusion flight time t f tPo and of the number of collisions COLP COL and crossed terms up to degree 3 Then for any sticking time ts a 2 exponential effusion release function like fo t1 t2 t could be fitted let us now call it E t4 t5 ts t in the way described in 1 9 1 The global release function R t would result from folding diffusion and effusion R t D rp geometry t 69 Et tz ts t 1 2 And the release fraction for a half life 2 would be RE 1 3 This is equivalent to making the Laplace transform of RF RF t o TORIS 1 4 5Which depends on the diffusion time constant and the geometry foil fiber particle 60 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL The key point is that the Laplace transform of a convolution of two functions D Q E is the product of the Laplace transforms of each function so the global release fraction RF is the product of the diffusion release fraction D RF and the effusion release fraction E RF The release efficiency
53. Particle tracking options x cell surf Only events in the POWDER FIBER D Ss Cn E gt Use sources userPRINT f routine COLLISION MODEL I What model for the treatment of the collisions Specular lt lt Ss B Knudsen Lambert recommended cosine law y Debye semi classic under development Custom BACK Go back fo previous menu For option C Custom modify the subroutine sources customcollision f at your convenience and recompile using tools make sh COLLISION MODEL II Use average energy or sample from coll law 86 CHAPTER 3 EXAMPLES RIBO USE Y Use average energy recommended N Sample from collision law slower BACK Go back 10 previous menu Y COLLISION MODEL III Sticking time s gt 0 gt every collision delays exactly ts s 0 recommended The output file includes a post analysis with several hypothetic ts Use 0 for noble gases lt 0 gt sample from law P t exp t ts ts It slows down calculations Only necessary if the number of collisions is low M E For ts ts X surface see examples 4 5 in the user routine customcollision f 0 Reading the source settings We then select to run in mode 2 standard option to have the standard individual output 3 and choose the normal collisions cosine la
54. RADIOACTIVE ION OPTIMISER USER MANUAL Mario Santana Leitner ii Contents Index List of Figures Preface 1 Monte Carlo code RIBO user manual 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 RIBO a MC code for isotope release optimization Overview 22 4 ban RS eee eRe UR oh E dod xe ois 1 21 Description of 12 2 Tnstall tion 5 sw CoA ES aD file 222252582253 SL BEE ARES 1 3 1 S rfacesc td s uw xXx eaman a RC er 132 Gero e she eh Pe ee Rex SSE OEE SESS 1 3 3 Birth of particles the Source card 1 34 The Tally end 1 3 5 Histogram card 0 Effusion with moving walls 1 Execute RIBO 225 09 4 eR e aR 1 5 1 The Isotope RElease Simulator IRES 1 5 2 Express Execution with a batch file 1 53 Debugging running errors 1 6 1 Introduction 1 6 2 Typical error situations messages Making 3D model views e User defined subroutines 1 8 1 User defined collisions customcollisionf iii ii vi 2 3 4 1 8 2 User defined desorption in powder customcollision f 1 8 3 User defined source distributions customsource f 1 8
55. adding updating total desorption time per particle tde IF surn eq 5 THEN 10 100 0 ts ELSE tO ts 1 1 25 X3 3 END IF tde tde 10 log l rand zero Frenkel Law COL6 1 COL6 1 1 0 CosineLaw type counter WARNING use positive ts with negative time you would be using twice the exponential law END 1 8 2 User defined desorption in powder customcollision f Very often the number of collisions in the powder is very large and therefore desorption becomes important for most isotopes If the mean sticking time in the powder differs from that elsewhere or if it changes from one region of powder to another the user can customize the user routine sources powderdesorp f in order to meet the particular requirements SUBROUTINE powderdesorption COLPOW celln tdep ts integer 4 celln 11 real 8 COLPOW ts d real 8 tdep i o external rand 1 8 USER DEFINED SUBROUTINES WRITTEN BY MARIO SANTANA LEITNER 2006 CUSTOMIZE THIS FUNCTION TO FIT YOUR PROBLEM USE add customized desorption time within a powder cell NOTE normally at the end of the history the amount ts COLP is added if tdep is zero don t try to sample from distributions i e
56. ali 10 1 5 zero 0 0 cosphi eos phij a b IF 11 0 5 THEN k V3scaV3 grad3 U3 U3 1 U3 1 2 k grad3 1 U3 2 U3 2 2 k grad3 2 U3 3 U3 3 2 k grad3 3 COL6 5 COL6 5 1 0 ELSE CALL cosineLaw grad3 U3 CALL Boltzmann U3 T A mode COL6 6 END IF COL6 6 1 0 Now adding updating total desorption zero 0 0 tde tde ts 108 1 V3scaV3 U3 grad3 norm U3 norm grad3 phi gt 60 deg grad3 1 U3 1 grad3 2 U3 2 grad3 3 U3 phi gt 60 deg existing function existing function time per particle tde ts provided at runtime e g IEA 55 56 57 58 59 3 60 61 5 46 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL EXAMPLE 5 Sample the VELOCITY vector angle speed with the cosine Law and M B thermalization USE customized STICKING as a function of position X3 surface and with a distribution function do not just use average value of the type exp t ts ts 15 1 55 The dependence with position is e g 10 ts 1 1 25 z The surface surn 5 is a COLD SPOT where t0 ts 100 0 zero 0 0 CALL cosineLaw grad3 03 existing function CALL Boltzmann U3 T A mode existing function Now
57. ata cond dat file as explained in 2 2 1 2 Write the input file to describe the geometry of the system using combina torial geometry refer to sections 1 3 1 1 3 2 Note that you must fill in all the cards and fields in the standard way for effusion calculations Surfaces Cells Source and Tally even if you are not going to need them all The function radia f is open for the user to customize the expression of the radiation as well as the temperature dependency of see section 2 3 4 and example at page 96 C i j k 2 contains the cell index of the boxel 11 7 k Boundary condi tions 64 CHAPTER 2 THREE DIMENSIONAL DIFFUSION MODULE 3 Adjust the source routines to match your specific requirements e Customize the initial concentration or temperature distribution by modifying the file sources C ST ART f as explained in 2 3 1 page 67 e If needed introduce time space direction dependences in the diffusion parameter by editing the routine sources Dij kt f printed in 2 3 3 on page 69 e Edit the file sources gener f page 68 to add time space region dependent sources or sinks of atom concentration or heat e For heat transfer calculations modify the subroutine sources radia f see sec 2 3 4 on page 71 to adjust the radiation function to your problem e Customize the printing routine sources user PRINT3D f see section 2 3 5 to print the necessary information
58. chal lenge to nuclear physics These together with other fascinating research lines in particle physics solid state physics and medicine demand utterly exotic and intense ion beams for which a global optimization of all relevant phenomena in beam formation has to be coherently conducted As a response to this request a Monte Carlo sim ulation code has been written to integrate diffusion and effusion under various vi pressure flows and conditions including the transport through continuous media and enabling diffractive and surface dependent effects emulating ionization in surface and plasma ion sources and finally reproducing the movement of ions under electro magnetic fields vii Monte Carlo code RIBO user manual 1 1 RIBO a MC code for isotope release optimiza tion Overview The Radioactive or Rare Ion Beam Optimizer RIBO is a scientific Monte Carlo simulation program focused on the optimization of radioactive ion beam produc It tracks the paths of atoms through ISOL targets from generation not included this step should be calculated with codes like MCNPX 3 MARS 4 5 FLUKA 6 71 to ionization and extraction It includes the following models e Diffusion from slabs fibers or powder e Diffusion in 3D structures with custom options NEW FEATURE Heat transfer in 3D multi body set ups including conduction radiation cooling heat deposition NEW FEATURE e Effusion in the
59. cluded in ionization eff Ion extraction time s 0 001 this factor is included in teff In presence of ion sources RIBO prints the ionization efficiency and the parame ters linked to the transport of the ions Number of histories N 2 1000 CPU time s J 130 CPU time history 0 13 s history CPU time collision 0 00184250808 s collision Duration of the simulation Connectivity matrix In this case the matrix of candidate links between cells is very simple 96 CHAPTER 3 EXAMPLES RIBO USE This means that the first cell is only connected to the second it is indeed so because it is embedded in the latter one that the second source is connected not only to the first but also to the third one and that the third cell is only linked to the cell number two Data analysis The results stored in the output file can be analyzed in various manners depending on the specific needs of the problem Usually the individual parts are imported into a worksheet in order to plot histograms with the possibility to observe the influence of several parameters on the delay time Whence a first column of results may contain the diffusion time and this may be modified proportionally as a function of 7 In the same way the effusion time vector may be adjusted to a particular speed of atoms proportional to T M 2 and the sticking time vector will be obtained by multiplying the array that stores the number o
60. demand the the flux of electrons and the program plION will take charge of the calculation of the electron impact direct ionization cross section subsequently asking for the energy of the electron beam the species to be ionized and the sought ion state If Surface ionization is chosen then the program will ask for the cell index of the first ion cell and for its corresponding surface ionizers then it will do the same for the second ion cell and so on Next a program called surfION will be launched and it will ask which is the number of the surface that corresponds to the ionizer and then it will demand the atomic numbers of the projectile and of the substrate or in their default the work function their mass number and the type of compound Boride Carbide Element 3 If absorption has been activated then after each collision to the walls the code will check if the particle condenses in the surface In order to do so a uniform random number is compared to the absorption probability on the surface which should have previously been introduced in the input file 19The database plion dat should be consulted to see if the wished ion is tabulated 1 5 EXECUTING RIBO 27 in the decimal part of the second input of each surface the integer part corresponds to the roughness parameter as explained in 1 3 1 From the fan o lonization the included in the f phenomena implemented in RIBO after having d
61. e quadric The family of quadrics comprises planes cylinders cones spheres hyperboloids With this base of surfaces almost any arbitrary shape can be approached with a moderate consumption of memory and computa tion time Smooth bends of tubes could have been modeled in a single surface a toroid However these belong to fourth degree surfaces quartics for which 33 parameters are needed This is too costly and rather unmanageable which means that such surfaces must be implemented in pieces A plane with an equation x 1 could be implemented as geometric entries 0000001001 46 2 a eua 22 A Coy DY A Caz ZH Cys Y zd Cuv Cy uz e g 200 05 would mean that the surface is not diffusive RC 200 Aluminum and that its absorption coefficient is 0 05 5 of the impacting atoms condense to the surface Toroids have to be approached by several cylinders Unlimited number of objects 8 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL The last number corresponds to the independent term C Naturally any propor tional equation would be equivalent e g 000000 2500 25 sphere centered at the origin of radius R would x y 22 C thus 111000000 R If the sphere is centered at zo yo zo then the equation transforms to y yo z zo 3 hence 111000 220 2yo 220 R2 xo yo zo Similarly ellipsoids and cylinders can be
62. e Cells card would look more or less like this 1f the number of elements is big the best method is to use a spreadsheet to generate surfaces and cells Cells n S1 S2 S3 S4 1 1 2 4 6 container outside target zone 2 2 3 5 0 Transfer line and or ionizer 3 6 1 7 0 4 6 7 8 0 space between foils 1 and 2 5 6 8 9 0 space between foils 2 and 3 3 1 3 Third example test inp In the standard distribution of files an example input file called test inp and possible output file called test out are provided In this section all steps will be thoroughly described input file runtime options and interpretation of the output file Thus the user shall be able to test the system and to get acquainted to most functions of RIBO Once results are recovered experiencing with the input file and runtime options is encouraged so as to gain a total control of the program prior to real case usage Input file The details of the problem are found in the input file Surfaces T x2 2 X X X Y Zz 1 0 5 298 0 0 0 0 0 0 1 0 0 15 2 0 5 298 0 0 0 0 0 0 1 0 0 5 3 0 5 298 0 0 0 0 0 0 1 0 00 10 4 0 5 298 0 0 0 0 0 0 0 1 0 1 5 05 298 0 0 0 0 0 0 0 1 001 6 0 5 298 0 0 0 0 0 0 0 0 1 i 7 0 5 298 0 0 0 0 0 0 0 0 1 1 ion source may span over multiple cells CHAPTER 3 EXAMPLES OF RIBO USE 8 0 5 9 0 5 Cells 51 1 8 2 8 3 2 Source Type M
63. e average intrinsic delay time is bigger than the average delay in the powder because it includes it EFFUSION NEUTRAL ATOMS EFFUSION IN VACUUM excluding powder fiber TOTAL 806 Average free path m 1 4589 Distance between collisions average 1 6666 cm highest maximum 6 8559 cm Average number of collisions 87 5384615 100 b1 0 b2 0 D 0 5 bl Stuck particles Emitted thermally b2 Surface stuck particles Thermally emitted D Inelastic scattering Debye surface phonons S Almost elastic scattering Specular reflection EFFUSION IN POWDER Average path m 3 3238 69 496 of total path Average collisions in the powder 221586 217 99 961 of total Distance between collisions in powder 3 1 EXAMPLES average free mean path 14 99998 um RESIDUAL GAS Mean free path m 3587 808 Average of collisions with residual atoms 0 04 EFFUSION OF IONISED ATOMS EFFUSION IN VACUUM excluding powder fiber TOTAL 194 Average free path m 1 3900 Distance between collisions average 1 6671 cm highest maximum 6 7662 cm Average number of collisions 81 4536082 100 51 0 96b2 0 D 0 968 bl Stuck particles Emitted thermally b2 Surface stuck particles Thermally emitted D Inelastic scattering Debye surface phonons S Almost elastic scattering Specular reflection EFFUSION IN POWDER Average path m 3 3322 71 047 of total path Average col
64. e cell limited option will be chosen at runtime RIBO Diffuse A diffusion toolkit DIFFUSE is a bash shell script that manages the core FORTRAN program DIF FUSE F It computes diffusion profiles analytically through the infinite series pro vided by Fujioka 14 and the second law of Fick 15 and for one dimensional geometries under variable and or non homogeneous conditions numerically first law of Fick Additionally it may compute the diffusion coefficient provided that a release fraction is known at any given time More details can be found in 11 2 5 1 Running DIFFUSE is trivial because it is fully interactive Fig 4 shows the first and second menu choices In fig 4 1 a options 1 3 5 permit to plot the drop of total concentration in the slab as a function of time 5 additionally shows the space dependency 2 D graph In any of the cases it is necessary to install PAW 8 to have the functions plotted Otherwise the user can take the output data written in the file profile dat and use his favorite data analyzing program 2 and 4 invert the diffusion function to extract the Diffusion coefficient from a fixed diffusion release situation Based on the chosen option DIFFUSE poses the following questions 107 108 CHAPTER 4 DIFFUSE A DIFFUSION EMULATOR Insert the number of space nodes A good compromise between precision and speed of computation would be between 10 and 1000 Insert the diffusion coefficient cm
65. eat diffusion conduction The different with respect to 5 is that in the surface of the objects instead of atomic desorption heat radiation takes place This option could be useful for coarse heat calculations of your target The next alternative concerns the physic model to be used for the reflection of atoms from the walls Specular reflection S may be used to represent reflec tions of light in systems of mirrors Lambertian reflections B follow the cosine model and thermalize the energy of the projectile to that of the surface and exact reflections D include information of the crystal lattice and of its vibrations The last choice C custom activates the user routine sources customcollisions f explained in detail in 1 8 1 on page 41 dogmate Ad um COLLISION MODEL I What model for the treatment of the collisions S Specular lt lt B B Knudsen Lambert recommended cosine law D Debye semi classic under development C Custom BACK Go back to previous menu For option C Custom modify the subroutine sources customcollision f at your convenience and recompile using tools make sh During runtime the user can decide between having the energy of the reflected particles sampled from the Maxwell Boltzmann distribution or fixing its energy Reflection models Including tempo rary sticking 30 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL to the ave
66. ecided about user is asked to specify which of the remaining steps will be simulation SELECT MODE 1 Diffusion 2 Effusion 3 Diffusion Effusion recommended 2 4 Conductance calculator Clausing Coefficient 5 Diffusion3D ALPHA VERSION 6 Heat transfer cond rad ALPHA VERSION 1 Diffusion This option uses the second Fick law analytically integrated from the 1 law for simple cases like foils cylinders fiber or spheres powder in uniform conditions More elaborated descriptions based in finite inte gration of 1 law are provided by option 5 see also chapter 2 or by the program Diffuse explained on page 107 Note also that even if diffusion is not explicitly asked for like in option 2 the output file computes the total release efficiency for effusion and diffusion for a range of diffusion time constants refer to section 1 9 2 on page 59 If the use ization is r only wants to simulate Diffusion this cannot be the case if Ion activated then an ulterior option asks whether it will be in grain or inter grain diffusion or both Diffusion Powder effusion Both D P B In any case one can obtain average parameters only or delay distributions 0 0Required if time histograms are to be plotted 28 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL CHOOSE THE OUTPUT MODE recommended 13 1 Only average figures 2 Diffusion delay distributio
67. ed to a pipe of 1 cm diameter with a bend of 90 The first example consists on bulb full felt with He at 300 K and connected to a pipe of 1 cm diameter with a bend of 90 Dimensions are shown in fig 3 1 It is indeed a simple geometry but attention has to be paid to define the correct subspaces unambiguously For that sake auxiliary planes are needed In fig 3 2 surface 2 is an auxiliary plane that will help to define the second cell if it was not used the program could not logically decide between the real cell and the dotted one In fact it would allow transmission of particles both to the right and to the left The same is valid for the crossing of pipes at the surface 4 The Surfaces card 76 3 1 77 Figure 3 2 Surfaces and cells for the first example Auxiliary surfaces are needed to define the correct subspaces would be something like this Surfaces n rc T x2 y2 z2 Xy XZ yz 7 C 1 0 5 300 1 1 1 0 0 0 0 0 0 9 0 2 0 5 300 0 0 0 0 0 0 1 0 0 0 0 3 0 5 300 0 1 1 0 0 0 0 0 0 0 25 4 0 5 300 0 0 0 0 0 0 1 1 0 8 0 5 0 5 300 1 0 1 0 0 0 16 0 0 63 75 6 0 5 300 0 0 0 0 0 0 0 1 0 4 0 And the Cells card would read as Cells n S1 S2 S3 S4 1 1 0 0 0 2 1 2 3 4 3 4 5 6 0 2 The Source card for a homogeneous distribution in the sphere should be Source Type M T Alpha nx ny nz x y Z S 4 300 180 1 0 0 0 0 0 78 3 1 22 Second exa
68. em This enables to progressively correct mistakes for instance first one can implement the system of tubes with a simple source point source then after debugging insert the target material foils 36 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL or powder and only after proper running implement a more complex source etc If this were done in one go it would be harder to disentangle the individual causes of the overall errors If a major error has been produced the program will terminate complaining about some input output error If this happened it could be due to any of these Causes e Some of the lines needed have been forgotten like the card names or the explanatory lines e A coefficient in the cells definition has been omitted typically some zero or the roughness coefficient or the independent term e All cells do not have the same number of elements some cell has not been completed with zeros e The source type is not compatible with the number of source parameters missing orientation angles or particle mass or temperature e An additional line has been written somewhere line breaking is not autho rized Once the input file is digested by the program it is preprocessed and a message summarizes the geometry telling the number of surfaces the number of cells and the maximum number of surfaces that contour a cell number of cells 2 highest number of walls in a ce
69. f collisions by the individual sticking times The array of results shall be divided in equally sized groups in order to com pute separate calculations merge them and compare them This procedure pro vides information about the variance of the results 3 1 4 Example of use of the 3D Diffusion 3D GRID heat3D t The RIBO distribution includes an example of the use of the 3 D diffusion pack age The input file targets heat3D t describes a simple case where a spherical target of W with a diameter of 3 cm is irradiated with a beam that deposits 10 kW cm3 uniformly over the sphere The sphere is under vacuum and exchanges heat with a spheric heat screen that encloses the W target The screen is made of Ta 3 1 EXAMPLES 97 it has an inner diameter of 4 cm and an outer diameter of 4 5 cm starting temperature of the W target is 500 K while the shield is at room temperature The example shows how to compute the evolution of the temperature of target and shield Figure 3 5 describes the main parameters linked with this example Ta em 0 19 2 49 W cm 58 J K T 0 298 10 kW cn J mu gt 9 scalebox0 6 Figure 3 5 The spherical W target and the Ta screen 98 CHAPTER 3 EXAMPLES OF RIBO USE Input file The input file for this example is very simple There are 3 quadrics the three spheres that describe the
70. fined functions for 3D diffusion custom3D f In addition a number of applications that help writing the input file and to analyze the output data are being developed and stored in tool s 1 2 1 2 1 The RJ Setup Description of files BO code is distributed with a number of FORTRAN files data libraries and text files The present distribution spans over the following files e DIFFUSE PACKAGE For diffusion calculations only fancy options D f diffuse e diffuse o diffuse sh functC f license pdf README timeN f A set of PAW based applications is under development 4 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL e MONTE CARLO For typical simple diffusion and all effusion calculations data Cond dat init dat plion dat sion dat Urz dat valves dat workf dat README docs effusion pdf dvi tex MANUAL batchinputs Several runtime batch files targets Several input files objects diffuse3D o inito powder o space o explicitEM o main o povray o readUrz o surfON o tools 3D view sh convgeom f make sh trajectory sh rotate f translate f sources Bn f customcollision f customsource f emittance f math f read Efield f userprint f custom3D f license README version Moreover for the graphical options Diffuse sh 3D RIBO sh and trajec tory sh the two free programs are required 1 Physics Analysis Workstation 8 2 Persistence of Vision Pov
71. having the gage at the end of a complex tubular system it can be first put at the end of the beginning section and then pushed forward to next section etc to ease the detection of the errors of geometry 1 6 2 Typical error situations messages Source error Reading the source settings invalid number incomprehensible list input apparent state unit 1 named inputfile t last format list io lately reading direct formatted external IO Aborted Reason 1 Too few parameters in Source card Input file error surfaces not read READING INPUT FILE temp t 1 reading surfaces list in end of file apparent state unit 1 named temp t last format list io lately reading direct formatted external IO Aborted Reason You wrote cells instead of Cells Input file error apparent error in source Post processing geometry DONE RUNNING x y z 0 302196309 0 00917545272 0 146321036 Error in source card source out of cell domain Reason 1 You forgot some comment line e g after Surfaces or after Cells 40 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL 1 You defined a source that falls completely off the geometry Input file reading error gt Reading source settings invalid number incomprehensible list input apparent state unit 1 named inputfile t last format list io lately reading direct formatted external IO Aborted Reason 1 You did no
72. ible result can be 1 T FLIGHT TIME IN VACUUM MI M2 M3 M4 5 0 2379 0 3249 0 4003 0 4643 0 5182 Fit to T i t exp t tl tl tl 0 23786 error O 2 7 17366 error O 3 25 86393 Fit to T i t I exp t tl exp t t2 tl 0 04672 t2 0 19999 error O 3 7 83485 error O 4 1 87617 The meaning of the results is f 1 1 are the five momenta for the free flight time 4 t ftignt obtained from the distribution In this category the flight time spent within a powder or felt mesh is not considered it is counted in 2 T FLIGHT TIME IN POWDER 1 first fit to a simple decaying exponential f t is done the fit parameter being t1 f error 2 is the relative error between the predicted M2 from f1 and the real one Idem for O 3 1 Idem for the second function f5 t4 t2 1 Diffusion is not included here because its function is in principle known 14 117The same is done for the flight time in the powder fiber if any and for the total effusion time 1 9 OUTPUT FILE 59 1 9 2 Computation of release fractions The individual delays for every history are printed in the output file as well as the particle state atom or ion extracted or absorbed and starting coordinates so that the user can compute the the release fraction as a function of whichever parameter e g diffusion time constant sticking time starting z coordinate
73. ight time reaches a certain thresholdP Setting a low threshold can be useful to have a fast scan of a rapid release peak sparing the long simulations of the tails of the release distributions For release fitting purposes however the threshold should be high in order to avoid annoying normalization issues 4Several detectors and forced paths can be used with the Complex Tally card described just after Unlike the other threshold times this one is not CPU time but physical flight time 1 3 INPUT FILE 19 The second number displays the number of histories size of the simulation and the third one the maximum elapsed time The simulation is finished as soon as any of these two events takes place maximum time OR maximum number of atoms histories WARNING Make sure that you don t use tabulators in the Tally card Complex Tally card In addition to the options provided in Simple Tally Card the user can specify sev eral sequences each of them composed of 1 or more conditions every condition is a surface crossing When ALL the conditions within any of the sequences are accomplished then the particle reaches the end The user can do the following 1 Specify to have various end surfaces that is to say several trivial sequences each composed by a single condition This is logic OR introduced by means of the character eg 0 20 10 20 6000 750 10 The particle will stop when it first reaches 10 or 20
74. ijkt f to customize the diffusion coefficient spatial or time dependence radia f to adapt the radiation function userPRINT3D f to determine the entities to be printed 2 3 USER DEFINED FUNCTIONS AND SUBROUTINES 67 2 3 1 Customizing the starting distribution of Concentration or Temperature CSTART f The FUNCTION CSTART R celln inmode real 8 R 3 CSTART character 12 inmode integer 4 celln real 8 CO CUSTOMIZE THIS FUNCTION TO FIT YOUR PROBLEM USAGE 3D diffusion problems starting Concentration Temperature may depend on the coordinates e g CO 273 CO 273 x R I1 R 2 It can also depend on the material celln or on both Variables R 3 absolute coordinates x y z cm of the volume boxel CSTART Starting concentration Temperature celln number of the cell inmode 5 gt mass diffusion problem 6 gt heat transfer IF celln ge 1 THEN IF inmode eq 5 THEN 1 0 ELSE IF inmode eq 6 THEN 500 0 starting temperature END IF ELSE outer space IF inmode eq 5 THEN 0 0 ELSE IF i
75. implemented Moreover arbitrarily ori ented figures may be obtained by applying the rotation equations over the coordi nates For instance an ellipsoid with an axis a in the x y plane forming an angle a with respect to the x coordinate be introduced by applying a rotation is replaced by cos a x sin a y and 7 by sin a x cos a y Therefore a cos o x sin a y bsin a x cos a y e 22 1 It should be remarked that the number of surfaces accepted by RIBO is unlim ited but naturally if simplifications conduct to less elements then the CPU power requirements will be lower and results will be obtained faster At this point some prospective users might be thinking that the level of compli cation involved in writing the input file is fairly high Responding to the demands of the first groups that have been exposed to these explanations a routine called convgeom has been written to assist in the generation of the Surfaces card When executing convgeom the user is asked to choose a surface type it can be a plane P a sphere S a Cylinder C an ellipsoid SQ a cone K or a general quadric equation GQ Depending on this first choice RIBO then asks to specify 1 3 INPUT FILE 9 the radius or the position of the plane or the orientation of the plane cylinder or the radii of the ellipsoid Then the user can decide to rotate the surface by specifying the two Euler angles or
76. ing read in customsource f And you can restrict generation to a cell source limited to a cell give cell number celln gt 0 gt generation limited to volume defined by celln celln 0 gt do not constrain to a cell celln lt 0 gt just generate a geometry plot This part is standard As for any simulation the code asks for the input file name and the output file name Then as usual it gives the chance to restrict the initial distribution of particles CSTART f 2 3 1 to one of the cells defined in the input file 3 1 EXAMPLES 101 SPECIAL EVENTS TERMINATION Please choose among these options 1 Crossing of end surface lst card Tally No ionizations in the system 2 Atoms can be ionized in a PLASMA ion source Histories of atoms and ions end when they cross the end surface detector 3 Atoms can be ionized in a SURFACE ioniser Histories of atoms and ions end when they cross the end surface detector 4 Atoms can be absorbed in the walls Trajec tories end at absorption or when crossing the end surface So what do you want 1 2 3 or 4 1 SELECT MODE 1 Diffusion 2 Effusion 3 Diffusion Effusion recommended 2 4 Conductance calculator Clausing Coefficient 5 Diffusion3D ALPHA VERSION 6 Heat transfer ALPHA VERSION CHOOSE THE OUTPUT MODE standard is 3 8 Only average figures 9 Print full ma
77. kK K FK K FK K K 2K 2K 2K K K FK K FK FK 2K FK 2K K K K FK FK K K K K K FK FK K K FK K 2K 2K 2K K 2K kK K K 2K FK 2K FK K 2K FK FK FK k K 2K K 2K FK 2K k K 2K K K 2K K K K K K CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL SUBROUTINE readEfield 2 real 8 x y z E 3 I CI CCK kk EXTRACTION FIELD PLASMA RADIAL FIELD A field E Er Ey penetrates a tube up to Length Upen 4 5 cm y The tube has radius Rmax 2 We have of the electrostatic potential U for grid with 17 radial divisions and 68 longitudinal divisions gt Ury 17 68 There is second radial field due to plasma effects Up this field behaves as Up r Upxexp r lambda lambda 0 01 8 Ury 17 68 r Rmax Er 8 Ey Ez Up lambda integerx4 i j 5 2 0 5 IF Ynax Udep lt Rnax THEN open UNIT 1 FILE data Urz dat STATUS OLD read 1 Ury close 1 230 0 100 0 Ury i j Ury i j 1 Udep 67 0 Erz30 0 100 0 Ury i j Ury i 1 j Rnax 15 0 END IF xxx need to define Up and lambda see Koe2000 Er Er Upxexp r lambda lambda 0 01 Ey Ey conversion to rectangular coordinates Ex Erx x abs r 1E 10
78. ladly accepted and acknowledged in this way RIBO will grow and the whole commu nity will benefit Presently these are the routines available in sources 1 8 1 User defined collisions customcollision f This routine is called if the user answers in the runtime question about the collision model SUBROUTINE CustomColl X3 grad3 T A surn rc ts tde U3 COL6 integer 4 surn 11 real 8 3 3 43 3 ts li 1 8 U3 3 COL6 6 tde i o real 8 SU3 3 SV3 3 SW3 3 la 3 real 8 phi alpha cosphi n zero k V3scaV3 norm al bl cl v t0 external rand FK K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K KK K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K K 42 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL WRITTEN BY MARIO SANTANA LEITNER 2006 CUSTOMIZE THIS FUNCTION TO FIT YOUR PROBLEM USE transport problems with special physics for the collisions HIGHLIGHTS customized law as a function of impinging angle desorption time sampled from Frenkel Law simulation of COLD spots desorption function of temperature position ACTIVATE USE THROUGH option C at RUNTIME to the QUESTION What kind of treatment of the collisions SSSpecular E B Knudsen Lambert D Debye C Custom VARIABLES INPUT X3 3 Absolute position at collision point x
79. lisions in the powder 222144 447 99 963 of total Distance between collisions in powder average free mean path 14 99998 um RESIDUAL GAS Mean free path m 3587 808 Average of collisions with residual atoms 0 06 91 These two tables tell about the effusive path of the atoms and ions when they were still neutral in the tubes and inside the powder It also prints the number of collisions with residual gas at the pressure that was specified at runtime STATISTICAL DATA Moment T i lt T i 2 li s 1 T FLIGHT TIME IN VACUUM MI M2 M3 M4 5 0 0015 0 0021 0 0027 0 0033 0 0038 Fit to T i t exp t tl1 tl tl 0 00154 error O 2 6 48439 92 CHAPTER 3 EXAMPLES RIBO USE error O 3 12 50578 Fit to T i t I exp t tl exp t t2 tl 0 00026 t2 0 00133 error O 3 14 89847 error O 4 21 85320 2 T FLIGHT TIME IN POWDER MI M2 M3 M4 MS 0 0034 0 0058 0 0079 0 0099 0 0118 Fit to T i t exp t tl1 tl tl 0 00343 error O 2 29 08787 error O 3 50 57320 Fit to T i t I exp t tl exp t t2 tl 0 00136 t2 0 00527 error O 3 30 52552 error O 4 44 10240 3 T TOTAL_EFFUSION_TIME FLIGHT STICKING MI M2 M3 M4 5 0 0050 0 0069 0 0089 0 0108 0 0126 Fit to T i t exp t tl tl tl 0 00498 error O 2 2 79223 error O 3 4 74298 Fit to T i t I exp t tl exp t
80. ll 6 number of bodies 7 1 6 DEBUGGING RUNNING ERRORS 37 The MC code then assembles the geometry thereby networking the cells that have a common interface and then it saves the result into an array By doing so each time that RIBO verifies whether a particle migrates from the current cell to any neighboring region this happens after every collision the amount of checks needed is reduced to the number of connecting cells listed in the connectivity matrix computed only one at the beginning That matrix printed separately in the file CCONM condenses a lot of information of the geometry of the system and therefore it aids to cross check the Cells and the Surfaces cards of the input file Matrix as The file contains a column with several groups of integer numbers every group 4 debugging tool starts by the cell number and it is followed by those cells that have a common interface with that given cell The CCONM matrix of the followed example has only two groups there are only two cells The first group says that cell number 1 first line is connected to cell number 2 second line the second group says that cell number 2 third line is connected to cell number 1 fourth line This case is quite trivial but it helps to underline two properties e The connectivity matrix is symmetric This does by far not mean that these cells are actually touching 38 CHAPTER 1 MONTE CARLO CODE RIBO USER M
81. made at this stage 1 A Cylinder category has not been included because random cylindrical birth distributions are a special case of Target with sigma 0 1 angles aperture cone and axis orientation should be given in degrees f The atom temperature in Kelvin expresses the energy velocity of the atoms If it is unknown then a good value is that of the wall temperatures since thermalization should fully have taken place after a few hundred col lisions 1 The central angle of emission is indeed a velocity unitary directing vector However normalization is not required it is done internally e g 2 1 1 not normalized can be given instead of 0 8165 0 4082 0 4082 normalized but not fully precise 1 For isotropic generation a 180 for focused beams a 0 In no case it can be omitted 1 Dimensions are expected like elsewhere in cm If the axis of the cylinder is 2 then 0 902 0 If the cylinder has its axis in 7 then 0 p 14 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL 90 90 if it is in 2 then 0 0 Note that these rotations only affect the geometry of generation and not the velocity vectors This means in particular that angles like 90 0 or 0 90 would also be valid for the two examples just shown In these trivial examples this does not matter but for more complicated cases the parity property avoids errors of sign 1 In total for the point source 1
82. mple Bigger files CHAPTER 3 EXAMPLES OF RIBO USE A more realistic example for the field of radioactive ion beams is that of a tar get made of thin foils Fig 3 3 sketches a target with a SPIRAL Christmas tree like shape 1 2 intended to dissipate the energy of the beam in steps Surfaces 1 2 and 3 are planes and 4 and 5 cylinders they delimit the target container and the transfer line cells 1 and 2 respectively The sixth surface is a cone and the remaining surfaces 1 n are planes these surfaces enclose the target slabs of thickness d and the spacing between them between two consecutive front faces The Surfaces card would look something like this the symbols L1 4 1 6 2 N n E PE E Y n Li L2 ma Figure 3 3 Sketch of a GANIL SPIRAL type target 1 2 L2 should have to be replaced by numeric values Surfaces n rc T x2 y2 22 xz yz x y 7 C 1 0 5 300 0 0 0 0 0 0 1 0 0 1 2 0 5 300 0 0 0 0 0 0 1 0 0 L1 plane 3 0 5 300 0 0 0 0 0 0 1 0 0 L2 plane 4 0 5 300 0 1 1 0 0 0 0 0 0 r1 2 big cylinder 5 0 5 300 0 1 1 0 0 0 0 0 0 272 small cylinder 6 0 5 300 r1 L1 2 1 1 0 0 0 0 0 0 0 containing cone 7 0 5 300 0 0 0 0 0 0 1 0 0 0 151 foil front 8 0 5 300 0 0 0 0 0 0 1 0 0 x0 d Ist foil back 9 0 5 300 0 0 0 0 0 0 1 0 0 0 2nd foil front 10 0 5 300 0 0 0 0 0 0 1 0 0 x0 d dx 2nd foil back 3 1 79 Th
83. n 2 Effusion The recommended and standard option is 2 because diffusion can be simulated separately with diffprof If only effusion is chosen in addition or not to ionization then the alternatives thereby available are those shown in example 3 3 1 3 3 Diffusion Effusion This option combines the first two The advantage is that the final release times diffusion effusion are obtained without need of convolving an analytical diffusion formula with a fitted histogram of effu sion The drawback is that diffusion is sampled and therefore it includes a stochastic artificial error 4 Conductance calculator Clausing Coefficient This option estimates the conductance between two sections in terms of the Clausing coefficient For technical reasons the surfaces cannot coincide with the endsurface defined in the card 1 Only average figures Introduce the beginning and ending surfaces Extracting con 5 Diffusion3D ALPHA VERSION This uses the combinatorial geometry duct i ors pen to define a 3D grid for a finite difference finite method integration of Fick s analytic vacuum first Law This allows to compute diffusion from arbitrary geometries calculations including distinct regions with different diffusing coefficients and other fancy effects Chapter 2 introduces this module 1 5 EXECUTING 29 6 Heat transfer cond rad ALPHA VERSION This is like option 5 but for h
84. n t use tabulators in the Tally card lThe sign criteria is the gradient to the surface 1 4 EFFUSION WITH MOVING WALLS VALVES 21 1 3 5 Histogram card 1 WARNING This card is OBSOLETE Please remove this entry from your input files and from the batchscript runtime options 1 4 Effusion with moving walls valves In some circumstances a few walls in the system are not static all the time This is the case when a valve is present in the system In order to deal with these objects RIBO reads over the file data valves dat and obeys the instructions thereby provided This file contains a table where every line describes the opening and closing sequences of a wall that communicates two contiguous cells These walls would normally behave as virtual boundaries in the sense that they limit different regions so they belong to the boolean logic definition but particles do not bounce with them but rather they cross them The user must therefore specify the closing time start time together with the ramp time to close fast valves may close in about 0 005 s the respective opening time time and opening ramp time the surface number of the boundary that is going to lose its transparency and the area where it is acting that is the two cells that it is communicating isolating For example the gate 1 that communicates cell 1 with cell 2 and cell 1 with cell 3 can be closed in the interval t 0 10 s between cell 1
85. nised atoms make it the outlet as ions The rest recombine The process takes about 1 microsecond in all IF you know these fields you can edit explicitly transport the ions by editing the functions readEfield f Bn f functions emittance f AND answering 1 0 now extract efficiency 0 1 extract time ms 0 2 1 3 SURFION starts and we type the atomic number of the effusing atoms and of the ioniser substrate We select the temperature of the surface and its composition pure chemical element and SURFION determines that the most likely event is positive surface ionisation and prints out the corresponding individual probabil ities Then the module for the IONIC TRANSPORT asks whether we want an explicit EM transport then we would need to customize the routines readEfield Bn f printed in section 1 8 4 or rather we will estimate the extraction with an average extracting efficiency and an associated extracted time We opt for the second and introduce the two values SELECT MODE 3 1 85 1 Diffusion 2 Effusion 3 Diffusion Effusion recommended 2 4 Conductance calculator Clausing Coefficient 5 Diffusion3D ALPHA VERSION 6 Heat transfer ALPHA VERSION CHOOSE THE OUTPUT MODE standard is 3 1 Only average figures Print all relevant events standard Like 3 individual desorption times Effusion time AND velocity direction
86. nmode eq 6 THEN 298 0 END IF END IF CSTART CO END 68 CHAPTER 2 THREE DIMENSIONAL DIFFUSION MODULE 2 3 2 Defining source and or sink terms gener f FUNCTION gener R CO t PAR ncell celln inmode integer 4 ncell celln li 12 inmode li 1 8 R 3 C0 t PAR 5 Dbq 7 ncell d real 8 dV CpRho real 8 gener o KK OK Customize this function to include heat or mass sources that depend on the position X or and time or and the concentration celln emissivity VARIABLES R 3 coordinates 2 DO NOT MODIFY VALUE CO Concentration par cm 3 Temperature t elapsed time s 1 dV dx dy dz cm 3 Useful to normalize generation PAR 2 concentration temperature at point X cm 3 T PAR 3 celln PAR 4 t desorption emissivity s 2 74 PAR 5 dt s inmode 5 mass diffusion 6 heat transfer Dbq 1 DO cm 2 K 0 s kO W cm 71 Dbq 2 DC cm 2 K 1 s kT W cm K 2 Dbq 3 DCC cm 2 k 2 s kTT W cm K 3 Dbq 4
87. of boxels outside vacuum 1814368 3 1 103 Maximum value at t 0 500 000 0 000 Minimum value at t Average value at t 0 398 7524 bem DE IENDEGONBDDIIONSSE E a What is the end condition for the 3 D problem Give condition number and associated value 1 tmax s Gom I X0 means stop when elapsed time 20 8 2 average average 0 e g 2 0 2 means stop when average Temperature wun decreases down 0 5 of the initial value 3 Tmin K e g 3 350 means stop when minimum Temperature goes below or grows up to Iming E 4 e g 4 1700 means stop when maximum Temperature 2 below Or grows up to imax 5 _of_steps e g 5 1000 means stop 1000 time steps 6 dT T max dt e g 6 1E 6 means stop when the maximum relative T alos per unit time goes below 1 6 5 NOTE Whatever condition the computation will be terminated if the computation time exceeds CPUmax s 4th entry in Tally input file Read Cond dat matrix found matrix is 2 700 0 000 0 000 0 250 170 000 0 000 0 000 2 490 0 000 0 000 0 190 58 000 0 000 0 000 INITIALISING TIME STEPS RUNNING 104 CHAPTER 3 EXAMPLES RIBO USE time s Cmax IM XM YM Cmin Im Xm Ym Zm C t C0 C t 0 00
88. or when its elapsed time exceeds 750 s or when the total CPU time for the simulation is greater than 6000 s 2 Follow only the paths that first go through hit a surface and then through another one and so on e g first through surface 10 AND then through 20 10 20 6000 750 10 Note that the order counts 20 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL 3 Specify the sense when crossing a surface e g end only when it crosses 10 positively gt 10 6000 750 10 4 Demand several crossings before end particular case of option 2 e g Stop after crossing surface 20 3 times 20 20 20 6000 750 10 Some remarks should be made at this point 1 Parenthesis are needed to delimit the boolean definition of end surfaces f S1 S2 Nmax Tmax Tpmax is equivalent to S2 51 Nmax Tmax Tpmax f S1 52 Nmax Tmax Tpmax is equivalent to 51 Nmax Tmax Tpmax 1 S1 52 Nmax Tmax Tpmax is NOT equivalent to 52 51 Nmax Tmax Tpmax 51 52 52 S1 Nmax Tmax Tpmax means that both S1 and S2 have to be crossed but the order is irrelevant 52 S1 S2 Nmax Tmax Tpmax will never score the second sequence because the first has priority the condition is read before 51 52 S2 Nmax Tmax Tpmax will score sequence 2 cross ing of surface 2 only when sequence 1 has not been fullfilled thus only when S1 has not been previously crossed 1 WARNING Make sure that you do
89. ores the values of C at 1 0 Highest Lowest Concentration temperature Imax Imin Pointers i j k of the boxel with Cmax Cmin Rmax Rmin Coordinates of the boxel of Cmax Cmin x y z cm cellmax cellmin the boxel of Cmax Cmin belongs to celln nbin number of boxels belonging to a cell sum sum of concentrations temperatures at time t 0 sum of concentrations temperatures at time 0 1 Print the evolution of the matrix C lot of memory DATA fileNOTyetOPEN TRUE IF fileNOTyetOPEN THEN fileNOTyetOPEN FALSE OPEN UNIT 26 FILE 26 out write 6 Creating file unit 26 END IF write 6 C PK ck ck ok ok ck K ok ok ok ok ck K K K K ck ck ck K K K ok ok ok K ck ok ok ok ok K K K K ck ck ck ck ck ck ck ok ck ck ck ck K K ck K K CK ck ck ck ck ck ck ck ck ck ck ck K K K K K K OK OK OK OK END 52 53 54 95 56 57 58 2 4 EXAMPLES OF USE 2 4 Examples of use Please check the dedicated example discussed in chapter 3 1 4 75 Examples of RIBO use 3 1 Examples 3 1 1 First example Geometry issues Figure 3 1 A simple example consisting on a bulb full of He at 300 K connect
90. path of atoms in the Fiber or Powder It will also ask for the probe spheres that should be used as macro steps for faster calculation mean free path powder fiber um e g UC powder 15 ZrO2 fiber 250 50 sphere probe radius um step path e g 800 recommended 6 x mean free path 300 sampling a limiting residence time in powder DONE tmax_sphere_of_powder ms 0 0183047683 Characterization of macrocollisions in powder Generating time and angle probability functions this may take some minutes Next the pressure inside the system is introduced in order to enable collisions if any between gas atoms Typically simulations are done for perfect molecular flow thus at nil pressure COLLISION WITH RESIDUAL NUCLEI Residual pressure torr 0 75 torr 100 Pa lt 0 gt molecular flow ideal vacuum P gt 0 gt collisions with residual gas if you do not know it but you know the mean free path anything gt 0 now If no more input is given the program will estimate the mean free path between atom collisions from statistical considerations Alternatively if the mean free path is known it should be provided 1 5 EXECUTING 33 RRRR EEEEE 5555 GGGG AAA SSSS R R E S G A RAS RRRR SSS GGG SSS R R E SG 5 R 5555 GGG 5555 INTRODUCE NUMBER n IF gt
91. r five cards each of them grouping dif ferent sets of information Every card is initiated by a key word which must not be This term of the computing jargon comes from the days when punch cards where inserted in early computers to transmit a set of data 6 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL changed it is case sensitive followed by a line that explains the inputs to come The explaining line can be edited but not deleted The next lines contain the core information of the card like the coefficients of the equations of the surfaces or the logic of the cells Extra line spacing within a card or between cards is authorized since the program skips blanc spaces Concerning the columns of text there are no constraints on the horizontal spacing of elements within a line However an even arrangement of elements in columns helps to clarify the input file and to find possible mistakes The first two cards define the entire geometry of the system walls and vol umes through which particles will effuse This means that outer elements where atoms cannot reach shall not be described e g if particles effuse through a tube then only the inner bore is given as input the outer surface defining the core is omitted First comes the card Surfaces corresponding to the walls of the sys tem Second the card Cells describes the elementary cells that are enclosed by the given surfaces Third the Source card which contains all informa
92. radius of the target and the inner and outer radius of the shield and 2 cells one for the target and one for the shield The details of the problem are found in the targets heat3D t Surfaces X2 Y2 72 XZ YZ X Y Z C 1 0 5 298 1 1 1 0 0 0 0 0 0 9 00 2 0 5 298 1 1 1 0 0 0 0 0 0 16 00 3 0 5 298 1 1 1 0 0 0 0 0 0 20 25 Cells 51 52 1 1 0 2 2 3 Source Type M T Alfa nx ny nz x y 7 R S 90 298 180 1 0 0 0 0 0 5 Tmax Tpmax 3 50000 750 10 In the present version the values of the Source card have no effect in the results for 3D diffusion but still it is compulsory to fill in all fields As for the Tally cards for the moment only the third one Tmax is used in 3D diffusion having essentially the same meaning as for effusion simulations As for the data files the simulation reads the specific capacity conductivity and emissivity from data Cond dat The example printed in 2 2 1 and included in the RIBO distribution already contains the values for W cell 1 and Ta cell 2 Note that this table can be refined by filling in the fields that show a dependence of first and second order with respect to the temperature or concentration More over if the temperature of the system varies a lot you may also want to introduce 3 1 99 the variation of the emissivity with the temperature You can easily do that by editing the function radia in
93. rage of the distribution The second option saves CPU time while pre serving precise results if the number of collisions is high enough central theorem of the limit for normal distributions but it may give way to notorious statistical errors in situations where few collisions take place DS e PETAT COLLISION MODEL II Use average energy or sample from coll law Y Use average energy recommended N Sample from collision law slower BACK Go back fo previous menu It has been shown that the number of collisions suffered by each particle in its path to the exit of the system or before ionization can be included among other output numbers Nonetheless in some circumstances it may be interesting to specify sticking in the release times The user can provide a positive ts which will be used as an exact fix number or a negative number ts lt 0 which will serve to sample exp zt fts a sticking time upon every collision from the exponential law p t The same comment about the CPU speed vs accuracy done for energy sampling applies here it is only worthwhile for a small number of collisions Note that the user routine sources customcollision f can cope with more complex schemes e g desorption time depends on the surface and on the position In particular this allows to define cold spots Sticking time s gt 0 gt every collision delays exactly ts s 0 recommended
94. ratory phase and starts the simulation At the completion of a bunch of 50 particles it prints to screen the number of simulated particles the number of extracted ions the number of surface absorptions here zero and the mean effusion flight path 3 1 EXAMPLES 89 Output file interpretation The resulting output file displays the following information comments are inter polated results after processing input file test inp reprocessing cpu time s 327 event 0 neutral particle ionised particle 2 neutral particle absorbed 3 ionised particle absorbed x0 0 70 birth coordinates ocell starting cell region td diffusion time in bulk s tPo effusion time in powder felt s COLP number of collisions in powder felt COL number of collisions elsewhere teff Flight time outside the powder s tde Total desorption time in vacuum s tdep Total desorption time in powder s tp Total effusion time s tp teff tPo tde tT Total release time per particle s tT tp td tdep desorption time in the powder s event 0 0 20 td tPO COLP COL teff tT 0 4 579 1 000 0 009 0 00000 0 00228 144017 6 56 0 0 00142 0 00086 0 4 284 0 399 1 000 0 00000 0 00006 3644 3 314 0 0 00540 0 00546 0 4 760 1 000 0 484 0 00000 0 00114 72076 3 76 0 0 00020 0 00133 Ok 0 4 946 0 440 1 000 0 00000 0 00872 551768 6 142 0 0 00623 0 00249 1 4 388 0 472 1 000 0 000
95. ray 9 1 2 2 Installation Execute the make sh script in tools if you need to recompile Otherwise no installation is required Then you should enable execution permission on the executable file that has been created 1 3 INPUT FILE 5 gt chmod x RIBO Finally you should edit your bash profile to include the path of the executable file or you can run it from the installed directory simply typing gt RIBO Depending on the shell and configuration typing bash or sh before RIBO may also work Once installed in order to use the program an input file has to be created The following instructions explain how to do this Alternatively you can test the installation with the input files stored in targets Give the name of the file and RIBO will search for in the directory tar gets if failed in inputs and otherwise in the directory from which you are executing RIBO Results will be stored in an output file whose name will be of your choice Before doing a first test it should be reminded that naming an output file with an existing file name will overwrite the old one 1 3 Input file The input file contains the information of the geometric arrangement of the target the starting properties and nature of atoms and the end conditions Choices about the physical models to be employed and determination of the output modes are decided interactively at run time The input file is organized in four o
96. ss Execution with a batch file One of the most useful tools is the execution through a batch file This file permits to execute the program routinely without having to reintroduce the interactive options When RIBO is executed all runtime options are automatically recorded in a file whose default name is batch If the user then wishes to re simulate the system it will be enough to type the exe file with lt batch at the end which means that all interactive data will be taken from the batch file The potential of this methodology is vast a user could edit the batch file change some input parameters probably also the output file name and save it as batch1 then repeat for different parameters for batch2 Finally a script like RIBO lt batchl RIBO batch2 or even a script with a loop could produce an enormous amount of data 24 CHAPTER 1 MONTE CARLO CODE RIBO USER MANUAL Before going deep into the runtime options you may want to test your RIBO distribution by running one of the batch files stored in the folder batchinputs 1 5 3 Runtime options At run time the user interacts with the program in order to define the input and output file names the type of output the models to be used and the options that shape out the results of the simulations In the first place the user introduces the file name of the input file and of the output file The input file name needs no particular extension
97. t exactly spell Surfaces 1 You misspelled the name of the input file Error in the cells definitions reading geometry storing surfaces bodies Storing cells regions invalid number incomprehensible list input apparent state unit 1 named temp t last format list io lately reading direct formatted external 1 Aborted Reason Missing elements in the region definition 1 7 Making 3D model views Making a 3D view of the target geometry is one of the fastest options to bugs in the geometry and to become aware of the target proportions Thi possible through 3D RIBO and Povray The first program executes find out 515 now BO and halts it when the Povray compliant geometry file pov is created Next Povray is called and the image file tga is created gt 3D view input t 1 1 8 USER DEFINED SUBROUTINES 41 The flat 1 specifies that the image file will be a thumbnail 2 or 3 would produce images with higher resolution The intermediate pov file is a text document with the combinatorial geometry in Povray format This file can be edited to include fancy features e g roughness and special optical effects Then 3 D can be run by specifying the proper file extension gt 3D view input pov 1 1 8 User defined subroutines RIBO includes a growing number of routines that the user can change and cus tomize It goes without saying that important user developments will be g
98. t is the temperature of the surface K 1500 add surface Y N N The generated input file would look like Surfaces n RC T x2 y2 z2 1 0 5 2000 0 0 0 0 0 0 0 707106 0 707106 0 1 2 0 5 1500 0 2500 0 749999 1 0 8660 0 0 0 0 0 4 It would now remain to define the cells 1 3 2 Cells card This card includes the logic expressions that assemble the previously defined surfaces into the delimiting elements that enclose cells these are referred to as regions in e g FLUKA 6 or GEANT4 12 Surfaces may extend infinitely and therefore bounds are required to define real elements thus the mechanism of cells It should be stressed that cells are finite subspaces This card has the following structure 1 3 INPUT FILE 11 Cells n SI S2 53 third line fourth line The third fourth fifth 2 lines correspond to the definition of the cells 1 2 3 n whose first element is precisely the cell number The following columns define the cell volume in terms of the bounding surfaces The MC code understands the cells as an intersection of the subspaces divided by a collection of surfaces the boolean union operation is not defined regions that require such an operator have to be split into several cells As an example of a simple cell cell 1 comprises the volume over the plane 1 under the plane 2 and inside the sphere 3 then The corresponding third line in the cells card is
99. t2 tl 0 00031 122 0 00469 error O 3 6 63791 error O 4 8 99565 These are the momenta up to order 5 of the flight time in vacuum in powder and the number of collision With this information you can in principle build up many fitting functions The code uses them to make fits based in exponential functions RELEASE FRACTION COMPUTATIONS T 1 2 ERF 0 1 2 9 10 0 3 2 26 98 3 1 EXAMPLES 0 6 2 43 95 0 1E 1 57 26 0 3 1 80 48 0 6 89 25 0 1 0 9327 0 3E 0 97 66 0 6E 0 98 82 0 1 1 99 29 0 3 1 99 76 0 6 1 99 88 0 1 2 90898 0 3 2 99 98 0 6 2 99799 0 1 3 99599 OTHER CONDITIONS ts s tao_d s 0 10 0 0 0 0 0 1 8 0 1E 7 0 1E 6 0 1 5 0 1 4 T 1 2 s DRF RF tsl RF ts2 RF ts3 RF ts4 RF ts5 RF ts6 000 000 000 000 000 000 000 000 1E 2 TOTAA 0 703 0 766 0 000 3 2 13 126 3 542 3 571 5 056 6 2 18 198 T999 7 934 8 056 1E 1 23 022 TIS ENT 0590 2003317 0 0 0 0 0 0 0 000 0 000 0 0 000 0 000 0 37 213 29 948 29 730 28 073 29 377 0 000 0 000 6 48 936 43 674 43 484 41 919 35 257 0 000 0 000 0 0 000 0 0 000 1 0 58 506 54 570 54 417 53 112 45 217 3E 0 78 059 76 231 76 154 75 476 69 799 67 202 000 6 0 87 058 86 027 85 983 85 590 81 998 66 568 1E 1 91 622 90 969 90 940 90 688 88 299 73 938 0 000 000 96 960 96 729 96 718 96 628 95 740 88 315 84 216 6 1 98 441 93 323 08731
100. tained in terms of other more familiar parameters _ k conductivity ps ILI ns 2 5 Cp p Specific Capacity 21 Density 35 The user is asked to split up the geometry in cells enclosed by quadrics The mass diffusion heat transfer properties of each cell must be defined by the user in the data file data C ond dat see 2 2 1 on page 64 Convection is not considered in this version but it could be easily included 2 2 INSTRUCTIONS USE 63 e s is the surface term which appears only in those boxels in the boundary between a cell and the vacuum This term has the following meaning d 1 exp dt ta E5 atomic desorption for surfaces cm s eT osp T c IE T heat radiation 2 6 Where for atomic desorption t4 is the average desorption time 5 and for heat radiation 7 is the emissivity 0 1 is the Stephan Boltzmann constant 5 6703 10 27 U is the temperature K of the boxel Ty is the average temperature seen by the boundary boxel By applying the equation 2 2 the Concentration Temperature in every boxel C i j k 1 can be monitored over time 2 2 Instructions of use In order to run a 3D atomic or thermal diffusion problem the following steps have to be cleared 1 Retrieve the conduction desorption and or radiation constants transform them into S I except length units cm and fill in the d
101. tion about the initial state of the atoms position speed mass and fourth the card Tally which gathers the end conditions of the simulation Usually time and velocity histograms are processed from the raw output data However in some cases it may be useful to get directly a histogram of the release times This can be solicited by a fifth card called Histogram 1 3 1 Surfaces card This card contains all the information of the walls surfaces that bound the effu sion paths of atoms In some Monte Carlo codes 6 7 these are known as bodies 1 3 INPUT FILE 7 The program works internally with the equations of quadrics which are intro duced in the Surfaces card The first lines look like this Surfaces RC T X2 Y2 72 XY X YZ X Y 7 third line fourth line The third fourth fifth 2 line have the information of the surfaces 1 2 3 n In each of these lines the first entry corresponds to the surface number 1 2 3 Numbering should be done in consecutive jumps of 1 unity starting with 1 The second entry contains simultaneously two parameters of the surface nature the roughness index RC only relevant if Phong 10 reflections are active 11 chapter 3 2 3 is stored in the integer part and the absorption prob ability in the decimal the temperature T in Kelvin of the surfaces is stored in the third entry and the nine remaining numbers fully define any surface of second degre
102. trix evolution SPACE C Use the custom routine userPRINT3D f The first of the questions has no impact in our future decisions We answer I The next question is crucial We choose 6 because we are dealing with a heat transfer 102 CHAPTER 3 EXAMPLES RIBO USE problem This triggers a preselected number of output formats 8 standard 9 full and C custom please consult section 2 3 5 3333 GGGG RRRR II DDDD 3 JD ID G R R II D OD 333 D D G GGG RRRR II D D 3 1D ID G GRR II D D 3333 GGGG R R II DDDD DEFINING 3D WINDOW FOR THE CALCULATIONS 1 starting vertex xmin ymin zmin cm e g 10 10 5 4 5 4 5 4 5 2 what are the full widths x y z cm 5 85 5 9 0 9 0 9 0 3 number of divisions in x y z e g 10 10 20 180 180 180 The module 3D takes over The user must define the 3D window where heat trans fer will act Vacuum will replenish the remaining space The first data input are the coordinates of the lower corner Xmin Ymin Zmin The 3D box will be defined by writing the 3 amplitudes deltaX deltaY deltaZ Finally the space will be evenly discretized in parallelepipeds when the number of divisions in each dimension is given Here we are in a symmetrical case so the number of division in each dimen sion is the same 1 WARNING The maximum dimension of the matrix is 300 300 300 INITIALISING GRID PLEASE WAIT Number
103. usually t is used getting at the fact that the file describes a target but inp or any other choices are also accepted WARNING The string of the input file should not be longer than 20 characters Name of the input file rectangle t The name of the output file again needs no particular extension out is quite intuitive o should be avoided as it may lead to confusion between the output files and the object assembled files WARNING The string of the output file should not be longer than 20 characters Name of the output file Beware it will overwrite the existing file results rectangle o The code used to ask whether a histogram was to be made t This option is now obsolete 17 searches for the input file in targets inputs and in running directory with this priority order 1 5 EXECUTING 25 If the user wishes to intersect the domain covered by Source see 1 3 3 with the volume restricted by a given cell it is then asked to specify the cell identifier Otherwise the value 0 will impose no restriction source limited to a cell give cell number celln gt 0 gt generation limited fto volume defined by celln celln 0 gt do not constrain to a cell celln lt 0 gt just generate a geometry plot The following module activates the ionization mode and termination mode Depending on the choice additional parameters will be demanded 5
104. v 73 List of Figures 3 1 3 2 3 3 3 4 3 5 4 1 A simple example consisting on a bulb full of He at 300 K connected to a pipe of 1 cm diameter with a bend of 90 76 Surfaces and cells for the first example Auxiliary surfaces are needed to define the correct subspaces TI Sketch ofa GANIL SPIRAL type target 1 21 78 The target and ion source of example 3 80 The spherical W target and the Ta screen 97 Screen captures of the interactive DIFFUSE program 110 Preface Currently the nuclear chart includes around 3000 nuclides distributed as 8 87 and a emitters stable and spontaneously fissioning isotopes A similar amount of unknown nuclei belongs to the so called terra incognita the uncertain region contained also within the proton neutron and fast fission driplines and thereby stable against nucleon emission The exploration of this zone is to be assisted by the use of radioactive ion beams RIB and could provide a new understanding of several nuclear properties Moreover besides pointing at crucial questions such as the validity of the shell model the dilute matter and the halo structure challeng ing experiments outside nuclear physics are also attended e g explanations of the nucleosythesis processes that may justify why the matter in the universe has evolved to present proportions of elements and which represents a major
105. vides clues on the potential chemical selectivity of each surface or on its surface ionization power Statistics of the free flight average distance between two consecutive colli sions average flight path from birth to ionization or up to extraction Statistics of the effusion in a powder or fiber 1f present average free flight path number of collisions Report of the effect of the residual gas average free path between two col lisions with a gas atom and average number of collisions per history Summary of the ionization scores ionization probability estimate of the ionization efficiency and error margin Module integrating computation times Authoring Some modules deserve a dedicated explanation For all the rest examples will illustrates a complete output file presenting almost all modules 1 9 1 Fitting of events to a release function Unless otherwise specified the code prints at run time the starting x y z and final conditions diffusion time fd effusion in powder effusion elsewhere ff number of collisions in the powder COLP number of collisions elsewhere COL 1 9 OUTPUT FILE 57 and ionic state of each history If the user wanted to build up a release distribution function out of the data then the histogram option could be of help However the ideal histogram binning depends of the distribution itself unknown at the beginning and of the quality of the results statistics
106. w thermalization without sampling and no initial sticking time Reading the source settings M M 5555 M M EEEEE SSSS H H MMMM E S H H 555 M E S H H H H STARTING MODULE FOR e diFFUSION IN POWDER 3 1 EXAMPLES 87 How many cells contain powder e g 2 1 what is the cell number of the powder cell 1 v 949 3955 mean free path powder fiber um e g UC powder 15 ZrO2 fiber 250 15 sphere probe radius um step path e g 800 recommended 6 x mean free path 90 sampling a limiting residence time in powder DONE tmax sphere of powder ms 0 0110181198 Characterization of macrocollisions in powder Generating time and angle probability functions this may take some minutes Most likely macro step pmax 7427 0 7427 bin 1 6 12 6 tm s thitax alphav thitav deg 6 05996603E 07 123 75 345 123 75 Next we specify that there is only one region containing powder cell number 1 We give the mean free path in the powder and the size of the macro steps MESH samples 100000 macro steps to fill a 4 d grid that will serve to sample paths within the powder COLLISION WITH RESIDUAL NUCLEI Residual pressure torr 0 75 torr 100 Pa lt 0 gt molecular flow ideal vacuum P 0 gt collisions with residual gas if you do not know it but you know the mean free path type anything gt 0 now
107. will start The connectivity matrix will be stored in file called CCONM and the results will be stored in the output file 1 6 Debugging running errors 1 6 1 Introduction Running a Monte Carlo code is similar to programming every little thing that is overlooked contains a potential bug and it will most likely induce an error at some point The time required to detect and to fix the bug will exceed the amount of work needed to avoid such flaws from the beginning This general recommenda tion concerns specially the implementation of the geometry It is strongly advised to take some time to do a sketch of the system as it will be modeled drawing and labeling each surface and marking the cells Eventually this step may already help to rise some questions about the optimality of the target and often new configu rations are immediately suggested Moreover the sketch helps to attain a logic numbering of surfaces and of cells this will in turn aid to write the input file and also to introduce future modifications Once the input file completed it should be reread cross checking with the sketch counting the number of cells and surfaces verifying that no space is undefined or multiply defined If the file has been cor rectly tabulated a fast glance will spot typing errors from the irregularities in the columns of data As what concerns the systematics another advice is to make several stages before reaching the full complexity of the probl
108. y could be weighted propor tionally to the spatial volume of the sources Analogue procedures are possible for complicated velocity spectra Particle birth within a cell runtime option Another possibility to work with more elaborate sources is to force generation inside a given cell This option is offered at runtime source limited 10 a cell give cell number celln gt 0 gt generation limited volume defined by celln celln 0 gt do not constrain 10 a cell celln 0 gt just generate a geometry plot A basic delimiting source is needed every time The primary container source chosen between Point Sphere Target and Box should include the entire source cell If the introduced cell number is 0 then the method remains inactive If it is negative then RIBO will not do any simulation it will just generate a plot of the geometry see 1 7 Otherwise particles will be sampled exclusively inside the selected cell This method can be universally used e g an infinite sphere contain ing a dodecahedron but in order to perform efficient simulations the volume of the primary source ought to match the cell dimensions as closely as possible this is similar to the concept of the rejection sampling technique 13 For instance sampling the birth of atoms homogeneously inside the volume of a I sector ofa cylinder cell 2 in the example below can be carried out first by defining a source

RADIOACTIVE ION BEAM OPTIMISER, RIBO. USER MANUAL.

Contents

Download Pdf Manuals

Related Search

Related Contents