5.3.17. Dump file source¶
By setting s-type = 17, information on particles recorded in a file can be used as sources. The dump file can be obtained by using the dump option in the [t-cross], [t-product], and [t-time] tallies. There are two alternative methods for treating the dump data, which can be switched by using the idmpmode parameter. By setting idmpmode = 0, each particle’s data recorded in the dump file is treated as an independent history. By setting idmpmode = 1, the correlations between particles at the time of data dump can be taken into account via statistical processing based on history information recorded in the dump file. For continuous calculation using the dump file based on the history information from the previous step, the use of idmpmode = 1 is recommended. [1]
The parameters related to this source type are shown below. The order of parameters is free. Parameters with (D=***) are optional.
value |
explanation |
Dump file name with full path. |
value |
explanation |
Number of dump data entries. If negative, the data are read as ASCII. If dump = 0, the dump data IDs are read from the standard tally output file. |
value |
explanation |
Special option for phase-space files. The beam starting point z0 is required. |
|
1 |
Without weight. |
2 |
With weight. |
value |
explanation |
Minimum x coordinate [cm]. |
value |
explanation |
Maximum x coordinate [cm]. |
value |
explanation |
Minimum y coordinate [cm]. |
value |
explanation |
Maximum y coordinate [cm]. |
value |
explanation |
Minimum z coordinate [cm]. |
value |
explanation |
Maximum z coordinate [cm]. |
value |
explanation |
(D=0) |
x component of spin. |
value |
explanation |
(D=0) |
y component of spin. |
value |
explanation |
(D=0) |
z component of spin. |
value |
explanation |
Direction cosine of the projectile relative to the z-axis. |
|
all |
Isotropic source. |
data |
An a-type subsection is required. |
value |
explanation |
(D=all) |
Azimuthal angle [degree]. |
all |
Randomly selected in the range from 0 to 360 degrees. |
value |
explanation |
(D=0.0) |
Solid angle spread [degree]. |
-1 |
\(\cos^2\) bias distribution. |
value |
explanation |
For mono-energy source, specify the projectile energy [MeV/n]. For an energy spectrum, use e-type = instead. |
value |
explanation |
mono-energy source |
Give the source energy [MeV/n] directly by e0. |
energy spectrum |
Specify the source energy distribution by e-type. |
value |
explanation |
(D=1.0) |
Weight of source particles. |
The following parameters can also be specified.
value |
explanation |
(D=1.0) |
Normalization factor of the source particle. |
value |
explanation |
(D=0) |
Time distribution. |
value |
explanation |
(D=all) |
Region restriction. |
value |
explanation |
(D=1000) |
Maximum retry number when reg is specified. |
value |
explanation |
(D=none) |
Transform number, or the definition of transform. |
value |
explanation |
(D=0) |
Control parameter for selecting the dump source mode. |
0 |
Each particle recorded in the dump file is treated as an independent history. |
1 |
Particle data in the dump file are processed while taking account of particle correlations in the previous step. |
value |
explanation |
(D=0.0 or 1.0) |
Control parameter for reuse of the dump file. |
0.0 |
The dump file is rewound and reused until the total histories are finished. Valid only with idmpmode = 0. |
> 0.0 |
The dump file is reused the number of times specified by this parameter. |
To use idmpmode = 1, the history number (nocas) and batch number (nobch) must be contained in the dump data.
Using the PHITS tally dump option, a dump file named ***_dmp is created in addition to the normal tally file with the file name *** specified by file = *.
Both files are required to use idmpmode = 1 and must be placed in the same folder.
The total number of histories of the previous step is adopted from the values of maxcas and maxbch written in this tally file, and the values of maxcas and maxbch given in the PHITS input file are neglected.
The normalization factor totfact is ignored when idmpmode = 1.
To use the normalization factor, totfact should be applied to the previous step for creating the dump file, at which point it will be reflected in the weights of particles recorded in the dump file.
The option idmpmode = 1 is incompatible with multi-source processing.
In addition, the calculation should not be terminated by using the stdcut parameter or editing batch.out.
The option dmpmulti controls the number of re-uses of the dump file. For instance, in the case of dmpmulti = 2.0, the dump file will be reused twice, i.e. information for each particle recorded in the dump file is reproduced twice and the particle is treated as two particles having half-weights based on different random numbers. The digits after the decimal point indicate the probability of creating another reproduction of the particle as stochastically determined by running the Russian roulette procedure for each particle. For example, in the case of dmpmulti = 2.3, the particle data are reproduced twice with probability 70% and three times with probability 30%. As a special mode, setting dmpmulti = 0.0 rewinds and re-uses the dump file until the total histories specified by (maxcas*maxbch) in the PHITS input file are finished. This mode is valid only with idmpmode = 0.
The restart calculation (istdev < 0) is essentially not applicable for dump source calculation. It is allowed only with dmpmulti = 0.0, in which case it should be noted that the calculation might give biased results unless the dump file contains sufficiently large particle data. Thus, re-calculation using a larger dmpmulti factor is recommended to achieve better statistics. If the particle data contained in the dump file are not statistically sufficient, the statistical uncertainties may not satisfactorily diminish even if a large dmpmulti factor is selected. In this case, it will be necessary to re-calculate from the previous step to create sufficient dump data.
The parameter dump specifies the number of dump data points per record. If this number is 0, the dump data IDs will be read from the standard tally output file, and it is not necessary to provide the IDs in the following line. If this number is positive, the data are read as binary data. Setting dump as a negative value will cause the data to be read as ASCII data. In the line following the dump = assignment, the data sequence of the record is input. The respective dump-file physical quantities are encoded with the ID numbers shown below.
physical quantities |
kf |
x |
y |
z |
u |
v |
w |
e |
wt |
time |
c1 |
c2 |
c3 |
sx |
sy |
sz |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
id number |
1 |
2 |
3 |
4 |
5 |
6 |
7 |
8 |
9 |
10 |
11 |
12 |
13 |
14 |
15 |
16 |
physical quantities |
name |
nocas |
nobch |
no |
|---|---|---|---|---|
id number |
17 |
18 |
19 |
20 |
In the table, kf is the particle kf-code, x, y, and z are coordinates [cm], u, v, and w are the unit vectors of the particle coordinate system, e is the energy ([MeV], or [MeV/n] for nuclei), wt is the weight, time is the initial time [ns], c1, c2, and c3 are the counter values, and sx, sy, and sz are the unit vectors of the respective spin directions. In the history-related IDs, name is the particle’s collision number, nocas is the batch history number, nobch is the batch number, and no is the cascade ID contained in the history. These are assumed to be in real*8 format for binary data and in n(1p1e24.15) data format for ASCII data.
For example, a record containing nine data points is given by
kf e wt x y z u v w
To read these data, the parameters are written as
dump = 9
1 8 9 2 3 4 5 6 7