MCNP CLASS SERIES (SAMPLE MCNP INPUT) Jongsoon Kim

Transcription

1 MCNP CLASS SERIES (SAMPLE MCNP INPUT) Jongsoon Kim

2 Basic constants in MCNP Lengths in cm Energies in MeV Times in shakes (10-8 sec) Atomic densities in units of atoms/barn*-cm Mass densities in g/cm 3 * 1 barn = cm 2

3 Simple sample problem

4 MCNP INP file One Line Problem Title Card Cell cards.. Surface cards.. Data cards.. All inputs lines: up to 80 columns Alphabetic characters: upper, lower, or mixed case Anything that follow $: a comment Comment lines Start C somewhere in columns 1-5 At least one blank A total of 80 columns long Blanks filling the first five columns : a continuation of the data from the last name card

5 Surface cards (Surface equations) Mnemonic Equation Card Entries PX x-d = 0 D PY y-d = 0 D PZ z-d = 0 D S ( x x ) + ( y y ) + ( z z ) R = x1 y1z1r

6 Surface cards in a sample problem C beginning of surface for cube 1 PZ -5 2 PZ 5 3 PY -5 4 PY 5 5 PX -5 6 PX 5 C End of cube surfaces 7 S $ Oxygen sphere 8 S $ Iron sphere

7 Cell cards IMP:N=1 1=> Cell number 1=> Cell material number, Material is described on a material card (Mn) => Cell material density Positive => Atom density in units of atoms/cm 3 Negative => Mass density in g/cm 3-7 => Specification of the geometry of the cell Combination with the Boolean intersection and union operators

8 Cell cards in a sample problem C Cell cards for sample problem :2:-3:4:-5:6 C End of cell cards

9 Data cards Mode Cell and surface Source specification Tally specification Material specification Problem cutoff MCNP card name MODE IMP:N SDEF Fn, En Mn NPS

10 Data cards (1. MODE card) Mode N N P P E P E N P E Neutron transport only (default) Neutron and neutron-induced photon transport Photon transport only Electron transport only Photon and electron transport Neutron, neutron-induced photon and electron transport * If the MODE card is omitted, mode N is assumed.

11 Data cards (2. Cell and surface parameters) IMP:N card Cell importance parameters For terminating the particle's history if the importance is zero. Fro geometry splitting if a particle moves to higher importance cell For Russian roulette if a particle moves to lower importance cell *IMP: N

12 Data cards (3.1 Source specification cards) POS = x y z Default is CEL = starting cell number ERG = starting energy Default is 14 MeV WGT = starting weight Default is 1 TME = time Default is 0 PAR = source particle type 1 for N, N P, N P E 2 for P, P E 3 for E

13 Data cards (3.2 Source specification cards) SDEF POS= CEL=1 ERG=14 WGT=1 TME=0 PAR=1 => Neutron particles will start at the center of the oxygen sphere (0, -4, -2.5), in cell 1, with an energy of 14 MeV, and with weight 1 at time 0 * SDEF POS=

14 Data cards (4.1 Tally specification cards) F1:P F1:E Surface current F2:P F2:E Surface flux F4:P F4:E Track length estimate of cell flux F5:P Flux at a point (point detector) F6:P Track length estimate of energy deposition F8:P F8:E Energy distribution of pulsed created in a detector

15 Data cards (4.2 Tally specification cards) Tally (Fn) cards F2:N 8 $ Flux across surface 8 F4:N 2 $ Track length in cell 2 Tally Energy (En) card E E2 1 12I 14

16 Data cards (5.1 Material specification cards) Mm ZAID 1 fraction 1 ZAID 2 fraction 2... m is the material number on the cell card Nuclide Identification Number (ZAID) : To identify the element or nuclide desired (ZZZAAA). ZZZ : Atomic number of the elements of nuclide AAA* : Mass number of the nuclide * For naturally occurring elements, AAA=000.

17 Data cards (5.2 Material specification cards) Nuclide fraction For H 2 0 M M For Air, Dry (near sea level)* M & Fraction > 0, atomic fraction Fraction < 0, weight fraction * From ESTAR (

18 Data cards (5.3 Material specification cards) The material cards for the sample problem M $ Oxygen 16 M $ Natural iron M $ Carbon

19 Data cards (6. Problem cutoffs) To terminate execution of MCNP NPS n History cutoff cards n is the number of histories to transport MCNP will terminate after NPS histories

20 Sample problem summary Sample Problem Input Deck C cell cards for sample problem :2:-3:4:-5:6 C Surface cards for sample problem 1 PZ -5 2 PZ 5 3 PY -5 4 PY 5 5 PX -5 6 PX 5 7 S $ Oxygen sphere 8 S $ Iron sphere C Data cards for sample problem IMP:N 1110 SDEF POS= F2:N 8 $ Flux across surface 8 F4:N 2 $ Track length in cell 2 E0 1 12I 14 M $ Oxygen 16 M $ Natural iron M $ Carbon NPS

21 Parallel Virtual Machine (PVM) Communication protocols to use MCNP5 with parallel capabilities Developed at Oak Ridge National Laboratory PVM must be started before MCNP can be executed $pvm pvm> quit Console: exit handler called pvmd still running $

22 How to run MCNP (I) MCNP uses several files for input and output File names cannot be longer than 8 characters File INP must be present as a local file MCNP will create OUTP and RUNTPE Default File Name in MCNP INP OUTP RUNTPE* Description Problem input specification Output for printing Binary start-restart data for expanded output printing, continue run, tally printing * After MCNP execution, RUNTPE has to be deleted.

23 How to run MCNP (II) MCNP execution line has the following form: 1 CPU mcnp5.pvm i=sim01 o=sim01o > sim01.out & 4 CPU mcnp5.pvm i=sim01 o=sim01o > sim01.out tasks 3x1 & sim01: MCNP input file sim01o: MCNP output file sim01.out: MCNP running status file > : What is printed on a monitor put into the following file. & : Background running

24 After MCNP execution Before sim01 After sim01.out, sim01o, runtpe

25 RUNTPE file It looks like junk! Just delete it.

26 Tally plot using RUNTPE

27 sim01.out (MCNP running status file) I like this line.

28 sim01o (MCNP output file)

29 Tallies F2 tally (1/cm 2 ) F4 tally (1/cm 2 )

30 Estimation of Monte Carlo errors(i) MCNP tallies are normalized to be per starting particle Printed output accompanied by relative error Estimated relative error defined to be one estimated standard deviation of the mean Sx The Central Limit Theorem states that as N approaches infinity 68% chance in 95% chance in x ( 1± R) x ( 1± 2R)

31 Estimation of Monte Carlo errors (II) Guidelines for Interpreting the Relative Error R Range of R Quality of the Tally 0.5 to 1.0 Not meaningful 0.2 to 0.5 Factor of a few 0.1 to 0.2 Questionable < 0.10 Generally reliable < 0.05 Generally reliable for point detectors Ref.: MCNP manual Relative error

32 Estimation of Monte Carlo errors (III) For a well-behave tally, R will be proportional to 1/ N where N is number of histories. To halve R, we must increase the total number of histories fourfold.

33 Practice 1. Running with a higher NPS Increase NPS from 1e5 to 1e6 at sim01 Open sim01 using pico Replace at NPS line with 1e6 Save as sim02 Exit from pico Running sim02 using parallel computing capability: $mcnp5.pvm i=sim02 o=sim02o > sim02.out tasks 3x1 &

34 Practice 1. 4 CPUs are running

35 Practice 2. At the sample problem, replace carbon with air in the cube box and compare results. (Density of air = g/cm 3 ) Atomic weight composition of air