MAR-2016 update

3-MAR-2016: Update to the old model.

Added routine kcsv4d to module io.f90 in branch cafe.old. This routine dumps to file the state of CA arrays for all FE with CAFE VUMAT properties. Selected states only are dumped.

The array is dumped as a formatted sequential file:

x,y,z,state
7.5459190756623711,.35097225136375326,.29142071312117274,1
7.7544983625411987,.42049868032336235,.70857928687882721,1
7.8240247915008077,.76813082512140785,.43047357104039091,1
8.0326040783796344,.21191939344453509,.22189428416156365,1
8.0326040783796344,.55955153824258053,.77810571583843635,1
8.0326040783796344,.69860439616179870,.50000000000000000,1
7.3678879248982057,-.14578446221460550,.23333783029371302,2
7.3678879248982057,-.14578446221460550,.30000337272028477,2
7.3678879248982057,-.14578446221460550,.36666891514685651,2
...

The header line is useful when the file is read into ParaView. I use CSV time series, e.g.:

zzz_ca_B.csv.1  zzz_ca_B.csv.16 zzz_ca_B.csv.22 zzz_ca_B.csv.29 zzz_ca_B.csv.35
zzz_ca_B.csv.10 zzz_ca_B.csv.17 zzz_ca_B.csv.23 zzz_ca_B.csv.3  zzz_ca_B.csv.4
zzz_ca_B.csv.11 zzz_ca_B.csv.18 zzz_ca_B.csv.24 zzz_ca_B.csv.30 zzz_ca_B.csv.5
zzz_ca_B.csv.12 zzz_ca_B.csv.19 zzz_ca_B.csv.25 zzz_ca_B.csv.31 zzz_ca_B.csv.6
zzz_ca_B.csv.13 zzz_ca_B.csv.2  zzz_ca_B.csv.26 zzz_ca_B.csv.32 zzz_ca_B.csv.7
zzz_ca_B.csv.14 zzz_ca_B.csv.20 zzz_ca_B.csv.27 zzz_ca_B.csv.33 zzz_ca_B.csv.8
zzz_ca_B.csv.15 zzz_ca_B.csv.21 zzz_ca_B.csv.28 zzz_ca_B.csv.34 zzz_ca_B.csv.9

and display data as points.

This simulation is crack propagation in a 3-point impact bend of a notched bar test, i.e. the Charpy impact test. The bar is 50x10x10 mm, i.e. 50 mm long and 10x10mm cross section. The "V"-shaped notch is 2mm deep and is normal to axis 2 (Y). The crack is propagating along axis 1. The Abaqus FE simulation of the Charpy test at -100^oC is shown below.

$The Abaqus animation of the fracture propagation in the Charpy test.$

In the animations below I dump only "dead by brittle" (state 2, yellow), "dead by ductile" (state 0, magenta) and "alive with GB carbide" (state 1, blue).

Another simulation at the same temperature. The CA simulation is stochastic, so each simulation prediction is unique.

UP | next

validate this page