Search Contact information
University of Cambridge Home Department of Engineering
University of Cambridge >  Engineering Department >  computing help


2. Jobs

Q2.1 : How do I run small jobs?

Use the following command
abaqus job=job-id interactive

Example : abaqus job=plate inter

Q2.2 : How do I go about running many small jobs?

Create a batch file (say aba-multrun.bat) with one line per analysis as shown below :
abaqus job=analysis-a interactive
abaqus job=analysis-b interactive
abaqus job=analysis-c interactive
abaqus job=analysis-d interactive
Then make the file an executable using the following unix command :
chmod u+x aba-multrun.bat 
Then type aba-multrun to execute the ABAQUS jobs one at a time while you are logged ON. This is only suitable for small jobs which only take a few minutes to run. These jobs will run one at a time and in the sequence in which these appear in the batch file.

This is preferable to submitting all the jobs at the same time (for example typing the above commands directly at the terminal without the interactive parameter). This will put a strain on the server and its resources and inconvenience the other users as well.

For medium to large jobs use the batch command available in the CUED teaching system.

If ABAQUS is installed on a PC (Windows XP O/S) then the above batch file requires a modification. Enclose the lines with a round brackets. in the absence of this bracket only the very first job in the file will be run.

abaqus job=analysis-a interactive
abaqus job=analysis-b interactive
abaqus job=analysis-c interactive
abaqus job=analysis-d interactive

Q2.3 : How do I run large jobs using batch?

One need to use the GRID-ENGINE to run long jobs of ABAQUS. see the answer to Question 2.12.

Q2.4 : How do I run ABAQUS/CAE?

Using the following command :
abaqus cae 

Q2.5 : How do I run ABAQUS/Viewer?

Using the following command :
abaqus viewer 

abaqus viewer database=cantilever

Example : abaqus viewer database=cantilever

Q2.6 : How do I get copies of the ABAQUS examples input data files (*.inp)?

There are 2 options. Cut and paste directly from the screen once from the on-line documentation of the Input Files. The other option is to use the fetch command along with the name of the analysis-ID. First find the name of the input file from the Examples manual (available on-line). This will copy the 1010101.inp file to the current directory.

For the Laminated Composite Plate failure example (Example 1.1.13 in version 6.3) the input files have the starting name damagefailcomplate. There are a total of ten files listed. Here * (wildcard character) can be used to complete the file name in the fetch command. abaqus fetch command.

Example : abaqus fetch job=damagefailcomplate*

This will fetch the following files to the current directory :

  inflating: damagefailcomplate_cps4.f  
  inflating: damagefailcomplate_cps4.inp  
  inflating: damagefailcomplate_cps4r.f  
  inflating: damagefailcomplate_cps4r.inp  
  inflating: damagefailcomplate_element.inp  
  inflating: damagefailcomplate_node.inp  
  inflating: damagefailcomplate_s4.f  
  inflating: damagefailcomplate_s4.inp  
  inflating: damagefailcomplate_s4r.f  
  inflating: damagefailcomplate_s4r.inp 

Q2.7 : How do I run a ABAQUS job which uses a user subroutine?

Using the following command :

abaqus job=job-id user=< subroutine-filename >

For the Linux Teaching System Computers give the extension .f and for PCs give the extension .for.

The user subroutine has to be in a separate file (say my_material.f).

Example : abq681 j=plate user=my_material

If there are more than 1 user subroutines then all the subroutines can be included in a single file.

Q2.8 : How do I run a user written post processing program which accesses the *.fil file?

Using the following command :

abaqus make job=job-id user=name-of-file
Example : abaqus make job=cantilever user=disp1
This will compile the user program in a file called disp1.f and then create an executable called cantilever.x. Type cantilever.x to run this program.

Q2.9 : How do I find out about the different execution procedures that are available with ABAQUS?

Type abaqus help and this will list all the abaqus execution procedures. These are listed below :

 Execution Procedure for ABAQUS/Standard and ABAQUS/Explicit
 abaqus job=job-name  [ analysis | datacheck | continue | help | recover |
                        convert={restart|select|all} | 
                        information={environment|local|memory|release|status} ]
                      [ input=input-file    ]  [ user=source-file ]
                      [ oldjob=oldjob-name  ]  [ fil={append|new} ]
                      [ globalmodel=results file-name ] [ double ]
                      [ memory=memory-size  ]  [ buffer=buffer-size ] 
                      [ interactive | background | queue=queue-name ]
                      [ cpus=number-of-cpus ] [scratch=scratch-dir]
                      [ subcomplex=subcomplex-name]

 Note: subcomplex is only valid on the Convex Exemplar 
 Execution Procedure for ABAQUS/Abares
 abaqus abares job=job-name [ restart=restart-name ]
                            [ beginstep=step-number ] [ endstep=step-number ]
                            [ increment={all|endstep|final|none|integer-list} ]

 Execution Procedure for ASCII translation of results (.fil) files
 abaqus ascfil job=job-name [ input=input-file ]

 Execution Procedure for on-line documentation 
 abaqus doc

 Execution Procedure for ABAQUS/Append

 abaqus append job=job-name oldjob=oldjob-name input=input-file

 Execution Procedure for ABAQUS/Fetch

 abaqus fetch job=job-name [ input=input-file ]

 Execution Procedure for ABAQUS/Findkeyword

 abaqus findkeyword [ job=job-name ] [ maximum=maximum-matches ]

 Execution Procedure for ABAQUS/Make

 abaqus make job=job-name [ user={source-file|object-file} ]

 Execution procedure for suspending or resuming analysis jobs

 abaqus [ suspend | resume ] job=job-name

Execution procedure for fixed format translator

 abaqus free job=job-name input=input-file

 Execution Procedure for input file upgrade utility
 abaqus upgrade job=new-input-file input=old-input-file

 Execution Procedure for output database file upgrade utility
 abaqus upgrade job=new-output-database-file odb=old-output-database-file

 Execution Procedure for the movie player utility

 abaqus movieplayer [ movie=animation-file ]

 Execution Procedure for ABAQUS Verification

 abaqus verify [ -all | -std | -user_std | -exp | -user_exp 
                 -cae | -viewer | -design | -make | -param 
                 -scripting | -adams | -cmold | -moldflow 
                 -catia | -ideas | -parasolid | -proe 
                 -log ]

Q2.10 : How do I find what the current settings are for the environment variables?

Type abaqus info=environment and this will list all the current setting of the ABAQUS environmental variables.

memory, local, release, status are other options on which you can get more information on.

For ABAQUS jobs you have run look for *.com file. For the current session in #the Teaching System use the printenv command.

For PCs type "set" in a MS-DOS command window at the DOS prompt (>).

Q2.11 : How do I change the current settings of the environment variables?

Create a file called abaqus_v6.env in the directory from which ABAQUS is run which contains lines of environment variables you want to change set equal to new values. However make sure that the computer on which you are running ABAQUS can support the changes. For example you can increase the memory used by ABAQUS. But you cannot increase this beyond what is available in the computer.

Example : Include the following lines in the abaqus_v6.env file to increase the size of pre_memory and the standard_memory used by ABAQUS.

For earlier versions of ABAQUS :

pre_memory="400 mb"
standard_memory="500 mb"

If running ABAQUS on a PC then make these to the system copy which will be in the ABAQUS installation directory :


Q2.12 :Can one run long ABAQUS jobs in the Teaching System Computers?

One needs to use the GRID-ENGINE to run long runs of ABAQUS. see ABAQUS on GRID-ENGINE

Q2.13 : When re-running a ABAQUS job getting the old files existing message. Is it possible to over-ride it without having to type Y?

This could also be a problem if ABAQUS is run in an optimisation loop many times automatically.

This can be done by adding the following line to the abaqus_v6.env in the directory from which ABAQUS is run (for example).


If running ABAQUS on a PC then make this change to the system copy which will be in the ABAQUS installation directory :


This needs to be used with care. Accidentally typing the name of a previously carried out analysis (which may have been a long run) which you didn't want to be deleted/overwritten can easily be lost with the above setting (as there is no safety net).

When not required change the setting / or comment it out / or delete it.

# comment it out
# ask_delete=NO

# Reset it


Q2.14 :How do I stop a ABAQUS analysis job I had submitted?

From a command prompt window (DOS Window) from the same directory from which the job was submitted. Re-type job submission command but this time add suspend to suspend the job execution.

ABAQUS job=plate suspend

This allows you to review matters. If you then decide then you want to terminate the job Re-type the job submission command but use the option terminate instead.

ABAQUS job=plate terminate

A run can be terminated using terminate option without having to suspend the job in the first place. A sup ended job can be continued by using the option resume.

ABAQUS job=plate resume

© Cambridge University Engineering Dept
Information provided by Arul M Britto (amb2)
Last updated: 23 June 2009