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

The MSC/PATRAN FAQ


2. Files/Databases

Q2.1 : After running PATRAN I have several files created in my home directory. Which files do I keep?

You need to keep the following files :
*.db,    *.db.jou

Example : plate1.db,  plate1.db.jou
These are the essential files. First is the patran database. The second is the journal file which can be used to re-create the database in most cases.

The following session files can also be kept. These can be used to run demos.

patran.ses.*

The following are the patran message files :

*.msg.*
Unlike a single journal file there are one session file per patran session. The following are the additional ABAQUS files :
*.inp
*.msg
*.dat
*.fil
*.log
*.com
None of the ABAQUS files need to be kept.


Q2.2 : Have set up the f.e. data. Ran the ABAQUS analysis. No errors reported in the *.msg.* file. However the *.fil does not show up under the list of "Results File" on the "Read Results" form. What do I do?

In the odd instant even if the *.fil exists it may not show up under the list of "Results File" on the "Read Results" form when you click on the filter button.

To check use a separate X-term window and go to the directory from which you are running patran from (this is listed in the top of the "Read Results" form. Type ls -l *.fil and this will list all the files with the extension fil in that directory. If the *.fil file exists then go back to Patran and click on the "filter" button again. If it does not list the *.fil file in question then type in the fil file name (including the full path name) in the box marked "Selected Results File" and click on the "OK" button. This hopefully should read the results from that *.fil file.


Q2.3 : After creating a separate directory for running Patran, I can't find any of the Patran files in that directory?

In the "Open/New Database" form it is possible to move to directories other than from which Patran was invoked/started (using the filter box). If you had not invoked Patran from that directory then the Patran files will not be placed in that directory.

For example if you are a first time user you may have a created a separate directory but forgetting to change to that directory before invoking Patran is a common mistake.

Go to your HOME directory using cd and then use the following command to search for the patran database.

find ./ -name "*.db"

The other Patran files will be found in the directory from which you invoked PATRAN. If necessary move these files to the desired directory using the mv command (make sure you are NOT running PATRAN while you are doing this).


Q2.4 : I have accidentally deleted the patran database (*.db) or it appears to be corrupted. What do I do?

First contact the computer operators (oper@eng.cam.ac.uk) to see whether it can be recovered from backup. This is the recommended choice.

If not you can use the corresponding journal file *.db.jou to re-build the database. For example if you have deleted the database plate1.db then look for the plate1.db.jou. If this file exists then start patran and then choose File / Utilities / Rebuild... and then choose the plate1.db.jou file and click on OK. This should re-create the database in most cases.

If the plate1.db.jou does not exist then you could try using the patran.ses.* file. This time choose File / Session / Play... and then choose the patran.ses.* file and click on OK. This is more likely to work if there was a single session file for the *.db in question ie all the work in creating the database was done in a single patran session.

If the work was carried out in several patran sessions then there will be one session file per session. Then these session files after minor edits to remove opening and closing of the said database can be run is sequence. However this method is frought with problems and the re-creation of the database may not be completely successful.


Q2.5 : The patran database I was using has grown into several Megabytes in size. What do I do?

First check the size of the Patran database (using ls -l *.db ). If at least the same amount of free disk space (use the quota or du command) is available then choose File / Utilities / Compact... and then choose the database. Then click on the OK button.

If not then you have to move the database into a temporary directory. Use the following commands to create a directory and move the database into it.

cd /tmp
mkdir uid ( Example : mkdir 94abc )
mv plate1.db /tmp/uid/. ( Example : mv plate1.db /tmp/94abc/. )

Then use the commands mentioned above. Once the database has been compacted quit from Patran and then move the database back to your home directory.

cd  (this should take it to your home directory)
cd p3 (if you are working on a separate directory called p3).
mv  /tmp/94abc/plate1.db .


Q2.6 : When I open existing PATRAN database the heartbeat continues to flash red indefinitely. What is the problem?

This can happen if using fvwm or one of its variants or Start or twm window manager in the HP teaching system. Find the process ID of the PATRAN process either using top or ps -eaf | grep "p3". Then killing the process off by typing kill -9 at a x-term window.

Exit from whatever window manager you were running and run the dtwm window manager and then open the PATRAN database in question. This should restore the database for use by the heartbeat turning to a steady green. However if this does not happen then the database is corrupt.

Then it might be possible to delete the database and then use the journal file (*.db.jou) to re-create the database. Under File... choose Utilities... and Rebuild and then choose the appropriate journal file.


Q2.7 : What do I need to specify to create a ABAQUS restart (*.res) file?

In the radio button Analysis form make the following selection :

Action : Analyze
Object : Entire Model
Method : Full Run

Click on Restart Parameters... . In the new form that pops up set

Restart Type : Write

Also set the parameter Increment between writing data to the appropriate number. Then click on the APPLY button.


Q2.8 : Why is that one has to give names to everything in PATRAN?

This is for easy grouping and assigning properties. If you have complex assembly with components having different physical and material properties then it makes it easier to assign different geometry groups when these components are created. Similarly the finite element entities when created could be in separate f.e. groups. This enables one to view and manipulate the groups individually and in combination.

The same applies when loading and boundary conditions are to be specified that these are also given meaningful names. Then these can be combined into one or more load cases.


Q2.9 : What is the significance of the names used for job, job step and load case in a ABAQUS analysis?

ABAQUS analysis consists of one or more STEPS. Each step consists of a load case. Whatever the loading and boundary conditions specified in a load case it is associated with a job step. A PATRAN job step corresponds to a ABAQUS STEP. The job steps have to be given names so that these can be selected. The names of the job steps also appears with each STEP in the ABAQUS input file.

If there are more than one step in the analysis then there should be one loadcase (and a job step) for each of the step. The loadcase name is used for identifying the results after the ABAQUS job has run and the results have been read into the PATRAN database ie during post processing.

The job name by default is the same as that of the PATRAN database. It is the name of the ABAQUS input file that will be created. For example if one is working on the database plate.db then the ABAQUS input file will be called plate.inp by default. However it need not be the same. For example if you are planning to run a number of analyses (say) for the purposes of a parametric study then the job name can be modified to represent this. Here "plate-a", "plate-b" are variants that can be used to represent different files but which is still associated with the riginal "plate" database.

© Cambridge University Engineering Dept
Information provided by Arul M Britto (amb2)
Last updated: 27 January 2009