- Akademische Forschung und Lehre
- Geräte des KIT
- Mitarbeiter des KIT
- Studierende des KIT (auch auf privaten Geräten)
- siehe Lizenzbedingungen
-
ANSYS Mechanical APDL
-
ANSYS Mechanical APDL is a finite element program for solving static and dynamic, linear and nonlinear strength problems, 2D flow problems and other field problems.
- Contact:
- Contact Person:
- Services:
About ANSYS Mechanical APDL
ANSYS Mechanical APDL is a finite element program for solving static and dynamic, linear and nonlinear strength problems, 2D flow problems and other field problems. ANSYS has its own integrated pre- and post-processors. The product has no node and element number limitations.
Current Licensed Software Version
ANSYS 2026R1
ANSYS 2025R2
Installation
The installation files and the installation document for the latest version of the software, along with network licensing information are available on our FTP server. ANSYS Mechanical APDL is also installed and available on our Linux clusters, BwUniCluster3.0 and Horeka to run computationally intensive simulations.
Note: Installation files for versions that are not listed above can be obtained upon request by contacting scs-contact∂scc kit edu.
Licensing
Network License: There are two separate ANSYS network license servers, both of which are currently active. The configuration details for both servers are provided in the network license configuration file available on the FTP server. During installation, please specify the appropriate license server name and port number as indicated in the configuration file, based on your intended use of the software. The two license servers serve distinct purposes and must be used accordingly:
- ANSYS Academic Research: Intended for non-commercial academic research activities conducted at KIT, including PhD research, Master's theses, Bachelor's theses, and other scientific projects.
- ANSYS Academic Teaching: Intended exclusively for teaching activities, such as lectures, courses, and exercises.
Please ensure that you select the appropriate license server based on your intended use of the software. Using either license for personal or commercial purposes is strictly prohibited. An active internet connection is required to use the software. If you are accessing it from outside the KIT network domain, please connect via VPN in order to reach the license server. For instructions on setting up and using the VPN, please refer to the link: Remote Access (VPN)
Basics
The ANSYS products are available on the bwUniCluster 2.0 (uc2).
The software can be installed on any institute machine.
Startup on the HPC-Cluster
In order to ensure a trouble-free operation of the computers, there are nodes on which interactive work can be done and which can be accessed e.g. through
ssh <username>@uc2.scc.kit.edu
logging in.
A module concept has been introduced on the HPC clusters. Therefore, before you can work with a product, you must load the corresponding module. You can get all available software modules from ANSYS by selecting
module avail cae/ansys
into the program. The module is called
2021R2: cae/ansys/2021R2
2022R2: cae/ansys/2022R2
2023R2: cae/ansys/2023R2
Load the module with
module load cae/ansys/2021R2
module load cae/ansys/2022R2
module load cae/ansys/2023R2
If possible, please always work with the latest versions.
However, a large computational run should always run on the nodes that are intended for it. This is regulated by special calls that request the resources for the job, such as CPU time, number of processors, size of main memory, etc.. These requests are handled by a job management system. Jobs can be submitted with the msub command. A detailed parameter list can be obtained by running
sbatch --help
is entered.
In addition to this command, there are other commands available which are described here.
Detailed information can be found on the web pages:
UC2: bwUniCluster 2.0 User Guide (Wiki)
Important: For batch jobs the first line of the input file must be /BATCH.
Parallel jobs on the bwUniCluster 2.0 uc2
There are now a total of 640 parallel tasks available.
First you have to create a small shell script:
Version 2023R2:
Intel MPI:
#!/bin/bash
module load cae/ansys/2023R2
export LD_PRELOAD=/opt/bwhpc/kit/cae/ansys/2023R2/ansys_inc/scc/strtok_proxy/strtok_proxy.so
cd Arbeitsverzeichnis/Workspace
export MACHINES=`/opt/bwhpc/kit/cae/ansys/2023R2/ansys_inc/scc/machines.pl`
ansys232 -dis -b -j lal -machines \$MACHINES < input_commands.txt
OpenMPI:
#!/bin/bash
module load cae/ansys/2023R2
cd Arbeitsverzeichnis/Workspace
export MACHINES=`/opt/bwhpc/kit/cae/ansys/2023R2/ansys_inc/scc/machines.pl`
ansys232 -dis-mpi openmpi -b -j lal -machines \$MACHINES < input_commands.txt
Version 2022R2:
Intel MPI:
#!/bin/bash
module load cae/ansys/2022R2
export LD_PRELOAD=/opt/bwhpc/kit/cae/ansys/2022R2/ansys_inc/scc/strtok_proxy/strtok_proxy.so
cd Arbeitsverzeichnis/Workspace
export MACHINES=`/opt/bwhpc/kit/cae/ansys/2022R2/ansys_inc/scc/machines.pl`
ansys222 -dis -b -j lal -machines \$MACHINES < input_commands.txt
Version 2021R2:
Intel MPI:
#!/bin/bash
module load cae/ansys/2021R2
export LD_PRELOAD=/opt/bwhpc/kit/cae/ansys/2021R2/ansys_inc/scc/strtok_proxy/strtok_proxy.so
cd Arbeitsverzeichnis/Workspace
export MACHINES=`/opt/bwhpc/kit/cae/ansys/2021R2/ansys_inc/scc/machines.pl`
ansys212 -dis -b -j lal -machines \$MACHINES < input_commands.txt
OpenMPI:
#!/bin/bash
module load cae/ansys/2021R2
cd Arbeitsverzeichnis/Workspace
export MACHINES=`/opt/bwhpc/kit/cae/ansys/2021R2/ansys_inc/scc/machines.pl`
ansys212 -dis-mpi openmpi -b -j lal -machines \$MACHINES < input_commands.txt
The working directory can start with $HOME, for example.
The default uc2 nodes have 40 cores.
If you want to use more than one node:
UC2:
sbatch -p multiple -N 2 -t 10:00 --ntasks-per-node=40 Shell-Script.txt
The ANSYS job will now be started with a time request of 10 minutes across the two nodes in Distributed Memory Mode and within one node in Shared Memory Mode. Please adjust the appropriate parameters to your problem size.
If one node is enough for you:
UC2:
Here you have to change the shell script a little bit:
Version 2023R2:
#!/bin/bash
module load cae/ansys/2023R2
cd Arbeitsverzeichnis/Workspace
ansys232 -b -smp -np 40 -j lal < input_commands.txt
and start the shell script with
sbatch -p single -t 10:00 --ntasks-per-node=40 -N 1 -t 10:00 --mem=4gb Shell-Script.txt
Version 2022R2:
#!/bin/bash
module load cae/ansys/2022R2
cd Arbeitsverzeichnis/Workspace
ansys222 -b -smp -np 40 -j lal < input_commands.txt
and start the shell script with
sbatch -p single -t 10:00 --ntasks-per-node=40 -N 1 -t 10:00 --mem=4gb Shell-Script.txt
Version 2021R2:
#!/bin/bash
module load cae/ansys/2021R2
cd Arbeitsverzeichnis/Workspace
ansys212 -b -smp -np 40 -j lal < input_commands.txt
and start the shell script with
sbatch -p single -t 10:00 --ntasks-per-node=40 -N 1 -t 10:00 --mem=4gb Shell-Script.txt
The ANSYS job is now started with a time request of 10 minutes and on the uc2 with a memory request of 4 GB on a node with 40 CPU cores in shared memory mode. Please adjust the appropriate parameters to your problem size.
If you want to use a GPU on the uc2:
Here you have to add one line to the shell script:
Version 2023R2:
#!/bin/bash
module load cae/ansys/2023R2
cd Arbeitsverzeichnis/Workspace
export KMP_AFFINITY=disabled
export MACHINES=`/opt/bwhpc/kit/cae/ansys/2023R2/ansys_inc/scc/machines.pl`
ansys232 -smp -acc nvidia -na 1 -b -j lal -machines \$MACHINES < input_commands.txt
Version 2022R2:
#!/bin/bash
module load cae/ansys/2022R2
cd Arbeitsverzeichnis/Workspace
export KMP_AFFINITY=disabled
export MACHINES=`/opt/bwhpc/kit/cae/ansys/2022R2/ansys_inc/scc/machines.pl`
ansys222 -smp -acc nvidia -na 1 -b -j lal -machines \$MACHINES < input_commands.txt
Version 2021R2:
#!/bin/bash
module load cae/ansys/2021R2
cd Arbeitsverzeichnis/Workspace
export KMP_AFFINITY=disabled
export MACHINES=`/opt/bwhpc/kit/cae/ansys/2021R2/ansys_inc/scc/machines.pl`
ansys212 -smp -acc nvidia -na 1 -b -j lal -machines \$MACHINES < input_commands.txt
Ansys is started with the request for a GPU (-acc nvidia -na 1). You then start the job with e.g.
sbatch -p gpu_4 --gres=gpu:1 -t 60:00 --ntasks-per-node=40 -N 1 --mem=200gb shell-Script.txt
start. If you want to use multiple GPUs, you must adjust the -na parameter in the shell script and the --gres sbatch command accordingly. As a rule, you should already achieve a good speedup with one GPU.
If you ever want to quickly start a job on a node and a core:
In addition, we have provided a small program that allows you to quickly start a job on a node:
usage: ans232job [-p PATH] [-c INPUT-FILE] [-T TIME] [-M MEM] [-j xxxx]
usage: ans222job [-p PATH] [-c INPUT-FILE] [-T TIME] [-M MEM] [-j xxxx]
usage: ans212job [-p PATH] [-c INPUT-FILE] [-T TIME] [-M MEM] [-j xxxx]
-p Name des Pfads in dem das Input-File steht. (notwendig)
Wenn ihre Dateien unter Ihrem HOME-Verzeichnis liegen,
muessen sie \$HOME voranstellen.
-c Name des Input-File. (notwendig)
-T n Laufzeit in CPU-Sekunden (notwendig
-M n Memorybedarf in MB. (notwendig)
-j Option um filenn.dat in xxxxnn.dat zu aendern (max. 4 Zeichen).
Example: ans232job -p \$HOME/ansysprojekt -c loch_k.txt -T 600 -M 4000 -j pr01
Example: ans222job -p \$HOME/ansysprojekt -c loch_k.txt -T 600 -M 4000 -j pr01
Example: ans212job -p \$HOME/ansysprojekt -c loch_k.txt -T 600 -M 4000 -j pr01
Info shell scripts:
Please note that shell scripts can only be executed if they have a so-called execute bit
-rwxr--r--. 1 xx4711 xxx 273 Apr 7 13:05 Shell-Script.txt
and have e.g. #!/bin/bash or #!/bin/sh in the first line.
You can achieve this by
chmod u+x <Shell-Script>
in the command line.
Start on own machines
ansys222 <parameter list>
ansys232 <parameter list>
ansys222 <parameter list>
ansys212 <parameter list>
will be done.
Documentation
The documentation is online
File Handling
The ANSYS default file names are of the form.
<jobname>.<ext>
For jobname, file defaults to file. The ext name extension is a file type specific addition with a maximum of 4 characters. Files are created in the current directory by default. Changing the default file name
The four-character default file name can be replaced with any other maximum eight-character string using the -j option. To assign a new file name ver1, ANSYS must be called in the following form:
ansys232 -j ver1
ansys222 -j ver1
ansys212 -j ver1
The new file names will then be ver1.<ext>.
ANSYS startup file
To create individual settings you can create a file with the name start.ans. This should basically be located in the current directory or in their login(root) directory. This is executed after the input of the first ANSYS command and acts as if they had entered /INPUT,start,ans. ANSYS looks for this file first in the current directory, then in the root directory of the current user, and only then in the installation directory. If the file is not found, ANSYS expects its next input. In this file all valid ANSYS commands can be used.