Command-Line Reference

Prints information about the simulation file: the version of Simcenter STAR-CCM+ with which that file was saved (in this example 19.04.###), number of partitions, file size, and number of iterations in the file:

STAR-CCM+ 19.04.### (linux-x86_64-2.5/gnu4.6) Serial 680.34 KB  Iter 66

Activates an .ini file that contains general and client options.

For the client, this option turns off broadcasting of queries to locate servers on the network; it also prevents the server from responding to any locator queries. The broadcast from the client is a UDP multicast, along with TCP pings, to known hosts. This multicast search is started by default when you launch the Simcenter STAR-CCM+ client without a simulation file. In Windows this option can cause network access warnings from your personal firewall.

Launches the Simcenter STAR-CCM+ workspace and starts a new simulation without accessing a simulation file.

Prints process IDs. An example of the output format is as follows:

Host ID 0 Rank 0 on marlin32 with process ID 7270
Host ID 0 Rank 1 on marlin32 with process ID 7422
Host ID 0 Rank 2 on marlin32 with process ID 7591

Activates Verbose mode, prints environment changes and sub-commands.

Requests a Lite Session license as opposed to a full license. Only a reduced functionality set is available.

Starts Simcenter STAR-CCM+ with the Power Session license feature.

Uses a Power Session license for the optimization session license.

Prevents main licenses from being used to enable cores for parallel computation. Only parallel licenses are used to enable cores for parallel computation.

Uses 10 DOEtoken licenses to launch a Simcenter STAR-CCM+ pre- and post-processing session.

Specifies to only use DOEtoken licenses for optimization studies. With this option, you can also use DOEtoken licenses for regular parallel runs in Simcenter STAR-CCM+.

Specifies to not use DOEtoken licenses for optimization studies.

Changes the order in which the DOEtoken license feature is checked out. Using this option checks out the hpcdomains license feature before the DOEtoken license feature for optimization studies.

Specifies a Power on Demand license key.

Specifies a DOE UUID (universal unique identifier).

Specifies a license path that overrides the default license path. When this option is used, Simcenter STAR-CCM+ uses the path to define a new CDLMD_LICENSE_FILE environment variable for the spawned server session.

Allows you to specify add-on licenses that can be automatically checked out.

Prevents add-on licenses from being checked out unless a model operation requires them.

Prevents add-on licenses from being retried when a significant operation is carried out.

Prevents add-on licenses from being released after importing geometry.

Starts Simcenter STAR-CCM+ in batch mode (without a GUI). The following additional commands are available:

• mesh — runs the meshing pipeline
• run — runs the solver until the stopping criteria are satisfied
• step — steps the solution
• <macro>.java — runs the specified Java macro file

If no command is specified, the default action is to run the solver. In each case, the simulation is auto-saved through the auto save feature (unless that property is turned off).

When applying simulation operations, if a chain of simulation operations is set active, the - batch without any [command] executes the operations in batch mode. If no chain of simulation operations is active, - batch steps the simulation by ignoring the simulation operations. However, once an additional command is included, the batch execution does NOT include simulation operations, no matter there is an active chain or not.

You can specify multiple commands within the -batch argument. In the following example, three macros are run in order before running the mesh pipeline and the solver:

-batch macro1.java,macro2.java,macro3.java,mesh,run

Generates a summary report of a batch simulation. The report appears in .html format. This report includes details about the computer system where the simulation was run, as well as details about the simulation itself.

Overrides the User macro classpath setting from the Options dialog. This option only affects the macro compilation in the current session of Simcenter STAR-CCM+.

This option permits you to add directory paths explicitly to the Simcenter STAR-CCM+ client library search path (LD_LIBRARY_PATH) for the client runtime environment.

Accesses a saved configuration of server options. (These options are also listed in the Saved Configurations list of the dialog that loads a simulation or creates a new one.)

Saves a configuration of server options and adds it to the Saved Configurations list of the dialog that loads a simulation or creates a new one.

Adjusts the size of the font that is used in the Simcenter STAR-CCM+ client workspace, such as dialogs and the object tree, for that session. Add the desired font size with this command.

For example, a command of starccm+ -fontsize 16 sets the font size to 16 points.

Specifies the host and port of a running Simcenter STAR-CCM+ server process to which the client connects. If the -host and -port options are used, a simulation file cannot be specified.

Overrides the default locale of your operating system. You can run Simcenter STAR-CCM+ in the following languages:

• English US: en:US
• Japanese: ja
• Chinese (simplified): zh
• Korean: ko
• Russian: ru:RU

Runs the macro file upon client startup. This option is similar to -batch in that a macro file is read, but in this case a GUI is opened. This option requires specification of a simulation file. Since it is impossible to record a macro that includes the opening of a simulation file, there is never a conflict.

Lets you specify a location, other than $HOME, in which the .star-*-*-* directory is written. Enter this command after client startup.

This option changes the location of key Simcenter STAR-CCM+ information, including the user settings directory and client log messages. Consider using this option when you do not have write access to your $HOME directory, or when you want to share modules.

Prevents the server process from shutting down after batch file execution has completed. By default, the client and server processes both shut down after the macro completes.

Specifies the hostmachine on which to start the server process. By default, when a new simulation is created or an existing simulation file is opened, the Simcenter STAR-CCM+ client starts a server process on the local machine. If the -on option is specified however, the client attempts to start the server process on the named host machine using the appropriate remote shell.

Additionally, this option can be used directly as a parallel option to specify the hosts that the parallel server uses.

Specifies the value of a global parameter that already exists in the target simulation. You can set scalar and vector parameter values using either fixed values or expressions, and can also set file paths for file parameters. Multiple instances are allowed.

If either argument (name or value) contains spaces or reserved characters such as '$', surround the argument in quotes, either single or double, depending on the command shell and your objective.

Examples

starccm+ -param MaxTime 10.0 demo.sim

starccm+ -param MinWidth 2cm demo.sim

starccm+ -param Sample [0.1,0.2,0.15] demo.sim

starccm+ -param MaxTime 10.0 -param MinWidth 2cm demo.sim

Linux Quoted Example

starccm+ -param MaxWidth '$MinWidth * 2' demo.sim

Windows Quoted Example

starccm+ -param MaxWidth "$MinWidth * 2" demo.sim

To set a file parameter for a specific simulation:

-param <file parameter name> <path to file> <path to simulation>

When you enclose the <path to file> in quotes, the path can use separators of either type (forward slash or back slash). Without quotes, the <path to file> must use forward slash separators for Linux, and back slash separators for Windows.

Example

starccm+ -param demoparameter demofile.txt demo.sim

Restores the default client layout (including window positions) and settings (options and mounted directories).

Prints server-elapsed and CPU times after a batch run. Specifically, it prints a summary of computational resource use of the server on close-down.

In serial:

Saving: /home/ncr/sims/suprBump@00077.sim
Server Elapsed Time :          5.362  (s)
Server CPU Time     :          0.830  (s)
Closing connection to server and exiting

In parallel:

Auto save: /home/ncr/sims/suprBump@00077.sim
Saving: /home/ncr/sims/suprBump@00077.sim
|Global Rank|Server Elapsed Time (s)|Server CPU Time (s)|MPI Overhead (s)|MPI Overhead %Elapsed Time (s)|
|          0|                  7.120|              1.250|           0.461|                         6.476|
|          1|                  7.120|              5.150|           2.371|                        33.302|
|          2|                  7.120|              5.120|           2.425|                        34.056|
|          3|                  7.120|              5.160|           2.451|                        34.421|
Maximum worker communication time on rank     : 3
Minimum worker communication time on rank     : 0
Average worker communication time             : 1.927
Standard deviation worker communication times : 0.847
Closing connection to server and exiting

On Linux, adds a user settings directory to your home directory with name optionally added to the folder name.

On Linux, specifies the NetBeans directory.

On Linux, reserves one core for running the client in a batch system. Use this flag if you perform a lot of post-processing while running in parallel.

On Linux, sets the client LD_PRELOAD environment variable.

This option allows you to override the automatic determination of the architecture of the machine. The specified architecture must exist within the installation and must be compatible with the machine architecture. If these conditions are not met, Simcenter STAR-CCM+ exits, or runtime errors occur.

For example, this argument can work when a 32-bit Linux installation is executed on a 64-bit Opteron-based architecture:

-arch linux-x86-2.2.5

This option can be used to aid the startup of parallel Simcenter STAR-CCM+ under the control of a supported Batch Management System (for usage see Using Batch Management Systems).

Currently the supported systems are:

• PBS (<system>=pbs): Open PBS, PBS Pro, Torque
• SLURM (<system>=slurm)
• Grid Engine (<system>=ge): Sun/Open Grid Engine and variants
• LSF (<system>=lsf)

Process distribution follows the same logic as for -machinefile. Implies parallel execution. Cannot be used with -machinefile or -on.

In the startup of a server, allows others to connect to it. (By default only the user who starts a Simcenter STAR-CCM+ server is allowed to connect to it.)

Messages are directed to the console rather than to the Simcenter STAR-CCM+ client.

This option permits you to add directory paths explicitly to the Simcenter STAR-CCM+ server library search path (LD_LIBRARY_PATH) for the server runtime environment.

The -ldlibpath option is required when Simcenter STAR-CCM+ relies upon third-party software that is installed in custom locations. In addition, consider using it for user code in the following instances:

• Linking against a math library like IMSL or NAG. (Getting access to the functions in them is one reason to write user code.)
• Most compilers that can build a .so file dynamically link it against their own runtime libraries. Add paths to these libraries, unless they happen to be the system libraries (as they are with GNU compilers on Linux).

Loads the specified simulation file as soon as the server starts. Without this option, the simulation database is loaded on demand when the client connects to the server.

Submits a cluster job to the Windows Server job scheduler. This functionality is available only on Windows Server 2012 machines.

Specifies server port for Simcenter STAR-CCM+ to open.

Specifies the number of ports the server can attempt to bind to.

Starts the Simcenter STAR-CCM+ server with an automatically generated port number.

• -rsh <option>
• -svrrsh <option>

When starting a server using -server, the -rsh or -svrrsh options default to ssh.

In a new client session the remote shell settings default to those set by the environment options, Generic Remote Process Launch Command and the Server Remote Process Launch Command.

• auto – (default) prefers supported native OpenGL over Mesa OpenGL; this option has the same effect as not using any -graphics option.
• native – enforces native OpenGL
• mesa – enforces Mesa OpenGL; uses the legacy Mesa graphics driver LLVMpipe.
• mesa_swr – enforces the OpenSWR Mesa driver which gives better performance than the legacy option. This option requires AVX CPU instruction set support (all CPUs released after 2012 support this instruction set).

Leverages multiple GPUs on a Linux system for local interactive or batch sessions, or for rendering on a remote server. The options are:

• N – selects a specific GPU in the system to use, for example -rgpu 1 or -rgpu 5.
• auto – cycles through all available GPUs on the system on a file in $TEMP.

-agpu is the equivalent argument for ad hoc rendering. For details about this type of rendering, see Additional Information on Remote Rendering.

For more information see Working with Multiple GPUs (Linux Only).

• auto – (default) prefers supported native OpenGL over Mesa OpenGL; this option has the same effect as not using any -rgraphics option.
• egl – enforces native OpenGL on EGL (Linux only)
• native – enforces native OpenGL
• mesa – enforces Mesa OpenGL; uses the legacy Mesa graphics driver LLVMpipe.
• mesa_swr – enforces the OpenSWR Mesa driver which gives better performance than the legacy option. This option requires AVX CPU instruction set support (all CPUs released after 2012 support this instruction set).

-agraphics is the equivalent argument for ad hoc rendering. For details about this type of rendering, see Additional Information on Remote Rendering.

Client option. Sets the number of threads N available for visualization tasks, including advanced rendering. When using this setting, consider the following:

• If you specify a high value for render threads, you can incur resource contentions that interfere with the proper functioning of your computer.
• Increasing render threads is expected to scale performance linearly. However, linear scaling can fail with high thread counts on hyper-threaded or multi-CPU systems, although performance can still increase. Complexity of the scene can also impact performance and prevent scaling.
• The value specified with the -rthreads command line option takes precedence over the corresponding setting that is available through Tools > Options.
• Cannot be used in conjunction with the -batchsystem option. If it is, then -rthreads defaults to a value of 1. If you attempt to specify a -rthreads value greater than 1 when using -batchsystem, Simcenter STAR-CCM+ generates a warning message.

Client option that tells the server to launch a renderer on a specified host:

• The argument hostname:port tells the server to use a renderer that was previously launched on some other machine (launched using starrenderer).

Client option that tells the server to launch a temporary renderer on a specific host:

• The argument hostname tells the server to launch a renderer on some other machine (hostname) and then use it. For details about ad hoc rendering, see Additional Information on Remote Rendering.

Server option which controls the number of threads N used by the rendering processes that can be spawned by the server for clients connecting with -rr. If Simcenter STAR-CCM+ is started with -rr -rrthreads N then the (batch) client uses remote rendering and the server that is launched uses N threads for remote rendering

-arthreads is the equivalent argument for ad hoc rendering. See Additional Information on Remote Rendering.

Assesses performance capabilities of your hardware. For detailed information about options, use -benchmarkhelp. For more information about benchmarking, see Performance Benchmarks.

MPI machine resource file name. Each hostname must appear on a separate line and can be appended by a colon and the number of processes on the host. Lines beginning with the hash symbol are interpreted as comments. The host loads are assigned in a round-robin fashion until the requested process count from -np is satisfied. Then processes are allocated to the hosts using that loading and filling all the CPUs/Cores of a host before allocating to the next host. Implies parallel execution. Cannot be used with -on or -batchsystem.

-mpi <driver>

-mpidriver <driver>

Specifies the MPI driver used for communication between processes. The default MPI driver on Linux is Open MPI and on Windows it is MS MPI. See Supported MPI Implementations for details.

MPI driver expert options.

Performs MPI tests and exits.

Number of processes to use. Implies parallel execution. If the -np option is omitted and -machinefile, -on, or -batchsystem is specified, the number of processes is assumed to be the number that the resource list prescribes.

Number of processors to use for co-simulation, where np is the number of processes for Simcenter STAR-CCM+ and npcosim is the number of processes for the other tool.

MPI resource list. Process distribution follows the same logic as for -machinefile. Implies parallel execution. Cannot be used with -machinefile or -batchsystem.

Activates parallel input/output on RAID-5 file systems. When this flag is included, the given simulation file is restored using parallel I/O, and all subsequent simulation files are saved using parallel I/O.

Specifies what binding policy out of off, bandwidth, and latency to use when binding CPUs.

You can request a report to display the binding layout by entering: -cpubind v

See CPU Binding for details.