diff --git a/doc/clusters/README.md b/doc/clusters/README.md
index 600fa5e..8f0e47d 100644
--- a/doc/clusters/README.md
+++ b/doc/clusters/README.md
@@ -66,5 +66,7 @@ instead of `$HOME`. More documentation on setting the config file can be found
 
 
 # Submitting jobs on cluster
-Once you have setup the solver on your desired cluster, you can submit jobs using the
-`submit_*.sh` batch submission scripts provided as reference.
+Once you have setup the solver on your desired cluster, you can submit jobs through the
+batch submission scripts `submit_*.sh`. Alternatively, you may use the `submit.py`
+provided here to generate these batch submission scripts, which may come in handy when
+scripting to generate large number of simulation cases for phase space exploration.
diff --git a/doc/clusters/submit.py b/doc/clusters/submit.py
new file mode 100644
index 0000000..f5c4121
--- /dev/null
+++ b/doc/clusters/submit.py
@@ -0,0 +1,133 @@
+"""Script for generating submission files for cluster jobs"""
+from typing import Optional, Literal
+
+SUPPORTED_CLUSTERS = Literal["stampede", "expanse", "bridges"]
+STAMPEDE_INFO_DICT = {"account": "TG-MCB190004", "compute": "normal"}
+EXPANSE_INFO_DICT = {"account": "uic409", "shared": "shared", "compute": "compute"}
+BRIDGES_INFO_DICT = {"account": "mcb200029p", "shared": "RM-shared", "compute": "RM"}
+CLUSTER_MAP = {
+    "stampede": STAMPEDE_INFO_DICT,
+    "expanse": EXPANSE_INFO_DICT,
+    "bridges": BRIDGES_INFO_DICT,
+}
+
+
+def create_submit_file(
+    program_name: str,
+    environment_name: str,
+    cluster_name: SUPPORTED_CLUSTERS,
+    output_file_name: Optional[str] = "%x_%j.out",
+    error_file_name: Optional[str] = "%x_%j.err",
+    partition: str = "compute",
+    num_nodes: int = 1,
+    num_tasks_per_node: int = 4,
+    memory: int = 249325,
+    time: str = "48:00:00",
+    verbose: bool = False,
+    mail_user: Optional[str] = None,
+    mail_type: Optional[str] = "ALL",
+    other_cli_arguments: str = "",
+    submit_filename: str = "submit.sh",
+) -> None:
+    """
+    The submission script can be roughly divided into the following parts:
+    1. sbatch directives / headers
+    2. printing details of launched job
+    3. module and environment preparation
+    4. running executable
+    """
+    cluster_info_dict = CLUSTER_MAP[cluster_name]
+    f = open(submit_filename, "w")
+
+    # (1) Write sbatch directives
+    f.writelines(
+        [
+            "#!/bin/bash\n",
+            "\n",
+            f"#SBATCH -J {program_name.replace('.py', '')}\n",
+            f"#SBATCH -p {cluster_info_dict.get(partition)}\n",
+            f"#SBATCH -N {num_nodes}\n",
+            f"#SBATCH --ntasks-per-node={num_tasks_per_node}\n",
+            f"#SBATCH -t {time}\n",
+            f"#SBATCH --account={cluster_info_dict.get('account')}\n",
+            f"#SBATCH -o {output_file_name}\n",
+            f"#SBATCH -e {error_file_name}\n",
+        ]
+    )
+    if mail_user:
+        if "@" not in mail_user:
+            print(
+                f"WARNING: mail notification might not work without full email domain."
+            )
+        f.write(f"#SBATCH --mail-user={mail_user}\n")
+        f.write(f"#SBATCH --mail-type={mail_type}\n")
+    # only expanse can set memory
+    if cluster_name == "expanse":
+        f.write(f"#SBATCH --mem={memory}M\n")
+    if verbose:
+        f.write("#SBATCH -v\n")
+    f.writelines(["\n"])
+
+    # (2) Print details about launched job
+    f.writelines(
+        [
+            "date\n",
+            "echo Job name: $SLURM_JOB_NAME\n",
+            "echo Execution dir: $SLURM_SUBMIT_DIR\n",
+            "echo Number of processes: $SLURM_NTASKS\n",
+            "\n",
+        ]
+    )
+
+    # (3) Prepare modules and environment
+    f.write(f"module reset\n")
+    if cluster_name == "stampede":
+        # Other mpi libraries are loaded by default (intel mpi)
+        f.writelines([f"module unload python2\n", f"module load phdf5\n"])
+    elif cluster_name == "expanse":
+        f.write("module load gcc openmpi fftw hdf5 anaconda3\n")
+    elif cluster_name == "bridges":
+        f.write(
+            "module load anaconda3 gcc/10.2.0 openmpi/4.0.5-gcc10.2.0 "
+            "phdf5/1.10.7-openmpi4.0.5-gcc10.2.0 fftw\n"
+        )
+    # activate python environment
+    f.write(f"source deactivate\n")
+    f.write(f"source activate {environment_name}\n")
+
+    # (4) Run executable
+    f.write("\n")
+    if cluster_name == "stampede":
+        f.write(f"ibrun python -u {program_name} {other_cli_arguments}\n")
+    else:
+        f.write(
+            f"mpiexec -n ${{SLURM_NTASKS}} python -u {program_name} {other_cli_arguments}\n"
+        )
+
+    f.close()
+
+    return submit_filename
+
+
+if __name__ == "__main__":
+    PROGRAM_NAME = "flow_past_sphere_case.py"
+    ENVIRONMENT_NAME = "sopht-mpi-env"
+    PARTITION = "compute"
+    TIME = "06:00:00"
+    NUM_NODES = 2
+    NUM_TASKS_PER_NODE = 32
+    MAIL_USER = "email@email.com"
+    CLUSTER_NAME: SUPPORTED_CLUSTERS = "stampede"
+
+    submit_filename = create_submit_file(
+        program_name=PROGRAM_NAME,
+        environment_name=ENVIRONMENT_NAME,
+        cluster_name=CLUSTER_NAME,
+        time=TIME,
+        partition=PARTITION,
+        num_nodes=NUM_NODES,
+        num_tasks_per_node=NUM_TASKS_PER_NODE,
+        mail_user=MAIL_USER,
+    )
+
+    print(f"Generated submission script file : {submit_filename}")