abaqus2py

This module is most important part of this repo, it used to run abaqus python script with out abaqus GUI. Ideally, this module can handle any abaqus simulation as long as you have a well defined abaqus script. However, in order to make your script work with this module, you need to modify your script a little bit.

script requirements

1. if you prefer to use functional code, you have to modify your script into the following format:

def simulation(sim_info):
    # note: name of the function is optinal, you can name it whatever you want
    # sim_info is a dictionary contains all variables you need to run your simulation
    # for example:
    # sim_info = {"youngs_modulus": 1000, "poisson_ratio": 0.3}
    youngs_modulus = sim_info["youngs_modulus"]
    poisson_ratio = sim_info["poisson_ratio"]
    # write your whole abaqus script below:
    ...

def post_process(job_name):
    # note: name of the function is optinal, you can name it whatever you want
    # job_name is the name of the job you just run

    # write your post process code here
    ...

    # finally, return the result you want to save into pickle file
    results = {"stress": stress, "strain": strain}
    with open("results.p", "w") as f:
        pickle.dump(results, f)

2. if you prefer to use objected-oriented code, you have to modify your script into the following format:

class Simulation(object):
    def __init__(self, sim_info):
        # sim_info is a dictionary contains all variables you need to run your simulation
        # for example:
        # sim_info = {"youngs_modulus": 1000,
        #              "poisson_ratio": 0.3,
        #              "platform": "ubuntu",
        #              "job_name": "job"}
        self.youngs_modulus = sim_info["youngs_modulus"]
        self.poisson_ratio = sim_info["poisson_ratio"]
        self.platform = sim_info["platform"]
        self.job_name = sim_info["job_name"]
        # write write other variables you need here
        ...
        # do simulation
        self.simulation()
        # do raw post process right after simulation
        if self.platform == "cluster":
            PostProcess(self.job_name)


    def simulation(self):
        # write your whole abaqus script below:
        ...

class PostProcess(object):
    def post_process(self, job_name):
        # job_name is the name of the job you just run
        # write your post process code here
        ...

        # finally, return the result you want to save into pickle file
        results = {"stress": stress, "strain": strain}
        with open("results.p", "w") as f:
            pickle.dump(results, f)

basic usage

Once you have your script ready as the following format, then you can put your script to a folder you want. Here I want to show a simple example about the folder structure of your own script and abaqus2py module.

├── abaqus_sim_main/
|   ├── case.py
│   │   __init__.py
│   │
├── rvesimulator/
│   ├── src/
│   │   ├── abaqus2py/
│   │   ├── microstructure/
│   │   └── ...
│   ├── docs/
│   └── README.md
├── LICENSE
└── .gitignore

The above case, the folder abaqus_script is parallel to the repo folder rvesimulator. Then you can run your simulation by:

from rvesimulator.abaqus2py.abaqus_simulator import AbaqusSimulator
folder_info = {
"main_work_directory": os.path.join(os.getcwd(), "Data"), # you can change where to do your simulation here
"script_path": ".../abaqus_script", # absolute path to your script folder
"current_work_directory": "data", # random name
"sim_path": "case", # important part, this is the name of your script
"sim_script": "Simulation", # important
"post_path": "case",# important
"post_script": "PostProcess",# important
}
sim_info = {
            job_name: "job", # important
            platform: "ubuntu", # important
            }
# initiate the simulator
simulator = AbaqusSimulator(sim_info=sim_info, folder_info=folder_info)
# run the simulation
simulator.run()
# read back the results
results = simulator.read_back_results()

Note

where folder_info is a dictionary indicates where to find your abaqus script and where to run and save your simulation. In the way, it has a bit strict requirement on the folder structure of your script. For the sim_info dictionary, it encode the design of experiment to abaqus simulation. Also, the job_name and platform are important keys in the dictionary, you have to include them in your sim_info dictionary.