Understanding the files generated by an optimization process

When an optimization process is executing, it creates two types of data that are saved in separate files—optimization results and analysis results.

Optimization results

The optimization results consist of the optimization variables and the optimization values. The optimization variables depend on the type of optimization you are performing.

Topology optimization

The optimization variable is the normalized material distribution variable (MAT_PROP_NORMALIZED).

Shape optimization

The optimization variables are the displacement variables:

  • DISP_OPT: a vector indicating the direction in which the nodes were moved during the shape optimization. Because of mesh smoothing and filtering, the vector may not be coincident with the node normal vector.

  • DISP_OPT_VAL: the magnitude of the DISP_OPT vector with a sign indicating the direction of the displacement—positive for growth and negative for shrinkage.

  • CTRL_INPUT: the value of the objective function at each node (stress, for example).

Sizing optimization

The optimization variables are the shell thickness and the change in shell thickness (THICKNESS and DELTA_THICKNESS).

Bead optimization

The optimization variables are the displacement variables:

  • DISP_NORMAL_VAL: the magnitude of a vector indicating the direction in which the nodes were moved during the bead optimization along the node normal vector.

  • DISP_OPT: a vector indicating the direction in which the nodes were moved during the bead optimization. Because of mesh smoothing and filtering, the vector may not be coincident with the node normal vector.

  • DISP_OPT_VAL: the magnitude of the DISP_OPT vector with a sign indicating the direction of the displacement—positive for growth and negative for shrinkage.

The optimization variables are saved in separate optimization (.onf) files after every design cycle. In addition, the optimization values, such as the value of the objective and the constraints, are written to a comma-separated text file (.csv) after each design cycle. (You are viewing the optimization values when you monitor the progress of an optimization process, as described in Monitoring your optimization process.)

Analysis results

The analysis results are the field and history data generated by Abaqus during the analysis. During each design cycle a new output database file is created. In the initial design cycle, Abaqus writes both field and history data to the new output database file; however, during subsequent design cycles, Abaqus writes only field data. Abaqus also generates data (.dat), message (.msg), and status (.sta) files during each design cycle. To conserve disk space and to speed up postprocessing, the output database file and the data, message, and status files are saved only at specified intervals; by default, after the initial, first, and final design cycles. (You specify the rate at which the files are saved when you create the optimization process, as described in Creating and editing optimization processes.) In addition, Abaqus saves the new input file that is generated during each design cycle.