251 LSDYNA GENERIC DOE GENERATOR
===================================
.. _251doegenericdoegenrator:


Introduction
==============

The LSDYNA_GENERIC_DOE_GENERATOR provides a structured and systematic Design of Experiments (DOE) generator workflow. This workflow is designed to plan, create, and execute experiments efficiently to analyze how multiple input factors influence output responses. It is applicable across various fields including engineering, manufacturing, pharmaceuticals, software testing, and data science.

.. thumbnail:: /_images/Images/251workflow.png
   :title:

   DOE Generator overview and workflow concept

|
|

Pre-requisites 
===============

To use this workflow, a general overview of the Workflow applications and Workers are recommended in addition to basic introduction to Ansys LS-DYNA solver. Please contact support@d3view.com to get more information on these topics if you are unfamiliar with these topics.


Core Concepts
====================


The workflow provides the user with the following tasks based on the available test data.

1. Design of Experiments (DOE): A methodical approach to test the effects of multiple variables on outcomes.

2. Workflow: A step-by-step procedure involving input setup, experiment generation, optimization, and result extraction.

3. Inputs: Divided into two main categories:

4. Basic Tab: Contains solver input files and task selection.

5. Machine Learning Tab: Contains variables related to DOE experiments and output configurations


Workflow Task
=================

.. thumbnail:: /_images/Images/251workflowtasktable.png
   :title:

   Task Type table

|
|

.. thumbnail:: /_images/Images/251workflowtasktable1.png
   :title:

   Task Type table

|
|

Workflow Execution Process
=================================

1. Setup Task:
 
Define simulation parameters and optimization targets.

Execute the workflow to prepare DOE values.

Important: Do not reset the workflow after execution to preserve updated DOE experiments.

.. thumbnail:: /_images/Images/setuptasktypepath.png
   :title:

   SET UP TASK TYPE WORKFLOW PATH

|
|

2. DOE Task (Machine Learning Tab):

Input variables and DOE experiment details are configured.

Running this generates an optimum dataset reflecting the best parameter combinations.

.. thumbnail:: /_images/Images/doetasktypepath.png
   :title:

   DOE TASK TYPE WORKFLOW PATH

|
|

3. Optimize Task:

Change task type to ‘Optimize’ and resume execution.

This step refines parameter values to reach an optimal solution.

.. thumbnail:: /_images/Images/optimizetasktypepath.png
   :title:

   OPTIMIZE TASK TYPE WORKFLOW PATH

|
|

Re-extract Results Task:

Select ‘Re-extract Results’ task type.

Run the workflow to retrieve final simulation data.

.. thumbnail:: /_images/Images/reextracttasktypepath.png
   :title:

   Re extract TASK TYPE WORKFLOW PATH

|
|

Workflow Outputs
=====================

Based on the selected Task Type, the workflow execution path automatically generates and downloads the optimum parameter dataset.
The optimum dataset corresponds to the selected optimization strategy (Near Origin, Pareto, or Simulated Annealing) and contains the best-performing design variable combinations identified during DOE or Optimization execution.


.. thumbnail:: /_images/Images/outputpaths.png
   :title:

   Download Optimum Variables Dataset

|
|


Introduction to the START worker
==================================

The START worker is identified as shown below and contains inputs classified into Basic and Machine Learning tab.  In many cases, only the Basic inputs are needed to be updated but the user is free to update the values in the Machine Learning section.


.. thumbnail:: /_images/Images/251startworker.png
   :title:

   START Worker Inputs

|
|