Dear E.A.Dinesh Kumara ,

In general, there are two methods for performing design optimization which are the followings:

The first method is using scripting or in another word coupling a numerical software (Ansys) with a mathematical model (Matlab or even Excel) for performing design optimization. In this method, you manipulate your inputs parameters in the numerical software and calculate the result. Then export the result to the mathematical model and predict what are the new values for the input parameters for the next iteration. You can develop linear or nonlinear model for predicting the new values (according to your needs). This is something like inverse design where Bendine Kouider also mentioned in his answer. this method is a very powerful method for design optimization where you can get very accurate results.

Second method is Design of Experiment (DOE) where you form a design table based on your input parameters, then run the numerical software (Ansys) for calculating the results. After calculating the results (outputs), you can calculate a response surface (or just an equation) for correlating the input parameters to output parameters. This way you can optimize your design. Depends on how many iterations you are performing and what method of DOE is used you can estimate your optimized value for input parameters.

Hope this answer help you.

Dear E.A. Dinesh Kumara,

I suggest you to see links and attached file on topic.

https://www.ozeninc.com/parameterize-everything-ansys/

https://www.padtinc.com/blog/how-to-use-lattice-optimization-in-ansys-mechanical-and-ansys-spaceclaim-19-2/

https://sites.ualberta.ca/~wmoussa/AnsysTutorial/AT/optimization/Optimization.html

https://books.google.dz/books/about/2014_International_Conference_on_Mechani.html?id=n-0RAwAAQBAJ&redir_esc=y

Best regards

Dear Mohamed-Mourad Lafifi,

Thank you very much for the answer.I'll refer the document that you have sent me.

I usually couple Ansys with Matlab to perform the optimization so I highly recommend that to you

Dear Bendine Kouider good idea

Dear E.A.Dinesh Kumara ,

In general, there are two methods for performing design optimization which are the followings:

The first method is using scripting or in another word coupling a numerical software (Ansys) with a mathematical model (Matlab or even Excel) for performing design optimization. In this method, you manipulate your inputs parameters in the numerical software and calculate the result. Then export the result to the mathematical model and predict what are the new values for the input parameters for the next iteration. You can develop linear or nonlinear model for predicting the new values (according to your needs). This is something like inverse design where Bendine Kouider also mentioned in his answer. this method is a very powerful method for design optimization where you can get very accurate results.

Second method is Design of Experiment (DOE) where you form a design table based on your input parameters, then run the numerical software (Ansys) for calculating the results. After calculating the results (outputs), you can calculate a response surface (or just an equation) for correlating the input parameters to output parameters. This way you can optimize your design. Depends on how many iterations you are performing and what method of DOE is used you can estimate your optimized value for input parameters.

Hope this answer help you.

Dear Hamidreza Ahmadi Moghaddam

I used the second method in my study

But you will need a lot of simulations, for a large number of variables (from 8 variables and more), especially in size optimization problem.

In the latest versions of the ANSYS program, it is available directly using the design of experiments (Workbench or APDL). I found it difficult in this technique.