It’s not hard; it’s a math problem!

Aug 10, 2021

Most design and process solutions are improved by subject matter experts’ experience and heuristics – the methods developed over years upon which their intuition is based.   

What if there were underlying algorithms and computations upon which the physical world was based – an inherent intelligent design? 

What if we could use mathematics and problem-solving approaches to make these improvements or come up with even better solutions? 

OmniQuest™ Iliad™ was built to enable just such capability.

We have shown how to use the DOE component to create a response surface using regression analysis. These relationships either create linkages that did not exist before or are a simpler alternative to pre-existing complex mathematic equations.

Here we will go a step further and use the response surface for mathematical optimization.

We will look at a special case of optimization, i.e., response surface-driven optimization, and how that through Iliad™ DOE.

Consider, the case of process settings optimization for injection molding using CAE simulation. The typical run time for a single analysis is a few hours. If the optimizer were to directly run the simulations, this study would have lasted days.

Instead, a quadratic response surface model was created using Iliad™ DOE and the optimizer converged in a matter of minutes.  This dramatic improvement was achieved by housing the DOE component inside an optimization component.

In the first iteration of the optimizer, the DOE component creates a response surface using a DOE design selected by the user. That implies as many simulation evaluations as the number of points in the DOE design.

For subsequent iterations of the optimizer, instead of running the simulations again, the DOE component calculates the responses using the approximate equations created in the first iteration. This option is enabled by checking the ‘Show Advanced Options’ checkbox.

The designing engineer can also reuse a response surface created in an archived Iliad™ database. Since most organizations already have records and data from previous experiments and CAE runs, the ability to apply Iliad™ to this earlier work ensures that nothing goes to waste – optimization builds upon organizational knowledge.

The importance of Iliad™ DOE cannot be overstated.  A simple tool that delivers powerful results. Contact us to see how the DOE component can improve your design and process solutions.