In order to achieve sustainable development and meet sometimes conflicting requirements,  technology and product optimization with high-level coupled simulation has a prominent role. Autodesk Moldflow, market leader for more than 40 years, has been providing proven injection molding simulations to tens of thousands of companies. Injection molding is a simulation solution and an excellent tool for product/mold/technology optimization tasks for injection molded single or multi-component polymer parts produced in a large series with short cycle time.

In the mechanical finite element analyses of injection molded technical plastic parts performed during development, it is challenging to take into account the changes in production, fiber orientation, and residual stresses. When designing products, ignoring manufacturing technology can be a major neglect. The various finite meshes used in conventional strength analysis and injection molding simulation as well as the different raw material databases make it more difficult to perform high-level coupled simulations. Autodesk has recognized that the simple further development of earlier functions is not enough for today's requirements, so it has begun to integrate the raw material data needed for strength analysis into Moldflow's raw material database.

Considering the nonlinear anisotropic properties of polymers, we were able to create complex simulations very close to reality.  However, during the simulation of the planned components, we may not get the expected results and additional iteration steps are required to optimize the product and/or technology. Within Moldflow, we can use the function Design of Experiments to test the effects of one or more variables according to our quality criteria. With the help of the experimental design, we explore the entire design space and maximize the amount of information available for analysis, allowing for an objective conclusion.    

From time to time, however, it may be necessary to modify external variables (e.g. product geometry) or to consider a quality criterion (mechanical load capacity) that we cannot directly control in Moldflow. By creating a multi-purpose optimization system, we can take into account conflicting requirements at the same time and find the best possible solutions. The solutions on the emerging Pareto front are equivalent, thus the decision is in our hands, which solution to choose from.

During the early analysis of an injection-molded product for Sony Visual Products Inc., it became clear that contradictory requirements had to be met. Due to the excellent light guiding capability, volume shrinkage had to be minimized, for which, according to experience, the size of the barrier and the distribution system had to be increased. However, the primary purpose of the new mold was to reduce the weight of the distribution system. Based on these criteria, manual optimization would have been a very lengthy process, and the built-in DOE function had its limits to perform the task independently due to the specificity of the finite element mesh. Therefore, the Japanese team, using the Mode-FRONTIER software, set up a multi-purpose optimization system for component optimization, allowing for the macro-level linking of different software and the objective evaluation of results. 

Simulation analyses for injection molding were carried out in Moldlfow and product geometry was modified within the PTC Creo application to provide the best possible geometry for the distribution system. As a result of successful optimization, the new mold has resulted in higher product quality and an annual saving of nearly 20% (8 tons) of raw material compared to the previous 8-nest mold, proving the success of the semi-automatic optimization process [1].  

With our current tools and our high-level knowledge, we are able to achieve substantial cost savings while ensuring higher quality and reliability, which is vital for companies competing in a fierce market. We are at the dawn of a radical change, where our natural human abilities will be extensively extended through the integrated use of our electronic devices that are able to learn. The next step in this change is the emergence of a new generation of our current passive command-based software, which will be able to find a solution to the problem at hand in a generative way based on our set of requirements and conditions.

​