Reinventing Additive Manufacturing with Simulation

Over the past century we have witnessed amazing milestones in manufacturing.  The most impactful achievements share one important commonality…they brought greater simplicity and automation to the design and production process. Consider the roll-out of Henry Ford’s moving assembly line in 1914, the launch of the first CAD software in 1954, and the arrival of industrial robots in 1973. Each of these revolutionary developments allowed manufacturers to hit the hypothetical fast-forward button on production, bringing better products to the marketplace more quickly while advancing society technologically.
Today we are witnessing the next mammoth milestone—the widespread adoption of additive-manufacturing technology, popularly known as 3D printing, in production. Invented for rapid-prototyping purposes in the 1980’s, 3D printing had many niche applications that were mostly non-structural, but the technology’s true potential remained untapped for decades.

This adoption of additive manufacturing is being aided, in no small measure, by the rapid advances in simulation technology for multiphysics optimization and predictive analytics. With design no longer constrained by subtractive manufacturing restrictions, a part designer can answer relevant questions: What is the functional objective of the part? Can we design a part with the same functional characteristics but use less material? Can we obtain the cost-savings from optimized additive parts? Engineers and designers are empowered to develop parts that are increasingly complex, more organic and lighter—all while meeting their performance requirements while using less time and resources! And process simulation solutions allow us to successfully print these unique designs by providing us with detailed analytics that help us predict potential failures and optimize the printing process parameters so that each part can be printed right the first time.

Most engineering parts that we traditionally think of are built from solid stock units of material of various regular shapes: blocks, beams, bars, and sheets. This means that they are internally continuous and usually homogenous in nature. A unique capability of additive manufacturing is the ability to manufacture parts with complex and heterogeneously variable internal sub-structures and properties at minimal additional production cost. These include repeating in-fill patterns and customized lattice designs. But designing those complex internal structures is in no way trivial, and has been a focal point in the additive manufacturing community for the past several years. It’s also particularly challenging for simulation. How do you capture intricate details at a part level and still have a computationally efficient model? Driven by representative volume elements (or RVEs) we can transform these detailed internal structures into continuum representations such that we can model them realistically in part-level simulations.

But, why stop there for innovation? Since the underlying variable that drives a topology optimization is the relative density of the material, we can now link it to RVEs and determine variable densities and material distribution in a fixed design space, a shoe insole for example. Users can apply single or multiple loadings, such as a pressure loading profile from a realistic human foot model, to represent static loading of an averaged human’s weight. Tosca will iterate with this load case to find the maximally stiff structure within a given mass constraint. The optimization yields a continuous field of relative density throughout the component. Amazingly, the optimization results often reveal that the maximum density region does not always correspond to maximum load location. This is the kind of insight that can only be achieved using simulation technology.

Let Simulation be your Guide to Additive Manufacturing Processes 

The additive-manufacturing machine provider market-space is ever expanding, from global conglomerates to startups, from hardware that handles metals to polymers. The process families can be widely divided into metal powder bed processes, polymer extrusion processes, binder-jetting methods and even advanced welding-like processes. Most AM processes require unique and detailed process analytics, such as temperature and distortion profiles, in order to get a better handle on process robustness and reliability. As this additive manufacturing ecosystem grows ever more complex, the engineering and research communities are looking at simulation as a necessary tool to mature these methods to a state of full-scale production readiness.
To address the community’s needs, we’ve developed an all- purpose simulation framework that gives them the flexibility to simulate parts built from different processes. The framework allows users to specify machine-dependent information (such as tool path, build environment, power input) as inputs in space and time, include support structures from their builds, analyze material behavior—while it computes the solution locally (micro-level) and globally (part-level).