Metal additive manufacturing (AM) processes are very complex and the process parameters required to fabricate quality parts can be very complicated and challenging to determine. For this reason, there is a continuous demand for AM simulations which can assist users to determine optimal process parameters in a timely manner. However, current commercial simulation packages are expensive and not designed with AM processes in mind, which makes it challenging to simulate an AM process. In addition, they are computationally intensive and usually require some high performance computing (HPC) system to run effectively. Besides, they take relatively more time, usually from several days to months, to simulate an AM process. This paper presents additive manufacturing simulator (AMS), a simulation tool built specifically with AM processes in mind, to address the challenges with current AM simulation tools. AMS is based on a three-tier client-server architecture, coupled with service oriented architecture's (SOA) publication and subscription model. It has the ability to scale to utilize available computational resources and automatically balances computational resources among multiple processors. Compared with current AM simulation systems, AMS improves performance of AM simulation considerably and can run on not only HPC systems, but also low-cost desktop computers.

Due to the high cost of each experimental run in additive manufacturing (AM), there has been a drive to develop simulations that can find the optimal processing parameters. The accuracy of these simulations is dependent on the accuracy of the material parameters recorded in literature. These reported parameters can vary widely resulting in differing simulation results. For that reason, it is necessary to determine which parameters are the driving parameters. This will allow for only critical parameters to be experimentally found resulting in more accurate simulations faster. This article uses a Plackett–Burman design of experiment to screen for the material properties with the greatest effect on the results of a thermal mathematical model of a laser-based directed energy deposition (DED) AM process. It was found that variances in the absorption of the laser at 880 and 922 K along with variances in the thermal conductivity at 922 K have the largest effect on the range of the response variables that were used to characterize the melt pool. Having a smaller impact on the results are the thermal conductivity at 1491 K and the specific heat at 733 K, and the remainder of the factors have a negligible effect on the melt pool characteristics within the simulation.

Additive manufacturing with metal components is a complex, and currently cyclic, process due to the physical phenomena that are occurring. These phenomena can be mathematically modeled in order to predict the outcome of a specific aspect of the build. Coupling the mathematical models can then be used to develop a complete simulation, which can produce estimates for a range of characteristics for a part built using additive manufacturing techniques. This chapter will investigate the main models used in the simulation of metal AM. These models will include the modeling of thermal behavior, fluid dynamics, stress, and a selection of other auxiliary models, which are necessary to complete the simulations. For each of the models investigated, the various modeling techniques that have been developed will be presented along with their limitations, validation techniques, and parameters necessary to model the process correctly.

Additive manufacturing (AM) simulations are effective for materials that are well characterized and published; however, for newer or proprietary materials, they cannot provide accurate results due to the lack of knowledge of the material properties. This work demonstrates the process of the application of mathematical search algorithms to develop an optimized material dataset which results in accurate simulations for the laser directed energy deposition (DED) process. This was performed by first using a well-characterized material, Ti-64, to show the error in the predicted melt pool was accurate, and the error was found to be less than two resolution steps. Then, for 7000-series aluminum using a generic material property dataset from sister alloys, the error was found to be over 600%. The Nelder–Mead search algorithm was then applied to the problem and was able to develop an optimized dataset that had a combined width and depth error of just 9.1%, demonstrating that it is possible to develop an optimized material property dataset that facilitates more accurate simulation of an under-characterized material.