Modern Machine Shop and MoldMaking Technology present ADDITIVE MANUFACTURING, a quarterly supplement reporting on the use of additive processes to manufacture functional parts. More at additivemanufacturinginsight.com.
Issue link: https://am.epubxp.com/i/497467
20 — AM Supplement Challenge Traditional approaches to process and performance modeling still requires signifcant experimental efort to quantify the efects of small, accurate, controlled and fast-moving energy sources across a wide range of geometrical variations. Since these correlations cannot be captured using traditional approaches, improved simulation is required to gain the understanding required to design better and optimize existing processes to enable the "next industrial revolution" sooner rather than later. ANSYS is arguably one of the best existing commer- cial tools for solving metals-based melting thermal problems. Using ANSYS, however, to simulate even a relatively small powder-bed-fusion (e.g., selective laser ical solutions shows promise for signifcantly reducing computational time. Future Advancements Tere are more unique algorithms being developed by 3DSIM personnel, including intelligent Cholesky, eigensolvers, degrees of freedom (DOF) reduction, and periodic and higher-order boundary condition solvers, that are being implemented into the 3DSIM C++ code, in part with DARPA funding. As these further enhance- ments are implemented, 3DSIM will be positioned to solve much larger problems even more quickly. In fact, 3DSIM claims that once its algorithms are fully imple- mented, it will be able to predict how a laser interacts with energy faster than a laser actually scans a part. As Table 1 - Time Comparison between 3DSIM and ANSYS Time Steps Ofsets sec/tme step sec/ofset Total (hours) 3DSIM 200244 11779 0.2 0.5 12.8 ANSYS 200584 N/A 67.54 0 3763.2 melting, SLM) thermal problem requires many years of computational time. Many traditional algorithms require each element and node to be accounted for during a moving mesh phenomenon, increasing the processing burden of the solution as a function of fne mesh movements. Furthermore, without sufcient time steps and fdelity, there is a discontinuous result that yields a less-than-realistic representation of the true as-built properties (Figure 1). Potential Solution Path Sets of algorithms and software tools like those being developed by 3DSIM show an initial advantage for speed, and thus capability, for simulating AM problems. In particular, the traditional requirement of accounting for each element and node at a fne scale is unnecessary with more intelligent algorithms such as those developed by 3DSIM. Given meshing and other diferences in algorithms between ANSYS and 3DSIM, results are difcult to compare, but Table 1 illustrates how a solver built for AM simulations using highly optimized algorithms and novel numer- such, future work to fully document the computational speed of 3DSIM's core solver and its prediction accuracy in light of these additional algorithms will be necessary. References 1 Referenced journal articles: "A Comparison of the Computa- tional Speed of 3DSIM versus ANSYS Finite Element Analyses for Simulation of Termal History in Metal Laser Sintering" (Zeng, K., et. Al.) and "A New and Efcient Multi-scale Simulation Architecture for Prediction of Performance Metrics for Parts Fabricated Using Additive Manufacturing" (Pal, D., Stucker, B.) 2 Examples given: temperature and phase dependent coef- cient of thermal expansion, elastic modulus, Poisson's ratio, yield strength and strain rate sensitivity, as well as the solidus temperature, solid state transition temperature(s) and phase(s) if present, latent heat of fusion, latent heat of vaporization, liquidus temperature, and vaporization temperature (including those of any low-elemental weight constituents) 3 Initial work with 2.5-dimensional symmetric meshes in the X and Y directions but extruded in the Z direction without any length-scale change are being supplanted by three-dimensional meshes such that the number of degrees of freedom can be reduced signifcantly and the quality of the solution can be made better. Other advantages of having a full three-dimensional mesh is to ease the use of an eigensolver.