High performance computational modeling for industrial applications

Professor Furlani (far left) and his graduate students implement models in UB Center for Computational Research to promote industrial innovation.

Research in Professor Ed Furlani’s laboratory involves multiscale modeling for the development of innovative materials and devices with features that are engineered from the nano to macroscale.

Ed Furlani, Professor

Furlani's students pursue fundamental research, but are also eager to obtain industrial modeling and design experience.

To this end, they develop computational models to establish proof-of-principle of device or process concepts in advance of fabrication, which accelerates product commercialization by reducing development time and cost.

In many instances, models are created using commercial computational software such as the COMSOL Multiphysics program, which enables customization of geometries, material properties, equation sets, etc. The students implement many of these models on high-performance computing clusters housed in UB’s Center for Computational Research (CCR), a state-of-the-art supercomputing facility. This enables highthroughput parametric analysis that is needed to determine optimum device or process performance.

The group has leveraged this capability for projects sponsored by companies such as Xerox, S. Howes, and Vader Systems, a local startup company. In the case of Vader Systems, students have developed models to simulate a unique additive manufacturing technology that enables 3D printing of custom aluminum parts on demand. This technology is based on a proprietary process called Liquid Metal Jet Printing (LMJP) that involves the liquefaction of a solid aluminum feed wire and the subsequent ejection of molten droplets using a pulsed electromagnetic field. LMJP, which mimics inkjet printing, is based on the principles of magnetohydrodynamics, i.e. the manipulation of conductive fluids using a magnetic field. Complex 3D structures are printed by depositing layers of molten aluminum droplets that coalesce and solidify into a predefined shape. Furlani’s group has developed models that simulate the electromagnetic and fluid dynamic (droplet ejection, coalescence, solidification) processes of LMJP as shown in the figure. Such modeling provides invaluable design experience for students. It also has direct economic impact by enabling companies to leverage faculty expertise as needed to develop new technologies and accelerate product development at reduced cost using student support.