Virtual benefits to materials science
Virtual reality and materials science may not seem like a natural pairing, but it might offer more than expected. UC Merced’s Dr Lilian Davila speaks to Khai Trung Le about its potential in teaching and research.
In the 1992 film, The Lawnmower Man, total immersion in virtual reality was achieved by being restrained to an aerotrim wearing skin-tight silver Lycra and a bulky headset. Fortunately in the last few years, virtual reality has managed to shed its iconic, cumbersome imagery with commercial releases from Sony, Facebook and HTC and Valve Corporation, centred on uses in remote communication, architecture and gaming. Dr Lilian Davila from the University of California, Merced, USA, Computational Materials Science Group, has pursued the use of virtual reality in materials science education and research for more than a decade. While the response has been upbeat at the University of California, it has failed to make headway beyond. What does the recent advance of VR mean for her efforts?
Initially a computer scientist, Davila said she first saw the potential of virtual reality following a doctoral project using molecular dynamics simulations for the study of structural transformations and properties of silica glass at high pressure. Recalling a computational group at UC Davis testing virutal reality systems for materials scientists, Davila collaborated with Dr Oliver Kreylos in 2004 on his open-source software, Nanotech Construction Kit (NCK), which enabled users to build models akin to Lego blocks or read data from different databases and edit the virtual system. ‘Not only could you modify the virtual structure in VR but there was also real-world physical behaviour based on inter-atomic forces used in the NCK modelling software - for example, mimicking realistic interactions between atoms, so you could describe the attraction and repulsion forces in a material system. You could “feel” these forces,’ Davila said.
However, wider adoption was hampered by the VR environment being solely available at the UC Davis facilities, with replicating the cheapest environment costing around US$150,000. But Davila saw the potential in the system, noting, ‘This was in 2004. Technology changes rapidly, and there was a good chance to reduce the cost of these VR environments.’
Davila and Kreylos reconnected in 2009, with Davila installing a new Kreylos developed system that was deemed more affordable (around US$8,000). After its use in a summer workshop focused on exploring carbon nanotube structures in VR, Davila was granted funding for the IDEAS project from the National Science Foundation, focusing on molecular dynamics simulation studies. Eschewing bulky headsets, the IDEAS platform uses 3D goggles and cameras to track the movement of a modified Nintendo Wii remote for atomistic models manipulation, and was made compatible with the LAMMPS open-source molecular dynamics simulator as part of the project’s bid to make VR systems more palatable for research purposes.
VR is not intended to replace conventional computational techniques, but to provide added value. Davila said, ‘If we only use conventional computing, we analyse results in a linear way. You work on your code, test and develop for efficiency, then run large batches of jobs and analyse data accordingly. When you have a VR system, you have added value to the research – yes, you can visualise models with existing packages, but for example if you are investigating the elastic behaviour of nanohelices, a VR system readily tells you when the helical structure is beginning to deform or break on turning or twisting, or benefit from an enhanced depth perception and visualise how defects affect overall properties. The main strength of molecular dynamics is that you are looking at the evolution of atomistic details. In the virtual world, you can “see” and interact with all features in the materials simultaneously.’
Learning to play
The greatest success of Davila’s system has been employing physical model-building procedures in education. Davila believes this allows her students to excel in spatial learning, by the act of building and drawing in the 3D space. She noted that ‘it can be difficult to grasp related concepts while fostering spatial learning, particularly for the younger students. You’re talking about directions and planes, impurities moving along vectors and planes in materials – computational scientists use vastly abstract ideas and codes. Virtual environments that mimic hands-on allow researchers to readily build models and analyse simulation results to advance knowledge and promote innovation.’
Virtual reality also has the potential to change the learning environment, including emulating multiple laboratory environments in VR and remote collaborations, but Davila expects this will only come about when familiarity with virtual reality systems is more widespread. ‘Since the cost of VR environments continue to drop significantly, this continues to be a good opportunity for collaborative projects. But in many cases, the building of a VR environment is not possible or practical. The main hurdle remains the integration of all components – not a trivial problem, but technology continues to provide new and different paths.’
Davila has not explored more recent consumer devices including Oculus Rift and HTC Vive, but is open to expanding compatibility with the IDEAS system with appropriate funding.
There remain obstacles to the proliferation of VR in materials science research. Whereas the bottleneck may previously have been from a technological perspective, Davila believes engineering faculties are more open to pursuing advanced technologies. One enduring obstacle remains in convincing scientists of its value. ‘Professors are not as open to collaborate with colleagues outside their field. But we’ve proven VR is a possibility, and we just have to be diligent about it. Using VR environments in materials science has a unique value, and research and funding is becoming more competitive. Given the next generation of students are visually oriented, it is certainly another motivation for educators and researchers to make use of VR systems as complimentary tools.’