Prof. William Curtin

Full Professor

After gaining a BS and an MS in Physics at Brown University, then a PhD in theoretical physics at Cornell University, he left the academic world for industry, working in the Applied Physics Group of BP (British Petroleum). There he addressed hydrogen storage in amorphous metal alloys and the mechanics of fiber-reinforced composites, to guide development of materials with enhanced performance. After seven years at BP, he came back to the academic world, but with less focus on physics. He settled down at Virginia Tech and for five years held a position as professor attached to two engineering departments: materials science and engineering mechanics.

He joined the solid mechanics group at Brown in 1998. 'Brown had an international reputation in solid mechanics. It was the best place for the type of research I wanted to conduct', explains the professor. At that point, he had the necessary skills to study the behaviour of materials at all levels. 'I had studied phenomena at the atomic and quantum levels during my PhD and at BP, I had modeled composites on the continuum scale. When I arrived at Brown, I was able to leverage these competences and work on multi-scale modeling.'

Prof. Curtin joined the EPFL as Director of the Institute of Mechanical Engineering in 2012.


Research Area

Current research areas of LAMMM are as follows:


 1.  Coupled Atomistic/Dislocation Dynamics in 3d (CADD-3d).  We are developing algorithms and codes to enable the intimate coupling of the open source molecular dynamics code LAMMPS and the open source dislocation dynamics code PARADIS such that dislocations can be fully resolved down to the atomic scale in some key region(s) of the material but be described by less-computationally-demanding continuum line models in the surrounding regions. 

2.  Coupled Quantum/Continuum methods (CADD-QM).  We are developing approaches for highly accurate coupling of density-functional-theory quantum methods to molecular dynamics or continuum representations of a material.  A local region where key chemical reactions are occuring is surrounded by a less-computationally demanding model (atoms or continuum) that faithfully transmits far field mechanical loads to the quantum region while providing a proper environment for accurate solutions of the electron density and ion positions in non-periodic quantum DFT domains.

3.  Coupled Atomistics and Discrete Dislocation Dynamics (CADD).  We continue to enhance the CADD model by implementing the full richness of discrete dislocation dynamics in 2d, with sources, obstacles, pile-ups, and representations of 3d forest hardening, while faithfully capturing local finite-temperature atomistic behavior  such as crack growth and/or crack tip dislocation emission.  We are also developing sub-cycling approaches and other methods to overcome the time-scale limitations of full molecular dynamics, which runs at the femptosecond time scale.


1.  Solute Strengthening in Metal Alloys.  We are using quantum mechanical predictions of solute/dislocation interaction energies as input to a new model for the finite-temperature, finite-strain rate flow stress of metal alloys.  The overall method has successfully predicted the flow stresses in several lightweight alloys Al-X (X=Cu, Mg, Mn, Cr, Si) and Mg-Al (basal slip), and we plan to extend the study to Cu alloys.

2.  Dynamic Strain Aging and Ductility in Al alloys.  We previously discovered the "cross-core diffusion" mechanism for solute-induced negative strain rate sensitivity in Al-Mg alloys.  Recently, we have developed continuum constitutive models that capture the physical consequences of this mechanism at dislocation scales, and we are now applying the full model to predict ductility in Al alloys.  Al-Mg alloys suffer from low ductility, generally attributable to dynamic strain aging, and our new studies can predict this ductility loss quantitatively as a function of temperature or strain rate.  Furthermore, combined with our models above for solute strengthening, we are now designing new alloys with enhanced ductility, which would enable more-complex forming processes in Al alloys and thus permit more widespread use of lightweight Al alloys in, for instance, automotive applications.  

3.  Hydrogen Embrittlement in Metals.  We are using molecular dynamics methods to probe possible mechanisms of hydrogen embrittlement in Fe and Ni.  We have recently found a new mechanism where in H aggregation around a crack tip, driven by the high tensile pressures near the crack tip, prevents crack tip dislocation emission, which is a key mechanism for crack tip blunting, leading to significant material toughening.  Predictions of the model as applied to Fe-H are consistent with a wide range of experiments showing embrittlement or lack of embrittlement in Fe-H depending on the H concentration, temperature, and mechanical loading rate.

4.  Size-dependent Plasticity.  We have developed a new concept, "Stress Gradient Plasticity", for understanding one origin of size-dependent plasticity in metals.  The Stress Gradient Plasticity theory derives directly from an analysis of the behavior of dislocation pile-ups in a stress gradient, and shows that the material length scale relevant for size effects is the spacing between dislocation obstacles.  A low-order continuum model that incorporates the stress-gradient-plasticity mechanism has been created, and has been used to predict size-dependent plasticity phenomena in bending of micron size beams, torsion of micron size wires, and nanoindentation.  We continue to pursue the application and implications of this model.

5.  Stress-driven Catalysis.  We are using computational models to understand the magnitude of changes in catalytic activity on a metal surface as a function of the applied strain.  This work supports an experimental effort that is using various nanostructured metal systems to impose large controlled and reversible elastic strains in metals within an electrochemical environment to accelerate reaction rates and/or control reaction paths. 

6.  Deformation of Mg and its alloys.  We are using atomistic and quantum methods to study the dislocation structures in hcp Mg, with the goal of understanding the role of solute additions in controlling relative deformation modes to enable high strength and high ductility.  We are currently computing the core structures of twinning dislocations and c a dislocations, and examining the interaction energies of various solutes with twin boundaries and dislocation cores. 

7.  Toughening in Carbon-nanotube Ceramic Composites.  We are using atomistic and continuum models to predict how the interfacial structure/bonding between a carbon nanotube (CNT) and a matrix influences the mechanical properties (strength and toughness) of these materials.  We are also investigating the role of CNT morphology - e.g. multiwall structure, waviness and length of CNTs - on the macroscopic behavior, with the goal of guiding tailoring of CNT sytems for high performance.  We are also investigating the formation of mechanically-interlocked CNT ropes achieved via irradiation to induce inter-CNT bonding in CNT bundles.   


William CURTIN
MED 3 1026 (Bâtiment MED)
Station 9
CH-1015 Lausanne