Javascript Tree Menu by Deluxe-Tree.com Research A BIT ABOUT OUR WORK

The problem of modifying an existing material to suit a specific need has posed itself repeatedly through history. The Middle Eastern potters who tin-glazed their ceramics or figured out that tin and copper together lead to much better tools and weapons than copper alone, were essentially initiating a tradition that finds expression today in the doping of semiconductors or fabrication of composite materials. While we justly celebrate the successes of our space missions and marvel at the designs of large particle accelerators or the intricacies of integrated circuits, it is necessary to remember that most of these would not have been possible had not been backed by an unremitting quest for creating materials that can withstand severe situations and --- a fact that is, in my view, not very well appreciated --- understanding their behaviour so that we can utilise and modify them effectively. I have created a timeline of landmarks in the development of materials (If you have a comment on this figure, please send me a mail).

What constitutes an understanding of a material? Because we use it in such large volumes, the example of steel is apt here. Steelmaking is definitely one of the most well studied processes. It involves studying the metallurgy of the process and also the whole economics of procuring raw materials for and running of large plants that, in these modern times, churn out humungous quantities of a mind boggling variety of steels. Steelmaking processes are supported on a bulwark of a thorough knowledge of the effects of microstructure on the properties of steel and ways of controlling the microstructure during processing. Moreover, we know such a great deal about defects occurring in steels that we are often able to use them to our advantage. Properties of steel produced under a variety of controlled procedures have been carefully calibrated over large ranges of physical conditions and very accurate computer simulations of its behaviour are possible. In fact, with all the knowledge at our disposal, it is possible to vary the yield stress of commercially produced steels by more than an order of magniude and fracture toughness by more than four times. However, possible variations in other properties like modulus, density, electrical conductivity are not as dramatic.

The example of steel, though inspiring, is also slightly misleading. Behaviour of most commercially important metals are well understood but there is hardly any room for complacency. There are various areas in which our knowledge is evolving even today. A case in point is the behaviour of very small metallic structures --- they seem to behave quite differently from large chunks of metals and we are only beginning to understand why.

Modern applications often require a material to meet conflicting demands. Classical examples abound. Concrete is weak in tension but we cannot avoid situations involving tensile stresses completely in a building structure; polymers are light but brittle; pressure pipelines need to be thin walled to ease handling but should be able to provide enough warning before failing; multimaterials are a nice way of harnessing the useful properties of different materials but the interfaces may become the weak link in their design, etc. More subtle conflicts arise in specific situations as demonstrated by the following (random) examples. Damping materials are supposed to suppress vibrations over a wide range of frequencies but should also have sufficient stiffness throughout this range; silicone hydrogel contact lenses are extremely permeable but are somewhat hydrophobic; streched silicon is supposed to have better transport properties that make them attractive for electronics but are more susceptible to crystal defects, and so on (send me an email if you have more suggestions).

It may sometimes be possible to juggle a myriad set of conflicting demands by devising a clever structural arrangement. For example, a readily available high damping material may be combined with other structural elements to produce a 'structural damping' arrangement that works over a very wide range of frequencies and meets stiffness requirements. Alternately, a high damping but compliant material maybe engineered, possibly at the micro scale, to gain sufficient stiffness while retaining the damping capabilities. This has been shown to be possible by performing microstructural level tweaks, the most basic being a composite of a high damping and compliant material with a stiff material. A more daring approach will be to engineer a long molecule, half of which contributes primarily to the damping properties and the other half to the stiffness properties of the material. The possibility of engineering the architecture at molecular or atomistic levels to achieve macroscopic, tangible technological goals is one of the most exciting developments in modern materials science.

We, in the Computational Mechanics Group at IIT Kanpur, exist with the belief that modern numerical methods can go a long way in enriching our knowledge of how microscopic mechanisms collude to produce macroscopic effects in materials, particularly in relation to their deformation and failure behaviours under various external conditions. We try and work with as real materials as is possible to represent computationally rather than with 'computonium', the ideal material of choice (and many a times wisely so) for many other workers in this field. Also, we do not envisage that models to seamlessly connect the micro to the macro scales are possible or even particularly desirable. We prefer to work at a particular level of detail, try to gain insights that enrich our understanding of the next coarser level and help us build models at increasingly larger length scales. After all, engineering design needs continuum level models and 'gross' things like stresses, strains etc to be computed. So, however detailed your microscopic model and understanding might be, it will be technologically useful only if it either helps in building models at the macro time and length scales or you happen to be interested in doing engineering on the material at the microscale itself. This approach of sequentially going up the length and time scale ladder makes us a bit like the blind men groping the elephant, but if we had allowed enough groping, who knows, the blind men might have figured out the truth!

There are numerous aspects of microscale knowledge that assumes importance when you want to understand materials. The deformation of a material under the action of externally applied forces is the result of many microstructural adjustments. Deformations often tend to get intense in a small region of the material. Fracture of materials most often occurs catastrophically at 'human' time and length scales but the material may actually be accruing microstructural damage due to intense deformations accumulating over a long time. The nature of damage accumulation or microstructural adjustments also depend on the underlying cause of the deformation, most importantly the temperatures and rate at which these events happen. It is sometimes important to know the micro-level story in order to keep the possibility of doing microscopic tweaks alive.

An essential feature of multiphase materials is the presence of interfaces, which add a new twist to the story. Interfaces may be required to be strong or weak depending on the end use of the material but understanding the chemical nature of the bonding at the interface is the key to imparting it with desirable characteristics. With the advent of materials with nanometer sized phases, engineering the interface has become the key element in designing these materials.

At the moment we are working on a variety of problems that have broadly been described in the last two paragraphs. Some of these problems require heavy computations and therefore, some work needs to be done on perfecting and honing computational methods. Moreover, experimental support for some problems is essential in order to address real materials. To do these we try --- and have had a reasoanble degree of success so far --- to sweet talk, cajole and/or threaten unsuspecting experimentalists into performing experimental simulations of our computational experiments.