To end this kind of a paper, which is full of ideas and history but not hard technical results, a “Prospectus” section seems more appropriate than a “Conclusions” section. So, what are we looking forward to? Here are some possibilities.
The advent of the personal computer back in the late 1970s and early 1980s has certainly had a profound effect on how we work and the kinds of research that can be done. The processing power of scientific mainframes has increased rapidly in tandem with the increase in power of the personal computer. Now scientific mainframes, as they were originally, have been replaced by large parallel machines that use large numbers of what are essentially modern personal computers. The 300+ processor machine at NIST on which we do our most intense computing is made up of individual processors that could serve very well as my desktop machine. Parallel computing is starting to move from the computer laboratory to the office desk – one can buy desktop machines now with two processors, and more processors are coming. The power of parallel processing will make it more and more possible for computer models to be accessed and used by more and more people who are farther and farther from the research laboratory.
In the light of the current research climate, we must squarely face the fact that, whether we like it or not, concrete is a nano-scale material. What I mean by this is not some ludicrous picture of tiny, 10 nm devices made of cement paste, but rather that the main “glue” that holds cement paste and therefore concrete together is calcium-silicate-hydrate (C-S-H), whose important structure and properties are determined at the nanometer scale. For example, if you want to understand and thereby intelligently modify the viscoelastic properties of concrete, you have to understand how C-S-H deforms over time under a load, because that is the only viscoelastic material phase to be found in concrete. There is certainly controversy about the role of intelligent design in the origins of life, but surely there is no argument that intelligent design is necessary for good concrete, and that means going to the nano-scale. People in industry and academia should be trying hard to see that some of the vast societal expenditure on nanoscience research is targeted towards concrete, an important material whose industry in the US alone is worth $110 billion per year [23].
I believe that we will continue to see a trend for the standard testing done in the cement and concrete industry to move from an empirical basis to a materials science basis, which will allow better understanding and control of manufactured materials, especially with the predictive power of computational materials science models that are enhanced by better characterization and measurement and knowledge of mechanisms. I really believe that we will see the Kepler-Newton scenario played out in the concrete industry, especially in the area of service life prediction, for which we desperately need good materials science, experimental and computational.
Let me end on a high note. I think that the research problems in concrete are fascinating and are as challenging and as “cutting edge” as any to be found in the realm of materials and beyond. Recently, there has started to be agreement with this point of view outside the field of concrete. In a national competition, a team of NIST researchers won a grant from NASA for 1,000,000 hours of computer time on one of the most powerful computers in the US, NASA’s 10,000+ processor supercomputer called Columbia [24]. The grant was to work on the modeling of suspension rheology; modeling that is aimed at fresh concrete. It was ironic that news of this award came on Friday, March 3, and Geoff Frohnsdorff passed away on Sunday, March 5. We did not get the news of our success to him in time. However, he knew about the proposal and had even suggested several items regarding the strategic and economic importance of concrete to our nation. Geoff would have been delighted, but not really surprised, that concrete research could successfully compete for and be considered worthy of this large amount of computer time on NASA’s top supercomputer. Without Geoff’s pioneering research and tireless vision, there would probably not be any computational materials science of concrete today, and so this article would never have been written and there would not be any possibility of a Kepler-Newton-type collaboration for cement-based materials.
Next: References Previous: Computational Materials Science of Concrete