Twenty plus years I ago I was a fresh-faced PhD student, and my then PhD supervisor, George Jackson was selling the science his group, including me, did. I remember that he was a fan of triangles like the one shown above. His group was doing both computer simulations and theory at the time, and he wanted to make the point that they complement each other, and help to understand and predict experiment.
On Monday, I have to sell the science we are doing, in this case to a couple of scientists from a paints company. So I am taking inspiration from my PhD supervisor, and hoping that this idea has not aged in the intervening twenty years.
Science is ultimately based on experiment, we measure the natural world and that tells us what is true and what is not true. But we need theories to model the natural world, so that we can understand it and make predictions. For example, we need to measure Jupiter’s position in the sky, but just measurements are not enough to predict where Jupiter will be on say 1st April 2018, for that we need a theory. Newton came up with one, and this theory combined with observations allows to predict where Jupiter will be next April.
But planetary orbits are unusually simple. Jupiter’s orbit is largely controlled by the Sun, so to a good approximation we just need to worry about two objects. Drying coatings have huge numbers of interacting paint particles in them, plus the water etc. They are much more complex. So our theories of coatings are pretty rough.
This is where computer simulation comes in. It can test candidate theories, and then if a theory passes the test, we can use it to predict new experiments with more confidence. We can also use simulation to try new ideas. So, as the diagram above tries to show, the three complement each other, with each having strengths and weaknesses. We will see how well this old idea goes down on Monday.
Chance & Necessity 