Simulations are computer-generated models that replicate real-world processes or systems to analyze, predict, or study their behavior and outcomes. They serve as valuable tools to gain insights in cases where experiments would be infeasible or simply impossible. For instance, predicting the motions and interactions of ten thousand abrasive filaments on a brushing tool.
However, every simulation needs to be validated to some degree, meaning it should be compared with experimental results. Because the processes we are attempting to model are almost infinitely complex, we must always make simplifications. Validation helps us determine which simplifications are appropriate and which should not be made in order to understand the process characteristics.
Even before validation, experimental values are needed as input for the simulation, each based on its own analytical model. For example, the abrasive filament simulation requires parameters such as elasticity, coefficient of friction, precisely measured filament diameters, damping ratio, and many more. Unfortunately, these experiments introduce additional errors into the equation. In the end, we are more surprised if reality is accurately predicted than if it is not.
With computers constantly improving in performance and becoming more affordable, numerical models can be refined with larger numbers of interacting particles or nodes. I am excited to witness this trend and see how much we can already predict using only numerical methods. This becomes crucial whenever we are unable to find analytical solutions because the mathematics we learned in school fails us.
Below, I have shared some of the simulations I developed for my studies or simply out of curiosity. If you are looking for a dating simulator, please refer to “Anima X.”