本讲摘要

The Lennard-Jones (LJ) fluid, named after mathematician-physicist-chemist Sir John Lennard-Jones (1894-1954), occupies a special place among fluids. It is an ideal entity, defined as the fluid whose particles interact according to the Lennard-Jones potential. This paper expounds the history of the LJ fluid to throw light on the tensions between theory and computational practice. In my presentation, I argue for the following claims. Firstly, the computational approach—even prior to the computer—pragmatically aims at prediction, rather than truth. Secondly, computer simulation methods, especially “molecular dynamics”, triggered a change in epistemology. Now, simulated model fluids became targets of investigation in their own right. The urge for prediction turned the LJ fluid into the most investigated fluid in engineering thermodynamics. Thirdly, in a surprising twist, recent work reports problems with reproducibility of LJ simulations. We argue that there is a yet under-examined layer of modeling steps leading from the ideal object to a simulation of its properties. Hence the LJ fluid, the allegedly most ideal of all fluids, a child of mathematical prediction, turns out to have a bricolage character in practice. All three claims are intended as an invitation to further discussion.

主讲人：Prof. Johannes Lenhard （凯泽斯劳滕理工大学（Technische Universität Kaiserslautern））

主持人：马大年 （北京大学科学技术与医学史系 助理教授）

主讲人介绍

Johannes Lenhard, born in 1964, studied philosophy and mathematics in Heidelberg and Frankfurt am Main, Germany. His doctorate in mathematics (1998) was complemented by a habilitation in philosophy at the University of Bielefeld in 2012. Since 2000, he works at the Technical University Kaiserslautern (from 2023 called RPTU) where he holds a Heisenberg (research) Professorship “Philosophy in science and engineering”. One focus of his research is on analyzing how the computer is used as a scientific instrument and how this use transforms scientific epistemology and rationality. Books worth highlighting: “Calculated Surprises. A Philosophy of Computer Simulation” (Oxford UP, 2019), and “Mathematics as a Tool. Tracing New Roles of Mathematics in the Sciences” (co-edited with Martin Carrier, Boston Studies in the Philosophy and History of the Sciences 327, 2017). His most recent book “Cultures of Prediction” (MIT Press, 2024, co-authored with Ann Johnson) uncovers the dynamic history of prediction in science and engineering over four centuries.