History-dependent friction

About the project

Friction is a topic of huge practical, technological and scientific interest that has challenged mankind for thousands of years. However, it still remains poorly understood, probably due to the inherent multi-scale and multi-physics nature of processes at the frictional interface. The empirical laws of friction where introduced by Amontons and Coulomb, and later refined into the rate-and-state friction law, which is commonly used today. The rate-and-state friction law states that the coefficient of friction that depends on the rate - on how fast the surfaces are moving relative to each other - and the state - how long the surfaces have been in contact and under what conditions. However, we have recently made a startling discovery: The coefficient of friction may also strongly depend on the history of the frictional contact, on how the two surfaces stopped relative to each other, changing the research focus from detachment to reattachment. In this project we will address how to reformulate the laws of friction to include the history of the contact - a history-dependent law of friction. We will determine under what circumstances history-dependent friction is important, develop a theory for history-dependent friction, test and apply this theory on atomic-, meso- and macroscopic scales and apply it to key problems in glaciology and geoscience.

A plot showing a slip pulse in a molecular dynamics simulation.
Illustration of a slip pulse in a molecular dynamics simulation shear rupture in silicate (SiO2). The top figure illustrates the displacement field and the bottom figures illustrates the propagation of the slip pulse by snap-shots of the local displacement over 4 picoseconds (Henrik Anderson Sveinsson).

In 2020, we have extended a one-dimensional model to address slip-pulse dynamics, which opens for new insights into how slip pulses affect friction and earthquake processes. We have also extended a molecular dynamics potential for the interaction of water and silicates and developed a machine-learning based approach to determine the model parameters under various geological conditions. The model is currently used to determine good parameter values to address hydro-fracture processes under realistic geological conditions. Simultaneously, we have built a new high performance computing cluster with 1100 cores to effectively model water-silicate systems over nano-second to micro-second time frames. We expect these models to provide new insights into fracture and friction processes in dry and wet silicates in 2021.


The Research Council of Norway


  • The Njord Center, University of Oslo, Oslo, Norway
  • CNRS, Laboratoire de Tribologie et Dynamique des Systèmes, Ecole Centrale de Lyon, France

Seals of project participants.



Cristin returned 'not found'

Published Apr. 28, 2021 2:46 PM - Last modified Oct. 20, 2021 2:10 PM