Currently, due to the limitations of continuum methods, discrete methods become increasing popular, however, to the best knowledge of mine, it seems that molecular dynamics is not prevalent in rock engineering. I do know some drawbacks of molecular dynamics methods, for example, the spatial and temporal scale restriction. Are there any other limitations of this method in rock engineering?
Previously, when I worked as a researcher at the university's aerospace department, I had a doubt that it simulated only with equations of fluid dynamics, and I calculated the molecular dynamics of the shock wave. In any case, I think that the most important issue is the selection of potential. I independently created the potential. Even in rock engineering, I think that it is the same thing that the kind of potential etc. is decided by what kind of phenomenon is trying to elucidate.
Dear Yuji
You are right. Currently, most molecular dynamics simulations about rock engineering are related to the fluid flow (diffusion, transport, etc) or surface characteristics (surfactants, roughness, etc).
Because molecular dynamics simulation is a force-field (also known as potential function) based method, therefore, choosing the appropriate potential function is of great significance.
Also, different from other materials, for example, polymers, other condensed matter, etc., the heterogeneity is the essential attributes of rock, therefore, heterogeneity should also be taken into consideration when dealing with rock engineering problem.
Thanks for your advise.
Regards
Mr. Zhaoyang,
Due to the horarium of 'computational quantum chemistry', 'theoretical chemistry' and 'theoretical physics'. This is valid to all disciplines, including already mentioned ones. There are also: medicine, medicinal chemistry, geochemistry and earth sciences, agricultural chemistry and relating sciences, etc.
Dear Bojidarka
Thanks for your answering. Yeap, I do know that molecular dynamics is feasible for rock engineering and it has been used for a wide range of disciplines, however, because my major is civil engineering, when I search the papers about the related topics, I found that most papers deal with fluid flow and surfactant problems, while the amount of papers related to rock mechanics is relatively small. Therefore, I wanna know the reasons for this phenomenon.
Regards
Zhaoyang
FYI
“ In recent years computational modelling has become an important tool to characterise properties of geo-materials and study their behaviour at geological conditions. The fact that materials can be analysed at the atomic level allows for a better understanding of many of the features observed in samples from nature. Computational methods make it possible to study materials at high pressure and temperature where experimental techniques fall short. In this line, properties such as equation of state, deformation at high pressures, defects and trace element incorporation for several materials have been performed in recent years.
However, the impact of atomistic simulations on Earth science is constrained by two factors: the accuracy of the model and the length and time scales of the simulation. In recent years most of the modelling has been based on very accurate quantum methods based on density functional theory (DFT). These methods make it possible to calculate properties of a wide range of materials with different compositions ab initio, i.e. without a specific assumption for the atomic interactions. Unfortunately, DFT is currently limited to systems containing a few hundreds of electrons and to dynamics with timescales of a few tens of picoseconds. In most of the cases this is enough to study relative stabilities, equations of state, crystal defects and even aggregation and dynamics of melts.
For large scale problems other methods such as empirical atomistic force fields are often used. They are based on a description of the system in terms of atoms (i.e. without explicit electrons) interacting with each other through a potential described by an analytic expression. The advantage of such methods is that they can allow for simulations containing up to millions of atoms and for simulation times on the order of nanoseconds. However, the accuracy of the method strongly depends on the functional form and on the choice of the parameters used to describe the interatomic interactions. Parameters are usually optimised to reproduce known experimental properties or high level DFT calculations. Simple additive potentials are often used; however, they fail to describe non-central forces. In the case of oxides, the high polarizability of O2- cannot be described by a pair potential. ”
cited from An atomistic model of MgSiO3 perovskite and post-perovskite phases
- Badges
- Science topic
- Similar topics
- Chemistry
- Theoretical Chemistry
- Computational Chemistry
- Molecular Dynamics