Questions on Integrating Gravitational and Tectonic Loading Effects in PyLith Modeling

PyLith
1

We aim to develop a comprehensive geodynamic model that simultaneously including gravitational (collapse) and tectonic loading. According to your 2D gravity and 3-D subduction examples in PyLith 2.2.1, we consider these processes in our complex geodynamic models with specific topographic and lower crustal viscosity variations.

In the configuration file (.cfg), we directly introduced gravitational effects and applied tectonic loading (GNSS velocity boundary conditions) along the western (W), eastern (E), northern (N), and southwestern (SW) boundaries, while designating the southeastern (SE) boundary as a free boundary. Notably, we exclude fault constraints from the simulation since the inclusion of them leads to model instability and failing to converge (attributed to static friction effects).

During the simulation, we observed significant vertical displacements (Vertical_displacements.jpg), raising concerns about the validity of our current modeling approach. We note that some PyLith examples employ an opposing force to counteract gravitational effects, thereby minimizing vertical displacement. However, this approach is unsuitable for our research objectives, as neutralizing gravitational effects would preclude the model from capturing gravity-induced horizontal displacements. Our primary challenge lies in effectively integrating gravitational and tectonic loading effects—ensuring that the model reflects gravitational collapse as well as accurately simulating tectonic loading responses.

We would greatly appreciate your professional insights and suggestions on improved modeling strategies. Your expertise could help us optimize the model design, better balance these two effects, and more accurately simulate the underlying physical mechanisms and evolutionary patterns in geological processes.
pylithapp.cfg (10.3 KB)
cd_gra.cfg (2.1 KB)

Displacements
Displacements509×314 38.6 KB

2

This is a difficult problem to completely understand. It sounds like you are trying to start a gravitational problem from a non-equilibrium state, but you don’t want the large vertical displacements that occur when you simply ‘turn gravity on’. These displacements occur as a result of having compressible material in your model. In the examples, we provide one method of avoiding these large initial displacements, which is to apply the computed stresses from a static simulation as initial stresses. Since your problem includes topography (and presumably density variations), you would need to first do a static calculation to compute the initial stresses. I think that the most reasonable way to do this would be to assume a nearly incompressible material, and use the resulting stresses as your initial stresses. You should also use the ImplicitLgDeform formulation, since otherwise the computed stresses will not take into account the deformed mesh. This may not give you the gravitational collapse you are looking for, but otherwise it will be extremely difficult to separate out the ongoing gravitational collapse from the effects due purely to elastic compressibility. Your other option is to not use initial stresses, and simply take the deformed state from your first step as your reference state (still using the ImplicitLgDeform formulation). In this case, you will still get large vertical deformations in your first time step, but these will decrease significantly in subsequent steps.