Use of structural forward models produced by the Distinct Element Method as a benchmark for restoration methods

Benjamin P. Chauvin and Andreas Plesch and John H. Shaw and Peter J. Lovely. ( 2018 )
in: 2018 Ring Meeting, ASGA

Abstract

Forward modeling and structural restoration are valuable tools used to understand the evolution of structures through time. The former defines pathways to explain the observed structural geometry from an initial undeformed paleo-geometry. The latter aims to retro-deform the present-day geometry, thereby providing insights into the processes of deformation and helping to validate structural interpretations. Both methods present distinct advantages and drawbacks. On one hand, complex mechanical behaviors can be used in forward simulations. However, forward methods often struggle to precisely reproduce natural geometries. Moreover, they rely on a paleo-geometry which is rarely known. On the other hand, structural restoration directly uses the present-day geometry, but its accuracy is limited by kinematic and/or geomechanical assumptions and simplifications (e.g., area or volume conservation, elasticity). In addition, there is no reference solution (paleo-geometry) to validate the result. In this paper, we seek to overcome these limitations by using realistic, mechanical forward models as the basis for structural restorations. The numerical forward models are developed using the Distinct Element Method (DEM), and enable us to investigate the importance of geometrical and geomechanical parameters in the development of specific classes of structures. Given that the boundary conditions, original geometries, and the displacement field for these models are perfectly known through time, we use them to test the effectiveness of different restoration methods. The DEM represents rocks as a granular material composed of numerous small particles that interact according to well understood mechanical rules, enabling us to produce geologically realistic structures with a self-emergent fault network. We apply the DEM to produce 3D compressional and extensional models, with and without mechanical layering. All models contain faults, fault-related folds, and growth stratigraphy. A frictional contrast on the base of the models, which acts as a detachment, enables us to produce non-cylindrical structures. All models are deformed by a vertical driving wall. The extensional models have an inclined base generating a gravitational potential which is consistent with some geological extensional domains. In compressional models, we show that the pre-growth thickness has an impact in the transition between the fault-related folding and the detachment folding. The presence of flexural slip interfaces promotes fault-propagation folds in the case of a thick pre-growth, and an asymmetry in the detachment fold in the case of a thin pre-growth. Our extensional models exhibit asymmetric graben development, with secondary normal faulting localized with rollover panels. In these models, mechanical layering delays secondary faulting within the primary graben. We sequentially restore an extensional, mechanically homogeneous DEM forward model with two different restoration techniques. The first method is geometric and based on the adaptation of the 3D mathematical formulation of the deposition (Wheeler) space. The second method is geomechanical and based on continuum mechanics and elasticity. We show that forward models built by the DEM constitute excellent benchmarks to qualitatively assess the success of restoration methods. Both are able to restore the first stages of our extensional model. However, the more the model is retro-deformed, the more there is a discrepancy with the reference solution, in particular around fault branching.

Download / Links

BibTeX Reference

@INPROCEEDINGS{,
    author = { Chauvin, Benjamin P. and Plesch, Andreas and Shaw, John H. and Lovely, Peter J. },
     title = { Use of structural forward models produced by the Distinct Element Method as a benchmark for restoration methods },
 booktitle = { 2018 Ring Meeting },
      year = { 2018 },
 publisher = { ASGA },
  abstract = { Forward modeling and structural restoration are valuable tools used to understand the evolution of structures through time. The former defines pathways to explain the observed structural geometry from an initial undeformed paleo-geometry. The latter aims to retro-deform the present-day geometry, thereby providing insights into the processes of deformation and helping to validate structural interpretations. Both methods present distinct advantages and drawbacks. On one hand, complex mechanical behaviors can be used in forward simulations. However, forward methods often struggle to precisely reproduce natural geometries. Moreover, they rely on a paleo-geometry which is rarely known. On the other hand, structural restoration directly uses the present-day geometry, but its accuracy is limited by kinematic and/or geomechanical assumptions and simplifications (e.g., area or volume conservation, elasticity). In addition, there is no reference solution (paleo-geometry) to validate the result. In this paper, we seek to overcome these limitations by using realistic, mechanical forward models as the basis for structural restorations. The numerical forward models are developed using the Distinct Element Method (DEM), and enable us to investigate the importance of geometrical and geomechanical parameters in the development of specific classes of structures. Given that the boundary conditions, original geometries, and the displacement field for these models are perfectly known through time, we use them to test the effectiveness of different restoration methods.
The DEM represents rocks as a granular material composed of numerous small particles that interact according to well understood mechanical rules, enabling us to produce geologically realistic structures with a self-emergent fault network. We apply the DEM to produce 3D compressional and extensional models, with and without mechanical layering. All models contain faults, fault-related folds, and growth stratigraphy. A frictional contrast on the base of the models, which acts as a detachment, enables us to produce non-cylindrical structures. All models are deformed by a vertical driving wall. The extensional models have an inclined base generating a gravitational potential which is consistent with some geological extensional domains. In compressional models, we show that the pre-growth thickness has an impact in the transition between the fault-related folding and the detachment folding. The presence of flexural slip interfaces promotes fault-propagation folds in the case of a thick pre-growth, and an asymmetry in the detachment fold in the case of a thin pre-growth. Our extensional models exhibit asymmetric graben development, with secondary normal faulting localized with rollover panels. In these models, mechanical layering delays secondary faulting within the primary graben.
We sequentially restore an extensional, mechanically homogeneous DEM forward model with two different restoration techniques. The first method is geometric and based on the adaptation of the 3D mathematical formulation of the deposition (Wheeler) space. The second method is geomechanical and based on continuum mechanics and elasticity. We show that forward models built by the DEM constitute excellent benchmarks to qualitatively assess the success of restoration methods. Both are able to restore the first stages of our extensional model. However, the more the model is retro-deformed, the more there is a discrepancy with the reference solution, in particular around fault branching. }
}