Stanford University

ERE Seminar: Pengcheng Fu, PhD, Lawrence Livermore National Laboratory — The role of computer simulations in researchers' cognitive processes towards scientific discovery: three stories related to geomechanics simulations

Monday, Feb 25, 2019 12:30 PM
Room 104, Green Earth Sciences Building, 367 Panama Street, Stanford
More Info:
ERE Seminar: Pengcheng Fu, PhD, Lawrence Livermore National Laboratory — The …
General Public, Faculty/Staff, Students, Alumni/Friends, Members
Energy Resources Engineering

The role of computer simulations in researchers’ cognitive processes towards scientific discovery: three stories related to geomechanics simulations

Pengcheng Fu, PhD
Staff Scientist, Computational Geosciences | Lawrence Livermore National Laboratory


Computer simulation’s role in science has traditionally been interpreted focusing on the relationship between simulation and the “objects” of research, namely math/physics/chemistry as well as data. This seminar discusses how computer simulation facilitates scientific discoveries by influencing the “subject” of research - that is to say, researchers - or more accurately, the cognitive processes of researchers.

The thesis of this talk is that researchers’ inherent cognitive biases often hinder discoveries, and an underappreciated role of computer simulation is to remind researchers of these biases. Nobel laureate Daniel Kahneman distinguishes two ways of thinking: System 1, which is fast, automatic, and intuitive, and System 2, slow, deliberate, and logic. Although most of us identify with System 2, System 1 actually dictates most of our thinking, even when we thought we are thinking hard! System 1 relies on the apparent coherence of a story to assess its quality, which often causes us to be stuck in a seemingly coherent but false theories. We argue that high quality, first-principle-based simulation is an effective means to engage System 2 to overcome our inherent biases and thereby to enable exciting scientific discoveries.

These arguments will unfold around three pieces of interesting research work performed by the author, which all originated from prevalent biases in their respective research communities: (1) the generalization of the so-called Mohr-Coulomb strength criterion to anisotropic granular materials, (2) hydraulic fractures in carbon storage caprock modeled as high-permeability zones, and (3) the migration of “steam chambers” in California’s diatomite reservoirs resulting in “surface expressions”. The talk will conclude with some practical tips for computer simulations to more effectively engage our critical thinking.

Dr. Pengcheng Fu is an earth scientist in the Computational Geosciences Group at Lawrence Livermore National Laboratory. He received his Ph.D. in geotechnical engineering from the University of California, Davis in 2009, and M.S. and B.S., both in civil engineering, from Tsinghua University, China, in 2004 and 2001, respectively. His main research interest is the development and application of massively parallel, multi-physics numerical models for a variety of subsurface processes with objectives of enhancing energy recovery while minimizing environmental impacts. Pengcheng was a member of the team that started LLNL’s GEOS code seven years ago. He currently works on the applications of GEOS in the stimulation of unconventional oil and gas reservoirs, engineered and hydrothermal geothermal systems, and CO2 sequestration.

maillinkedindouble carrot leftarrow leftdouble carrotplayerinstagramclosecarrotquotefacebooktwitterplusminussearchmenuarrowcloudclock