Case 97-11

Sandia National Laboratories

Micromechanics of Failure in Brittle Geomaterials

A Case Study of the R&D Value Mapping Project

Institute for Policy Research and Development

School of Public Policy

Georgia Institute of Technology

Atlanta, GA 30332-0345

Unedited Draft

This case was written by Susan Silver under the direction of Juan D. Rogers and Barry Bozeman. Comments or questions should be directed to Barry Bozeman (404-894-0093; juan.rogers@pubpolicy.gatech.edu). The research was sponsored by the Department of Energy, Office of Basic Energy Sciences, Contract ER45562. The interview was conducted by Barry Bozeman and Hans Klein. The views represented here are the case author’s and do not necessarily represent those of the Department of Energy, Sandia National Laboratories, or Georgia Institute of Technology.

The Micromechanics of Failure in Brittle Geomaterials

Project Summary

This case outlines the research conducted by Joanne Fredrich at Sandia National Laboratory. BES funded research has been the basis for the development of geomaterial pore imaging analysis software which creates 3-D rock specimen reconstructions. It is this 3-D imaging technique that resulted in publication in Science. Researchers investigation provides a detailed understanding of the micromechanical processes associated with the failure of brittle Geomaterials. Knowledge of the failure behavior of rocks is not only relevant to oil and gas exploration, there are additional application to the underground disposal of nuclear waste, and drilling technology.

Although the BES research comprises only about 30% of Fredrich’s research at S

Her work is funded from multiple sources. In fact, Fredrich notes in the interview the difficulty separating BES sponsored research from other funding sources because much of the applied research she is doing builds on the basic research funded by BES. NASA and DOE programs: SERTA and LDLD also sponsor her research. Kimberly Clark, Shell and Chevron, Union Pacific,the Natural Gas and Oil Partnership are industry sponsors in which her applied research is based on the pore imaging technique.

In this particular case a BES project manager brokered cross-disciplinary collaboration. Fredrich’s modeling technique provided a way to transfer complex physical material data to a computer. Computational Physics collaboration took Fredrich’s Geophysical modeling technique from computer visualization to computer simulation. Computational modeling and simulation work has been conducted using the supercomputing capabilities of the teraflop computer at Sandia National Laboratory.

Sources: The Geomaterial imaging case is drafted based primarily on interview notes prepared by Barry Bozeman in an interview conducted by Hans Klein and Barry Bozeman, with SNL employee Joanne Friedrich. Supplementary information (e.g. grant proposals, summaries, reviews) were not available. The technical analysis is based on Fredrich’s article Science article"Imaging the Pore Structure of Geomaterials". Additional technical background information regarding geophysics was based on Q&A with GA Tech Geology graduate student Laurie Cabena.

Project Description

          This project seeks to provide a fundamental understanding of the effects of grain boundary structure and cementation, damage, state and load path on the deformation and failure of Geomaterials. Researchers carefully measure brittle materials mechanical behavior and changes (cracks) that occur under high pressure and stress. These measurements are the basis for further theoretical analysis.

Technical Background

Pore connectivity is key in analyzing the properties of solids and developing theories of brittle materials failure. The pore structure of rocks is varied and complex. Physical rock properties such as permeability, electrical resistivity, seismic velocity, elastic moduli, and strength all are affected by pore size and the distribution of pores. The connectivity of pores is relevant to predict fluid flow properties or assess events such as movement, damage, or electric conductivity.

          Geophysical laboratory tests apply stress and strain to solid materials. At various stages of failure, researchers apply different loading conditions to the materials being analyzed which result in cracks. The micromechanical failure processes are further elucidated and characterized quantitatively using light microscopy, laser scanning confocal microscopy and scanning electron microscopy². Computational physics is especially relevant to researchers work. Although observation of damage state is an important parameter in theories of rock failure, computational methodology is particularly needed in volumetric loading in which researchers apply specific, minute degrees of damage to the materials being examined.

          A 3-D technique to analyze the geometric complexity of pore space is one of the major outputs of this project and the basis for additional work analyzing the properties of brittle materials. In this technique, the researcher begins by slicing a thin polished planar section of the rock specimen. During the planar preparation, the researcher must carefully induce the damage that might obscure analysis while taking the cross-section. Once the extremely thin cross section is taken, a tracer substance is injected into the specimen void space. The tracer is used to create a pore cast.

          The Laser Scanning Confocal Microscopy (LSCM) used to scan these images was common in imaging biologic solids but Fredrich explains that recent advances and optical and electronic microscopy technology has only recently made this method relevant to a wider variety of solids. 3D computer software developments enable researchers to now take 2D images and convert them to 3D images.

The basis for materials analysis lies in the geophysics discipline however computer modeling applications involve computational physics. Sandia National Laboratory has a relatively large group of geophysicists, eight PhDs. Maybe 10-15 U.S. universities focus on geophysics. Europe has a stronger base in geophysics because there has not yet been much of a convergence of oil and gas interests. Some primary geophysics competitors to researchers at Sandia National Laboratory researchers can be found at Los Alamos, Livermore, the U.S. Geologic Survey, MIT, Stanford, Stoneybrook, California - Berkeley, University of Minnesota, Brown, and the University of Colorado.

Project History

Joanne Fredrich is the central researcher working on the BES geophysics project examining the Micromechanics of Failure in Brittle Geomaterials. The other researcher on the project is Teng-fon Wong, their association can be traced to her undergraduate school career. Fredrich attributes her career path into Geophysics to a class she took with Wong in 1983. He was also on her thesis committee. She describes their work as collaborative and they have kept in touch outside of their geologic work together. He knows her husband, she used to housesit for him Wong. Fredrich and her husband have visited him at Stoneybrook.

Joanne Fredrich defended her PHD in October of 1990. While in graduate school most of her work was funded by the National Science Foundation and the U.S. Geological Survey. In 1991 she went to work as a post doc for industry in 1991. Once that post doc finished, Fredrich’s job search was drawn out. At that time the job market was tight and even the oil and gas companies were cutting back. Personal rather than professional factors weighted higher in her decision to take an industry job with Terratech, in Utah, a company that provides mapping and image processing services to the natural resource businesses and the geographics community. In fact Fredrich states that just out of an MIT PHD program, a job in industry was below her own expectations as well as the expectations of her colleagues. A relationship with the Department of Energy BES division began while she was working for industry at Terratech.

BES’s reluctance to fund industry research played a role in the beginning Fredrich’s career at Terratech. She wanted to split her Terratech time between basic and applied research. Eventually she received basic funding from BES and Terratech supported her because they wanted to keep her happy. While at Terratech Fredrich had a joint appointment at the University of Utah and this enabled her to use the University of Utah’s facilities.

The reproach against applied research also carried over into her initial interaction with the BES division of the Department of Energy. Once at Terratech she began to float research ideas to BES program manager Bill Luth. Luth told her that BES does not really fund industry. If she was going to pursue BES funding further, he advised so don’t try to do for profit work under BES because it sets off the university community. Luth suggested she try to work for a university or national laboratory.

Fredrich left Terratech when presented the opportunity to work for Sandia National Laboratory. In 1993 she found out a staff person was leaving so she applied for a position at SNL. It was not difficult for her to leave Utah. Terratech was not financially strong and she was not excited about the prospects of teaching. While in Utah her husband was not working and he was taking care of the baby. Sandia provided not only and outstanding career move it seemed like a more stable source of income since she was supporting her family. Her first proposal went to BES in 1992, about the same time she was interviewing for the Sandia National Laboratory position. She began working for Sandia in 1994, the same year she received funding from BES.

While still with Terratech, it was interaction with a University of Utah undergraduate student that led her to apply the laser microscopy tool to brittle geomaterials failure analysis. Most people in biology had already used this laser microscopy tool and applied it to bio-materials. The 3-D technique was not part of her original BES proposal, and was an unintended output of BES research. This technique became the common thread in Fredrich’s research for other clients and continued BES work. Fredrich first started on the 3-D technique at Terratech and continued the work once she moved over to Sandia National Laboratory

Changes and restructuring at Sandia National Laboratories had an impact on Fredrich’s research. Fredrich came to Sandia while it was in the midst of downsizing to respond to macro defense industry changes affecting laboratories¹. Also, Sandia was trying to be competitive price wise for private industry work. At first Sandia actually seemed like a less stable working environment than what Fredrich left. Less than 18 months out of Terratech, Fredrich was told she was on a list of potential staff cuts. This did not happen but she indicates it had some impact on the collaborative nature of her research.

Not long after that first round of proposed cuts, Fredrich benefited from implementation of DOE’s integrated job structure that created a new technical staff title which was attached to a slightly better salary band. Despite the benefits of receiving the new title, she indicates that even this shift affected collaboration. There were some individuals that did not get this new title and in a sense this undercuts teamwork. Restructuring also lead to monetary separation policies which meant that Sandia could cut a research and give that funding to someone else. Fredrich explains that the impact has been that her work is probably more contained and her focus more internal. She still tries to offer colleagues input in project development stages however, these environmental changes make it more difficult for her to go out of her way to help peers. Fredrich’s internal focus is probably not just a matter of less reward for collaboration. She funnels ideas and leaves them because her work load has increased leaving her less time to explore possibilities.

Hiring constraints probably have some impact on project collaboration as well. Although Fredrich has received approval to hire, many of the qualified individuals are foreign nationals and it is difficult to hire them. Unlike in Utah with Terratech, university interaction and input was not an option at Sandia. Joanne has not explored adjunct possibilities because the University of New Mexico does not have a Geophysics or Geomechanics department.

Fredrich received a significant amount of response to her article in Science magazine from people within as well as outside her field. Beyond just reprint requests there were non-traditional people contacting here to image soil, paper, bones and ceramics.

Joe Marshal of NASA contacted Fredrich and sent her a sample of materials to see if the imaging technique would work on tile samples. on. about working with her to examine heat shields and how they relate to solid space. This 3-D imaging was key to geometrically map heat shield geometrical information about this. Her analytical technique described in Science was digitally based and permitted pulling out geometric parameters. The new way in which she prepared and scoped samples was based on the BES basic funding but it lead to other interactions funded by other organizations. This work for NASA helped Fredrich establish a broader based materials track record. Scoping NASA’s materials and prepping samples and imaging the materials was performed as part of BES funded research. The work with NASA is ongoing.

The NASA interest in materials analysis lead to funding from the Laboratory Directed Research & Development program (LDRD). It was at this point that computational physics began to play a larger role in her work and she began working with a Civil Engineering post doc at Sandia National Laboratory.

BES was instrumental in bringing researchers together. In August of 1995, Bill Luth of BES sent Fredrich a proposal by Brent Lindquist to review. He was working with volumetric data at BNL and NSLS working to develop an algorithm to extract properties of structure. She gave him high reviews on his proposal. However his work related to physics not rock physics. His strength lied in his computational work but he did not have a feel for the heuristics of the geologic applications. Luth suggested to Fredrich many times that she should get in touch with Lindquist. She explains that had it not been for the proposal Luth forwarded she would probably not have read his work otherwise.

After this peer review she e-mailed Lindquist and they began informally collaborating. This collaboration improved the image mapping capabilities and his knowledge contributions guided them to one-to-one mapping of fluid flow capabilities.

Not only did they share a common tie with MIT, at the time they also shared an office. Collaboration between Fredrich and this post doc was not immediate, they put a project together after getting to know each other. This post doc is the first one that the group has had in 10-15 years.

In the summer of 1996 he and Fredrich worked together and put out a proposal which was funded by Laboratory Directed Research & Development Program (LDRD) for $1.2 million dollars. They were surprised to get this engineering project because they were competing with other computational people. LDRD asked Fredrich and Lindquist to combine efforts across SNL and so Fredrich and Lindquist became the lead on the LDRD contract. Lindquist works with Fredrich as a subcontract. His background is Civil Engineering and he specializes in computational modeling of loads being transferred among materials and geometry volumetric modeling.

Lindquists’ computational contributions have improved upon the 3-D imaging technique. Specifically, improvements of the 3-D technology now allow for more complex mapping and cross sections and one-to-one mapping of fluid flow.

          The 3-D imaging technique described in Science in April of 1995 was the primary output of BES sponsored micromechanics failure research. This key output used to study rock failure has been applied to other scientific problems. Lindquist also developed a computer source code under LDLD. There is no intellectual property agreement on this code at this point. Fredrich commented that she thought it was naive on the part of SNL to offer the code free one year and then start charging a royalty for it.

          In August of 1997 the group received funding from SERTA, the joint DOE/DOD/EPA program. The work is related to waste containment and fluid flow. All of the fluid flow analysis is conducted using the 3D technology developed under BES. Specifically SERTA wants to know how changing fluid transmittal ability of clays stabilizes structure. Clay is porous but absorbent because it is less permeable than sandstone. Clay pores are generally not as interconnected as in brittle materials.

A researcher from the University of Colorado and David Betzel of the 6600 division of Sandia National Laboratory are also working on this SERTA project. Fredrich is the Co-PI on this project and the woman at the University of Colorado is the PI.

          Kimberly Clark is also interested in their work. Fredrich’s post doc colleague Lindquists has received low levels of funding from them for years. She was pleased to get the call from Kimberly Clark especially since LDRD measures success by customers taking advantage of what has already been developed. This company wanted analysis of materials similar to tiles from NASA. Kimberly Clark also showed interest in fluid flow modeling. Researchers pitched an integrated approach involving the code to characterize geometries and imaging. However they just wanted a basic modeling approach. Kimberly Clark has indicated that they might provide some funding in the end for more complex simulation like approach to modeling. At this point LDRD money funds this research.

          A large part of Fredrich’s work at Sandia National Laboratory is with the Fossil Fuel Division. She explains that unlike her other areas of research there is not much technical overlap with the BES project. This area of her work is much more industry oriented and is funded out of the Natural Gas and Oil Partnership. However there is a small technological tie to the BES project. The imaging techniques and some studies on the failure of geomaterial has been relevant to a particular set of problems in which there were subsidences in oil a foot a year and surface sink was a threat.

          The gas and oil industry industry problems have provided the opportunity for the largest simulation work performed by their group to date. Most of their oil and gas industry outputs have been published papers and conference presentations. 3-D imaging is being used to do some problem solving work for Union Pacific. The company has a reservoir in Texas that has a spotty oil well with low poracity. Imaging helps map fractures from which oil can flow.

Outputs and Impacts

          The laser scanning confocal microscopy imaging technique in this case was an unintended output of BES research yet this became a core competency for applied research for additional clients. Benefits of this basic research output have diffused to several different clients in many different disciplines. The imaging technique has been more central to some clients than others, but it is still used by Fredrich in her work in other areas. For instance, although the technique was conceived to analyze the brittle properties of geomaterials, as is the case in client work for SERTA, this technique is now being adapted to examine fluid flow properties of clay, which is relevant to waste containment.

The 3-D imaging approach to materials analysis plays a role in continued work for the following clients:

   NASA

   LDRD

   SERTA / DOE/DOD/EPA

   Kimberly Clark

   Union Pacific

   Natural Gas and Oil Partnership

Additionally, publication in Science magazine was a major milestone which provided Joanne Fredrich and her work exposure both within and outside Geophysics. This ultimately lead to an increasing emphasis and need for interdisciplinary work in the computational physics discipline.

Thus far outputs have been:

   Publication in Science magazine

   New findings that led to further theoretical analysis. Under low confining stresses, open microcracks act as nucleation sites for growth of new microcracks which interact and lead to failure.

   1 post-docs

   One student went to work for industry as a consultant. Works for Failure Analysis Inc. and they consult for oil companies

Analysis of DOE progress report materials or back up materials may provide some more specific information about outputs.

Inputs

          Joanne Fredrich did not start working for Sandia National Laboratory until 1994. The basis for the 3-D imaging technique and the ’95 publication in Science magazine lies in her work conducted while she was working for Terratech. A central imput to her work came from a University of Utah undergraduate biology student. Bill Luth of BES brokered collaboration between Fredrich and Lindquist. This was a significant imput to her work because the changes and shifts in Sandia National Laboratory seemed to inhibit collaboration.

Conclusion:

          This case reads like a Management of Technology case study description of how an R&D laboratory capitalizes on basic research investments. The client based focus of a privately run national laboratory probably has something to do with the researchers ability to apply BES basic research outputs to other research areas.

          The BES funding came to Fredrich relatively early in her career. For this reason the impact is probably greater in terms of providing her a "materials track record". The BES program manager provided input that was instrumental in Fredrich’s research path. Beyond just brokering collaboration in complementary area, her initial contact with BES program manager in the pre-funding stages, while she was still at Terratech proved beneficial to her career choices.