SOLICITATION NOTICE
B -- Professional services to provide physics-based earthquake model in modern object-oriented computer code.
- Notice Date
- 6/3/2004
- Notice Type
- Solicitation Notice
- Contracting Office
- U S GEOLOGICAL SURVEY, APS BRANCH OF ACQUISITION AND GRANTS 3020 STATE UNIVERSITY DR. EAST, MODOC HALL STE 3001 SACRAMENTO CA 95819
- ZIP Code
- 20192
- Solicitation Number
- 04WR14NOSOLICITATION
- Response Due
- 6/14/2004
- Archive Date
- 6/3/2005
- Point of Contact
- ANN MARGRAVE PURCHASING AGENT 9162789329 amargrave@usgs.gov;
- E-Mail Address
-
Point of Contact above, or if none listed, contact the IDEAS EC HELP DESK for assistance
(EC_helpdesk@NBC.GOV)
- Small Business Set-Aside
- N/A
- Description
- NA The U.S. Geological Survey (USGS) intends to negotiate on a sole source basis with Invisible Software, 2058 Coastland Avenue, San Jose, CA 95125. The proposed contract is to provide a comprehensive physics-based earthquake model in modern object-oriented computer code in collaboration with USGS scientists as follows: SCOPE - When slip occurs on a fault, it changes the distribution of stresses in the surrounding rock. Afterwards, the stresses undergo a slow change due to viscoelasticity. The amount of slip is determined by stress-induced forces acting on the fault by friction, roughness, and other factors. The primary goal of this contract is to develop computer code that calculates the amount of slip on the fault, the immediate changes in stress, and the evolution of stress over time caused by viscoelasticity. The calculations will be performed using the finite element method (FEM). The program shall address the following requirements: STATEMENT OF PROBLEM - COMPUTATIONAL DOMAIN. The computation takes place in a 3-D "coordinate space". A non-linear transformation maps the coordinate space into the "physical space" where the earth resides. Because the transformation is nonlinear, the coordinate space is a non-Euclidean space. In particular, it is a Riemannian manifold with a curved metric. The coordinate space is constructed in such a way that the gravitational equipotential surfaces are the level surfaces of one of its coordinates. This ensures that the top and bottom surfaces of the computational domain are locally horizontal, and that there is no gravitational loading across the sides of the domain. The program must be able to accept a geodetic model of the earth's gravitational field, construct the non-linear transformation between coordinate and physical spaces, and calculate the Riemannian metric of the resulting curved manifold. FEM FORMULATION. The finite element method is carried out in the computational domain. Because the domain is a curved manifold, it is necessary to use a tensorized version of FEM instead of the standard version. In the tensorized version of FEM, all physical equations must be expressed in tensors in a form appropriate for curved space. All volume and surface integrals must be performed using curved-space formulations involving the metric and Levi-Civita tensors. The program must be able to generate a mesh surrounding a specified fault system. Cell boundaries are positioned to coincide with the fault surfaces, and the mesh includes special nodes that represent slip along the fault surfaces. The program will include a linear solver able to handle very large problems and able to solve sparse linear systems of several million equations on a desktop PC in a few minutes. The solver must be able to impose constraints, for example, constraining the motion of certain nodes to lie in a fault surface. The constraint support must be intrinsic to the solver (for example, not relying on the addition of fictional "springs"). It must be possible to modify constraints "on-the-fly" without recomputing the coefficient matrix. The FEM code will be constructed in object-oriented form, and will support cells of varying shapes, including hexahedra, tetrahedra, triangular prisms, and quadrilateral pyramids. The FEM code will be sufficiently general to handle any arbitrary cell shape, needing only a suitable description of the desired shape. It will also handle both linear and quadratic cells. ADDITIONAL CAPABILITIES. The program will be able to calculate the evolution of stress over time, caused by viscoelastic relaxation of the rock. It will support both Newtonian and non-Newtonian viscoelastic materials. The program will be able to calculate the effects of friction on the fault surface by calculating the normal and tangential forces at each node and making them conform to a fiction model. A special type of quadratic cell is required for this calculation, since standard quadratic FEM cells to not allow physically correct calculation of surface forces. As a validation tool, the program will incorporate an analytic dislocation solver, which can be used in simple cases to provide analytic solutions for comparison to the FEM solutions. The analytic solver will implement a solution similar to the well-known Okada formulas, except modified and extended so that the fault slip is given by an FEM shape function. An additional validation tool will be a built-in patch-test module that will test various cells and combinations of cells to ensure they behave correctly. The program will include a graphics module able to draw images on an internal bitmap and export them to TIFF files. The images will be used to visualize the FEM meshes and other program output as needed. The program will be able to handle slip along non-planar faults by constraining motion along the fault surface in such a way that it produces the correct total value for the seismic moment tensor at each FEM node. GEOPHYSICAL EFFECTS. The program will support both flat-earth and curved-earth computations. The program will include a geodetic model which takes into account: the oblateness of the earth; centrifugal forces due to earth's rotation; and the variation of gravitational acceleration with depth. It will account for gravitational body forces and for the effects of friction and roughness along the fault. The program will allow for specifying the elastic parameters of the crust, and allow them to vary with location. It will calculate the effects of the viscoelastic behavior of the lower crust and permit the elastic parameters, viscosity, and non-Newtonian exponent to vary with location. In general, the program will be carefully constructed so that additional geophysical effects can be added over time, without rendering the existing code obsolete. PROGRAMMING CONSIDERATIONS - The program will be written in C++, and will be designed to be portable. There are several dimensions to portability. A. Code portability. The code will adhere to the ISO C++ standard, and not use functionality specific to any one operating system. The program will also avoid arcane or obscure code constructions (whose support might vary amount different compilers) and programming practices that are known to introduce portability issues. Where system dependency is unavoidable (e.g., file naming conventions), such code will be isolated in one place for easy modification. B. Data portability. Strict control of I/O on a byte level (so that data files are byte-for-byte identical) will be maintained to ensure data-file portability on different systems. Where the code must read files prepared by other programs (i.e., text files created by a text editor), the code will be flexible enough to accept a variety of different formats, and such code will be isolated in one place so additional formats can be added easily. C. Command portability. When programs perform a large set of functions, users often write script files to conveniently issue a complicated series of commands. Since script files tend to be non-portable, this can make it difficult to get the program to perform the same task on different systems. This problem will be avoided by defining a simple script language, and building in code to read the script and execute the commands. D. Time portability. In order to write code that will execute on future computer systems, running operating systems that are yet to be developed, the program will not rely on any facility that has a significant risk of becoming unavailable in the foreseeable future. E. Self-diagnostics. The program will have built-in diagnostics to run tests of the various program modules and verify that they are operating correctly. This will allow each new version of the program, and each port to a different system, to be immediately tested by running a test script. F. Scientific portability. The contractor shall ensure program portability to a variety of seismotectonic settings, as well as ensure access to scientists with more modest computer skills. G. Deliverables. 1. Data identified below are to be in digital report format, and can be transmitted either by e-mail or as anonymous ftp data transfer. Period of performance: 07/01/2004 through 03/31/2005. Invisible Software can provide a professional computer scientist with broad expertise in math and physics, who is actively developing a 3-D FEM model to simulate earthquake processes, so has a detailed understanding of the mathematical and physical basis for our rapidly evolving scientific observations and deductions. This combination of skills is unique in the computer science field and provides a highly qualified match for writing programs that use advanced math to analyze and model interconnected physical processes. Thus, the vendor can quickly and efficiently construct the core computer model in modern object-oriented code, and then add program modules that both develop the complexity of the analysis and allow custom-designed earthquake studies. The vendor is an expert in writing computer programs in C++ and is experienced in all phases of the software design cycle, from concept to architecture to coding to testing to documentation and deployment; and then to upgrading and revising for new operating environments, new hardware, and new customer requirements. The vendor is familiar with 3-D FEM models in general, and the specific elements required to fully develop a comprehensive earthquake model. In addition, the vendor could immediately begin writing computer code to construct the required products, and has the skills, expertise, and training required to complete this project in a timely fashion. THIS NOTICE IS FOR INFORMATION PURPOSES ONLY. THIS IS NOT A REQUEST FOR COMPETITIVE PROPOSALS. Information will be considered solely for the purpose of determining whether to conduct a competitive procurement. Determination to compete this requirement is solely within the discretion of the Government; no additional synopsis will be published. Contact: Ann Margrave, (209)742-5079 - email amargrave@usgs.gov NOTE: THIS NOTICE WAS NOT POSTED TO WWW.FEDBIZOPPS.GOV ON THE DATE INDICATED IN THE NOTICE ITSELF (03-JUN-2004); HOWEVER, IT DID APPEAR IN THE FEDBIZOPPS FTP FEED ON THIS DATE. PLEASE CONTACT fbo.support@gsa.gov REGARDING THIS ISSUE.
- Web Link
-
Please click here to view more details.
(http://www.eps.gov/spg/DOI/USGS/USGS/04WR14NOSOLICITATION/listing.html)
- Place of Performance
- Address: San Jose, CA
- Zip Code: 95125
- Country: USA
- Zip Code: 95125
- Record
- SN00596532-F 20040605/040603212649 (fbodaily.com)
- Source
-
FedBizOpps.gov Link to This Notice
(may not be valid after Archive Date)
| FSG Index | This Issue's Index | Today's FBO Daily Index Page |