We propose a new synthetic finite element regression analysis for the determination of the geostress field in engineering construction areas. In this method, numerical simulation, regression analyses and a small number of actual measurement stress data are combined in a synthetic evaluation of the geostress field. Stepwise regression is first performed to screen the major factors determining the geostress field and to ensure the completeness and significance of these factors. Bound regression is subsequently performed to analyze quantitatively the influence of each major factor on the geostress field and thus ensure the accuracy of the results. As an experimental case study, the proposed method was applied to analyses of the formation of the initial geostress field in the Baojiadian coal field, and the results were found to agree well with actual conditions.


The geostress state of a construction area is a major factor in determining engineering stability. Geological hazards are usually closely related to the geostress environment. With economic growth in China, there are an increasing number of large engineering projects, such as excavations of underground caverns, large mining/oil fields, nuclear waste disposal projects, and large hydraulic/hydroelectric facilities. Because many of these projects involve a complicated high geostress environment, accurate information of the geostress state of the construction area is critical in project design. Therefore, better understanding of the nature of rock/solid media and their interaction with engineering architectures can provide scientific knowledge as well as practical guidelines for the rational design, construction, and operation of such engineering projects. The present tectonic stress field has been studied from various perspectives, such as those of tectonic evolution, crustal deformation, and focal mechanisms. For a given location or area, these approaches can provide the approximate direction of the principal stress, but not the exact value of the stress.

This content is only available via PDF.
You can access this article if you purchase or spend a download.