Event Calendar
Prev MonthPrev Month Next MonthNext Month
ASCE-SF Geo-Institute Dinner Meeting
Scott's Seafood
2 Broadway
Oakland, CA 94607

View additional information
Thursday, November 15, 2018, 6:00 PM - 9:00 PM PST
Category: Other

San Francisco Geo-Institute Dinner Meeting

The Probabilistic Design Approach and Its Application to Geotechnical Problems
(with Emphasis on Earth Dam and Levee Filter Design)

By Sangho Lee

Abstract:

Various probabilistic methods have been applied to geotechnical problems such as probabilistic settlement estimates, Monte Carlo stability checks of retaining walls, probabilistic earth dam/levee filter design, probabilistic geotextile pore characterization and simulated random microcrack pattern in sedimentary rocks. The probabilistic approach is a powerful tool to estimate anticipated feature (mean and variation) of analysis outputs based on probability distribution shapes of model input parameters, identified from prior cumulative field investigation results. The deterministic method usually assumes the worst case scenario of model inputs and can produce overly conservative design resulting in excessive construction costs. Consequently, deterministic design methods may not fit in well with sustainable development in public and energy sectors, demanded by new a global economy paradigm.

The probabilistic approach is the key to finding the optimal solution between the economic and safety aspects. Backward erosion or piping phenomena of fine particles within cohesionless soils cannot be predicted with current deterministic dam/levee filter design which simply compares two selective particle sizes from the base soil and filter gradations. Accurate evaluation of base soil internal stability is very important in order to predict filter performance after dam/levee construction. Many filter design failures have stemmed from ignorance of progressively changed field conditions, such as inhomogeneous seepage flow generation and a subsequent change of soil relative density and gradation, which contrasts unfavorably with the homogenous and uniform flow regimes presumed in the design stage.

The author has introduced a new probabilistic filter design method for improvement of filter performances in terms of both retention and drainage purposes. This probabilistic method can handle entire grain size distributions (GSD) of base soil and filter material to compare the constriction size distribution (CSD) of pore structures between two discrete media, which is closely related to the mechanisms of internal stability against "piping" and a self-filtration process called "bridging". The proposed model does not as yet cover the reliability issue of filter design to consider the variety of native soil conditions. However, this probabilistic method, using thefull base soil GSD and relative density, can help understand the piping and filter mechanism more clearly and lead to more advanced stochastic models capable of dealing with the range of site soil features identified from the field investigation.

About the Speaker:

Dr. Lee has worked on many local, state and federal civil design/construction projects and is familiar with international/state building codes and regulations and engineering practice standards such as AASHTO, FHWA, DOT , Army Corp and FERC manuals. He has also managed the Army Corp/AMRL certified materials testing laboratory since 2009. Dr. Lee has published and/or presented 15 technical papers in international conferences and peer-review journals in the fields of civil/geotech/environmental engineering with nine publications dedicated to practical applications of probabilistic methods to solve real engineering problems.

Register

Professionals: $60
Students: $20
Walk-Ins: $70, Walk-ins are welcome, but please pre-register if you can!

REGISTER NOW