Geomechanical models are indispensable for reliable design of engineering structures and processes and hazard and risk evaluation. These models are however far from perfect. Errors are introduced by fluctuations in the input or by poorly known parameters in the model. To overcome these problems an inverse modelling technique to incorporate measurements into the deterministic model to improve the model results can be implemented. This allows for observations of on-going processes to be used for enhancing the quality of subsequent model predictions.

In geomechanics several examples of inverse modelling exist where the improved model of the system is obtained by minimizing the discrepancy between the observed values in the system and the modelled state of the system within a certain time interval [1,2,5,6,7,8,9]. This requires the implementation of the adjoint model. Even with the use of the adjoint compilers that have become available recently, this is a tremendous programming effort for the existing geomechanical model system.

Therefore, the Ensemble Kalman filter [3,4] is implemented to overcome this problem. The Ensemble Kalman filter is already rather common in meteorology and oceanography and the state of the system is analysed each time data becomes available. An additional challenge is to incorporate the influence of the heterogeneity of the ground.

Very promising results of a conceptual example, based on the construction of a road embankment on soft clay, are presented. Also the effect of incorporating measurement noise is shown as well as the importance of the location of the observation points. In the near future, the Ensemble Kalman filter is not only used for a straight forward identification of the elastic modulus E and the K₀ parameter of several soil layers below the embankment, but also incorporates the uncertainty of the exact location of the boundaries between the subsequent layers.

1

Cividini, A., Jurina, L. and Gioda, G., "Some aspects of 'characterization' problems in geomechanics". Int. Journal of Rock Mech. and Min. Science and Geom., vol. 18, 487-503 (1981). doi:10.1016/0148-9062(81)90513-1

2

Cividini, A., Maier, G. and Nappi, A., "Parameter estimation of a statistical geotechnical model using a Bayes' approach". Int. Journal of Rock Mech. And Min. Science, vol. 20, 215-226 (1983). doi:10.1016/0148-9062(83)90002-5

3

Evensen, G., "Sequential data assimilation with non-linear quasi-geostrophic model using Monte Carlo methods to forecast error statistics". Journal of Geoph. Research, vol. 99, C5/10, 143-162 (1994). doi:10.1029/93JA02082

4

Evensen, G., "The Ensemble Kalman filter: theorethical formulation and practical implementation". Ocean dynamics, vol. 53, 4, 343-367 (2003). doi:10.1007/s10236-003-0036-9

5

Gens, A., Ledesma, A. and Alonso, E., "Estimation of parameters in geotechnical back analysis - II. Application to a tunnel excavation problem". Computers and Geotechnics, Vol. 18, No.1, p. 29-46 (1996). doi:10.1016/0266-352X(95)00022-3

6

Gioda, G. and Maier, G., "Direct search solution of an inverse problem in elasto-plasticity: identification of cohesion, friction angle and in situ stress by pressure tunnel tests". Int. Journal for Num. and Anal. Methods in Geomechanics, vol. 4, 15, 1832-1848 (1980).

7

Gioda, G., "Some remarks on back analysis and characterization problems in geomechanics". Proc. 5th Int. Conf. On Numerical Methods in Geomechanics, Nagoya (1985).

8

Honjo, Y., Liu, W.T. and Soumitra, G., "Inverse analysis of an embankment on soft clay by extended Bayesian method". Int. Journal for Num. and Anal. Methods in Geomechanics, vol. 18, 709-734 (1994a). doi:10.1002/nag.1610181004

9

Honjo, Y., Wen-Tsung, L., Sakajo, S., "Application of Akaike information criterion statistics to geotechnical inverse analysis: the extended Bayesian method". Structural safety, vol. 14, 5-29 (1994b). doi:10.1016/0167-4730(94)90004-3

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £135 +P&P)