Technical Report NTB 90-02

Numerical modelling of the creep behaviour of clays with emphasis on tunnels and underground openingsA critical review of the state-of-the-art

This report presents an interpretive overview and critical assessment of the state-of-the-art for numerical modeling of the creep behavior of clays. The overview and assessment is focused upon application to underground openings. Field and laboratory observations of time-dependent behavior, constitutive modeling of creep behavior, and numerical implementation of constitutive equations are addressed. A critical assessment of the ability of existing models to predict aspects of creep behavior relevant to waste repository design and suggestions for improved analyses that can be developed with existing technology are provided.

Observations of creep behavior in the laboratory and in the field are reviewed to provide a background upon which to assess the adequacy of existing constitutive equations and numerical models for reproducing important facets of creep behavior in clays. The lack of data on creep under long term drained conditions is noted. Attention is also called to the lack of significant data on the creep of stiff clays and clayey rocks, on temperature effects, and on creep under drained conditions for other than one-dimensional stress-states.

Both heuristic and mathematical constitutive models are reviewed. Heuristic models provide a basis for evaluation of the required parameters for the continuum mechanics based mathematical models. The continuum mechanics models are required for numerical analysis. It has been demonstrated that, by using iterative and incremental analysis, virtually any viscous or inviscid continuum mechanics material model can be adapted to consider time-dependent behavior.

Available numerical models for numerical analysis of geotechnical problems involving creep deformations are reviewed. Models for thermo-mechanical coupling are also addressed in this review. Cases where creep-inclusive analyses have been applied to analysis of prototype behavior are cited. However, the lack of well documented case histories of time-dependent deformations over significant time spans is identified as a major obstacle to model verification.

Recommendations are made for an alternative design approach capable of guaranteeing the very long term mechanical integrity of the liner.