Given the need to evaluate the older infrastructure for higher design hazards in order to ascertain the performance in multiple sigma events, the use of nonlinear tools for the prediction of the performance of concrete structures is becoming increasingly popular. Smeared crack models are the primary tools to simulate the performance of dams due to such properties as mesh independence and robust behavior, as well as being dependent on only a few parameters discernible by common laboratory tests. In contrast to the use of linear elastic models, a general overview or unifying suggestions on the use of nonlinear analyses tools for design or evaluation of dam structures is hard to come by: two design engineers can reach completely different decisions on the behavior of a system due to the complexity and unpredictable nature of the analysis tools. The goal of this article is to provide a general set of suggestions and guidelines in the use of cracking models for the prediction of the behavior of concrete gravity monoliths. The study hopes to establish this premise in two stages. First a set of calibration studies were conducted for the accurate computational prediction of the response of a dam monolith for 4 different controlled static and dynamic laboratory experiments. For practical purposes, “deviations” from the calibrated models are more important as they signify design assumptions or decisions by the engineer. Therefore, in addition to the calibration study given above, a sensitivity study is conducted in the second phase of the study in order to determine the effect of the deviations from “calibrated” models on the analyses results. In this fashion, a general framework is suggested for the dependability of nonlinear FE results for analyses as well as the setting of the bounds on the “interpretation” on such type of analyses.
