vedi mappa – Sito Web – info email
The new ICOLD Committee for Computational Aspects of Dam Analysis and Design, appointed in the 1987 Executive Meeting of Beijing, addressed its attention to the more pressing needs of the dam engineer’s profession in this field.
It was apparent that a shift of focus has taken place in the last years. Performing a numerical analysis is, seemingly, easier and easier, insofar as automatic computational tools (powerful hardware at reasonable costs, abundance of efficient software) have taken away the drudgery and length of such tasks.
However, shedding this burden from one’s mina, one is more and more aware that computations are only a part of the long road of dam design/analysis. Refined mathematical models, calling for precise values of physical parameters, only highlight our quite wide fringes of uncertainty; deeper and deeper understanding of real behaviour, with all its attending non-linearities and discontinuities, reveal the incompleteness of even the most up-to-date formulations. Besides, the reliability of existing software is too often simply taken for granted, rather than verified.
Last but not least, fitting together the delicate, multi-disciplinary process of “design” with the rigorously specialized job of computational analysis is by no means a trivial matter. Likewise, linking the results of computations to operational decisions is a big jump, and one, moreover, which is not amenable to formalization.
In this context, my personal opinion is that two main points of debate will emerge in the coming years: on one hand, the reliability of dam analysis software (and this topic is addressed by Benchmark Workshops such as the present one); on the other hand, the correct interpretation of computer analysis results in the broader framework of the design process and of the safety appraisal.
In fact, a difficult change of approach is needed (and is partly in the making). From a world of apparently certain rules and formalized knowledge, we are moving with ever increasing awareness to what I would like to call “the rational management of uncertainties”. For this, however, not all the proper tools are yet in existence.
If our ad-hoc Committee will succeed in at least initiating a meaningful activity in one, or both, of the above-mentioned problem areas, I think it will have performed a welcome and useful role for dam professionals in general.
The present Benchmark-Workshop was conceived as the first step in a continuing activity of systematic, critical comparison of existing softwares in the several provinces of dam analysis problems. If, as we hope, it will be successfully carried out, it should be followed by a large number of similar B- W.S, to be held on a regular basis under the auspices of ICOLD, as an international forum in which to produce a completely documented set of test-cases and expected results together with confidence bands.
The publication of these results, possibly as Bulletins, or periodic Reports, of ICOLD, would then constitute a permanent reference source – periodically updated when and if necessary – for everyone who is involved in the critical use of computational tools.