ABSTRACT:

The paper deals with the specific failure potentials in tunnelling and underground work. By reviewing the experiences gained by excavating tunnels especially in difficult ground, it can be shown that most of the actual structural and functional failures are not covered by the usual engineering approach of analysing structural models. Therefore, engineers should develop rational means in addition and beyond those of numerical analyses. There is an urgent need for improved methods firstly to identify possible failures and secondly to apply adaquate risk assessments to these potential failures.

RÉSUMÉ:

L'article traite de possibilites de dommage specifiques pour des construction souterraines. Gràce aux experiences faites pendant des creusements, particulièrement dans les terrains difficiles, on peut montrer que la plupart des cas de dommage reels ne sont pas inclus dans les analyses conventionnelles des structures. A cóte des analyses numeriques il faut donc developper supplementaires. Il y a un besoin urgent des methodes ameliorees, qui donnent les moyens pour identifier les cas de dommage possibles et en plus qui fournissent des critères d'une estimation des risques dans ces cas.

ZUSAMMENFASSUNG:

Der Aufsatz befaβt sich mit den fuer den Bau unterirdischer Hohlraume spezifischen Schadensmöglichkeiten. Auf Grund der Erfahrungen, die beim Vortrieb, insbesondere bei schlechtem Baugrund, gewonnen wurden, kann gezeigt werden, daβ die meisten der realen Schadensfalle nicht durch die ueblichen Nachweisverfahren von Ingenieurbauwerken erfaβt werden. Daher sollten Ingenieure auch methodische Wege entwickeln, die ueber die numerischen Berechnungen hinausgehen und sie erganzen. Es besteht ein dringender Bedarf fuer verbesserte Methoden, die erstens mögliche Schadensfalle identifizieren und die zweitens fuer diese Falle Risiko-Assessments bereitstellen.

1 INTRODUCTION:

We engineers are promising, well ahead before the structure or design is exposed to an actual test, that by applying our methods of design, analysis, and construction our engineering structure will fulfill its functional purpose and will neither fail nor causing damage. The design and the excavation of an underground work should be safe and economical during the projected life time of the structure as well as in all its construction phases. Engineering design, therefore, is concerned with avoiding failures, running no risks. Yet, running no risk at all is bad engineering, because of the higher costs and the missing innovations. Progress and risk-taking are inseparable. Moreover, excavating a tunnel in a hitherto unknown ground is always an innovation in the sense of constructing a prototype, mastering ever new situations. Good engineering, therefore, implies taking risks, yet knowing the potential hazards, and preventing failures nevertheless. By reviewing our experiences and pondering upon the actual engineering work of a specific tunnelling project, some questions may arise: What is our approach for finding all possible hazards and failure potentials? Rather intuitively: a sudden idea in a sleepless night? Or do we provide sufficient time for a systematic screening of the entire work to be done to find all hazards? Do we take - or have - time enough for exploring all the possible harzards, for weighing their damage potentials and putting them in an order of priority? Experiences prove that overlooking, just simply being unaware of a dangerous situation is the most critical case. In general we apply safety analyses only for those situations for which we have developed a tool for expressing safety by numbers. The design methods as given by codes and standards cover only structural failures, even not all the possible ones. Yet, what is our risk assessment approach for those hazards which defy any analytical evaluation? These questions can be raised, of course, for all kinds of engineering structures. However, they are especially relevant for tunnelling and underground works.

This content is only available via PDF.
You can access this article if you purchase or spend a download.