Remerciez-le!

Remerciez @Admin pour avoir partagé cet document gratuitement, de la manière la plus simple, en partageant sur les réseaux sociaux.

White, D. J. (2008). Ge ´otechnique 58, No. 5, 413–421 [doi: 10.1680/geot.2008.

White, D. J. (2008). Ge ´otechnique 58, No. 5, 413–421 [doi: 10.1680/geot.2008.58.5.413] 413 Contributions to Ge ´otechnique 1948–2008: Physical modelling D. J. WHITE* This paper reviews the major contributions to Ge ´otechni- que that relate to physical modelling, including develop- ments in modelling technology, important experimental observations, and the resulting advances in geotechnical engineering. An increasing proportion of the papers published by this journal involve physical modelling, conducted either at 1g or in a geotechnical centrifuge. Over the 60 years since Ge ´otechnique was ﬁrst published, experimental techniques have advanced signiﬁcantly, im- proving the realism of small-scale simulations, and rais- ing the quality and detail of the measurements that can be made. These techniques are reviewed, and some of the consequent advances in relation to foundations, tunnels, retaining walls and slopes are highlighted, as reported in the pages of Ge ´otechnique. KEYWORDS: centrifuge modelling; historical review; model tests La pre ´sente communication passe en revue les principales contributions a ` Ge ´otechnique portant sur une mode ´lisa- tion physique, y compris des de ´veloppements dans la technologie de la mode ´lisation, d’importantes observa- tions expe ´rimentales ainsi que les progre `s re ´sultants en inge ´nierie ge ´otechnique. Une proportion croissante de communications publie ´es dans le pre ´sent journal com- porte une mode ´lisation physique, effectue ´e a ` 1 g ou dans une centrifuge ge ´otechnique. Au ﬁl des 60 anne ´es qui se sont e ´coule ´es depuis le premier nume ´ro de Ge ´otechnique, les techniques expe ´rimentales ont effectue ´ des progre `s conside ´rables, en optimisant le re ´alisme de simulations a ` petite e ´chelle, et en renforc ¸ant la qualite ´ et le de ´tail des mesures qui peuvent e ˆtre effectue ´es. Ces techniques sont passe ´es en revue, en mettant l’accent sur certains des progre `s qui en de ´coulent, sur le plan des fondations, des tunnels, des murs de soute `nement et des pentes et talus, qui ont e ´te ´ reporte ´s dans les pages de Ge ´otechnique. INTRODUCTION Since the birth of Ge ´otechnique, physical modelling has matured as an experimental technique relevant to geotechni- cal engineering. The key milestones of this development are described in the pages of Ge ´otechnique, which has been chosen by many involved in physical modelling as the repository for their best work. In this paper, the major contributions to the development of geotechnical physical modelling are highlighted, and some of the resulting ad- vances in the theory and practice of geotechnical engineer- ing are described. A total of about 200 papers, representing approximately 6% of the Ge ´otechnique archive, are concerned primarily with physical modelling, and many others make reference to this body of work. However, during the ﬁrst 20 years of Ge ´otechnique, from 1948 to 1968, only 10 papers de- scribed physical modelling: that is one every second year, representing 1–2% of the journal. Most of these early contributions describe model tests conducted in large tanks—generally of sand—which aimed to establish the forces on retaining walls and piles. These models were not intended to replicate any particular ﬁeld-scale equivalent; they were aimed at understanding generic modes of behav- iour. In 1970, the Rankine Lecture delivered by Roscoe (1970) included a description of the 5 m radius geotechni- cal centrifuge that had recently been commissioned in Cambridge, UK—a machine described as ‘terrifying’ by de Josselin de Jong, in his vote of thanks. Roscoe showed how the progressive failure of a kaolin slope could be simulated in the centrifuge. Later that year Lyndon & Schoﬁeld (1970) published the results from a similar experiment conducted using the geotechnical centrifuge at UMIST in Manchester, using London Clay. Their post- failure cross-sectional view—drawn with the dimensions multiplied up to the ﬁeld-scale equivalent—is at ﬁrst sight indistinguishable from the many cross-sections of ﬁeld scale slope failures found in the early volumes of this journal (Fig. 1). The challenge set out by Roscoe was that ‘the only satisfactory way of truly modelling to scale a prototype problem, in which the self-weight of the soil is signiﬁcant, is to use a centrifuge.’ Over the following 40 years around 90 papers on centri- fuge modelling have been published in Ge ´otechnique—23 in the past 5 years. Many early developments in centrifuge techniques took place in the UK, in Cambridge and Manche- ster. Ge ´otechnique contains many of the key publications emerging from this work, together with numerous contribu- tions from the international centrifuge modelling community. However, Roscoe’s intermediate clause—‘in which the self-weight of the soil is signiﬁcant’—should not be forgotten. Signiﬁcant contributions to Ge ´otechnique also include physical modelling of geo-environmental processes and small in situ testing tools, which can be simulated in conditions that replicate ﬁeld-scale behaviour without the inconvenience of an inhospitable centrifuge environment. Furthermore, as described later, many important aspects of geotechnical behaviour have been elucidated through small-scale model tests conducted at 1g—taking advantage of the easier control of events compared with the centri- fuge. Two key developments have advanced the art of geo- technical physical modelling over the past 50 years. The development of the centrifuge in the 1970s allowed the realism of physical modelling to be enhanced, through the correct modelling of self-weight stresses. The subsequent development of miniaturised electronics and microcompu- ters has led to enhanced methods of data acquisition, control, and image analysis. The reﬁnement of these techniques continues to yield dramatic improvements in the utility of physical modelling. More realistic simulations can be conducted, and more detailed observations can be gathered. Discussion on this paper closes on 1 December 2008, for further details see p. ii. * Centre for Offshore Foundation Systems, University of Western Australia. Downloaded by [ Universidad De Chile] on [24/11/16]. Copyright © ICE Publishing, all rights reserved. PHYSICAL MODELLING TECHNIQUES Geotechnical centrifuge development Approximately half of the 200 physical modelling con- tributions to Ge ´otechnique make use of a centrifuge in order to ensure that the stress levels in the model are comparable to ﬁeld-scale conditions. The majority of these papers de- scribe research conducted in Cambridge or Manchester, in the groups led by Professor Andrew Schoﬁeld and Professor Peter Rowe respectively. Schoﬁeld and Rowe pioneered the use of the geotechnical centrifuge in Europe, in parallel with developments in Japan and following earlier work in the USSR, which was at that time unknown in the West. The idea of using a centrifuge to correctly model civil engineer- ing structures in which self-weight forces are signiﬁcant can be traced back to the French engineer E ´douard Phillips in the nineteenth century, as described in Ge ´otechnique by Craig (1989). The earliest mention of centrifuge modelling in the pages of Ge ´otechnique is the ﬁnal section of Roscoe’s Rankine Lecture, delivered in 1970. Despite leading a research group focused on the development of theoretical models for soil behaviour, he argued boldly that ‘with the centrifuge it is possible to obtain answers immediately to full-scale pro- blems without having to appeal to, or wait for the develop- ment of, any theory’ (Roscoe, 1970). In a letter to Ge ´otechnique, Golder (1971) relates a more light-hearted attempt to test soil using centrifugal force, which was conducted on the lawn outside the UK Building Research Establishment in 1936. Rowe’s Rankine Lecture, given in 1972, was concerned with the identiﬁcation of soil fabric during site investiga- tions, but concluded with a description of the second geotechnical centrifuge built in Manchester—at the (then) Victoria University of Manchester (Rowe, 1972). Unlike most physical modelling, Rowe’s work relating to site- speciﬁc situations frequently involved using intact samples of natural soil, which were built into models placed within the centrifuge. This approach followed rationally from his conclusion that strength and consolidation testing of natural soil elements in the laboratory should be conducted in cells sufﬁciently large to accommodate representative amounts of the natural fabric—leading to the consolidation device now known as the ‘Rowe cell’. Applying this logic to the centrifuge and his particular interest in earth embankment dams led him to design a machine sufﬁciently large to accommodate soil models that are 1 m 3 2 m in plan. Some of Rowe’s most signiﬁcant centrifuge work, con- ducted with Craig, contributed to the development of the large gravity platforms deployed in the North Sea in the 1970s and the Oosterschelde storm surge barrier. These studies inﬂuenced the ﬁnal form of these structures, and provided the necessary performance data to support the design (Smith, 1997; Craig, 2002). Ten years after Roscoe’s Rankine Lecture, his successor, Schoﬁeld, delivered the Rankine Lecture (Schoﬁeld, 1980), focusing on centrifuge operations in Cambridge. Twenty-six years later, in 2006, Professor Robert Mair—Schoﬁeld’s successor—delivered the Rankine Lecture (Mair, 2008), and also described extensive centrifuge modelling studies con- ducted in Cambridge. Working alongside Schoﬁeld and later Mair, Professor Malcolm Bolton, another strong proponent of centrifuge modelling, has made more than 40 contribu- tions to Ge ´otechnique, many of which are concerned with centrifuge modelling. The research conducted by the groups in Cambridge and Manchester, and uploads/Litterature/ 2008-white-contributions-to-geotechnique-1948-2008-physical-modelling.pdf