The story of recessed and back-loading cutters, which is integrated with the evolutionary development of Mixed Ground TBMs, commenced some forty years ago. It began with the California State Water Project in the U.S.A. then moved to a project in Melbourne, Australia, to Sila in Italy and from there to the Buckskin Mts Project, U.S.A. From Buckskin to Russia, finally ending with the Selby Project in the U.K.
Diverse ground on the California State Water Project tunnels, influenced the development of shield excavators which had some degree of flexibility built into their basic design. This enabled them to cope more readily with anomalous terrain.
In the same way, in their turn, did the Melbourne & Metropolitan Board of Works south-eastern trunk sewer tunnels in Melbourne, Australia, inspire the design of the first hard rock hybrid machine. However, there was one important difference; the M.M.B.W. had, to some extent, a forewarning of the type of country which would be met along the way. Test holes indicated, with some accuracy, the various types of ground conditions which could be expected. They showed that water-bearing Tertiary sandy sediment over Silurian bedrock lay in the Kew to Moorabbin tunnel route, while mud and sandstone strata was expected in other sections. There were also indications of badly faulted and fractured siltstone, blocky sandstone and plastic clay. Lightly fractured sections of fairly good cohesive material were interspersed with areas of soft running ground and bands of heavily fissured rock combined with soft sticky clay. To say the least, a very discouraging picture.
In addition to the provision of a canopy or shield, mounted over the cutterhead and extending as far as the tunnel spring line, the Robbins machine used for this job was fitted with a domed boring head. This latter component, with its protruding cutters was unsatisfactory when blocky material was encountered. Instead of crushing and removing the spoil as they had been designed to, the cutters tended to grab the material in chunks and pull it across the face. The net result was that unstable ground ahead of the TBM was disturbed in much the same manner as had occurred in the first section of the California State Water Project’s Newhall Tunnel.
Taking note of suggestions put forward by the Board’s staff – in particular Frank G. Watson (Workshops and Plant Services Engineer) and Alan J. Neyland (Project Engineer), the problem was finally solved by the replacement of the domed head, sliding support and steer shoes at the front of the machine, with a new revolving spoked-wheel type cutterhead, and a full circle flexible slotted shield which rapped completely around the cutterhead support. This important component, which was manufactured in Australia, later came to be known as the ‘Melbourne-type’ head.
Later, when a TBM was required for the Orichella and Timpagrande tunnels in Sila, Southern Italy, with ground conditions ranging from 69 to 207MPa, it was realized that two types of TBM would be needed – namely, a mechanized shield for those sections where the ground was unstable, and a machine with side wall grippers for those areas where the rock fell into the higher compressive strength range.
Obviously the purchase of two TBMs would be costly. To meet the situation Carlo Grandori, at that time, Managing Director of the Italian company S.E.L.I., suggested that the Robbins Company should design and build them a special prototype TBM embodying, in one unit, as many of the essential characteristics of the two types of machine as was practicable.
Following the Melbourne and Metropolitan Board of Works’ experiences in Australia, Carlo Grandori decided to equip the Italian machine with a ‘Melbourne-type’ cutterhead, to enable it to cope with the anticipated sections of fractured and blocky strata which lay ahead. It featured an open six-spoked wheel-type cutterhead fitted with one tri-disc and 28 disc cutters.
The expression ‘one man’s meat is another man’s poison’ might be aptly applied to the two Projects; i.e. the one in Melbourne and in Italy. In both cases badly fractured and blocky ground was anticipated and encountered. However, in Melbourne, this was mostly Silurian mudstone and sandstone, whereas in Italy it comprised Sila granite – a much harder substance, and perhaps therein lay the difference.
Grandori reported that hard boulders were found embedded in a softer matrix in some sections of the badly fractured Sila Granite formations. Large pieces of rock frequently fell from the face in these areas. They passed through between the spokes and jammed themselves firmly behind the cutter arms, where they caused extensive damage to the muck loading and conveying system. When this occurred the wedged rock had to be manually broken and removed, a costly and time-consuming operation.
The situation was tolerated until the machine entered a particularly bad area and a major face collapse jammed the cutterhead preventing further progress. This induced Grandori and his team to remove the cutterhead and effect suitable modifications to it which, hopefully, would enable the machine to proceed more efficiently. Basically the modifications consisted of closing up the large gaps, which existed between the cutting arms, with heavily welded steel bars arranged concentrically between the spokes. These formed a strong grille or screen which prevented rock fragments measuring more than 8 inches in diameter from passing through into the muck-handling and conveying equipment. The completed grille-work was approximately an inch from the tunnel face which, in Grandori’s opinion, was the key to the modification’s success, as this distance represented the average depth of rock spall produced by the disc cutters.
The development of the Robbins TBM for the Buckskin water tunnel in Arizona came as a direct result of Grandori’s experiences with the tandem shield in Italy. The Buckskin machine was a 7.16m diameter completely shielded rock boring machine consisting of two shields which telescoped together for the boring stroke and yet were flexible enough to permit normal steering control.
According to Eugene G. Murphy (Manager for the Buckskin Mts. Project) the tunnel started in well-cemented Andesite rock formation which varied from 2m cube blocks to crushed material, visicular Andesite, agglomerate and tuff. However, as the tunnel progressed, the calcite and gypsum cementation between the blocks was replaced by clay, or the interfaces became completely devoid of any adhesive material. At the 460m mark the overbreak was seen as some 6m high.
The collapse of large blocks caused structural damage to the machine, constantly plugged the muck hopper and affected the general stability of the TBM.
As Murphy so aptly described it, the TBM was, in effect, acting as a giant horizontal blender, no work being performed by the cutters other than the mixing action. Although cutter wear was minimal, the cutters were nevertheless failing structurally due to their impact with blocks of rock. To overcome these problems the cutterhead face was advanced a distance of a foot and the consequent gap between the new plate and the original cutterhead was filled with a cement grout, so that the thrust against the new cutterhead could be handled.
A ring made from 24 inch diameter pipe, with inch thick walls was welded around each cutter, the edge of the pipe being trimmed to fit it to the spherical shape of the cutterhead. These and other modifications including circumferential rings that were welded to the cutterhead between the outer gauge cutters and the inner circle of cutters, and the provision of block-outs around the gauge cutters and muck buckets all formed part of the secondary breasting skin arrangement tried in the Buckskin tunnel.
Although the problem of overbreak and blocking appeared to have been solved, a new difficulty became apparent. The shield skin plate designed with a horizontal joint 3ft below the spring line, was now being deflected by the blocky ground at the crown and sides. This deflection was severe enough to hinder the erection of segments within the tail shield. So, once again, the machine was stopped to enable five stiffener beams to be installed in the tail shield skin plate, while the plate itself was increased in thickness from 1 to 1½ inches. Blocky, soft and various grades of ground in between were then successfully handled by the TBM.
As a result of the experiences gained on the Buckskin Project the cutterhead of the Robbins fully shielded hard-rock machine destined for work in Russia during 1979-82 featured cutters protected by breasting plates and rings. But, later, when the Robbins’ Selby machine was commissioned, it was note-worthy because it was the first full-face TBM with recessed cutters and was also the first to use rear change or back loading cutters.
Having moved from projects in Melbourne, Sila, and Buckskin to ones in Russia and the U.K. as we traced the evolutionary development of hybrid machines, we now return to the S.E.L.I. company and the DSU-TBM produced by them in 2001 to cope with a wider range of ground conditions than could be handled by conventional single-mode TBMs. Its basic features included the ability to adjust the excavation diameter of the cutterhead, cope with squeezing ground, treat the ground ahead through the face and operate in single or double shield mode. That unit led to the development of the Seli-DSU/EPB TBM.
The most significant component of this latter unit was its cutterhead, which was designed to be converted from DSU to EPB and vice versa, while in the tunnel. It consisted of bolted plates fitted between the cross arms, thus closing the gaps separating them. This allowed the machine to operate in rock-cutting mode. Recessed cutters were mounted on the cross arms. When the machine was operated in the EPB mode the plates between the cross arms were removed. The gaps thus created then permitted the ingress of material to the drum-chamber, situated immediately behind the cutterhead arms. In this mode an auger or screw conveyor was fitted to the invert for the removal of the compressed material in the drum chamber, in much the same manner as a conventional EPBM would operate. Where coarse granular material was encountered the machine could be converted into a Slurry unit by injecting fines into the hyperbaric chamber, thus further extending the use of this type of TBM.
Recently a 3.1m diameter Robbins unit with the capability of conversion from soft ground EPBM to medium/hard rock TBM was constructed for Wastewater Tunnels in Istanbul. The machine used disc cutters and a conventional belt conveyor for medium to hard rock strata or tungsten-carbide drag bits and a screw conveyor for soft ground.
And finally when the Swedish/French JV of Skanska/Vinci and Herrenknecht AG were confronted with the very difficult mixed ground terrain prevailing beneath the Hallandsås ridge, they designed a special TBM to handle these complex strata. Dubbed the ‘Asa’ the unit was capable of operating in either the open mode as a hard rock machine with a conventional TBM spoil conveyor or in the closed mode as a Mix-Shield with slurry removal to a surface treatment plant. The Mixshield evolved from the ‘Hydroshield’ and ‘Hydrojet Shield’ slurry machines introduced by Wayss and Freytag Ag during the early 1970’s. This ‘convertible’ unit was developed in conjunction with Herrenknecht GmbH (who was the Licence Holder of the Wayss and Freytag Mixshield’ patent). Its main purpose was to be able to rapidly convert the machine, in long runs, to varying modes so it could handle a wide variety of soil conditions. These transformations needed to be rapid and, if possible, undertaken underground within the narrow confines of the tunnel itself. It was envisaged that such a unit should be capable of being changed to the slurry, earth pressure, or dry mode of tunnelling, with or without compressed air.
Robbins ‘Grandori’ tandem mechanized shield The Robbins Melbourne machine as supplied Depicting part of the secondary breasting skin arrangement tried in the Robbins Buckskin TBM The Herrenknecht TBM for the Hallandsas Project
