DUE TO the large spans and high intensity of ground loading, the concrete chamber needed to be reinforced. Steel reinforcement was unsuitable as outlined in the previous portion of the paper. As a result, the preferred design option for reinforcement was glass reinforced plastic (GRP) bars. GRP when excavated by the TBM disintegrates into strands, which can be handled by the slurry treatment plant and TBM slurry pumps. Furthermore, it has a comparable ultimate yield strength to traditional steel reinforcement, so it would be able to deal with the high bending and shear stresses imparted on it by the conservative ground loading model.

GRP is not a widely used reinforcing material, especially in the UK due to its limited ductility and had to be procured from Sireg of Italy. Sireg was the only manufacturer in Europe that had the manufacturing experience and associated test data to validate its quoted design capacities. Detailed in the same basic way as steel, the GRP could not be conventionally bent to small radii and all bending had to be completed in the factory at the point of manufacture. This created an additional problem regarding the delivery on the design. The GRP reinforcement scheduling had to be right the first time as it required a long lead time and could not be manipulated in the formwork on site once delivered. As GRP is not a widely-used structural reinforcing material no codes of practice cover its use under either Eurocodes or British Standards. The only useful guidance available is produced by the Institution of Structural Engineers (IStructE). The guidance forbids capacity derived from plastic behaviour and states that only elastic behaviour can be relied upon. This is due to the brittle failure nature of the GRP as it does not slowly yield in a ductile fashion as steel does. Therefore when structurally designing using this material the author had to derive from first principles all of the elastic stress blocks for the walls, roof and floor sections, and use an unusually high factor of safety for material (3.6) to account for the brittle failure risk. The guidance discusses the possibility that with more testing and research this material factor is likely to be significantly reduced. The insitu behaviour of this structure potentially offers data that could be used to further advance the confidence in this country in using GRP as a serious alternative to traditional steel reinforcement.

To compare to the structural design values above, the author calculated the structural capacities by hand from first principles. Modelling first elastic behaviour in stress blocks and then undertaking a staged M-N (moment & axial force) assessment to calculate the combined axial-moment and shear capacities for each of the structural members, to determine the most efficient GRP reinforcement layout. The GRP reinforcement became very congested due to a number of key factors:

  • High material factors of safety for the tensile GRP reinforcing bar, leading to a design tensile strength of 235N/mm2 compared to 500N/mm2 for steel
  • Higher material factors of safety for the shear GRP reinforcing bar, leading to a design shear strength of 55.6N/ mm2 compared to 250N/mm2 for steel
  • Restrictions on bend radii of the GRP bars leading to more laps than would be normally expected
  • Extra reinforcement required for the lifting condition The GRP bar, even though expensive and difficult to detail due to bending and placement restrictions, does have one distinct advantage over steel. It is not affected by corrosion and is therefore tremendously durable, although this is not of any relevance to this application as the chamber is sacrificial and only in place awaiting the TBM’s arrival for a relatively short period. In this case it is only the TBM’s ability to cut the GRP and not traditional steel reinforcement that means it is the primary choice for reinforcing the concrete chamber.

Casting and placement of the Chamber

The concrete could be easily cast on site, using the existing skills of the site team that had just finished working on the reinforcement for the connection shaft base slab. A C50/60 concrete mix was selected for the box culvert although concrete strength was more critical for the placement of the chamber at pit bottom than for its structural performance, which will be discussed in more detail below. The chamber was to be cast in two pieces; the main twin bore culvert (chamber) and a back panel. The front of the chamber was to be left open to allow for working space to facilitate tool changing on the TBM. Casting of the box culvert took place next to the connection shaft, within ‘one lift’ reach of the connection shaft into which the chamber was to be positioned. The culvert and back panel were both cast on their backs, to minimise formwork and falsework. This simplified the construction element, but complicated the design work following casting, as it would be necessary to rotate both pieces into their installation orientations for placement in the shaft. In order to rotate the culvert from its horizontal casting position to the vertical the author had to design bespoke steel trunnion lifting points, these used in conjunction with two cranes, allowed a tandem (rotation) lift to be carried out.

Lifting analysis, design and testing

The structural design of the chamber had been modelled for ground loading with the culvert in the vertical condition. The lifting condition of top and tailing the chamber introduced a complex varying load case that had to be assessed and checked against the structural capacities of the culvert. To do this the author modelled the chamber being lifted on lifting points at four different rotation angles using SAP 2000 another finite element package in order to generate the bending moments and shear forces the chamber would experience during the lift. The lift angles modelled were 0, 30, 60 and 90 degrees. When compared to the structural capacity of the system the lift condition necessitated further shear reinforcement being added in the base section but was less than the ground load condition for all of the other structural elements.

The trunnion pin design was limited by the spacing of the vertical reinforcement in the culvert wall sections. This limited the diameter of the lifting pin to 100mm. The reinforcement could not be locally cut to accommodate the lifting pins as there was very little redundancy in the system due to the aforementioned high reinforcement material safety factor and the concrete elastic behaviour criterion. The design of the trunnions was undertaken to a factor of safety 2.0 in line with current industry lifting regulations and best practice. This is in addition to the BS5950 imposed load factor of 1.6. As a result of the weight of the culvert being 22.8t (un-factored) this pushed the grade of steel required for the trunnions to S460. The high weight of the chamber determined that the 100mm diameter trunnion pins had to be grouted into their 110mm diameter seating holes, which were formed using plastic piping. The surrounding encapsulation grout was required to ensure that the very high local stress distributions could not build up due to the trunnions effectively acting as restrained cantilevers.

The back slab was cast with two holes to facilitate the rotation from its horizontal casting position to the vertical position using strops passed through the holes. This arrangement when combined with recycled tyres placed under the pivoting edge of the slab (to mobilise enough friction to stop it sliding whilst being rotated) allowed a single crane to lift the back slab into the vertical position and place it in the shaft. Once in the shaft the back slab was lined up with the culvert and the secured in position with four concrete key pins. The purpose of these pins was to ensure that as the TBM approached and the torque of the machine imparted a lateral load into the ground, the back slab would not move relative to the culvert. They, like everything else in the structure could not contain steel so they consisted of 90mm diameter concrete cylinders formed from the same mix as the remainder of the structure, inserted into 110mm diameter holes on the culvert.

Once placed on the shaft base, the chamber was filled with Thanet Sand. This was undertaken to aid short-term stability whilst filling and compaction was undertaken around the outside of the structure. Unlike the fill material everywhere else in the shaft this was plain Thanet Sand that could be easily hand excavated by the miners when they first encounter the chamber, after having entered the chamber through the TBM head intervention doors. The CBGM, as it’s bound together, would have taken too long to remove from the chamber and would have added impact on the tunnelling programme.

The trunnions were grouted, which made removing them difficult after the placing of the culvert in the shaft. They had to be removed but removal had to be done very carefully so as not to damage the critical GRP reinforcement located at close vertical spacing in the outer walls. The site team cut the outside plates off the trunnions then stitch drilled around them so that they could be pulled free from inside the chamber.

To verify the lifting system, testing had to be carried out. This was completed to satisfy the LOLER and PUWER regulations as well as MVB’s own lifting guidance in line with accepted best practice. The testing was subdivided into:

  • The use of certified (coded) welders to weld the end plates on the trunnion pins.
  • Independent non-destructive testing of critical welds on the trunnion lifting pins, in the form of dye penetration testing.
  • Inspection and sign off by the designer (myself) of the chamber and slab before the lift.
  • Production of a Declaration of Conformity for the Trunnion pins as they are classified under the European Directive 2004/42/EC for lifting equipment, LOLER and PUWER.
  • A ten-minute test lift on the chamber once rotated into its vertical position 100mm off ground.

Conclusions

The crossing of the shaft and the changing of the TBM tools from within the box culvert went very well. The box culvert concept achieved all of its objectives. The TBM safely transited the connection shaft and the nose cone and cutting equipment were changed in a safe environment provided by the head intervention chamber. A significant time saving on the programme was achieved and valuable experienced gained on the crossing of shafts affected by contamination. The author learned a great deal from the design and sequencing of the works and is proud of how the system and the GRP reinforced intervention chamber in particular performed. In hindsight, the only item the author would have changed was the pipework arrangement, which became blocked, but this thankfully did not significantly affect the crossing, only slowing the initial shaft emptying.