un artículo muy ineteresante que les comparto

The Collapse Of The Ramsgate Walkway
J. C. Chapman

1. Introduction
1.1 General

The life of ship to shore structures is related more closely to the life of a ship - say 20 years - than to the notional 120-year life of a highway bridge. Yet the operating conditions of ship-to-shore structures are exceptionally onerous. The support system must accommodate the tidal range, wave effects, and occasional ship impact. Large horizontal movements must also be accommodated.

Berth 3 had been operating for several years with a single-tier vehicle linkspan supported at the seaward end on a pontoon. Transverse inclination (roll), caused by waves, wind, or eccentric loading, was assumed (in accordance with common practice) to be accommodated by the torsional compliance of the linkspan. The relocation of Belgian ferries to Ramsgate required the provision of a second linkspan above the existing span. The seaward end was supported on a portal frame erected on the pontoon, which had been designed to allow for that possibility. A covered walkway was also required, so that foot passengers could board safely without interrupting the flow of vehicles (Figure 1). The seaward end of the main walkway span was supported on a cantilever platform on the outside of the portal frame. The shore end was supported within an aperture in the passenger access building.

Figure 1a	Figure 1b

1.2 Procurement

The existing pontoon and the vehicle linkspan were designed by an experienced and respected firm of naval architects and consulting engineers. They, and the fabricators and constructors, were both subsidiaries of a Swedish engineering group. The contract was for 'design-and-build' and a similar arrangement was adopted for the extension.

The existing pontoon and linkspan had been certified by Lloyds Register, who also conducted periodic surveys on the structures. It was a condition of the contract that the design, fabrication and construction of the new linkspan and walkway would also be certified by Lloyds Register.

1.3 Design

The walkway consisted of box section trusses. The cladding of the roof, lower sides, and floor, consisted of 6mm steel plate, welded to the trusses and cross-members, so the walkway was very stiff flexurally and torsionally. The shore bearings consisted of steel plates to which low friction pads were attached (Figure 2). The bearings could slide on steel plates fixed to the webs of channel sections, between the upstanding flanges. The side clearance enabled the walkway to rotate about a vertical axis. The bearing was attached by vertical brackets to tubes within which stub axles, projecting transversely from the end of the walkway structure, could rotate.

Figure 2a	Figure 2b

The axles were welded within the central bore of vertical discs, which were site welded to the walkway structure. The seaward bearings were of similar design, but the right-hand bearing had a vertical pin projecting downwards through the horizontal plate of the cantilever platform. Vertical movement was limited by a welded collar (Figure 3).

Figure 3

Thus the articulation system allowed for the three translatory freedoms of the pontoon and two of the rotational freedoms, but there was no allowance for roll, which could be accommodated only by a bearing lifting from its seating, as was observed in service.

2. The Problem
2.1 Collapse

At 00.45 a.m. on 14 September 1994, whilst passengers were boarding a ferry, the seaward end of the main walkway span dropped, without warning, about 10m to the pontoon deck. The violent deceleration on impact caused the death of six passengers and serious injuries to seven others. The walkway had been in service for only 4 months. There were no unusual external circumstances.

The right-hand seaward bearing, including the axle, remained on the platform, being retained by the pin. The welds connecting the axle to the disc had fractured. The left-hand seaward bearing struck the deck and was propelled into the dock, whence it was subsequently recovered. The walkway structure itself remained intact.

As part of the certification procedure, the walkway had been tested under the design loading, without apparent damage. One month before the accident the walkway had been inspected by the insurers' inspector, who had been trained to inspect ship-to-shore structures; he had found nothing amiss.

2.2 The investigation

After the collapse, the bearings and part of the support platform were taken to the laboratory of the Health & Safety Executive for examination. Calculations made by FKAB and by Lloyds Register were also examined, and HSE took statements from witnesses.

The welded connection of the axle of the right-hand seaward bearing to the disc had failed in fatigue, thereby detaching the walkway from the bearing (Figure 4). The connection of the left-hand seaward axle had failed largely in fatigue, and final fracture had occurred when the bearing hit the pontoon deck. Both shore axles had extensive fatigue cracking but were still connected. This confirmed that an incident during construction involving the right-hand seaward bearing was not the cause of failure, because when the incident occurred the shore bearings were not loaded. The lubrication of the axles was found to be deficient, partly through poor design and partly because some automatic lubricating devices had not been fitted. However the shore bearings had retained the original lubrication, and the slideways were also lubricated.

Figure 4a	Figure 4b

The prime cause of failure became apparent from the calculations of the designers and checkers. Both had independently made almost the same conceptual mistake. Instead of calculating the axle moment at the weld connecting the axle to the disc, from the reaction times the lever arm (nominally the distance from the centre of the bearing to the disc), they had assumed the moment to be zero; they had designed on the basis of shear alone. A smaller but still major error was that they had both assumed that all four reactions were equal, notwithstanding the rotational movement of the pontoon and the large torsional stiffness of the structure.

The design concept was inept; in the linkspans, the bearings were placed in the obvious position - under the girders. Because the bearings were not fabricated in the same factory as the structure, drawings for the bearings were separate from those for the structure. There were no assembly drawings, and this could have increased the risk of incorrect design assumptions.

2.3 The Trial

The trial took place in an annexe to the Old Bailey, before a jury. The checkers pleaded guilty, the Swedish designers and fabricators did not plead and did not attend, so the Port of Ramsgate was the only defendant. All parties were charged under the Health and Safety at Work Act, and Ramsgate was also charged under the Docks Regulations. The CDM Regulations were not in force at the time of the accident. All parties were found guilty, and the designers/fabricators, checkers and Ramsgate were fined in the proportions 5 : 2 : 1. HSE's costs were allocated approximately equally.

3. How practice was changed

The risk of competent designers and checkers making the same simple errors might be regarded as so remote that no further action needed to be taken. Although those errors were the cause of the accident, the investigation and trial did suggest that further steps should be taken by the ports industry to protect their passengers and themselves against accidents, and to ensure that they would be better served by their advisers and suppliers. Accordingly, funds were raised to produce a CIRIA Good Practice Guide on procurement, operation, and maintenance of linkspans and walkways. This 230-page document [3], drafted by Posford Duvivier with guidance from a steering group, was published in 1999. Many recommendations were made, one of the most important being that the ports industry should set up a confidential reporting scheme on undesirable events or performance. The reports would be circulated to the industry without attribution. Such a scheme is now in operation through the Port Safety Organisation. Also, Posford Duvivier, with the PSO, have instituted periodic meetings of the industry, where safety matters and new developments are discussed. This is especially beneficial because ship-to-shore structures involve civil, structural, mechanical, and electrical engineering, as well as naval architecture.

The design, procurement and operation of the walkway, the investigation, and the trial are fully described in [1], in which various recommendations are also made. The paper gave rise to lengthy discussion, which is reported in [2].

4. Key references

1. Chapman, J C; Collapse of the Ramsgate Walkway, The Structural Engineer, 76 No.1, 7 January 1998.

2. Collapse of the Ramsgate Walkway discussion, The Structural Engineer, 78, No. 4, 15 February 2000.

3. CIRIA, Safety in ports: Ship-to-shore linkspans and walkways, London, Construction Industry Research & Information Association, Report C518, 1999.

This feature was compiled by John Chapman [F].

14 March 2001

LA INGENIERIA CIVIL
HERRAMIENTA PARA EL DESARROLLO HUMANO