M Ogle, G Bowden

BS EN 1090-2:2008 AND PD 6705-2:2010

Martin Ogle, consultant to TWI, Cambridge, UK Geoffrey Bowden, Highways Agency, UK

Abstract The paper gives a brief summary of the development of UK specifications for materials and workmanship for steel bridges including BS EN 1090-2 which represents a major change in UK practice. One of the main features of the new specification is the extensive list of choice items and additional information on which the specifier needs to make technical decisions. To assist in this process BSI is in the process of publishing PD 6705-2 which is described briefly in the paper.

Background to Development of UK Material and Workmanship Specifications for Bridges

British engineering has been at the forefront of the development of iron and steel bridge construction for nearly two centuries and there are many thousands of examples all over the world still doing good service after more than a century. This is a testament to both the design and the construction standards on which these bridges were based. The first British Standard for steel girder bridges BS 153 was published in 1922 in two volumes containing five parts, namely: Part 1 Materials and Part 2 Workmanship, in the first volume[1] and Part 3 Stresses, Part 4 Details of construction and Part 5 Erection in the second volume[2]. These standards covered riveted and bolted girders made with mild steel. Revisions of these two volumes were made in 1933 and 1937 respectively. The most extensive amendments were made in the 1950s with the regrouping of the Parts and the inclusion of specifications for railway and highway loading. The Parts were re-grouped into three volumes; the first containing Part 1 Materials and workmanship and Part 2 - weighing, shipping and erection (1958)[3], the second containing Part 3A loads (1954)[4] and the third containing Part 3B stresses and Part 4 design and construction (1958)[5]. Thus all specifications for requirements for fabrication and erection (Parts 1 and 2) were contained in one volume. Part 4 covered design of constructional details and not manufacturing aspects. The extent of interest in this standard at that time can be judged by the fact that 45 Government departments and scientific and industrial organisations were involved in its supervision and/or preparation. The main new features of the 1958 version of Parts 1 and 2 and subsequent revisions in the 1960s were clauses dealing with high strength friction grip bolts and arc welding. These made extensive use of the many new British Standards which had been published for these methods of construction in the 1950s and 1960s.

M Ogle, G Bowden

In the early 1970s the buckling failures of various box girder bridges highlighted the limitations of the BS 153 design and specification Parts, where used for this new form of construction, as opposed to plate girders (for which BS 153 was written). The resulting extensive research carried out in the early 1970s revealed the importance of welding residual stresses and geometrical imperfections in affecting buckling resistance of panels and stiffeners. The scope of BS 153 was not extended to cover box girders. Instead new execution requirements were published in the Interim Design and Workmanship requirements for the box girders which arose from the Merrison Committee inquiry and its subsequent recommendations. At the end of the 1970s work on redrafting the new limit state bridge design code BS 5400 was completed. This took into account all the recent work on box girder design and manufacture and included that form of construction. In 1980, BS 153 Parts 1 and 2 were superseded by BS 5400-6[6] which included references to important new welding standards, BS 5135[7] (welding practice) and BS 4870-1[8] (weld procedure approval). BS 5135 contained extensive rules for the avoidance of hydrogen cracking, which was highlighted in the catastrophic failure of the Kings Bridge plate girders in 1962. BS 5400-6 also included detailed requirements for geometrical tolerances. In the 1980s concerns were expressed by some fabricators over difficulties experienced in meeting the strict weld quality standards in BS 4870 consistently, which were also imposed for normal production welds. The BS bridge committee set up an investigation into the matter and after a lengthy examination of the fracture mechanics implications and the detection and measurement capability of commercial non-destructive testing (NDT) techniques, ultrasonic testing (UT) in particular, it was agreed that relaxations from the quality required for procedure approval could be justified for production using fitness-for-purpose (FFP) principles, particularly where fatigue stressing was not high. On the other hand it was concluded that there were difficulties of assuring in practice by non destructive means that the quality of workmanship in production, of welding in particular, was adequate to meet the stringent requirements needed when fatigue stressing was high. In 1999 BS 5400-6 and BS 5400-10[6] were revised to take account of these findings. BS 5400-6 included three levels of weld quality depending on the degree of fatigue stressing. BS 5400-10 provided designers with a method for assessing which quality was necessary, (in terms of the fatigue stressing), and also amended the detail classification tables to warn when the higher classifications might not be reliable unless special attention was paid to NDT procedures. The results of this work were used to produce FFP based weld quality standards for other applications. This included BS 5950-2 for buildings[9] and ISO 10721 - the new International Standard for structural steelwork, Part 1 covering materials and design and Part 2 covering fabrication and erection[10]. BS 5400-6 was withdrawn by BSI in April 2010 in line with CEN requirements and is replaced by the new European based standard BS EN 1090-2. This ended nearly 90 years of development of standards reflecting best British practice for steel bridge manufacturing.

M Ogle, G Bowden

Background to Development of European Execution Requirements for Bridges

The CEN technical committee responsible for development of the Eurocodes for structures design CEN/TC250 was not responsible for the development of the associated specifications for materials and workmanship (now referred to as Technical requirements for execution in Euro-English parlance). This separation of responsibilities represents a major difference from British practice. When the draft Eurocode for steel structures DD ENV 1993-1-1[11] was published in 1992 reference was made to such European product standards as had already been published by CEN. The standard for Execution of steel structures was at that time without an EN number. A different CEN Committee, CEN/TC135 was set up in the late 1980s to draft the execution standard for structural steelwork, which was eventually published by CEN as ENV 1090[12] in six parts, starting with Part 1 in 1996 which gave general rules and rules for buildings. The five other parts were published over the next four years including Part 5 which gave supplementary rules for the execution of bridges. The UK was well represented on the CEN working groups preparing these different parts. The ENVs were subject to two year Enquiry Periods after which National Standards Bodies submitted comments. On receipt of the comments CEN/TC135 decided to merge all six parts into one common part, which became EN 1090-2[13] under one working group. Two other working groups were set up to produce a new Part 1[14] on certification of conformity and a new Part 3[15] covering the execution of aluminium structures. The UK was active in all three working groups. Thus the requirements for the execution of steel bridges were no longer separated from other structural steel applications as they were in the British Standards. Throughout the whole process of developing EN 1090-2 it was clear that there were fundamental differences in practice across Europe and that these could only be resolved by introducing a wide selection of choice items and leaving many gaps to be filled in by the user. A further factor affecting the development of the standard was that there was no funded Project Team for the drafting work, as there had been with all the Eurocode parts. These factors inevitably had a delaying effect on the programme, with the result that the final Enquiry draft pr EN 1090-2[16] was only put out for comment in 2005 by which time many of the key parts of Eurocode 3 had been finalised. This was in spite of the fact that EN 1993-11[17] made it clear that the validity of the design rules were dependent on execution of the work being in accordance with EN 1090. EN 1090-2 was essential for defining the necessary quality and testing standards which would have a significant effect on the reliability of the design resistance values (as determined by the various M values which had already been prescribed by each member state in their National Annexes to EN 1993). A further novel feature which was introduced into pr EN 1090-2 and which did not appear in ENV 1090, was that of Execution Class. This was a device to enable flexibility to be introduced, due to the wide range of potential applications. This effectively created a set of widely differing standards within the document which could be selected by the user either for the whole structure or different parts of it. Informative advice on the selection of execution

M Ogle, G Bowden

class was included, involving a matrix of different parameters which included three levels of consequence class (CC), which in BS EN 1990[18] are linked to three classes of reliability (RC), each with its own target reliability index. A further parameter was the new Service Category (SC) which was only defined in qualitative terms, (eg whether or not the structure or components were deemed to be susceptible to fatigue). The BS Bridge Committee was very active in preparing comments on pr EN 1090-2, which included fundamental concerns about the new Execution Class matrix and its consistency with the reliability principles in BS EN 1990. They also prepared proposals for clauses to address their concerns and to improve the compatibility with BS EN 1990. The comments included proposals for specific testing and fitness-for-purpose assessment of welds, based on principles used BS 5400-6 and BS 5950-2. These comments were considered by the BS Committee responsible for BS 5950-2 to be relevant to steel structures in general and were included as such with all the UK Enquiry comments. Whilst many of the UK comments were accepted by CEN/TC135, the UKs fundamental concerns about the use of Execution Class and the need to have specific fitness-for-purpose testing requirements were not, and no reasons for their rejection were provided. This was a major contributory factor leading to the UK casting a negative vote to the final version of pr EN 1090-2. Sweden and Finland, who had also been very active in the CEN working group also voted negatively, but the final weighted vote was still sufficient to approve it for publication. The resulting BS EN 1090-2:2008[19] now supersedes BS 5400-6 and BS 5950-2. It represents a substantial change in specification of materials and workmanship for steel structures in general not only bridges. The new standard is considerably longer than its predecessors and includes many more clauses requiring user decision as can be seen from the table below.

Reference no. Specification Version, year Number of pages Size paper version Clauses requiring User decision Normative references Normative content

BS 153: Parts 1 and 2 1958 25 A5 20 25 100%

BS 5400-6 1999 38 A4 39 65 90%

BS EN 1090-2 2008 203 A4 200 177 95%

Table 1. Developments in scale of British and European based materials and workmanship specifications for steel bridges over the last half century

The main benefits of BS EN 1090-2 over its predecessors are as follows:

It is up to date with all the new BS EN reference standards which have emerged over the last 20 years. It addresses higher strength steels, cold rolled hollow sections and cables.

M Ogle, G Bowden

It addresses laser and plasma cutting. It includes qualification requirements for welding co-ordinators, welders and weld inspectors. It addresses quality management systems, including documentation, plans etc which lays the basis for approval of factory production control (FPC) systems of fabricators. It includes a wider range of fastener types and methods of tightening of preloaded fasteners. It covers a wider range of geometrical tolerances. It covers surface treatment for corrosion protection.

The main disadvantages of BS EN 1090-2 as far as the specifier and/or purchaser are concerned are as follows:

It does not set a clear cut standard, as its predecessors did. It contains a range of standards. Selection of an inappropriate standard by a user can result in a bridge whose structural integrity could either be seriously at risk or could be unnecessarily expensive to produce. The informative guidance on selection of Execution Class is based on a premise that the UK objected to on grounds that it was technically unsound (as discussed above). The user frequently has to supply additional information for which he/she may not have sufficient experience to undertake with confidence. Control of weld quality relies entirely on one of three choices of FPC systems. The extent and methods of weld testing are very loosely defined and not specific to each bridge. The acceptance levels are arbitrary and are not based on FFP criteria. These are major differences from BS 5400-6:1999. As a consequence the BS Bridge Committee agreed to develop a published document (PD) to assist specifiers, employing BS EN 1090-2 for steel bridge work, in making sound choices when selecting their technical requirements. This document has been approved for publication by BSI and is described below.

PD 6705-2 Recommendations for the Execution of Steel Bridges to BS EN 1090-2

PD 6705-2[20] was prepared by the working group of the BS Bridge Committee responsible for maintaining BS 5400-6 and its successors. The working group has been closely involved with the developments of EN 1090 over a period of 20 years and has included design experts closely involved with EN 1993, bridge procurement agencies, and steel fabricators. PD 6705-2 has therefore benefitted from an intimate knowledge of the main technical aspects in both BS 5400-6 and BS EN 1090-2, including the fundamental issues raised by the UK at the 2005 Enquiry stage. The prime objective of PD 6705-2 is to provide users with sufficient technical advice that they can make appropriate choices and supply essential additional information. It is not intended to supply ready-made specification clauses for immediate verbatim addition to a contract specification, although some of the material may be transferred across with appropriate editorial adjustment.

M Ogle, G Bowden

PD 6705-2 only addresses key technical user decision issues which have a significant effect on the structural integrity of steel bridges. This applies primarily to the resistance of members and joints to buckling, fracture and fatigue failure, and is intended to ensure that the controls on quality are consistent with the M values selected by the UK in BS EN 1993 (including the various PDs referenced in the National Annexes). PD 6705-2, as a consequence, addresses approximately a third of the total 200 user decision items in BS EN 1090-2, (some of which do not relate to bridges, others do not have a significant effect on structural integrity and others do not need additional information). The main features of PD 6705-2 are:

It is intended to cover all types of bridge from the smallest and simplest footbridge to the most sophisticated and complex structure in high tensile steel with highly dynamic loading. It is not intended to cover contractual matters, only technical ones. The user need only address those recommendations which are relevant to the application(s) in hand. It provides information on the background to the development of EN 1090-2 and its main differences from BS 5400-6. It provides explanatory notes on the background to many of the recommended choices and additional information. It extends the concept of a service category from a qualitative one to a quantified one, whereby the performance requirements are clearly defined in terms of specific levels of static and cyclic stressing, using principles similar to those use in BS 5400-6 and BS 5400-10. The term quantified service category has also been used in PD 67053:2009[21] which provides guidance on the use of BS EN 1090-3:2009 [22] (execution of aluminium structures), and which uses similar principles to those in BS 8118-1 and -2[23] (now superseded by EN 1999[24] and BS EN 1090-3 respectively). Tables giving specific methods and scope of testing of welds and simplified FFP based acceptance criteria for each method are given for the various quantified service categories. These are similar in layout to those in PD 6705-3. The FFP based acceptance criteria provide definite rejection limits, which are not defined in BS EN 1090-2.

It is expected that there will be many specifications written in future for a wide variety of bridge applications, all based on BS EN 1090-2. These are likely to vary from the very simple to the very detailed. Some may be project specific, others may be for generic types of bridge or be intended to embrace the full range of bridge applications. The degree of refinement of the resulting specification based on BS EN 1090-2 and PD 6705-2 can be decided by the user. The way in which the guidance in PD 6705-2 is used to construct these specifications will therefore vary according to the appropriate balance needed between simplicity of application and avoidance of unnecessary overspecification.

Conclusion
BS EN 1090-2 represents a very substantial change in UK practice in terms of specification of materials and workmanship for steel bridges. PD 6705-2 is considered to be an essential aid to those using the new European Standard for preparing specifications.

M Ogle, G Bowden

Acknowledgements
The authors would like to acknowledge the extensive contributions of the members of Working Group 6 of the BS Bridge Committee B/525/10 over many years during the development periods of EN 1090 and PD 6705-2.

References
[1] [2] [3] BS 153, 1922: Girder bridges: Part 1. Materials, Part 2. Workmanship, British Standards Institution, London. BS 153, 1922: Girder bridges: Part 3. Stresses, Part 4. Details of construction, Part 5. Erection, British Standards Institution, London. BS 153, 1958: Specification for steel girder bridges: Part 1. Materials and Workmanship; Part 2 Weighing, shipping and erection, British Standards Institution, London. BS 153, 1954: Girder bridges: Part 3. Loads and stresses. Section A: Loads, British Standards Institution, London. BS 153, 1958: Steel girder bridges: Part 3B. Stresses; Part 4. Design and construction, British Standards Institution, London. BS 5400. 1980: Steel, concrete and composite bridges: Part 6: Specification for materials and workmanship, steel (revised 1999); Part 10: Code of practice for fatigue (including Amendment 9353, March 1999), British Standards Institution, London. BS 5135, 1974: Process of arc welding of carbon and carbon manganese steels, British Standards Institution, London. BS 4870, 1968: Specification for approval testing of welding procedures: Part 1: Fusion welding of steel, British Standards Institution, London.. BS 5920-2, 2001: Structural use of steelwork in building: Part 2: Specification for materials, fabrication and erections, rolled and welded sections, British Standards Institution, London. ISO 10721: Steel structures: Part 1: Materials and design (1997); Part 2: Fabrication and erection (1000), International Organization for Standardization, Geneva. DD EN 1993-1-1, 1992: Design of steel structures: Part 1: 1. General rules and rules for buildings, British Standards Institution, London. ENV 1090: Execution of steel structures: Part 1: General rules and rules for buildings (1998); Part 2: (1998); Part 3: Supplementary rules for high strength steels (1997); Part 4: Supplementary rules for hollow section structures (1997); Part 5: Supplementary rules for bridges (1998); Part 6: Supplementary rules for stainless steel (2000), European Committee for Standardization, Brussels. EN 1090-2, 2008: Execution of steel and aluminium structures: Part 2: Technical requirements for the execution of steel structures, European Committee for Standardization, Brussels. EN 1090-1, 2009: Execution of steel and aluminium structures: Part 1: Requirements for conformity assessment of structural components, European Committee for Standardization, Brussels. EN 1090-3, 2008: Execution of steel and aluminium structures: Part 3: Technical requirements for aluminium structures, European Committee for Standardization, Brussels.

[4] [5] [6]

[7] [8] [9]

[10] [11] [12]

[13]

[14]

[15]

M Ogle, G Bowden

[16]

[17] [18] [19]

[20]

[21]

[22] [23] [24]

pr EN 1090-2, 2005: Execution of steel and aluminium structures: Technical requirements for the execution of steel structures: CEN TC 135 Draft Document N414 for Enquiry, European Committee for Standardization, Brussels. EN 1993-1-1, 2005: Design of steel structures: Part 1-1: General rules and rules for buildings, European Committee for Standardization, Brussels. NE E 1990, 2002: Eurocode Basis of structural design (including UK National Annex for Eurocode 0), British Standards Institution, London. BS EN 1090-2, 2008: Execution of steel and aluminium structures: Part 2: Technical requirements for the execution of steel structures, British Standards Institution, London. PD 6705-2, due for publication 2010: Structural use of steel and aluminium: Part 2: Recommendations for the execution of steel bridges to BS EN 1090-2, British Standards Institution, London. PD 6705-3, 2009: Structural use of steel and aluminium: Part 3: Recommendations for the execution of aluminium structures to BS EN 1900-3, British Standards Institution. BS EN 1090-3, 2008: Execution of steel and aluminium structures: Part 3: Technical requirements for aluminium structures, British Standards Institution, London. BS 8118, 1991: Structural use of aluminium, British Standards Institution, London. EN 1999, 2007: Eurocode 9: Design of aluminium structures, European Committee for Standardization, Brussels.