Proceedings of the



- Problems and Possibilities -

Delft University of Technology, The Netherlands October 27th, 1999

Assoc. Prof. Leo G.W. Verhoef Editor

Publications Office Faculty of Architecture Delft University of Technology

The Netherlands



Publications Office Faculty of Architecture FaculLy of Architecture

Delft University of Technology Berlageweg I

2628 CR Delft

The Netherlands

Translation: Mrs. Lorraine van Dam-Foley

Lay-out: Wido Quist and David Keuning

Copyright 1999

Renovation and Maintenance ISBN 90-5269-274-2

All rights reserved. No part of this volume may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical photocopying, recording or otherwise, without prior written permission from the publisher.



Assoc. Prof. L.G.W. Verhoef

W. Kippes W. Miiller

Prof. A. de Naeyer EG. Hofmann Prof. R. Oswald

A. Avdelas

Prof. M. Zador

R. di Giulio

L. Giorgo

B.M.B. Korompai M.Mun

Prof. S.M. Rodrigues Lopes J. Sroczynska

Prof. R. Zarnic

Prof. S. Chandra

1. Holmstrom

Prof. EH. Wittmann Prof. A.E. Long

1. Maxwell

Local Organising Committee:


Assoc. Prof. L.G.w. Verhoef

D. Keuning W.J. Quist

Delft University of Technology

Schloss Schonbrunn, Wien Ostcrrcichischcs Bauinstitut, Vienna Gent University


AI Bau Aachen

Aristotle University of Thessaloniki Budapest University of Technology Univcrsita degli Studi di Ferrara Universita degli Studi di Firenze Renaissance

Laboratorio Nacional de Engenharia Civil Universidade de Coimhra

Politzchnika Krakowska

University of Ljubljana

CTH Goteborg

KTH Stockholm

ETH, Ziirich

Queen's University of Belfast Historic Scotland, Edinburgh

Delft University of Technology

Delft University of Technology Delft University of Teehnology

The Netherlands

Austria Austria Belgium Germany Germany Greece Hungary Italy Italy

The Netherlands Portugal Portugal




Sweden Switzerland United Kingdom United Kingdom

The Netherlands

The Netherlands The Netherlands

The International Congress on Urban Heritage and Building Maintenance - Problems and Possibilities, Part J, is sponsored hy:


Delft University of Technology, The Netherlands


European Co-operation in the Field of Technical and Scientific Research, Belgium




In the 1ge century what had been the traditional building techniques up to that time were considerably changed owing to the use of cast iron ami later wrought iron and steel. In fact, the industrial revolution, initially based in England, would not have been possible without these materials. Not only the development of the product, in which the difference between iron and steel lay purely the fact that steel could be tempered, while iron could be hammered into shape, but also the manufacturing processes. These progressed via the puddling furnace, introduced by Henry Cort in 1784, which made possible the production of wrought iron, the Bessemer converter 1855, and the Siemens-Martin open hearth process ,md led to continuous improvements in building technology. In this period the technology for rolling of profiles was also developed. Even now the improvement of the quality of steel continues. Very high strength steels and steel foam for use in the construction industry are in the development phase.

Looking back into the past, we can see with what ingenuity the iron and steel, each with its own characteristics, were used in buildings, bridges, and towers. Certainly, the initial phases when cast iron was used bore witness to the daring with which this brittle material was used in building. The first cast iron structures were therefore mainly subjected to compression loading. Cast iron columns could be made in advance and used just when they were needed. In consequence, quicker construction methods emerged, so the columns could be given beautiful and efficient forms, as in the Stock Exchange, in London, constructed between l846 and 1849 and the flax mill in Shrewsbury in 1796. Bridges too, the oldest of which is the bridge over the Severn at Coalbrookdale, in England, built in 1779 and still in use, demonstrated the arch construction. Here the idea of loading the structure on compression was the basis of the design. For industrial buildings, where an open plan was important for the production processes, buildings were constructed in which the building envelope and the facades were built of brick or partly of stone. Cast iron columns were used first, later followed by steel columns, while the beams were made of steel. Floors were usually made of masoruy in the form of trough vaults. An important development was the joining technique, first using rivets, later bolts and then welding.

Left: Beautiful cast iron column structure (stock excange, London) Right: Efficient cast iron collums (flax mill, Shrewsbury)

The arguments in favour of using iron and steel were, and still are, their safety factor, their low dead weight, the ease of adaptation, their relative insensitivity to settling, the short construction time, the accuracy attained, up to the millimetre. and. above all, the possibility to create free spans and make free forms, In the Netherlands. a load-bearing steel skeleton was used for the first time in 1925 whereas in France Menier chocolate factory at Noisiel-sur-Marne built in 1871-1872 had already been constructed on a complete steel skeleton. This is very clearly expressed in the facade in which the spaces between the bars are filled by decorative masonry. Recently this building has been renovated as part of a total


renovation plan for the area and buildings of the Nestle Company, by the French architect Philippe Robert. In places where the steel beams in the floors with masonry trough vaults were well protected against very high temperatures, the same can be said for the facade of this building

Chocolate factory (Monier)

Bergpolderflat (van Tijen)

The Bergpolder block of flats, in Rotterdam, designed by the architect W. van Tijen is considered one of the highpoints of Dutch urban heritage of architecture in steel. It was designed with the objective of industrialising housing construction. Many attempts have been made to use industrial methods to fabricate dwellings, but up to the present, these have not met with success.

Left: Main office "van Leer" company

Right: New second floor connected to existing columns

With regard to the maintenance of buildings that form part of the urban heritage, the first question that arises is how can these buildings be saved. A typical problem is that usually no calculations of the original construction exist and that we do not know the properties of the materials used. Research should indicate the structure of the materials and, at the same time, whether specific renovation techniques, such as welding, can be used. The assessment of the safely of older constructions, specifically those in which cast iron beams have been used, demands the attention of experts in this field. The welding of cast iron still presents problems, certainly when the metal needs to take up any tension.


Jacking down system

When the functions of buildings are increasingly often changed, the loading regime also changes and understanding of the safety factors becomes essential. Older techniques such as riveting arc no longer used, so sometimes-artificial renovation tricks are used. For example, rivet heads arc welded onto profiles to suggest that a riveted construction has been used. At present once again there is indeed again a new structure in which riveting has been used; the 'Schwebcbahn' in Wuppertal. COllusion resulting

from unsuitable structural detailing can be a big problem. This is partly related to the lack of attention paid to thermal bridges in the past, as a result of which condensation occurs on the steel profiles .The profiles were certainly not always well protected. Another reason for the problem is that steel profiles were sometimes placed so close together that crevice corrosion could easily occur. However difficult this may be to achieve, we possibly need adapt structural details so that there are accessible spaces between profiles for monitoring and maintenance purposes. Alternatively, changes may be made so that moisture can no longer accumulate in parts of the Steel construction of the extension of the John structure.

Jay College of Criminal Art (New York)

In many cases, we find that buildings with older steel construction are of great significance. Architects take pleasure is using this type of mainly utilitarian building for residential functions. The constructive clarity of the iron and steel structure has great architectural meaning. In places where fire may playa role, the safety of people demands very unusual treatment. Often this involves the creation of an additional support system by means of walls or extra columns, used in such a way that the steel structure remains visible. Alternatively, sprinklers or foam forming coatings may be used, The way of strengthening the structures also demands the attention of the engineer and here the leitmotiv is always that the intended operation is not blocking future adaptations of the building when the function will change again. In the meantime, many buildings have been adapted and we begin to obtain an overview of the potential for the usc of existing constructions.


Left and centre: GEB, Amstedam (now under renovation) Right: Technickmuscum, Delft (van Zwol)

The well-known American architect, Marcel Breurer, used steel in designing the head office for the 'Van Leer' company, which manufactures drums and packaging materials. The interior had a layout similar to that of an open plan office for the two storey building, which had relatively high ceilings, and was the forerunner of the tend to use opcn plan otfices. With the reduction of the staff of the company, the building became too large and this was a reason to move to new premises and to renovate the building before offering offer the whole or part of it for rent. The building had considerable architectural beauty, which led to its being given the status of 'modern architectural monument' in 1985. The architectural company of Broek and Bakema designed the renovation operation in which it was stipulated that the 'building envelope' should remain unchanged. However, to make the most efficient use of space the two-storey office was to be transformed into a three-storey office. For this, a new floor was connected to the existing structure of columns and a supplementary structure was added. The columns above and below the existing first floor were cut through, as well as part of the columns under the first floor. After this, wing by wing, the first 1100r of the building, including the ceilings, and lighting elements etc., was lowered to its new level with the aid of jacks. In view of the carbon content of the steel, an unusual operation whereby the junction plates were welded onto the existing columns was feasible.

All in all, buildings with cast iron, wrought iron and steels structures can be successfully recycled. Nevertheless, because of the many unknown factors, especially in relation to the quality of the steel used in the elements and connections, it is necessary to make site specific preliminary investigations. These are buildings that are dear to architects who want to reuse this type of structure and thereafter to their users, in view of the relative ease with which they can be put onto the market after renovation.

Delft, 1999-10-10

Leo G.W. Verhoef



Cost, Iron and Steel Assoc. Prof. Leo GW. Verhoef 11
Problems in refurbishment and re-use
of existing iron and steel structures Prof. l.W.B. Stark 17
Problems and possibilities - cast iron,
wrought iron and steel M. Bussell BSc 27
The safety basis for the rehabilitation
of steel structures Prof. G. Sedlacek 41
Evaluation of structural safety of
existing structures, based on actual
material properties and geometry II. F.S.K. Bijlaard 51
Strength, stiffness and stability
adaptations 11'. GG Nieuwmeijer 55
Engineering Conservation -
cast & wrought iron Gl.O.WaUis BSc. 67
Refurbishment of the first mono rail
in the world Ing. K. van del' Burg 79
Restoration of historic iron structures
using epoxy bound laminates Prof. A. de Naeyer 91
Fire resistance of cast iron structures
in historic buildings Ir. L. Twilt 99
Thermal insulating glazing in steel
frames Ir. F.W.A. Koopman 107
Learning from liberty Th. Prudon PhD 117
Cautious repair of historical cast iron
constructions Dr.-Ing. R. Kapplein 123 IX


Assoc, Prot: Leo G,w. verhoe]

Delft University o/Terilllology, Farulry of Architecture, Tile Netherlands

The co-operation of nationally funded scientific and technical research organisations under the name 'COST' includes 28 countries, Almost all the European countries arc involved and work together. For the last 10 years organisations of non-' cost' countries have also participated in the individual actions.

While most of the international scientific programmes are organised 'top-down', the 'COST programme is organised 'bottom-up'. Bottom-up means that each institution such as a research centre or a university can propose a 'Cost' action on a specific topic if it needs intemational co-operation to raise the subject to a higher level.

Now 17 areas are active under the 'COST' umbrella, with, in total, about 300 separate actions. One of these is the COST C-5 "URBAN HERITAGE - BUILDING MAINTENANCE" action, which is lies within the area of Urban Civil Engineering (UCE).

The action C-5 embraces European towns and cities in which we find a large number of valuable buildings. From a historical, architectural or esthetical point of view, the majority of the buildings are less important than the highlights, which are labelled as cultural heritage. Nevertheless, it is worthwhile to preserve the urban setting with its neighbourhoods, buildings and streets as an important part of the historical heritage. It is also worthwhile to determine the characteristics of buildings, to identify decay factors, and to find solutions that will extend their lifetime by using both traditional and new technologies. The most important solutions are those that are economic in terms of cost and sustainability.

The two principle intentions of the C-5 action are:

- to evaluate criteria for determining the historical importance of individual buildings and of neighbourhoods in our towns and cities;

- to determine the effectiveness of the methods and techniques used in preservation work.

This general task is sub-divided and tackled by five working groups.

Working Group 1 'Evaluation of the urban heritage'

The approach chosen was to set up a system to establish data for each member country,

For example, to provide the names of responsible institutions and to get specific information about the way the established data are presented in each country and about the availability of established data;

After that to collect and to evaluate the available data concerning urban heritage

For example, the number of buildings in relation to their age, their size in m2 and m>, structural typology and present use.

The last aim is to collect data about the physical condition of buildings and potential renovation measures.

The difference in the proportions of dwellings dating from a specific period from country to country is striking. Figure I shows that the dwellings dating from the period before 1919 make up only 10% of the total number of dwellings in the Netherlands, while for the UK, this percentage is 27. There are many more older buildings in the UK.

For the COST C-5 countries, it has been calculated that 300 billion Euro (figure 2) is needed for the

Iron & Steel


The age of the dweUing stock

Minimum maintenance, refurbishment and renovation needs

I::~~::~---------:IC;~ comb • f~ ;~,~: I '~:



F~'l~~~,rjr':f: C' :'~'~'; -'-'-. __ .. -,. 1 ,Ct'K> ,ODD

~'5'OIl""d t,1/,7

~-l(ifp1n-:n i rt:-~-.:JwJ 6Cr;;~ ,SOD ~!. ?!jL,

'~'_"'~~~_~ n •••• _ •• _~~ •••• ..".~".'W~ .

~""~'~"lgD 2,:697

i~~;?:;'~C5oG0Ut1f;tI?6 1 12,576·,Oj~1 2fii'd'f

1 Milr;'f\u.:nlll s-: t tUJlJ~.~ Mu.m.o:; 1'1 m;~nt ;:-;~."';;(>..,,';1~I{d'l nceee ;W3.EdJ3~ '1 tH.127

I of t~lt!- build'l;ag ~t{}I:}. in th~ C{;,llit Cs.~01.IntJl(Io'


Figure 1: The age of the dwelling stock

Figure 2: Minimum maintenance, refurbishment and renovation needs

JrOJI & Steel

maintenance, refurbishment and renovation of the housing stock. For Europe as a whole, therefore, the sum required will be at least 600 billion Euro. It is clear that we arc dealing with gigantic costs, but also with gigantic opportunities for the market. It is a major task for the EC to generate solutions that will deliver value for money. Already, at this moment in the Netherlands the market for maintenance and renovation exceeds that for new buildings.

Working Group lA 'Framework evaluation criteria'

Because of the immense scope of related topics within the action C-S, a task force has been set up to design an outline framework. In other words to determine how the assessment process can be integrated.

Working Group 5 'Lime technology'

- Working Group 5 collaborated with eurolime. A real form of collaboration now exists between the two groups. There is also a desire for stronger collaboration with rilem. The working group is quite successful, having realised a network linking interested parties in Europe. They have assessed lime production techniques and technical information and they were allocated a three-year funding to investigate the consolidation of lime-kilns in the Raphael programme. They have discovered that up to now 90% of the research has been concentrated on additives and only 10 % on the lime itself. Last but not at least, they found it necessary to formulate a Eurocode for repair.

Working Group 2 'Inspection and diagnosis'

Working Group 3. "Methods and techniques of rehabilitation".

The task of Working Group 2 is to determine the condition of the buildings. Tests are necessary to determine the condition before, during and after interventions.

The methods and techniques used for inspection and diagnosis are divided into non-destructive, slightly destructive and destructive ones. Important case histories of projects in the different countries have been collected. These case histories give an overview of the object, the test period, scope of analysis, test methods and results. The case histories themselves will be used to see what kinds of inspection methods exist in the various countries concerned and also to indicate why different methods are in usc.

An other way of dividing the three types of tests:

- the first type concerns maintenance. A regular check is needed to monitor the quality of the building.

This check may be visual but there are other non-destructive, tests that yield more scientific information.

- the second type is needed after the damage has been noticed. Every time advice to repair is given it is necessary to have a description of the original material and how this has been determined; an explanation of the reason(s) for the deterioration and how this has been researched; and the materials chosen for repair, with their properties and the reasons for their use.

- the third type has to be used if the function of the building has to be changed in the case of renovation or re-use of the building.

The term 'buildings' is understood to include parts of buildings and the materials used in them. There is an agreement between Working Groups 2 and 3 to look at the same building parts or elements that demonstrates the strong relation between both these working groups.

frail & Ste«]


Roughly speaking, Working Group 3 deals with methods and techniques for rehabilitation. To limit the task, both working groups decided to pay attention to maintenance and re-strengthening and to focus this on seven areas of interest, these are:

Restrengthening of foundations as part of a building;

Restrengthening of cast iron, wrought iron, steel; concrete; brickwork; stonework, wood and plaster.

- One of the most important objectives is to delineate and study problem areas in which specific improvement techniques can be applied. The techniques are studied within the context of the areas in question and not separately because some techniques, such as injection, can provide solutions in many different areas. For example injection can be used to fill gaps between metal and bolts and it can block pores, preventing rising damp. Moreover, this technique can also be used to create cohesion in the soil and thus increase its bearing capacity. To maintain a clear approach, areas that require improvement are defined and the techniques that may provide the requisite solutions are investigated. The international approach that is inherent in the COST- programme extends the scope and permits comparison of the solutions found in the various countries involved.

In this intemational congress today we are also looking towards the same goals, which are:

- determination of the condition of buildings, in other words determination of problems;

- determination of the most suitable research techniques:

- determination of criteria for the evaluation of intervention work;

- establishment of criteria for determining the effectiveness of methods and techniques used in preser-

vation work;

- determination of effectiveness of the assessment technique employed;

- promotion of case studies;

- development of research action areas and topics.

For reasons of efficiency, today the goals are limited to: - problems

- research/testing

- restrengthening

- cases

In such a simple approach, it seems that the only materials that matter are iron and steel. Of course, the different subjects (problems; research/testing; restrengthening and the cases) will show that the reality is much more complex. Often, from a structural point of view, the static system and the way forces are acting on it determines the Iorm and the dimensions. However, after a building has been in used for some years, time will show that the influences of temperature and more especially moisture have major effects on the future need for maintenance. During the structural design period in the past, building physics was seldom understood. Indeed, it is only in recent decades that building physics has come to the fore and been established as important basic knowledge [or architects. For this reason, previously little or no attention was paid to the need to protect iron mid steel structures. It was common usage in the Netherlands and also in other European countries, to place beams made of iron or steel in facades made of brickwork or stonework. Through the mortar joints, rainwater could easy reach the metal beams, which in the best cases were protected by a coating of red lead coating, but often had no protection at all. Moisture from the inside of the building also produced condensation on the relatively cold surfaces of the beams with the same effects. To determine the problem is not difficult; often it is even visible, on the facades. Even the damage caused by corrosion often has minor effects on the structural system. From the aesthetic and maintenance points of view, this is not acceptable, especially because we know


Iron & Sleel

that it is an ongoing process. II is not difficult to protect iron or steel beams against the affects of moisture, but improving the present structures that moe encapsulated in masonry is a complex process. This is certainly so when we want an effective cure that will also fulfil our wishes in technical and economic ways as well as an aesthetic way. Therein lies the real task for Cost in relation to Iron and Steel.

Figure 3: Damage caused bij steelbeams often visible in brickwork

Figure 4: Flat by van de Broek en Bakcma Left: Building under renovation

Right: Tipical problem-detail

iron & Steel



wooden floor




Iron & Steel


Prof if: JWB. Stark,

Delft University of Technology, Subfoculty of Civil Engineering, The Netherlamls


In this paper an overview is given of potential problems to be expected in refurbishment and re-use of existing metal structures. First attention is given to the appraisal process. Then methods are given for collecting the necessary information.

Aspects of the properties of cast-iron, wrought iron and structural steel are discussed. Then analysis is covered and finally some problems which may occur during the execution are presented.


Fortunately, the last decades there is a tendency to consider refurbishment and re-use of existing buildings a good alternative to demolition and re-building. Also there is more attention for the conservation of human heritage.

The consequence is that architects and engineers are more frequently confronted with the task to report on the potential for re-use of a building and to give advice on repair. strengthening and alterations.

In the professional education and training this subject is generally not well covered, so the consultant often starts this job with lack of specific knowledge and experience. So it was a good idea to collect available information and establish the European "state of the art", as is the target of the COST-CS programme.

The item of to-day's session is iron and steel. I have been requested to introduce this session by presentation of potential problems.

The first problem is that I do not consider myself an expert in this field. During my career at TNO I was of course involved in a number of relevant projects, but this only forms a minor part of my total experience.

As I expected that this may be the case for many of us, I still accepted the invitation under some pressure of my esteemed colleague, Leo Verhoef.

As requested, I will focus on the presentation of problems. Answers and solutions will be given by later speakers in this session.

Another restriction is that this paper is concerned with iron and steel structures for buildings. Bridges, piers and towers are not considered, although nice examples of rehabilitation of old bridges exist.


An appraisal has to be carried out to determine whether an existing structure in its present form or altered is adequate to fulfil the requirements.

In figure 1 a flowchart of an appraisal process is shown.

An appraisal may of course concern several aspects, but here it is limited to a structural appraisal. Structural adequacy is focused on strength, stability, serviceability, durability and fire resistance.

/1'011 & Steel


Propos cd use I'ropo,;"d iih('rationc Propo.,('d 11l'gmdinj!

Building [{'gldmiun,


..... J' ..


fill. OK

1n"IC51 igathm

Dimensioll'. Malerials Connections Corrosion jJ,:,iix[s

Figure I: Schematic overview of an appraisal process

For new structures the structural adequacy is normally demonstrated by means of the design calculations. In these calculations it is normal to develop a mathematical model of the structure. Material and dimensions of members and connections are chosen in such a way that the applied loading and other imposed conditions can he resisted safely. Properties of materials and members are usually taken according to approved design tables and product standards.

Design procedures given in design standards or codes of practice are followed. In this way, the verification will be acceptable to building control authorities.

However in the appraisal of an existing building, the situation is quite different. There is no free choice


Iron & Sleel

of the structural system. It is "as built". The dimensions of members, the properties of the materials and the detailing of connections have been determined in the past and the information is often not available and modem product standards do not apply. Even if the original drawings and/or other project spccifications are available, clements may have been repaired. replaced or even removed in the past. Materials may contain defects, corrosion may substantially reduce the member dimensions and clements may have been distorted by mis-use.

So it will normally be necessary to collect the data required for assessment of structural performance by a survey.


The target of a survey is to collect all data required for assessment of structural performance. This includes the geometry of the structure and its elements and details, the identification of materials and fasteners and its condition and defects.

Sometimes original drawings and some form of project specifications are available (see figure 2).

:~~r!:I;.Bn,~ C '. "

\",,_t1;!tr:; .. tFf)l.t1..::1...:~ta LI.1'n d(:~ft ,r."li-\t...~l.

, .B/.:"';,.!n~ if: 1/7.8


Figure 2: Sometimes original drawings are available r 1].

Iron & Slee!


This information is of great value, but should be handled with care because the structure may have been altered in the past without formal registration of these alterations. T experienced that essential structural elements as for example tic rods, restraining elements, fasteners and even columns were removed. Also elements may have been replaced by others with different dimensions and material properties. So it will often be necessary to make a measured survey of the building. This will include the overall dimensions of the structure and assessment of the structural system, influenced by properties of supports, hearn-to-column connections and restraining elements. Particular important is how the stability of the structure is arranged for.

The dimensions of the members should be measured. It is worth wile to have available old section tables as for example given in [4]. Open sections can easily be measured, provided these are not encased or cladded. Normally a tape, a scale and a micrometer are sufficient to do the job. To determine the wall thickness of closed sections such as cast-iron circular hollow columns and rolled tubular sections small holes have to be drilled. One should bare in mind that especially in cast iron sections the wall thickness may vary considerably over the length and over the cross-section. Also ultra-sonic equipment may he used to determine wall thickness.

The dimensions of connection details should be determined including the thickness of gusset plates and the number and diameter of bolts and rivets. The diameter of bolts can normally be measured because the threaded part of the bolt is projecting beyond the nut. To determine the diameter of the shank of rivets is more difficult because a head is formed on both sides. However an experienced inspector can estimate the shank diameter from the form and the dimensions of the head. This estimate may be checked by grinding the head of one or two rivets. Be aware of fakes. Glued bolt heads and rivets formed from putty have been found in past renovation projects.

Identification of the metal can start with visual observation. The first step is to divide in cast iron, wrought iron and steel. Dating of the structure may support separation of wrought iron from steel. In l3] helpful information is given on the visual characteristics of the three materials, Sometimes it may be difficult to distinguish between wrought iron and early steel, also because there is an overlapping period of use.

Sampling will support the evidence of visual inspection and provide more detailed information about structure, chemical composition and structural properties of the material. For sampling, small pieces should be obtained by sawing or core drilling from lightly stressed areas of the structure.


Figure 3 (I): Original bolt in cast-iron column tl] Figure 4 (1'): Corrosion ncar connection [2]


Iron & Steel

The condition of the structure is established by visual inspection normally carried out in combination with the measurements. In many cases accessibility of the structure is a problem. This is especially applicable for the determination of corrosion damage. Those places which are now difficult accessible for inspection were most probably also not accessible for inspection and maintenance and so arc places where corrosion can be expected. Examples are steel components huilt in masonry or under gutters. Also hases of columns covered by concrete and badly detailed connections are potential places for severe corrosion.

Apart from corrosion other defects may be detected by an inspection of an existing structure. A number of possihle problems are listed below.


- Embedded or encased iron and steel

- Poorly designed details

- Leakage of gutters and windows

- Chemical attack

- Delamination of wrought iron


- Bolt and rivet distortions

- Missing connectors

- Defective welds

- Fractures and cracks in welds


- Blowholes and porosity in cast iron

- Delaminations in wrought iron

- Fractures and cracks in material


- Damage caused by impact

- Permanent set caused by overloading

- Buckling, elastic or inelastic

- Out-of true rolling

- Distortion during erection

- Deformations caused by setting of foundation

Material properties

The properties of materials relevant for analysis of stuctural adequacy are:

'" Strength, expressed in values for yield strength fy and ultimate tensile strength fu * Stiffness, expressed in rhe value of the modulus of elasticity E

'" Ductility, expressed in the maximum strain at failure eu

For modern structural steels these properties can be derived from a stress-strain diagram measured in a standard tensile test.

In analysis it is normally assumed that the values are the same in tension and compression. However this is not the case for cast iron. As an illustration in figure 5 stress-strain diagrams for cast iron in tension and compression are qualitatively shown. For comparison a stress-strain diagram for structural steel is also given.

11'011 & Sled





... "" _-- ~ - .......... St.-uduralSteel ..







Cas! Iron


Figure 5: Stress-strain diagram for cast iron

For modern steels of various grades minimum guranteed values for the relevant structural properties m-e given in product standards. In design codes material safety factors are given to derive the design values from the values in the product standards. For structural assesment of existing buildings this information is not so easy available. Properties of structural metals have changed considerable over the time as a consequence of improved production technology. And properties of cast iron and wrought iron differ substantially from structural steel.

If re-analysis is required. then in a first stage approach the relevant material properties are normally estimated at aminimum expected value. Dating and information on the origin of the material can he of help to fix these values.

Fortunately in ealier times the adopted safety margins were considerably higher than nowadays and the analysis methods and structural knowledge was less developed. Due to these circumstances it is well possible that the result of this sort of first stage analysis is positive in spite of the conservative underestimation of the material properties.

Only in case of doubt it may he necessary to determine the actual material properties more exactly by testing of samples from the structure.

Just as an indication the characteristics of the three types of metals are discussed now in general and some indictive values for the relevant material properties are given.

Cast iron

Almost all cast iron used in historic building structures has been grey cast iron. Typical characteristics are:

* Casted in a mould. Complicated shapes can be achieved. '" Good compressive strength.

* Low tensile and flexural strength. '" Low ductility.

'" Sensitive to production defects as flaws, blowholes shrinkage cracks and dimensional inaccuracy. '" Welding very difficult if not impossible,

'" Relative good corrosion resistance.


Iron & Steel

Nodular graphite iron is a modern form of cast iron with a higher tensile strength and better ductility, It is suitable to produce replicas for replacement of defective cast iron components,

Indicative values ( in N/mm2 ):

!yalue of: IRange of actual values -
Indicative design value
Ultimate tensile strength r 65 - 280 30
Flexural strength fU,b 115 - 315 50
Ultimate compressive strength f 550 - 750 200
Modulus of elasticity - tension E I 65 - 90 65
Modulus of elasticity - compression Ee 85 - 90 85 Wrought iron

Typical characteristics are:

* Good compressive and tensile strength.

* Good ductility in case of "longitudinal deformation"

* Due to the laminar structure not well suited for loading transverse to the surface

* Butt welding possible, fillet welding not recommended because of risk of lamellar tearing * Relative small section sizes

* Relative good corrosion resistance.

The mechanical properties of wrought iron are in the same order of magnitude as those of early mild steel, Indicative values arc:

Design yield stress: fy = 220 N/mm2

Modulus of elasticity: E = 200,000 N/mm2

Structural steel

Typical characteristics are :

* Good compressive and tensile strength. * Good ductility

* Old steels can generally be welded, They may have a relative high "carbon equivalent", Choice of welding material and welding procedure should preferably be based on test welding, For a first stage analysis the values for the mechanical properties can he estimated as the values given in present standards for the lowest grade of mild steeL


The objective of structural appraisal is to show that the structural system is adequate to carry the expected loads with sufficient safety against failure and to perform well under service conditions, An important difference with a new design is that an existing building has given an historic proof of being able to perform satisfactorily, It could be stated that no further analysis is necessary if the structure is not modified, if there is no significant des integration and if the expected loads for future use are not higher than in the past. However this docs not mean that the structure is in conformity with the building regulations. Such a conformity check may be required by the elicnt or the building control authorities.

Iron & Sire!


Re-analysis is anyhow necessary when the structural system is modified for the intended future use. This is illustrated by the example in figure 6, where the iron columns and part of the beams are lateral not supported.

Figure 6: Re-use of existing warehouse for appartments. Columns and part of beams lateral not supported.


The loadings used in the structural assesment generally have to be based on the current standards. Dead load and imposed loads are often not a problem unless the intended use requires much higher imposed loads than the original use. In that case it can be considered to reduce the dead Load by using leightweight construction and light partition walls. In some cases also modification of the restraint conditions may help. If that is not sufficient strengthening of the structure is necessary.

For roof structures the wind loads given in the current standards may lead to big problems. Often these structures have not been designed for substantial upward loading caused by wind suction.

Wind tunnel testing can help to reduce the design load for a specific structure at a specific location. Sometimes it is suggested that variable wind loading may have caused fatigue damage in an existing steel structure, This was the case for the renovation of the "Kurhaus" in Scheveningen. I assume that the contractor would have been happy when he had been allowed to demolish the historic cupola. That would have simplified the process of renovation considerably. However after careful inspection and fatique analysis we could prove that the original steel structure was fully adequate. In general wind loading may be regarded as a quasi-static load unless vibrations occur.


[roll & Steel


For modelling of the structure the normal principles used in nowadays structural design can be used. However it should be realised that designers of historic buildings mostly not explicitely designed for resistance to sidesway and lateral stahility of structural elements. Often it was assumed that resistance to lateral loads and lateral stability was assured by the interaction with floors. roofs or walls. For some structures even the glass was used as a stabilising element. In this respect modern design concepts using structural glazing can be considered as a revival of an old concept.

So if a first step simplified analysis indicates that an apparent sound structure is inadequate it is often useful to reconsider the structural model and take in account, whereever permitted, the contribution of "secondary elements".


In accordance with modem so-called "Euro-slang" the word execution is used here for repair, strengthening and replacement.

Some potential problems in this phase of the process are: * Replacement.

In historic huildings elements severely damaged or corroded should preferably be replaced by a similar section. However often these arc not availahle anymore. In that case, depending on the restauration philosophy a replica can be used or deliberately a clearly visible modem section.

* Repair of connections.

As indicated earlier in this paper welding of older material and in particular cast iron is tricky or even impossible. Not many steel fabricators are able to do true riveting. The use of cup-headed bolts may he an altemative. Rivets can be replaced by injection bolts.

* Structural adequacy during execution.

As indicated before in older structures the so-called "secondary elements" as floors, roof cladding and walls may play an important role in providing stability to the structure. So removal of these elements may cause dangerous situations and should only be carried out after consultation with the designer.

Safety and environmental requirements.

Progressively these conditions influence the methods and the costs of renovation projects. This is particularly so if the work is carried out while the building is still in service.

The users have to he protected against hinder by noise and dust.

Just as an example in figure 7 a work unit is shown used for repair of the roof of railway station Amsterdam CS.

Iron & Sled


'.'L=.: ..... ., .. ~, ..... < .. ~.·£yj~r.:l;;1~-: -

_~~_.\ tfI :.'l.~~l ,~

" .. O/~"- . -;.- _.,: .. -,~>-:)_::::..S:?!:.~l!!L__.


''';'' ,

I \~i

I I l'

---'-----f,---:--~·t' ~·'4-.-+----. '"'--~.~. ~

V~ -:~

-P-",'"-r.~-~lJ ,~,~,,,";

Figure 7: Work unit used for repair of Amsterdam CS


III Douna, C.,Koneman, L., Restauratie Station Hollands Spoor, Den Haag, Bouwcn met Staal 101,1991 [21 Brink, Gv.d.Raay, 1.Th, van, Rcnovatie ovcrkappingen Amsterdam CS, Bouwen met Staal 105,1992 [31 Bussell, M., Appraisal of Existing Iron and Steel Structures, Steel Construction Institute, 1997

[41 Bates, W., Historical Structural Steelwork Handbook, British Constructional Steelwork Association,1984

[5] Kluft, D., Verbouwing pakhuizcn aan de Cruquiusweg in Amsterdam, Bouwn met Staal 98,1991


Iron & Steel


M. Bussell, BSc(Eng), MASCE

Consultant, formerly 01'e Amp & Partners, London, United Kingdom


I am grateful for the invitation to offer my ideas and experiences in this field to an audience in Delft. As a major aim of this Congress is to stimulate the exchange of experience and to leam from each other, I hope you will find it useful that my paper deals largely with British practice in structural iron and steel. Also, the approach to iron and steel construction has much in common with the approach to other building materials, and so it is logical to consider a general approach.

Maintenance, refurbishment, and adaptive reuse of existing buildings today occupies many architects and engineers. This reflects a remarkable change in attitudes. Until a generation ago, a redundant building in an urban area was usually demolished (and many still are). The same fate often befell many pleasant buildings that were still serviceable (and this too, sadly, still happens). Thirty years ago, an architect or engineer would probably inspect an existing building only as a prelude to letting its demolition contract. Today, however, the building professional is often working on a maintenance strategy or a report on the potential for refurbishment or reuse of such a building, and needs to approach its structure knowledgeably so as to respond informatively to its possibilities,

This can cause problems. Modem engineering training and codes of practice focus on the design of new construction using modem materials. However, the existing building, usually at least half a century old, was built using materials and techniques now generally obsolete. Its structure may contain cast iron or wrought iron (no longer used) or early steel (made without modem quality control), designed either by rules of thumb or by formulae which have now probably been forgotten or supplanted in the minds of practising engineers. In order to make informed decisions about the work required, it is necessary to understand the historical development and methods of manufacture of these materials, their properties, and their evolution in structural use.

The care of the urban heritage covers a spectrum of activities. From one extreme to the other these include:

* Regular maintenance tasks to ensure the daily 'health' of an occupied building serving its original


'" Minor repairs

* Refurbishment for continued use, or adaptive reuse once its original function has become redundant * Major repairs

In Britain, planning regulations apply to work that alters or affects the function, layout, or appearance of buildings. In addition, there are legal controls for work affecting historic fabric, administered by local authorities and by national bodies: English Heritage, CADW in Wales, and Historic Scotland. There are three categories of such control:

Iron & Steel


* For listed buildings (including structures such as bridges), which generally are still occupied and/or in use

';' For Scheduled Ancient Monuments, which generally were not built for occupation (e.g. Stonehenge), or are now unoccupied or mined, although some still-used structures are Ancient Monuments (e.g. The Iron Bridge of 1779 at Coalbrookdale in Shropshire)

* For Conservation Areas - areas of historical and/or architectural merit whose character is to be conserved or enhanced

Consent will be needed for any major work, and may even be necessary for minor work such as repainting doors and windows of a house in a Conservation Area, where an unsympathetic choice of colour or texture can impair the overall character. The picture is complicated. For those interested, an English county council publishes a valuable guide to British historic buildings law that is regularly updatcd l.

English Heritage's very sensible view is that historic fabric should in general undergo 'honourable decay', or grow old gracefully, and not be returned continually to some (often notional) 'pristine' condition.

Approach to Work on the Urban Heritage Introduction

The scope of work can vary from minor maintenance to total refurbishment or reuse with major alterations. The following table compares the characteristics of the two broad categories into which such work can be divided.

Maintenance and minor repair Refurbishment, reuse, and major repairs
Initial condition survey, followed by regular Investigation of existing construction
Maintenance schedule Drawings and specification for work required
Routine, regular repainting and minor works 'One-off event, possibly on large scale
Maintaining or repairing existing fabric Possible major alterations
May require heritage approval, but often will not Almost always requires heritage approval
Does not require structural appraisal/assessment Almost always requires structural
appraisal/assessment There are analogies with medicine. Inspections and investigations represent the doctor's examination; maintenance and repair represent 'medicine' for the patient; refurbishment, rense, and major repairs are 'surgery'. While on this theme, I wonld add that the two most 'dangerous' times in a building's life are when it is not looking after itself properly - the building is derelict, and vulnerable to the elements and vandalism - and when it is undergoing surgery. Too many buildings have been bumt down due to carelessness during what is often only minor maintenance work, and special attention must be given to protecting the fabric against fire, water, vandalism, and the like.

The procedure to be followed is basically the same in all cases, although the emphasis on the different stages will obviously vary, The philosophy is to 'plan the work, then work to the plan'.


Iron & Stee!

Define objectives and implications of the intended work

Objectives range Irorn repainting to a major refurbishment or alteration for reuse. There must be a clear agreed brief. The work will have non-technical implications. If the building is occupied. access must be provided when it is empty, or arrangements made to minimise disruption to occupants. Getting safe access to all relevant parts of the structure may need ladders, towers, hoists, platforms or even scaffolding. (Exterior structure may be a particular problem.) Opening-up concealed structure requires tools, removal of rubble, and subsequent making good. It is usually noisy and messy.

Make initial inspection, and appraise using immediately available information

This first acquaintance with the structure is invaluable. Major problems should be apparent, and the overall 'feel' of the structure and its condition can be gathered. Information to hand at this stage may be documentary (reports, drawings, etc) and/or oral (notable events, accounts of past problems and remedies, etc). Photographs and sketches made now will be useful in discussing and planning subsequent stages.

In maintenance work, the first essential is a comprehensive condition survey, identifying all building components, the materials of construction, and their condition. An inspection regime can then be set up (see below).

Carry out any immediate action

If the structure is obviously at risk then precautionary action may be essential. This could involve propping and/or preventing access if the structure appears dangerous. Less obvious risks include leaking pipes or other causes of progressive deterioration in contact with the structure. Labour should be on hand to do any necessary tasks, such as ad hoc cleaning, clearing a blocked rainwater outlet, or mending a broken window.

Obtain and review all available documentary information

Information specific to the building and its structure may exist in many different places. The benefits from finding original drawings, or even calculations, must be weighed against the time involved in seeking out sources (often with no useful result). In medical terms, the structure is the patient; documents are the patient's medical records.

Carry out inspections and/or investigations

Inspections and investigations are time-consuming. The aim should be to get just as much information as is both necessary and sufficient to allow an informed appraisal to be made. Thorough records must be made by sketches, notes, photographs, audio and video recording. Samples removed for testing must be clearly identified as to origin.

Inspections for maintenance and minor repairs

This work comprises routine inspections, annual general inspections, and less frequent detailed inspections.

The routine inspection should be the responsibility of an identified individual, and involves a walk-over (with binoculars and other aids as necessary, to allow clear sight of all surfaces) to check that all is as it should be. The inspection is basically a visual task, although opening-up may be needed if signs of distress me noted. Frequency of routine inspections will vary from daily to monthly, depending on the probable consequences of a defect or Iailure, In addition, 'active observation' and fault-reporting should become part of the culture and routine of all working in the building, so that any defects can be dealt with promptly. This should be clearly defined as everyone's responsibility - not just a job for the maintenance staff.

11'011 & Steel


For general and detailed inspections, close access will be needed. Access costs can be high, so these should be combined with repairs and repainting already scheduled, and any work found necessary during the inspection. The annual general inspection comprises a visual inspection of all main elements. It assists in the planning of maintenance activities and expenditure, and in particular will highlight priorities for maintenance tasks. The detailed inspection will involve access to all parts of the exterior to update the condition survey, typically at five-yearly intervals.

investigations for refurbishment, reuse, and major repairs

Investigations could involve some or all of the following operations, while a more limited study might involve only one or two:

* Measured survey of building

* Detailed measured record of elements, connections, etc * Identification of materials (see Appendix)

* Location and recording of defects (defects are covered by other papers to this Congress)

'" Trial pits and other investigations of foundations and ground conditions, or underwater (e.g. in pier structures)

* Installation of monitoring devices, particularly if movement appears to be taking place

Carry out appraisal/assessment of construction Philosophy

Appraisal or assessment is a key task in refurbishment and reuse. From it comes all views and decisions on the condition of the fabric, and on the work needed. It involves the exercise of engineering judgement, combined with awareness of material properties and behaviour, and a practical approach to the behaviour of structures, components, and connections. When dealing with older structures, the engineer has to face the added complication that the materials as then used are now effectively extinct for new work, while some structural forms, components or connection may be archaic. The ultimate conclusion from an assessment has to be: does it/will it work? Deciding 'yes' or 'no' - or indeed 'maybe'puts a burden of responsibility onto the appraising engineer.

An important difference must be recognised between scientific and engineering approaches. A scientific approach would seek to establish the 'true' behaviour of a structure, relating this to the effects of 'actual' loadings and service conditions. An engineering approach, although similar in comparing behaviour against loadings and service conditions, is concerned essentially with establishing whether the structure is adequate rather than with identifying its true behaviour (if indeed such a unique set of stresses and strains could be identified). If a structure can be shown to be adequate by a simple, conservative assessment then a more detailed assessment is not necessary. Load testing may be considered if further assessment docs not prove adequacy. Finally, repair, strengthening, or replacement must be investigated.

An existing structure is a body of information (perhaps supplemented by documents). It stands, with all faults, and clearly - by not collapsing - it has found a way to resist loads and withstand the effects of deterioration and defects, albeit with perhaps some evidence of distress. In theory, a full appraisal would demand complete dismantling of the structure to retrieve all the information - an obviously absurd and literally self-destructive approach. Any realistic appraisal must thus be a compromise. More positively, structural appraisal offers scope to determine the actual state of the structure as it was realised. This is significantly not how structures are designed. In design it is normal to develop a mathematical model of a structure that appears to resist applied loadings and other imposed conditions in a sound stable way. This is verified by calculations, using a code of practice to ensure that the resulting design accords with current good practice and will be technically acceptable to building control authorities and others concerned.


11'011 & Steel

The need for calculations

It might be assumed that calculations will always be needed to demonstrate structural adequacy. This is not so. The appraising engineer should ask these questions after carrying out investigations to an appropriate extent:

"' Is there evidence of problems such as instability, undersize or missing members or connections, or gross distortion?

* Is there evidence of distress such as excessive deflection, corrosion, distortion, cracking or fracture? * Is there activity that could produce or aggravate distress in future (for example leaks to cause corrosion, exposure to aggressive chemicals or seawater, exposure to substantial dynamic or repeated loading)?

"' Will continued use, or proposed new use, increase loadings and other actions on the structure?

If the answer in all cases is 'no' then it is probably unnecessary to make calculations.

The American saying "if it ain't broke, don't fix it" is relevant when making an assessment. This is particularly true for listed buildings, buildings in conservation areas, and indeed any structure of architectural or historical interest, where unnecessary intervention such as repair or strengthening may be both damaging to the present character and, indeed, unacceptable to the relevant heritage authorities.

Building regulations and original design methods

In 19th century Britain, design of iron structures was based on simple elastic analysis to obtain design forces; the ultimate tensile strength (UTS, derived from tests); and factors of safety applied to the UTS to give permissible stresses. At this time the yield strengths of wrought iron and mild steel were not regarded as relevant when determining permissible stresses, The table below gives typical factors of safety. These are high, reflecting the uncertainty of structural behaviour, the variability of materials, and also the brittleness of cast iron compared with the ductility of wrought iron and steel. Factors arc higher for live load than for dead load, a forerunner of today's partial load factors.

Before the London County Council (General Powers) Act in 1909, there were no statutory controls

Material Element Dead load Live load
Cast iron Beams Columns 5-6 8-9
5-7 8-10
Wrought iron and Steel Beams Columns 3-4 5-6
4-5 6-7 over the design or construction of building structures in cast or wrought iron or steel. The earlier 1894 Act had made some reference to metal frames, but until 1909 the masonry wall thicknesses were still prescribed in relation to overall building height (itself restricted in the Act), as if they were fully loadbearing. In practice of course a metal frame would carry virtually all 1100r and roof loads, and indeed much of the wall self-weight also. A framed structure designed accordingly was therefore beneficial only in allowing fast erection of the building skeleton; after that, the lleshing-out had to be done with masonry of substantial thickness. (The Ritz Hotel of 1904 in Piccadilly was the first such framed structure in London.) The 1909 London Act gave detailed rules on (elastic) design, imposed loadings, permissible stresses, and minimum metal thickness. It must be emphasised that these stresses arc effectively lower-bound figures; any iron or steel used would be adequate at these stresses.

iron & Steel


The 1909 Act rules for metal frame construction were effectively reprinted in the London Building Act 1930, although by 1909 the use of cast and wrought iron had already effectively ceased in new building,

BS449, the use of structural steel in building first appeared in 1932, with many subsequent revisions, and remains available today as a permissible stress code of practice for structural steelwork,

National building regulations were not introduced into England and Wales until 1965, Before then, building control was enacted through local by-laws, although few dealt with iron or steel structures until more recently, and then usually referred to BS449. From 1965 onwards, the Building Regulations have accepted designs to BS449, and the later BS5950, as complying with the regulatory structural requirements for buildings.

Other design guidance

The Institution of Structural Engineers published various reports on steelwork in buildings, notably in 1927 (five years before the first BS449) in the form of a model code. This and other such publications are described more fully in sections of the BCSA Historical Structural Steelwork Handbook 2.

Loadings used in assessment

The loads for a present-day assessment should be based on codified figures for dead, imposed, and wind loads. However, when actual thickness, densities, and other dimensions of construction are known, these may be used both for the actual values used for dead load, and (where relevant) for considering partial load factors. Locating heavier features such as paper stores and machinery on ground-bearing slabs may allow the general floor areas to be designed for light loadings, eliminating or reducing the need for strengthening or replacement of existing structural fabric.

Modelling the structure

The study should begin with a simple assessment, refining this only when necessary to show that the structure is adequate. The model used for analysis should be correspondingly simple for an early assessment, being refined subsequently. These may of course be little to refine: an unclad simply-supported steel beam with a central point load will have little scope for redistribution of midspan moments, composite action with concrete encasement, or load sharing. If the steel strength is known and an assessment shows the member is clearly inadequate, then it is. But often, if not usually, there is scope for refining the model and making second or even third assessments.

Cast iron

Cast iron should always be assessed by elastic section analysis, since failure is invariably brittle, abrupt, and catastrophic Serviceability loads are to be used; in limit state terms, (f = 1.0.

The first, simple, assessment is usefully made with the aid of the 1909-1930 London Act permissible stresses for cast iron. These figures (shown in the table below) are normally acceptable to London District Surveyors, and other British building control offices, without the need for any materials testing. If this check is satisfactory then the member assessment is adequate. A second assessment may be made using the Highways Agency guide BD21/973. Strictly this is directed at the assessment of highway bridges and structures, but it offers generally applicable explicit advice on the assessment of cast iron (and indeed wrought iron and steel) structural elements. Once again this method of assessment does not demand any materials testing (except to justify enhanced strengths to be adopted for wrought iron and steel). Again the approach uses permissible stresses for cast iron, and generally combines this with serviceability loads.


Iron & Steel

Material Permissible stresses in N/mm2
Tension Compression Shearing Bearing
Cast iron 23 124 23 154
Wrought iron 77 77 62 lOR
Steel 1]6 IJ6 85 170 If further work is needed to justify the member, it becomes necessary to sample materials for strength tests. For such third assessments it is prudent to use a factor of safety of at least 3 on a 95% confidence limit UTS. It must be accepted at the outset that the results may lead to lower permissible stresses than might have been obtained already from the first and second assessments! A load test, strengthening, replacement, or relief of load are ways of achieving adequacy.

Assessment of cast iron columns may be made in the first instance using tables from the 1909-1930 London Act (which allows for slenderness effects), and then from BD21/97, which uses the RankineGordon formula.

Wrought iron and early steel

Early steel is taken here as being steel produced before the 1950s. First and second assessments will follow the approach in the 1909-30 London Acts and BD2l/97, broadly as described above for cast iron. Wrought iron and mild steel both being ductile, it is possible using BD21/97 to make allowance for some plasticity in the analysis for strength. This clearly represents a significant departure from the elastic analysis and permissible stress approach of the 1909-30 London Acts, in which the permissible stresses were based on the UTS divided by a factor of safety of four. BD2l/97 aligns itself closely with the Institution of Structural Engineers report on appraisal of existing structures-l, which goes into detailed consideration of the issues affecting choice of partial factors, etc, BD21/97 gives partial factors for materials and loads, and a clear recommendation for the yield strengths to be used for wrought iron and mild steel.

Property Wrought iron Mild steel
Partial material factor «gm) l.20 l.05 -l.3
(varies for different stresses)
Partial load factor (gf): structure 1.05-1.15
Partial load factor (gf): other dead loads 1.2
Partial load factor (gt): live loads 1.5
Characteristic yield stress (N/rnm2) 220 230 Figures for the characteristic yield stress of wrought iron and mild steel show that the two are virtually identical at the ultimate limit state - dictated by yield strength rather than by UTS. The 50% higher permissible stress of the 1909-30 London Acts for mild steel compared with wrought iron is not wrong, but is based on comparisons of UTS.

For a third assessment, a modern limit slate code of practice may be used, adopting the suggested yield stresses given above or obtaining a value by material tests. Wrought iron may be assessed in the same way, provided allowance is made for its lower E value. In such cases it is recommended that the partial material factor for wrought iron (1.2, as given in BD2l/97) is retained. The same value, 1.2, is recommended for steel when making such assessments, to allow for the greater variability of older steels.

frail & Sled


Modern steel

Modern steel here is taken as steel from the 1950s to the present day. Such steel has generally been produced and designed in accordance with relevant modern standards. Subject to checking date of origin (and evidence for the use of high tensile steel), such structures may be first assessed to the contemporary design code of the time - which in Britain until 1985 was on a permissible stress basis. A second assessment would be based on a modern limit state code, again using available data on steel strength; a third assessment would be based on the same code, using a value for yield strength derived from tests.


Connections have, throughout the use of iron and steel in building structures, been made generally in one of two ways: by direct bearing (vertical and/or horizontal) and by basically ductile fixings such as wrought iron or steel rivets, wrought iron or steel bolts of various types, and welding. The assessment of connections can be addressed in one of three ways: using contemporary permissible stress approaches; using modem, sometimes modified limit state or pseudo-limit state approaches; and using the latter approach supported by results from materials testing. (The first two assessment methods are conceived to avoid materials testing, which is particularly relevant to connections where, by their nature, removal of samples for testing may have unacceptable consequences for subsequent strength.)

The behaviour of connections is, as noted, essentially ductile assuming appropriate materials have been chosen and used. Connection assessment is generally basic in form. A first assessment may be all that can prudently done for connections to cast iron members, in view of the risk of brittle failure. For wrought iron and steel a second assessment could be based on a plastic model, with member and connections assumed capable of yielding up to full plastic joint behaviour. Care should be taken where parts of the connection are at a distance from its centre of rotation, when the remote components might fracture before full plasticity has been mobilised throughout the conneetion.

Identify work needed

Existing defects or deterioration can be remedied by maintenance, repair, or renewal, and can be minimised and dealt with in the future by a programme of planned inspection and maintenance.

The needs of refurbishment, reuse, or minor repair (e.g, inereased loading) can be met - if necessary - by strengthening or replacement. For heritage structures, load testing is a valuable way of justifying the adequacy of existing structural fabric, but it is only worth doing when calculations produce an inconclusive result. It is a waste of resources to test a structure that is clearly inadequate.

Report on findings and recommendations

The 'report' is the culmination of the study. Its title, format, and purpose will vary depending on what the work is and for whom it is being done. The report may form a schedule as part of the maintenance manual for the future care of the building, it may record the building's condition and defects (for example if the report is for a prospective purchaser); or it may make proposals for the intended refurbishment or adaptation of the building.

For maintenance work, a maintenance schedule should be prepared. This will indicate what planned maintenance is required at what notional frequency, although the actual timing of such work will often vary depending on inspection findings and the timing of other work in the vicinity. Procedures for inspections and maintenance work should be documented alongside the original condition survey, so that future work can be planned and implemented on a systematic basis. This record, described in BS7913: 1998, Guide to the principles of the conservation of historic buildings5 as a conservation


Iron & Steel

manual, may also contain plant maintenance and operating instructions, health and safety information, warranties, and details of the designers, manufacturers, contractors and specialists who have contributed to or worked on the building. A computer-aided system will allow the database to be used as an active management tool, for example by giving a 'prompt' on inspection and maintenance commitments to be undertaken in the near future. Between inspections, a log book should be maintained for the routine recording of inspections, and maintenance and repair work.

Carry out the necessary work

The work can be considered in the following categories, of increasing scale and scope.


Cleaning has two roles: to remove potentially damaging deposits; and to restore appearance. The former includes clearing blocked gutters and plant growth on masonry, which can lead respectively to water damage and to physical erosion of mortar joints. Bird droppings can damage fabric, but are not usually removed for this reason alone. Airborne dirt can damage metal components such as window frames, but again they are more likely to be washed down for the sake of appearance, periodically as part of the regular window cleaning routine.

It is often preferable to 'undercleari' rather than to 'overclean', to avoid the risk of damaging original fabric by over-vigorous cleaning techniques and abrasive materials (such as grit-blasting of cast iron, which can remove the durable vitrified surface). English Heritage supports this approach. BS7913 advises: "While light cleaning and redecoration can form part of a normal maintenance regime, wholesale cleaning should seldom be undertaken on aesthetic grounds alone, and should be considered only where surface deposits can be shown to be harmful to the fabric".


Repainting is effectively preventive maintenance, by enhancing protection against the weather, particularly for iron and steel components. Of course it also restores appearance. Proper preparation is essential for durable paintwork.

Access for repainting is often not easy. Health and safety issues must be considered. Safe access to higher levels will often require expensive scaffolding, etc. The combination of difficult access and relative frequency of repainting makes this the single most expensive maintenance task, so to be costeffective, this work should be co-ordinated with other work at higher levels such as inspections and repairs.

Minor repairs

Minor repairs inelude work to make good local damage such as slipped or damaged roofing slates, making good local rot damage in a timber window frame, or repairing a torn lead flashing. This can usually be carried out using small-scale and/or short-term access methods (and can often be combined with inspection).

Works for refurbishment and reuse

These works are considered in other papers to this Congress. Approaches based on British experience are described in a Steel Construction Institute guideo,

Major repairs and replacement

Major repairs and replacement are similar to works for refurbishment and repair, as they too are characterised by larger scale and longer duration. In general, effective planned maintenance should

Iron & Stee!


minimise the need for 'unplanned' major works, although some will be necessary due to the finite lifespan of building materials - and there is always the possibility of some serious event (e.g. a bomb) causing severe damage to the building.

Liaise throughout with involved parties

This is very important for heritage structures. Early discussion with the local conservation officer and the building control officer (if necessary) will establish the 'rules of engagement' for work on the building, just as it will help relationships if everyone affected by the work is kept informed of what is going on (not least the present occupiers of the building).

In Britain, building regulations require structural appraisal of buildings to be undertaken in many situations where there is a change of use, particularly if the new use is 'public' (e.g. a hotel) or floor loadings in the new use will be increased. Approved Document to the Building Regulations, A: Structure7, cites BRE Digest 366, Structural appraisal of buildings for change of use8, to which the author contributed. This follows the same procedure as described above.

Guidance on Maintenance and Minor Repairs

An active and planned approach to maintenance is very desirable so that potential problems can be identified and dealt with before they become more serious and more expensive to remedy. 'A stitch in time saves nine' or, as William Morris put it in regard to historic buildings, 'Stave off decay by daily care'. A planned approach minimises disruptions to the use of the building, as inspections and routine maintenance work can be arranged as far as practical to suit the occupants. 'Reactive' maintenance to deal with emergencies, on the other hand, will necessarily have to be carried out quickly and will almost certainly be more disruptive because the scale of work is likely to be greater.

Sound general guidance on the approach to building maintenance is contained in the UK in BS821 0: 1986, Guide to building maintenance management9. Technical guidance on maintenance and lifespan of materials and components gets variable coverage in British Standards, but there m'e numerous other reliable publications on the subject. Among these, Conservation of historic buildings by Bemard FeildenlO deals with both procedural and technical guidance on the maintenance of historic buildings, while Building maintenance and preservation: A guide to design and management edited by Edward D. Mills I I contains contributions on various aspects of maintenance.

Of particular relevance is BS7913, which addresses routine maintenance and housekeeping (under the heading of Systematic care) as well as repair and restoration work. Appendix D offers guidance on conservation manuals, log books, and the five-yearly detailed inspections. An English Heritage publication Developing guidelines for the management of listed buildings (1995)12 gives advice on a collaborative approach that can be developed between building users and the local planning authority. (Listed building consent may have to be obtained from the authority prior to undertaking alterations, or even repairs that affect the appearance or fabric of a listed building.) This approach can be useful, in order to define what work may be done on the building without such consent being necessary, and to establish 'ground rules' for more substantial changes.

None of these documents is mandatory. but together they can be said to represent 'best practice' in Britain.


Iron & Steel

Guidance on Refurbishment, Reuse, and Major Repairs General guidance

A guide to the appraisal of existing structures was first published in 1980 by the Institution of Structural Engineers, with a second edition in 1996 (Ref. 4). It provides excellent guidance on the procedure and process of appraisal and the main methods of survey and testing. In 1986 the Construction Industry Research and Information Association published a guide to the structural renovation of traditional buildings13. This covers building structures of the late 18th to early 20th century, with guidance on the approach to structural appraisal. It has informative sections on construction in masonry, timber, and cast and wrought iron and steel, as well as advice on foundation investigation and strengthening and on facade retention.

Specific guidance on iron and steel

Dealing specifically with the appraisal of existing iron and steel structures is the guide published in 1997 by the Steel Construction Institute (Ref. 6). It attempts to combine guidance on appraisal procedure with a comprehensive treatment of the ferrous structural materials and their characteristics and usage in building structures. A case history using the approach set out in this document is presented at the Congressl4.

More recently, the Institution of Civil Engineers has published a guide on the structural appraisal of iron-framed textile mills15. This is a fruitful collaboration between a historian of technology and a civil engineer who has led recent research on the properties and behaviour of structural cast iron at the University of Manchester Institute of Science and Technology. It includes results hom some remarkable load tests on an existing iron frame in a mill awaiting demolition, which give encouragement that these structures are more robust than might be thought, and shows how they can easily be strengthened to satisfy modem requirements for structural integrity.

Much useful information on the properties of iron and steel, early section dimensions, original loadings, design stresses, and design methods me contained in a historical structural steelwork handbook published in 1984 by the British Constructional Steelwork Association.

Supplementary guidance on hridge assessment is given in an Advice Note issued on behalf of the Highways Agency and other transport authorities in the United Kingdom16.

Other useful anglophone sources are listed below (17 18 19 20 21 22 23 24.)


1. The Cambridgeshire guide to historic buildings law (9th edition). Cambridgeshire County Council, Shire Hall, Castle Hill, Cambridge CB3 OAP, 1995.

2. Bates, W. Historical structural steelwork handhook. British Constructional Steelwork Association, 4 Whitehall Court, London SWIA 2ES, 1984.

3. Highways Agency et al. The assessment of highway hridges and structures, Departmental Standard BD21/97. The Stationery Office, POBox, 276, London SW8 5DT, 1997.

4. Appraisal of existing structures (2nd edition). Institution of Structural Engineers, 11 Upper Belgrave Street, London SWIX 8BH, 1996.

5. BS 7913: Guide to the principles of the conservation of historic buildings. British Standards Institution, 389 Chis wick High Road, London W4 4AL, 1998.

6. Bussell, M. Appraisal of existing iron and steel structures. Steel Construction Institute, Silwood Park, Ascot, Berkshire SL5 7QN, 1997.

7. Department of the Environment and the Welsh Office. The Building Regulations 1991: Approved

Iron & Steel


Document A: Structure. The Stationery OiTice, 1994.

8. Digest 366: Structural appraisal of existing buildings for change of use. Building Research Establishment, Garston, Watford, Hertfordshire WD2 7 JR, 1991.

9. BS 8210: Guide to building maintenance management. British Standards Institution, 1986.

10. Feilden, B. M. Conservation of historic buildings. Butterworth Architecture, Linacre House, Jordan Hill, Oxford OX2 8DP, 1982.

11. Mills, E. D. (ed.) Building maintenance and preservation: A guide to design and management, Butterworth, 1980 (and later editions).

12. Developing guidelines for the management of listed buildings. English Heritage, Customer Services, 429 Oxford Street, London WIR 2HD, 1995.

l3. Structural renovation of traditional buildings. Construction Industry Research & Information Association, 6 Storey's Gate, London, CIRIA, 1986.

14. Bussell, M. N. and Robinson, M. J. Investigation, appraisal, and reuse of a cast-iron structural frame.

The Structural Engineer, 76(3), 3 February 1998, pp. 37-42 (published by the Institution of Structural Engineers).

15. Swailes, T. and Marsh, J. Structural appraisal of iron-framed textile mills: ICE design and practice guide. Thomas Telford Publishing, 1 Heron Quay, London E14 4JD, 1998.

16. Highways Agency et al. The assessment of highway bridges and structures, Advice Note BAI6/97.

The Stationery Office, 1997.

17. Chrimes, M. and Thomas, R. Reading list on cast iron, wrought iron and steel frame construction.

The Institution of Civil Engineers, Library and Information Services, I Great George Street, London SWIP 3AA.

18. Smith, R. J. (ed.) Innovations in steel: New life for old buildings around the world. International Iron and Steel Institute, Rue Colonel Bourg 120, B-1140 Brussels, Belgium, 1993.

19. Trebilcock, P., Lawson, M., and Smart, C. Adaptability in steel. British Steel Sections Plates & Commercial Steels, Steel House, Redcar, Cleveland TS I 0 5QW, 1994.

20. Brereton, C. The repair of historic buildings. English Heritage, 1991.

21. Michell, E. Emergency repairs for historic buildings. English Heritage and Butterworth Architecture, 1988.

22. Porter, A. et al. The behaviour of structural cast iron in fire: A review of previous studies and guidance on achieving a balance between improvements in fire protection and the conservation of historic structures, In English Heritage Research Transactions: Volume 1: Metals, James & James (Science Publishers) Ltd., 35-37 William Road, London NWI 3ER, 1998, pp. 11-20.

23. Lawson, R. M. and Newman, G .. M. Fire resistant design of steel structures - a handbook to BS5950:

Part 8. Steel Construction Institute, 1990.

24. Ferris, H. W. (ed.) Iron and steel beams 1873 to 1952. American Institute of Steel Construction, One East Wacker Drive, 400 N0!1h Michigan Avenue St. 3100, Chicago, IL 60601 2001, USA, 1990.


[ron & Steel

Appendix: Identification of Structural Materials

The structural material will generally be grey cast iron, wrought iron, or steel. These materials can be distinguished principally by appearance. Dating may he helpful to support visual evidence - particularly in separating wrought iron (in use from the late 1840s to 1890s) from steel (used from the 1880s to the present day). Visual inspection and dating will, with experience, allowing cast or wrought iron or steel to be identified. Sampling will be needed if confirmation is needed, especially in distinguishing wrought iron and mild steel, which look very similar, Sampling will give definitive confirmation of material, hut may not he necessary if the visual and dating evidence is sufficiently strong.

The tables below offer guidance on recognition by appearance, and a general timescale for British use of structural iron and steel.


Remember that repeated overpainting may disguise a characteristic texture; it may be necessary to remove paint locally or find an area free of paint. Also, corrosion may distort the surface texture; corrosion itself has a distinctive texture.


Britain pioneered the structural use of cast iron in the late 18th century. After this, continental European nations quickly adopted this new material, and (especially in France, Germany, and Russia) were leaders with Britain in developing and applying wrought iron, and later steel, to structural use also.

It is often possihle to date the huilding itself from documents including original drawings, and from the visual evidenec of its architectural form and detailing which - with experience - can place the building within a decade or two. Care is needed when, as is common, a structure has been altered in its lifetime or where structural elements have been reused.


Drilling swart' may he sufficient where wrought iron or steel are suspected, and can be analysed chemically by a specialist testing house. Or a sample may he obtained for metallographic inspection by sawing or core drilling. Percussive drills must not he used on cast iron, known or suspected, because of the risk of shattering. A piece about 2Smm square is ideal, taken from a lightly-stressed area. Flame cutting may damage or weaken cast iron, while with wrought iron or stecl the resultant heat affected zone will give a misleading picture of the metal's structure.

Main period of use in Grey cast iron Wrought iron Steel
Bridges 1770s to 1860s late 1840s to 1880s 1 88 Os to today
(arches and beams)
Beam; 1790s to c.1870 late 1840s to e.1900 1880s to today
Columns 1 790s to c. 1910 c.18S0 to c.1900 1880s to today 11'0/1 & Steel


Visua I Characteristic Grey cast iron Wrought iron Steel
Often pitted from mould:
blow- holes; straight
Surface te xtu r lines or steps from Smooth Srmoth: possible
(uncorroded) junction of half moulds millscale
(e.g. along axis of
hollow circular column)
Possible delamination,
Surface texture Uniforrnl y rough with rust flaking off in Possible delamination as
(corroded) layers (flat sections) or wro ught iron
triangular wedges (rods)
Relatively thick sections Thinner, sharper profile:
Element cross- often of ornate profile: Thinner. sharper o,l,U,L, ], or T section;
sectional profile beamsvs, 1, T, ] sections; profile:o,I,L,T, ], or Z solid circular columns,
colurrns cruciform (+) or sections later rectangular or
hollow circular section circular tubes)
Re-entrant corners As wrought iron, except
Corners of eleme nt rounded to control Flange corners sharp, tor recent l-sections with
shrinkage cracking other corners rounded sharp 9()O external
Rectangular or
trapezoidal in beams,
with larger tension Usually tapered flanges
flange and small/absent on l-sectious, thickest at As wrought iron, except
Flange section compression flange: web; equal flange sizes; for recent l-sections with
flange width and constant flange section parallel tl unges
thickness may vary along
element (largest at
Often v uri es along
length: beam, wi th Constant along length
variable web profile and unless compound As wrought iron; recent
integral stiffeners: beam(s) or plate girder plate girders may have
Element profile COlOO1l1S often with when web profile may web and flange plates of
orruue Classical heads, vary (plate girders only) various thicknesses and
spigots, extended 'tables' and flange plates depths bun-welded
supporting beam" increase in number and together
baseplates and size towards midspan
intermediate stiffeners
l-sections up to O,3m l-secrions up to O,6m
Large beams (over say deep, occasional! y, deep before 1962, up 10
O,9m or more since then;
Section size 10m) often cast in slightly more; deeper solid circular columns
sections, bolted together sections plate girders up 10 O,3m diameter;
at tlanged junctions nude up of plates, flats, hollow tubular columns
and angles up 10 O,5m side length
Rivets (up 10 1950s);
bolts in clearance holes
Rivets for all built-up (earlier square, later
sections; bolts; flats, hexagonal heads);
Connection methods Coach bolts typically bars and rods sorretimes welding (20th century);
harrmer-welded together close tolerance bolts
in older structures (since First World Wm');
high strength friction grip
bolts (since 1950s)
Maker's name and
Identification on location often cast onto Rarely, at intervals on Often, at intervals on
element elerrent; load capacity roUed sections rolled sections
sometimes given 40

Iron & Stee!


Prof G. Sedlacek. B. Kiihn RWTH Aachen. Germany


The basic material data used in structural design with steel are strength and strain properties, fy' fu' A;, do,

fabricability (weldability, coldformability, ... )

temperature resistance (fire, heat, low temperature) corrosion resistance (coating, painting, weathering steel, ... )

The rehabilitation of steel structures needs a design that addresses these items. A particular risk from old steels originates from its suspected brittleness.

The method to avoid brittle fracture is in general related to the temperature resistances, in particular to the resistance to tension in association with low temperatures, see section 2.

The main steel property relevant for the tension resistance with low temperature is the fracture toughness that is specified for new steels in the material standards by the minimum energy A, (e.g. Av = 27 Joule) for Charpy V-impact tests at a given testing temperature TA" e.g. Tm = -50 DC. For such new steels (produced after about 1960) a reliable method is available in ENV J 993 - Part 2 - Design of Steel Bridges, that correlates this temperature Tm with the fracture mechanical toughness K" so that a safety check to avoid brittle fracture may be performed, when the action effect KapPld is known. This method is presented in section 3.

For old steels however, e.g. for steels used in rivetted structures delivered to the material standards of that time, toughness values were not specified, so that they must be produced from test pieces taken from the structure. In these cases fracture mechanical toughness values as J, or K, may be determined directly and the use of any correlation to the Charpy- V-energy A, is not necessary. The method used in these cases is given in section 4.

Relationship between the toughness-temperature curve and the tension resistance-temperature curve for notched tension specimens

For ferritic steels the toughness property is a function of the testing temperature as illustrated in fig. 1.





: Elasto-plaslic behaviour

I of steel slructures wilh

r flaws and weed discon-

: tinuities

Validity of the design rules

r • r



Temperature transition behaviour: Upper shelf behaviour


m Temperature

Figure 1: Shape of the toughness-temperature curve for ferritic steels

Iron & Sled



'K~PP\d ~ KMill•d
~ 1
K:pp1 s 20 + [70{eXp TEd - T27J ;;SOC-lHlI }+10 ](b~~J"4
I Te-d ~ Tcc 1
(K,pp],d -201 (bOff )'
T&] ~ [T27J -18"C]+52ln kR6 .I 25 10 +.i~.Tn
70 Assessment scheme

~ Lowest airtemperature in combination with Oed

Tm1n :::-2SoC

~ Radiation loss


~ Influem::e of material toughness: 1;oo=I;7-18[OC]

~ Influence of stress, crack imperfection and member shape and dimensions

(__l(.",,_-20)'(~)~ _lOJ

k -p 25

l!.Tf =52·Cn Jj~ 70 roC]

Figure 4: Transformation into a verification formula based on temperature values and final assessment scheme

s-tnnuence of the strain[rate. J'" sr ~ 1440-1,(1), in.;'-

~ 550 .eo

... Additive safety element "'J;~-7°C(fiirp~3,8)

This assessment format allows to balance the action TEd with the resistance T Cd that is composed of various additive terms including T27J and the safety term dT, Also strain rate effects and effects from cold forming may be considered by additive tempcrature elements (temperature shifts).

The procedure in figure 4 is the basis of the method for the choice of material to avoid brittle fracture specified in ENV 1993 - Part 2 - Design for steel bridges and proposed for EN 1993 - Part 1.10 - Design for fracture. K"PPld has been determined in a safesided way for a severe fatigue detail, see figure 5. Assuming a semielliptical surface crack at the hot spot that may have grown from its initial detectable size (aJ2co) to its design size (a/2e) from fatigue loads equivalent to one quarter of the full fatigue damage applicable to the detail. Such a fatigue damage may bc assumed to represent the service load between at least two subsequent inspections that should be carried out for structures subjected to fatigue.


Iron & Steel

Fig. 5: Fracture mechanical model chosen for determining the permissible plate thickness in Euroeode 3 - part 2

Hence the fracture mechanical safety assessment used for the choice of material does not only represent an ultimate limit state check for low temperatures; it is also a check for damage tolerance because it verifies that the structure is safe in periods between inspections as long as the assumed detectable fatigue cracks (aj2,o) are not found at the inspections. Such damage tolerant structures may be kept in usc even when the nominal fatigue life of the structure determined with the usual fatigue strength from Wohler-curves is exceeded.

Of course the method can also be used with material data and the member geometry as given for a specific case in the course of a unique verification.

Fracture mechanical safety checks for old steel structures

When minimum toughness values arc not available for old steel structures, the material data should be determined from plate samples as given in figure 6 , from which both strength values and Lvalues may be obtained.

Figure 6: Miniaturised plate sample


~ 2 - specimen for chemical analysis

3 - 1/2 CT sample for fracture mechanics test

4 -load direction in the tension tie

Figure 7 gives a scheme for the material identification of old steels from specimens in view of production methods. Wrought iron has similar properties in chemical composition and microstructure as low strength-low alloy steels of today and is applicable to the fracture mechanics safety assessment. Puddle iron has a totally different microstructure, which can be characterized as laminar type, built up from ferrite and slag.

Based on a statistical evaluation of the chemical and the microstructure properties of 407 specimens

Iron & Steel


S~oI St«l. ",O<!oe¢d from. law - 1930

lOCi" Sill In<! ... ._ or>41Qf Mn<Cl,I% 0>0.5%

Irorn riveted bridges in Germany it could be concluded that the strength and toughness data of wrought iron can be treated as a statistical homogenous population. The statistical distribution of the material strength has been derived from 205 tests at O°C and 283 tests at -30°C. Table I shows the results in terms of mean values, standard deviations and fractiles. The log-normal distribution fits best for the statistical description of the characteristic strength.

Figure 7: Identification scheme for old steels

Rrt. R_ R ... R~ A Z J"" J<rll
T["C] -30 --30 0 0 0 0 -30 0
Typ Log. Log. Log. Log. NV NV Weih. Weib.
II x~ 257 385 248 374 26 57 17 30
"UI 310 446 293 423 34 66 62 91
375 516 345 479 41 75 Log. '" Log-normal distributed. NV = normal disuibuled,

Wcib. = Weibull distributed (3'parameter), R..:; Yield Strength, R. c Yield Stress. A. = Fracture Elongalion. Z:::: Reduction of Area. J,,.,. Fracture Toughness ace. [16]

Table I: Characteristic values of strength. fracture strain and fracture toughness distributions for wrought iron from 412 tests


11'011 & Sleel

For safety assessments the characteristic values Rei = 257 N/m2 and Je';1 = 17N/mm may be used unless more informations are available.

As the fracture toughness is rather low, the safety verification should be performed on the basis of J values without the inclusion of correlations which due to their scatter would enhance the necessary safety factors and hence reduce the economy.

The verification may be performed using either a finite element analysis of lapPi for the detail and crack configuration given or using standard formulas for fracture mechanical models that can be applied to represent the details, see figure 8 and figure 9.

r . intcgnlMn path around th~ crack tip W • energy Q~nsity

r ' sm:>!> ~relQT

<Is • ""om,;:m of (be imegrauon 1'",(11 Ii • displacement vector

FE . idealisation at the crack tip


<;'r., ''1'



=R" .(l-~W)

Figure 8: General procedure (FEA with J-Integral) or simplified procedure (with formula for basis acture mechanical models) illustrated for a plate with a through crack (centre crack CCT)

lron & Steel



lm!i~1 "mck ~r;t;:: -1.; ~ (I)' 10.1/1 u=Uov..-, C"::'::2-::k :"i.i~ ..!.--"'~ eli width:W 02


lmUd n<1.CK ~"E: u., (D. ;0,.2 max. aJ]O\\.·_ cruel, size: It"",~:4~ (\'2 Plute \ViiIl!>: W" L 1-(';2




Initial nad "i.m. ',,', (D-10r:;? mnv ;ilk'i'J.:. cnu-k :~l:l&~' ~I- c/:! [';"le width: \V = I.e J II )!2

Figure 9: Association of fracture mechanical models with structural part of built up cross-sections (here angles)

The procedure is in general iterative, because for a given material toughness Jc<it determined for the structural element a critical crack aC<<l size has to be found, for whieh the safety equation Japlld < ] cntd is ful1llled, see figure 10.

Figure 10: Determination of acrit by iterative variation of a-values


11'011 & Steel

Once Jccitd is known the permissible intervall time between inspections may be determined from da = ace;t - ao by calculating the crack growth da from the fatigue load, figure 11.

Figure 11: Principle for the determination of minimum service time N(.p)

The usual assumption for detectable crack sizes which most frequently are used in the verifications is given in figure 12.

Figure 12: Assumption for the initial crack size ao for a) angles, b) plates covered by angles

In case the steel structure is not subjected to fatigue load the values in figure 12 may also be used unless other assumptions are made depending on the detection methods applied.


In the rehabilitation of steel structures the question arises whether the use of general design rules for steels may be restricted because of a suspected brittleness of the material originating either from the material production or fatigue effects.

The paper gives a method how to perform a safety check to avoid brittle failure both for new steels, for which toughness values are specified in material standards, and for old steels, for which these values are in general not known a priori.

By the safety check to avoid brittle fracture the structure is proved to be "damage tolerant", This feature

Iron & Sfeel


permits to keep the structure in use even when the full nominal fatigue resistance of a structure is reached.


[I[ G. Sedlacek, N. Srranghoner, M. Schafers, W. Dall, P. Langenberg, B. Kalinowski, J. Brozzetti, A. Nussbaumer: Composite Bridges Design Improvement for High Speed Railways, Final Report ECSC Project 7210/SA/128.

[21 ENV 1993-2-Euroeode 3 Design of Steel Structures, Part 2: Design of Steel Bridges, CEN 1997

[3] G. Stetzel, G. Sedlacek, P. Langenberg, W. Dahl: Material Identification and Verification Method for the Residual Safety of Old Steel Bridges, IABSE Report Page 241-251.


Iron & Steel

Evaluation of structural safety of existing structures, based on actual material properties and geometry

fI: F.S.K. Bijlaaard

Head Division 'if Structural Fnginfering

TNO Bnilding (/1111 Construction Research. Rijswijk, The Netherlands


This paper presents an overview of the structural assessment of existing structures in comparison to that of newly designed structures. The role of measuring geometrical data and testing the material properties is explained. The method provides objective criteria to judge the safety of existing structures.


In most cases the structural safety of a structure is only evaluated in the design stage. This is to obtain a building permit. The evaluation is mainly based on using strength formulae in design codes.

For existing structures the structural safety is evaluated most times only if the material degradation (e.g. due to corrosion) endangers the integrity of the structure or in case that the use of a structure is changed and the loadings become more severe than they were in the original design.

This paper describes the starting points for the evaluation of the structural safety of existing structures together with the dillerences and similarities with the design of new structures.

This paper is based on literature reference [I], [21 and [31.

Safety level

Safety level for newly designed structures

The safety level of a structure can theoretically be described in the chance of failure during a relevant time period. This can also be expressed in the reliability index B.

This reliability index ( is in direct relation with the chance of failure (P) of the structure:


B = 2.0 P = 0.023

B = 2.6 P = 0.0047

B = 3.0 P = 0.0013

B = 3.6 P = 0.00016 B = 4.0 P = 0.000032

In principle it is possible to calculate the chance of failure of a structure in the design process, but it is rather complex because it is dependent on many parameters which are of stochastic nature such as actions due to wind and snow or 1100r loading, material properties such as yield level and tensile strength, imperfections such as eccentricities and out of straightness.

For the day to day practice it is only for special structures, like offshore structures, a custom to take these stochastic parameters direct into account. For more regular structures like onshore building structures most designs are based on design codes which on their turn are based on a certain reliability index level. This is realised by using safety classes, prescribed loadings, prescribed load factors, strength formulae and material properties which are normalized and prescribed material factors.

These parameters and factors are chosen such that at least a certain reliability index B is reached.

Iron & Steel


Safety level for existing structures

The evaluation of the structural safety of existing structures differs on some points from that of newly designed structures:

- increasing the safety level of existing structures is in many cases much more costly than for structures which are still in the design stage;

- the remaining period in which an existing structure has to function is most times shorter than the standard reference period of 50 years for new designed structures;

- the possibility to know the actual material properties and geometry of structural components via measurements is present.

In principle there are two reasons to put a condition to a reliability level for a structure, namely economy and safety for people.

The first leads to an economical optimization of the sum of building costs together with the product of damagc and chance of failure.

The second leads to minimize the risk of loss of human lives to a level significantly lower than other risks people run in daily life (driving cars, going by airplane, ctc).

Starting points for these considerations is statistical information on accidents,

In the case that the remaining period for a structure is shorter than that for newly designed structures the (-factor needs to be the same as for newly designed structures,

This shorter period docs result in a decrease of the design loads which have to be taken into account. The int1ucncc of time plays also a role in the aging of the materials used. In case of steel structures corrosion is a factor which have to be considered.

In the design stage many parameters are only known on paper.

Due to the variations in material properties and geometrical measures the coefficients of variation have an int1uence on the safety factors.

For existing structures there are two possibilities:

a) The old specifications do exist and there is no reason to doubt those values. In that case these design values can be used.

However having the existing structure present there is the possibility to establish the actual values for those parameters by measurements and testing.

These measurements often lead to higher values for the actual strength of the structure because in the design lower bounds for measures and material properties arc used.

b) The old specifications do exist, but there are reasons to doubt them because ofthe possible low quality of the realization of the structure, or the old specifications do not exist,

In this situation and to avoid unacceptable risk one has to use either the lowest reasonable values for those specifications or one has to establish the actual values by measurements and testing.

From these described situations it is clear that the calculated safety is not a properly of the structure as such, but goes along with our knowledge from the structure and the loadings acting on it.

Lack of safety can in fact be lack of knowledge, which can be re-evaluated by measurements and testing and using more advanced calculation methods.


Iron & Sleel

Measurements and testing Non destructive testing

In some cases it is not desirable or even possible to take material samples hom relevant places in the structure to determine the actual material properties.

Via the relation between NDT based measures such as hardness and for instance the yield point of steel the last can be established. From these always limited number of measured values a representative value for the material properties can he established by using statistical evaluation.

Geometrical measurements can directly give information about the actual situation. From these data the actual structural safety can he evaluated.

The actual structural safety can also be evaluated by full scale loading tests. It is advised to increase the test loading gradually step by step and the monitor the deformations carefully to prevent pre-mature failure. By analysing the measured deformation it can be decided whether or not to apply the next loading step in the test.

Destructive testing

Where possible well established destructive testing procedures can be used to determine the actual values for the relevant material properties such as the yield level, tensile strength, strain after failure, fracture toughness, etc.

After statistical evaluation of these test data, representative values of the relevant material properties can he established and used for analytical evaluation of the structural safety of the existing structure. For establishing the material properties of steel, such as the yield stress level and the fracture toughness it is referred to EN 10025 and 10 113-1. For the determination of properties of bolts it is referred to ISO 898-1 and ISO 898-2.

In these codes further reference is made to codes where actual testing procedures are prescribed. With respect to weldability it is referred to NEN 6774.


In Amsterdam an old steel structure, which had served as a storage warehouse for many decades, called the Brazilie warehouse, was renovated and made suitahle to serve as shopping centre [4].

By doing this a link was created between the past and the present.

The hearing structure is an old riveted steel structure consisting of columns and trusses made in 1916. The structure was broken down piece by piece because under the shopping centre a car park was built. Having the structure demounted there was an opportunity to inspect the steel components.

TNO Building and Construction Research carried out a technical inspection and found that the conservation of the steel structure was very good due to the fact that the warehouse was always used as a storage for coffee. Only structural components at the outside of the building were corroded severely and had to be replaced by new ones.

The other components were randomly investigated to establish the relevant material properties. The steel grade appeared to be comparable with S 235/R.

Based on the research results on the material it was decided that the material was not suitable for welding and the stress level needed to be limited to 60% of the yield stress level of S 235/R.

The old roof with concrete plates was replaced by a light weight roof of steel plates and isolation material. The side walls were completely replaced by modem shopping facades.

For reasons of fire protection it was not needed to protect the steel structure because there were enough escape routes and the fire resistance of the steel structure was sufficient.

Iron & Steel


All measures taken resulted in a modern shopping centre showing the beauty of an old riveted steel structure of the past, perfectly fit for use in the next century (see figure I).

Figure 1. Cross-section of the Brazilie shopping centre


A comprehensive methoci of evaluation of structural safety of existing structures is available to structural engineers in design offices and checking bodies.

This method allows us to apply objective criteria in judging whether an existing structure is fit to be used for an other goal, can be maintained for its original purpose, can be maintained for the rest of its economic life-time, hut only under restricted usc or even leads to the conclusion that the structure is not fit for further use and needs to he closed immediately.


[ 1] A.C.W.M. Vrouwenvelder "Assessment of existing structures"

JCSS probabilistic model code, TNO Bouw report no. BI-88-48

[2] A.C.W.M. Vrouwenvelder, P.E. de Winter

"Beoordeling van de constructieve veiligheid van bestaande bouwwerken" TNO Bouw report no. B-91-0832

[3] F.S.K. Bijlaard, J.W.B. Stark, G. Sedlacek "Determination of Design Resistance from Tests" Eurocodc 3, Annex Z, 1988


Iron & Steel


Jr. G.G. Nieuwineijer

Delft University of Technology, Facility of Architecture. The Netherlands


In the 19th century construction methods differed from those used today. Calculations were simple or sometimes not even made at all. Regulations were few or nonexistent, and engineers often worked from their own experience and insight.

Old structures do not usually satisfy current regulations and insight with regard to strength, stiffness and stability. Indeed, we must ask ourselves whether it is necessary that they should do so. Modifications may be necessary in relation to strength and stiffness when the original structure fails to meet modem requirements or because it has been weakened by corrosion, Adaptations can be made by strengthening existing elements, changing joints and by the addition or replacement of structural elements.

Ideas about stability were also different in the 19th century. Supplementary measures to historic structures include the addition of stability diagonals or the use of roof or facade elements such as diaphragms. Many old buildings are important for historical or cultural reasons. When restoring such buildings it is necessary to consider what they will look like later. It is necessary to consider whether it should be obvious that restoration has taken place or whether the original appearance should be retained to show how buildings were previously constructed. Often restoration is carried out pragmatically and the restoration methods used depend on the restoration philosophy and the means available. The objective should be to execute the adaptations in an honest way so that no falsification of history occurs.


Structures must conform to regulations relating to strength, stiffness and stability. When a structure is being designed, the appropriate sizes for the elements can be determined by calculation. Usually the regulations provide sufficient guidance in relation loads, material properties and the required safety margins for the calculations to be made. In the past things were different. When older structures are recalculated, they do not usually satisfy all the requirements relating to strength and stiffness. Moreover, the structures may have been weakened by corrosion.

In addition to strength and stillness, a structure must also be stable. That means that it may not collapse. The stability in the horizontal sense must be guaranteed. In the 19th century, ideas were different from current ideas, especially those relating to stability.


The first methods for calculating trusses were introduced by J.W. Schwedler and C. Culmann in 1851. These were analytical methods. However, the graphical methods of Culmann (1864) and Cremona (1872) were much more suitable for trusses.

In the Netherlands, the calculation of roof trusses started only in the third quarter of the 19th century, although calculation for bridges had begun a little earlier. Before that, dimensions were determined from experience. When calculation began, the estimation of the loads was left to the engineer himself. One of the earliest calculations in the Netherlands was that made for the roof spans of the platform of the Delftse Poort II Station in Rotterdam (1877). Although the estimates fell short, the complicated calculation was carried out correctly, At that time the bridge over the Lek at Kuilenburg (1868), now Culemborg, with a main span of 154 m, was the biggest bridge in the world. It was constructed as a triple truss, the calculation of which was not understood at that time.

11'011 & Steel


Figure I: Platform roof, Delftse Poort II station, Rotterdam, 1877. The complicated analytical calculation was carried out correcLly. The assumption that the end bearing was a stiff joint was not correct, Stability trusses were simply omitted for aesthetic reasons.

Figure 2: Railway bridge, Kuilenburg, 1868. The bridge is constructed as a triple truss, which could not be calculated at that time. It was assumed that each of the trusses I-III took one third of the loads.

Moreover, there was little insight into metal fatigue under dynamic loading. Whether building on the basis of experience or of early methods of calculation, engineers took some precautions. The test loading of roof trusses and bridges was general usage.

When testing trusses, only vertical downward loads, whether evenly or unevenly distributed, were considered. Little or no account was taken of wind loads, which may also be directed upwards. From the design of the platform roof span of the Gare Montparnasse, Paris (1840) it can be seen that it was not designed for horizontal wind loading.

For buildings with a complex shape, or those on which the immediate environs exert great inlluence, the regulations for wind loading provide little guidance. Only by investigating on site or by using models in wind tunnels can the actual loading be approximated. When investigating on site, one can consider placing strain gauges or deformation gauges at critical points. After sufficient information has he en acquired for various wind speeds and directions, the data can be processed statistically. The anticipated wind loading can then be predicted. By using wind tunnel experiments, it is possible to determine coefficients of wind loading by subjecting the model to various wind speeds. In long structures, considerable temperature stresses may occur if no provision is made to accommodate them. These may give rise to damage, especially at the ends of the structure.


Iron & Steel

Usually, government regulations stipulate that structures must meet the current regulations relating to strength. When people may he on or underneath the structure, this seems to be self-evident. However, we must ask ourselves whether it is indeed necessary to satisfy all these regulations.

With regard to fire proofing regulations in The Netherlands, a distinction is made between existing and new buildings. In the Netherlands too, the safety requirements for classic automobiles relate to the year in which they were built, while aeroplanes are built with a considerably lower safety coefficient than that required for bearing structures in huildings.

Figure 3: Platform roof, Gare Montparnasse, Paris 1840. In the calculation wind loading was not taken into account.

Figure 4: (Ie diagram for cast iron, wrought iron and steel.

When considering safety, one can look at the failure mechanism. A beam that is overloaded will show this by greater hending. An overloaded column may suddenly buckle and thus fall without warning. Lateral instability may also occur unexpectedly. A lower coefficient of safety can be used for bending than for lateral instability.

The importance of a specific structural element may also playa role. Failure of a secondary beam is of less consequence than the collapse of a column. An important consideration is whether the loss of an element leads to progressive collapse or to the redistribution of forces.

The type of material may also be important. Steel has a clear area in which plastic deformation takes place, after which strain hardening occurs. The large plastic deformation may indicate overloading. Wrought iron has a less clear now area and its failing strain is lower than that of steel. Cast iron fractures suddenly at tension with a very small strain.

This does not mean arguing in favour of different regulations limn those applying to new buildings. For historic constructions in which great value is attached to the original construction, existing regulations can be interpreted in a sensible way. Assumptions, calculation rules and methods can be refined and interactions with elements that are usually disregarded can be taken into consideration. If this does not lead to a satisfactory solution, one can think about accepting a lower safety factor in specific incidental situations.

Further considerations relate to restrictions on loading and, for bridges, speed limits. Maximum loading and speed must therefore be clearly indicated. In areas with high snow loading it may be possible to introduce heating so that the snow does not lie. In extreme conditions, a building must be closed to the public,

When a structure fails to meet the safety requirements the following measures may be taken: * strengthening of the existing structural elements

* changing the joints

* the addition of structural elements * replacement of structural elements

Iron & Sleei


The Palm House in Kcw Gardens (1848) consists of a central section with two wings. The arched wrought iron beams are l-shaped in cross section. They arc curved deck beams with hulb shaped lower flanges. Research in preparation for the restoration, which was completed in 1989, revealed that they did not satisfy current requirements. They were strengthened by welding strips onto the upper and lower sides of the bulh profiles and by changing the joints with the foundation. The strips on the upper sides of the profiles are obscured by other structures while those on the underside remain visible.

~= I :~

Figure 5: Palm House, Kew, 1848. The I-bulb profile is strengthened by welding steel strips onto the top and bottom. Originally the I-bulb profile stood in a cast iron base plate and the space between the main beam and the base plate was filled with lead. Strengthening has been achieved by embedding the lower end of the beam in concrete, producing a rigid joint with the foundation.

The beams of the wings originally stood in cast iron base plates. The space between the main beam and the base plate was filled with lead so that the end bearing behaved as a hinge. By embedding the lower ends of the beams in concrete, a rigid joint with the

foundation was created, with a more favourahle distribution of forces in the beams. After completion, this change was invisible. Corroded parts of the lower flange were removed and new ones were welded into position. In places where the body of the beam or the junction between body and flange were in a poor state, strengthening plates were welded on. After painting, these changes

in the structure were inconspicuous, but they do consciously show the restoration in an honest way.

In the restoration of the platform roof spans at 's-Hertogcnbosch (Den Bosch) Railway Station (1896), strengthening and changing of the joints has taken place. The joints in the tops of the roof truss could be approximated as a hinged joint. In the restoration, the lower flanges were linked by a strip so the

connection hecame a stiff joint.

At the same time, the sections to the left and right of the top were strengthened by welding on strips, This was also done in the truss joints.

For bridges, adaptation must be found by reducing the span. At the land ends of the tu bes of the Conway Tubular Railway Bridge (1848), new piers were built. These piers are in historic style and the way that they lead the forces into the body of the tube gives the impression that they are part of the original StlUC


I Figure 6a: Platform roof s'Hertogenbosch Station, 1896. Strengthening the top of the roof tIUSS.



frail & Steel

";t.L. 1Hi

Figure 6b: Platform roof s 'Hertogenbosch Station, 1896. Detail of roof old/new

Figure 7 (I): Conway Tubular Railway Bridge,

Conway, 1848. New piers were built in historicstyle to reduce the span

Figure 8 (r): Albert Bridge, London 1873. Extra piers have been placed in the middle of the main span

The piers introduced into the centre of the span of the Albert Bridge in London (1873), which was originally an cable stay bridge, are worse. In 1884, the bridge was strengthened by suspension chains, making it into a combination of a cable stay bridge and a suspension bridge. In 1973, the extra supports, constructed in historic style, were placed in the centre. Both piers are highly detrimental to the bearing system, but at least they permit continued use by modern traffic.

The platform roof spans of the Hollands Spoor Station, s ' Gravenhage (1893) consisted of two bays. The arched beams are trusses with a v-shaped strut pattern. Although the trusses meet the current stringent regulations a number were so severely affected by corrosion that replacement was necessary. The objective was to retain their appearance, even though modern materials and techniques were being used. During the restoration a quarter of the roof was destroyed by fire, so the deformed trusses were replaced in the same way. Because the greater part of the replaced structure is high above the traveller's level, these changes are scarcely noticeable. It is an honest reconstruction and only those with a special interest will recognise that restoration took place around 1990. Fortunately, it was decided not to weld on rivet heads or, worse still to glue on plastic imitation rivet heads, which might have made the structure appear more original.

Iron & Steel


t';'i>H ..... ",.+ ..

Figure 9: Platform roof, roof, Hollans Spoor Station, s'Gravenhage, 1893. Cross-section of the upper edge of a beam before and after restoration

Figure LO: Suspension bridge over the Menai Strait, NW Wales, 1826. The bridge before and after 1940, when the suspension chains were radically changed.

The deck of the Menai Strait Bridge in NW Wales (1826) was originally suspended from four chain cables, each of which consisted of four chains. In view of the increase in the traffic load and the danger of collisions with the changes between the traffic lanes, in 1940 all sixteen chains were replaced by two cable chains. These are on the outer sides of the bridge and each is made up of two chains. At the same time, the deck was also changed. The new chains are made of high quality steel and although the shape of the links remains about the same, they are bigger. From a distance the bridge looks the same as it did before, but seen from close to its image has been considerably changed.


In addition to requirements relating to strength, there are also requirements relating to stiffness. These may he a visual and psychological in nature. With inadequate stiffness, the finishing of a building may cause problems or the drainage of the precipitation may be impeded. In bridges, fatigue may playa role


//'011 & Sled

Figure l l : Tacoma Narrows Bridge, Washington, 1940. Because the construction of the hridges was too weak, torsion oscillations developed led to its collapse.

Figure 12: Block of flats in Delft. The building was stiffened by the addition of extra struts near the end facade

11'011 & Slee/

When the sagging of a roof is clearly visible under loading this will appear unpleasant and may create an impression of insecurity, even though its strength is not in question. The finishing construction must be able to withstand the deformation caused by the sagging of roofs and floors. Large-scale sagging may cause cracking in partition walls, while doors and windows may stick and panes of glass may crack. Precipitation remains standing on sagging roofs, which increases the chance of leakage, while the extra weight causes even further sagging. The possible accumulation of water is a determinant for the required strength. In the horizontal sense too, demands are made on the stiffness of a building. With inadequate stiffness, movements resulting from wind loading will be appreciable. Such horizontal movements may he unpleasant and induce a feeling that they are unsafe. Here too, the finishing construction may cause difficulties.

The psychological aspect also plays a role in bridge design. The railings are usually much stronger than they need to he from the point of view of strength. Strong vibration and hig oscillations must be avoided, hoth from the psychological point of view and in view of the danger of fatigue. Stiffness is particularly important for suspension bridges,

The maximum permitted deformation for a structure must be determined. For buildings, regulations in which the maximum bending of beams and the maximum horizontal deformation of buildings are defined provide guidance. In addition, it is wise to investigate what the finishing construction can bear and whether precipitation drainage is safeguarded. This includes determining how much movement can be accommodated in joints between the panels of partition walls and how much deformation door and window frames can bear before they stick.

This is usually more difficult for bridges, The dynamic hehaviour of large bridges and of their parts should be investigated. In the past bridges, in particular, often had a structure that was too weak. Although a lesson could have been learnt from the past, fatigue caused the collapse of the Tacoma Narrows Bridge. Even while it was under construction in 1940, movements that were not taken seriously OCCUlTed. During a storm shortly after its completion, torsion oscillations developed which caused the failure of beams in the deck and the suspension cable. This led to the collapse of the bridge deck. Stiffening measures are often the same as those used to strengthen structures that have been mentioned previously. Two examples of this are given below,

A hlock of flats in Delft, built in 1968 was too weak in the longitudinal direction. The building was later stiffened by the addition of extra struts near the end facade.

The Menai Straits suspension bridge in NWWaies (1826), with its span of 579 ft (almost 180 m) has been stiffened several times. The light wooden deck had already been replaced by a stiffer wooden deck in 1839. In 1940, this deck, together with the suspension chains, was again replaced by a stiffer one. This indi-


cates that there has been a continuing struggle with the stiffness of the bridge. The designer Thomas Telford thought that it was impossible to build suspension bridges with spans exceeding 600 ft.


When structures are being designed. stability is an important consideration. The stability must be guaranteed or in other words, the structure may not collapse. In modern steel structures, stability is usually ensured by the use of stability diagonals. In large halls, these usually consist of crossed diagonals, the size of which is related exclusively to tension. In multi-storey buildings and bridges, Kvstruts are also used. Sometimes stability is obtained by using portal frames with rigid comers. So far, we have been concerned only with external stability but structural elements can also be internally unstable. This happens when they buckle or when lateral instability or folding occurs. To prevent the buckling of flanges of beams they must be supported at regular intervals. Connections known as stability diagonals are used for both beams and the upper edges of bridges.

In the 19th century, there were different ideas about stability. Less attention was paid to it and stability diagonals were frequently omitted because they were thought to be ugly. Both analyses of buildings and desk studies reveal that there was often something wrong with the stability of structures.

Diagonals were sometimes present but not well executed. In the platform roof of Haarlem Railway Station, the diagonals are fastened to a ring that is not stiff in its own plane. If there are no diagonals in the roof of a hall, this function may be taken on by the roof plates. This is often the case with tongued and grooved wooden covering and is certainly so when the covering is laid diagonally. The corrugated sheets that were frequently used in the past, and even glass, may have some function in providing stability. This may cause problems, especially when glass is used. Glass is a strong material but it has a brittle failure character and high notch sensitivity. It is thanks to the glass that the greenhouse at Bicton Gardens in Devon remains standing without any problem. In the exhibition halls in the Pleasure Gardens (Pare du Cinquanenaire) (1888), Brussels, there are no stability diagonals where there are glass plates. During heavy storms the windows sometime crack. Owing to its high strength, under some conditions glass is suitable for use as a stabilising diaphragm. This is the case in the pavilion designed by Benthem and Crouwel for Sons beck 86, that now stands on the premises of the Autotron in Rosmalen. In addition to its use as a stabilising element, glass is also used as a bearing clement in the columns.

Figure 13: Modern hall with stability struts. These struts in the planes of the facade and roof ensure that the hall is stable in both longitudinal and transverse directions and cannot collapse.

Figure 14 (r): Platform roof Haarlem Railway Station 1908. The stability struts are fastened to a flexible ring and therefore function poorly.


Iron & Steel

Figure 15: Exhibition Halls Pare du Cinquantcnairc, Brussels, 1888. There are no stability diagonals where there are glass plates. Some windows arc broken or replaced.

Figure 17: Liverpool Street Station, London, 1875. Only the old trusses could be retained. The secondary and tertiary girders, sheet glass and roof sheathing are replaced by modem equivalents. An interesting combination of old and new.

Iron & Steel

Figure 16: Pavilion Baltard, Nogent sur-Marne. During the reconstruction Stability-struts were added.

Figure 18: Hoge Brug, Zwolle, 1883. The connection between the upper edges is extended by a (rave on each end.


Further research in this field is now underway at Delft University of Technology. In particular laminated glass, which is mounted in silicon adhesive, seems to have a good potential. If the stability of an existing structure is not up to standard, stability diagonals can be added or the stahility can be provided by stiff planes.

In 1975 the demolition of the Central Market Hall (1866) of Paris was started. One of the pavilions was dismantled and reconstructed in Nogent-sur-Mame, near Paris. During this reconstruction, measures were taken to ensure stability. On the facades heavy struts were mounted in the corners on both the ground 1100f and in the roof structure. Diagonals have also been added under the roof panes, which are made of profiled steel sheets. The struts in the comers are pronouncedly visible.

If supplementary measures in the ostensibly stiff facade surfaces had really been necessary, they could have been taken in a more elegant manner. The profiled steel plates in the roof could also have worked as diaphragms, so the diagonals there would also have been unnecessary.

There were no stability diagonals in the platform roof of Padding ton Station, London (1854). The roof plates were replaced several times, the last renovation being in 1990. On this occasion, coated COITUgated iron with the same profile as the original galvanised corrugated iron sheets of 1854 were used. The plates are fixed with self-tapping stainless steel screws and work as diaphragms.

To prevent horizontal instability, stability diagonals are fitted hetween the upper edges of the Hoge Brug, Zwolle (1883). For reasons of safety, during the restoration in 1993, these were extended by the addition of a trave on each end. This was permitted by the profile of the free space and the upper edge is now laterally supported at more points than was previously the case.


When restoring old structures it is often necessary to made modifications with regard to strength, stiffness and stability. Because many such structures are important for historic and cultural reasons, these modifications must he carried out with care. It is important to decide how much of the restoration work may be seen, considering whether the original appearance should be retained, to show how buildings were previously constructed or whether the changes should be visible. The choice will depend on the historical value, the state of the structure and the available financial means. The modifications should he made in an honest way so that no falsification of history occurs.

When minor changes have to be made to valuable structures, these changes must be executed in such a way that they are only visible to the experienced eye. If many elements must be replaced this can be done in such a way that an interesting combination between old and new is created.


Material from the archives of the History of Structural Design Group, Faculty of Architecture, Delft University of Technology.


Iron & Steel

iron & Steel



Iron & Steel


Eur.lng, cu.owau». ESc. C.Eng. MIMecil.E. Dorothea Restorations Ltd,

Bristol, United Kingdom


This paper covers some practical techniques used in the conservation of cast and wrought ironwork used for structural and decorative purposes. It is based on the experience of Dorothea Restorations Limited over 25 years working on historic structures throughout the United Kingdom.

Cast iron is a brittle material poured molten into sand moulds where the shape has been formed by a wooden pattern. Wrought iron is ductile and malleable, formed by forging and rolling. As their properties compliment each other they were often used together, cast iron for compression and wrought iron for tensile members. Both were "hand made" and the control of their quality during manufacture was subjective rather than scientific. Original components may therefore be of variable quality with defects such as blowholes, inclusions, and cracks in castings, and in wrought iron laminations, slag, and defective fire-welds.

Assessing the Problem

It is therefore important to survey the metalwork fully to establish the quality of original materials and the nature and extent of their deterioration, by answering the following questions:

- Which parts are cast/wrought iron?

- How were the structural loads intended to be taken originally?

- How are they taken now?

- Cautious repair of historical cast iron constructions- What was the original strength/integrity of the


- How have they deteriorated through

- Deformation, (settlement, applied loads, rust-jacking, etc.)

- Mechanical or thermal shock (impact, fires, past repairs, etc)

- Corrosion, ( and possible graphitisation in cast iron)

This exercise should be undertaken diligently whether for a humble gate or a grand railway station, to avoid more conservation work being carried out than is absolutely necessary. Pay particular attention to:

- Footings

- Fastenings, joints and interlocking parts

- Water traps, pipes, columns

- Thin castings

- Splash zones

- Adjoining metals


Minimal intervention is the fundamental aim, so as to cause minimum loss of evidence. We aim to: - Retain as much existing material as possible

- Repair and consolidate, rather than renew

- Use reversable processes

- Add material to reinforce, strengthen, prop tie and support

Iron & Steel


- Usc traditional materials anel techniques

- Make new materials distinguishable from originals

- Record before, during and after conservation work.


Both hot and cold techniques arc available. Heat is provided either by a high electric current, which forms a spark or 'arc', or by a flame fuelled by a mixture of gasses, usually oxygen and acetylene. These processes use very high temperatures and so have two serious drawbacks:

- They can be dangerous if applied in historic huildings,

- They can be destructive if applied to cast iron which, being brittle, often cannot accommodate

expansion and contraction.

Cast Iron Wrought Iron Site Works
Electric Arc W:lding (3000°C, MIG, TIG, Stick) Not Advised OK OK OK
Gas Welding (900DC, Fuel Gas & Oxigen) OK OK Difficult OK Cold repair techniques (especially suitable for castings): - Straps, plates

- Threaded studs

- Dowels, pins - glued or force-fitted

- Stitching

Case Studies

The following case-studies illustrate some of the well-established techniques in regular use: - Wrought Iron - Waterlilly House, Royal Botanic Gardens, Kew, London

- Wrought Iron - The Curvilinear Range, Glasnevin, Dublin, Ireland

- Wrought Iron - Queen Victoria's Royal Railway Shelter, H.M. Navel Base, Portsmouth

- Wrought Iron - Salisbury Cathedral, Salisbury, Hampshire

- Cast Iron - Boy and Swan Foundation by the Coalbrookdale Compagny of Ironbridge

- Cast Iron - "Pissotierre"

- Cast Iron - Seaside Picr Column

- Cast Iron - Palace of Westminster, Houses of Parliament, London

D.R. Ltd would be pleased to advise or assist on any historic metalwork project.


Iron & Sled

A Waterlily House, Royal Botanic Gardens, Kew, London,

AI 1852 Glasshouse by Richard Turner designed for the newly introduced Victoria Amazonica waterlily, The structure is 14.5m square, with cast iron columns and beams supporting a wrought iron trussed roof.

A2 Principal trusses are "bulbous tee" sections as commonly used Ior the frames of ships. Flanges had generally corroded seriously but the remaining sections were in sound enough condition to he repaired.

A3 After careful recording of dimensions, the roof was deglazed and dismantled to eaves level. To ensure every component was returned to its original position, two tags were wired on to every part. Metal tags were used to withstand grit-blasting.

A4 Recycled wrought iron plate was welded across the rusted flanges, arc welds being run simultaneously on both sides to minimise distortion. Trusses were pre-bent to allow for distortion during weld contraction.

A5 Wrought iron has a laminar structure, so there is a danger small cracks could develop in it due to weld contraction. All welds were checked by magnetic-particle testing in which a strong magnet is placed near the suspect area and magnetic ink used to reveal defects.


A6 Structural castings were repaired by plating with stainless steel hedded on epoxy putty to avoid contact between plates and ironwork. Stainless steel is agressive to iron and will corrode it in the presence of moisture. Iron work was cleaned locally and plates applied hefore general blast cleaning so that the final paint system was applied all over all components as one operation to seal out water.

B The Curvilinear Range, Glasnevin, Dublin, Ireland.

Bl&B2 Glasshouse by Richard Turner dating from 1842-9, built using twenty different sections of wrought iron glazing bar. The refurbishment of the structure required new bars of every type.


B3 Glazing bars would originally have been rolled, but the cost of twenty sets of new rolls would be astronomical. A forging process was therefore developed to enable us to manufacture every section to a tolerance of O.5mm using reclaimed wrought iron.

Iron & Sleel

C Queen Victoria's Royal Railway Shelter, H.M. Navel Base, Portsmouth

C I Built as a terminus to serve The Queen as she transferred to her royal ship en route to Osbourne House on The Isle of Wight, this topheavy structure suffered severe corrosion to its main posts due to sea spray.



I' ,

/ .'


C2 New sections of reclaimed wrought iron were formed into "tee" sections, rivettcd back to

Iron & Steel

C3 using hot-set rivets forged pneumatically, matching the originals. The structure was jacked up and new posts were welded to existing sound ironwork to replace corroded areas.

C4 Further pneumatic tooling, called a "rivet buster" is used to shear off the heads of rivets when dismantling a structure. The rivet heads shoot off like bullets which is a somewhat dangerous process, hut great fun!


D 1 Salisbury Cathedral, Salisbury, Hampshire

D I The Salisbury Cathedral, Salisbury, Hampshire

',I It

D2 Late mediaeval wrought iron tic bars 80m long, forged in sections pinned together encircle the tower. Further wrought iron ties were added during last century, so the early ironwork is not now structural, but is of historic interest duc to its great age, size and general condition. Only one bar showed any sign of failure, and was stabilised by arc welding a small plate across the gap, allowing access for painting all round. Forked ends were carefully filled with epoxy resin to exclude water, and thoroughly painted.


D3 Lead paint is best removed by use of needle guns and must be disposed of in accordance with health and safety legislation. Metalwork should then be 95% clean and can be blast cleaned more safely, but

D4 Primers must be applied on the same day as ironwork is blasted to avoid rerusting. In damp conditions wrought iron can start rerusting in 10 minutes. For exposed locations we use 2-pack epoxy and urethane paints whieh must be applied to a specified thickness, measured both by the operative who applied the paint, and by an independent inspector using properly calibrated equipment.

Iron & Steel

E Boy and Swan Fountain by the Coalbrookdale Company of Ironbridge, early 19th century.

," "'1

, . ~:~,~: ~ .' . ~ ,

f ... -+ • I '.. '~

'-'," - \ ,

-, '

El This exquisite sculpture is one of only three known to exist in the world. one is in America, one in The Museum of Iron, Coalbrookdale, and this one in a private garden in Worcestershire. It is complete except for minor damage where water has entered joints and expanding rust has pushed off sections of cast iron.

El This exquisite sculpture is one of only three

known to exist in the world, one is in America, one in The Museum of Iron, Coalbrookdale, and this one in a private garden in Worcestershire. It is complete except for minor damage where water has entered joints and expanding rust has pushed off sections of cast iron.

11'011 & Slee/

E3 The Swan's body had cracked and was repaired by adding a cast iron plate moulded from a wooden pattern carved to fit the ironwork. The wings were bolted through the broken body shell into this plate, clamping the weak components back together.

E4 Missing cast iron components can always be replicated by moulding their shape with vinyl rubber backed by plaster. From these a glass fibre pattern is made and taken to the foundry for casting in iron.


F "Pissotierr e"


F lOne of our more unusual conservation projects; a gentleman's toilet in central London, made of cast iron panels only 6mm thick.

F2 Conservation work included a broken panel which was repaired by bolting to a backing plate of galvanised steel 2 mm thick bedded on resin and sealed round. If such a thin plate of cast iron had been welded, it would have cracked repeatedly; welding is not recommended for thin sections.

iron & Sleet

G Seaside Pier Column

G I For sections of cast iron over 6mm thick we recommend the use of metal stitches wherever possible. The process is applied cold using pneumatic equipment and so is very suitable for site-work. Stages arc as shown on sketch G2 .

.. -- .. -. ----~-~_j

G2 The stitching process:

1 The broken casting is clamped together and hole drilled accross the crack using a jig 2 The holes are joined by a specially-shaped pneumatic chi scI.

3 Special Dorothea stitches are driven into the slots in layers through the metal's depth 4 Interlocking holes are drilled along the line of the crack, then

S Tapped (i.e. threaded), and

6 Studs are screwed in to seal the crack. 7 The surface is finally dressed off, and

8 Painted, resulting in an invisible repair, achieved without heating the components.


H Palace of Westminster, Houses of Parliament, London

H I The Palace was completely rebuilt to the designs of Sir Charles Barry following a serious fire in 1834. Fireproof construction was used throughout including bronze and lead for window frames, and cast iron for roofs and ventilator towers such as the one illustrated. The Clock Tower (Big Ben) is also predominantly cast iron.

H2 Many repairs have been carried out over the years including this early form of stitching to a decorative leaf. This is the blacksmith's traditional way of repairing cast iron using a wrought iron stitch.


H3 Many fastenings had corroded away, and thin castings were cracked where water had entered and caused rusting and expansion. Missing components were recast using originals as patterns, and these, together with original fragments, were refixed. They were bedded on 2-part resins and secured by stainless steel screws suitably painted to avoid direct contact with cast ironwork. Ev-

H4 Many components had been damaged near fixing holes either through movement or overtightening of screw fastenings. Thin castings were repaired by gas welding using a bronze filler rod.

Iron & Steel



H5 Wrought iron is more durable than steel, and cast iron is more durable than wrought. Like all building materials both will deteriorate if not maintained properly, like the flag mast and its support structure 100m up on the Palace of Westminster (illustrated). Given basic care and maintenance, historic cast and wrought ironwork can continue to serve long into the new millennium, continuing indefinitely to grace our buildings. Alternatively, without proper maintenance the sort of metalwork illustrated in this paper will gradually be lost.

/1'011 & Sled



Iron & Steel


lug, K. Villi der Burg, Hol/und;" bv,

Krinipe» aid l.lsel, The Netherlands

Project history.

The Wuppertal mono rail (Sehwebebahn) is the first mono rail system in the world and has a length of 13.3 km. At the end of the 19th century a number of concept studies were made, of which the winning concept from Mr Eugene Langen was built from 1898 to 1903,

After an operation period of about 100 years, the system has to be refurbished,

For reason of the historic value of this project, the decision was taken to rebuild this public transport system according to the modem design requirements, however original configuration as well as the connection details needed to be maintained.

It goes without saying that rebuilding a riveted structure in this form and of this magnitude has become new to the steel industry, Therefore, questions need to be answered like how old fashioned riveting techniques can be re-used and yet meet the to-days technical requirements in terms of design, production techniques and corrosion protection.

In this presentation information is given about:

1. Public function.

2. System description,

3, Reasons for the refurbishment. 4, Design information,

5. Fabrication.

6. Corrosion protection system

7. Stations,

Public function

At the time, the system was built for reason that the valley area between Barmen and Eberfeld was already quite populated and thus required a public transport system, Already in those days, the infrastructure and existing transport methods simply did not allow to transport the 300,000 habitants of this highly industrial area.

Later, the industrial revolution attracted much more people to this area which grew out to one city, Wuppertal, Now the Wuppertal mono rail system fulfills the function of one of the main public transport systems needed to meet the traffic requirements in this area,

Daily 70.000 passengers make usc of this mono rail transport system.

After 100 years it is fair to say that this mono rail system was initially built to allow for transport in the area. Now it can be stated that, despite the enormous infrastructure developments realised in the 20th century, this system has maintained its function due to the excessive increase of mobility of the public. Like in many other cities, an efficient public transport systems is the only answer to control city traffic congestions. Consequently, at the time the system was built to get the public to its destination without having a good infrastructure. These days, despite the good infrastructure, the system is needed to transport the public to its destination for reason that due traffic congestion reason it is time consuming to go to the inner-city.

Iron & Steel


Figure 1 :General View of the Mono rail system

System description

This mono rail system is a public transport system of 13,3 kIn length, running through the city of Wuppertal. However, a substantial part is situated above the river Wupper.

It consists of 468 bridges varying from 7 to 33 meters, of which a significant part is curved 10 follow river Wupper and to follow the configuration of the city. The bridges are supported by 468 support structures consisting of hinged and fixed structures.

The total weight of the structure exceeds the 25.000 tonnes.

Stutze Nr. 45'



(in - Ausfobrtsversolz = N 10 rnrn

Figure 2: Cross section


[ron & Sfee{

Reasons for refurbishment. Technical reasons.

The refurbishment program consist of a complete replacement of the system, for which activities the mono rail system is periodically taken out of operation to allow for partial replacement.

The main reasons for this radical step are that the steel structure's life time has come to an end, from fatigue and loading as well as from a corrosion protection point of view,

Further, modern technical requirements have to be implemented, such as the increase of the train speed, passenger capacity per train, frequency of operation. Simultaneously the electrical and control systems are brought to the today's technology standards. One of the further requirements is noise reduction. From the new bridges already installed, it has been demonstrated that new design has resulted in a significant noise reduction.

Corrosion protection.

Like for all steel structures, the reduction of maintenance costs is one of the design criteria's. However, at the time, the paint technology, environmental requirements and the ratio labour costs to material costs was totally different. History has leamed that already in 1910, which is 7 years after project completion, the total structure had to be repainted. A situation which is now totally unacceptable.

So far, a crew of 20 men is almost permanently employed to do paint repair work. This resulted in the situation that the existing structure has on many places a paint coat thickness of between 3-4 rnm. The extremely thick coat system has been cracked due to its inflexibility and thus has therefore lost the majority of its corrosion protection function.

Further, the structure has many inaccessible area's.

Therefore, removing of these paint layers is not only very difficult to do, but will result in major environmental problems, since a possible contamination of the river Wupper with paint chemicals, for instance lead, will not be accepted.

Further, the costs for paint removing and repainting were estimated to the value of DM 230,- million, a figure for which a substantial part of the mono rail system could be renewed.

The corrosion problem combined with the technicalities and risks earlier specified resulted in the decision that the refurbishment program is in fact a program of total renewal.

For obvious reasons, much attention was paid to the corrosion protection system for the new structure which is, as stated earlier, a riveted structure. Moreover it is a structure with difficult accessible area's.

European influence.

It has widely been acknowledged that the Wuppertal mono rail structure is a demonstration of the high level of German construction techniques already available in 19th century. Therefore the European Parliament made a significant contribution to this rebuilding program.

This fact, in combination with the advantage that for a replacement of an existing structure no planning permission procedure is required, resulted in a project go-ahead, this is a re-doing of a construction process which already was realised 100 years ago. Since then, men have walked over the moon, we moe in the possession of computer technologies, faxes and e-mail.

Iron & Steel


It of course has to he accepted that these days the project requirements have become much higher than the requirements of 100 years ago, but even though during its current design and construction process, the Gelman construction abilities of the 19th century has come full to light.

Technical information. General Design information

As the result of the mono rail geometry, the majority of the bridges are unique.

Further for the curved bridges in particular, every horizontal bracing in a bridge is unique as well. This requires a stringent design and production regime.

In order to satisfy the German rules, stress calculations are to be made from every bridge and support structure. Consequently much attention was paid to the stress calculation and design process .

.. ,t..:i69 '

J ~p!Je~C:I'clWJ~lIq :i.t M01t:1 SaW::k1:'

IIffi1l«tlLhi,ooml UH.·l$1l~n9#l1g. &~f*~"'~ptt L __ fOhd

DUll! ...... f;;ip:p~GhmnmD -Ah;"d#!'1lo~ I~ I' 5J 1i1l:Lefi.libm-OU~ffrt -UIHEi.Mooma'fiI'llt!


0_ f«lllfl<l~n:t tnf.m r~.".",., Til> 1 .Q2IiOO



Od D.(IIUil

nma .. 'IlIl!>!tf~ t.~ .. J,mn 1 yp ) G2GtO

7t!_11,.1Jl 1S,m

[h. B1blem nl'_W5WY2.11 lH



""",(Q~,,, INc, II

Oi),{(iJJI: fffi~c Al'b';1nrf.i1\9 Zcl: leb1e AndnlUnQ W,lN,lt.\'i.' A~dttlJny:Hllrufld

1l>1tig1upfi'!'!fI:·,11flUlion !I'iJug:II'Jflfi'l'l'RHu.mnlf!r



1,4". Ii


trn"L. "I lI!<llr_ I

."" ..

Figure 3: TIP/Schweb: Location information

The well advanced computer systems and the CAD/CAM systems resulted in the position that the design as well as the fabrication information is fully computerised, The design output, given on CD's, is the input to the machines for fabrication.

This has avoided a manual input from data to the production machines which undoubtedly would have lead to production problems due to human errors in data transmission would lead to misalignment of rivet holes.


Iron & Steel

An impression of the various design requirements is given in the different models hereunder specified.




a) Fixed points above water.

b) Hinged point above water.

c) Fixed points land based.

d) Fixed points land based.

Figure 4: Types of Support Structures.

Figure 5: Calculation model of a bridge system

Figure 6 (r): Stress model of a curved bridge

Computer requirements. For the stress calculations:

a) 468 three- dimensional framing models are required to be made with about 200 elements and 700 loading cases each to achieve the necessary structural integrity the bridges (figure 5)

b) 468 three- dimensional framing models are like wise to be made for the columns;

the number of elements is, according to the type of column, between 700 for hinged and 250 for fixed columns. (figure 4)

Iron & Steel



In order to achieve a stress calculation which meets the standards required by the checking engineer, about 1000 pages per bridge or eolumn are required. Therefore, about I million pages of stress calculations are to be made. This corresponds to a memory storage capacity of 5 GB. Therefore, the stress calculation and design activities arc very lahour intensive and so is the fabrication process.

For the design:

a) design of one three- dimensional CAD-model each including all design details and NC- data for all the 468 bridges and columns (figure 7)

h) development of all workshop drawings based on this information ( about 25 drawings per bridge and 15 drawings per column).

In total, these are about 1000 three- dimensional CAD-models and about 20000 full size workshop drawings.

This requires a memory storage capacity of 54 GB for the CAD- models and 30 GB for the workshop drawings. In addition, 3 GB will he required for NC data and part lists.

All in all this is a challenge for the information processing which can only be done by the use of the today's available CAD- and database- technologies.

Model development. General.

In TIP/WSW Schweb we can find two different models, the geodesy model and the structure model. The geodesy model consist of the route axis's information including the bridge levels and the axis's of the support structures with the coordinates of the base.

The structure model consists of the calculation model (called base model) and the design model.

11 is possible to show the models in different representations like symholic lines or the real three dimension model.

Layout of the mono rail system

The system design starts with the data of the route axis's. The data are store in TIP/WSW Schweb in digital form. The elements of the route descrihe the surface contour line and curve on the ground. (figure 3). All parameters for the description of the different structures arc stored. The bearings are of a fixed or moveahle type. All this information is used for the exact geometric description of every group of parts. For example for the automatic output of a hridge TIP/WSW SCHWEB the data and digital model from the system catalogs are required.

Stress analysis and determination of section sizes.

The stress model is derived from the basic model descrihe before. Eccentricities of the system lines and the truss joints are automatically taken into consideration. The types and geometry of all elements are described in dependents of the route and the system types.

On the basis of this information, the stress analyses are made.



The design development process and implementation to the fabrication process, as specified in par 4 of this document, meets the latest computer design development possibilities.

However, how is this to cope with the old fashioned riveting technique. The times of a noisy hammering process and throwing hot rivets in the air have gone.


Iron & Steel

These days the rivets are electrically heated. The riveting process is done hydraulically. This not only results in a controlled production program, hut it is now fully repeatable production process, so that the reliability of every rivet can be guaranteed. In the old days, hammer checks were made to control whether or not the rivet is .. loose".

The hydraulic process, however, guarantees sufficient plastic deformation of the rivet and so the proper filling of the rivet hole and proper deformation of the rivet head.

Fabrication details.

Figure 7: General Arrangement of a bridge structure.


... ~,-........ t.~ ...

16 N! fl


Figure 8: Connection details

Iron & SleeT






Fig 9 Bridge bearing structure

Fig 10 Cross section

Fabrication process

It goes without saying that the production process to he followed for such a project requires a very accurate work preparation. Only for the 2.1 km bridges made in the Hollandia works over 600.000 rivet connections are to be made. The accurateness of the design and fabrication process is therefore determinant for the economics and quality of the project. Further a constant now of materials and design information is needed to allow for a continuous production, which is essential for an economic fabrication process.

The fabrication process, needed for this project, is the combination of the production disciplines needed


Iron & Steel

for the fabrication of large steel structures. A production process which requires quite a logistic control of materials, fabrication and quality management of thousands of different parts and a single bridge manufacturing process which requires a totally different production regime and production control, generated from a limited number of pieces to be assembled.

Therefore the bridge parts are being prefabricated in diagonal assemblies and complete upper bridges. This process is done at the Hollandia steel structure manufacturing plant.

The steel members are produced on CNC sawing and drilling lines. Hole diameters are being kept smaller to allow for reaming at a later stage. The individual sections are galvanised and provided with the first coat. Assemblies are made by temporary bolts, the hole diameter are being reamed 1 mm larger in size than the rivet diameter.

It should be noted that this reaming process is required to achieve smooth and aligned holes, which is required to satisfy the fatigue requirements. The accurateness of the CAD data is such that there are practically no alignment problems. Therefore, the reaming process is limited to a small increase of the hole diameters.

The riveting tool is in fact a U-shaped brace provided with rivet heads, a hydraulic cylinder and controls. In a prefabrication situation this tool forms part of a fixed positioned machines.

In the final assembling situation, this tool is hanging in a balanced situation in a crane to allow for easy maneuvering,

Much attention is to be paid to the riveting sequence. When the proper riveting sequence is followed it is possible that, even at the final assembling phase, this tool has access to all the rivet connections. This avoids a manual riveting process.

Final assembling has taken place in the bridge manufacturing plant, after which the corrosion protection system is completed. The fit out will take place at the fabrication yard so that all the parts can be erected directly from the truck without the execution of further assembly activities.

The corrosion protection system. General

For many projects it is experienced that the corrosion protection system is one of the last subjects of discussion of a design development.

It occurs quite often that corrosion protection experts are being consulted at the final design phase, after the phase at which the designer has secured the structural integrity of the structure. This is a stage in which very little room is left for design modifications to allow for a proper corrosion protection system.

It has to be recognised that a good corrosion protection system combined with a modern design forms part of the structural integrity as well!

This under the condition that the design details are as such that it can take the paint properly.

It may be of interest to know that, unlike in the Netherlands, British Standards do exist for long life (10 years) and very long life (20 years) corrosion protection systems, a system which has very little maintenance requirements.

This can be treated as an advantage, however, it equally can be treated as a disadvantage. This for reason that with such a code in place, incentives for new paint systems developments are taken away, which results in a possible obstruction that modem paint techniques are implemented in new codes to be developed.

Iron & Steel


Therefore on some projects, the writer of this presentation experiences a blockade of further paint development techniques, Moreover, having this standard in place, it is not an early invitation to consult paint experts during the design phase.

It is widely accepted that a good corrosion protection system is one of the key factors for low future maintenance costs. The quality of a corrosion protection system is determined by the environmental conditions during shot blasting, the cleanliness of the steel SA 2,5 and the cleanliness of the painted area's. Projects brought to site in one piece on which the shot blasting and painting is done in-doors under climate conditions will result in the lowest future maintenance costs.

It is therefore very impractical for the project owners that these sorts of features do not form part of the bid evaluation process in European tenders. In other words, clients are normally not prepared to pay additional money for additional quality for which they pay the penalty at a later stage.

The mono rail project

For the mono rail project, the designer was confronted with the given situation that the design is poor from a corrosion protection point of view. Therefore, a duplex system has been chosen of which the life time is about. 2,5 times longer than the lifetime of a solely galvaniscd structure.

However, each bridge has about 6.000 rivets and thus 12.000 rivet heads which has damaged the galvanised coat due to its hot riveting process.

Questions needed to be answered as:

What is the degree of damage to the galvanised coat due to riveting, how is it to be repaired and what is to be done to achieve a very long life corrosion protection system.

Application of the corrosion protection system.

The corrosion protection system starts with the ordering of the right material quality.

The chemical composition is such that the best galvanising quality can be achieved so that uncontrolled growth of zinc layer is avoided. This growth will take place when the parts are too long in the zinc bath. However, due the varying plate thickness of a particular member, the parts need sufficiently long to be in the zinc bath to allow for the proper temperature, necessary to achieve the right zinc layer thickness. Therefore, a compromise had to be found to achieve on every part the required zinc coat, without having the uncontrolled growth as specified earlier.

After sawing and drilling or fabrication of the individual members, the system to be followed is: - shot blasting to achieve a good pattern to take the galvanising

- pickling and galvanising, the zinc thickness is 120 micron, which thickness is considerably more than

the standard coat thickness of galvanised structures. - sweep blasting

- application of the first coat of 40 micron.

After assembling and riveting completion: - shotblasting of the rivet heads.

- application of gO micron of zinc paint to repair the rivet heads and the damaged area's

- water cleaning of the structure, the repaired area's in particular

- application of a 2 component polyurethane paint of a coat thickness of 80 micron.

- application of a 2 component polyurethane paint of a coat thickness of 80 micron.


Iron & Steel


The galvanised layer will melt when it is brought in contact with the hot rivet.

Therefore, zinc pearls will be formed around the rivet head. In order to limit the zinc pearl building, the rivet head should not be made too hot.

Immediate removing of these pearls is essential to avoid pit corrosion at a later stage.

General remark.

In Hollandia's view, for many projects it applies that the clients do not make sufficient distinction in the 2 different functions of a corrosion protection system, i.e. the corrosion protection as a technical function to protect the steel structure and the cosmetic or architectural function to satisfy the esthetic requirements.

Therefore it is often experienced that the final coat is applied on site. However, the quality of the final coat is determined by the environmental conditions during paint application and the degree of cleanliness of the surfaces to be painted.

Experience has shown that, when large surfaces are needed to be cleaned and particularly in cases of difficult accessible area's, the site coat has more a cosmetic function than a corrosion protection function.

Consequently, if in a corrosive protection system the final coat is site-applied its capabilities has already been reduced, due to reasons specified earlier.

As the result of contamination, the cosmetic quality has lost its function relatively shortly after project completion, so that the corrosion protection function remains.

It is therefore that Hollandia strongly recommends to apply the complete paint system during the fabrication process under climate controlled conditions. Certainly when the possibility exists to ship completed pieces to the site.

Only a very few customers acknowledge this fact and are prepared to spend extra money to transport completed pieces to the site, having in turn the advantages of lower future maintenance costs and a longer expected life time of the structure.

Project particulars.

At shot blasting after riveting, damaging of the surrounding zinc layer can not be avoided.

With a proper shot blasting pattern, a good degree of roughness, achieved before galvanising will result in a good adhesion of the zinc coat and thus will prevent from coming off the zinc coat.

Further measures are taken in terms of the shot blast materials, equipment and blasting process.


With the renewal of the mono rail system, the stations were renewed as well.

The design is of an entirely different nature. The modem welded structures form part of the bridge structure and are therefore to be built and installed progressively with the replacement of the mono rail.

For obvious reasons, the same corrosion protection system was applied to this steel structure. The time will tell, whether or not differences can be found in the corrosion protection performances of old live ted structures compared with modem welded structures.

Iron & Sleel


Figure 11: One of the new railway stations Photo Mark wohlrab, kumen

References :

* Stahlbau No.3 - March 1999 (special edition)

Photo,' Mort: Wohimh, Kamen

'" Thcrmisch verzinken No.3, page 44,45 - September 1999

(including the 2 Photo's by Mark Wohlrab, Kamen)

* Publication of Dipl.-Ing. Andreas Bermel, Dr.-Ing. Wolfgang Bilstein, Dipl.-Ing. Ulrich Pfingst and Professor Dr. Ing, Burkhard Weber - October 1999


/1'011 & Steel


Prof A. de Naeyer

Postgradnute course ill architectural preservation, Institnte o] Architecture, Antwerpen, Belginm Assoc. Prof Ghent University, FlICI"!)' Civil Engineering, Department ofArchitecture, Gem, Belgium

Not being a material engineer, nor a specialist in metallurgy, it was the "preservation praxis" who pushed me to some research in the field of old iron structures. I said "preservation" or restoration, that means that kind of operation which tries to preserve the authentic historical material as long and as complete as possible, using all availiable techniques, even at the expense of aesthetic homogeneity or structural uniformity, Too often in restoration praxis, we see that old and somtimes very valuable iron structures are demolished or substituted by new ones. Uncertainty about strenght and stability of deteriorated materials, or ignorance about specific behaviour of ancient structures, and also all kind of interventions because of new safety regulations or modern comfort, usually end in substitution of the old structure by a new one, or, if it deals with a protected building, by the copy (or a almost - copy) of the old structure (figure I). This is regrettable from preservation point of view; and it looks to me that the metallurgic industry is not that interested in renovation of old structures (maybe they consider the market too small), although there is an enormous field of applications, and a lot of very fine, but often very threathened, 19th century buildings.


;" ;.

Figure I: Royal glasse houses, Brussels

For the two other main traditional buildi, ng materials, stone and wood, specific restoration techniques have been develloped since many years. I am thinking of the applications of epoxy and polyurethan resins for restrenghtening wooden structures or renewing timber beam-ends, in order to preserve the healthy part of the original beam as big as possible; I am thinking of the various products of so-called stone and concrete "repair mortars" or "artificial stone mortars" based on lime, cement, or different kinds of chemicals, I think on all kind of stone surface consolidants .... All of this products were develloped to maintain the maximum of the original and authentic material.

Iron & Steel


Why such products clon't exist for iron structures? One may say: reproduction of the old type of iron is impossible while the modem production processes are totally different from earlier times and much too complex to change them! And also stress and strain situation in iron sections can not compared with those in wood or stone, or even in concrete. Of course, it is not realistic to change modern production lines; but I can image the re-opening of some small traditional furnace (maybe in one of the many museums of industrial archaeology) for reproduction of iron products, following the ancient techniques, materials and labour (as it is done e.g. for lime production or brick production). But I realise, this may not he very realistic!

Based on this considerations, we first of all studied the old production methods and qualities of old iron After that, we had to evaluate available techniques to repair and restrenghten old iron structures. It was done in a coherent program of student thesises (I) and a limited series of test benching under the direction of my colleague prof. Joris DEGRIECK, head of the section "Mechanical Constructions and Production".

Our limited time does not allow to present the first part of those studies; lined on ancient production techniques and their consequences for the products characteristics. Neither can we enter into the different aspects of the necessary preliminary control, testing and diagnosis on the ancient metal structures. (2)

My topic of today treats the possible (or impossiblel) repair or exchange of a small part within a bigger load hearing element, preferably such a part as the connection with foundations or the main framework of the structure, This parts are the most exposed to all kinds of deterioration agents, and are the first for going in failure! The classic (soft) intervention in case of such corroded or unreliable part, was to cut it out and to substitute with a similar one, or to clean and reinforce by douhling it with similar material. The idea was to look for alternatives for those "prothesises" by using new, modem and "easy" materials, which were not so sensible to corrosion,

In analogy with epoxy resin techniques used for the repair of timher and concrete or brick structures, and the good results concrete industry gets using fibre mixing in the mortar, we thought it useful to look after possibilities with modern composite materials as "Fiber Reinforced Plastics" in the repair of metal structures, e.g. for filling up surface cavities, for replacing too heavy corroded parts, and even for reinforcement (by applying them or in liquid, or in a solid state). Of course, also other materials for the matrix could have been considered, but within the very limited time and money, we decided for epoxy resin as a polymer, because of the existing and experienced analogies.

But, as already mcntionned, mechanically speaking, wood, concrete and iron are very different materials. To indicate the most evident differences:

* stress concentrations in iron clements are considerably higher than in wooden or concrete ones (steel elements as I or H sections, me specially designed for taking high stresses at the extreme points of the sections);

* the porosity of iron is nihil in comparison with wood or stone! For this reason, adhesion by glueing will be very difficult or even impossihle; it can be improved by treating the surface (making it rough or using special coatings);

* iron structures are much more stiff and rigid than wooden ones. Any substitutional material must have comparable elasticity properties, otherwise it will get haircracks very soon and water will enter through this capilars and corrosion proces will start again!


Iron & Steet

So, the type of substitution material, his composition and compounding will be of the biggest importance to guarantee some succes. In our case of a composite material (= matrix + fibers), his mechanical characteristics will have to approach those of the old metal as close as possible!

Thc quality and behaviour of the proposed material, FRP (Fiber Reinforced Polymer) will depend from: - the type of fibers (carbon- , glas- , metallic - , ceramic-fibers, ... ),

- the direction in which the fibers arc orientated (this is determining in great extend the mechanical

identity of the prothesis, as well for strenght as for stiffness or elasticity / usually there is a big difference between elasticity orthogonal on the fiber-direction and those following fiber-direction. E.g. stress by forces in the same direction as the fibre can reach (with C fibers) the same as the best steel qualities i.e. +- 250 MPa , - but orthogonal on the fibres it has to be reduced sometimes to 1/5 i.e.+- SO MPa or even less!) Specialists confirm that the production of high resistant FRP, similar to steel, should be no problem!

- the type of matrix (thcrmoplast, thermohard, metal, carbon, ceramics, glas, .... )

- the type of the border surface of the fibre and his capacity of adhesion with the matrix substance.

The effectiveness of the combination of iron and FRP depends on:

a. the type of iron and structure : the chemical properties (% C, S, P, Mn, Si, N, ... ) and material structure of the iron (cast, wrought, rolled, .. ) and the mechanical loads in the structure (pression, tension, flexion, shear, ... ), ( NB : the direction in which the fibers are woven and the way in which different layers are composed can guarantee a very high mechanical resistance, perfectly comparable to the quality of the original iron or modem steel )

b. eleaning level and - technique (mostly sand or ironblasting) and adhesion capacity on the original iron surface; The adhesion is not only a question of the iron surface (roughness, porosity or microcracks), but also of the type of polymer (surface tension, viscosity) and the type of hardening and glueing-process of the liquid polymer (solidification / evaporation of solvents / chemical binding after putting a conversion layer - mostly fosfates - on the metal)

e. accessibility of the place where the intervention has to be applied (has it to be prepared in the workshop, or can it be done on the spot itself, has the structure to be - partly - dismantelcd ? ... )

d. climate conditions after application ( e.g. exposition to humidity or heat: under a metal roof, you get extremes of - 30 a up to + 80 a C ), especially important for the long term behaviour of the composite.

e. resistance of the composite material towards fire (especially for load bearing structures), towards light (U.v. radiation), and characteristics conceming creep, relaxation, cohesion of both materials, ...

All this characteristics have to be tested in a long and systematic testbenching program. Within the small scope of student thcsisses, this was impossible because of financial reasons. To start at least with something, prof. DEGRIECK and myself tought it advisable to test the linking possibility and the strenght of the junction between the ancient iron and the new composite material. One could think about three linking possibilities: l. glue, 2. screws or pins, 3. form-connection. Connection by glueing was not kept because other experiments, although not in the same conditions, had shown not lasting very long, and connection with screw bolts was also withdrawn while the fixation of screws on old and often deteriorated iron sections usually is complicated and little reliable. The only option left was a so-called "form-linking" (i.e. the composite material is used as a real substitue for the original material and is precisely fitted within the borders of the original section); this was definitely the best looking solution from the preservation point of view as well as from the aesthetic point of view.

Iron & Steel


Iron sections mostly work within mono-axial stress situations; for this reason it is advisable to use mono-directional laminates as in figure 2. Based on the specific calculation techniques for laminates. it is possible to devellop any material with specific values of strcnght and strain, even considering different behaviour in x , y and z direction. Other applications (e.g. reinforcement of concrete with outside glued steel sections) have

already proved that carbonfibre reinforced epoxy resin (CtER) can be used as a good alternative in iron structures as it has an elasticity similar to steel (210 GPa) and strenght which is even higher than normal steel. II

looked logic to propose also GER, hut, as this composite was not available in the laboratory at the moment of testing, we accepted testing with "E-glas (= glas fiber with specific dielectric characteristics) fiber reinforced polyester". This composite is less strong, but the tests could give us the same usefull information.

For a possible application on a building site, we tought that a cavity in a ancient section, caused by extreme corrosion, could be filled with a kind of prefabricated central core (made of the same composite, but it can he done also in other material), wrapped and glued with different layers to the original section. One could also think of different types of winding and even tieing with knots. But this could be an idea for futher experiments.

Figure 2

In our applications, we used for fixation of the core material to the historic material the so-called "prepreg" technique. "Pre-preg" (= previously impregnated binding material) arc strips of grouped or winded fibers which is already impregnated with matrix-material and already partly polymerised. It is very sticky, and is placed on the two parts of the structure which have to be connected (the end of the ancient material and the end of the new core) in a direction which can be ehoosen. Once this sticky prepreg is placed, one has to create the conditions (heat, pressure, kathalysators, adding other components, ... ) in order to complete the polymerisation and full hardening. This technique is often used for reinforcing existing structures (in porous materials). Another advantage of "prepreg" is his flexibility and adaptability for being "filled-in" in the irregularities of existing structures or sections.

Within the frame of his diploma thesis, civil engineer-architect Andries RANSON (1), made a small series of tests on a flat piece of normal construction steel, and, as already said, with prepreg of E-glas fibre in epoxy resin. We summarise the results:

I. Hardening of the prepreg : should happen in a well controlled environment and conditions depending from the type of matrix material. In our case of a thermohard resin, some heat and pressure has to he added to initiate the hardening process (by cross linking). This process can only be effective when any creation of gas hubbIes ( coming from solvents, thinners or even water vapour) is avoided. For this reason, the prepreg part has to be put in under-pressure condition or vacuum suction. This was tested, first in an autoclave where heat and pressure can he precisely controlled, and secondly in normal building site conditions, by heating the material in a traditional oven, after having put it in a plastic bag for creating vacuum and the necessary pressure. But within the scope of our applications (huge metal structures), the use of an autoclave or heating oven is rarely possihle and will create problems in praxis. One could coneieve also other matrixes or other chemical combinations, even of the thermoplasttype (although the need for the factor "heat" is higher in this last case). Anyway, "heating" on the building site has to he eliminated, and full polymerisation should be based on selected hardening components. A pressure difference of 1 bar (easy to obtain with the vacuum bag) was already sufficient for our materials; and this can he realised on a building site without any prohlem.

2. Putting a intermediate layer (a one component elastic polyurethane coating) on the core of the prothesis and on the steel, to prevent stress concentrations (or uniforming elasticity behaviour) or to im-


Iron & Sleei

prove the cohesion between elements, did not gave the result we hoped. One tought to compensate with this intermediate layer the possible differences in deformation (settlements) of the different materials and the anisotropy of the matrix, but maximum tension value without intermediate layer was clearly higher than those with this layer. This indicated that such intermediate layer was more acting as a lubricant and his action as a elasticity improver was to neglect.


"":::~ .C:·':l[' .~~:3

-.~~ .

. . . .

::::_ •••.•..••.. , .••.•... ,....... . ..... '::"'';';:\,J.. _~ __

.• ,....-.-

.... -.-... >~h"Ii"E":' n.';:;~' ~~(:;uiJmptlmf"

r~'.}'" .:;:.::-::m_L-·::_,~:::·:::::::·:::::~::·~~ .. i .. -» .. "" ..... :~::~.ii

--r- t'r.:..!1 ..... ~~

.~~~~iiiiiili~. 1:;-'~lfX~

!! F}Q;.\r'~

ge;:;l.ill~i.;'::":-';';"(j ~:"Irl<'> n-eesren m;:;l

Figure 3: Tensile test

.... ,._ .. _ .. _.,___.. ;a-J

_ .. _-" (l';~'~I'_(-J : ...

lron & Steel


3. As most of the metal structures are frames or trusses, longitudinal forces (compression or tension) are the most common ones in the bars. Three type of tensile-tests were executed, all with slightly different conditions of connection :

3.1. Connecting the prothesis after narrowing the end section of the original steel (figure 3) : very early delamination of the composite: first small cracks started already with 5 to 6 kN ( = 23 MPa), complete fraction (along the line of connection) arrived with 40 kN

3.2. Connection without narrowing the end section: same strenght and same fraction values as before, but bigger deformation of the composite. Evidently, the composite slided more easy on the section end of the steel.

3.3. Connection by changing the winding direction of the prepreg strips over 10 0 (and doubling the winding to avoid possible torsion as a result of tension). The nominal maximum tension value was +- 25 % lower than in the two former cases ( fracture after +- 30 kN in stead of 40 kN)! Because of the inclination of the fibre structure, evidently a part of the tension had been taken by the matrix, and that could give a plausible explanation for the lower strenght.

4. Two flexion test ( figure 4 ) Iearncd that the connection was strong enough to take the same t1exion stresses as the original steel section .


Figure 4: Flexing

Figure 5: Compression

]5. Compression tests in similar conditions as the tensile tests. Also here, delamination of the composite occured rather quickly (+- 12 kN), always at the point of connection between the two materials. Final fraction was due to buckling. The inclination of the fibre direction with 45 0 was very useful! in this case and resulted in an augmentation of the maximum compression strenght of +- 20 %. (figure 5) Hook- or angle sections are more adapted than Hat steel sections for laking compression stresses. The testing should be extented for those applications, but was not possible within this smal! project. (figure 6)


Iron & Sleel

.. .-."

.: ... ' :;..,


Figure 6: Angle sections


1. The results of the testing were not so successfull.Small cracking in the composite material start already at rather small tensile stress. In praxis, this will cause again water infiltration and continue (or recommence) corrosion. Delamination of the composite is also a problem which has to considered, but the few tests indicate that the composite material can be made strong enough (comparable to the metal), but the weakest point is the link between the ancient metal and the prothesis of composite. One has to study on better connecting techniques, maybe by making holes in the old structure and sewing together with composite strings (using also prepregs ) and even by using special noding techniques as are used by fishermen for their nets. - figure 7

Figure 7: Sewing and noding

Figure 8: Relaxation

2. Great uncertainty remains concerning the long term behaviour of the composite and the binding.

Plastics, in general, are very sensitive to creep, to relaxation (= fracture after repeated loading conditions) (figure 8 ) , in which temperature and humidity are very important. There are numerical calculation models for having an idea about life time of those materials, and they have to be considcrcd carefully.

3. The specific chemical composition of each type of ancient metal was left out of consideration in our small project (e.g. % of C, S, Mn, P, ... ). But this aspect also is important, e.g. about possibilities for elaborating surfaces of the old metal, or for the selection of a specific type of matrix. This aspect has to be deveUoped in a further program.

fran & Steel


4. The special design and great variety in dctailling of 19th century metal structures are also an important problem by preparing adapted core - models for the different prothesises. The junction and joints of different nodes (which are the first pieces to replace) are very delicate and can be copied only by manual production. Extrusion or pulltrusion processes are not adapted for those type of most irregular one-piece production!

Notwithstanding the limited results and the many practical problems for application on a building site, I remain convinced about the opportunities of Fibre Reinforced Polymers in this field. It should be the subject of a general research project, in which different combinations of fibres, polymers and connecting techniques have to be tested systematically. I hope this COST network can offer good possiblities for exchange and cooperation.


S. WILLEMS: "Renovatietechnieken bij het herstel van metalen draagstructuren", Univ. Gent, 1997; A. RANSON: "Restauratie van staalconstructies met behulp van vezelversterkte kunststof', Univ. Gent 1998 ;

F. DELACROIX : "Technisch vooronderzoek bij renovatie van historische metaalstrucruren", Univ. Gent 1999


see a.o. : M. BUSSELL: "Appraisal of existing iron and steel structures", Publications Steel Construction Institute, Ascot, 1997, pp. 170


11'011 & Steel


Ir,L Twilt

TNO Bnihling atul Construction Research, Delft, The Nctherlonds


ln the 2nd half of the last century cast iron was often applied as construction material in buildings. In view of the relatively low tension strength, the application was mainly confined to centrically loaded columns. With the introduction of steel in the beginning of the 20th century, cast iron was pushed aside as construction material in buildings.

Nowadays it is generally recognised that the preservation of historical buildings is of great importance. ln this category, buildings with cast iron structural elements, play an important role [11. Obviously, also such 'monuments' need to meet certain safety requirements. An import aspect in this respect is structural fire safety.

The assessment of the structural fire safety of cast iron structures is difficult because the available information on the fire behaviour of cast iron is limited, whilst the fire design assumptions (if any) often are not well documented. A complicating factor with regard to protective measures is that generally the cast iron construction - in view of its "monument function" must remain visible. In other words: fire insulation by spray or boards, which is quite commonly applied for steel structures, is no realistic option for cast iron in historical buildings.

This paper deals with the needs and options of structural fire safety of cast iron in historical buildings. To understand these, it is necessary to quantify the behaviour of fire exposed cast iron structural elements. For reasons as explained above, the discussion will be confined to columns under centric loading. A further limitation is that only so-called 'grey' cast iron will bc taken into account, since this was commonly used for columns around the turn of the century, i.e. in the period from which most of the historical building with cast iron stem [1]. With a view to analyse the need for structural fire safety, first however, some attention will be given to required safety levels.

Required levels of structural fire safety

Fire safety in buildings is concerned with achieving two fundamental objectives: - to reduce loss of life;

- to reduce property or financial loss in or in the neighbourhood of a building fire.

In addition to these, there is an increasing concern regarding the protection of: - the environment

- infrastructures of vital importance

- cultural heritage

Note that, with respect to protection of cultural heritage, old buildings in general and even relatively new buildings with cultural significance may have values which can not be measured on a monetary scale. Their protection against the damage caused by fire is considered to be one of the basic moral duties of the society. Often historic buildings are also of direct measurable value to society, e.g. as sights attracting tourists.

Iron & Steel


In engineering terms, the protection of cultural heritage may not be that different from protection of any other values. Sometimes, however, historic buildings call, for a "special treatment". An example is the above mentioned complication that cast iron structural clement should remain visible, thus excluding fire protection measures which would otherwise be quite acceptable.

In most countries the responsibility for achieving fire safety objectives is divided between government or civil authorities who have responsibility for life safety via building regulations and insurance companies who arc concerned with property loss through their fire insurance policies.

To date, the use of a conventional fire scenario based on the ISO standard fire curve is common practice in Europe and elsewhere. The gas temperature time development according to ISO standard curve is

1 ego.
2: 1)1)0
2_¢c) presented in figure 1 [2J.

Figure 1: The ISO standard fire curve

The period of time during which a structural clement can fulfil its load bearing function under ISO standard fire conditions is called "fire resistance with respect the load bearing function", in this paper simply referred to as "fire resistance". The fire resistance is expressed in minutes or hours. Related classes are identified by the letter R.l Structural fire safety levels in buildings are mainly, but not exclusively, based on this approach and are verified against standard fire tests or a numeric simulation of them.

Required levels are specified in National Codes and normally depend on factors such as: - type of occupancy;

- height and size of the building;

- effectiveness of fire brigade action;

- active measures such as vents and sprinklers (but not in all countries).

In the national approaches, safety levels specific for historic buildings are normally dealt with on "case by case" basis. On a EU level there is - in this respect- no guidance what-so-ever, the main reason being that defining (fire) safety levels is seen as the responsibility of the EU member stales.

An overview of fire resistance requirements in various European countries as a function of the above factors, is given in the Annex 1 [3]. For a global summary, refer to table 1.


Iron & Steel

fry~P;;- of building ~ ~ ~ ~-- ~ ~ ~ -~~-~~ --- -
-- ISO -fire class
lOne storey No or low requirements Possibly up to R30
12 to 3 storey No up to medium requirements Possibly up to R60
More than 3 storey Medium requirements R60 to R120
!TaT1 buildings High requirements IR90 and more Table 1: Variations in required fire resistances

Allhough quite a large variation in requirements exists, one may conclude that in most countries the required fire resistance is not beyond, say, 90 to 120 minutes. If requirements are set, the minimum value is 30 minutes (some countries however have minimum requirements of 15 or 20 minutes). Intermediate values are usually given in steps of 30 minutes, leading to a scheme of 30, 60, 90, 120 .... minutes. The following general features may be identified:

- no specified fire resistance requirements for buildings with limited fire load density (say: 15-20 kg wood equivalent/m-) or where the consequences of collapse of the structure are acceptable;

- fire resistance for a specified but limited period of time, where the time requirements are mainly intended to allow for safe evacuation of occupants ane! intervention by rescue teams;

- fire resistance of the main structure to ensure that the structure can survive a full bum out of combustible materials in the building or a specified part of it.

A full fire engineering approach in which compartment and steel temperature are calculated from a consideration of the combustible material present, compartment geometry and ventilation is becoming more accepted and has shown considerable savings in fire protection costs in specific cases [3J. In such an approach also the effect of active fire safety measures (sprinklers, detection) can be accounted for in a rational manner. This is especially important in cases where significant requirements hold and the use of passive measures is less appropriate, e.g. if the use of fire insulation by boards or sprays interferes with the need to keep the cast iron construction visible.

The behaviour of cast iron at elevated temperatures

For a theoretical assessment of the fire resistance of cast iron columns, the thermal and mechanical properties of cast iron at elevated temperatures must be known. Comparison of these properties with those of steel will provide insight into the extent to which the fire behaviour of cast iron differs from that of steel. The behaviour of steel structures under fire conditions is well known.

The following thermal properties are of interest: the specific heat (cp) and the heat conductivity. These properties are presented in figure 2 as function of the temperature [4]. For reference, also the corresponding properties for structural steel are presented. Note that especially the heat conduction of cast iron is practically identical to that of steel. This means that the for steel common assumption of a uniform temperature distribution, will hold for cast iron as well. At room temperature, the specific heat of cast iron is nearly the same as for steel. However, the increase as function of temperature is less significant than for the specific heat of steel, Hence, under otherwise similar conditions, the heating rate of a cast iron will be faster than for a steel.

Iron & Steel


Sign up to vote on this title
UsefulNot useful