Article - Issue 11, February 2002

Engineering architecture: The art of engineering structures from the Pantheon to the Jubilee Line

J Robert Benaim FREng

Download the article (728 KB)

Major structures of the past were designed by ‘master builders’ who combined the roles of architect and engineer. Engineers are now seen as being concerned principally with the stability of structures while architects are responsible for appearance. In reality, civil engineering designers are continually faced with aesthetic choices as they craft their structures. Robert Benaim demonstrates that engineers can produce a distinct form of architecture, based on the refinement of their understanding of structural behaviour, and on a search for rationality and economy.


My father was an artist, a painter, and he never could understand what I did. Whenever he had to introduce me, he did so, proudly, as an architect. He knew that architects were responsible for the design and construction of the beautiful buildings that surrounded him in his native Florence, and had understood that I also was responsible for the design of buildings, ergo …

Although I was embarrassed at the time, I now believe that his misunderstanding illuminated the truth. Engineering and architecture have been artificially separated. Certainly, those who built the Florentine palazzi combined the skills of architect and engineer.

The Pantheon and the Middle Ages

I would like to go even further back in time, to 120 AD and the effective start of modern building, in fact to the Pantheon in Rome. The roof is an amazing 44 m diameter spherical concrete dome. To minimise the thrust on the walls it uses lightweight aggregate, with the density and thickness of the concrete reducing towards the crown. The underside of the dome is relieved by caissons, which have the triple role of saving materials, further reducing weight while maintaining stiffness and decorating the public space.

The dome rests on concrete walls that are 6 m thick and some 30 m high. At their base they are relieved by chapels made within their thickness. Again, these chapels have multiple functions, economising on materials, displacing the centroid of the walls outwards to improve their resistance to the thrust of the dome, and decorating the temple. The upper lifts of the walls are more massive, and one has to assume that the designer knew that this weight would help resist the thrust of the roof. There is no evidence that the designers added any of the basic building materials for aesthetic effect, but in fact made features of their measures to save materials. It has most certainly proved to be durable. This beautiful, sophisticated building demonstrates the complete integration of architecture and engineering, the creation of a master builder.

The Pantheon is an early example of what I like to call ‘engineering architecture’. Its appearance has its roots in the understanding of materials and structures, and in a search for economy and rationality rather than in aesthetic rules or transient fashion. Only those with adequate technical knowledge can achieve this form of architecture. The contrast with traditional architecture is emphasised by the classical entrance that is applied to the side of the drum.

I aim to convince you of the existence of engineering architecture as a discipline that is separate and different from architecture as it is conventionally understood. This distinction is not, in my experience, generally accepted.

The Pantheon exhibits three important attributes of engineering architecture:

  • the multiple roles of design features

  • the expressiveness of the structural options adopted

  • economy in the realisation.

These same three attributes were demonstrated by the architect/ engineers, who, in medieval France, progressed in a remarkably short time from building modest Romanesque churches to superb gothic cathedrals. Cathedrals were daring structures, at the limits of knowledge, and built economically, as the funding communities were poor and made great sacrifices to afford them.

The proportions of the cathedrals were determined by the geometric and structural logic of the vaults and arches, and by the dictates of stability. The result is a totally original form of architecture called by Viollet le Duc ‘l’Architecture Raisonnée’. This means architecture governed by reason, rather than by rules of taste and fashion, and is basically synonymous with my term of engineering architecture.

There was no separation of tasks between those who knew how to make stable and successful structures, and those who decided what the buildings should look like. These master builders combined the talents of architect, craftsman and engineer. Clearly there were specialised skills that the master builders needed to call on to complete their task. Sculptors were commissioned to provide the statues, and painters to decorate the walls and vaults with frescoes or mosaics.

The Italian Renaissance and the Industrial Revolution

Later, in the Italian Renaissance, there remained the same lack of distinction between engineering and architecture. Of course, the word ‘engineering’ had not yet been coined, but the architects of the time were responsible both for the appearance and stability of their structures.

As an example, in the fifteenth century, Brunelleschi demonstrated his engineering skill in the detailed design and the methods of construction of the dome of Santa Maria del Fiore in Florence, the largest dome that had been seen since the Roman Pantheon. In addition, he invented the cranes and other machines necessary for its construction. He was also responsible for the design of the Pazzi Chapel, at Santa Croce in Florence. This small chapel is considered one of the gems of Renaissance architecture, and is in reality a triumph of drafting, of solving extremely elegantly the geometric problems of situating a dome on a rectangular building. An understanding of engineering and architecture were seamlessly combined in such a man.

Subsequently, architecture and engineering gradually grew apart. As stately homes and municipal buildings became more in demand, presumably the emphasis in architecture moved towards the function and appearance of buildings, towards their comfort and grandeur. Clearly, the design of Italian palazzi, French chateaux and English stately homes did not generally need a very high level of engineering skill. Their structure could be taken for granted: it no longer represented a real challenge. As a result, most architects probably became de-skilled in terms of their structural expertise.

In the eighteenth century came the Industrial Revolution. The demand increased for functional buildings and structures, such as pumping stations, ports, lighthouses, canals and bridges that did require the builder to exercise state-of-the-art knowledge of statics and of the properties and strength of materials. The structurally de-skilled architects probably could not respond to these demands, and consequently a new breed of builder evolved, who could work at the limits of knowledge, and who was always seeking to push forwards, the boundaries of what was possible, just like the cathedral builders of old. Many of these early civil engineers had little or no formal training, and were certainly not trained as architects. However, they were very interested in aesthetics, and, in the tradition of the master builders, the appearance of their structures was closely related to the nature of the materials, and the decoration was that of craftsmen.

The modern context

The final parting of the ways of architecture and engineering is a relatively recent phenomenon. It was probably twentieth-century education, with its disastrous separation of the arts from the sciences, which confirmed the schism between the technical and aesthetic components of design. Architecture drifted away from building science towards the fine arts, while engineering education became ever more mathematical and specialised.

Vitruvius, writing in the first century BC, described the architect as having limited knowledge of a wide range of skills. This allowed him to understand and control the diverse specialists involved in any new building venture. Although the range of skills is now somewhat different (for instance, engineers are no longer called on to learn music so that they can adjust the tension of the twisted rope springs of a ballista), this definition is just as relevant to engineers today. The designer of a bridge should take responsibility for all aspects of the structure, which includes its appearance. Design decisions that have both aesthetic and technical components are best taken by one mind.

However, a totally different view has been gaining ground among clients and engineers. This was illustrated most clearly at a joint engineering/ architectural seminar held in Paris some years ago concerned with the design of bridges on the TGV Méditerranée. This high-speed railway passes through some of the loveliest countryside of Provence, and a great effort has been made to design structures that are beautiful. At this conference, the almost unanimous consensus was that engineers should abdicate their aesthetic role; they should practise as technicians, leaving aesthetics to the architects. Although such depressing sentiments have not been expressed so clearly in the UK, many in the profession and in client bodies hold the same view.

Owing to the specialisation of engineering education, which has divorced engineers from art, and indeed from the history of their profession, they have not surprisingly lost confidence in their aesthetic judgement. Many engineers do need the support of architects.

However, the essence is not whether architects are involved in the design of civil engineering structures, but whether the appearance of these structures is defined by the engineer’s logic or by the architect’s fancy. If the latter, the appearance will be reduced to rootless taste, to whimsy, producing ‘fairground structures’, parodying engineering, and usually grossly extravagant. In my experience, good architects appreciate these problems, and, as long as they are collaborating with an engineer whom they respect, and who has the commitment to work hard to understand the behaviour of the structure, they are happy to help the engineer produce a design of quality.

Engineering and art have very much in common. One of the central characteristics of art is an attention to detail, an understanding of the importance of nuances, which outsiders may consider obsessive. However, it is this concern with minutiae that makes the difference between art that fails to achieve its aim, and that which succeeds. This is very close to the method of an engineering designer, who needs continually to refine a concept, to check repeatedly their understanding of the mechanics of the problem, and gradually, painstakingly to reduce a structure to the simplest expression of the applied forces. Antoine de St Exupery’s aphorism that a work of art is perfect not when there is nothing to add, but when there is nothing more to take away, is equally applicable to engineering design.

Engineers need to learn to cultivate this painstaking return to first principles, this intellectual analysis that allows them gradually to eliminate all received ideas, ending up with a clear understanding of the structural issues. Such clarity is the foundation for engineering architecture.

The Byker Viaduct

Three of the principal attributes of engineering architecture, as introduced when discussing the Pantheon, are multiple roles for design features, the expressiveness of the structural options adopted, and economy in the realisation. These are well illustrated in the design of the Byker Viaduct. The twin columns of this rail bridge splay at their base to increase the stability of the bridge, and are relieved by arches. These arches remove excess material, improve the appearance for pedestrians in the valley, and they were sized so that the largest precast deck units could fit through, in an innovative construction sequence. The twin columns provide:

  • a support that can flex under length changes of the deck and remove the need for bearings

  • fixity to the deck to improve its economy

  • a stable base for the balanced cantilever construction.

This bridge was economical as well as being elegant, and in scale with its nineteenth-century neighbours. It was a successful collaboration between engineer and architect, with the engineering logic being expressed in its appearance.

The Ah Kai Sha Bridge

In order to illustrate the relationship between technical and aesthetic design decisions, I will describe briefly the thinking that lead to the shape of the towers for a cable stayed bridge I designed recently.

The Ah Kai Sha Bridge has a main span of 360 m, is exceptionally wide at 42 m, and carries ten lanes of traffic on the top deck, and six lanes inside the box. The towers provide the height necessary to attach the stay cables that support the deck. They also carry the loads to the foundations, and give stability to the bridge under the effects of wind and earthquake. All bridge decks expand and contract with changing temperature, and either the deck has to be separated from the column by bearings to allow this, or the columns have to be made sufficiently flexible. Here, the deck has been fixed to the columns, which have been split into two leaves to make them flexible.

The towers cantilever from the foundations and rise 100 m into the air. Their tapering shape is determined by the progressive increase in weight applied to them by the stay cables and by the deck, by the forces imposed by expansion and contraction of the deck, and by the wind and seismic loads which gradually increase their effect towards the tower base.

Each leaf of the tower has a dumbbell shape, which is the most efficient use of material. The dumb-bells become solid at the junction with the deck to resist the concentration of forces that occurs in that zone.

Most cable stayed bridge towers have at least two cross beams which make the towers into portals. These cross beams are heavily reinforced, and slow down the construction of the columns. The last straw that made us decide to omit cross beams was the great width of the deck, which would have required the beams to be very substantial. Support for the deck, in the absence of cross beams, is provided by powerful prestressed brackets, which offer little obstruction to slip-forming.

The towers are outside the deck, and, as a result, the stay cables all pull slightly inwards. The combined effect of these pulls is very significant, and would require larger columns if they were not propped apart at the top. The prop is designed to be made on the bridge deck and winched up into place and so has to be as light as possible. It has to resist buckling under compression, and bending under its own weight. The connections of the prop with the tower must be sufficiently small that they do not attract bending moments. The shape of the top prop reflects faithfully these various constraints and actions.

Every significant dimension of these towers had a technical, rational justification. None were chosen for appearance alone. However, the appearance of the towers was present in the mind of the designer at all times: ‘l’architecture raisonnée’.

As a comparison, the Annacis Bridge solved some of the same problems differently. This splendid bridge is much narrower than Ah Kai Sha, and has only a single level of traffic. The deck rests on bearings carried by the lower cross beam. The towers are outside the deck, as at Ah Kai Sha, and hence the cables would all pull inwards. Here, the designers have cranked the towers inwards above the deck, so that the stay cables pull concentrically, and consequently a top strut is not required. The transverse stability of the towers is provided by the portal action of the two cross beams. I believe this comparison underlines the engineering source of the different appearance of the towers for these two bridges.

The Jubilee Line

The Jubilee Line extension gave many opportunities for engineering architecture arrived at by teamwork between engineers and architects. Major underground excavations require the skills of engineers, while the organisation of the stations needs the particular skills of architects. At North Greenwich station, an analysis of the pedestrian flows allowed the replacement of the traditional full-width concourse slab with a suspended walkway, opening up the volume of the station. Also, an analysis of the measures to avoid progressive collapse in the event of the destruction of one of the columns lead to their arrangement in a truss pattern, which was adopted by the architects as a major design feature.

Although I have been using examples from my own domain of civil and structural engineering, the principles of engineering aesthetics are by no means limited to this field. For example, the beauty of the World War II fighter aircraft the Supermarine Spitfire was the result of an engineer, faced with a totally functional brief, solving his problems and refining his design, tempered by a sense of proportion and beauty.


I hope I have demonstrated that engineering architecture can be created by engineers alone or in collaboration with architects. It cannot be created by architects alone, or by architects who impose their vision on engineers. Its essence is that it reflects rational engineering thought, expressed with elegance.

Faith in the existence of engineering architecture, and courage to persevere in the difficult road of rational design is easily found in a study of the works of great engineers, past and present, such as Robert Maillart, Pier Luigi Nervi, Eugene Freyssinet and Christian Menn, to name but a few.

These, and many other engineers who are designing structures of beauty by concentrating on refining the engineering, are the best proof of the existence of this special form of architecture, which has its roots in the knowledge of materials, in the understanding of structures, in the expertise in the building process and in the obsessive search for rationality and economy.

Their structures were not, and could not have been, the work of architects. The inspiration for their shapes, for their beauty, came from a source that is not accessible to someone not trained as an engineer.

The worst possible outcome of the ongoing debate on the way to improve the quality of our civil engineering structures would be that clients conclude that design teams must be lead by architects, and that engineers should provide a structural service, as for the design of routine buildings; this is the direction in which our French colleagues appear to be heading.

If one rejects that route and accepts the existence and the value of engineering architecture, then the industry needs to be organised to foster it. Our current system of engineering education stifles creativity. Exceptional engineers, such as those cited above, break through and express themselves. If we wish to improve the appearance of our civil engineering structures, we need to respect the qualities of those engineers, understand their art, and organise our profession to make it easier to emulate them. This requires principally that the training of engineering students who intend to become designers of major works would need to be fundamentally changed to allow them to understand and assume their architectural heritage and responsibility.

Engineering architecture is not a new concept. The search for rational and economical solutions has driven the design of some of the greatest functional buildings and civil engineering structures of antiquity, of the middle ages and of modern times. The sine qua non of this form of art is the highest level of engineering competence allied to artistic sensitivity.

After obtaining degrees at Imperial College, Robert Benaim specialised in the design of pre-stressed concrete structures for seven years in France. He then spent 11 years with Ove Arup & Partners. In 1980 he formed Robert Benaim & Associates with offices established in London, Hong Kong and Malaysia. He withdrew to private practice in 2000.

[Top of the page]