This page complies with "XHTML 1.0 Strict" but should be intelligible in non-compliant browsers.
Site navigation
[Home.]
["The art in structural design."]
["Aesthetics of built form."]
[Work of Jörg Schlaich.]
[John Monash's engineering prior to WW1.]
[Combined References] (for the 'Design of Structures' section of this website)
Preliminary notes
When this paper was published, some readers interpreted it as an attack on those who practised conventional methods of teaching the 'design' of structures. Some, alas, saw it as an attack on particular departments of civil engineering. Admittedly, the tone of the article was designed to goad academics into taking a new look at their work; but I envisaged it more as a 'call to arms' which, despite my own attempts to break out of the mould, applied to myself as much as others.
With less justification, I was accused of being "anti-theory". In fact, my position was (and is) "the more theory the better - and the safer" However, theory should not crowd broader issues out of the syllabus, and individual theories should be presented in the context of their development and applicability.
When I submitted my published work for recognition as a PhD in 1995, I added notes to the paper to explain sources and provide additional reasoning. These have been included here, slightly modified, but the references are now very much out-of-date. [A.H. Jan 2008.]
The original paper follows:
1. Summary.
In recent years much attention has been paid to developing the study and teaching of structural design. This paper discusses further opportunities for establishing the field as a recognised academic discipline and looks at some of the obstacles which present themselves. It is assumed that an academic discipline has three aspects: research, scholarship, and teaching. It is suggested that there be greater concern with descriptive material in all these fields, and a wider application of special teaching methods. It is noted that these modifications cannot be successful without changes in attitudes of staff and students. There is a need to recruit into both areas some individuals with a wider range of interests than is now the norm.
2. Introduction.
As this paper is intended to promote discussion, an informal tone has been adopted. Statements and arguments are based on personal experience of a small number of universities (mainly in Australia and the United Kingdom) and a selective reading of relevant literature.
I adopt the view that a true understanding of the design and behaviour of structures cannot by gained by studying formulas and computational processes. There has been considerable discussion of the need to improve upon conventional teaching, and many initiatives have been reported. Much work of a scholarly nature has been published which looks at structures from a wider perspective. It is possible that by building on this foundation we could achieve a fresh definition of the discipline. There is no intention in this to deny the importance of rigorous and complex analysis and purely technical knowledge. It is one of the characteristics of engineering which sets it apart from many other fields of endeavour and provides many of its challenges and satisfactions. There will always be a need for practitioners and academics who specialise in this work. It is assumed that these points are taken for granted throughout the following discussion.
3. The Traditional Approach.
3.1 Disadvantages.
As readers are well aware of the strengths of the traditional approach, this brief review concentrates on its disadvantages.
Not surprisingly, the accent is on equipping students with the ability to do things. The often unspoken aim is to produce a graduate who will be able to work as a sort of human computer. There is little effort to prepare the student for managerial responsibility. In the field of structures it is assumed that there is little point in discussing questions of overall form because the raw graduate will not have responsibility for such matters. It is sometimes argued that it would be harmful to encourage an enthusiasm which might be dashed when the graduate realises the long apprenticeship that awaits. This may be a reason why it is not normal for academics in engineering faculties to research and marshal information about the factors which govern the form of structures. Or it may be that this aspect of scholarship is not considered a suitable field for serious study by engineer-academics because it does not involve mathematics and experimentation. The result is the anomaly that students of architecture are often exposed to a wider general knowledge of structural forms and behaviour than engineering students.
Note inserted for PhD: My favourite examples at the time of writing were Howard (1966), Salvadori and Heller (1975), and Siegel (1962). [For details see PhD Bibliography.]
Thus the emphasis in traditional engineering education is on teaching 'how' rather than 'what' or 'about'. Harris (1980) notes that there is an old joke that "engineers know everything about what they do except why". A former colleague maintained that his undergraduate education was characterised by a session on the 'design' of reinforced concrete spread footings in which the lecturer did not think it necessary to explain what a spread footing was, what its purpose was, or why a designer might choose to use it in preference to any other sort of footing. The presentation was limited to the standard procedure for calculating the depth of concrete and the area of reinforcement required for a given set of loads, given plan dimensions, given material properties, and given safety factor. That was called 'design'. Brooks (1989) writes that "the … adage that engineering is an exact science based on assumptions is perhaps truest in structural engineering". Yet traditionally we, and those who write our textbooks, tell students very little about those assumptions and concentrate instead on the apparent certainties of computational procedures.
The attitudes which contribute to this state of affairs are not confined to staff. They are widely shared by students who have graduated from their final years of schooling having studied only mathematics, physics and chemistry. It is common for such students to express contempt for any descriptive subject which contains matter that cannot be quantified for automatic manipulation by formula. If they are presented with descriptive material about buildings, no matter how important and essential to the design process, many become restive. If the lecturer writes a few formulas on the blackboard, they will happily copy them down under the impression that they are now doing some 'real engineering'.
Even essential subjects such as Materials are liable to be dismissed as "mere memory work". Colleagues who teach the subject complain that such students are liable to recognise certain key words in examination questions and simply perform a 'memory dump' onto their examination script of all concepts remotely connected with the key words. The resulting answer shows no real comprehension of the links between the concepts, and no ability to organise them into a coherent and relevant response to the question.
On the other hand, as J. Cowan (1986) has noted, conventional examination questions in reputedly 'rigorous' subjects such as the analysis of structures tend to fall into classes which students can learn to recognise. It is thus easier to cope with conventional quantitative questions than with descriptive questions which require real understanding. I have marked several examination papers in which a frame analysis, including the drawing of a complex bending moment diagram, has been performed perfectly, while on the next page the student has shown complete lack of understanding of the bending moments within a cantilever carrying a point load at its end. (In this classic mistake, the triangle is placed the wrong way round, so that the moment at the support is shown as zero and that at the free end as PL.)
It is natural that most of our present cohort of students, who have been reared on conventional questions in high school and the early years of university, strongly support the status quo. However, the status quo in many schools is changing. There is a trend towards methods of teaching and assessment which are more suitable to open-ended problem-solving. This is continuing despite strong opposition from those who equate academic rigour with closed-endedness. Students who emerge from this process may expect from staff a level of general knowledge and understanding of structures which many of us do not possess, and will be bored with conventional closed-ended questions in practice classes and examinations.
There is evidence that our preoccupation with quantitative subjects does not necessarily produce the best result. Richards (1984), an architect, has noted that engineers
"seem to have been trained to think of design as a linear process, where, by synthesising all data known at the outset, the final design will eventually emerge. This is a common cause of difficulty in synchronising the inputs of engineers and architects working on the same design problem … The input offered is unnecessarily accurate, and too slow, to be useful at the early stages of the process."
He asks, "Is this because engineers are trained by a method suitable for the handling of only a small number of variables, and these variables easily quantified?"
Undervaluing descriptive material has other dangers. It is common in faculties of engineering to provide 'service courses' for electrical or chemical engineers in topics such as structures. A common justification is that they give such students "an idea of how structural engineers think" and thus make for better cooperation in practice. However, what is provided is usually an elementary course in quantitative analysis which has been designed as a first step in the civil engineering syllabus. The dangers were illustrated for me when a practising mechanical engineer asked for help because his design of a three-storey rigid frame supporting major items of mechanical plant had been rejected by the local council. He expected me to convince the council of the error of their ways. He had analysed the frame using the moment distribution method without allowance for sway, and had used Euler's formula for the design of the columns. The entire frame was already fabricated and about to be assembled.
In his undergraduate course this engineer had encountered only elementary 'how to' courses about structures without completing the later-year courses which give civil engineering students some introduction to real design concepts and codes of practice. The latter, for all their faults, do at least encapsulate the collective wisdom of the profession in bridging the gap between theory and reality. An 'about' (or 'descriptive') course specially designed for such students would be much safer and more successful in teaching them 'how structural engineers think' than any course in basic theory of mechanics. Developing a descriptive course of that nature would present a considerable challenge to the lecturer. It would be interesting to know if any exist.
There is an equal need for subjects which describe the challenges, the responsibilities, and the creativity of our profession in a way that can be apprehended by students of business, commerce, and law. The lack of such subjects must have much to do with the general ignorance of engineering amongst other professions and the public, and the resulting lack of status. With the increase in interdisciplinary degrees which allow students to cross the traditional barriers between engineering, law, commerce, and languages, it is important to ensure that the traffic is not all one way, with engineers reaching out in all these directions, but the others remaining ignorant of our discipline.
A more descriptive approach to the teaching of structures provides greater flexibility in the scheduling of subjects. This is becoming important as more institutions drop the 'pass-by-year' system in favour of a subject-based credit-point system. Questions arise such as 'Is it possible for students to study Timber Design 3456, in which they may encounter a continuous beam, if they have not yet passed Structures 2345, in which the analysis of continuous beams is covered?' Some staff feel that the answer is 'no' because they associate an 'understanding' of continuous beams with an ability to derive their shear force and bending moment diagrams. However, if students have at least an elementary knowledge of bending moment and shear force diagrams gained from a first year course in simply-supported beams, it must be possible to tell them what the bending moment diagram of a continuous beam looks like, or to explain the principle of WL/10. They can then proceed to learn the basic rules for sizing in timber as well as a great deal more about its properties, and about design and form in timber structures. When these students later come to the analysis of continuous beams they will see more point in their studies because they have encountered a practical application for the results.
Similar considerations apply when teaching about code provisions for a very complex phenomenon such as flexural-torsional buckling due to bending moment, or local buckling of plates in compression (e.g. flanges of steel I beams). There is not time to follow through the relevant theory. The best that can be done is to describe or demonstrate the physical behaviour and the major parameters which influence it, and then simply quote the formula.
3.2 Indicators of the need for change.
Practitioners have complained for some time about the quality of graduates emerging from the traditional course. A consultant in London said some time ago that they were so lacking in creativity that he had seriously considered appointing "numerate arts graduates" to his practice.
PhD Note. This was based on a private communication (1987) from Mr Andrew Smith at the Bartlett School of Architecture, University College, London, who was working with Alan Baxter and Associates. The remark may have been made in exasperation, but is significant nonetheless.
Discussing the prominence given to analytical skills, Jay (1984) has suggested that "… engineering courses tend to perpetuate those qualities that the educators find admirable and that were found admirable in them." Armitage in the same discussion (1983) concluded "… the present balance of academic curricula needs to be shifted away from scientific and mathematical processes towards awareness, understanding, communication, and design." Many academics share this viewpoint. The importance and usefulness of qualitative understanding of structural behaviour has been forcefully argued by Brohn (1982, 1984). The number of papers which continue to appear describing innovations in teaching methods and the encouragement of creativity in students, show that there is a lively interest at many teaching institutions.
Recent developments in computing have major implications for our profession. Packages for structural analysis have made redundant the art of the experienced analyst who had developed a 'feel' for which one of half-a-dozen 'manual' techniques would be most appropriate to a given problem. What is still required is the art of modelling, mainly in finite elements, but this is not necessary for routine structural design.
PhD Note. My syntax here was confused and the statement could be faulted. It is true that finite element analysis is not required for the design of routine structures. It is also true that the analytical models used in routine design are well established by convention and relatively clearly defined, and that little conscious thought need be applied to modelling. However, some form of modelling must always occur.
We are now seeing the production of 'design' packages which carry out member sizing of complete buildings according to the relevant code of practice. It is possible to imagine architects and even developers carrying out their own 'design' work for conventional buildings. When such packages are connected directly to others which produce drawings and specifications, it will be interesting to see what changes occur in the work of the majority of structural engineers. Expert systems are also becoming available which will provide advice to the architect and layperson, as well as the practising engineer.
Those who set examination papers are facing the fact that hand-held calculators now have the potential to store routines for solving quantitative questions, or the equivalent of five thousand words of 'crib' notes. It is possible to avoid the consequences by banning such calculators from the examination room, but disturbing questions remain. If the skills and knowledge we are examining may be programmed, stored, and regurgitated by a machine, what real value do they have?
3.3 Hopeful signs.
There are signs of new trends in scholarship, research, and teaching about Structures which offer hope of a more appropriate approach to the discipline. There is not space to make full mention of even those that are known to me, or to include full bibliographic references. However, the work of people such as Billington, Happold, Leonhardt, Medwadowski, Salvadori, and Schlaich is typical.
Reviews of building practice are becoming more common, e.g. Morris (1978) and Orton (1988). Given the lead time required for research, and for preparation and publication of such books, the information they contain must be somewhat out-of-date by the time it is published. However, they do serve a great purpose in drawing together information, identifying basic principles and trends in design, and presenting these in a coherent and accessible fashion. Unfortunately, while some lecturers may recommend this type of resource material as suitable background reading for their overworked students, I suspect that very little of it is actually used in class because of the attitudes to descriptive material described above. The result is that the majority of students see no benefit in reading it when they could be gaining credits by completing urgent assignments involving conventional subject matter.
4. Subject matter of the proposed discipline.
4.1 General Knowledge.
In practice, the content of a newly-defined discipline and of the courses associated with it will be limited by several factors. Current academic staff are specialists. To be a generalist is be to second-class. Descriptive material is accorded less intellectual merit than quantitative material. Analysis is valued over synthesis. Many academics feel unable or unwilling to teach and research practical design. The time available to engineers, students and academic staff is limited and student/staff ratios are higher than ever. There are unresolved questions concerning the 'teaching' of engineering design: can it be 'taught' and if so, how? In preparing the following ideal list of proposed contents for the discipline, these practical considerations have been set aside for the time being. The approach has been to ask 'What should an educated engineer know about the world of structures?'.
An essential element must be a knowledge of the way in which design is organised: how building and industrial projects are initiated, and the role of the various participants in the process. Building economics is another important aspect that is often omitted from courses in civil engineering. The objective would not be to produce graduates who could step into a new job and immediately turn out an accurate cost estimate (if such a thing exists); but to ensure that they are aware that cost estimates are done; have some idea of how they are done; and realise that economics is an important constraint on design.
The educated engineer should know about the functional and aesthetic objectives of buildings and structures; the nature and origin of the loads that must be resisted; the opportunities and constraints offered by current processes of fabrication and construction and by site conditions; and the criteria for success in terms of function, strength, serviceability, and cost. He or she should know about considerations of risk and the application of safety factors, and the responsibility which the profession shoulders in this regard on behalf of society. This would touch on attitudes governing provision for rare events.
Structural engineers cannot be considered educated unless they have some knowledge of the process by which theories are developed. They should know the approximations and assumptions that lie behind the formulas and sizing procedures they use. They should know the accuracy achievable, and the range of parameters (or practical situations) over which each theory or formula is reliable. They should have studied the art of modelling the physical behaviour of structures and loading cases.
The means whereby theories are moulded into a set of codified rules is a fascinating study in itself. It encapsulates much of the philosophy of the engineer. It illustrates the need to generalise, simplify, and approximate, and the ingenuity which has been applied in arriving at rules which strike a balance between generality and economy. The paper by Kerensky, Flint and Brown on the BS153 rules concerning lateral stability of steel beams is an excellent example of this, and there are many more. These studies would lead naturally to a consideration of the merits and disadvantages of codes.
PhD Note. Relevant papers at the time of writing included Kerensky et al (1956), ACI-ASCE Committee 326 (1962), Siess (1960), and Harrison (1968). [For details see PhD Bibliography.]
Background knowledge of this nature would help graduates when they confronted the common problem of the structural system which does not quite conform to the assumptions contained in the code rules. They would have a better idea of the extent to which they could bend or extend the rules without undue increase in risk.
An educated engineer should have a wide knowledge of structural forms and how they relate to the functions of structures. There is much information on buildings and their functions, particularly in German. So far, however, most of this work has been left to architects, because engineer-academics have been preoccupied with computational work.
PhD Note. My favourite example at the time of writing was the E+P (Entwurf und Planung) series of books published by Callwey of Munich. (Many of these were published in English under the heading Design and Planning by van Nostrand Reinhold, New York.) For many years the Architectural Record contained a series of articles under the heading of The Architectural Record Building Types Study. See also the section on Sources in Chapter 5 of The Art in Structural Design.
The result is that there is little if any discussion of the logic behind structural design decisions, and little coverage of heavy industrial and engineering structures. The subject of structural form is better served, although we are again heavily indebted to the architects. The development and maintenance of such a body of resource material, whether in the form of books or expert systems, is a task for scholarship in the field of construction and design. Much material may be amassed from periodicals, but it would be better obtained directly from the designers concerned. A general knowledge of structures, including typical sizes of members and overall dimensions, or an expert system knowledge base, would enable graduates to check computer output more reliably, and go some way to replacing the hard-earned experience which once ensured that gross errors were not allowed to pass.
An educated engineer should have some knowledge of the history of our profession. Teaching would need to be skilful, because few students with a background in mathematics, physics, and chemistry have a natural interest in the subject. History prior to the mid-nineteenth century is considered totally irrelevant. The best approach might be to start by teaching something that happened last year as history. In this way the line between 'case study' and 'history' may be made less definite, and it may be possible to teach backwards, so that history creeps up on the students unnoticed! To discern the relevance of the lessons of history could be an enlightening experience for all concerned. The study of history for its own sake would be a suitable topic for academics, postgraduate students and the exceptional undergraduate. A knowledge of recent case studies should in any case be an integral part of the discipline. A great deal is to be learned from an analysis of failures but, as Harris (1980) has pointed out, it is most important that successes be given at least as much prominence.
The student's general knowledge of the world of structural engineering should include a knowledge of personalities and organisations. This should comprise simple factual knowledge of 'who is who' and 'who designed and constructed what'. It could also include the sort of study of personalities recommended by Billington: a study of creative designers at a reasonably intimate level to try to determine what qualities have made them special. [Billington (1970), p.121.]
4.2 Design process.
Students should be familiar with the attempts which have been made so far to describe and explain the workings of the design process. These come under the heading of design methodology. There is an extensive literature on the subject, although it relates mainly to mechanical engineering and architecture. A brief introduction is included in Holgate (1986a), Chapter 13 to Chapter 17.
Such descriptions are no substitute for design experience or even for simulated experience, but a knowledge of them will help students grasp and interpret their experiences of design and increase their ability to file this knowledge away for future recall. Students could be given at least a simple three-stage model of the design process such as 'definition, generation of trial solutions, evaluation' and told something about design strategy (direction and balance), exploration of design space (cycling) and the 'patent' methods for improving design efficiency such as morphological charts, decision trees, brainstorming, and synectics. They could be given some practice in applying these to simple structural design projects.
It would be useful to introduce some consideration of developments in computer packages and expert systems. The emphasis would be on understanding the construction of such systems, the limitations which result, and the effect of these facilities on the work of the engineer.
4.3 'Design ability' and creativity.
Educated engineers should know that being a good designer involves much more than creativity. They should know the importance of having a good business sense, and of being able to work with others, and something about the social dynamics of groups. This sort of knowledge and related skills can be fostered only through experience, but it might again be useful to discuss them in the abstract to provide concepts to assist the student and researcher to analyse and store their acquired experience. Such concepts might include aspects of design strategy such as 'tolerance of ambiguity' (resisting the temptation to settle immediately for the first solution that springs to mind); deciding when to stop searching for and testing trial solutions; deciding whether to try to rescue a good idea that seems to be going wrong, or to give up and strike out in a new direction. It is important to teach about the "burden of decision" (Harris) and the level-headed courage necessary for innovative design. Such material could lead on to discussion of the role of the junior engineer in a design organisation, and the need for information to be passed up the chain of command as well as down. As Cowan (1981) notes, little can be achieved by telling someone how to design [Cowan, 1981, p.748]. Therefore design as a lecture-room discipline must be run in tandem with design projects. Cowan's students developed an ability to "stand outside themselves" in order to record and analyse their own strategy while they were tackling design projects. It seems possible that this ability could be enhanced by a lecture-room discussion of the subjects mentioned above. More could be achieved by considering the list of questions which Armstrong (1983) suggests engineers should ask themselves throughout their life in practice.
Similar considerations apply in treating individual and group creativity as an academic discipline (see Harris 1980, Holgate 1986a, Cowan 1981, and Armstrong 1983). An examination of the problem of defining and measuring 'design ability' and creativity would provide an opportunity to discuss and question its nature.
Measures have been based on personal 'success' or recognition; on an estimation of the quality of the product in terms of commercial return, popular acclaim, or peer judgement; on identifying processes or behaviour considered to be typical of creative people; and on simple tests similar to those used for intelligence. An interesting question is whether creativity and design ability in engineering are different in nature from their counterparts in scientific research and in architecture. A valuable contribution could be made by studying the characteristics of creative people. Students might be encouraged to emulate them, or at least to tolerate their personal characteristics which are, apparently, not always amiable. (This could be tempered by an examination of the importance of cooperation and teamwork in engineering projects.) More likeable characteristics attributed to creative people are sensitivity to problems, a tolerance of ambiguity, a high level of motivation, the level-headed courage referred to earlier, intuition, and an ability to change their mental frame of reference appropriately (lateral thinking). Such studies could be tied in with the study of individual engineers referred to under the heading of general knowledge.
Exercises in techniques for removing 'blocks' to creative thinking could be practised or studied. These include self-consciousness, fear of failure, cultural conditioning, and the type of 'linear' thinking that is encouraged in the conventional education of scientists and technologists. It would be important to stress the value of play and the fact that it is permissible, and even profitable, to enjoy oneself while designing. Exercises in brainstorming and synectics have already been mentioned. Projects demanding creativity, such as the making and testing of structural models in common household materials could also be pursued. It is gratifying that the latter are now reasonably commonplace.
PhD Note. For more on the above see Chapters 13 to 17 of The Art in Structural Design.
4.4 Aesthetics.
An important descriptive subject which could be included in the discipline is the aesthetics of built form. Few courses in civil engineering contain even a mention of this subject. As far as is humanly possible all aspects of structural design should be taken into account simultaneously and aesthetics should be at least equal among these (Schlaich 1986). Some people are innately more sensitive to the visual world than others. However, experience with courses at postgraduate level suggests that all may develop whatever natural ability they have. This may be done by means of lectures and discussion of the principles of visual analysis and the appreciation of architecture.
PhD Note. This statement was based on experience with a subject entitled 'Structural Engineering and Ideals of Architecture' which I offered in the 1990s as part of the 'Masters by Coursework' degree at the Department of Civil Engineering, Monash University.
4.5 Analysis and comment.
A wider and deeper knowledge of the design process and of our stock of completed projects may lead to a growth of criticism of particular designs, and this would present difficulties. The critic of structural projects would be open to charges of affecting the earning capacity of practitioners. Ligo (1984) attributes the superficiality of much architectural criticism to the fact that in the 1920s in the USA any criticism of the real functioning of a building would bring a law suit claiming loss of business. [Ligo, 1984, p.16.] I have been told that even criticism at the superficial level is currently influenced in Australia by threats of legal action.
PhD Note. This information came from Professor Conrad Hamann, Dept. of Visual Arts, Monash University who cited the cases: Seidler vs Fairfax 1984 dismissed, and Andrews vs Fairfax (1980) which was upheld.
However, it is possible that regular comment on structural design would raise the level of appreciation and understanding amongst the general public and even within the profession. While much of the general criticism that we encounter is negative, ill-informed, or highly subjective, there is a strong tradition of thoughtful and constructive criticism in the arts which has its own body of theory. Several forms of criticism are identified. In one, the aim is simply to present the facts surrounding the creation of an artefact as objectively as possible, so that the observer is thoroughly informed. It is thus a matter of analysis and education rather than judgement.
PhD Note. I depended for this on treatises such as Beardsley (1958) pp.454-89. Ligo (1984) states (p.3) that types of criticism are discussed in Weimer, D. R. 'Lewis Mumford and the design of criticism' Arts and Architecture Vol.79, Sept 1962, p.14.
An informed critic would be aware that a multitude of factors affects design decisions and that no-one could be sure of having discovered them all, even with the cooperation of the designer. In addition, many facts are of a politically sensitive nature or are commercial secrets, and if they are revealed at all will be communicated 'off the record'. Thus any assessment is liable to be based on inadequate evidence. In any case, the normal journal article does not allow room to provide proper coverage. The critic would have to offer criticism in this light (tentatively) and expect some of it to be negated by the revelation of more facts.
The practical-minded critic would also know that different people have different perceptions of the same events, and quite different recall, and that much of what happens in the development of a project is beyond the control of even the principal participants. There is a high degree of randomness in the acquisition of knowledge, the testing of hypotheses, and in the order in which external events and inputs impinge on the design. The need to put a finite end to development of a design may cut it short at a critical time and result in the loss of a potentially better solution.
Thus the concept of 'responsible journalism' is particularly applicable in this context. Scholars and researchers would need to be careful how they reported projects, and would need to build up gradually a reputation for fairness and discretion.
In the worlds of art and architecture there is a complex relationship between practitioners and critics. Practitioners usually gain financially from the publicity and may derive some benefit from the public or private interaction of minds. Informed criticism would also help educate the general public and especially the legal profession about the difficulties of engineering design and the concept of risk appraisal. It would acquaint them with the fact that engineers must routinely make decisions under uncertainty, either because knowledge is not available or because of the problems of organisation and communication within a large group of interacting parties.
The subject of analysis and comment on real projects would perhaps be most suitable for scholarship and research work undertaken by academics and postgraduate students.
In the traditional engineering course undergraduates do mainly quantitative work that is either right or wrong. There is little room for difference of opinion. On the other hand, architecture students regularly experience reviews of projects by academics and practitioners. This must have something to do with the vigour and success with which practising architects are able to defend their viewpoints in project meetings. If engineering students are encouraged to tackle some projects on the basis of qualitative knowledge it will be possible to give them this experience to their lasting benefit. Naturally, criticism should be fair and balanced. As Harris has said, "it is doing them a favour to stiffen their backbones", but criticism must not be destructive, and the most important thing is to foster self-criticism.
Thus, if comment by scholars is not acceptable to practitioners, then perhaps more self-criticism could be encouraged. Some designers are willing to be quite frank, at least in private, about the fact that given a second chance they would have tackled a project, or certain aspects of it, in a different way. Also, it should be fair to ask a designer: 'If you were designing that building or bridge now, rather than twenty years ago, in what way would it be different?' The answer could provide many useful lessons for scholars and students, and an insight into the designer's methods and philosophy.
5. Scholarship and research.
A major switch of emphasis in the discipline of Structures would alter the balance between computation and description in scholarship and research. Scholarship is considered here as the acquisition and preservation of knowledge, and the continuing maintenance and updating of the common 'data bank'.
There appears to be little observational research going on into the actual processes of structural design in professional practice, at either the level of the individual designer or of the team. (One instance is the work of Addis at the University of Reading.) Suitable topics might be: the source of the designer's knowledge and skills, its extent, and its nature, and how it has been developed; the action of groups, and problems which arise due to factors such as communication problems and personality differences; the element of randomness in design strategy (see above); and the conditions under which specialist advice is sought. Research engineers could cooperate with psychologists and sociologists to investigate some of these questions, and especially the nature of innovation and creativity in structural design. These aspects of the discipline would lie part-way between technology and the humanities.
PhD Note. In making this statement I overlooked the investigations of researchers working on the development of computer-based 'expert systems'. An early example is Baker and Fenves (1987). Designers were asked to 'think aloud' while tackling typical problems in the design of structures. An example of more recent [1991] work is the PhD thesis by Moore (1991) which has been reported in AI Review, 5, 255-71 and Civil Engineering Systems, 8, 81-6, both 1991. The work of Dr Addis was described in a private communication. He spent several months in a design office studying the modus operandi of structural designer engineers. To the best of my knowledge [as at 1995] this work has not been directly reported, but is reflected in Addis's general work on the nature of design mentioned in the PhD Bibliography.]
[Note, 2003.] Some of William (Bill) Addis's more recent work is listed on the
Structurae website.
Orton provides a good example of how the quantitative and the qualitative may be combined. [Orton, 1988.] For each of the buildings he discusses, he combines technical facts with a summary of the rationale governing the choice of form and structure and an appreciation of the architectural qualities. Scholars could prepare texts describing and discussing the work of individual engineers, similar to those produced on famous architects. These would of course have a more technical emphasis. A case-study approach would be appropriate, following the genesis of the structure and its development throughout the design process, taking account of the full range of factors which influenced its form.
Practising engineers could make a great contribution to the accumulation of accounts suitable for case study. However, it will be necessary to persuade many of them that it is worth their while to spend time on documenting their projects. They are often keener to meet the challenges and rewards of their next project than to analyse the details of a previous project which in their terms is already 'history'.
Scholars will encounter the problems of investigation and reporting discussed above. When two of my students interviewed a local structural engineer about a multi-storey building project they noticed that he was reserved when speaking of the architects. He later told me privately that he felt the architects had an unprofessional approach to their work but, because they supplied him with most of his livelihood, he had been unwilling to reveal this to the students. It may be equally embarrassing to reveal the way in which decisions have been influenced by political factors, or the personality of the participants in the project. It is unfortunate that this sort of information is currently found almost exclusively in judicial investigations of failures. [See e.g. Victoria, Australia (1971).]
While it would be idealistic to think that designers could ever be completely open about such things, it might be beneficial for the profession if the general public and other professions were made aware of the complexity of engineering design and the important role played in it by human factors. This could be done by the intervention of skilled reporters and analysts whose judgement could be trusted by practitioners.
6. Possible initiatives.
6.1 Innovations in teaching techniques.
A number of special techniques have been tried for transmitting 'general knowledge' of Structures and Design, and some of these will be reviewed here. Various types of design project, ranging from conventional exercises in sizing through to tests of creativity, have been reported at length in the literature and will not be included.
It is always effective to entice students out of the classroom. A simple means is to develop a 'structures trail' along the lines of the 'mathematics trails' developed for school children. The concept is similar to that of the a car rally. Students are required to answer a list of questions concerning the characteristics of local buildings and other structures, such that they are obliged to visit them, take measurements, and think about the way they work and their approximate proportions and stress levels. Smith has used a similar technique at the Bartlett School of Architecture using historical (mainly nineteenth century) structures as well as modern ones. The differences in design practice make more evident the influence of material properties and available techniques on structural form.
PhD Note. The reference above was to the work of Mr Andrew Smith at the Bartlett School of Architecture, University College, London (observed in 1987).
Another method is to ask teams of students to investigate the planning, design, and construction of a local structure by talking to the major participants in the process: clients, architects, design engineers, contractors, local government officials, and sometimes representatives of the National Trust. The students become highly motivated. Besides learning a great deal about the way the real world works, they also gain in self confidence. They sometimes unearth sensitive information and must be counselled on dealing with this.
PhD Note. Instances ranged from engineers' uncomplimentary remarks about architects on whom they depended for their "bread and butter"; to inter-office jealousies; and in one case mis-placement of reinforcement in a freeway overpass bridge.
Theoretically, it should be possible for students to prepare case studies on the basis of what has been published about building projects in books and journals. Experience shows that this is rarely satisfactory. Usually it is found that the available material in engineering journals is entirely technical, while that in architecture journals is entirely subjective.
A valuable technique is to have students act out the various roles involved in the development of a building project. Sessions of this nature have been run at Monash University for some years (Holgate 1976, 1987c), but there is still much room for development of the technique. It has proved very good for motivation, for giving students a picture of the wide range of factors involved in planning and design, and for teaching them about the interaction of the various participants with a degree of reality that would be impossible in a lecture situation.
In Melbourne, the Association of Consulting Engineers Australia and the Royal Australian Institute of Architects mount regular role-playing demonstration sessions for tertiary institutions. In this case the participants are practising professionals (developers, accountants, architects, and engineers) and they simulate the progress of a building project by loosely following a prepared script. This ensures that certain chosen characteristics of the process are demonstrated.
The experience with student role-play has highlighted the need for engineering students to be schooled in simple 'rules of thumb' for the initial sizing of members, and to be given access to charts for preliminary design. Before this was realised, the other players were obliged to wait for up to twenty minutes while the 'structural engineer' worked out the depth of a single beam. It can be very difficult to convince students that the sizing methods which loom so large in their 'design' courses are not suitable for such occasions. It is equally important for academics to realise that these rules of thumb need not follow studies in routine code-based procedures for analysis and sizing, logical though that might appear (see Warner 1989).
6.2 Pedagogic considerations.
This is not the place to go into detailed considerations of teaching technique, but some general comments are appropriate. Teachers of general knowledge of structures and design would need to take note of Armstrong's warning (1983) against encouraging the development of limited physical or mental 'libraries' of details and forms.
Finniston's emphasis on teaching engineering practice by the study of 'engineering applications' is a good one - provided the general principles within each application are brought out and understood. Structural engineering can be seen as a vital, developing, self-fulfilling profession, constantly presenting new challenges to its practitioners, and opportunities for greater service to the community, or it can become a rather dull, stale craft, applying and repeating details of analysis and selection without any real creative impulse. [Armstrong (1983) p.23.]
There is general agreement that design is best learned through practice under the supervision of someone able to give continuous advice and criticism, or at least through graded exercises which develop to simulate real-world design as closely as possible. In this it is comparable to the practical arts such as music, painting and sculpture. To be a good practitioner requires a certain amount of native talent and a strong motivation, but there is evidence that any degree of innate talent may be exercised, refined, and enhanced by the experience of formal training, especially in the form of studio work. Conversely, it may be suppressed in a learning environment which is heavily structured and encourages the passive absorption of information; where analysis is valued more highly than synthesis; and where an element of fun is considered inappropriate to the serious business of engineering.
Greater emphasis on learning and discussing the qualitative aspects of engineering will present major challenges in student assessment and grading. Engineer-academics will need to learn how to set and mark papers which require discursive answers. The present cohort of engineering students is not accustomed to reading large amounts of descriptive material or to making notes on it during lectures. I have tried to overcome this by handing out concise lecture summaries, but have found them memorised and regurgitated almost word-perfect in examination scripts. In 'open book' examinations for my post-graduate course on the relationship between structural engineering and architecture, I have encountered scripts completed by practising engineers which consisted entirely of paragraphs copied down word-perfect from texts and lecture summaries. The people concerned claimed they were incapable of expressing the material more cogently than was done in the source. Answers consisting of plagiarised paragraphs strung together are offered even to questions which ask for a personal assessment of a topic. Many students evidently feel that every question must have a single, discoverable true answer, and that they have no hope of competing with the experts in the field.
Traditional written examinations are a reliable measure only if the ground they will cover has been defined reasonably clearly at the start of the course. This is contrary to the spirit of open-ended exploration which is most suitable to the learning of descriptive material and the encouragement of design ability and creativity. Viva voce examinations offer an alternative, but have major disadvantages. They could perhaps be used in conjunction with written examinations.
In the sciences and engineering it is often claimed that subjects which involve a quantitative approach and a certain level of mathematics (or even arithmetic) are more 'rigorous' and 'intellectually demanding' than descriptive subjects. The latter are sometimes described as 'soft options' which may be picked up at any time simply by reading a good book. People in the other camp (mainly the humanities) may retort that technology demands too little reflective effort to be allowed on a university campus. The issue is highly complex. Minds are largely made up and debate seems futile. My own opinion is that any subject which is done well and pushed to its limits by intelligent and creative people is capable of offering the highest intellectual challenge. The amassing of a personal store of facts is not intellectually demanding when these have already been researched and categorised by others, but this must be a precursor to the application and analysis of that store of knowledge and its subsequent expansion and reorganisation. This process involves perception of links between facts, the development of extrapolations, and the drawing of hypotheses and conclusions. Studies of creativity suggest that at the forefront of all types of intellectual endeavour the modes of thinking required are very similar. At the other extreme, mediocre work is too often accepted for publication. It may be that, because generalist and descriptive publications are more accessible than scientific and technical ones, it is easier for non-specialists to identify mediocrity.
6.3 Administrative measures.
Commercial pressures and the desire to 'get on with the next job' will presumably prevent practising engineers from making a major contribution to research and scholarship in the discipline. The task will therefore be left mainly to academics. It would be ideal if administrative action within tertiary institutions could bring about greater academic recognition for this type of work. However, similar attempts over many decades to give teaching equal status with research have proved fruitless. Possibly some change will occur naturally now that much arithmetical analysis and even member sizing is done by computer. This has led to a swing of interest towards artificial intelligence and expert systems. As a result many scientists and technologists are beginning to realise that the business of acquiring, encoding and manipulating descriptive knowledge is far more complex and intellectually demanding than they had imagined.
A further administrative challenge is that of fitting the additional material into the tertiary course. One answer would be to provide less of the technical and analytical material that presently dominates the course. The other would be to allow greater choice and specialisation, so that some students may continue with the present syllabus while others develop their knowledge of structures to an extent that suits their interests and personal qualities. The majority of graduates could easily survive in practice with less mathematics than is taught in traditional university courses. Although there is a need for a core of engineers who have specialist knowledge, many more are required who have a reasonable level of competence in technical areas, but are highly creative and adaptable, able to relate well to clients and architects, and who know when to call in specialists. This should be recognised and catered for by flexibility in course structure and accreditation. It may be time to abandon the concept of the tertiary degree as a three- or four-year standard course leading to a 'ticket' which states that the holder has studied more or less the same subjects as his or her colleagues. A document which stated simply what subjects the graduate had passed over any period of years might be of more value to employers. It would allow them to assess suitability for particular positions and, given a more flexible syllabus, could allow them to state their own preferred prerequisites for their particular type of business. The growth in continuing education would provide for individuals and firms who wished to develop their skills in new directions to suit changing job opportunities and conditions in industry.
An alternative measure would be to move towards degree courses in Structural Engineering in which students would receive less coverage of water and transport subjects than they do in current Civil Engineering courses. They could, of course receive more coverage of topics relating to architecture, aesthetics, and building construction. This would bring the material closer to that of some degrees in Building.
Whatever course were adopted, it would be necessary for administrators and politicians to recognise that teaching techniques which encourage initiative and creativity are very demanding of staff time and mental energy. The studio method of teaching design is very labour-intensive. With class numbers of 80, and an overall staff/student ratio approaching 1 to 15, there are good reasons why many staff cling to the old, less demanding methods involving one-way transmission of lecture notes by means of chalk, blackboard, and the occasional colour slide. Efforts to raise the status of teaching in academe should be continued with renewed vigour, but radical initiatives will be required if anything is to be achieved. Under present conditions, teaching is a charity. Academics invest time and effort in it to the detriment of their careers.
It would be equally important to change attitudes towards the intellectual value and difficulty of descriptive work. Many departments contain a member of staff who has occupied senior positions and has a wealth of the sort of knowledge and experience under discussion here which is wasted because of the low status accorded to qualitative knowledge in the culture of university engineering faculties.
Perhaps the only way in which these cultural attitudes could be changed would be through administrative action to alter the type of person admitted as academics and as students. As Jay (1984) has said
"… the best endeavours of educators … will not alter by very much the degree of excellence of the engineer of the future … unless we take a fresh look at the type of person recruited to the profession."
It should be possible to attract a student with a wider range of interests than the traditional Mathematics, Physics, and Chemistry by introducing into the course the sort of material suggested above, which puts the technical facts and computational procedures in context within the whole discipline of structural design. A promising development is the 'combined degree' which allows simultaneous study in two fields such as engineering and economics, engineering and law, or engineering and science. At Monash University, engineering may be combined with an Arts degree in language (including a full 'major', with background subjects of a cultural nature). It may also be combined with a major in Visual Arts, with an emphasis on the appreciation of architecture.
An interesting suggestion made by Armitage is that everyone has a store of useful skills and knowledge acquired intuitively since childhood, and that as many of these are relevant to function as an engineering designer, engineering faculties should test for them as part of normal admission procedures.
The failure of engineering faculties in developed countries to attract sufficient numbers of native postgraduate students is usually attributed to the poor financial returns. While this is a major factor, it is also true that many graduates and engineers consider conventional doctoral studies to be irrelevant and uninteresting. It might be possible to recruit some of these to carry out research work into the design process both in the abstract and in terms of recording the work of individuals and firms. This would be scholarship more like that found in the faculties of Arts and Architecture. If there were strong objection from those who maintain that mathematics is necessary for intellectual rigour, a different name could be found for the resulting qualification. A shorter period than the usual three years of the PhD, or provision for part-time research, might attract people who are already working in industry, with beneficial results for all concerned. The general environment in which conventional PhD students work might be much improved.
7. Conclusion.
In writing this paper I have consciously adopted an idealistic approach because I feel this can be a useful preliminary to lowering one's sights and deciding what is practicable. My main aim has been to initiate a discussion on the matters outlined above, and I would be pleased to hear from anyone who has a point of view to express, either privately, or in the form of a contribution to this journal.
Note, 2003. Looking back, this does seem very utopian. As more topics crowd into the civil engineering course the time available for teaching structures has diminished, while leading-edge structural engineering has become increasingly complex. There would be a case in a traditional university environment for the creation of specialist Departments of Structural Engineering in selected universities. However, the mindset known in Australia as "economic rationalism" makes it unlikely that this would occur in the under-funded public system. The answer in this environment might be a privately-run "staff college" for middle-level structural designers, perhaps organised and financially backed by employers.
Combined References for this part of the website.
Footnotes added for thesis presentation, 1995.
[The art in structural design.] [Aesthetics of built form.] [Work of Jörg Schlaich.] [John Monash's early engineering.] [My personal page.]