Why are structural grads uninterested in Risk and Reliability?

Some of the materials were first drafted in my LinedIn early on.  But the character limit of LinkedIn makes it only a place to show opinion and not a full argument.  Here is my relatively complete argument:

CV8311 Risk and Reliability for Engineers Fall 2016 was completed. The course project presentations were terrific. The presentation titles give you a glimpse of the coverage.

  1.  Generalized linear modeling in road safety
  2. Probabilistic life cycle costing for P3s
  3. Post-event risk analysis for Walkerton Disaster
  4. Project contingency management
  5. Stochastic deterioration modeling of water mains
  6. Resilience of downtown transportation network under major sport events
  7. Pile resistance model and LRFD
  8. Dam safety
  9. Vulnerability of electric transmission network under ice storms
  10. Criticality of transportation network
  11. Reliability of soil nail walls

A former colleague of mine commented that “nice to see that transportation is well represented,” and I returned, “Sadly structures are so under represented!” This triggered me this blog article.

Although Structures Researchers contributed one of the two reliability methods, called structural reliability methods, in reliability engineering, very sadly structural students at my school are not interested in taking this subject as one study of their urgent need.  This happened not only in this graduate course, but also in CVL609 Civil Engineering System where I usually spend at least three hours to discuss the structural reliability method.  They were not excited.  This is ironic, for safety is such a big concern for the profession.  Although, as the Professional Engineers Ontario states, public safety is the paramount issue of professional engineers in general,  safety is more important an issue in structural engineering than in almost any other engineering disciplines.  Thus, any serious structural engineers should take this question into a serious consideration. When I was an undergraduate students, my professor asked the class this question: given that there is no absolute safety, how safe is safe enough in design? It was that moment when I started to realize that research could be something for my career. 

The reason our students aren’t interested is, on my speculation, was that they are deeply duped by codified design. Many of them are paralyzed by the misconception that as long as they follow the code they are safe. They don’t bother to challenge the codes; they don’t bother to be challenged by real engineering problems where codes don’t apply.  Our students rarely asked why the load and resistance factors are taken value as they are specified in the design code.  They rarely think why those factors are different between the Building Code and Highway Bridge Design Code, even though they are designed the same concrete girders, one for an industrial plant and the other for a highway bridge.

Many structural engineers take the design factors for granted.  The don’t understand, or don’t bother to understand why the factors are taken as they are.  When situation changes, they can easily make mistake. One of my colleagues once was puzzled when supervising his graduate student how a resistance reduction factor can be greater than one!  He rushed to conclude that the student must have made some computational mistakes.  After several rounds of checks with both simulation and iterative methods, the poor student came to me for a ‘judgment’.

Many believe that the best structural engineers are those who know all the very details of the design codes and standards — Yes, a lot of our students are not even able to understand the clauses and apply them in practice.  They don’t understand that actually the best structural engineers are those who can challenge the codes and standards.  Structural engineering research, in essence, is to keep pushing the frontier.  In this regard, design codes and standards are indeed the real battle fields for structural engineering researchers.  Because of this, modern design codes and standards, unlike the Hammurabi Code, are often the result of compromises of different schools of thoughts.  I used to think that the design codes and standards are often a good starting point of research in structural engineering.  I should now adjust my thought to add that the design codes and standards must be the starting and ending point of any good research in structural engineering.

Broadly speaking, there are only three key battle fields in structural engineering: (1) new materials, components, and systems, (2) models and modeling, and (3) risk and reliability. While the first two are explicit and noisy, the risk and reliability battle field is often implicit and silent.

A few examples may help understand what I just said.  In seismic design of frame structures, a basic design philosophy is “strong columns, weak beams.”  The rationale of it is to ensure plastic hinges occur in beams before in columns.  If plastic hinges occur in columns at first, a storey-failure mechanism will form and the whole structure will easily collapse.  In addition, it is commonly agreed that beam hinges absorb more hysteric energy than column hinges do, because the former is subject to smaller or even negligible axial forces.   To achieve “strong columns, weak beams”, design standards often introduce an artificial moment magnification factor for columns.  Recall that in structural analysis, the bending moments at the ends of the column and beam in one joint must be in equilibrium.  The moment magnification factor creates an artificial equilibrium in the strength of the two members to ensure, in deterministic view, that plastic hinges will not form in columns before beam hinges are developed.  The design standards suggest a value for such a magnification factor.  Depending upon the standards, structural materials, and design earthquakes, this factor ranges from 1.1 to 1.3.  But question is, to what extent can this factor avoid a storey-failure mechanism?  Is it enough? More importantly, how do we develop our argument?  Similar issues apply to many other structural design philosophies such as ‘strong shear, weak bending’ and ‘strong shear wall, weak frame.’ 

Another situation where we have to go back the first principle of structural reliability is expansion, rehabilitation and retrofitting of existing structures that may or may not have been damaged by uses and external events.  Existing structures were designed as per outdated design codes and standards.  But expansion, rehabilitation and retrofitting need to satisfy the new codes and standards.  The incompatibility of reliability of old and new standards may pose additional challenges that can be solved with a solid understanding of the embedded reliability in the standards.

Risk and Reliability deals with the philosophy of engineering.  This is no simple subject. It requires not only a solid understanding in the subject matter — being it structures, transportation, environmental, geotechnical, construction and infrastructure management — but also probability, statistics, and optimization, which many engineers are weak.  It roots back to the very first principle of engineering: the tradeoff of safety and economy.  Of course, beyond economy, we can also add additional criteria such as sustainability and equity.  The drills of the first principle can really sharpen your mind and provide a great confidence in your future consulting services.  It is just like the study of elasticity.  Although rarely is this subject itself applied to solve real structural problems, without knowing it, you often do not feel comfortable in your finite element analyses — Or you do not even know you did not understand the results, which is even worse!

What really worries me is that if our students are so happy with knowing only how to follow codes and standards without knowing why the codes and standards were so developed, there is no ingenuity and innovation.  One colleague of mine told me that the reason he did not choose structures as his specialization when he was an undergrad was he dislike the codified design in structures.  I never had this impression when I studied my undergrad back China.  I thought this over, and compared with a couple of courses I have taken in Canada.  I tried to take a concrete design course, and a steel design course, both at graduate levels.  I could not stay there, because the instructors kept teaching how to apply the clauses while providing little explanation why the clauses were so specified.  You may argue the instructors may not the best ones; but I can ensure you they are the best ones in Canada.  I got a sense that the students in Canada were trained to follow the codes and standards while in China, at least in my university, we were taught to know how, why, and why not.  This is another irony because China in many outsiders’ view is a communist place where people are supposed to do whatever are told to do, and people in western societies are supposed to have more liberty in doing things. 

There are three levels of knowledge in engineering education: how, why, and why not. One thing I honestly dislike is the emphasis of the so-called ‘hands-on experience’. In my view, if your mind are not on, no matter how many hands are there, you can hardly build up your insights.  Innovations result from the curiosity of ‘why-not’.  For this, put your minds on!

This entry was posted in Risk & Reliability. Bookmark the permalink.