Straining at a Gnat

Near the beginning of my undergraduate soil mechanics class, the instructor taught us the mnemonic device "Structural Engineers are Weird Guys" to help us remember the equation S*e = w*Gs.  It was hilarious and it will help me to remember that equation until my dying day.  (It was also ironic, considering that this guy was like the alpha nerd, but that's a story for another day.)  In my experience with structural engineers, there is a very bimodal distribution of capability in engineering among them.  They're all either really bright and capable engineers, or quite engineering challenged.  There are almost no average engineers (so to speak) in the mix.

Yesterday, I felt like I got to experience both ends of this distribution simultaneously.  In my structural dynamics class, the professor showed us the computer model of a 12-story reinforced concrete building in Berkeley, California that he had been working on for several months.  He had gone to great ends to carefully model all of the columns, slabs, and shear walls in this building.  He built a 3-dimensional computer model that had thousands, if not tens of thousands, of lines of code to carefully track all of the components of the response of this fairly complex building to an earthquake that actually happened in the area in 1989.  On an ordinary personal computer, the analysis of the response of this structure to maybe 20 seconds of earthquake motion took the better part of a month to complete.  The instructor of my class then wrote probably another couple thousand lines of code to make a graphical presentation of the numbers he was coming up with.  This presentation was slick, too.  He had a model of the skeletal structure of the building in both 2D and 3D moving according to the model outputs along with graphs of the accelerations, displacement, forces, and moments in the building all varying simultaneously as the model time progressed.  I've included a low-quality (i.e. one that won't have the sponsors of this research seeking out my name to put on the cease and desist order) picture of this graphical representation.  It took not only a good bit of intelligent thought, but quite a bit of dedication to make this high-quality of model.

Crazy-complicated model with rigid base.

So what's wrong with it that I say it's both smart and dumb at the same time?  Like many other analyses that structural engineers perform, the instructor of my class used a rigid base on this model.  Basically, he assumed that the soil on which the structure sits is much stiffer than the materials used in the building.  Indeed, he assumed that something commonly referred to as "Bay Mud" is significantly stiffer than concrete and steel.  I'd think this next part would be pretty intuitive, but I'll spell it out anyway: that's a terrible assumption.

Yes, I am a geotechnical engineer, and, like most geotechnical engineers, I think that the soil on which a structure sits deserves much more attention than it often gets.  And yes, soil-structure interaction is the topic of my research and I am a little biased in that sense, but hear me out.

I know that the effects of the soil that supports a structure can often be neglected for simple structures that are subjected to simple loading.  Under these conditions, the effects of soil-structure interaction don't warrant an in-depth geotechnical analysis for the design of the structural members.  (There should still be some investigation for shear strength and settlement, but not necessarily for structural design.)  However, a 12-story reinforced concrete building is not a simple structure, neither is earthquake loading a simple load case.  I really can't justify, in my mind, spending months to build an enormously complicated model that doesn't reflect reality as much as is reasonable because I've omitted a key factor that clearly affects the reality of the situation.  If I'm going to go to all the trouble of building a crazy-complicated structural model, I would at least try to model some flexibility in the soil.  It would make the analysis much more reflective of reality and would probably only add a few hours to an analysis that's already going to take several weeks to run anyway. 

I respect structural engineers.  I think that the work they do in practice and in academia is just as important as the work that other engineers, including geotechnical engineers, to society.  However, they, like other engineers, often make very complicated, precise analyses for structural members while, at the same time, minimizing or totally ignoring important constraints and boundary conditions.  I think it's important for all engineers to take a step back occasionally and look at the entire scope of the system they're designing instead of focusing so intently on the one specialized concept to which they are assigned to attend.  This way, we, as engineers, can avoid straining (pun intended) at a gnat and swallowing a camel in our designs.