Rendering of Experiment Setup

On Wednesday, 12/11/08, I attended a dynamic test of a 110-foot long bridge model at the University of Nevada Reno, Large-Scale Structures Laboratory.  This test is part of series of tests to investigate the performance of various technologies to improve bridge performance in earthquakes. The principal investigator is my former advisor, Dr. Saiid Saiidi, but the work is part of a group of researchers that span across the country: University of California San Diego, University of California at Berkeley, Florida International University. The testing is part of an overall network of research: The George E Brown, Jr. Network for Earthquake Engineering Simulation (NEES) that is sponsored by the National Science Foundation.

The model bridge is about ¼ scale of an actual structure with realistically modeled joints and abutments. The test model has three two-column “bents“ (piers to the non-bridge engineers)  each on its own hydraulically-activated shake table. Each shake table is capable of mimicking earthquake accelerations in two directions.

Each bent in the model is designed to test a different technology – one was set up to test a base isolator design – essentially a special rubber and metal doughnut – to take energy from the earthquake motion and dissipate it as material deformation in the isolator. This method is similar to the way a car’s shock absorber works. Another bent was designed to review the performance of high-strength reinforcement – nickel-titanium alloy. The third bent was designed to test post tensioning (pre-compressing beyond regular structural weight) of the columns.

The test starts rather slowly – a white noise of motion is first fed to the shake tables to calibrate the sensors on the model. This model had already had some tests performed before we were invited to see the “big” test.  The big test caused the model to sway dramatically and some concrete cover on the isolator bent crumbled. Unfortunately for major drama fans, none of the bents failed in a heap. This was fortunate for the audience because we were not all that far away from the model. It really wasn’t dangerous – there are backup safety frames to prevent a collapse.

This was of particular interest to me as I was the BJG lead engineer for designing the lab structure itself. The “strong floor” is three feet of high-strength concrete reinforced with #14 bars at 12” on center top and bottom. There is a basement below the floor for equipment access. The floor has a 2’ x 2’ grid of pipes to allow test fixtures and experiments to be attached to the floor using high-strength threaded rods. This design has proven to be very useful as it is the basis for a similar lab in Korea and for Simpson Strong Tie’s test facility in Stockton, California.

I watched the test standing next to a professor of electrical engineering who was asking questions about the test fixture and methods. It’s been some time since I have been involved in anything like dynamic testing or evaluation but most of my old Master’s work came back through the haze of time to answer his questions. Or at least it sounded good… I did note that the equations for structural stress, strain and motion are of the same form as equations for voltage, amperage and induction. (and that, Ladies and Gentlemen, is all I remember of differential equations) Indeed, the professor stated he could model the experiment on the floor as a circuit of capacitors, inductors and resistors.

The test accelerations are based on Loma Prieta and Slymar Earthquake ground motions. Everyone always wants to know what magnitude earthquake a test represents and the answer is  – none. Really the answer is unknown – Richter magnitude is measure of energy released in an earthquake event. Structures are affected by the resulting accelerations or ground motions – the relationship between the two is highly variable and depends on hypocenter location (the actual break in the ground at depth) and the material types that the energy must pass through. I noted as I watched the local evening news reported the acceleration as equivalent to a magnitude 8.

As this was reported on TV, it must be true.

See this link http://nees.ce.unr.edu/telepresence/ to get more information on the test program and tune into webcasts of the testing.

UPDATE

See this link to a story on NPR regarding the experiment.