Science Buddies: "Ask an Expert"

My question refers to the response of wood columns carrying gravity load when subjected to lateral drift due to earthquake or wind loads. For the experiment, I am turning the column (i.e., beam-column) on its side and clamping one end in a vice making a cantilevered beam-column. I load the beam-column with a gradually increasing point load, V, to simulate a lateral deflection due to the earthquake or wind load on the building. I will do several experiments with different test specimens. The first experiment has no other loads (just V). For subsequent experiments, I will apply a constant compressive load P to the beam column to simulate a gravity load. (I anticipate that I will use a spring device to apply this constant compressive load.) Each experiment has a different compressive load P. I will measure the beam-column lateral deflection as the lateral force V gradually increases. The purpose of these experiments is to see the influence of the gravity load P on the lateral response of the beam-column. My teacher wanted me to ask an expert to see if the point load V is a good way to simulate lateral drift in a column due to a wind or earthquake load. Thanks…

ThomasD wrote:My teacher wanted me to ask an expert to see if the point load V is a good way to simulate lateral drift in a column due to a wind or earthquake load.

IMO: Static simulation/analysis methods are poor indications of dynamic load behaviors. Modeling a single post is also a poor simulation/analysis of dynamic load behaviors.

Take for example the 1989 Loma Prieta earthquake where the Cypress Street Viaduct on the Nimitz Freeway Interstate 880 double deck collapsed. A single column lateral drift model would not have predicted that collapse. In fact, the engineering studies by the engineers who designed that structure used more sophisticated models that did not predict a structural weakness.

That earthquake was a case where the 10 to 15 second surface-wave magnitude 7.1 caused standing waves in the structure which amplified the bending forces and micro fracture cracked the concrete away from the reinforcing rods pancaking the structure at the in phase wave point or points and then the pancake impact wave went through the structure unloading the balancing forces which fractured the adjacent columns. A major defect in the modeling was not taking into account any pre-existing stress fractures from dynamic traffic loads or heating/cooling cycles.

The data you get from any static test is still valuable information and is needed information to create a good dynamic model, it just isn't all the data needed to build a good model and it certainly isn't sufficient to predict all dynamic load behaviors by itself.

-Craig