Earthquake Computer Simulation Provides More Detailed Response Data than Laboratory Test Results at a Lower Cost

Earthquake destruction can cause massive death and expensive structural damage, which most often leads to business interruption, loss of housing, insurance instability and loss of jobs. When designing a building for a location prone to seismic activity, the potential for earthquakes guides much of what civil engineers need to consider in a design. Engineers have often relied on laboratory tests performed with scale models to try to predict how building designs will hold up in an earthquake and guide the optimization of the design. Since laboratory tests are time-consuming and expensive, an international team of civil engineers conducted a project that compared laboratory results to results from ALGOR’s Mechanical Event Simulation (MES) software and discovered that MES may not only reduce the need for multiple iterations of laboratory tests, but also provided more detailed response information than traditional laboratory sensors.

An international team of civil engineers compared results from ALGOR’s Mechanical Event Simulation (MES) software (upper left and lower right) with the results of a laboratory seismic table test (lower left). Engineers superimposed a Fast Fourier Transform (FFT) graph from ALGOR’s Monitor utility over a graph produced by laboratory testing equipment (upper right). The team discovered that MES not only reduces the need for multiple iterations of laboratory tests, but also provides more detailed information than laboratory sensors with less “noise.”

Modeling the Seismic Table Test

To prove the usefulness of MES, the civil engineering team simulated the effects of an earthquake on a water tower on both a scale model in the laboratory and with an MES model. To facilitate comparison of laboratory and computer simulation results, the MES needed to exactly replicate the laboratory experiment. Therefore, the 3-D MES model contained both the water tower and the seismic table.

The team’s computer-aided engineering analyst, Patricia Belles of Universidad Nacional del Sur in Buenos Aires, Argentina, began by building a 3-D MES model of a water tower in ALGOR’s Superdraw III, a precision finite element model building tool. She was assisted in her work by Pablo Vicente Legazpi of CAEsoft Consulting in Madrid, Spain.

In the MES model, the water tower structure consisted of beam, plate/shell and kinematic elements. The beam and plate/shell elements represented the structure of the water tower, while the kinematic elements represented the reservoir of water. The seismic table was represented by kinematic, general contact  and actuator elements.

“Kinematic elements behave dynamically like regular solid elements and can transmit forces; however, stresses are not calculated for these elements so processing times were greatly reduced,” explained Legazpi. “Since stresses in the water reservoir and seismic table were not of engineering concern, we used kinematic elements to shorten the length of time needed to process the simulation. This was important because the event would take 4 seconds and we needed to output results for every 1,000th of a second.”

The general contact elements represented the static compensators that stabilize the seismic table. The motion of the event was generated by the actuator elements, which represented the hydraulic cylinders present in the seismic table at the laboratory. Actuator elements enable engineers to realistically simulate complex computer-controlled movement over time.  In using these actuator elements, engineers can specify contraction, extension and rotation values over time to drive motion between the connected parts.

The MES analysis provided displacement, stress, acceleration and velocity output. The engineers used ALGOR’s Monitor utility to produce graphs of these results over time and to produce a Fast Fourier Transform (FFT) graph, which would subsequently be compared to the laboratory results.

In the Laboratory

The seismic table test was then conducted in the laboratory of CEDEX, Spain’s official Civil Engineering testing laboratory, where Manuel Pastor, leader of the Numerical Analysis Department, and Francisco Navarro, a member of Spain’s National Seismic Commission lent their expertise. Accelerometers and strain gauges recorded the laboratory results throughout the event.

“We were looking for the laboratory and simulation results to correlate closely,” said Legazpi. “Not only was that goal achieved, but, more importantly, we found that MES actually helped us to understand what we observed in the laboratory tests, because the computer simulation could capture results at smaller time increments than was possible in the laboratory. While laboratory tests will undoubtedly continue to be used as the final method for design certification, we concluded that Mechanical Event Simulation results can be more detailed, and therefore, are invaluable during the design phase.”

Optimizing a structure in the design phase is beneficial because of the expense and time involved in performing laboratory tests. The seismic table used in this experiment costs around $2 million U.S. dollars and there are other costs associated with testing, such as building the scaled structure and dedicating at least one engineer to setting up the instrumentation, which often takes several days. By using MES to optimize a design, engineers can reduce laboratory testing down to a single certification procedure.

Looking to the Future

Although the results of this experiment prove the effectiveness of MES for civil applications, Legazpi is most encouraged by the potential MES holds for simulating more complex events. “The future of MES for civil applications lies in the fact that MES can go beyond simulating an earthquake with a spectrum,” said Legazpi. “Once validity has been obtained with scale models, MES can model and simulate the true, full-sized structure. In addition, MES can incorporate real-world factors like contact with other structures, damping and soil-structure interaction, all of which cannot be tested in the laboratory.” The capability to perform more realistic simulations will enable civil engineers to design safer structures for use in earthquake-prone sites.