Dr. Franke’s research group has been focused on three major research projects this year:
- Developing simplified analysis methods to closely approximate the results from high-end probabilistic liquefaction (i.e., triggering and post-liquefaction settlement) and lateral ground deformation (i.e., lateral spread and seismic slope displacement) procedures
- Working with researchers from Oregon State University to develop probabilistic liquefaction triggering and lateral spread displacement maps for Utah County, UT
- Working with optimization and computer vision research experts to develop automated flight path planning and anomaly detection algorithms for monitoring large/long infrastructure features (e.g., roads, pipelines, transmission lines, levees) using unmanned aerial vehicles (UAVs)
The efforts put forth by Dr. Franke’s research assistants have produced significant results. For example, his students Braden Error, Lucy Astorga, Kristin Ulmer (former student), and Levi Ekstrom (former student) have created a simple-to-use Excel spreadsheet that incorporates all of the simplified analytical liquefaction methods that they have developed, as well as mapped liquefaction reference parameters for all six U.S. states that have supported their research. All that is required from a user are the latitude/longitude coordinates of his/her site and the typical soil subsurface information that is traditionally required for liquefaction assessment. Dr. Franke is now being invited across the country to share these findings and the simplified spreadsheet with various geotechnical engineering groups and chapters, and he hopes that these simplified methods will eventually become part of seismic geotechnical design codes.
Scatterplot comparison between a newly-developed simplified method for computing probabilistic lateral spread displacements and the full probabilistic lateral spread method (after Ekstrom and Franke 2016). The simplified method provides a near-perfect approximation of the full probabilistic method across a wide range of seismic loading levels, hazard levels, and representative soil/topographic conditions.
As another example, Dr. Franke’s graduate student Derek Wolfe just completed some valuable research involving the integration a non-GPS light distancing and ranging (LiDAR) sensor onto a small UAV platform. This sensor can utilize a relatively new remote sensing technique called laser odometry and mapping (LOAM) to rapidly develop 3-dimensional point reconstructions of the built environment, all without the need for GPS location of the UAV platform or the LiDAR sensor! To test the sensor, Derek received authorization to perform a single UAV flight inside of LaVell Edward’s Stadium to see how effectively the LiDAR sensor mapped the structure. The 3-minute UAV flight consisted of a single lap around the football field at an elevation of about 30 feet above the ground. The resulting 3D point cloud was developed after approximately 5 minutes of processing by a third-party LOAM provider. Based on subsequent total station surveying performing by Derek and other researchers, the point cloud model showed both horizontal and vertical accuracies of 3 cm or less. This type of remote sensing technology applied in the real world could revolutionize the way that many parts of infrastructure are inspected and monitored.
3D intensity-based point cloud of the LaVell Edwards Stadium developed from small LiDAR sensor mounted onto a UAV and a relatively new remote sensing methodology called LOAM. Notice how the light intensity was able to detect even the tire tracks in the grass on the field!