Testing Explosion Source Models From Yield and Depth Analysis of Chemical Explosions Conducted in Alluvium

Date: 4/24/2019

Time: 03:00 PM

Room: Cascade I

We have developed the envelope yield method (Pasyanos et al., 2012) to estimate explosion source parameters for a variety of sources (chemical and nuclear), yields (sub-ton to ktons), depths (deeply buried to above surface) and distances (local and regional). Using the method, observed coda levels are tied to direct phase amplitudes. When corrected for propagation, these amplitudes have a physical basis to yield and depth through explosion source models such as Mueller-Murphy, Denny-Johnson, and Walter-Ford. While these models have been developed and validated using hard rock, the models are less mature for explosions in alluvium, which show important differences in the low-frequency moment rate level as well as in the high-frequency roll-off above the corner frequency. Changes in the material properties also affect the far-field displacement. Variations among the various source models and properly accounting for material properties have important implications for yield determination and source identification of small events and in uncalibrated media. In Phase II of the Source Physics Experiment (SPE), chemical explosions are now being conducted in Dry Alluvium Geology (DAG). We use local and regional data for DAG-1 (908 kg equivalent TNT @ 385 m depth) and DAG-2 (50997 kg TNT @ 300 m depth), and legacy U.S. nuclear tests conducted in alluvium (e.g. Riola) for which we have ground truth (announced yields and depths from DOE NV-209). We use the recordings to test several variations of the material model of alluvium for existing source models to best match the known GT.

Presenting Author: Michael E. Pasyanos

Authors