Abstract
The citizen science app, MyShake, represents a new source of global
seismic data that can be provided with much higher density sampling than
a traditional seismic network, and at a relatively low cost. Because
phones are most frequently located in buildings, the signals they record
include structural response information in addition to ground shaking.
This work explores the capabilities of MyShake as a remote damage
detection tool within the field of structural health monitoring. There
are several common, empirically determined parameters, such as modal
frequency and inter-story drift, which characterize buildings and their
response to shaking. Structural damage causes these values to deviate
from their ‘healthy’ baseline. Using data collected during smaller
excitations, we aim to generate a database of healthy values for
buildings in which MyShake users live and work. Then, when a significant
earthquake occurs, data from the mainshock and aftershock coda seismic
records can be analyzed to detect deviation from this baseline,
permitting a rapid, remote diagnosis of damage that could provide
critical information to engineers on the ground. To gauge the expected
capabilities of MyShake while the system is still relatively new, we
compile a database of waveforms collected by the Center for Engineering
Strong Motion Data (CESMD), which has instrumented buildings across
California. Available CESMD data spans several years, building types,
and earthquake magnitudes. After ‘downgrading’ these data to emulate the
quality of waveforms collected by MyShake, we determine how big and how
close an earthquake needs to be to excite buildings so that key
structural health parameters can be determined. These data also inform
our expectations for the range of values that typify each building type
and size, permitting a gauge for the reasonableness of our results from
true phone data. Demonstrative examples from this CESMD-based database
in addition to real MyShake examples will be presented.