Pyroshock data analysis—Analysis, editing, and validation

Abstract

The most common form of pyroshock data analysis is the shock response spectrum (SRS), which is defined and briefly discussed. The most common way to compute the SRS of a time history is with a digital recursive filter that simulates the response of a single-degree-of-freedom (SDOF) system to the pyroshock excitation. This filter is described together with its strengths and limitations. Sometimes additional characteristics of the time history are used to supplement the SRS. The most common is a definition of duration. If the time history is assumed to be a sum of exponentially decaying sinusoids, a decay rate can be used. Often, the time duration is defined as the duration until the waveform decays to 10% of its peak value. More recently temporal moments have been used to define the temporal characteristics of the waveforms. Flaws in pyroshock data are common, including accelerometer zero shifts, clipped data, dropouts, and instrumentation overloads. Data validation procedures that expose many of these flaws are outlined. Good pyroshock data are difficult to acquire. As a general rule, flawed data should not be used, but sometimes we are forced to use flawed data. Procedures are briefly described for editing flawed data that will result in credible, although not necessarily correct, results.