Wavelet Homepage Table of Contents
Our
ability to find a solution to the signal analysis puzzle hinges on finding a
transient event on the CVR caused by a mechanical flaw.
In order to determine how the signals to the Beech 1900c differ from that
of a plane with no mechanical flaws, we need to be able to compare the Beech
1900c recording to a control case. Dr.
Stearman has provided us with such a case with the CVR of an airplane crash in
which there were no mechanical flaws. This
is a recording in which the airplane crashed due to pilot error, and not a
structural problem with the plane. For
this reason we cannot list the type of plane, nor describe the events leading up
to the crash.
Despite
the fact that we cannot discuss the events on the recording, we can note that
there was only one point in the silent track that the signal deviated outside of
the normal range. At this point there was one high frequency spike event
followed roughly one second later by a mid frequency double “bump” as
pictured in Figure 42 along with the wavelet decomposition.
Figure 42: Single Transient Event on Nominal Tape
In and of itself, this single event tells us little about what to look for on the Beech 1990c recording. However, the fact that a “healthy” plane had this one transient event certainly makes one suspect that transient events themselves are not the norm for a recording such as this. What is significant is that the single transient resembles some of the jerks seen in the spring 1998 Q testing, except that the second “bump” in this case is out of phase. This did not happen in the earlier testing.