Imagine you're tuning a piano. You don't need to strike each key with a mathematically perfect force — you need to hear clearly what comes back. If your ears are sharp and the room is quiet, an imperfect strike tells you everything about the string's behavior. That's the core argument of TB 26-09: the shape of the hammer hit matters far less than the quality of the recording. This technical bulletin from NASA Langley, authored by Joel W. Sills and colleagues, addresses a surprisingly persistent orthodoxy in structural modal testing: that a hammer tap must produce a clean half-sine pulse to be valid. The bulletin states flatly that this is wrong. An instrumented hammer test does not require an idealized single half-sine waveform. When force and response are measured accurately, crest ripples and separated contacts remain known inputs — they don't invalidate the record. The physics argument is clean. A half-sine pulse is broadband but contains spectral zeros at odd multiples of the fundamental frequency (f/fo = 3, 5, 7, …, where fo = 1/(2T) for pulse duration T). That means a 'perfect' half-sine actually has holes in its excitation spectrum. Meanwhile, the supposedly messy waveforms — crest rippling and separated contacts — can fill those notches, sometimes providing better spectral coverage than the idealized pulse. Figure 3 in the bulletin normalizes spectra to DC and shows this directly. The uniqueness theorem does the heavy lifting. If an exact model and test article receive the same measured force history and initial conditions, uniqueness guarantees identical responses. The waveform shape changes which frequencies and amplitudes get excited, but it does not break the physics. An approximate model may show different errors depending on the input spectrum, but that's a model problem, not a measurement problem. The key reference is Majed, Sills et al. (September 2026), which derives the exact-model case for both linear and nonlinear structural systems. The recommended practice boils down to four checks: measure the complete applied force with adequate bandwidth, calibration, dynamic range, and time synchronization; verify response measurements aren't clipped or saturated; confirm the excitation covers the required frequency, amplitude, and nonlinear-response ranges; and reject a test only for an identified measurement, excitation, or coverage deficiency. Not for waveform aesthetics. This matters because the half-sine myth costs real money and schedule time. Test engineers repeat hammer taps chasing a clean waveform when the data they already have is perfectly adequate. In contexts like pre-pyroshock modal surveys (the bulletin references NASA-STD-7003) and flight hardware qualification, unnecessary re-tests mean unnecessary risk to hardware and unnecessary delays. The bulletin is a formal institutional correction to a widely taught but unsupported heuristic. The limitation is scope: this is a technical bulletin, not a peer-reviewed journal paper. It synthesizes existing theory (Ewins 2000, McConnell & Varoto 2008, ISO 7626-5:2019, Nayfeh & Mook 1979) and a companion paper (Majed et al. 2026) rather than presenting new experimental results. The argument is sound on first principles and supported by the referenced derivation, but the bulletin itself does not include independent experimental validation across a range of real structures.
