Your record. A clearer response.
Inspect an acceleration record, explore shaking intensity, and calculate linear response spectra. Open the link, choose your data, and run your analysis.
Acceleration history
Analyzed motionMethods, assumptions and numerical verification
SI calculations use g = 9.80665 m/s². Arias intensity is π/(2g) ∫a²dt, integrated by the trapezoidal rule. D5–95 uses interpolated cumulative-intensity crossings. Zero motion has undefined significant duration.
Linear SDOF response uses u″ + 2ζωu′ + ω²u = −ag with zero initial displacement and velocity. The browser solver uses average-acceleration Newmark with substeps at most Tmin/80, or Tmin/160 for convergence checks. This differs from the Python package’s exact state-transition solver. Peak sampling and integration errors remain. Substeps cannot restore missing input bandwidth.
PSA = ω² max|u|. Absolute Sa = max|−2ζωu′−ω²u|. Spectra include a one-input-step ramp from the final acceleration to zero, then the selected zero-input tail. Intensity measures use the record interval only.
Fourier amplitude is dt × |FFT(a)|, with zero padding to the next power of two. No taper or one-sided doubling. Plot drawing uses range-preserving reduction for large records. Downloads retain all samples.
For research testing. Check input provenance, sampling, and numerical convergence. This tool does not perform site response, liquefaction prediction, or design-code checks.
File format and first test
CSV example: first row “time_s,acceleration_g”, followed by numeric time and acceleration pairs. Choose the appropriate columns. Time must increase uniformly. For acceleration-only CSV, select “Use supplied dt”. AT2 requires an NPTS and DT header. Multi-channel V2/V3 files need separate extraction.
Start with the demonstration. Doubling the amplitude scale should double PGA and linear spectral ordinates, quadruple Arias intensity, and leave D5–95 unchanged. The demonstration is a synthetic signal, not an earthquake record.