In general, the aim of acoustic emission (AE) assisted tensile tests of composite materials is to identify and characterize the damage and failure modes of the specimens. This paper presents a fiber-bundle-cells (FBC) based statistical model, which provides a possible solution to the problem of characterizing the mechanical and failure behavior of the material. The model, based on the results of mechanical tests and AE measurements, decomposes the measured AE event number and tensile force-load time processes into components corresponding to the different damage modes. The AE events belonging to different failure modes are described by inhomogeneous Poisson point processes, while failures are modeled with the breakage of fibers as elementary parts of the sample. Hence damage modes can be characterized with the number fraction, and the tensile strength and signal energy distributions of the components. Moreover, the variation of the number fraction of the intact or damaged fibers as a function of the load time can be calculated and depicted as well. As reliability function or a kind of damage map, it reveals the mechanical load-bearing ability of the material tested. The applicability of the model is demonstrated by compact tension testing and the comparison of short and long glass fiber reinforced VERTON PP sheets and injection molded wood fiber reinforced PP composites.With the help of fast Fourier transformation (FFT), the frequency components of the single AE signals can be characterized by the amplitude or power spectrum in the frequency domain. Their description parameters are the bandwidth (B), the mean frequency (fm), the frequency of the maximum amplitude component (fP), and the amplitude (AP) or energy (UP) of the maximum peak. Similar parameters can be obtained from the mean spectra of a series of AE signals, and histograms and/or cross-plots can be calculated from the data of single AE signals. 2,4 Special techniques can provide information in both the time and the frequency domain, resulting in time-dependent spectrum statistics. 2,4 The classic method is the short time FFT (SFFT) technique, which uses a moving time window. In the case of wavelet transformations (WT), AE signals are analyzed with the help of various wavelet components. The signal transformations mentioned above provide additional or in itself usable information about the sources and failure modes in the frequency domain [21][22][23][24][25][26] or both in the time and the frequency domains [27][28][29][30][31][32][33][34][35][36] . In the latter case, wavelet transformations based on Haar 28 or Gabor 27, 33 wavelet components as well as Hilbert-Huang transform 30 can be used to decompose the AE signals and calculate the energy of the components. [27][28][29][30][31][32][33][34][35][36] In the case of thin plates, modal analysis can be applied to study and separate extensional and flexural Lamb waves. 37,38 The finite element method (FEM) can give a good base for wave discrimination. 38 The localization of a single AE signal in the spe...