Detecting The Location Of Defects Of Wood Based On Modal Analysis

In this paper, the experimental materials are tree species (larch and poplar) with much natural defects. The object of the test were wood beams and knots and cracks in these wood beams were natural defect. A fast Fourier transform (FFT) analyzer was used to analyze the modal of specimens, and the natural frequency and vibration mode of wood beam were obtained.Then calculating elastic modulus, shear mode and modal shape of wood beam, and analysis the different of natural frequency and modulus between wood beam with defects and wood beam without defects. Fitting the sine curve of modal shape of specimens by MATLAB, and the determination coefficient (R2), the error sum of squares (SSE) and the mean square error (RMSE) were represented smoothness degree of curve fitting, and The larger R2 and smaller SSE and RMSE, the higher degree of curve fitting. The area (S) between fitting curve and modal shape was represented degree of curve smoothness, and the larger S, the worse curve smoothness. And then, every wood beam was processed into four small specimens, and small specimens were used for experimental modal analysis to research the relation between the defect area and density. After that, the specimens was used to calculating modal analysis by computer software ANSYS, and using the different unit, meshing and calculation method to simulate repeatedly, finally, the optimal analogy method of calculating modal analysis was obtained. At last, comparing the results of calculating modal analysis and experimental modal analysis to verify the experiment results. The conclusion of two kinds of modal analysis experimental is as follows:(1) Defects can change the wood physical properties, and the natural frequency and the elasticity modulus would decrease with the increase of defects. Fitting the sine curve of modal shape of specimens by MATLAB found that R2 were more than 0.9429. From the first to the sixth order of modal shape of specimens, SEE, RMSE and S would increase with the decrease of defects. It is illustrated that the higher order of modal shape, the lower degree of fitting and smoothness. Degree of variation of specimens with defects was larger than that of specimens with defects for specimens which were processed from the same board. It could judge whether the area of the specimen natural defect is greater than 5 cm2 by the value (S) of degree of curve smoothness. It is illustrated that the area of natural defects of specimen is larger than 5 cm2 when the mean value of smoothness (S) of the first six order modal shape is larger than 1.(2) Knots have an effect at the peak of first, second and third order modal shape wood beam. The curve of fourth, fifth and sixth order modal shape of wood beam would change in the place of knots, and the change was curve monotonicity and not degree of smoothness. However, first order modal shape of wood beam with cracks was asymmetry, and the difference of first tapping point and the last tapping point were larger than 0.4, and the imaginary frequency charts which were obtained by tapping the point located on the crack show two peaks. It showed that it could be qualitative, quantitative to detect the defects and detect the location of knots and cracks through natural frequency, elastic modulus, modal shape and imaginary frequency chart.(3) The setting of poisson’s ration has no effect on the result of modal analysis when make the calculating modal analysis by ANSYS. Wood beam element was set as solid 186, and wood model were meshed by hexahedron. The natural frequency (ff) obtained by calculating modal analysis were larger than the natural frequency (fs) obtained by experimental modal analysis through the simulation results.ff was set as abscissa and fs was set as ordinate to make binomial fitting, and then the natural frequency could been adjusted by the dependent equation. The change rule of modal shapes that obtained by ANSYS and FFT were the same, and it was verified that could be qualitative to detect wood natural defects and detect the location of these defects by modal analysis.