Criterion for Objective Determination of Threshold Value in Filtering Surface Roughness Signal of Solid Wood After Machining with Fast Fourier Transform (FFT) Based Filter
Research subject and fields:
The article presents the method for an objective determination of threshold value, needed for filtering out the anatomical roughness signal, with filter based on fast Fourier Transform (FFT), from surface roughness profile after machining. The method includes experimental preparation of solid wood surface by cutting in such a way to get a surface that can be considered to represent only anatomical roughness, with no other influence. Experiments were performed on radial cross section of solid oak wood (Quercus robur L.) so that the results could be compared with roughness profiles that were previously obtained in experiments after sawing with circular saw. From these samples and based on frequency analysis of anatomical roughness signals, the threshold value was determined to be 1.6 μm. The average value of Ra parameter of anatomical roughness for specimens of radial cross section of solid oak wood was 2.1 μm with standard deviation of 0.3 μm. The importance of choosing adequate sampling length in threshold determination, based on frequency analysis of anatomical roughness signal, was also established.