Procedure for data preparation for the correct evaluation of the geometrical structure of the surface in functional analysis

Parametric results describing the geometric structure of a surface are influenced by many factors. One of the most important is the correct preparation of input data. This is related to the type of Surface levelling, the selection of a suitable measuring section or the applied filtration. The data preparation procedure depends on the adopted measurement technique: contact or optical. Depending on the type of data acquisition, it is necessary to implement supplementary procedures. Each of the applied operations on data directly influences the obtained values of individual spatial parameters 3D and surface parameters 2D, such as roughness, waviness or base profile. In this paper, different surface levelling options and selected filtering methods were analysed. In the case of optical measurements, the influence of filling in non-measured points was also considered. The results obtained allowed us to develop of a data preparation procedure and to determine the influence of the different steps on representative parametric values. It was found that the most significant factor influencing the parameter values obtained is the removal of outliers. In this study, based on two measurement techniques: contact and optical, a procedure for sequential processing of input data was prepared to assess the impact of the steps performed on the resulting values of parameters describing the surface structure. The procedure is based on one of two main types of surface levelling: levelling with the least squares plane (LS) and levelling line by line (LbL), along with the impact of one filtering method – removal of outliers (RO). In the case of optical measurements, the impact of filling unmeasured points (NM) was also considered. It was found that the most significant factor affecting the obtained parameter values is the removal of outliers (RO). The relative differences between the data for which this operation was not applied are 13 % for contact profilometry and approximately 1% for optical profilometry. On the other hand, the procedure of filling unmeasured points (NM) has the least impact – however, this is only applicable in the case of optical measurement techniques. The choice of the appropriate levelling method can affect the final results. Differences between these methods are generally small (up to 5 %), but in certain situations, they can be significant. The obtained results allowed for the development of a data preparation procedure and determining the impact of various stages on the representative parametric values.