On the usability of different optical measuring techniques for joint roughness evaluation

The roughness of rock discontinuities is an important input parameter for mechanical models of rock masses. To reliably calculate roughness indices, adequate representations of the surfaces are required. Various optical measuring approaches have been applied in the past. However, many studies lack information on resolution and accuracy of the resulting surface meshes. These qualities are yet important, as they explicitly affect the deduced roughness metrics. Often, the sensors do not achieve the given precision and accuracy. Moreover, no technical standards presently exist for roughness evaluation from optical measuring approaches. Therefore, previous studies are difficult to compare. To overcome these issues, this study offers a comparison of four different techniques and sensors. Here, the focus lies on laboratory use and evaluation of micro-roughness, meaning sample sizes up to 20 cm in length. Stationary structured light scanning (SLS) serves as the reference method. As results, the surface models from dense image matching are very consistent with the reference. Their calculated roughness values accord to a high degree, both for 2D and 3D indices. In addition, roughness indices deduced from models acquired with manually operated SLS show deviations from the reference yet within an acceptable range. Instead, terrestrial laser scanning turned out to be not suitable for micro-roughness evaluation, at least at laboratory scale. Furthermore, in this contribution, an algorithm is applied, which can retrace all possible profile measurements directly from the triangulated surfaces. That way the ambiguity of the profile-based roughness measure joint roughness coefficient (JRC) is made visible.