Employing non-contact sensing techniques for improving efficiency and automation in numerical modelling of existing masonry structures: A critical literature review

Abstract

This paper presents approaches for the employment of non-contact sensing to enhance both the efficiency and reliability of numerical modelling of historic masonry. It commences with a thorough review of the high-level numerical modelling approaches of historic masonry. Following, the accuracy and cost-effectivity of available non-contact sensing techniques are reviewed for surveying masonry structures. These are: a) the total station; b) the laser tracker; c) Structure-from-Motion (SfM) photogrammetry; and d) terrestrial laser scanning (TLS). Then, strategies of automatically developing geometric models (i.e., numerical models before structural analysis) from geospatial data are reviewed, considering their potential for automation and usage. These were based on the employment of: a) point clouds; b) meshes; c) non-uniform rational basis splines (NURBSs); d) building information models (BIMs); e) orthoimages; and f) discrete points. Primarily, the review found that high-level numerical modelling approaches such as the continuum and block-based models are highly effective, but necessitate accurate geometric data for reliable results. To bridge this gap, the potential of emerging technologies such as SfM photogrammetry was found to significantly improve the efficiency and robustness of high-level structural analysis, through providing geometric data accurately and with a low cost. Moreover, the cloud-based (i.e., with a point cloud) and image-based (i.e., with an orthoimage) approaches of converting geospatial data into numerical models were also found the most effective, for continuum and block-based modelling respectively. This contribution demonstrates the potential to employ novel digital technologies such as non-contact sensing techniques to improve the efficiency and robustness of high-level numerical modelling approaches.