A group of Israeli researchers are working on developing algorithms to predict the damage states of concrete buildings with masonry infills.
After an earthquake, Terrestrial Laser Scanning (TLS) can capture point clouds of the damaged state of building facades rapidly, remotely and accurately. A long-term research effort aims to develop applications that can reconstruct ‘as-damaged’ BIM models of reinforced concrete (RC) framed buildings based on their ‘as-built’ BIM models and scans of their ‘as-damaged’ states.
This paper focuses on a crucial step: generating an initial ‘best-guess’ for the new locations of the façade structural members. The output serves as the seed for a recursive process in which the location and damage to each object is refined in turn. Locating the ‘as-built’ structural members in the ‘as-damaged’ scan is challenging because each member may have different displacement and damage.
An algorithm was developed and tested for the case of reinforced concrete frames with masonry infill walls. It exploits the topology of the frames to map the original structural grid onto the damaged façade. The tests used synthetic datasets prepared from records of two earthquake-damaged buildings. In both cases, the results were sufficiently accurate to allow progress to the following step, assessment of the individual structural members.