Development and testing of retrofit solutions for hollow-core floors in existing buildings

Abstract

The collapse of precast concrete flooring components in Statistics House and varying levels of damage to precast floor units in many other buildings during the Kaikoura earthquake has increased concerns about the performance of these elements in earthquakes. While the details for these floor systems have been improved in new buildings, support conditions for units in existing buildings designed before 2006 are likely to lead to significant damage and potentially collapse in design level ground motions. Buildings with precast floors comprise a large percentage of the commercial building stock in all New Zealand cities, with likely over 60% of commercial buildings in Wellington falling in this category. There are increasingly more residential buildings with older precast floor details as more buildings are being converted from commercial to residential in Wellington CBD. A Wellington Fault event will undoubtedly lead to multiple floor collapses in numerous buildings throughout Wellington. Assessing the likely performance of these floors in an earthquake is a challenge for engineers. While guidance has recently been developed for the seismic assessment of buildings with precast floors (so-called Yellow Chapter), engineers will urgently need direction on retrofit approaches to address vulnerable buildings. In particular, concern has been raised that seat angles, already provided as a retrofit for several buildings, could potentially lead to unintended negative moment failures and collapse of hollowcore floors. This research identifies under what conditions such unintended failure modes may be triggered and provides a retrofit solution where vulnerability to negative moment failure is identified. An experimental investigation was directed at issues related to 200 mm deep hollowcore units (known as loss of seating (LOS) and Negative Moment Failure (NMF)) that could lead to casualties in earthquakes. A focus was on identifying seating connection details that would lead to the unfavourable failure mechanisms and validating retrofit options to remediate existing floors at risk. A previously used retrofit known as the “seating angle retrofit” to avoid LOS was also examined to determine if it would promote NMF in cases where it was installed “hard up” against the bottom of the hollowcore unit. It was found that the relative flexibility of the most commonly used seating angles reduced the severity of NMF promotion. This was a good outcome as it meant that many existing cases of the retrofit in New Zealand buildings will not require further remediation. However, it was found that seating connection details with stiffer seating angle retrofits or strong or short starter bar configurations are prone to NMF. These cases represent a smaller subset of floors in New Zealand but will require additional retrofit. Three retrofit strategies were tested to fix NMF prone cases. These were: • Cutting starter bars at the interface between the hollowcore unit and support beam to release restraint and demand on the unit at the end of the starter bars – where negative moment cracking initiates. • Post-installing bars into the unit topping to increase the strength of the unit at the critical section for crack initiation for NMF. • Lowering the seating angle retrofit by 10 mm to remove the additional restraint it imposed on the unit. All of these retrofits proved successful for preventing NMF. The diaphragm weakening side effects of cutting starter bars means that this retrofit is only appropriate for some areas of a floor though.