Summary
The historic church building was reassessed and rated above 33%NBS (previously 14%NBS). This meant that the building was no longer earthquake prone. The DSA (Detailed Seismic Assessment) and outcome were peer reviewed by international experts and resulted in the preservation of this historic church – a happy outcome for both the client and local community.
Background
The client is a Catholic Diocese established in the late 1880s, with many parishes covering parts of New Zealand. The client owned and operated an old church, which is a Reinforced Concrete (RC), Unreinforced Masonry (URM) / timber building.
The building was assessed and rated at 14%NBS by Engineer A in a Detailed Seismic Assessment (DSA). Engineer A is a major international engineering firm. The DSA was undertaken by Engineer A qualitatively in accordance with the relevant European guidelines. There was no detailed quantitative analysis undertaken by Engineer A as part of the DSA.
However, a major earthquake shook the building, which only suffered minor damages. From the observation of the building and damaged areas on site, Engineer B believed that a detailed quantitative analysis would be required in the DSA to understand the true behaviour of the RC, URM / timber structure. As a result, the client engaged Engineer B to reassess the building using a different approach.
Building Description
According to the original specifications, the church building was constructed in the late 1930s with a combination of RC, URM and timber. It can be considered as a single storey building even though there is a small gallery on the upper floor near the front entrance. The church consists of a rectangular long body with a long nave and a square sanctuary. On the east and west sides of the church there is a single-storey symmetrical wing for sacristies, porches and confessionals. There is also a small basement area. On the east side of the building, there is a four-storey bell tower.
Figure 1 – Typical floor plan of the church
The seismic load resisting system within the building consists of the following:
The timber roof structure (diaphragm and truss actions)
The ground and upper timber floor structures (diaphragm actions)
RC ground floor slabs
RC bell tower structure
RC buttresses around the church
A series of transverse frames (each includes a main timber roof truss and two RC wall piers supporting the truss)
Longitudinal moment-resisting frames formed with the RC wall piers and the ring beam located on top of the URM infill walls on each side of the church
URM infills and facades: in-plane and out-of-plane actions
What did Engineer A do?
Engineer A conducted a Detailed Seismic Assessment (DSA) on the church building. The DSA was mostly qualitative in accordance with the relevant European guidelines. There was no detailed quantitative analysis undertaken by Engineer A as part of the DSA.
However, Engineer A undertook intrusive investigation on site. Material samples were extracted from the URM walls at several locations to determine construction details used for the URM walls. In addition, reinforcement scanning analysis was performed to establish steel reinforcement details used in the reinforced concrete.
Engineer A reported an overall seismic rating of 14%NBS for the building. It was limited by the critical structural weakness being the out-of-plane failure of the front façade. There were no detailed modelling and calculations carried out to support the DSA outcome.
How did Engineer B improve the work?
First, Engineer B carried out a visual inspection of the building and found that it only suffered minor damages from a major earthquake that shook the building. Then, Engineer B decided to understand the true progressive collapse behaviour of the structure by explicitly modelling and solving the full range of linear and non-linear behaviours of the structure under a range of three real earthquake ground motions selected and scaled by the project geotechnical engineer in accordance with the relevant loading standard.
For the seismic assessment of URM structures, the Finite Element Method (FEM) is typically
used in computer-aided 3-D modelling and analysis. Materials are modelled as a continuum, and elements are connected at nodes. It is assumed that the elements at the common node move with the same displacements. This assumption is not correct when element separation occurs in a progressive collapse situation during a major earthquake and, in this case, elements should be considered to move independently. Even though multiple node IDs can be used at expected separation locations, this technique often leads to inaccurate stresses at these nodal separation points; the potential cracks in elements and the effect of element separation on the building stiffness have also to be predicted using special techniques such as ‘smeared cracks’ and ‘discrete cracks’ approaches, both with limitations. For orthotropic materials such as masonry, it is not always possible to predetermine the locations of cracks in the elements in a progressive collapse situation.
Due to the limitations and weaknesses of the traditional FEM, Engineer B decided to use a new method that is capable of predicting the continuum and discrete behaviour of URM structures to a higher degree of accuracy. Using this advanced method, the structure was modelled as an assembly of small elements made by dividing the structure virtually as shown in Figure 2. This new method has been verified extensively by engineers and scientists around the world through experimental testing and observations of real buildings after major earthquakes.
Adjacent elements are connected by a set of springs including one normal and two shear springs located on the contact surfaces, representing the material behaviour including the full range of linear and non-linear behaviours. These nonlinear springs represent stresses and strains of a certain volume of the material. When the average strain value at the element face reaches the predefined separation strain value, all the springs at this face are removed, and the adjacent elements are then separated and may come into contact again later with new springs generated to represent the contact / collision behaviour.
It is a stiffness-based method, in which an overall stiffness matrix is formulated, and the governing equations including the stiffness, mass and damping matrices are directly integrated to obtain displacements and rotations. Based on this method, the structural collapse behaviour can be tracked passing through all the stages of the application of loads: elastic stage, crack initiation and propagation in tension-weak materials, reinforcement yielding, element separation, element collision (contact) and collision with the ground and with adjacent structures.
The final mathematical model of the subject building was built with approximately 150,000 solid finite elements (each with six degrees of freedom) and millions of non-linear matrix springs representing brick, mortar, concrete, steel and timber. Every brick, mortar joint, RC and timber elements were modelled explicitly and carefully in accordance with the as-built structural information and findings of site investigations and material testing. The full range of URM linear and non-linear behaviours were analysed and understood, including the following:
Out-of-plane arching action mechanism
Cavity-tie axial / flexural hinging
Pier rocking
Diagonal tension
Spandrel cracking
Joint bed sliding
Toe crushing
Interface behaviour of different materials
Dynamic interactions of separate members, parts and portions
Movement and final resting position of debris for planning of egress route
Figure 3 – In-plane failure modes of URM walls
Figure 4 – The out-of-plane arching action mechanism with various boundary conditions
Results
There were several time-history analyses completed. After each one, several steps were taken to assess the results. First, the final state of each material was determined at the last step of the time-history analysis. The URM elements that were more critical may be in one of the following states:
Fully elastic
Tension / compression failure (post-elastic state)
Open cracks
Complete separation of elements
Then, areas of the failure and open cracks were carefully identified from the analysis results. Those areas were assessed in terms of wall drifts and crack patterns to the performance-based acceptance criteria at three different levels: Immediate Occupancy (IO), Life Safety (LS) and Collapse Prevention (CP). They were defined in accordance with the relevant standards and guidelines. The final step was to check the ground material and identify any bearing failure of the foundations.
As expected, it was found that the critical structural weakness was the out-of-plane failure of the front URM façade. The overall rating of the building was limited by this failure and is greater than 33%NBS, meaning that the building was no longer earthquake prone.
What would happen if the building was subjected to a real earthquake greater than 100%NBS? A range of progressive collapse behaviours were obtained from further analysis.
FIGURE 5 – FINAL COLLAPSE OF URM CHURCH SUBJECTED TO REAL EARTHQUAKE GREATER THAN 100%NBS
Figure 5 – Final collapse of URM church subjected to real earthquake greater than 100%NBS
Conclusion
In conclusion, the existing church building was rated above 33%NBS (previously 14%NBS). This meant that the building was no longer earthquake prone. The DSA and outcome were peer reviewed by international experts and resulted in the preservation of this historic church – a happy outcome for both the client and local community.