A number of concerns were raised regarding the design and construction of reinforced concrete (RC) walls following the 2010/2011 Canterbury earthquakes. A lack of distributed cracking was observed in several modern lightly RC walls, such as the Gallery Apartments (see Figure 1a). The potential issues with lightly reinforced panels combined with examples of poor detailing and panel fixings also lead to renewed concerns around the seismic response of precast concrete wall buildings. Lastly, the potential non-ductile response of older lightly reinforced concrete walls was also a contributing factor to the collapse of the PGC building (see Figure 1b).
|a) Gallery Apartments (Credit: Des Bull)||b) PGC building (Credit: 3 News)|
Figure 1: Damaged lightly RC walls in Christchurch
This project will build on research into the design of new lightly RC walls that was initiated under the Natural Hazards Research Platform. A series of experimental tests and numerical modelling will be used to verify the behaviour of existing wall designs, as well as to investigate improved design procedures and details. The detailed research objectives are: • Determine minimum reinforcement requirements and deformation capacity of for lightly reinforced walls. • Determine the deformation capacity of older singly reinforced walls. • Evaluate the capacity of precast walls with grouted connections and identify improved connection details. • Evaluate out-of-plane deformation capacity of base connections for singly-reinforced walls, including bi-directional loading.
Testing and outcomes
Minimum reinforcement limits
A combination of detailed numerical modelling and experimental tests were used to investigate the seismic behaviour of the Gallery Apartments walls and walls compliant with NZS 3101:2006 (A2). As shown in Figure 2, the models confirmed that the as-built grid-F wall in the Gallery Apartments was likely to only form a single flexural crack. A greater number of primary flexural cracks formed when the vertical reinforcement was increased to satisfy the current minimum requirements in NZS 3101:2206 (A2). However, additional reinforcement is required at the wall ends to ensure a large number of secondary cracks form.
|(a) As-built grid-F||(b) NZS 3101 (A2)||(c) Additional reo at ends|
Figure 2 Modelled walls
A total of six test walls were tested to investigate the seismic behaviour of RC walls with distributed minimum vertical reinforcement in accordance with current provisions in NZS 3101:2006. The test setup used to tests the lower portion of a 50% RC wall is shown in Figure 3.
Figure 3 Test setup
The behaviour of all six test walls was controlled by 1-3 large flexural cracks at the wall base, as shown in Figure 4. The failure for all the six test walls was controlled by vertical reinforcement buckling and subsequent reinforcement fracture. The experimental results confirmed that current minimum vertical reinforcing limits in NZS 3101:2006 (A2) are insufficient to ensure that a large number of secondary cracks form and are only suitable for walls designed for low ductility demands.
|(b) Extent of flexural cracking|
|(a) Overall condition at 2.5% drift||(c) Concrete crushing and bar buckling at east end|
Figure 4 Photo of wall C1 at the end of testing
A second phase of the tests are currently underway to investigate the seismic response of RC walls with additional reinforcement at the ends of the wall, in accordance with the proposed amendments to minimum vertical reinforcement requirements for ductile RC walls in NZS 3101:2006 (A3).
Precast wall connections
Experimental tests are currently underway to assess the in-plane and out-of-plane response of precast wall panel connections. A review of manufactured precast concrete panels was undertaken in order to develop a comprehensive understanding of the common typologies for connections between precast concrete panels and foundations. Examples of three commonly used wall-to-foundation connection details are shown in Figure 5.
(a) Threaded inserts (b) Metal duct (c) grout sleeve
Figure 5 Cross sections of typical wall-to-foundations connections
A series of panel tests were completed to investigate both the monotonic and cyclic out-of-plane response of typically constructed threaded insert wall-to-foundation connections. As shown in Figure 6, the test panels did not perform well with the flexural cracks in the panel propagating vertically into the joint region behind the back of the inserts and the connection started acting like a pin. In many cases the panel did not reach its full flexural capacity prior to the onset of this failure mode.
| || || |
| (a) Panels setup for testing||(b) Connection failure ||(c) Strut and tie model |
Figure 6 Out-of-plane tests of threaded insert connections
A further series of out-of-plane tests are currently underway to investigate improved connection designs. Following these tests a bi-directional tests will be conducted on the best and worst designs to highlight the potential interaction between out-of-plane damage and in-plane capacity.
An experimental programme was developed to assess the seismic behaviour of precast concrete panels connected to the foundation using grouted connections. An example of the typical behaviour of a wall with a grouted metal duct connection is shown in Figure 7. The wall behaviour was characterised by the opening of the joint at the wall-to-foundation interface with fracture of the vertical reinforcement at large drifts. Damage was observed around the metal duct at the conclusion of the test, but the loading was not demanding enough to generate significant spalling or failure of the reinforcement lap splice.
|(a) Test wall||(b) Damage at the connection|
Figure 7 In-plane tests of grouted metal duct connections
In-plane tests of precast panels with grouted connections are ongoing. The currently planned tests will be conducted on larger panels and with an appropriate axial load applied. Test panels with grout-sleeve connections will also be undertaken.