Seismic retrofit of URM structures (NZ)

Parapets: the highest-risk element in any URM building

Robert Hudson opened with a pair of photographs from Christchurch showing two adjacent buildings that both went through the 2011 earthquakes. One was retrofitted with braced frames, concrete shear walls, wall-diaphragm connections, and related techniques. The other was not. The retrofitted building is still standing. The unretrofitted building is not. The point, as Robert framed it, is that the work in this space, when it is designed well and built as designed, does work.

The opening technical section focused on parapets, which Robert characterised as the element with the highest per-unit life-safety risk in URM buildings. A parapet is effectively a lump of unreinforced bricks at the top of a building, experiencing amplified accelerations and largely disconnected from the main structure. The collapse data from Christchurch is unambiguous: on a graph of parapet height versus percentage of as-built parapets that failed, every bar is at or near 100 percent.

The reason code-based rocking assessments underestimate the risk is a common material detail: a damp-proof membrane at the parapet base. Codes in New Zealand, Australia, the US, and Canada assess parapets using a rocking mechanism. Where a DPM is present, the DPM acts as a sliding surface rather than a rocking hinge. Parapets slide off the building at accelerations well below the rocking threshold. Vertical seismic accelerations further reduce the effective weight stabilising the parapet, and short stub parapets can simply be ejected upward.

Among the retrofit approaches tested and observed in Christchurch, bond beams tied into and over the top of parapets had a roughly 50 percent failure rate. Robert described this as a technique to avoid: when bonded parapets do fail, the connection to the structure below means they take material with them. Steel strongbacks performed substantially better at around 25 percent collapse rates, with most of those failures tracing back to the wall below the parapet failing out-of-plane, or to anchor installation problems, not the steel brace itself. Timber strongbacks tested at 1.6g in both directions in cyclic loading. Robert stated directly that timber is a completely appropriate material for this purpose. For parapets up to around 800mm high, post-installed rebar using mechanical anchors rather than coring provides a lower-cost alternative that a local builder can complete with a hand drill.

Chimneys, diaphragm connections, and timber strongbacks

Chimneys present the same fundamental problem as parapets: unreinforced masonry at the top of a building under amplified acceleration, often pre-cracked and structurally isolated. Steel bracing for chimneys introduces uplift, which masonry handles poorly. Keeping braces horizontal rather than diagonal reduces uplift, and connecting at multiple heights along the chimney lowers the risk of a single shear failure cutting across the masonry at the brace connection.

Where a chimney is straight and not intended for ongoing use, post-tensioning is very effective. A threaded rod runs down the full height, bearing against a steel plate at the top and connecting to a cast concrete base tied into the surrounding structure with anchors at the bottom. Applying compression to the masonry dramatically improves performance. Robert showed shake-table testing comparing an unreinforced chimney failing at 0.42g to the same configuration post-tensioned and surviving 2g of input acceleration.

Wall-to-diaphragm connections are the first intervention in any URM retrofit. Without them, the wall behaves as a cantilever off the foundation, and out-of-plane failure follows at relatively low acceleration. Full-scale subassembly testing has produced reliable design guidance. One detail worth checking after connections are installed is mortar bed shear failure directly below the connection anchor, which can occur in buildings with very soft mortar. Where mortar strength is very low, the wall can revert to cantilever behaviour despite a well-designed connection above, because the mortar fails in shear beneath the anchor point.

Timber strongbacks are widely used across New Zealand largely because the installation is accessible. Standard sections from a local hardware store, cleat plates, mechanical anchors, and a drop saw are all that is required. The ability to adapt everything on site with hand tools matters when dealing with the variability of existing masonry buildings. Composite action between the timber and the masonry is possible to quantify, and a formula exists, but the increase in capacity is small and one-directional. Robert's general advice is to assume all lateral load goes into the timber. For large-span buildings, particularly partially or ungrouted concrete masonry warehouses, timber spans become impractical and hot-rolled RHS steel sections take their place.

In-plane retrofit systems: concrete shear walls and CLT

Concrete shear walls have very positive collapse statistics from post-earthquake reconnaissance. They are effective but expensive, primarily because of the cost to install shear studs into existing masonry. The cost breakdown Robert described from contractors pricing this work often runs nine to one: shear stud cost versus concrete cost. Optimising the reinforcing layout and shear stud spacing, rather than the concrete thickness, is where meaningful cost savings are found.

Cross-laminated timber panels have seen adoption in New Zealand retrofits, partly driven by research partnerships with Italian universities, where heritage buildings are older and the breathability constraints of masonry are better understood. CLT is lightweight and can be prefabricated: panels arrive on site and anchors are set through them into the masonry. Plywood works on the same principle and is considerably cheaper. It requires more material to match CLT capacity, but the cost difference often makes it the better choice. Both materials can be combined with timber strongbacks in a single installation: set the strongback locations, lay the plywood over them, and drive anchors through both layers in one pass.

Cavity walls can be strengthened for out-of-plane loading by installing semi-rigid veneer ties to create composite action between the inner and outer wythes. This effectively increases the wall's structural thickness without any internal work. For buildings where composite action alone is not enough to reach the required NBS percentage, post-tensioning can be added, but it must be done in conjunction with the composite tie installation. Applying post-tension force to a cavity wall without composite action creates an eccentric load that can cause the veneer to pop out. The two techniques together create a single section capable of accepting the compression.

Anchor installation in existing masonry: what the field data shows

Fifty percent of all failed retrofitted URM buildings examined after Christchurch had some number of failed epoxy anchors. This is not a New Zealand-specific finding. Robert showed a photograph from the 2014 Napa earthquake in California of a wall-diaphragm connection where a cluster of seven epoxy anchors had only three still holding bricks, the rest pulling out cleanly. The US requires continuous special inspection for epoxy anchor installation in existing masonry — an independent inspector standing on site for each anchor installation — and the problem still occurs.

Robert identified three causes. The first is hole cleaning. In soft masonry, including brick, clay tile, sandstone, and limestone, a pressurized air gun tends to fracture material from inside the hole and multiply the dust problem rather than remove it. A vacuum is the correct tool. The second cause is an incompletely filled hole. The third is loading the anchor before its chemical curing time has elapsed, which shears bonds before they have fully formed and leaves the anchor permanently below capacity.

For mechanical anchors, two parameters govern selection in multi-wythe brick masonry: embedment and diameter. Embedment must engage every wythe in the wall, with a minimum of 10 inches (approximately 200-250mm) as the requirement now being codified in the US through the masonry society's existing masonry standard. Short anchors that reach only the first wythe are vulnerable to whatever mortar surrounds that single brick, and mortar quality in existing buildings is highly variable. On diameter, Robert identified approximately 12mm as the threshold above which the installation stress from a mechanical anchor significantly increases the probability of splitting the masonry unit. A split unit loses essentially all of its anchor capacity.

Cavity wall tie inspection

Existing cavity ties present an inspection challenge that depends on the mortar type. In buildings with lime mortar, the mortar is more permeable than the surrounding bricks and draws moisture into the mortar joint. This means the mild steel wire ties corrode most severely in the mortar joint, which is the one location a borescope camera through the cavity cannot reach. The tie may appear intact in the cavity portion while the section embedded in the mortar joint has corroded through. The only way to confirm tie condition in a lime mortar building is to remove sample veneer bricks and physically inspect the ties in the mortar joint. The US building code mandates this approach before signing off on existing cavity ties. Most other jurisdictions do not require it by code, but the masonry conditions that drive the corrosion pattern apply regardless of jurisdiction.

In buildings with cementitious mortar, the moisture behaviour is different: condensation tends to accumulate in the cavity itself rather than being drawn into the mortar bed. For these buildings, a borescope camera through the cavity provides a meaningful assessment, and the images give a reliable indication of tie condition without removing brickwork.