Preliminary Report on Bridge Damage from the Darfield (New Zealand) M7.1 earthquake of September 4, 2010 – Draft of 2010-09-23

by Michel Bruneau[1], Myrto Anagnostopoulou[2], and Alessandro Palermo[3]

Overall, bridges in the Canterbury area have suffered little damage, and this damage was mostly limited to the Christchurch and Kaiapoi areas. A number of factors contributed to this behavior.

  1. Most bridges in the Canterbury area are small to moderate spans; such spans are recognized to generally exhibit a more sturdy seismic response, due largely to their symmetry and limited reactive mass.
  2. Like many buildings and other infrastructure in the areas subjected to the earthquake excitations, bridges were generally designed to resist forces substantially larger than the demands imparted by this particular earthquake.
  3. Bridges shared a number of common design features that gave them high seismic resistance.

Regarding this third point, Figure 1a‐c shows the Gayhurst Road bridge over the Avon river in Christchurch, in a neighborhood where extensive soil liquefaction took place. River banks had multiple sand boils, the ground surface exhibited many large open cracks due to lateral spreading of the soil, and many residential homes suffered extensive damage due to differential settlements. At that location, in spite of this extensive liquefaction and soil movements, that bridge did not suffer any significant structural damage, except for abutments where vertical cracks in the wall at the edge of bridge deck width developed under the applied soil pressures. This can be explained by the sturdy monolithic structure of this bridge, in both its longitudinal and transverse directions; wide wall piers provided a stiffness and strength largely in excess of the values needed to resist severe seismic excitations transversely to the axis of the span, and continuity of the superstructure from abutment to abutment provided a rigid behavior in the longitudinal direction. Note that while the bridge’s wall piers were likely supported on piles, damage to such piles would be difficult to identify if at all present (this would incidentally be the case for all the bridges inspected as part of this earthquake reconnaissance visit).

The longer Bridge Street bridge connecting South Brighton to Christchurch behaved similarly, with the difference that, in this case, the lateral forces due to lateral spreading at the abutments rotated those abutments about a contact point at the slab level, resulting in a residual rotation of the abutments as well as distortion and sliding of the neoprene bearings (as seen in Figure 2a‐c) – the original location of the bearings could be clearly seen on top of the abutments at the point of girder supports (Figure 2d‐e). As a result soil lateral spreading and settlement, both abutments moved closer to each other, the West Abutment (from Bexeley Road) moving in contact with the deck.

Some bridge not endowed with similar strengths and stiffnesses to resist the demands applied to them as a consequence of the liquefied soils did not fare as well. For example,

  • A pedestrian steel truss bridge over the Avon River in Christchurch buckled axially under the axial thrust imposed by the movements of its abutments due to the lateral spreading action of soils on top of liquefied layers (Figure 3a‐b). This buckling introduced compressive forces in the bridge truss chord. These forces, unanticipated in the original design, consequently led to local buckling of some chord members, twisting of the superstructure, and uplifting at a support location above one of the braced pier in the river.
  • A pedestrian arch bridge also over the Avon River twisted slightly at its southern abutment; slight spalling was observed at center span of one of the arches, and a sign posted the bridge as closed due to hazards created by utilities (power pipes) carried by the bridge and exposed at the abutment by the liquefied soils (Figure 4a‐b). Horizontal cracks parallel to the axis of the bridge were also observed in the soil at the abutment and along the road, as evidence of local
    resistance of the bridge at that location to the lateral spreading of the soil, as well as horizontal cracks in the abutments in response to the pressures excerted by lateral spreading of the soil.
  • A disused railway bridge turned into pedestrian use in Kaiapoi (north of Christchurch), consisting of steel girder beam supported on timber pile bents, lost supports over part of its length (Figure 5a‐c). Many of those piles split along the vertical plane in which lied the bolts used to connect the spans to the pile bent. Evidence of the lateral pressure on the piles was provided by the large gaps between the soil and piles on one side of the piles.

The Mandeville pedestrian suspension bridge in Kaiapoi, built in 1874[4] over the Waimakarini River, also suffered substantial damage when one of its towers rocked longitudinally beyond the point of stability above its support, resulting in a downward span failure in the direction of the falling tower, and concurrent beam fractures in the opposite upward moving span (Figure 6a‐b). Cursory inspection of the steel connectors between the timber towers and its supporting timber pile bents did not lead to a clear understanding of the mechanism providing stability to the structure under normal unbalanced load conditions. Lateral spreading may have also partially contributed to this
failure, albeit in a secondary way.

In some instances, while the bridges remains essentially intact, the approach spans partially or totally failed, making access to the bridge more difficult or impossible. For example, the twin continuous bridges at the Chaney overpass on Highway 1 north of Christchurch were found to be structurally sound, and tied to their abutment walls to prevent unseating there. However, due to liquefaction at the site surrounding the bridge, the approach to the southbound lanes of that Highway 1 settled by a few inches (Figure 7a‐b). After a brief closure for inspection, this busy route was reopened with signage reducing the speed to 30 km/h (down from 100km/h) for the safety of motorists driving across that bridge.

The approach to the River Bridge, in Lincoln (approximately 8 miles from the fault), exhibited a dramatic evident of settlement from its original perfectly horizontal alignment (Figure 8a). Again, cracks in the approaches parallel to the axis of the bridge provide evidence of the resistance against lateral spreading provided by this short monolithic span (Figure 8b).

Many bridges at similar distance (or closer to the fault), where soil liquefaction did not occur, did not suffer damage. For example, the Highway 1 bridge across the Selwyn river (Figure 9a), less than 3 miles from the fault, and the railroad bridge adjacent to it, were operational. As the main North‐South route from Christchurch to south of the island, this bridge consisted of multiple simply supported spans tied together in pairs using reinforced steel wedges over wall piers supported on piles (Figure 9b). Note that similar ties were observed in many other bridges on major thoroughfares, such as the Avondale Rd Bridge in Christchurch shown in Figure 10 for example.

Finally, note that soil settlement and lateral spreading at bridge abutments often detrimentally affected the utilities carried by those bridges. For example, fracture of a sewer pipe across the Pages Rd Bridge near Brooklands north of Christchurch, contaminated the river (Figure 11a‐b). Incidentally, flexural cracks were observed on the abutment piles exposed as a consequence of soil lateral spreading (Figure 11c).

[1] Professor, Dept of Civil, Environmental, and Structural Engineering, University at Buffalo, Buffalo, NY

[2] SEESL Structural Engineer, Dept of Civil, Environmental, and Structural Engineering, University at Buffalo, Buffalo, NY

[3] Senior Lecturer in Structural Engineering, Dept. of Civil and Natural Resources Engineering, University of Canterbury, Christchurch, New Zealand

[4] Historic Kaiapoi and Mandeville Bridge

Leave a reply

You must be logged in to post a comment.