I-35 Bridge Collapse

Some general notes about bridges.

Bridge failure is always due to the supporting structure unable to support the loading forces applied to the structure.

Bridges have always had designed in to them a safety factor. For example a reasonable estimate of the total traffic loading will be estimated i.e 200 tonnes and a safety factor of 50% will mean that the bridge will support 300 tonnes before collapsing.

Cheap and poorly designed bridges have lower safety factors than more expensive better designed bridges.

The maximum load a bridge can withstand is a function of the load in which an individual tie or strut can support.

Bridges made from steel have very well known load failure characteristics. Tie and struts manufactured from steel will deform, buckle or lengthen before failure. The paint work will show cracking and blistering on the structural strut, tie members or jointing sections. Over a short period of time this will show up as localised surface corrosion.

Joints using riveted construction can give rise to crack initiation. This problem is well understood and any design will/should compensate for this problem together with proper bridge maintenance and inspection.

Creep failure in steel bridging structures is generally attributed to poor quality or cheap steel specified in the design or used in the construction phase. Not really an issue in well designed steel bridge structures.

Bridge maintenance is crucial. Inspection to ensure there is no corrosion of joints and structural members is important. Although generally in a steel bridge structure this generally means just keeping the bridge in a good painted condition and ensuring there are no loose or missing rivets and the joints of the ties or strut members. Inspection will be looking for cracked paint and localised corrosion.

The design load of bridges with respect to bridges carrying road traffic are generally designed for symmetric loading. There is a design assumption that a road traffic bridge, which is designed to carry say 200 tonnes with a 50% safety factor will have the loading distributed evenly along its length. The safety factor should/will compensate for uneven or asymmetric load distributions.


Some personal observations relating to the I-35 bridge.

The bridge is of an unusual design. The width of the upper road section is unusually wide (to accommodate 8 lane traffic) in comparison to the distance between the supporting concrete columns at each end of the steel arch. The steel arch is unusually shallow and is constructed from what looks like very small section I steel beams.
The bridge supported steel construction is very cheaply constructed being constructed from mostly I section steel in a conventional lattice steel beam riveted jointed construction.
There appears to be some evidence of some corrosion at the ends of the steel arches where they terminate at the supporting concrete pillars.

Bridge failure I suspect would have been due to the following causes.

The bridge was overall a poor design and cheaply constructed with a small safety factor (hence the 4 out 9 mark) but has stood without failure for 40 years.


Corrosion of the riveted joint section above the concrete pillars indicates possible recent loads in which the proof stress of the strut members of the arch above the concrete supporting pillars may have been exceeded, leading to permanent structural deformation. The inspection and maintenance program should have picked this up though and the bridge should then have been declared unsafe. The safety factor of the bridge would have now become critical and the bridge management should have either closed the bridge or reduced the bridge loading by only having single file road traffic and ensuring no heavy vehicles such as articulated lorries be allowed on to the bridge.

Recent resurfacing work I suspect is the main factor relating to the bridges collapse or more precisely the management of the resurfacing work being carried out. The critical factor in the I-35 bridge collapse is now the loading of the bridge with a poor reduced safety factor. The thickness of the newly laid road surface concrete (wet concrete being heavier than set concrete and if an extra 1 or 2 inches more concrete could easily have amounted to a huge increase in the bridge loading thereby reducing the safety factor even more), the type of concrete (high strength lightweight concrete which is more expensive may have been specified in the original specification but normal concrete may have been used this time around during the recent maintenance road surface construction).

This combined with today’s heavier road vehicles being distributed asymmetrically on one side of the bridge (the width of the road section being much wider than the support would have lead to a high degree of torsional forces being placed on the bridge structure for which I suspect it was not designed, at least not asymmetrically) would have led to the main I beam arch strut above the concrete pillar failing due to the twisting forces or the heads of the rivets popping due to the exceptional sideways forces at the strut end above the concrete pillar on the same side of the above traffic. (Tubular steel sections can resist torsional forces much more readily than I beams)

There really is no mystery to bridge failures. There is just poor design, poor maintenance, poor materials selection, but ultimately just poor management.