Civil Engineering Disasters – The Hyatt Regency Walkway Collapse

July 23, 2012

While the spectacular demise of the original Tacoma Narrows Bridge can be considered the largest civil engineering disaster in terms of physical size, it doesn’t hold a candle to the cost in human lives of the collapse of the Walkway in the Hyatt Regency. The total lives lost through that disaster were 114, with over 200 more hospitalized due to injuries. This collapse wasn’t an error in the design or calculations to back up that design, but rather in the execution of the design by the fabrication company who built the walkway and lack of project management by the civil engineers on the project.

As originally designed, there were three walkways, spanning the atrium and connecting the two halves of the second, third and fourth floors. The walkways for the second and fourth floors were stacked one over the other, while the one for the third floor was offset to one side. These walkways were suspended from the ceiling on metal rods, which attached to box beams that crossed underneath the walkways.

In the original engineering drawings, the same threaded rod ran from the ceiling, down through the fourth floor walkway and continues down to the second floor walkway. Large nuts, backed by washers were to connect the threaded rods to the box beams. The box beams themselves were two steel “C” channels, which were welded together.

Due to the difficulty of manufacturing a threaded rod that long, and the risk of it becoming damaged in shipment and installation, the steel fabricator who was building the walkway decided to change the design, using two rods in each location, rather than one. The first rod would run from the ceiling down through the upper walkway and attach there. The second rod would run from the upper walkway, down to the lower one, hanging it from the upper walkway’s beams.

On the surface, this seems like an acceptable solution to the problem, making the design easier to fabricate and install. Since the same size steel rods were being installed for both the upper and lower sections, it would seem that the rods should be able to carry the weight. However, this did not take into account the amount of weight placed upon the nuts on the bottom side of the upper walkway’s beams.

In the original design, those nuts only had to support the weight of the upper walkway. None of the weight of the lower walkway would be transferred to those nuts, but rather directly to the hanger rod by the nuts on the bottom of the lower walkway’s beams. Although there was an error in the calculations, the nuts on the upper walkway could marginally accomplish their task if you eliminated any safety factor.

In the modified design, those critical nuts no longer had to carry the load for only the upper walkway, but the entire weight of the lower walkway as well. That added weight meant that the nuts could only carry 40 percent of the dead load. Add the weight of the people jumping and swaying as they watched the dance contest in the lobby, and the nuts didn’t stand a chance.

To envision the effect of this, think of a rope handing from a tree, with two people handing from that rope. Each person is holding onto the rope, supporting their own weight. However, if the lower person grabs hold of the upper person’s leg and releases the rope, the upper person now has to support the weight of the two of them. The possibility of that person’s arms failing and the two falling is greatly increased.

In addition to the overload, the holes for the box beams had been drilled through the welds holding the two “C” channels together, the weakest point in the box beam. When the wreckage was examined, it was found that the pressure on the nuts not only caused the nuts to fail, but the holes in the box beams to deform, allowing the nut to pass through.

Forensic engineering discovered that this failure was actually a series of four failures on the part of the engineering team. The first failure, which may have been survivable, was the miscalculation of the weight of the walkways, and the ability of the nuts to support them. The second failure was a failure of communications between the engineering firm and the fabricator who made the design. Thirdly, nobody ever calculated the difference in loads and stresses made by the design change. Finally, the engineering team never checked the final work by the fabricators, to ensure that they had followed the design in all details.

The civil engineer’s responsibility doesn’t end when the drawings are completed. Engineering involvement throughout the construction process is essential to avoid similar disasters. Contractors often make design changes, whether it is the replacement of welds with bolts or changing the materials used. These changes are made by non-engineering professionals, based upon their experience, but without the benefit of knowing how to calculate the effects of their change. In some cases, those changes can cause disaster.

Had the civil engineers who had responsibility for this design completed their responsibilities, this disaster could easily have been avoided.