Storm-surge barriers (75-m long, 13-m wide, and over 8-m high) designed by a Netherlands-based engineering firm deploy when water reaches 0.5 meter above average sea level. MSC.Marc simulated the deployment and pinpointed high stresses in the tube. The 33-ton dam is made of rubber reinforced with 16-mm-diameter nylon cord.
Storm-surge barriers (75-m long, 13-m wide, and over 8-m high) designed by a Netherlands-based engineering firm deploy when water reaches 0.5 meter above average sea level. MSC.Marc simulated the deployment and pinpointed high stresses in the tube. The 33-ton dam is made of rubber reinforced with 16-mm-diameter nylon cord.

Design firm Hollandsche Beton-en Waterbouw had never built such a structure, so the concept had to be validated. MSC.Software Professional Services, Santa Ana, Calif., simulated the dam using FEA software MSC.Marc and suggested some design changes.

The dam is actually a large rubber tube anchored to the river bottom. The deployment process consists of opening the dam's interior to water while air compressors provide positive buoyancy. The air keeps the dam above the water level. When the storm surge is over, the dam deflates and the membrane collapses into a base or sill on the river bottom, where rollers position it over the width of the base.

"Stresses in the first analysis were too high at clamp points," says Maarten Oudendijk, project manager with MSC.Software. "But peak stresses vanish far away from the boundary because the stress is better distributed. We needed to fasten the membrane so it would not break at high stress points."

The first analysis clamped the tube model so it could not move. Real-world tubes move when pulled on. "A second simulation modeled the clamp like a spring, letting it move a little," says Oudendijk. The simulation determined the force required to start moving the tube in the clamp and showed the mechanism would work without high stresses.

"The relatively thin material makes it necessary to apply loads and boundary conditions to keep the analysis stable," says Oudendijk. One such condition was a consistent and high-tensile force to stretch the membrane. The simulation also showed the inflation and deflation procedure.

Designers created a 1/25-scale model in a hydraulics lab to verify the simulation's accuracy. Oudendijk's team found two stable positions for the dam. Another plus: transitioning from one position to the other lowered the loads on the dam.

The model and computer simulations agreed well. But the real test came in October 2002 during one of the worst storms of the last 30 years. Wind gusts reached 72 mph. The dam deployed in less than hour and held back the storm surge.

-- Paul Dvorak