High-performance computing toughens slender sign structures

	Sougata Roy, a senior research scientist in the ATLSS Center, is working to develop sign structures that resist fatigue cracking.

The countryside is a facade of tranquility as you drive out of town. Lazy clouds drift across a deep blue sky. Fields of tall grass bend slightly to the whoosh of a light breeze.
For the highway signs guiding you to your destination, however, the picture is hardly peaceful. In fact, says Sougata Roy, the signs and the structures that hold them up are doing quiet battle with that deceptively light wind.
Storms, gusts of air, and wind-induced phenomena such as “galloping” and “vortex shedding” make life stressful for the slender poles and masts that support directional signs, lights and traffic signals.
Inside these structures, the welded connections joining poles and masts are subjected to repeated stresses that lead over time to deterioration in strength and fatigue cracking.
And when they collapse or fail, the structures make life hazardous for motorists and expensive for state highway departments.
Roy, a senior research scientist in Lehigh’s ATLSS (Advanced Technology for Large Structural Systems) Center, takes an avid interest in the sign structures. He is developing specifications for the design and fabrication of new sign and mast structures and for the retrofitting, or repair, of existing structures. In a four-year, $900,000 project, Roy and his colleagues and graduate students are testing the welded connection details, or joints, in cantilevered signs, high-level lighting and traffic-signal structures that are used in all 50 states.
The project is funded by the American Association of State Highway and Transportation Officials (AASHTO) and the Federal Highway Administration (FHWA) through the National Cooperative Highway Research Program (NCHRP). Lehigh was awarded the contract over six competing research groups. Roy, who developed the proposal and leads the project, is co-principal investigator. ATLSS director Richard Sause is principal investigator.
The researchers are subjecting the sign, signal and high-level lighting structures to fatigue tests in the ATLSS Center and in Fritz Lab, whose facilities for testing full-scale structures are among the largest in the world.
They are also developing mathematical models of the structures and their welded connections, and running simulations utilizing Lehigh’s high-performance computing (HPC) facilities.The models are based on finite element analysis, a technique for solving differential equations numerically in which researchers tackle a large problem by solving small segments of it and connecting these solutions over the entire domain.
The lab tests and computational simulations complement each other, says Roy.
“Because lab tests are expensive and time-consuming, you can do only a limited number of them. Simulation enables you to extend your findings, but it goes only so far. Even the most sophisticated calculations will not produce the right result until you can validate your mathematical model with experimental results.
“You can then use mathematical modeling to project the results of lab tests to a similar, but distinct, set of tests. As more experimental data comes in, you refine your model to improve its predictions.”
Flexibility at a cost
The poles and masts that support highway sign, signal and high-level lighting structures are thin and light. These qualities promote flexibility but also cause the structures to rock continuously from every tiny push from the wind. A steady light breeze can magnify this rocking, depending on the structures’ natural frequency of vibration, and cause two unstable phenomena. Galloping occurs when a sign bobs up and down, while vortex shedding occurs when a mast undulates back and forth. These movements of pole and mast are self-sustaining.
“Because they are so light and lack structural damping, sign, signal and high-mast structures shake easily and keep on shaking,” says Roy. “Even a 10- to 20-mile-per-hour wind can cause vigorous shaking.”
The consequence of all these phenomena, says Roy, is fatigue, or the degradation of material strength as the cumulative effect of repeated stresses. A further result is fatigue fracture, or the fatigue-induced failure of a structural system.
Fatigue fracture in sign, signal and high-mast structures is a growing concern. In 1990, a motorist was killed in Michigan when a falling sign crashed through his car’s windshield. A storm in 2005 in Colorado caused more than 20 signal structures to collapse. Iowa estimates that half of its high-mast structure inventory is cracked.
“Before the 1990s, wind-induced fatigue was not explicitly considered when these sign structures were designed,” says Roy. “It was only after the failures of the 1990s that fatigue design of the structures got the attention it deserved.”
Degrees of freedom
Roy’s group is focusing primarily on the two locations in a structure where fatigue fracture is most often reported – the connections joining arm and pole, and pole and foundation plate.
“Welded connections in general are more fatigue prone, so they are the focus of our research,” Roy says. “The stress at the tube-to-plate transition increases rapidly and makes the welded connection more susceptible to fatigue cracking.”
The researchers hope to optimize the cross-sectional shape and size of several variations of these connections so they perform more robustly in service.
“We will change the relative stiffness in these connections so they sustain less damage from fatigue,” says Roy.
In the last 15 years or so, says Roy, advances in computational capabilities have enabled structural engineers to analyze and simulate more realistically the response of an entire structure and design to different types of failure modes. Advanced finite-element analysis has also enabled researchers refine models, predict structural behavior more accurately, and optimize designs.
Roy makes use of two HPC architectures in Lehigh’s Computing Center. The SMP (Symmetrical Multi-Processor) computing facility contains a large number of processors and a shared memory in a single machine. The 40-machine Beowulf cluster enables parallel processing of full-scale structural systems by parceling parts of one large analysis to many different machines.
Roy and his group expect to conduct 100 to 110 physical tests on a total of 80 sign, signal and luminaire structures. So far, they have tested 50 specimens. They have also conducted simulations of 18,000 mathematical models using Abaqus, a suite of finite-element analysis software programs. Each model contains about 40,000 degrees of freedom, a term that refers to the number of equations the computer must solve at each iteration.
“Lehigh has historically excelled at developing new ways of doing fatigue and fracture research,” says Roy. “This project has been very computer-intensive. That helps us create a good balance – a research model with experiments substantiated by analytical work. Ultimately, it enables us to expand the scope of our studies.
“Our goal is to define the infinite life threshold of the various critical connections in these structures so that we know that no fatigue-induced failure will occur during their design lifetime. This will enable future standards to suppress the fatigue-fracture limit state of failure for these structures. This will free designers to worry about other predominant failure modes, such as vehicle collisions and corrosion.”
Other members of the project are graduate students Yeun Chul Park and Reilly Thompson; John W. Fisher and Ben T. Yen, professors emeritus of civil engineering; and Eric Kauffman, manager of manager, materials testing and characterization at ATLSS.