They are as common as billboards, as important to motorists' safety as the seat belts they wear, and as innocuous as the mile markers and reflectors that line our highways. Uninspiring, yes, but without guardrail systems, the carnage on the nation's roadways would be even more gruesome than the 42,000-plus victims that automobile crashes claim each year.

For that reason, the Federal Highway Administration (FHWA), in conjunction with a number of state departments of transportation, is working to improve guardrail systems as part of a much larger highway safety campaign being orchestrated by the U.S. Department of Transportation (DOT). Safety, to paraphrase Secretary of Transportation Rodney Slater, is priority one, and guardrails save lives.

As part of its safety strategy, FHWA recently enacted more stringent guidelines governing crash barriers along federally funded highways. Crashes involving vehicles that run off the road account for roughly one-third of those 42,000 annual deaths, and DOT is committed to reducing that figure.

FHWA is particularly concerned about the guardrail systems that protect the opening between side-by-side bridges - openings that often lead to depressed roadways running perpendicular to the bridges above. Such openings are euphemistically known as "elephant traps," but in reality, falling into the elephant trap could lead to a catastrophic crash with loss of life for the occupants of the vehicle and for any unsuspecting motorists below. One crash barrier that is used to prevent out-of-control vehicles from falling into the elephant trap is known as the "bullnose" guardrail system, so-named because of its distinctive U-shaped design.

Historically, three systems have been used to protect against median hazards: crash cushions, open guardrails, and bullnose guardrails (closed guardrail envelopes). Crash cushions, which involve rigid barriers with cushions on each end, are effective but expensive. Open guardrail systems are less expensive, but because of their length, they create their own hazards for motorists. Bullnose systems wrap the guardrail completely around the hazard and are the least expensive of the three designs.

State DOTs are particularly fond of bullnose systems because of their low cost and strong safety potential. But to use bullnose systems, engineers need to bring them into compliance with a provision in the National Cooperative Highway Research Program (NCHRP) Report No. 350 (Recommended Procedures for the Safety Performance Evaluation of Highway Features) that requires crash tests to include light trucks (up to 2,000 kilograms or approximately 4,400 pounds). NCHRP Report 350 has been adopted by FHWA as a required standard for the crashworthy acceptance of roadside safety appurtenances used on federally funded roadways.

This requirement is a reflection of the enormous growth in the popularity of light trucks and sport utility vehicles, which together now outsell passenger cars. No longer is it sufficient for a bullnose guardrail be able to withstand a head-on collision with a large automobile. It now must be able to accommodate the full impact of a two-ton behemoth charging into it at 100 kilometers per hour (62 miles per hour).

Indeed, earlier guardrail designs accommodated the vehicles - primarily large two-door and four-door passenger sedans - that the automobile industry produced before the Corporate Average Fuel Efficiency (CAFE) standards were introduced following the energy shortages of the 1970s. And after the larger passenger sedans were removed from the fleet, barriers were built to handle the smaller and lighter cars. But today's light trucks and sport utility vehicles are heavier and higher off the ground than cars. All of which is to say that federal and state highway engineers were forced to rework the bullnose design to reflect the changing consumer tastes in automobiles.

In 1997, the Midwest Roadside Safety Facility (MwRSF) began a multiyear research project with a goal of developing an NCHRP Report 350-compliant bullnose guardrail system. Ideally, the bullnose guardrail either deflects passenger vehicles parallel to the roadway, or in the event of a head-on collision, it "captures" the vehicle much like a safety net. Unfortunately, the first two full-scale crash tests - one involving a small passenger car and the other a light truck - resulted in mixed results. While the bullnose system successfully captured the automobile, the light truck plunged through the guardrail.

In the past, such failures often placed an enormous financial strain on crash-test programs. Corrections were made based on observational data; more tests were run with more cars crashed into guardrails; and additional observations and corrections were made. If the engineers were lucky, their suspicions were borne out in subsequent tests, and the problems were solved. If not, more and more tests were conducted until the correct solution was found. And because FHWA protocol requires a minimum of seven successful tests for certification of a bullnose guardrail system and the cost of each test ranges from $15,000 to $25,000, crash-test budgets could quickly escalate.

The second phase of the MwRSF project was started earlier this year and also resulted in a failed crash test - again involving a head-on collision with a light truck. However, this time data were gathered and put to use in an innovative new system that fundamentally changes the way future crash tests will be analyzed and performed.

Enter LS-DYNA, a computer analysis system that employs nonlinear impact finite element code to reconstruct the test in the virtual world of digital electronics. In other words, engineers plug in the data from the real-world crash tests and then simulate the tests over and over again in an attempt to determine what failed, why it failed, and how best to correct the problems.

The brainchild of John Hallquist, who created its predecessor, DYNA3D, while at Lawrence Livermore National Laboratory in the 1970s, LS-DYNA allows crash-test engineers to digitally model real-world tests without ever leaving their computers. Originally used as a tool for simulating underground nuclear tests and to determine the vulnerability of underground bunkers to nuclear missile strikes, DYNA3D proved to be a versatile tool for other impact simulations, albeit those with far less destructive capacity than nuclear blasts.

For several months, engineers at the University of Nebraska's Center of Excellence used LS-DYNA to run simulations designed to strengthen bullnose guardrails to withstand a direct assault by a light truck. (There are three others centers of excellence at the universities of Cincinnati and Iowa and at Texas A&M University.) A major advantage of LS-DYNA is that it enables researchers to run less expensive crash tests on the computers, and then, they can go out and rerun the test in the field. With some real-world crash-test data, the software can be very successful and allows researchers to make lots of changes that would be very difficult and expensive to do in field tests.

In some ways, the use of computer technology to model crash tests is simply mirroring whatautomobile makers have been doing in recent years. When Detroit comes out with a new car design, it spends a long time in the virtual mode before the car is actually built, and LS-DYNA is the automobile industry's code of choice for crash-test work.

The benefits of using LS-DYNA aren't just financial although that's obviously a very important consideration in these days of austere government budgets. The system permits engineers to capture data they might ordinarily miss. No matter how many cameras you have recording a crash test, you can't possibly capture all the relevant and essential data.

LS-DYNA also permits engineers to root out specific problems on which it would be difficult to concentrate in field tests. Like peeling away the layers of an onion, LS-DYNA affords engineers new perspectives into the myriad forces at work in a crash test. With this software, engineers can stop the simulation in any frame, look at it from any perspective, and even remove parts of the car to get a better look. For example, for a wheel-snagging problem, all of the car except for the wheel could be removed, and engineers could focus on the wheel during the test. You can't do that in the real world.

LS-DYNA permits engineers to move beyond the old-school practice of concocting remedies by relying primarily on experience and first-hand evidence in analyzing a crash test. The software provides an in-depth analysis of the forces at work, and with the software, engineers are starting to predict the results ahead of time based on the computer tests.

Indeed, once the design concept was verified by LS-DYNA, subsequent field tests using a light truck were successful. During a typical head-on collision with a bullnose guardrail, the rail - a thrie beam having three humps - is bent inward to capture the vehicle and slow its progress. The bending strength of the guardrail's thrie beam and the attendant breakage of wooden posts at the ground line is sufficiently strong to cause rapid, yet safe, deceleration of the light truck. In earlier tests, enormous stresses in the top two humps of the thrie beam caused the guardrail to rupture, permitting the truck to pass through.

To simulate a new test, the University of Nebraska-Lincoln (UNL) downloaded a crash test model of a C2500 pickup truck from the National Crash Analysis Center (NCAC) at The George Washington University-Virginia Campus in Ashburn, Va., to use in the simulation. Incorporating about 10,000 "elements" to be analyzed in the simulation, the model enabled engineers to study the various stresses and other dynamics that occurred during the collision with the bullnose. After several attempts, the failed field test was successfully replicated with LS-DYNA, enabling engineers to conduct a kind of "lessons learned" analysis of the test.

The simulation indicated that the failure of the bullnose could be traced back to long longitudinal slots cut into the valleys of the triple-hump thrie beam. These slots were purposely introduced to "capture" the front ends of vehicles of varying sizes - from small cars to light trucks. Without the slots, small cars tend to "under ride" or submarine under the beam, and larger (and higher off the ground) light trucks tend to "override" or run over the beam. During field testing, when the light truck collided head-on into the beam, the beam separated about the slots as intended into three "ribbons," each wrapping around the front of the truck and preventing override. However, the lower ribbon, as it separated downward, contacted the rotating front tires of the truck and was subsequently run over by the truck. With only the upper two ribbons still available for stopping the light truck, the beam failed due to overload.

After some tinkering, it was thought that reinforcing the thrie beam with two 19-mm cables would sufficiently strengthen the bullnose to withstand the truck collision without endangering the vehicle's occupants. A simulation of the cable-reinforced design showed that the system would successfully stop the truck.

More importantly, the subsequent field test confirmed the LS-DYNA test results: the cable-reinforced bullnose guardrail system not only withstood the truck collision, but it also provided excellent protection for the truck's occupants. The field test substantiated LS-DYNA's predictions.

One of the few drawbacks to LS-DYNA is the time required to get the system ready. While a real-world field test can be set up and run in two weeks, the initial time for a good simulation can take six to 12 months. However, it is important to note that those times drop substantially once the simulation is set up. Subsequent tests can be set up in about a month.

The current program is being supported by FHWA and eight state DOTs. FHWA is paying for the LS-DYNA data analysis effort, and the states are footing the bill for the actual crash tests. The recent test of the bullnose system with a light truck was successful, and some of the credit for that success goes to the software.

This is the first time a major real-world problem involving crash barriers has been solved using simulation. The use of LS-DYNA to design an acceptable bullnose guardrail system is a real success story, and given the promise of simulation, it is certainly only the first of many such successes.

For additional information about LS-DYNA and its role in supporting a new bullnose guardrail design, contact Martin Hargrave at (202) 493-3311 or e-mail (martin.hargrave@fhwa.dot.gov) or John Reid at (402) 472-3084 or e-mail(jreid@unl.edu).

Dr. John D. Reid is an assistant professor of mechanical engineering at the University of Nebraska-Lincoln (UNL). He is also the director of the DYNA3D Center of Excellence at UNL. Before joining the faculty at UNL in 1993, he worked at General Motors Corp. for eight years - the last three in safety and crashworthiness. He received his bachelor's degree, master's degree, and doctorate in mechanical engineering from Michigan State University.

Martin W. Hargrave is a research engineer in FHWA's Safety Design Division in the Office of Safety and Traffic Operations Research and Development. He conducts and manages research associated with FHWA's DYNA3D finite element research program. Before joining FHWA in 1979, he worked for 17 years in varied engineering assignments for private sector companies. He received a bachelor's degree in mechanical engineering from the University of Alabama, a master's in engineering from Pennsylvania State University, and a master's in civil engineering from The Catholic University of America.

Doug Rekenthaler Jr. is a freelance writer and editor. His experiences as a writer and editor include cub reporter covering Capitol Hill and Pentagon news beats; managing editor responsible for 12 newsletters that covered a wide array of communications technologies; founder of the multimedia industry's first daily fax news service; and corporate communications manager for America Online Inc., the largest commercial online service in the world.