Free Software Predicts How And When Steel Beams Will Buckle

The computer tool could become increasingly important as construction rules change to accommodate innovations in structural design "To keep costs down, many people today are looking for maximum strength with the minimum amount of materials," says Schafer, an assistant professor in the Department of Civil Engineering. "So very often, you end up with what we call thin-walled structures. But instability or buckling can cause these structures to collapse or 'fail'. If you press down on the top of a plastic straw, there's basically only one way it can buckle. With more complicated thin-walled structures, there are many more ways this can happen. Engineers need to predict how and when this will happen, so they can design buildings that won't buckle under a particular load. This software is a tool that does just that."

Schafer's software, called CUFSM, is available for free downloading on his Web site: http://www.ce.jhu.edu/bschafer.

The civil engineering researcher recently updated the program to provide a far more user-friendly interface. It is available in a stand-alone version for users of Windows and in another version that is compatible with the popular MatLab software, which is available on virtually all computer platforms. The Web site also features tutorials and examples to help designers and students learn to use CUFSM.

The computer program is an extension of research Schafer began while earning his doctorate in structural engineering from Cornell University in 1997. (The software's name is short for Cornell University Finite Strip Method because Schafer developed his first version of the program at that school.) At Johns Hopkins, Schafer's research focuses on the behavior and design of cold-formed steel sheet metal bent into various shapes. Cold-formed steel has become a popular structural alternative to timber and is used extensively in low-rise buildings. It is utilized in a variety of applications, supporting floors, roofs and walls in industrial, commercial and residential buildings. The more general category of thin-walled structures can range from the previously mentioned industrial and residential buildings to box girder bridges, ship hulls and aircraft skins, as well as buried structures such as tanks, pipes and culverts. "The idea is to provide maximum strength with the minimum material," Schafer says. "The design is critical. Even small changes in the geometry can affect the strength of the structure and how it might buckle. Placing little bumps or folds in the members, using corrugated versus flat sheet metal -- all of these things can make a difference."

To test how such changes in geometry and materials will affect the stability of a thin-walled structure, an engineer can enter this data in the CUFSM software to determine how much compression and/or bending the structure can tolerate before it buckles. In this way, the sturdiest and most cost-effective design can be developed long before the structure is built.

Currently, civil engineers must adhere to rigid building codes that severely limit their design options, Schafer says. These rules were adopted to ensure that structures are safe and to maintain a level playing field among competitors. "These highly prescriptive building codes accomplished that, but they also took away the opportunity for structural innovation," Schafer says. "My software can help bring it back by giving engineers a way to predict buckling for structures that don't fall within the rules."

The Johns Hopkins researcher says the construction industry is already working toward revised guidelines that will allow new shapes of strong yet thin steel beams to help reduce costs. "Knowing this is going to be allowed is encouraging people to manufacture more innovative cross-section pieces," Schafer says. "This modeling technique will make a big difference to the industry. It fundamentally changes what you can design."

To prepare the next generation of structural designers for this change and to teach them how buckling can occur, Schafer also uses the software as a teaching tool in his undergraduate civil engineering classes at Johns Hopkins. "This is how engineering design will be done in the very near future," he says.

Schafer's research has been funded by the American Iron and Steel Institute and the Metal Building Manufacturers Association.