Value in steel design

    Figure 1: Labor productivity index for U.S. construction industry and all non-farm industries (Source: Paul Teicholz, founding director of the Center for Integrated Facility Engineering at Stanford University)

    The worldwide economic downturn places greater demands on the design and construction industries as banks and other lending institutions require enhanced price definition and cost control. Financing will distribute to projects that provide the greatest value for the lowest associated risk. Consequently, little tolerance for waste and inefficiency can be expected in the foreseeable future.

    As reported by Preston Haskell of the Haskell Company in his groundbreaking 2004 white paper, while productivity in non-farm industries has more than doubled in the United States since 1964, the construction industry has gained by less than half this trend despite tremendous technological advances. Labor productivity has actually declined in the construction industry during the same time period, according to Paul Teicholz, founding director of the Center for Integrated Facility Engineering at Stanford University; see Figure 1. For the structural steel industry, one major culprit is the inefficiency created by poor or incomplete connection design practice. In 1995, William Thornton, then chief engineer with Cives Steel Company, presented the T.R. Higgins Lecture at the American Institute of Steel Construction (AISC) National Steel Construction Conference, outlining many shortcomings with steel connection design that result in considerable cost growth to typical structural steel frames. Areas of concern included poorly defined load criteria for the design of simple shear connections, moment connections, and bracing connections. Thornton also attacked the methodology used to design bracing connections, and the use (or abuse) of stiffener and doubler plates in columns at moment connections when unnecessary. It can easily be argued that nearly all of Thornton’s concerns remain present 14 years later.

    Uneconomical or incomplete connection design information is predominantly the result of the outdated design-bid-build business model, which nearly guarantees a lack of communication between the design and construction teams until design is substantially complete. Without fabrication and erection preferences from the construction team, engineers are forced to produce “one-size-fits-all” designs. Some engineers do provide complete connection design details and schedules that allow for considerable flexibility in connection type as well as optimization to detailed minimum load criteria. However, developing such details and schedules involves significant time, which has unfortunately made this approach rare.

    Segregation between design and construction eliminates an effective check and balance between design decisions and true cost. Consequently, it becomes easier for parties to the contract to look out for their own interests to a greater extent than the value of the project. Many structural engineers have taken the mentality of expending the fewest number of engineering hours possible on each project. Understandably, such an approach has been taken both to maximize profitability and out of necessity. While engineers may be evaluated on many criteria, certainly one of primary consideration is fee. Unfortunately, many of the entities engaging engineering services, be it architectural firms or ultimately owners, have little understanding of the insignificance of the structural engineering fee when compared with the cost savings resulting from a complete and economical engineering design.

    A good example
    As an acute, but applicable, illustration, a peer review was recently performed on a group of three similar 180-foot-tall stair towers for an industrial project. The tower designs used vertical braces on four sides, primarily in two-story X configurations with pipe bracing members. Connection design work was delegated to the fabricator using bracing forces in terms of a percentage of brace capacity, and gravity beam end reactions in terms of 60 percent of the total uniform load-carrying capacity of the beams. Seismic design provisions and detailing did not apply. Creating a design model for one tower and verifying design criteria and member designs required approximately four hours. An additional six hours were spent performing the peer review.

    In the peer review, it was determined that the worst-case brace sizes, and thus connection forces, were used throughout the design. This resulted in the vast majority of the lateral system being significantly overdesigned for both strength and stiffness. Instead, single and double angle bracing members could economically be substituted for the pipe braces. This would allow for members to bolt directly to gusset plates in the field rather than shop connecting braces to connector plates, which were later field bolted to gusset plates. Additionally, single diagonal configurations could be used at the end bays in lieu of the two-story X configurations with no penalty to brace size.

    It was also determined that use of 60 percent of the total uniform load capacity for the gravity beams in the end bays resulted in reactions approximately 650 percent of those determined from the actual analysis. Double angle connections, end plate connections, or single angle/plate connections with extensions below the beam were required to meet the overly conservative end reactions. Any coping of the beams would necessitate doubler plates to meet beam block shear, net shear, and cope bending limit states.

    The redesign allowed for significant cost savings and vast simplification of the connection details. Altering the brace types, reconfiguring the brace configurations in the end bays, and using the actual beam reactions for the gravity connection design eliminated 22 percent of the braces and approximately 170 associated field connections, as well as nearly 380 shop connections. Connection forces were greatly reduced such that more economical connection types could be used. Six additional hours of engineering time resulted in potential savings of nearly half a million dollars!

    While this illustration is influenced by the significant amount of vertical bracing, the design decisions and load criteria are frequent to designs for many commercial and industrial projects. The overdesign that seems to have become common place in the building industry is foreign to other industries. For example, manufacturing engineers are historically measured on their ability to refine designs to the minimum requirements of performance criteria while adhering to life safety requirements.

    Puma Steel. Figure 2: Example tower brace connection comparison: simple changes in design and detailing can save significant construction costs.

    Why this is happening
    Whether due to schedule, fee, building code complexity, or other factors, structural design drawings today appear to incorporate more and more typical details that conservatively cover a greater number of conditions with less engineering effort. For example, except in high seismic situations, seldom is the full development of member or component strength necessary. However, complete-joint-penetration groove welds or requirements for connections to develop the strength of main members are frequent byproducts of reduced engineering time. Designers seem to all but ignore the fact that these types of connections are the most expensive to provide.

    In addition to the negative impact reduced engineering effort has on project economics, the potential development of a dependence on the typical details produced is an inherent concern for engineering firms. As design manuals and codes continue to thicken with each edition, it becomes increasingly difficult, if not impossible, for an individual to be an expert on a large breadth of structural engineering topics, particularly in all materials. An increased tendency toward less flexible, typical details, as well as an associated resistance to deviating from those details, is understandable. However, the competency to perform, or even review, connection designs that fall outside of a firm’s typical standards is ultimately lessened.

    It is understood that, given the opportunity, engineers often receive little help regarding connection preferences from less sophisticated fabricators. However, preferences go far deeper than simply welded or bolted connection with more sophisticated fabricators, particularly as structures gain in complexity. Some very large fabricators may even vary preferences from project to project depending on the equipment and labor resources available at a given time. Therefore, lack of flexibility in connection type can be of considerable detriment to project cost and schedule.

    Puma Steel. Figure 3: Typical non-seismic moment connection comparison.

    In an effort to reduce engineering hours while at the same time attempting to provide flexibility in connection type, some engineering firms, particularly in the Eastern half of the United States, delegate connection design to the fabricator. However, many pitfalls often exist with this approach including incomplete and/or conservative design criteria, limited flexibility in connection type, incompatible or inefficient member selection in relation to connections, and resistance to deviation from typical standards as mentioned previously. In some cases, even when connections meet the criteria outlined in the construction documents, design criteria or connection components are modified during the approval process, ultimately adding cost to the project. While the load paths through connections may be complex, if the engineer of record can develop criteria to review a connection, then it is reasonable that such criteria can and should be adequately conveyed to the fabricator within the construction documents. The inability to do so can again be attributed to the minimal time put toward the project as rationalized previously.

    What is being done about it?
    For the reasons stated above, steel connection design appears to be developing slowly into a lost art in consulting engineering. Out of necessity, or as part of business strategy, many sophisticated fabricators employ or retain engineers with expertise in connection design. However, because of difficultly in coordinating economical connections with member types and sizes, and with extracting reasonable and adequate load information, some fabricators are going further by engaging consistent engineering teams that design to specific fabrication and erection standards. Design-build and design-assist are early involvement models that help to facilitate this approach.

    Figure 4: Historic material industry market share in the United Kingdom (Source: Richard B. Barrett, president of the British Constructional Steelwork Association)

    Historical trends in the United Kingdom support the effectiveness of these early involvement models within the steel industry. In a 2008 presentation, Richard Barrett, president of the British Construction Steelwork Association, indicated steel market share on multi-story, non-residential structures has increased from 33 percent to 71 percent since 1980, essentially switching places with concrete; see Figure 4. Within the same duration, performance-based design-build delivery of structural steel frames has grown from almost non-existent to 40 percent of the construction delivery market share.

    The steel industry should not tolerate decreased market share in the United States simply because of poor structural design. Early fabricator involvement in projects and increased engineering time and fee with feedback from the fabricator are feasible solutions to the problem. Design-build, design-assist, and integrated project delivery are effective methods of facilitating these solutions. Handled properly, each has the potential to reduce project costs and add overall project value.

    This is a call to arms for the structural engineering community — be the frontrunners in an effort to educate owners, developers, architects, and construction managers. Eliminate those who do not understand or appreciate the value an integrated, quality structural design brings to every project. Those who fail to work toward enhanced value could find themselves among those eliminated.

    Patrick S. McManus, S.E., was a consulting engineer with Martin/Martin, Inc., before taking the position of head of engineering and detailing at Puma Steel in Cheyenne, Wyo., in 2006. He can be reached at Rex I. Lewis is president of Puma Steel and has more than 40 years of experience in the steel industry. He is heavily involved with AISC and can be reached at Both authors have served terms as the Wyoming director-at-large for the Design-Build Institute of America.