Those who provide structural design for new building structures know well that the art of designing buildings is much more than computing stresses and strains for the structural members. Structural design includes integration of creative structural concepts into the architecture, coordination with mechanical systems, material selection, analysis, design checking, and documentation; as well as developing complete and coordinated construction documents. This has to be accomplished in concert with the architect and other designers on the team, often at a fast pace with design iterations and changes, and always with real-time understanding of how the design decisions affect the final project cost.

Early use of computers was generally limited to only a part of this process such as the analysis of large and/or complex structures to provide the engineer with member stresses and deformations. Today, there has been a proliferation of commercial computer software to assist the design engineer with various facets of the overall design process, including structural analysis, design, and code checks; system optimization; and document development, to name a few. Although each software application may be very comprehensive for its specific task, and may increase the productivity for that task, the engineer must integrate each separate application into a comprehensive design process.

The process, and in particular how we achieve quality design and design documents, presents the engineer with challenges when trying to incorporate disparate computer tools. Increasingly, software developers attempt to integrate more and more of the design process into one computer tool. This provides efficiency and simplification for the user, but also brings new challenges with respect to transparency, technical validation, staff training, and, ultimately, quality of design.

Market demand for commercial software

Given enough time and money, engineers today should be able to achieve excellent structural design without computers for the vast majority of building structures.

However, today’s market does not give the structural engineer that luxury. It is safe to say that computers and state-of-the-art design software is a necessity in the modern structural engineer’s office. Not because the complexities of the structure demand it, but because the market demands the productivity that such tools make possible. As code-required load and design rules get more and more complicated, we will need to increasingly rely on computers for productivity gains. Whereas our predecessors before computers needed practical means and methods to arrive at a design, we need similar practical means and methods to check our designs derived from computer software and to help us better understand the structure.

Consider the computational efforts needed to show compliance with modern building codes. Earlier codes had more simple methods for developing loads; fewer load combinations, and often, much simpler member stress checks. The complexities of modern codes lend support to the argument that current computer-aided design software is not only a means of achieving productivity. Design software is also a practical necessity to avoid computational errors given the context of structural design, which today includes complex codes, complex designs, and fast-track construction.

The impact of computers

Structural analysis computer programs have been in use, at least in academia, since the late 1960s. Through the 1970s and early ’80s, computer use by practitioners was generally limited to analysis of large and/or complex structures. Designers would develop a design using approximate analysis methods and preliminary design techniques. This preliminary structure was then analyzed in detail with the computer to obtain member-end forces and deformations. Verification included checking the reactions and external forces for statics, interpreting the deformed shape, and comparing results with the preliminary and approximate analysis. The designer then performed design checks on the members. If the designer deemed it necessary to change members’ sizes, he or she needed to use his or her understanding of structural analysis to determine if another computer run was necessary. Generally speaking, for the building designer, the computer programs of this era were much less a design tool than an analysis tool. Compared to today’s software, they were tedious and awkward to use with little in the way of graphical results. Although young, inexperienced engineers may have helped the designer with checking input and output, it was necessary that the computer analysis be managed by an experienced designer or analyst to avoid colossal wastes of time and money.

During the 1980s and into the early 1990s, scaled-down versions of these main-frame analysis programs became available for use on the personal computer (PC). Early PCs, by today’s standards, had extreme limitations of memory and processor speed. The analysis programs were still awkward to use with input and output files, and the DOS operating system. But the computer time was now free and the input files were often text files that the designer could modify with text editors on the computer. Some graphical displays were available, such as displaced shapes and mode shapes.

By the late 1980s and early 1990s, the use of the computer for analysis was no longer limited to the most complex structures. PCs were getting faster, were less expensive, and had more memory and computing power. There were still plenty of firms doing ordinary and small buildings without computer analysis, but firms that were using PCs for analysis were now using them for most structures as a productivity tool. Many firms were inventing ways to post-process the output files for convenient use with spreadsheets that did the design checks. With a PC on the desk of every designer, the role of a senior, experienced designer overseeing the use of computer analysis diminished in many firms. The "penalty" for computer model mistakes by inexperienced designers no longer included computer time and computer money.

Through the 1990s, PCs continued to get faster with leaps in available memory, while also becoming less expensive. With the introduction of the graphical user interface of Microsoft Windows, designers now had powerful graphic tools to generate, display, and check the computer models. At the same time, there were rapid advancements in a second generation of computer-aided design and drafting software. By 2000, with the Internet going mainstream, e-mail and intra-office networks became commonplace, and there was a PC on every person’s desk in every engineering firm. By then, even the most sophisticated general analysis software was configured for the PC and there was a proliferation of special-use design software for the structural engineer. Throughout the design process, the designer used various software packages in an ad hoc manner mixing in hand calculations here and there. It was very typical, for example, that one program was used for frame analysis, another for floor framing, another for checking columns, and yet another for analysis and/or design of foundations.

Development of a third generation of structural engineering software began in earnest in the late 1990s and recently became conventional. This generation integrates analysis and design checks into the same model. In some instances, the designer does not even need to input preliminary sizes, as some software has design and optimization routines for member selection. Explicit input and output files are no longer used. The designer builds the model interactively with the software, visualizing the model in 3-D as it is built. The designer usually has many options for viewing the output, including member-by-member reports, color-coded member stress displays, animated mode shapes, and many others. Software is available that allows the designer to model all of the structural systems integrating the analysis and design of the floor systems, the columns, the lateral load systems, and the foundations. Consequently, structural engineering firms have seen huge productivity gains with such software.

Market pressures on fees for building design are compelling building designers to take full advantage of productivity gains offered by the automation of today’s building design software. The capabilities of this software include automation of framing layout, generation of loads and code-specific load combinations, member selection and optimization, embedded checking for conformance with material design specifications, and generation of framing plans. It is now practical to provide complete framing plans for the entire building during schematic design and to explore alternate designs without severely impacting the design schedule. Value engineering ideas are tested and implemented late in the design phase or even during construction.

Software verification and transparency issues

It is my observation that concurrent with software improvements, many experienced engineers, who were designing structures in the 1970s and 1980s, have developed a concern that the software tools are in the hands of engineers lacking the understanding of the behavior of structures and the methods used by the software. The civil engineering profession has been concerned about the misuse of computer software since the early use of commercial software and much has been published about checking results.

Although the concerns are still appropriate today, if not more so, prior efforts were focused primarily on the analysis of structures with general-purpose analysis software. The recent, rapid advancements in the automation and integration of analysis, design, and drafting in special-purpose building design software has generated a need for us to re-examine how we ensure proper and responsible use of such software.

An issue facing the structural design profession is verification of this third generation of commercial design software. For most common building design projects, simple equilibrium checks, combined with sound engineering judgment and experience, are adequate to find deficiencies or errors in the software that may surface due to that project’s specific parameters. However, designers still rely considerably on the software doing what it claims to do without errors. Designers are operating under the assumption that any errors of significance to the design will be detected during their review and validation of the results. Thus, designers become comfortable with particular software over time as they use it, and verification of software is limited to validation of results on a project-by-project basis.

Generally speaking, the software’s code is not available to the public. Even if it were, most practitioners do not have the programming experience to make much use of it in regards to verification. What designers know about the operations and function of software is limited to what is published with the software and what they can infer from the available output reports. Designers should look for software that provides detailed explanations of operations and functions, as well as output reports that provide step-by-step member capacity development and code check results and not just a final stress or demand/capacity ratio. This allows the designer to penetrate the program and better understand the embedded design process.

There are few, if any, current standards or codified requirements for structural design software embraced by the structural design community. Organizations such as the Structural Engineering Institute (SEI), Council of American Structural Engineers (CASE), and National Council of Structural Engineers Association (NCSEA) could partner with leading structural design software companies to explore ideas about guidelines and/or standards for qualifying and verifying software claims on operations and functions. Perhaps our industry can develop a voluntary certification program where software can receive industry certification after thorough review by both design experts and computer programmers. In addition, perhaps software companies are certified as maintaining good practice with respect to quality assurance and documentation.

Impact on staff and staff training

Young designers today may not appreciate the process and time involved in the design of structural members without the use of computer software. Before computers, the experience and aptitude of the structural designer had a tremendous impact on the time and quality of design. The good designer knew the design codes so well that he could eliminate many load combinations and design checks that would not control the design. Experience also came into play in evaluating the impact of a proposed change by the architect, and, if necessary, into defining the scope of any re-analysis and re-checking of code provisions for the areas affected by the change. For most projects, the degree of structural optimization relied on the experience and judgment of the designer.

Most seasoned engineers have a sense of pride about their ability to check the computer results, if they have the skill to do so. Whether our young engineers currently have those skills, or will develop them, depends upon the effectiveness of training and on the effective use of commercial design software. We must ensure that the amazing productivity gains that current and future design software provide become a means to free the engineer from the drudgery of code checking, numerical accounting, and the anxieties of mathematical errors, and provide the engineer with a greater, not lesser, understanding of the structure and its performance.

Management of today’s design firms must address training issues in light of the automated nature of the software. For example, young designers do not become familiar with member design specifications by frequent use as was the case just 10 years ago. Skill sets needed for design with current software are much different than the skill sets needed when design was done by hand calculations. Designs can now be accomplished using automated software without the user necessarily being intimately familiar with the design specification and, perhaps more critically, without the user being familiar with the fundamental philosophies embedded in the design specification. When the design checks were performed by hand, practitioners read the design specification, codes, and user commentary to perform the calculations. In addition, industry material organizations such as the American Concrete Institute (ACI), Precast Concrete Institute (PCI), and American Institute of Steel Construction (AISC) provided design guidelines and sample solutions to aid and teach the designer. With today’s automated design checks, young engineers need explicit training on the design codes and standards as they may not become familiar with them through routinely performed hand calculations.

Future trends

The leading software vendors will continue to advance their structural design software. Most have, or are developing, a system or suite of software so that the designer can become familiar with one vendor’s tools and use them for various structural materials and structural systems. The vision of the vendors is to have interoperability within their suite of software. For example, the designer will model the entire building in a modeling module, and use other software modules from the vendor to design foundations, frames, shear walls, flat plates, connections, base plates, etc., without further input of loads or structural geometry. In addition, the vendors will move to integrate the drawing production process with their suite of software. The leading vendors already have tools, with varying degrees of sophistication, for exporting data from their software to CAD systems.

Building Information Modeling (BIM) software solutions are already developing relationships with structural analysis and design software to achieve interoperability with them. The vision is that the model developed in the BIM will include all pertinent engineering data. Data then flows from the BIM to the analysis and design software with results coming back to update the BIM. Proponents of the BIM concept see this new process as a means to reduce risk from the now-fragmented process.

Just as the structural design software became more intelligent and comprehensive, so too will BIM technology become more capable by incorporating or associating with analytical tools and other software, automating the design process, and even automating the design documentation process. No one knows the ultimate impact of BIM software on the professional services market, but it will certainly make issues of software verification and transparency more acute.

Conclusion

The first and second generations of structural software led to productivity gains in analysis tools, but had little impact on the overall design process, especially with respect to how the structural engineer interacted with the rest of the design team. The third generation we are using now, by integrating the analysis and design of building-whole systems, has had a major impact on the design process. It is debatable if structural engineers have yet fully adapted best-practice procedures for software that can almost fully automate the design and analysis of a building. And, as we move to the fourth generation of structural software that is tied into BIM technology that allows automatic generation of drawings and reports, we must define our roles and responsibilities for what well may be a paradigm shift in our industry.

James C. Parker, P.E., is a senior principal at Simpson Gumpertz & Heger Inc.—national engineering firm that designs, investigates, and rehabilitates structures and building enclosures. He is located in the firm’s Waltham, Mass., office and can be reached at 781-907-9000.

Photo caption: Engineers at Simpson Gumpertz & Heger Inc., used computers in the early 1970s to analyze large and complex structures.

Photo credit: Simpson Gumpertz & Heger Inc.