Half-baked designs seem especially prevalent in underground construction, as I know firsthand. During the past 45 years, my company, Tri-State Drilling, has drilled tens of thousands of shafts into the ground. We know that no matter how many test borings or investigations are done, you simply can't foresee the tremendous variety and bizarre combinations of variables in the subsurface.

Let me give you an example. We were asked to install drilled shaft foundations for a large power transmission tower in the Midwest. Drilled shafts are great foundations for these towers. They're relatively straightforward to design, and we've installed hundreds of them. But in this instance, the project engineer tried to optimize the design, mistakenly thinking that he could save some concrete and money by applying a little innovation. Instead of calling for straight vertical shafts, the design called for battered (inclined) ones with bell-shaped (flared) ends anchored in bedrock.

MINING NIGHTMARE. We warned the clients before we bid that the job would be tricky, but they asked us to give it a try anyway. Instead of a standard drilling job, the project turned into a mining nightmare. Drilling at an incline through rock required extremely heavy tools--so heavy that they tended to turn downward. Instead of a battered shaft, we got a hook-shaped foundation. It was also difficult to keep the belling tool upright. We ultimately creating a lopsided bell at the end of the hook. Consequently, the "better mousetrap" design was scrapped after the first shaft and completely redesigned. But still, the failed attempt at battered bells ultimately added 30% to the cost of the foundations.

I'm not against innovation. But in this case, the engineer went way beyond the limits of practical optimization and swung the pendulum to "the unconstructibles"--that assemblage of half-baked designs that we, as an industry, need to avoid. There are myriad ways to avoid unconstructibles. For example, the design-build method of procurement allows contractors more input than straight design-bid-build into constructibility decisions. But design-build is not appropriate for every job. Simple constructibility reviews are appropriate, however, throughout the design process. Input should come from the contractors who will build, or at least bid on, such projects.

But besides merely reviewing half-baked designs in feeble attempts to interject constructibility, we need to cure the problem once and for all, starting with fundamental engineering education. Colleges and universities already do a pretty good job of training engineers how to apply the proper equations. They teach engineers to apply the proper factors of safety to account for unknowns in material behavior, empirical constants, and other variables. What they do not teach is constructibility. Why? Engineering curricula are jammed with teaching numerical analytical methods, leaving little time for practical matters. Many professors themselves do not understand how civil structures are built, and therefore cannot teach constructibility.

To help compensate for such knowledge deficits, I'd like to propose creating and using "factors of constructibility"--not divisors to make designs even more conservative, but a new way of approaching design. The factors-of-constructibility curriculum could put flesh on the bones of engineering theory. It could place much more emphasis on case histories, in which students could learn valuable lessons from successful projects and learn how to avoid horror-story problems. This new curriculum could teach the "how" and "why" of civil construction, not just the "what." It could make constructibility a recognized and routine part of every design.
COOLEGE DRILL Educators learn about drilled foundations.

If faculty aren't prepared to teach constructibility, they need to look outside. Here's where the construction industry can step up to the plate. Routinely, contractors should give classroom or after-hours case-history presentations, invite students to their job sites, and provide curriculum resources to faculty. And other trade organizations should replicate Faculty Workshop 2000, the highly successful collaborative industry-academic program sponsored by adsc: The International Association of Foundation Drilling.

TEACH THE TEACHER. In a week-long workshop last July at Colorado State University, adsc provided 57 leading civil engineering educators with the knowledge and resources to teach the most up-to-date design and construction technologies for drilled shaft foundations and anchored earth-retention systems. This week, at adsc's equipment exposition and annual meeting in San Antonio, we will remind our members of the value of holding such teach-the-teacher workshops.

Half-baked designs now show up across civil, mechanical and electrical engineering projects. If we made constructibility reviews part of the design process, "unconstructibles" could be replaced by innovative, cost-effective techniques.

Jim Melcher is the chief operating officer of Tri-State Drilling, Inc. in Hamel, Minn., and the current president of Dallas-based ADSC: The International Association of Foundation Drilling. He may be e-mailed at jmelch@attglobal.net.