Mr. Mladjov’s reflections on the AASHTO LRFD Code are not consistent with the basic philosophy and foundation of the modern AASHTO Code. Load and Resistance factors are not safety factors. His translation of load and resistance factors to the abandoned safety factor concept is a throwback to older codes – codes that were shown to produce highly variable reliability that resulted alternatively in wasteful and questionable designs.
In the instance of seismic design, his application of additional load factors to a displacement based design is not relevant to the underlying ductility design basis. And in the general case for AASHTO as it is applied across the US, target reliability for seismic design outside of high seismic regions often does not warrant the same displacement design approach as that in California. In much of the US, force-based design may be both sufficient and consistent with the overall target reliability assigned for bridge structures.
While the target reliability for the Code is generally assumed to be a beta of 3.5 (failure of ~1 in 5000), actual targets vary from about 2.5 (~1 in 200) to 4 (~1 in 30,000) or more. There are both practical and philosophical reasons for the range, many of which relate to historical performance. Designing a foundation slope in front of a bridge might not be practical for a beta of 3.5, and designing a non-redundant connection is generally not endorsed at 3.5. Importance factors are currently discretionary and warrant attention. However, confusing historical ‘safety factors’ with importance factors can result in reliability levels that are not reasonable.
While redundancy and ductility are important in establishing reliability, mobilizing these attributes in contemporary design does not lend itself to the application of arbitrary load factors, either for reliability or economy owing to the variety of structural systems in use today. Designing to unwarranted levels of conservatism might provide engineers emotional comfort, but it does not serve society well. For every excessively conservative major bridge design there will be another bridge never built, or worse yet, perhaps a dozen old bridges that do not get retrofitted.
There are no zero-risk strategies. LRFD is a framework for deliberate reliability-based design. Just as planes must be safely designed with an efficiency that allows flight, bridges need to be safely designed with efficiency to optimize our resources. If you assume your life span is similar in duration to that of a new bridge, you are far more likely to meet your end by just driving your car, without crossing a bridge at all. While one might argue that the variances assumed for certain loads or resistance warrant discussion when applying the Code, validation through adherence to the LRFD principles is important in order to achieve both safe and economical outcomes.
David Goodyear, P.E., S.E. NAE
Author’s Response
I appreciate Mr. Goodyear’s comments as it allows me to clarify some of the article’s statements. Contrary to the claim that the article’s positions are not consistent with the underlying philosophy of the modern AASHTO Code and that the term “safety factors” is not used correctly, these positions are founded upon the most recent AASHTO code (AASHTO 8th Edition, September 2017) and its basic LRFD method. The article follows the LRFD structural design philosophy to maintain the factored resistance (the strength of the entire structure and all its elements) above the maximum demand from the worst possible combination of loads on the structure. The ratio of the strength to the demand represents the safety of the structure, where the nominal resistance is reduced by multiplying factors < 1.0, while the loads are increased by multiplying factors >1.0. These multipliers are the safety factors prescribed by the design codes and specifications for both building and bridge structures. Accordingly, the article refers to all strength-reducing or load-demand-increasing factors as safety factors, for simplicity. The claim that the article is inconsistent with the AASHTO (i.e., LRFD) philosophy is unfounded.
The article did criticize the current AASHTO values for operational, redundancy, and ductility factors. It recommended enhanced values for operational and redundancy factors while suggesting that nonductile structures, elements, and connections should not be allowed in earthquake-prone areas. The commenter does not give an opinion on these recommendations, but states that “Designing to unwarranted levels of conservatism might provide engineers emotional comfort, but it does not serve society well.” In principle, I agree with prioritizing efficiency; however, there is nothing in the article that justifies the claim of promoting “excessively conservative design.” The code factors discussed in the article do not apply to structures designed according to good engineering practice with sufficient path of load distribution and without nonductile elements. Here, as in other publications, I strongly support efficient, economical design and construction. As I have written elsewhere, “With available funding that pales in comparison to the amount needed, engineers working on infrastructure-related projects have a professional obligation to produce high-efficiency projects to ensure maximum impact is obtained from the available funding… Today, given the extreme necessity for replacing or strengthening substandard bridges, the responsibility of engineers to design and build efficient, less expensive structures becomes as essential as life-safety requirements. An efficient project optimizes delivery resources, thus allowing for the repair or replacement of other substandard bridges that may have life-safety issues.”
A good illustration of the issues mentioned above is the recent replacement of the East Span of the San Francisco – Oakland Bay Bridge. This replacement resulted in record-high project costs, more than 14 years of construction (January 2002 to October 2016), and a substandard structure, vulnerable in the case of a serious earthquake. The omission of the operational, redundancy, and ductility factors, already part of the AASHTO code, did not help the responsible engineers and authorities reduce the cost, and the unfortunate design concept resulted in a bridge significantly more expensive and not as efficient as it should have been.
However, public safety and human life should be the highest priority. As the article states, “It is not acceptable that the current safety of some bridges on essential roads, carrying 200,000 to 300,000 vehicles per day, should be less than the safety of a two-story building with 30 to 40 occupants.” In response to the commenter’s statement: “For every excessively conservative major bridge design there will be another bridge never built, or worse yet, perhaps a dozen old bridges that do not get retrofitted,” has no place in a highly specific, technical professional discussion. Surely, such factors are in the realm of state and local governments’ budgets, assessment, and planning. Even so, engineers are achieving efficient and economic structures by using highly efficient design concepts and solutions, not by applying lower safety factors.
The commenter does not state an opinion on the essential proposal that State Departments of Transportation should amend their requirements to correspond to AASHTO’s specifications; not using lower loads or lower safety factors than those specified by AASHTO.
Roumen V. Mladjov, S.E., P.E.