Euro-SiBRAM’2002 Prague, June 24 to 26, 2002, Czech Republic
Session 8 – Comment
Codes, Databases, Software and Application of Internet
Discussions and debates continue in the engineering community about working stress design, limit-states design and probability based design.
One of the first fully limit-states design code was introduced in the Ontario Highway Bridge Design Code in 1979. However, these codes exist only for short and medium span bridges. Although the Canadian Highway Bridge Design Code, in the year 2000, included a brief section on the design of long-span bridges, there is no comprehensive code specifically devoted for long-span bridges. As such, we must seize the opportunity to introduce reliability-based design philosophy in our approach for long-span bridge design. Benefits include consistency and uniformity in safety level for all designed bridge structures. Based on a well-calibrated limit-states design code, a healthy balance between safety and cost can be met. Due to the uniqueness of the static dynamic material properties and the probabilistic characteristics of individual bridges, even with the use of the limit-states design code, some may still be under-designed and others over-designed. For this reason, it is best to use the probability based method itself, rather than using only the code.
The probability-based method is a more reliable realistic approach for design, particularly in the evaluation of existing structures. However, one of the major problems in using the probability based design is that, in some cases, there is a lack of reliable data. Thus, we need to treat the available data in an intelligent and cautious manner. We need to review the possibilities that we have with data management. One needs to assess as much data as possible and then proceed with a level of flexibility. MonteCarlo is one of the available tools with which to do this. Although we have entered the Internet age, in practical design, we still use the concepts introduced in pre-computer design. Several good programs have been developed for use in computer-based probabilistic design analysis. We need to provide these creative tools to the designers and promote the use of the same. Designers need to move from an analytical to probabilistic way of thinking.
In North America, there is already a move towards performance-based code and objective-based code. With rapid development occurring in the IT industry, we must seize the opportunity to provide input to this change and move towards a uniform direction of code development. Options include: working stress design; full limit-states design with partial factors; full limit-states design with alternatives of reliability-based design philosophies; or full reliability-based design philosophy.
My suggestion is that now is the time to introduce the full probability-based design method into future design codes including both the design and evaluation of existing structures. However, one must keep in mind that reliable data and distribution patterns must be developed and used in the development of these codes.