News Detail

Judgment in Reinforced Concrete Analysis and Design

10/1/2008

The ultimate role of analysis is to forecast the behavior of an as-yet unbuilt structure to be certain that it will be able to comply with preestablished performance criteria during its life span. These performance criteria include strength, stiffness to account for unwanted deflections and excessive slenderness, durability, serviceability, and others. Even though it sounds simple, the complexity involved in meeting the performance criteria could be considerable even for a simple structure. To make this endeavor feasible, numerous simplifications must be made, both in the definition of what is required of the structure, generally expressed in terms of strength and deflections, and in the description of the behavior of the structure through analysis. This is especially true for the design of reinforced concrete structures and the reason why engineering judgment and experience are required.

We are witnessing tremendous advances in the field of structural engineering that have caused a reduction of safety factors compared with those used before the advent of the computer, increasing the designer’s responsibility. Meanwhile, the time allocated for structural design during the planning stage of a project has been substantially reduced in recent years. The use of fast-track project management schemes has subtracted even more of the available time to ponder and explore better structural solutions.

The engineer’s obligation is to understand the methodologies employed both in analysis and design, and to be aware of their backgrounds and limitations. Building codes are increasing in complexity but are not a replacement for engineering judgment and experience. The fact that codes cover more and more material is only a reflection of advances in structural engineering. All this points to the fact that nowadays the engineering judgment required from the structural designer is probably greater than at any time in the past. The diversity of tools at his or her disposal and the accumulated knowledge obtained through research—of which engineers may not always be aware—have accentuated the problem.

The design process has been transformed by computers to a point where the structural engineer can proceed from the analysis stage to structural drawings without having seen, or studied, the results of the intermediate design stages. The main problem brought on by this possibility is the lack of a final stage of "evaluation of expected behavior" of the structure. If a peer review of the design is required, this should be the major concern of the reviewer.

When the diverse loads routinely used today come into play, it is evident that the description of the behavior under the effect of any of them must be based on corresponding modeling parameters. The main conclusion is that no single set of modeling parameters in a structural analysis can describe the behavior of the structure under the diverse effects that must be taken into account in a modern design. A very good example is seismic design for which model codes require analysis using cracked sections but do not define the stage at which the cracking must be evaluated (first cracking, yield, post-yield). ACI 318-08 clarifies this issue for the first time through new Section 8.8, Effective stiffness to determine lateral deflection. The amount and pattern of cracking for gravity loads are different from that observed in a structure subjected to lateral loads and depends on how the stage under evaluation was reached. It is evident that engineering judgment comes into play when assigning realistic stiffness values for checking whether gravity deflections meet allowable limits.

The primary drawback of automatic analysis-design procedures is that the engineer is generally satisfied using a single analysis to take all effects into account. This means that the description, through analysis, of the behavior of the structure will be good for some effects and deficient for others. If actual structural performance evaluation is the goal, then the results of the single modeling parameters analysis are just the starting point, and not the final result.

In 1935, Hardy Cross appropriately concluded in "The Relation of Analysis to Structural Design" (Proceedings, American Society of Civil Engineers, V. 61, Oct. 1935):
"It may be well now to divert the attention of structural designers from the endless elaboration of analytical technique to the more important matter of interpretation of analyses."

He was right 73 years ago and he is still right!

Luis E. García
American Concrete Institute
luis.garcia@concrete.org

Back to Memo List