Size Effect on Mean Strength and Failure Risk of Concrete Structures: Its Introduction into Design Codes and Practice, and Scientific Basis

1 Abstract The project has two objectives: (1) to improve the reliability, durability and structural efficiency of infrastructure by transferring the results of previous basic research on fracture mechanics of size effects in concrete structures into: a) design practice, b) design codes, c) standardized materials testing, and d) commercial computer programs; and (2) research the probability structure of design code ensuring tolerable failure probability under 1 in a million. This is important for large-span box girders of prestressed concrete bridges and cable-stayed bridges, supporting piers, column footings and foundation plinths, as well as for road and airport pavements, earth-retaining walls of highways, dams (which often carry highways) and tunnel linings. The design practices for ductile failures are not affected, but those guarding against brittle or quasibrittle failures, which are currently still based on the plasticity-based theory of limit states, need to be updated for size effect according to fracture mechanics and random strength statistics. Furthermore, a standard for concrete fracture testing needs to be introduced. With only two exceptions, the commercial software for simulating concrete structures is generally not yet based on fracture mechanics and eschews the size effect. Although the basic theory (which earned the writer his election to NAS) is ripe, simple design formulas need to be developed, proper material testing procedures instituted, and simple ways of implementation in commercial computer codes proposed and promulgated. Moreover, the relevant committees of ACI, ASTM, RILEM and FIB, the specifications of which are generally followed by AASHTO, need to be persuaded. This is not only a scientific-technical problem but also an educational one, difficult and politically time-consuming. Previous ITI support has already led to several successes: 1) After years of argumentation, the report of RILEM Committee QFS (chaired by the writer) was finally approved and published in RILEM journal Materials and Structures. 2) The writers' criticism of unsafe extrapolation to large sizes (and a warning about potential legal liabilities for ACI) engendered sufficient concerns in ACI code committee 318 to cause rejection of an unrealistic solicited proposal by ACI Committee 445 regarding size effect in shear. 3) A competing (unsolicited) proposal by ACI Fracture Mechanics Committee 446, prepared by the writer, passed (after revisions and split votes) unanimously in the committee and is now submitted to ACI-318 code committee for judgment. Furthermore, the theory of energetic-statistical size effect in shear was reduced to a simple design formula and a method to …