Evaluation of current crack width calculation methods according to eurocode 2 and fib model code 2010

Abstract

The background theory for the crack width calculation methods according to Eurocode 2 and fib Model Code 2010 is discussed to evaluate the applicability for the more general case of relatively thick beams, slabs and shells. Essentially, the formulas originate from the maximum transfer length and the difference in steel and concrete strains over this length. It is shown that the formulas are based on both a slip and a no-slip theory, two theories using exactly opposite assumptions. The slip theory assumes that a physical slip occurs in the interface between concrete and steel and, also, that plane sections remain plane. The no-slip theory assumes that no physical slip occurs between concrete and steel and, thus, that plane sections no longer remain plane. The theories were merged pragmatically in an attempt to describe the physical reality related to cracking. This resulted in a formula for the transfer length composed by two linear terms. Such a formulation, however, leads to inconsistencies that opposes the basic principles in solid mechanics. It is argued that these inconsistencies limits the application for the more general case. The observations in this paper suggests that a more robust and consistent calculation method should be formulated. A possible way is by improving the bond assumptions in the interface between concrete and steel, and thoroughly studying the geometry and configuration of cracks experimentally and theoretically.