| Back |
ENERGETIC-PROBABILISTIC SIZE EFFECT ON
STRUCTURAL STRENGTH
Zdeněk P. Bažant
In contrast to fluid mechanics, the problems of scaling and size effect in
solid mechanics have not come to the forefront of attention until late in the
last century. The classical view that any observed size effect was statistical
was reversed during the 1980s. As is now widely accepted, quasibrittle materials including concrete, rock, tough
ceramics, sea ice, snow slabs and composites exhibit major size effects on the
mean structural strength that are deterministic in nature, being caused by
stress redistribution and energy release associated with stable propagation of
large fractures or with formation of large zones of distributed cracking. The
lecture begins by reviewing the general asymptotic properties of size effect
implied by the cohesive crack model or crack band model, and highlights the use
of asymptotic matching techniques as a means of obtaining scale-bridging size
effect laws representing a smooth transition between two power laws. A new
asymptotic matching technique combined with dimensional analysis is proposed
and used to justify the size effect laws. Attention is then focused on size
effects observed in fiber-polymer composites failing either by tensile fracture
or by propagation of compression kink bands with fiber microbuckling.
The size effects in polymeric foams and sandwich structures are
also discussed. A nonlocal model for incorporating
the Weibull-type statistical size effect due to local
strength randomness into the energetic size effect theory is described next,
and the predictions of the combined nonlocal
energetic statistical theory are compared to experimental evidence. Nonlocal probabilistic analysis of the size effect on the
statistical distribution of nominal strength of structures is outlined and
discussed from the viewpoint of the extreme value statistics. Implications for
the design of hulls, bulkheads, decks, masts and antenna covers for very large
ships, and for the design of large load-bearing aircraft fuselage panels, are
pointed out. Adaptation of the stochastic finite element method to cope with
extreme value statistics of energetic-statistical size effect is described, and
its importance is demonstrated by analysis of some famous disasters,
particularly Malpasset Dam. Finally, some problematic
features of size effect in designing reinforced concrete structures against
shear failures are pointed out.
References:
Bazant, Z.P. (2004). "Scaling theory for quasibrittle
structural failure."
Bazant, Z.P. (2002). Scaling of Structural Strength. Hermes Penton
Science,

Photo by W.F. Pfeffer
Born and educated in
IA-FRAMCOS and of IA-CONCREEP; Division Director in IA-SMiRT;
is a member of US Nat. Comm. on Theor. & Appl. Mech.; and chaired various committees in ASCE, RILEM,
ACI, SES and IA-SMiRT. He is an Illinois Registered
Structural Engineer. Among his honors: 5 honorary doctorates (Colorado, Lyon,
Milano, Karlsruhe, Prague); SES
Prager Medal; ASME Warner Medal; ASCE Newmark
Medal, Croes Medal, Huber Prize, Lin Award and Lifetime
Achievement Award; RILEM L'Hermite Medal; Am. Ceramic
Soc. Roy Award; Torroja Medal (Spain); Solin and Stodola Medals (Czech
Rep., Slovakia); ICOSSAR Lecture Award; Medal of Merit (TU Prague); SEAOI
Meritorious Paper; Best Engrg. Book
of the Year (SAP); ISI Highly Cited Scientist (www.ISIhighlycited.com); and Guggenheim,
Humboldt, NATO, JSPS, Kajima and Ford Fellowships.
| Back |