Title:
Low-Cycle Fatigue Effects on the Seismic Performance of Concrete Frame and Wall Systems with High Strength Reinforcing Steel
Author(s):
Kuanshi Zhong;Gregory G. Deierlein
Publication:
CRC
Volume:
Issue:
Appears on pages(s):
Keywords:
seismic,design,high-strength, reinforcing,steel,concrete,nonlinear,degradation,bar,yieling,buckling,fracture,
DOI:
Date:
8/27/2019
Abstract:
This is one of several companion projects supported by the Pankow Foundation to support the
development of seismic design requirements for the use of high strength reinforcing steel in concrete
buildings. In particular, this study examines the effect of high strength reinforcement on the nonlinear
system response and risk of bar fracture in concrete moment frames and walls subjected to earthquake
ground motions. The risk of reinforcing bar fracture is a particular focus, given concerns raised regarding
the difference in cyclic ductility between conventional and high strength reinforcement. The analyses
utilize data from cyclic testing of high-strength reinforcing bars to evaluate bar fracture and data from tests
of concrete beam, column and wall components to validate the simulations. These models are used to assess
the influence of reinforcing bar properties on the seismic performance of special moment frame and wall
systems under earthquake ground motions, including the effects of degradation associated with reinforcing
bar yielding, buckling and fracture.
Data from over 250 tests of reinforcing bars conducted in a companion project (Ghannoum and Slavin,
2016) is used to calibrate a fatigue-fracture model based on a Manson-Coffin formulation to relate
cumulative effective plastic strains to bar fracture. Effective plastic strains are measured over a gage length
associated with the bar slenderness based on lateral reinforcing tie spacing, which implicitly includes bar
buckling. Statistical regression is used to calibrate three material parameters (αf, Cf, and εf) to capture the
correlation of combined buckling and fracture-fatigue resistance to steel yield strength, tensile-to-yield
strength ratio, and bar slenderness. The resulting fracture-fatigue damage model is described by a
lognormal probabilistic distribution, calibrated to represent the median estimate of fracture with a
dispersion of 0.5.
The response of beam, column and flexural wall components is simulated using fiber-type elements that
incorporate inelastic behavior of reinforcing steel and concrete and bar slip. Numerical integration points
are selected such that the bar strain demands correspond to an 8-inch gage length that is consistent with
the effective strains in the bar fatigue-fracture model. The detailed member analyses and reinforcing bar
fracture model is validated using data from companion Pankow projects, including tests of beams (To and
Moehle, 2017), columns (Sokoli et al., 2017), and T-shaped walls (Huq et al., 2018) with conventional (Gr.
60) and high-strength (Gr. 80 and 100) reinforcement. Overall, the simulated forces, deformations,
reinforcing bar strains and fracture indices compare well with the measured test data.
Multi-story moment frame seismic systems are modeled using beam-column members with concentrated
hinges, where the resulting member deformation histories are interpreted using the detailed fiber-type
elements. The concentrated hinges employ a strength and stiffness degrading hysteretic model which has
five backbone-curve parameters and one cyclic deterioration parameter. The backbone curve parameters
are determined by aggregating moment-rotation response from fiber-type cross-section analyses with
member shear and bar-slip relations. To improve the reliability of these analyses for identifying the
influence of reinforcing bar properties on response, the model backbone and cyclic deterioration
parameters are calibrated through statistical regression to data from over one hundred previously
published tests, including specimens with normal and high strength reinforcement.
The calibrated component and fracture-fatigue models are used in incremental dynamic analyses (IDA) to
evaluate the performance of special moment frame and wall systems under strong ground motions. In one
set of analyses, the story drifts, cumulative plastic strains, reinforcing bar fracture risks are evaluated under
MCE ground motions for high-seismic sites in San Francisco and Seattle, where the results are adjusted to
account for characteristic ground motion spectral shape and duration. In a second set of analyses, the
CPF Research Project RGA #02-16
Final Report
4
FEMA P695 approach is used to evaluate the safety margin against collapse, with and without the effects
of reinforcing bar fracture.
Alternative frame and wall designs are compared to evaluate the effects of reinforcing bar strength, tensileto-yield ratio (T/Y), and stirrup spacing. Three T/Y ratios are considered for each steel grade (1.3, 1,4, and
1.5 for Gr. 60; 1.2, 1.3, and 1.4 for Gr. 80; 1.1, 1.2, 1.3 for Gr. 100) and three stirrup tie spacings (s/db equal to
4, 5 and 6) are considered for each type of steel. This results in 27 different combinations of steel grade, T/Y
ratio, and s/db ratio for each building type. The study includes 4-story and 20-story concrete special moment
frames and 8-story and 42-story concrete shear wall systems designed to conform to current seismic design
standards.
Key results of the structural component and building archetype analysis studies are as follows:
- As one would expect, replacing Gr. 60 reinforcement with higher strength bars tended to decrease
the building stiffness, increase the fundamental period, and increase the earthquake-induced drifts.
The moment frames with Gr. 80 and Gr. 100 reinforcing bars experienced maximum story drifts
that were about 10% to 20% larger, respectively, than those in the benchmark buildings with Gr.
60 reinforcement. Story drifts in the wall systems with Gr. 80 and 100 reinforcing bars were about
5% to 10% larger, respectively, than the Gr. 60 benchmark design. These results suggest that the
design story drifts should be evaluated using models based on transformed section properties,
which reflect differences in steel areas for higher grade steels.
- Among the three reinforcing bar material properties and design parameters that were
systematically varied in the analyses (yield strength, T/Y ratio, s/db ratio), the T/Y ratio has the most
significant effect on the risk of reinforcing bar fracture, followed by s/db ratio and yield strength. It
should be noted that the characteristic T/Y ratios tend decrease with increasing yield strength, so
these effects are correlated. But the sensitivity studies indicate that change in fracture behavior is
more directly driven by the T/Y ratio. Variation in reinforcing bar toughness, beyond that captured
in the yield strength and T/Y ratio parameters, is treated as a random variable that is reflect in the
dispersion of the facture model.
- Comparative studies of frame and wall systems with variable steel reinforcement indicate that
structures with Gr. 80 and 100 reinforcing bars with T/Y = 1.2 and tie spacing s/db = 5 have similar
cyclic strain demands and probabilities of reinforcing bar fracture (under MCER ground motions)
to benchmark counterpart systems with Gr. 60 bars with T/Y = 1.3 and tie spacing s/db = 6. In most
cases, the reduction in the tie spacing (s/db from 6 to 5) tended to offset the increased tendency for
fracture in higher strength bars with lower T/Y ratios (T/Y of 1.2 versus 1.3). Where the T/Y ratios
and tie spacings were controlled to T/Y = 1.2 and tie spacing s/db = 5, the probabilities of bar fracture
under MCER ground motions ranged from about 2% to 5% in the moment frames and 42-story shear
wall. Considerably larger fracture probabilities of about 10% to 15% where observed in the 8-story
shear wall, due to higher cyclic strain demands. In cases where the T/Y ratio is reduced to 1.1 and
the s/db tie spacing is relaxed 6, the bar fracture probability roughly doubles compared to cases with
constraints on the permissible bar properties and tie spacing, i.e., T/Y = 1.2 and s/db = 5.
- Comparison between fracture index (FI) demands for the frames and walls at story drift ratios up
to 4% further confirmed that the fracture probabilities of Gr. 80 and 100 bars are roughly equivalent
to the benchmark case (Gr. 60 with T/Y = 1.3 and s/db = 6) by limiting the minimum T/Y ratio to 1.2
and maximum spacing s/db to 5 (or possibly s/db of 4 in the 8-story wall where the fracture demands
are largest).
CPF Research Project RGA #02-16
Final Report
5
- FEMA P695 Collapse Risk analyses indicated that the risk of collapse under MCER motions is
comparably between the systems with Gr. 60 versus Gr. 80 or 100 steel, provided that the T/Y > 1.2
and tie spacing s/db < 5 in cases with high strength reinforcing bars. Overall, the risk of collapse
under MCER motions is negligible (<1%) in the 42-story shear wall and on the order of 5% to 11%
in the other systems. As noted previously, control of the T/Y ratio and tie spacing is especially
important in the 8-story wall, where the contribution of bar fracture to collapse risk would
otherwise increase to values (MCER collapse probabilities up to 30%) that are in excess of the
specified requirements of ASCE 7.