F. Michael Bartlett

Bond Strength Variability in Pullout Specimens with Plain Reinforcement

Bond strength results from 252 plain bar pullout specimens are presented. Parameters investigated include: concrete compressive strength, bar size, bar shape, concrete cover, and bar surface roughness. All load-slip curves displayed a characteristic shape: the maximum tensile load occurred at a very small slip (∼0.01 mm) and the load then dropped asymptotically to a residual value as the slip increased to 10 mm. Empirical equations based on least-squares analysis are presented to predict maximum and residual average bond stresses. The load-slip curve can be represented with load as a linear function of the logarithm of slip. The average bond strength was 0.98 MPa for as-received bars, and increased by 124% to 2.2 MPa for bars sandblasted to simulate realistic surface roughness. Coefficients of variation were 8% for maximum average bond stress and 24% for residual average bond stress.

Statistical Analysis of the Compressive Strength of Concrete in Structures

The relationship between the in-place compressive strength of concrete in structures and specified strength Fc is examined via the use of factors F1 and F2. Factor F1, the ratio of the average strength of standard 28-day-old cylinder specimens to the specified strength, is evaluated using data from 3,756 cylinder tests representing 108 concrete mixes produced in Alberta, Canada, between 1988 and 1993. Factor F2, the ratio of average in-place strength to average cylinder strength, is evaluated using core and cylinder data representing 108 concrete mixes with strengths less than 55 MPa that were studied by others. A statistical description of the compressive strength of concrete in structures is derived that accounts for the inherent randomness of Factors F1 and F2 and also the typical strength variation within a specific structure. The probability of the in-place compressive strength of concrete in a 28-day-old column being less than Fc is approximately 13 percent. It is likely that a recalibration of the load and resistance factors for the design of new structures in Canada based on these findings would yield greater factored concrete strengths than are currently in use.

Bond Stresses Along Plain Steel Reinforcing Bars in Pullout Specimens

Although recent research has recognized the existence of relationships between bond stress, slip and bar force, mechanics-based analytical prediction models for bond of plain reinforcement based on pullout specimens have not yet been established. This study establishes mechanics-based relationships between bond stress, bar force, slip at the unloaded end of the bar, and slip along the length of plain steel reinforcing bars in pullout specimens. Two 200 mm (7.9 in.) diameter by 800 mm (31.5 in.) long pullout specimens reinforced with instrumented built-up hollow reinforcing bars were tested. The derived mechanics-based relationships show that bond stress is a function of relative bar slip, slip is a function of bar force, and bar force is a function of bond stress. It is shown both analytically and experimentally that bond stress magnitudes vary along the length of the bar at all applied loads. Maximum pullout resistance was observed just before slip initiated at the unloaded end of the bar, and the bond resistance subsequently reduced as slip increased. The location of the peak bond stress shifts from the loaded end toward the unloaded end of the specimen with increasing applied load. This study also demonstrates that a previously-developed theoretical two-step bond stress-slip model captures the essential features of bond behavior observed experimentally in pullout specimens once the authors accounted for the debonding adjacent to the loaded end of the specimen.

VARIATION OF IN-PLACE CONCRETE STRENGTH IN STRUCTURES

The variation of in-place strength in a structure is due to within-batch variation, batch-to-batch variation, systematic within-member strength variation, and systematic between-member strength variation. Batch-to-batch variation is particularly significant for cast-in-place structures and may either inflate the within-member variation if each member is cast from many batches or inflate the between member variation if each member is cast from a single batch. Values of coefficients of variation that represent the overall variation of the in-place concrete strength in a structure vary from 7% for one member cast from one batch of concrete to 13% for a structure consisting of many members cast from many batches of cast-in-place concrete. Multiple regression analysis techniques are used to assess the systematic variation of the strength of concretes in laboratory specimens cast from one batch of concrete. Statistically significant systematic strength variation is detected over the height of 32 of 43 columns with average strengths from 2,200 psi (15,169,000 Pa) to 5,200 psi (35,854,000 Pa). Typically, the top region was 3-14% weaker than the region in the middle, and the bottom region was 3-9% stronger than the region in the middle. Significant systematic variation of the in-place strength is also detected in 20 of 26 beams, blocks, slabs, and walls with average strengths from 2,200 psi (15,169,000 Pa) to 17,000 psi (117,215,000 Pa). Investigation of ultrasonic pulse velocity and pull-off test data from building columns and bridge girders corroborates the findings of the investigation of elements cast in the laboratory.

Precision of in-place concrete strenghts predicted using core strength correction factors obtained by weighted regression analysis

Abstract A two-step method for converting a concrete core compression test result to the in-place strength of the corresponding volume of concrete is presented. The strength of a non-standard core is first converted to the equivalent strength of a standard core, and then the standard core strength is converted to the equivalent in-place strength. Strength correction factors required for these conversions, obtained from weighted linear and nonlinear regression analyses presented elsewhere, are summarized. The accuracy of the predicted in-place strength is affected by the inherent error of the core strength measurement itself, and by the uncertainty of the various strength correction factors. It is shown that confidence intervals on the estimates of the strength correction factors obtained by regression analysis underestimate the true model error because the underlying models are imperfect. Instead, the accuracy of the strength correction factors is determined by a weighted regression analysis of ratios of observed-to-predicted values which accounts for the non-uniform variances of the dependent and independent variables. The coefficient of variation of the in-place strength predicted from a test of a 100 or 150 mm diameter core is between 4 and 5.5.%. If the in-place strength is predicted from a test of a 50 mm diameter core, the coefficient of variation of the predicted in-place strength is approximately 12.5%. These error estimates do not account for possible variation of in-place strength throughout the volume of the element being cored.

Bond in Flexural Members with Plain Steel Reinforcement

Two concrete T-beams reinforced with instrumented built-up hollow plain bars were tested to investigate the effect of flexural cracking and bond loss on the flexural and shear behavior. The beam with a flexural reinforcement ratio of 0.98% exhibited arch action due to bond failure that initiated when the applied load reached 60% of the failure load. The beam with a reinforcement ratio of 0.33% exhibited predominantly beam action until failure initiated by yielding of the longitudinal reinforcement. When beam action was the primary shear-carrying mechanism, the observed bond demand was greatest within the transition zone between elastic-uncracked and elastic-cracked behavior. The transition from beam action to arch action caused a marked reduction offlexural stiffness that indicated bond loss in beams with plain reinforcement.

Analysis and design of rehabilitated built-up hybrid steel compression members

Compression members in steel bridges designed before 1960 may be deficient according to current design code requirements and so require strengthening. This paper explores the response of steel wide-flange columns reinforced with new steel flange cover plates, accounting for: residual and locked-in dead-load stresses; different yield strengths of the original W shape and the new cover plates; initial out-of-straightness; and load eccentricity at the member ends. A refined numerical analysis model is formulated and validated that computes the compressive resistance from principles of equilibrium, compatibility, and force–deformation relationships. Parametric studies conducted indicate that the capacity of the reinforced column is relatively insensitive to the magnitude of the locked-in dead-load stress in the original member, and that the strong axis capacity is insensitive to the magnitude of the residual stresses in the original member. A preliminary approach for designing strengthening for these members ...

Alternative Load Paths in Steel through-Truss Bridges: Case Study

Conventional design and evaluation procedures usually classify steel through-truss bridges as single-load-path structures; however, their historic performance has demonstrated considerable structural resiliency. This paper presents a study of the Grand River Bridge, a Pratt through-truss bridge in Cayuga, Ontario, Canada, that systematically investigates the load paths not conventionally assumed in design and evaluation. The bridge with individual truss members removed was analyzed using nonlinear finite-element analyses to investigate the alternative load paths and associated critical members and responses. A system reliability analysis was conducted to evaluate the failure probabilities of the bridge accounting for the member resistance uncertainties. The bridge was found to be sufficient to carry its nominal dead load, even for the worst case of removal of an end post. The end bottom chord and hanger vertical were found to be the critical tension chord and web member, respectively. If multiple alternative load paths exist, the variability of the system collapse load is much less than that of the critical member resistance.

ARIMA-based investigation of resistance of member composed of elements with correlated strengths

A method for computing the variance of the strength of a compound member consisting of identical mass-produced ductile elements is presented that accounts for correlation between the strengths of the individual elements. The individual element strengths are represented as a first-order, zero-mean, autoregressive time series and the associated variance, covariance and correlation are computed. The variance of the overall strength of the compound member is derived, and can be computed efficiently without requiring derivation of the element strength correlation matrix. The sensitivity of the variance of the overall strength of the compound member to the number of elements and the degree of correlation between their strengths is investigated. The common assumption of fully correlated strengths is increasingly conservative as the number of elements increases or the degree of correlation between their strengths reduces. Two practical examples illustrate the method. The first also quantifies approximately the impact of the degree of correlation on the resistance factor of the compound member. The second also indicates how readily available data, such as the within-batch strength variation of individual members, can be used to estimate the degree of correlation of their strengths.