Deformation capacity and moment redistribution of partially prestressed concrete beams

Ductility is a measure of the ability of a material, section, structural element or structural system to sustain deformations prior to collapse without substantial loss of resistance. The Australian design standard, AS 3600, imposes minimum ductility requirements on structural concrete members to try to prevent premature non-ductile failure and hence to ensure adequate strength and ductile-type collapse with large deflections. The requirements also enable members to resist imposed deformation due to differential settlement, time effects on the concrete and temperature effects, whilst ensuring sufficient carrying capacity and a safe design. Current AS 3600 requirements allow a limited increase or reduction in elastically determined bending moments in critical regions of indeterminate beams, accommodating their ability to redistribute moment from highly stressed regions to other parts of the beam. Design moment redistribution limits and ductility requirements in AS 3600 for bonded partially prestressed beams are a simple extension of the requirements for reinforced members. The possibility of premature non-ductile failure occurring by fracture of the reinforcement or prestressing steel in partially prestressed members has not adequately addressed. The aim of this research is to investigate the overload behaviour and deformation capacity of bonded post-tensioned beams. The current ductility requirements and design moment redistribution limits according to AS 3600 are tested to ensure designs are both safe and economical. A local flexural deformation model based on the discrete cracked block approach is developed to predict the deformation capacity of high moment regions. The model predicts behaviour from an initial uncracked state through progressive crack development into yielding and collapse. Local deformations are considered in the model using non-linear material laws and local slip behaviour between steel and concrete interfaces, with rigorous definition of compatibility in the compression and tension zones. The model overcomes limitations of past discrete cracked block models by ensuring compatibility of deformation, rather than strain compatibility. This improvement allows the modeling of members with multiple layers of tensile reinforcement and variable depth prestressing tendons having separate material and bond properties. An analysis method for simple and indeterminate reinforced and partially prestressed members was developed, based on the proposed deformation model. To account for the effect of shear in regions of high moment and shear present over the interior supports of a continuous beam, a modification to the treatment of local steel deformation in the flexural model, based on the truss analogy, was undertaken. Secondary reactions and moments due to prestress and continuity are also accounted for in the analysis. A comparison of past beam test data and predictions by the analysis shows the cracking pattern and deformation capacity at ultimate of flexural regions in reinforced and partially prestressed members to be predicted with high accuracy. The analysis method accurately predicts local steel behaviour over a cracked region and deformation capacity for a wide range of beams which fail either by fracture of steel or crushing of the concrete. A parametric study is used to investigate the influence of different parameters on the deformation capacity of a typical negative moment region in a continuous beam. The structural system consists of a bonded post-tensioned, partially prestressed band beam. The primary parameters investigated are the member height and span-to-depth ratio; relative quantity of reinforcing and prestressing steel; material properties and bond capacity of the steels; and lastly the compression zone properties. Results show that the effects of the various parameters on the overload behaviour of partially prestressed beams follow the same trends as reinforced beams. A new insight into the local steel behaviour between cracks is attained. The deformation behaviour displays different trends for parametric variations of the local bond capacity, bar diameter and crack spacing, when compared to past analytical predictions from comparable studies. The discrepancy in findings is traced back to the definition of the plastic rotation capacity and the sequencing of the yielding of the steels. Compared to the other local deformation models, the current model does not assume a linear distribution of strain at a crack. The current findings highlight an important difference between predicted behaviours from different deformation compatibility requirements in local deformation models which has not yet been discussed in the literature. The local deformation model evaluates the relationship between maximum steel strain at a crack and average steel deformation over a crack spacing for the entire loading history. The total steel percentage, hardening properties of the steel and concrete strength are shown by the model to have the greatest effect on these steel strain localisation factors. Section analysis, as currently used in design, can be improved with the proposed simplification of the relationships to identify and quantify the effects of steel fracture on deformation capacity and strength. The numerical effort required to simulate the overload behaviour of practical beam designs with multiple reinforcement elements and a prestressing tendon are currently too great to be used in an extensive numerical study. The numerically more efficient smeared block approach is shown to accurately predict the ultimate carrying capacity of prestressed beams failing by crushing of the concrete. Consequently, this method is adopted to study the allowable limits of moment redistribution in the present investigation, Simplified relationships of the steel strain localisation factors evaluated in the parametric study of deformation capacity is used to predict maximum steel strains and premature failure. The limits of moment redistribution in bonded, post-tensioned partially prestressed band beams are explored by comparing the design load and predicted carrying capacity, for different section ductilities and design moment redistribution. In addition, the effects of different concrete strengths, up to 85 MPa, along with as three reinforcing and prestressing steel ductilities are quantified and compared to current Australian and international design requirements. Limitations in the carrying capacity are investigated for different reinforcement and prestress uniform elongation capacities. More than one thousand beam simulations produce results showing that current design moment redistribution and ductility requirements in the Australian design code for concrete structures (AS 3600) are sufficient for normal strength concretes (less than 50 MPa). A suggestion for design moment redistribution limits, section ductility requirements and steel ductility limits is made for members constructed from higher strength concretes. A special high steel ductility class is proposed for both the reinforcement and prestressing steel to allow moment redistribution in higher strength concrete. No moment redistribution is proposed for members reinforced with low ductility (Class L) steel. An increase of the current elongation limit of Class L steel from 1.5 % to 2.5% is suggested to ensure strength and safety. An increase in the current ductility requirements from fsu/ fsy=1.03 and elongation equal to 1.5% to fsu/fsy=1.05 and 2.5% elongation for low ductility Class L steel is suggested to ensure strength and safety.Thesis (Ph.D.) -- University of Adelaide, Dept. of Civil and Environmental Engineering, 200