Title:
Efficiency of Strut-and-Tie Model for Design of Reinforced Concrete Deep Beams without Web Reinforcement
Author(s):
Kondalraj R and Appa Rao G
Publication:
Structural Journal
Volume:
119
Issue:
3
Appears on pages(s):
233-247
Keywords:
deep beam; diagonal cracking load; strut-and-tie model; strut efficiency
DOI:
10.14359/51734494
Date:
5/1/2022
Abstract:
The efficiency of ACI 318 and AASHTO strut-and-tie models for designing the reinforced concrete deep beams without web reinforcement is discussed. A database of 232 tests on reinforced
concrete (RC) deep beams without web reinforcement was chosen for the investigation. Based on a review of the experimental database, the depth of flexural compression derived using linear elastic analysis appears to be an acceptable approximation for deep beams. Compared to AASHTO and ACI 318-19, ACI 318-14 suggests a higher strut efficiency factor of 0.51 for struts without web reinforcement. When the shear strength limit of ACI 318-14 is considered, the shear strength is estimated with a mean strength ratio of 0.69 and a 4% overestimation. To reduce the overprediction, ACI 318-19 reduces the strut efficiency factor to 0.34 from 0.51, which also decreases the mean strength ratio to 0.51 and eliminates the overprediction of capacity. AASHTO’s strut efficiency of 0.45 exhibits a 5% strut efficiency factor estimated from the database. Analysis of the experimental database reveals that the strut efficiency factor decreases with an increase in concrete strength. The lowest strut efficiency by ACI 318-19 may result in an unsafe estimation for high concrete strength. As a result, the strut efficiency factor as a linear function of concrete strength has been presented, which predicts a deep beam capacity with high accuracy—less than 3% overestimation. This study also examines the factors affecting the diagonal cracking load. The beam depth affects the diagonal cracking load, which was ignored in AASHTO. The improved AASHTO formula, accounting for the size effect, predicts diagonal cracking load with the same accuracy as AASHTO’s original expression while reducing overestimation to 5% from 13%.