For concrete repair and strengthening projects, licensed design professionals are often challenged with developing an appropriate assessment and testing scope to characterize the as-built and current conditions of a structure. Performance-based standards for assessment, repair, and rehabilitation of existing concrete structures, such as ACI CODE-562, require representative measurements to be performed. There is always the trade-off between decreasing uncertainty by performing more samples versus limiting the number of samples to control costs and potentially reduce schedule. This session will present real world examples of how different practitioners have made decisions on sample size for different properties based on statistical considerations, and how this was accounted for in determining the characteristic properties. The impact on the project performance will also be discussed.
Learning Objectives:
(1) Define populations for representative sample sets;
(2) Procedure for determining an appropriate test sample size to meet project objectives;
(3) Factors that affect the sample size;
(4) Determine appropriate confidence limits for various measured parameters.
Optimizing GPR Sampling Density for Representative and Reliable Assessment of Concrete Bridge Members
Presented By: Wael Zatar
Affiliation: Marshall University
Description: Performance-based standards for the assessment, repair, and rehabilitation of existing concrete structures—such as ACI CODE-562—emphasize the need for representative and sufficient measurements to ensure reliable evaluations. In this context, determining the appropriate sampling density is critical for balancing reliability with testing efficiency. This study examines the effect of sampling density, represented by ground-penetrating radar (GPR) scan spacing, on accurately identifying rebar and tendon locations in prestressed concrete bridge beams. Each scan line was treated as a sampling iteration analogous to a statistical sample size in condition assessment. A decommissioned box beam was scanned using transverse spacing intervals of 4, 8, 12, 16, and 24 inches, and automated data processing quantified rebar depth and spacing. The results demonstrated that scan spacings up to 12 inches produced nearly identical measurements to those from denser scans, confirming that, beyond this threshold, additional data collection offers minimal benefit. These findings provide quantitative guidance for determining appropriate test sample sizes in nondestructive evaluation of concrete bridge members. This work establishes a statistically informed framework for optimizing GPR scan spacing, ensuring reliable results while minimizing data collection effort and supporting more practical, scalable bridge evaluation and repair practices.
Case Studies on Sample Selection for Assessment Before Rehabilitation
Presented By: Ali Abu-Yosef
Affiliation: Pivot Engineers
Description: This presentation explores the difficulties in reliably defining populations when evaluating existing structures before rehabilitation. The choice of population parameters influences the sample size and the scope of the investigation. To illustrate the effect of population selection, we will describe two case studies of major rehabilitation projects in which defining the affected population influenced the scope and extent of the investigation. The first case study will focus on a durability assessment in which measured parameters were evaluated using statistically based methods. The second will present a feasibility study for strengthening an existing structure. Both case studies will describe specific types of testing performed on the existing structures, how distinct populations were defined within each structure, and how sample sizes were determined.
Statistical Techniques to Reduce the Size of Concrete Testing Programs
Presented By: Benjamin Estrich
Affiliation: Martinmartin Inc
Description: This presentation looks back at a concrete repair project with a large non-destructive testing program, completed in 2021. Martin/Martin prepared structural concrete repair documents for two existing, 20-year-old, cast-in-place post-tensioned parking structures in Denver CO, encompassing approximately 200,000 SF of parking. Unique project features included extensive x-ray imaging of in-place post-tension system anchorage components to limit concrete demolition at the site. This presentation will use statistical techniques to look back at the project to determine if the sample size of the non-destructive testing program could have been reduced by varying the accepted confidence interval. The presentation will also discuss the number of destructive test samples necessary to calibrate the non-destructive test data.
Statistical Approach to Assessment of PT Structures
Presented By: Joshua White
Affiliation: Wiss, Janney, Elstner Associates
Description: Post-tensioned bridge systems have provided safe and efficient construction solutions for the transportation world. Although significant cost savings can be attributed to the use of such systems, failed tendons due to deterioration conditions may cause reduced safety and load capacity, costly rehabilitation actions, and traffic disruption if construction, maintenance, or repairs are not performed properly and in a timely manner.
This presentation focuses on how using a statistical framework can guide the assessment of such structures, particularly when used with state-of-the-art nondestructive evaluation methods. Case studies will be reviewed that discuss practical limitations and advantages of various evaluation methods.
Evaluating the Condition of Unbonded Post-Tensioned Tendons: Balancing Investigation Scope and Sampling Strategy
Presented By: Diego Romero
Affiliation: Walker Consultants
Description: The evaluation of unbonded post-tensioned tendons in parking structures is a critical component of structural restoration, particularly when deterioration is suspected but not visibly evident. This presentation outlines a structured approach to developing and executing a destructive investigation program aimed at assessing tendon condition and informing restoration decisions.
Key elements of the investigation include determining appropriate sample sizes and selecting between targeted and random sampling strategies. Targeted investigations focus on areas of known or suspected distress, while random sampling aims to provide statistically representative insights across the structure. The presentation will explore the merits and limitations of each approach, emphasizing how sampling strategy influences the reliability and scope of conclusions drawn.
A case study will be presented involving a large-scale destructive investigation of a multi-level parking structure. The investigation included tendon exposure, documentation of corrosion and distress, and statistical analysis of findings. Statistical methods will be used to back-check the results and evaluate how sample size and sampling approach affect the confidence level in assessing the overall condition of the structural system.
The session will conclude with practical recommendations for engineers and restoration professionals on designing effective investigation programs, balancing investigative rigor with logistical constraints. Attendees will gain actionable insights into optimizing destructive investigations for post-tensioned systems in aging parking structures.