The periodic assessment of our existing concrete infrastructure is a crucial part of maintaining appropriate levels of public safety over long periods of time. It is important that realistic predictions of the capacity of existing structures can be made in order to avoid unnecessary and expensive intervention work. Assessment is currently undertaken using codified models that are generally readily applied to infrastructure with simple geometric and reinforcement details that conform to design methods for new structures.
This approach presents two significant challenges for prestressed structures: 1) design and construction practice has changed significantly in the past 50 years, and modern codified approaches can be incompatible with historic structures; and 2) deterioration of exposed soffits can lead to reduced cover to internal prestressing strand. Unless appropriate reductions are used in assessment of a structure with such problems, unnecessary load restrictions, or major strengthening or reconstruction work may be required, despite having carried a full service load since its construction.
There are currently no widely accepted methods for the prediction of peak and residual capacities in prestressed concrete beams with inadequately detailed 7-wire strand. This paper presents a completely new prediction methodology, validated against new experimental results from 31 novel semi-beam tests. The proposed models for peak load, residual load, and bond stress-slip modelling provide reliable, accurate, and conservative results. Their results demonstrate feasible and appropriate capacity reduction factors for use in the assessment of existing concrete infrastructure.
Original languageEnglish
Pages (from-to)1-16
Number of pages16
JournalEngineering Structures
Early online date14 Feb 2017
Publication statusPublished - 1 May 2017


  • Asssessment
  • Bridges
  • Prestressed concrete structures
  • Half joints


Dive into the research topics of 'Anchorage and residual bond characteristics of 7-wire strand'. Together they form a unique fingerprint.

Cite this