construction · infrastructure · family: it worked in the lab
the bridge rustswhere you cannot look
Post-tensioned concrete tendons cannot be inspected for corrosion once grouted — failures are catastrophic and without warning
Problem statement
Post-tensioned (PT) concrete is used in approximately 50% of new bridge construction and in thousands of parking garages, stadiums, and high-rise buildings worldwide. Steel tendons — the structural elements that give PT concrete its strength — are encased in grout inside ducts after tensioning. Once grouted, these tendons are completely inaccessible to visual inspection. Corrosion of tendons (from grout voids, chloride contamination, or hydrogen embrittlement) progresses invisibly until catastrophic tendon rupture occurs, often without warning. Tendon corrosion that no inspection method could detect has already caused a complete collapse (the Ynys-y-Gwas bridge in Wales, 1985 — the segmental PT bridge failed without warning; Woodward & Williams 1988) and emergency closures short of collapse (London's Hammersmith Flyover in 2011, numerous FDOT bridges in Florida), in each case discovered only when the structure showed visible distress or failed outright.
Why this matters
A single corroded tendon in a PT structure can carry 200–500 kips of force. When it ruptures, the load redistributes suddenly to adjacent tendons, potentially triggering progressive collapse. The Florida DOT alone discovered corrosion-induced tendon failures in 40+ bridges built after 1983, leading to emergency closures, weight restrictions, and a statewide inspection mandate. The UK's Highways England identified tendon corrosion as the highest-priority inspection gap in its bridge stock. Yet no reliable method exists to assess tendon condition through the concrete cover and grouting that surrounds them.
What’s been tried and why it hasn’t worked
Magnetic flux leakage (MFL) can detect section loss in ungrouted tendons but is severely attenuated by the steel duct and grout — sensitivity drops to only detecting >50% section loss, far past the point of useful intervention. Impact-echo and ultrasonic methods can locate grout voids (which indicate corrosion risk) but cannot determine whether corrosion has actually initiated in the tendon. Electrochemical methods (half-cell potential) work for conventional reinforcing steel near the surface but cannot penetrate to tendons buried 6–18 inches deep inside electrically isolated ducts. Acoustic emission monitoring can detect wire breaks in real time but requires pre-installed sensors and continuous monitoring — it cannot assess existing structures retroactively. Ground-penetrating radar locates ducts but cannot image tendon condition within them.
What would unlock progress
A nondestructive sensing modality that can penetrate concrete cover + metal duct + grout to assess steel tendon condition with sufficient sensitivity to detect early-stage corrosion (5–10% section loss) before structural capacity is compromised. Candidate approaches include high-energy X-ray/gamma radiography (used in industrial pipe inspection; gamma radiography has already been used on a Florida bridge to locate displaced ducts, and radiography can locate grout voids, but it cannot show corrosion until a considerable part of the steel section is lost), guided ultrasonic waves propagated along the tendon from access points, or electromagnetic methods operating at frequencies that penetrate the grout-filled duct system. The key challenge is signal-to-noise: the concrete, duct, and grout layers that protect the tendon also shield it from every known sensing modality.
Entry points for student teams
A team could construct a mock-up PT beam with intentionally corroded tendon segments (induced by accelerated corrosion prior to grouting) and systematically test the sensitivity of available NDE methods (MFL, impact-echo, GPR, acoustic emission) against known defect states. A signal processing team could explore whether machine learning models trained on multi-modal NDE data (combining several weak signals) can detect tendon corrosion that no single method resolves alone. Relevant disciplines: nondestructive evaluation, signal processing, structural engineering, materials science.
Genes
Structural cousins: stuck the same way, in other fields
Sources
PCI Committee on Post-Tensioning, "Guide for Grouting of Post-Tensioned Structures," PCI Journal, 2019; Corven Engineering, "New Directions for Florida Post-Tensioned Bridges," FDOT Final Report, 2002; R. J. Woodward & F. W. Williams, "Collapse of Ynys-y-Gwas Bridge, West Glamorgan," Proceedings of the Institution of Civil Engineers, Part 1, Vol. 84, 1988, pp. 635–669, Accessed 2026-02-25; Woodward & Williams citation corrected and re-verified 2026-08-20. go to source ↗
Working notes
Notes made while writing and checking this brief, shown as written. They use working shorthand.
The "not-attempted" tag reflects that the problem has been known for 30+ years (first PT corrosion failures documented in the 1970s) but no sustained R&D program has targeted through-grout tendon assessment — the response has been design changes to prevent corrosion (improved grout, encapsulated strand) rather than detection methods for existing structures. The "ignored-context" tag reflects that the original PT system design made inspection impossible by design — accessibility was not a design consideration. Related briefs: construction-bridge-visual-inspection-subjectivity (both are bridge inspection gaps), construction-shm-existing-building-stock-gap (similar challenge of assessing existing structures).
Reconciliation 2026-08-20: the genome no longer carries `failure:not-attempted` for this brief; the current tags are `failure:lab-to-field-gap, ignored-context`. The note above is kept verbatim as the original tagging rationale. Same date: the third source citation previously conflated two UK cases ("Collapse of Hammersmith Flyover Tendon Failures," ICE Forensic Engineering, 2015 — no such paper). Corrected to Woodward & Williams 1988 (Ynys-y-Gwas, the actual collapse); Hammersmith Flyover suffered tendon corrosion leading to emergency closure and repair in 2011–2012, not a collapse, and the Problem Statement now says so.
Reconciliation 2026-09-29: A site review read this write-up against the sources it cites. Of the three, one could be read in full. The Corven Engineering report for the Florida Department of Transportation was fetched on 2026-09-29 from https://fdotwww.blob.core.windows.net/sitefinity/docs/default-source/content/structures/posttensioning/newdirectionsposttensioningvol1.pdf and the matching addresses for volumes 2 to 10B (the copy of Volume 1 at www.fdot.gov downloaded but could not be opened). All eleven files were turned into text and searched; the repair history in Volume 1 (section 1.3) and the review of testing methods in Volume 9 (Appendix B) were read line by line. The Woodward and Williams paper is behind a paywall, so only its citation record was fetched, from https://api.crossref.org/works/10.1680/iicep.1988.179 and https://trid.trb.org/View/259373. The PCI guide has no link and a web search did not find a document with that title, so it was not read. (1) The Source line gave the Corven title as "New Directions for Florida's Post-Tensioned Bridges." The title page reads "New Directions for Florida Post-Tensioned Bridges"; the Source line now matches it. (2) The text said radiography was "used in industrial pipe inspection but not adapted for field use on bridges." Volume 9, Appendix B.2.10 describes gamma-ray radiography used on a Florida bridge near West Palm Beach to find displaced ducts. It says radiography can locate ducts and grout voids, and that it cannot detect corrosion "unless there is considerable loss of section or many broken strands or wires." The sentence now says that. Confirmed against the Corven report and left as written: tendons sit in grout inside ducts, and internal tendons cannot be inspected or replaced; corroded tendons failed on Florida bridges (Niles Channel in 1999, Mid-Bay and the Sunshine Skyway piers in 2000) and were found during inspections; grout voids and salt-contaminated water are named as causes; magnetic flux testing works on external tendons and is not suited to internal ones; impact echo can find voids; the electric half-cell method is not recommended for tendons; radar cannot see through metal ducts; acoustic monitoring needs a network of sensors and a period of monitoring. The Woodward and Williams citation matches the TRID record (title, authors, volume 84, pages 635–669, August 1988). Crossref and the publisher page spell the title "Collapse of Yns-y-Gwas Bridge, Glamorgan," which looks like a data-entry error, so the citation was left as it is. NOT checked, because the Corven report does not contain them and the other two sources could not be read: that post-tensioned concrete is used in approximately 50% of new bridge construction; the 200–500 kips carried by one tendon; "40+ bridges built after 1983" in Florida, and the emergency closures, weight restrictions and statewide inspection mandate said to follow (the Corven report describes tendon failures on three bridges, grout voids on a fourth, and emergency inspections of one, and gives no count of bridges); the statement about Highways England; the ">50% section loss" limit for magnetic flux testing; the 6–18 inch depth; the 1985 date and "without warning" for Ynys-y-Gwas; the 2011 Hammersmith Flyover closure, which rests on the note above; and the first failures in the 1970s. These are left as written. Genes untouched.