What are the as-designed, as-detailed and as-built versions of a structural connection?
The as-designed version of a structural connection exists in the engineer’s calculations and the contract drawings, often as a required capacity rather than a complete detail; the as-detailed version appears in shop drawings prepared by the fabricator or a detailer, which convert that requirement into specific plates, welds and bolts; and the as-built version is what erection actually produced, including any field modification. The as-designed, as-detailed and as-built versions of a connection should agree, and frequently they do not. Establishing all three versions is the first analytical task in a connection failure investigation.
How does delegated connection design go wrong?
Delegated connection design goes wrong when the forces the engineer of record gives the fabricator are incomplete. Connection design is commonly delegated to the fabricator, with the engineer of record specifying forces and reviewing submittals, and that arrangement works when the forces given are complete. Delegated connection design fails when a required force is omitted, when the engineer of record specifies only a shear reaction for a joint that also carries axial load or moment, or when a stiffness assumption in the frame analysis is not communicated. The connection is then correctly designed for the wrong demand.
What does the shop drawing review record show in a connection dispute?
The shop drawing review record establishes what each party to a structural connection knew and when. Shop drawing submittal review generates a documented exchange: what was submitted, what was marked up, what was resubmitted and what was finally approved. A connection built exactly as an approved shop drawing shows sits differently from one built contrary to it, and a detail that was questioned in review and then approved unchanged sits differently again. The shop drawing review file is the single most useful documentary source in most connection disputes.
What do requests for information (RFIs) reveal about a connection failure?
Requests for information (RFIs) mark the known problems: an RFI is a record that somebody encountered something they could not resolve, and the RFI response is a record of what they were told to do. RFIs concerning fit-up, interference, member size changes or connection geometry at or near the failure location are directly relevant to a connection failure investigation. The absence of an RFI where a field change plainly occurred is itself informative, because it indicates the field change was made without engineering review.
How are undocumented field modifications to a connection identified?
Undocumented field modifications to a structural connection are identified by comparing the physical evidence on the recovered connection against the approved detail, and that physical evidence is usually visible on the recovered connection. Fit-up problems get resolved on site under time pressure, and the resolutions are not always documented: holes are flame-cut or reamed oversize, bolts are omitted where they cannot be reached, plates are trimmed to clear an obstruction, stiffeners are left out, and welds are added or substituted for bolts. Each of these field modifications changes the capacity of the connection, sometimes substantially.
What does the recovered hardware from a failed connection show?
The recovered hardware from a failed connection carries direct evidence: bolt grade markings against what was specified, hole sizes and condition against the detail, weld sizes and profiles against the drawing, and whether the specified number of fasteners is present. Flame-cut holes, mismatched bolt grades within a single joint, undersized welds and missing elements are all findings that need no analysis to state, and they are commonly the substance of the case.
Why is the erection stage a separate question in a connection failure?
The erection stage is a separate question because a structural connection may be complete in its final condition and inadequate during erection, when only some bolts are installed, welds are not yet made, and the bracing that gives the frame stability is absent. Collapses during erection frequently trace to this temporary erection-stage condition rather than to the permanent design, and analyzing the joint only in its finished state misses it entirely. The erection sequence and the temporary bracing plan are the relevant documents for the erection-stage question.
What do inspection records establish in a connection failure investigation?
Inspection records establish what was verified on a structural connection: what was checked and what was found. Shop and field inspection reports, weld inspection records and their nondestructive testing results, bolt installation verification, and any nonconformance reports all document that verification. A defect in a joint that was inspected and accepted raises a question about the inspection; the same defect in a joint that was never inspected raises a different one. Both questions are relevant, and the distinction usually matters to the outcome.
What should be preserved after a connection failure?
After a structural connection failure, the connection itself, adjacent unfailed connections of the same type, and the complete documentary set should all be preserved. The failed connection should be kept intact, cut out well away from the joint itself rather than through it, with all fasteners, plates and fragments retained and their positions recorded. Adjacent unfailed connections of the same type establish what the population actually looked like. The complete documentary set is the contract drawings, shop drawings with review markings, requests for information (RFIs) and their responses, change orders, erection procedures and all inspection records.
This article is general technical orientation, not a failure analysis, an engineering opinion, or advice on any specific matter. Determining the cause of a particular incident requires hands-on examination by a credentialed expert.