Managing Repairs, Modifications & Configuration Changes in Integrity Assessments
How repairs, temporary fixes and plant modifications are controlled so that recovered local margin does not create an unrecognised new load path or invalidate the wider integrity case.
A Repair Changes the Structure
A repair is not merely the removal of a defect; it creates a new structural configuration. Welded patches introduce residual stress and stiffness changes, clamps redistribute load through contact and bolts, composite wraps add anisotropic load paths, and replacement sections alter weld locations and geometry. The post-repair integrity assessment should therefore treat the repaired asset as a new configuration with its own failure modes rather than assuming the original design basis automatically applies.
Separate Temporary Mitigation from Permanent Repair
Temporary measures may be entirely appropriate when their authorised period and controls are explicit. A clamp, local support, operating restriction or monitoring arrangement can reduce risk while a permanent solution is prepared. Problems arise when temporary measures remain indefinitely without reassessment. The technical record should state the design life of the measure, inspection requirements, environmental limits and the event or date that triggers replacement or review.
Understand the Load-Path Change
The central repair question is where load travels after the modification. Added stiffness can attract load from adjacent degraded material, a new support can change thermal reactions, a reinforcement plate can shift peak stress to its termination and a local weld can restrain deformation that was previously free. Free-body diagrams and simple stiffness estimates are often useful before detailed FEA because they expose the intended and unintended load paths.
Represent Installation Effects
Repair performance depends on how it is installed. Bolt preload, clamp fit-up, weld sequence, heat input, cure condition, bond-line thickness, interference and installation temperature can all affect the residual state. If these effects materially influence the load path or fatigue/fracture margin, include them or bound them. A model of an ideal stress-free repair can overstate benefit when the real installation introduces residual stress or distortion.
Assess the Original Damage Under the New Configuration
The original defect does not always become irrelevant after reinforcement. Load may still pass through a cracked or thinned region, and future growth can continue beneath a repair. The analysis should determine whether the damaged material is unloaded, partially shared or still critical. Inspection access after repair also matters: a solution that hides an active degradation mechanism can make future integrity management more difficult.
Check New Failure Modes
Repairs can introduce local buckling, fastener bearing, adhesive peel, fatigue at terminations, weld toe cracking, galvanic corrosion, fretting or overload of adjacent supports. These may govern before the original defect does. The substantiation should therefore include a failure-mode review of the repaired configuration, not just a repeat of the original defect assessment with higher section thickness.
Control Materials and Fabrication Quality
Repair material properties, weld consumables, heat treatment, adhesive batch, laminate architecture and surface preparation may all be safety-significant inputs. Acceptance should include the inspection or quality-control evidence needed to demonstrate that the installed repair matches the analysed configuration. Where workmanship variability is large, conservative allowables or proof/commissioning tests may be appropriate.
Update Models, Drawings and Asset Records
The integrity case can only remain valid if future engineers know what was changed. As-built repair geometry, material, installation records, concessions and inspection results should be incorporated into configuration-controlled records. The next FFS assessment should not need to rediscover a repair from field photographs. Digital models should likewise be updated where the modification changes system stiffness or loads.
Commissioning and Post-Repair Evidence
Depending on the function, post-repair evidence may include NDT, leak testing, pressure testing, strain measurement, torque/preload verification, dimensional survey or monitored operation. Testing should be designed to verify the assumptions that matter rather than performed as a generic ritual. A successful proof event does not demonstrate infinite future life; it validates a defined aspect of the repaired configuration.
Change-Control Checklist
- Repair objective and authorised period are defined — Temporary and permanent status are unambiguous.
- New load path is demonstrated — Global and local reactions are understood.
- Installation state is represented — Preload, residual stress, fit-up or cure effects are addressed as relevant.
- New failure modes are screened — Terminations, joints and adjacent structure are checked.
- Configuration records are updated — Future assessment can reproduce the as-installed state.
Decision Principle
A successful repair restores or controls structural function while leaving a traceable and inspectable integrity basis. The measure is not whether a local stress plot improves; it is whether the modified load path, materials, installation state and future degradation can be defended for the authorised service period.
Every repair is also a configuration change. The integrity case must follow the structure that now exists, not the structure that used to exist.
Engineering judgement — governing sensitivities
For Managing Repairs, Modifications & Configuration Changes in Integrity Assessments, the most useful review question is not simply whether the solver has produced a plausible contour or scalar result, but whether the model preserves the link between observed degradation, the governing failure mechanism and the acceptance criterion used for continued service. This is where apparently small modelling choices can change the engineering conclusion. The analyst should identify the variables that can move the governing response, separate physical uncertainty from deliberate conservatism, and show that the selected modelling fidelity is proportionate to the decision being supported. Where the response is close to an acceptance boundary, sensitivity cases should bracket credible changes rather than apply arbitrary percentage perturbations.