What does a laminate wrinkle wind turbine blade defect look like?

A wrinkle is a region where the fibres of a laminate depart from the straight path the design assumes. Out-of-plane waviness, the out-of-plane wrinkle proper, is a fold through the thickness: the plies rise over a bump and fall again. In-plane waviness is a lateral swerve of the fibres within a ply, visible on the surface as a wave in the fabric. Both are described by an amplitude, a wavelength and a maximum fibre misalignment angle, and an aspect ratio relating amplitude to wavelength is the usual measure of severity (Wind Energy Science, 2022).

The defect matters most where the laminate is thick and loaded in compression along the fibres: the spar caps, the root build-up and the transition between them. A field study that sectioned commercial blades recorded out-of-plane waves with a mean maximum fibre angle of about 8.6 degrees and a maximum of 39 degrees, and in-plane waves with a mean of about 26.7 degrees (Nelson, Riddle and Cairns, Wind Energy Science, 2017). Those angles were measured on sections that had failed at the flaw and, as the authors note, probably include plastic deformation, so they are an upper bound rather than the as-manufactured geometry.

In inspection reports the same defect appears as fibre waviness, spar cap wrinkle, out-of-plane wrinkle, ply wrinkle or marcelling. When we compare findings across blades, the first task is to establish that the reports describe the same thing.

How do wrinkles form during blade manufacturing?

A wrinkle is a blade manufacturing defect with a mechanical cause. Fabric that is longer than the path it must follow has to go somewhere, and it goes out of plane. The literature attributes wrinkles to tooling geometry, operator error, skewed fibres, poor draping in curved regions, compaction and temperature gradients during cure (Wind Energy Science, 2022).

Ply drops and mould geometry

A spar cap tapers by terminating plies, and each ply drop is a step the plies above must climb. With heavy infusion fabrics, of the order of a millimetre per ply, the step is not small. If several plies end at one station, the covering plies bridge the step and a fold forms on the thin side. Ply drops weaken the laminate even without a wrinkle: in thin carbon-prepreg coupons tested at Montana State University, a different material system and process from infused glass, double ply drops reduced static strength by approximately 45 per cent in tension and 42 per cent in compression (Samborsky et al., 2006). Where the mould changes curvature, at the root transition or along the maximum chord, unidirectional plies bridge concave radii rather than conform to them; the gap closes under vacuum and the excess fabric becomes a wave.

Layup handling

Unidirectional glass and carbon fabrics have little resistance to in-plane shear and almost none to buckling along the fibre. A roll laid slightly off axis, a ply pulled to correct its position after tacking, a stack disturbed while the bag is fitted: each is enough to skew a length of fabric.

Infusion and consolidation

Ply wrinkle infusion defects arise during resin infusion itself. The resin front can move loosely tacked fabric, a bag leak can lift part of the stack, and uneven consolidation leaves thickness steps for the following plies. Thick unidirectional laminates are the most exposed, because the length mismatch between inner and outer plies over a bend grows with thickness. In a spar cap tens of millimetres thick the wrinkle can be entirely internal and invisible at the surface, which is why acceptance depends on process discipline rather than final inspection alone.

Why does a spar cap wrinkle cut compressive and fatigue strength?

A unidirectional laminate carries compression through fibres supported laterally by the matrix. Misalign the fibres and the load acquires a component that shears the matrix and rotates the fibres further, a self-reinforcing process that ends in a kink band. Fibre misalignment therefore affects compressive strength far more than tensile stiffness.

The wrinkle knock-down is well documented in coupon tests. In the field study above, coupons with in-plane waves loaded in compression reached 37 to 44 per cent of the control strength (Nelson, Riddle and Cairns, 2017). Laminates with artificial out-of-plane wrinkles of two severities lost 54 and 37 per cent of their stiffness in the wrinkled section (Wind Energy Science, 2022), and a review of blade failure mechanisms reports a fall in strain to failure from about 1 per cent for straight carbon fibres to 0.6 to 0.8 per cent for wavy ones (Mishnaevsky, Materials, 2022).

Two cautions apply when carrying these fibre waviness spar cap numbers from coupons to a blade. First, the coupon knock-down is conservative: the flaw crosses the whole specimen width, whereas in a blade the surrounding laminate can redistribute load around a local wave (Nelson, Riddle and Cairns, 2017). Second, fatigue governs service life. The damage sequence seen in tests is matrix cracking at the ends of the wave, delamination that frees the plies to straighten, and fibre failure at the peak. Under cyclic load that delamination grows long before ultimate strength is reached, so a spar cap wrinkle that survives the extreme load case can still fail after years of ordinary operation.

How are laminate wrinkles found and measured?

The reference method is destructive. A cross-section cut through the wrinkle, polished and photographed, gives amplitude, wavelength and maximum angle ply by ply, and it is the evidence on which an acceptance decision or a root cause analysis ultimately rests.

Ultrasonic phased array is the practical method for spar caps. With the right probe frequency the length and deviation of a wrinkle can be evaluated, and the method is applied in manufacturing as well as in service (Wind Systems Magazine). In our experience its sensitivity falls with depth in thick glass laminate, so the deepest part of a spar cap is the least well covered.

Active thermography and shearography are fast over large areas but say little about depth or angle. X-ray radiography and computed tomography resolve the fibre paths directly, but on blades they are confined to cut-outs and laboratory coupons. Internal visual inspection, by technician or crawler camera, sees surface waviness and the resin-rich ridge a large wrinkle often leaves, but not a wrinkle buried under straight plies. A clean internal report is not evidence that a spar cap is free of wrinkles.

What are the acceptance criteria, and why does one wrinkle repeat across a batch?

Blade design standards require the designer to account for manufacturing tolerances in the design strength. Both IEC 61400-5 and DNV-ST-0376 place that requirement on the certified design, with an explicit focus on the effect of manufacturing tolerances on design strength (DNV). The manufacturer's own specification then defines what may remain in the laminate: typically a maximum fibre angle or aspect ratio and rules by region. A wrinkle inside that limit is one the strength calculation already assumed; one outside it is a defect the calculation did not cover.

Whether a wrinkle is isolated or serial turns on why it was there. A random cause, a single handling error on one shift, produces one wrinkle in one blade and does not repeat. A process cause, such as a new fabric lot, a drift in tacking or vacuum practice or a revised work instruction, produces the same wrinkle at the same station in every blade made while the cause persisted: one plant over one production window. A design cause, a ply drop schedule or transition geometry that cannot be laid without bridging, produces the wrinkle in every blade of that type from every plant that follows the drawing.

The second and third are serial defects. The signature is repetition: the same detail, span position and morphology in cross-section. Establishing it is a matter of manufacturing evidence: ply books, traveller records, non-conformance reports, mould and station allocation, dates and supplier lots. A finding in a second blade at the same station is suggestive. A cause identified in the plant that explains every finding is conclusive.

How is a fleet-wide blade defect exposure established and handled?

Once a serial defect is suspected, the owner, the manufacturer and the insurer each need the same three things: a defined population, a measured condition and a decision rule.

Checklist: defining the exposed population

  • The plant, moulds and production window behind the blades with findings, and what changed at its start and end.
  • The serial range that maps to that window, including blades shipped to other sites and spares.
  • Whether the cause is process-driven, limited to that plant and window, or design-driven and present in every blade of the type.
  • What inspection each blade in the range has already had, and whether that method could have seen the defect.
  • Which contracts attach to the range: supply agreement, warranty, serial defect clause and operational insurance.

The inspection campaign and the decision

The campaign has to be designed to answer a question, not merely to look. That means a method that can detect the morphology already seen in cross-section, at the depth at which it occurred, demonstrated on a reference sample before the field work starts. A sample inspected by a method that can see the defect is worth more than a whole population inspected by one that cannot.

Each blade then ends in one of four outcomes: accept, within the design allowance; monitor, with a re-inspection interval and a trigger for action; repair, where a repair design can restore the load path; or replace, where it cannot or where the number of findings makes repair uneconomic. The population decision follows from the blade decisions, on structural grounds first.

Warranty and insurance positions

Blade warranty serial defect provisions usually define a serial defect as the same defect occurring in more than a stated fraction of a delivered population within the warranty period, and extend the remedy from the failed blades to the whole population once that threshold is met. The threshold and remedy vary between agreements and should be read before the campaign is designed, because how the population is defined and how findings are counted can decide whether the threshold is reached. Operational insurance policies commonly exclude the defect itself while covering resultant damage, so one finding can be a warranty matter for the wrinkle and an insurance matter for the rupture. In a transaction, an open serial defect question on a blade type is a due diligence red flag even before any failure.

Frequently asked questions

Is every laminate wrinkle wind turbine blade defect a serial defect?

No. A single wrinkle with a random handling cause is an isolated defect. It becomes a serial defect when the cause lies in the process or the design, so that the same wrinkle repeats at the same station across a production batch or a blade type. The distinction rests on manufacturing evidence, not on the severity of the first wrinkle.

How big does a wrinkle have to be before it matters?

It depends on where it is, how deep it is, what load the station carries and what the design assumed. Coupon data show that compressive strength falls steeply with the maximum fibre angle, so angle and aspect ratio matter more than length. The manufacturer's acceptance criteria, derived from the certified design, set the limit for each region of the blade.

Can an out-of-plane wrinkle be found from inside the blade?

Sometimes. Internal visual or crawler inspection sees surface waviness and resin-rich ridges on the inner face of the spar cap, and is a useful first screen. A wrinkle buried under straight plies is not visible from the surface, and confirming its absence needs ultrasonic inspection or a cut-out.

What evidence establishes a serial defect for a warranty claim?

Repetition with a common cause: cross-sections or NDT findings from more than one blade showing the same morphology at the same station, and manufacturing records that explain why. The contractual definition then determines whether the count of affected blades meets the serial defect threshold.

How Apex Wind can help

We are an independent blade engineering consultancy in Denmark. When a wrinkle has been found, or a blade has failed and a wrinkle is suspected, our blade failure investigation work establishes the mechanism from the failed structure and the manufacturing evidence, states whether it is isolated or serial, and defines the exposed population. Where a blade type with a known wrinkle history is part of an acquisition or a refinancing, our technical due diligence puts the serial-range exposure into terms the transaction can use. We sell no repairs, hardware or inspection products and take no referral fees, so our conclusions are our own. Contact us to discuss a finding.