Leading edge erosion or coating failure?

Both show as loss of the outer surface along the leading edge, and both are often reported as erosion.

Erosion is driven by droplet impact, so it follows relative velocity. It starts outboard, typically beyond half radius and concentrated in the outer third, and it is worst at the stagnation line, spreading symmetrically to either side as it progresses. It begins as isolated pinholes, which grow and merge into patches, then into a continuous stripped band. The edges of the damage are ragged. Severity increases towards the tip in a gradient.

Coating failure, by contrast, is a bond or application problem, so it follows the process rather than the physics. It appears where the coating was applied badly: in patches with sharp, often straight or curved boundaries, at repair edges, at masking lines, around the root where the coating system changes, and frequently inboard where erosion cannot explain it. It lifts and flakes in sheets rather than pitting. It can appear at any radius, and it does not show a tip-ward gradient.

The distinction changes who pays. Erosion is wear, and its rate is a site and blade design matter. A coating that has delaminated in sheets at 40 percent radius is a product or application defect, and if it appears on many blades from the same plant, it is a serial defect.

Lightning damage or impact damage?

Both produce localised damage near the tip.

Lightning leaves heat and pressure signatures. Look for scorching and soot, a puncture or burst rather than a dent, delamination radiating from the attachment point, split trailing or leading edges near the tip where internal pressure vented, melted or pitted receptor surfaces, and damage to the down conductor or its connections. Damage away from a receptor is the important finding: it means the protection system did not capture the strike, which is a system question under IEC 61400-24 (IEC 61400-24:2019), not simply a repair.

Impact, from hail, debris, a bird or a transport contact, leaves a mechanical signature: a compression mark, crushed core under an otherwise intact skin, no soot, no burning, no receptor involvement, and no internal damage away from the point of contact.

A strike log or a lightning detection record puts a date against the first. Its absence does not rule it out; the network does not catch everything, and blade attachments are not always logged by the turbine.

A gelcoat crack or a structural crack?

This is the pair that carries the most risk, because the surface appearance is similar and the consequences are not.

Coating and gelcoat cracks are shallow, tend to be short, often craze in networks or run in fine parallel lines, and do not follow a structural feature. They frequently appear over thick laminate or at geometry changes where the coating is thick and the substrate is stiff. They do not open under load, and they do not grow steadily year on year.

Structural cracks follow the structure. The features that should escalate a finding:

  • orientation transverse to the blade axis, over or near the spar cap, on either surface
  • location at a ply drop, a thickness transition, a repair boundary or the root transition
  • a crack that runs off into the trailing or leading edge, rather than stopping in open shell
  • fibre visible in the crack, or any opening with depth
  • a measured increase in length since the previous inspection
  • a matching feature on the opposite surface or visible from inside

A crack that runs along the spar cap edges on the outside frequently marks the boundary of the laminate underneath. It may be a coating effect over a stiffness change, and it may be the surface expression of a delamination. It is the classic case where the image cannot decide, and the answer comes from tapping, from ultrasonic testing, or from looking inside.

A bond line crack or a laminate crack?

Both appear as a line of damage, often near an edge.

Bond line cracks run along a joint, which means they are straight over long distances and follow the geometry of the shell join, the shear web foot or the leading edge bond. They appear at the trailing edge near maximum chord more than anywhere else, because that is where edgewise fatigue loads the bond. The crack faces show adhesive, sometimes with a clean separation from one substrate, which indicates the bond never formed properly rather than that it fatigued.

Laminate cracks cross fibre and follow load, not geometry. They may start at a bond but then turn into the shell. The crack faces show broken fibre.

Distinguishing them decides the repair and the cause. An adhesive separation with a clean laminate face on one side is a process defect: contamination, an unprepared surface, an adhesive that did not wet. A fatigued bond with arrest marks along its growth is a design or load question. These are different conversations with a manufacturer.

Old repair or new defect?

Inspection reports routinely record a repair as damage, and occasionally record damage as a repair.

A repair shows a boundary: a scarf ring, a change in surface texture or gloss, a visible patch edge, filler, and often a colour difference that is more obvious in flat light than in sun. What matters is not the repair itself but its edges. A crack restarting at the boundary of a previous repair is one of the more serious findings available, because it says the repair did not restore the load path or did not remove the original defect.

This is also where the maintenance log earns its place. A repair found on the blade and absent from the log is a finding in its own right, and it is common after transport and installation damage that was made good on site.

What the image cannot tell you

Some distinctions cannot be made from photographs at all, and a report that claims them is overstating:

  • depth of erosion, beyond the coarse question of whether laminate is exposed
  • whether a surface crack has a subsurface delamination behind it
  • wrinkles, dry fibre and porosity inside a spar cap, which need ultrasonic testing or a cross-section
  • bond void extent behind an intact surface, which needs tapping, thermography or an internal inspection
  • whether a crack is growing, without a previous measurement to compare against

The correct entry in a report for these is that the finding requires a closer method, with the method named. An inspector who writes "no structural damage observed" from external images alone has reported the limits of the method, not the condition of the blade.

Frequently asked questions

Can lightning damage be confused with manufacturing defects?

Yes, and it matters commercially, because policies and warranties treat them differently. A strike can open an existing weakness, so the same blade can show a lightning event and an underlying defect. The sequence is established from the fracture surfaces and the internal condition, not from the external photograph.

How do you tell a serial defect from ordinary damage?

By repetition with a signature. Ordinary damage scatters across a site. A serial defect appears at the same radius, on the same surface, at the same structural feature, on blades sharing a plant, mould set and production window. One blade is a repair; the same finding at the same place on many blades is an exposure.

Is thermography useful for distinguishing these?

It is useful for finding subsurface voids, delaminations and water, which the camera cannot see, and it complements visual inspection rather than replacing it. Its results depend heavily on conditions and on the operator, and it should be specified with an acceptance procedure rather than bought as a service.

Does a crack that has not grown in three years still need repair?

Not necessarily on urgency grounds, but it still needs a category and a documented decision. A stable crack is evidence about the load, not a guarantee. Stability should be recorded as a measurement, so that the judgement can be revisited if the growth rate changes.

How Apex Wind can help

We are an independent blade engineering consultancy in Denmark. We review inspection findings and establish what the damage actually is through blade failure investigation and root cause analysis, including the laboratory work where fracture surfaces and cross-sections settle the question. We run blade training for inspection, engineering and claims teams who have to read these findings themselves. We sell no inspections, repairs or hardware and take no referral fees. If a finding is ambiguous and the decision is expensive, contact us.