A full-scale blade static test produces a clean number: the rig applied this force, at this span position, and the blade did not fail before the target. The number is true and it is also incomplete, because the blade never sees "a force". It sees that force divided across whatever the saddle or the pad is touching, and the division depends on the fit between a curved composite shell and a steel fixture that was built to a drawing.
When that fit is poor, the test still passes its load target and the strain gauges still report a plausible field. What changes is margin. The peak contact pressure under one edge of a saddle can be several times the average, and local compressive damage — a crushed core, a delamination starting under a pad — appears at a load the analysis said was safe. The failure then gets attributed to the blade, when the load introduction was the variable.
Three ways a rig puts load into a blade
The mechanics differ enough that they should not be discussed as one thing.
Saddles and cradles
A saddle wraps the blade over a span of a metre or more and pushes normal to the surface. Contact is nominally distributed, but the shell is curved and the saddle is not perfectly matched to it. The result is load concentrating along one edge of the wrap, at the point where the two curvatures disagree most. The blade is supported everywhere and loaded nowhere near evenly.
Loading pads and hydraulic actuators
A pad introduces load over a small area, often deliberately, to represent a concentrated lift or a mooring point. Here the peak is expected to be high, and the question is whether it is high in the intended place and whether the pad edge is digging in. A pad that is slightly too small, or that is allowed to tilt, produces an edge load that the design never accounted for.
Root fixtures and embedded inserts
At the root, load goes through a steel hub into a laminate that carries it through embedded inserts or a bonded collar. The interface is inside the structure and is the hardest of the three to inspect. Bolt preload distribution and the flatness of the mating face decide whether the laminate around each insert sees the load it was designed for, and this is where a fixture that is out of flat by a fraction of a millimetre changes the local stress state.
Why the force reading does not catch it
The load cell in the rig sits upstream of all of this. It measures the total force it delivers, and it is usually accurate to a fraction of a percent. The quantity that decides whether the blade is damaged is a pressure distribution two mechanical interfaces downstream, and nothing in the standard instrument set measures it.
Strain gauges do not close the gap either. They report the response of the structure at the points where they are bonded, and they are excellent at that. But a gauge bonded 200 mm from a saddle edge reports the strain that arrived there, not the pressure that caused it. If the saddle is loading on an edge, the gauge will show a concentration and the natural interpretation is "the blade is locally stiff here" rather than "the fixture is loading on an edge". Both explanations fit the same trace.
| What you have | What it tells you | What it cannot tell you |
|---|---|---|
| Rig load cell | Total force delivered by the actuator | How that force divides across the contact |
| Strain gauges on the shell | Local structural response where they are bonded | Whether the cause is the structure or the fixture |
| Actuator pressure | What the hydraulics were commanded to do | Anything downstream of the pad |
| Contact pressure film or array | The field between the fixture and the blade | Whether the structure would have coped anyway |
These are complementary, not competing. The point of adding a pressure map is that it sits at the one interface every other instrument is blind to.
How to measure at the load introduction point
The method is the same one used for surface pressure on a loaded blade, applied to a different interface: a thin sensing layer between the saddle or pad and the blade surface, read while the load is applied. Three practical details decide whether the reading is worth having.
- Thickness budget. A 0.3 mm layer is not nothing when it sits between a steel saddle and a composite shell. It changes the local stiffness and slightly redistributes the contact it is measuring. For a test whose purpose is to characterise the fixture, that is acceptable and repeatable; for a qualification test where the blade's own response is the deliverable, it has to be justified or removed.
- Coverage versus resolution. A saddle contact is long and narrow. Covering the whole wrap at a coarse pitch tells you which edge is loading; covering a 300 mm section at a fine pitch tells you the shape of the peak. The first answers "is my fixture wrong", the second answers "how wrong". Start with the first.
- Fixed read time. Composites under a saddle creep, and so does the sensing layer. Read at the moment the target load is reached, on every test, or the comparison between two fixtures is really a comparison between two hold times.
The measurements that change decisions most often are not taken during the qualification test at all. They are taken during rig set-up, at 20–30 % of target load, on a dummy or on the first article, where an uneven saddle shows up clearly and can still be shimmed. Waiting until the blade is at 100 % to discover that a saddle loads on one edge is an expensive way to learn it — and whether that check needs a wired array or a sheet of single-use indicating film is itself a decision worth making deliberately.
Where this advice stops
A pressure map at the load introduction point does not certify a blade test. It characterises one interface at one moment in the loading sequence. It cannot replace the load calibration that the standard requires, and it does not tell you whether the blade as designed will pass.
It also does not resolve the root fixture properly. Embedded inserts and bonded collars put the interesting interface inside the laminate, and a surface film cannot see it. For root details the relevant measurements are bolt preload distribution, face flatness and strain in the surrounding laminate — different instruments, different question.
Finally, the film changes the thing it measures, and at load introduction this is not a rounding error. A 0.3 mm layer between a saddle and a shell moves the contact patch. Where the purpose is to compare two saddle designs, that shift is the same for both and the comparison holds. Where the purpose is to reproduce the load path of a service condition, the layer should be characterised and then removed.
The habit worth building
Most blade test programmes measure the blade thoroughly and the fixture hardly at all. The asymmetry is understandable — the blade is the article under test — but it is backwards in one respect: the fixture is the part that is easy to change. If a saddle loads on an edge, a shim and a re-survey fix it in an afternoon. If the same condition is discovered as a delamination after the test, the conclusion is a blade problem, and a blade problem costs months.
Recording contact pressure at the load introduction points does not make the test more accurate. It makes the test attributable: when something does go wrong, you know whether you are looking at the article or at the rig.
Frequently asked questions
Can the film stay in place for a full static test to failure?
It can be in place, but it will not survive the failure event usefully, and at high loads the layer itself becomes part of the interface. The useful reading is the field at target load, not the last frame before a bang. Plan the measurement around the load steps you care about.
How many load introduction points should be instrumented?
Start with the one that is hardest to fit and the one closest to the root. Those are where curvature mismatch and stiffness gradients are largest, and therefore where an uneven contact is most likely. Adding all of them on the first campaign usually produces more data than anyone reads.
Does this replace a pressure-sensitive film check?
No — it is the same family of method with a different instrument. Single-use indicating film is perfectly adequate for a one-off "is this saddle loading on an edge" check. A wired array is worth it when you need to compare load steps, compare fixtures, or keep a record across a campaign.
What load level should the check be taken at?
Low enough to be safe, high enough for the contact to be established — typically 20–30 % of target. Below that, the saddle may not be fully in contact and the map describes the approach rather than the working condition.