On a curved blade surface, a flat sensing film reads low at the edges

A blade shell is curved in two directions. A flat sensing sheet pressed against it touches along the crown of that curvature and lifts away toward its own edges. Nothing is wrong with the sensor or with the clamp: the geometry does the lifting, and it does it predictably.

What the map actually shows

Pressure falls toward the edge of the sheet. Part of that fall is real — contact pressure genuinely drops where the surface curves away. Part is the film never being pressed, so the reading is the pressure of a gap. From the map alone the two are indistinguishable, which is why a near-zero at the edge should be read as "no information here", not as "no load here".

The sealed edge makes it worse

The seal around the sensing area cannot bend. Over a radius of a few hundred millimetres its outer fibre sits above the laminate by an amount set by the radius and the seal width, so the last few millimetres of the map are the least trustworthy part of it. This is the same stiff-step problem that appears when a vacuum bag crosses a repair step or a scarf toe.

Three ways to get a usable map

Cut the sheet so the sealed edge falls outside the contact patch. Where the clamp allows, add a conformable pad — it redistributes load, but it also changes the interface, so use the same pad in test and in reference. And read the map for gradients while a load cell carries the total force; the two together give shape and magnitude, which neither gives alone.

Curvature is part of the part

On a strongly curved section, treat edge-zone values as lower bounds and say so in the report. The useful move is not to eliminate curvature but to know where a pressure map stops being a measurement — and on blade surfaces that boundary is the edge of every sheet.