20,480 points across a 2.5 m² blade panel
Five long sheets and ten compact collectors merged into one continuous map over 1040 × 2480 mm, then compared against the aerodynamic model.
Client: a blade manufacturer in East China · name withheld under NDA · Published with the customer's consent to anonymised disclosure.
Background
A blade section had to be qualified against a structural model. Strain gauges gave the global load path, but they could not separate a genuine load peak from a localised contact or bonding effect — the two look identical in a single-point signal.
The requirement was concrete: a pressure field dense enough to locate where the load actually concentrated, over an area of roughly 2.5 m², without machining the test article.
Configuration choice
Three standard builds cover the same 1040 × 2480 mm surface. Config A uses ten 490 × 520 mm sheets for the finest granularity; Config C uses five 490 × 1040 mm sheets with five larger PXS128 collectors. This project selected Config B: five long sheets, ten compact PXS064mini collectors.
- Fewer seam lines — five sheets instead of ten halves the number of joints crossing the measured surface
- Cheaper single-sheet replacement — a damaged sheet is swapped without rebuilding the whole panel, at 4096 points per sheet
- Compact collectors — PXS064mini at 90 × 40 mm and under 100 g fits around the fixture, not on it, at the cost of two units per long sheet
- One software session — all ten collectors are grouped through the docking hub and merged into a single map
| Specification | Value |
|---|---|
| Sensing film | PD-FS, 0.3 mm, piezoresistive |
| Points per sheet | 4096 (64 × 64), active area 487 × 1034 mm |
| Pressure range | 400–6000 kPa |
| Accuracy / repeatability | ±5% FS / ±5% FS |
| Response time | < 1 ms |
| Collectors | 10 × PXS064mini, USB to one PC |
| Sampling | 20 Hz, 12 bit |
Installation sequence
Five working days on site, excluding film lead time:
- Day 1 — survey and layout. Surface scan, sheet layout drawing, cable routing agreed with the test owner
- Day 2–3 — surface preparation and bonding. Cleaning, adhesive application, sheet placement respecting the 1.5 mm sealing zone (3 mm at combined seams)
- Day 4 — electrical and acquisition. Collectors mounted off the test article, ten channels verified one by one
- Day 5 — calibration and baseline. Per-sheet calibration values stored with the project, first full-field snapshot recorded
What the data showed
The pressure maximum sat 60–80 mm inboard of the position predicted by the model, and it was localised rather than distributed. Because the map covers the whole panel, the deviation could be attributed to a specific region instead of triggering a full-panel re-examination.
Honest limits of this setup
Points worth stating plainly before anyone plans a similar project:
- 20 Hz is not dynamics. This is quasi-static and steady-state work. Blade vibration or transient impact needs a different acquisition chain.
- 10,000-cycle durability changes the plan. For long-duration cycling, sheet replacement and recalibration should be budgeted from day one.
- The film measures what it touches. Thick paint, heavy contamination or an uneven bond under a sheet reads as pressure error, not as surface condition.
- Cabling is a real constraint. Ten collectors mean ten cable runs; the layout drawing is where most of the risk lives.
Why the client is not named
A blade manufacturer will not usually let a supplier publish which section of which blade was instrumented — that single fact says something about a structure still in development. What the customer agreed to is narrower and more useful: publish the configuration, the method and the finding, leave out the identity.
So the page carries no logo, no plant name and no quote. If you are evaluating a similar project, the configuration drawing, the full pressure dataset and the deviation report can be reviewed under a mutual NDA — that conversation starts on the contact page.