Background
A curing press platen is never perfectly flat — and it is least flat under load at temperature. The usual check, whether paper pull or a feeler gauge at the edges, samples a handful of points and says nothing about the middle of the plate.
That matters when a process change is being decided. Thickness variation and void content follow the pressure field, not the nominal tonnage. If the field is unknown, the only way to settle the question is to run parts and inspect them afterwards — slow, destructive, and statistically thin.
The requirement: flatness of the loaded platen over the full working area, in one measurement, before committing to the change.
Configuration choice
Config A — ten 490 × 520 mm sheets, ten PXS064mini collectors, 40,960 points over 1040 × 2480 mm. It is the finest granularity of the three standard builds, and it is the right one here because a platen is a flat field rather than a strip.
- 7.6 × 8.1 mm point pitch — fine enough to see a local low spot, not just a regional trend
- 40,960 points — roughly an order of magnitude more than a strip layout, at the cost of more seams
- Nine internal joints — accepted deliberately: on a flat plate the seams can be laid over solid backing, and their effect is repeatable rather than random
- Compact collectors — PXS064mini at 90 × 40 mm sits clear of the press frame
| Item | Value |
|---|---|
| Sensing film | PD-FS, 0.3 mm, piezoresistive |
| Sheets / points | 10 × 4096 = 40,960 points over 1040 × 2480 mm |
| Point pitch | 7.6 × 8.1 mm |
| Pressure range | 400–6000 kPa, configurable |
| Accuracy / repeatability | ±5% FS / ±5% FS |
| Collectors | 10 × PXS064mini (90 × 40 mm, < 100 g) |
| Sampling | 20 Hz, 12 bit |
| Operating temperature | −30 to 70 °C, 10–90% RH |
How the measurement was taken
- Sheets laid over the working area, edge to edge, respecting the 1.5 mm sealing zone and 3 mm at combined seams
- Load applied at production pressure, then held while the full field was recorded — the interesting state is loaded, not open
- The load case repeated so the pattern could be separated from noise: a repeatable field is a property of the tool, a one-off low reading is a sample artefact
- Per-sheet calibration stored with the project, so a later re-check is comparable
What the data showed
The deviation was not random — it followed the platen's own structure. Some regions sat above the target band and others below, and the pattern repeated across runs. That is what makes a process decision possible: a repeatable field can be corrected at a named location, while a single low reading can always be argued away.
The change was therefore decided on a map of the loaded platen rather than on a sample of finished parts.
Honest limits of this setup
- The film is 0.3 mm and it is in the gap. On a platen this matters most: the reading is the pressure across the film, not the metal-to-metal gap. It should be accounted for in the tolerance budget
- Temperature is a real constraint. The film is rated to 70 °C. A hot platen needs a controlled cool-down measurement, or the result carries an offset
- Seams cost accuracy. Nine joints are nine places where a reading is approximate; if a low spot lands on a seam, re-check it with a shifted layout
- Static, not dynamic. 20 Hz captures press closure and hold, not press bounce or ram dynamics
- Edge effects. Pressure at the very edge of the sheet is not as reliable as in the middle — plan overlap, not abutment
Why the client is not named
Publishing which press and which process step was qualified would say something about a production line still being tuned. The customer agreed to publish the configuration, the method and the finding, and to keep the identity out.
No logo, no plant name, no quote. The layout drawing, the full-field dataset and the deviation report can be reviewed under a mutual NDA — start on the contact page.