Ring Crush to Edge Crush (Whitsitt Relation)

Sometimes the components were tested and the combined board was not. Ring crush — a strip of paper rolled into a short cylinder and crushed on end — is a standard component test, and this regression estimates the combined board's edge crush value from the ring crush of the two liners and the medium.

The relation carries an ugliness that McKee's does not, and the page states it on its face. McKee's constants are dimensionless because the exponents were chosen to cancel the lengths, so 5.874 is a pure number that works in any consistent unit set. This one cannot do that trick, because the "+ 12" is an ADDITIVE term. An addition inherits the units of the thing it is added to, and no choice of exponent anywhere can make it dimensionless. It is 12 pounds-force per inch, which is 2101.5 newtons per metre, and if you find this equation quoted without its units attached — as it very often is — that source has published a number that is simply wrong for most of its readers. This site stores the constant in SI and converts your input, so the answer is right whichever unit you work in.

Two things to get right before the sum goes in. The medium's ring crush must be multiplied by its flute take-up factor: a flute of medium contains more paper than the flat distance it spans, and leaving the take-up out understates the sum by roughly a third of the medium's contribution. And the sum is a total — the regression only ever sees one number — so it is indifferent between a strong liner with a weak medium and the reverse, while a real board is not.

Treat the result as an estimate of a test, not as a test. It is a fitted line through one population of board, and it knows nothing about your combining adhesive, your flute geometry, or how hard the board was crushed on the way out of the corrugator. Where an ECT figure actually matters — which is anywhere a compression prediction depends on it — test the combined board.

And now the thing that ought to be better known: there is no current ASTM test method for ECT at all. D2808 was withdrawn in 1991 and nothing replaced it. What exists in practice are four preparations — waxed, unwaxed, necked-down and clamped — and THEY DO NOT CORRELATE. ISO 3037's own introduction says as much: the non-waxed method reads lower than the waxed, neckdown and clamp methods. The consequence is worth stating carefully, because it is routinely muddled. Converting an ECT figure between pounds-force per inch and kilonewtons per metre is a UNIT conversion, exact, and this site does it. Converting between test methods is not a conversion at all — there is no factor, nobody has one, and an ECT number without its method attached is incompletely specified.