The wooden bar (a rolling pin) was used as a foot brace, the large wrench as a two-handed handle. I exerted enough force to stretch the cord by about 25%. I estimate the force was easily 100 lbf, probably significantly greater. The cord under test was kept shorter than the distance between my hands and the foot bar by the heavy grey rope shown, (5/16 inch / 8 mm polyester braid) limiting the energy storage in the test cord for safety, and allowing convenient use of a ring as a point of attachment for the test cord. The grey rope was attached at both ends with anchor bends. The test cord was attached to the ring with an anchor bend. The test cord was held by the wrench by five or six turns pressed under one of the two hands gripping the wrench.
For untying, I used only fingers and fingernails. In their working forms, each of these bends has a collar around each of its two standing parts. (The standing parts are the long bits sticking out of the knot. The short bits are the working ends) The untying strategy was to nudge the collars around the standing parts away from the knot toward the ends of their respective standing parts-- toward the standing ends.)
I performed 10 tests on each of the bends. For each bend, for half of its tests, the knot was carefully set up into compact working form before each test. For the other half of its tests, the knot was left slightly loose (in particular, the collars around the stading parts were left loose.). I reused each pair of test cords 2 or 3 times. Beyond the measures just described, the bends were tied by the commonly known procedures for tying them, as portrayed at the linked descriptions, paying no particular attention to the detailed shape of the result except to ensure that the resulting bends looked generally the way someone familiar with each bend, but not expert would expect it should look. In particular, no attention was paid to the orientation of the working ends, except for the expectation that the knot as a whole should look as it familiarly does.
Results:
Out of 10 tests on each bend:The Ashley bend, ABOK 1452, jammed rock hard 10 times. The bend had to be cut out of the cord if I intended to reuse it.
The Carrick bend ABOK 1439, jammed zero times. It was always easy to untie.
The Zeppelin bend jammed zero times. It was always easy to untie.
The Alpine Butterfly Bend jammed zero times. It was always easy to untie.
The Hunter's bend , ABOK 1425A, jammed rock hard 10 times. The bend had to be cut out if I intended to reuse the cord.
None of the these bends showed any slip at all after residual looseness had been taken out of them. I tried to hold maximum tension for at least 5 seconds in each test.
Conclusions:
Ashley's bend and Hunter's bend often jam when loaded to the point where substantial stretching occurs in the cord in which they are tied.The Carrick bend, the Alpine butterfly bend, and the Zeppelin bend, either don't jam at all, or jam so infrequently that a jam could not be observed in these tests. These three bends seemed to be equally easy to untie in the rather extreme circumstances of these tests, namely very easy. The collar around each standing part nudged easily away from its bend toward its corresponding standing end, with just a little easy encouragement from a fingernail, after which the bend could be pulled apart easily using the loosened collars as handles. The ease of untying these bends in these extreme circumstances suggests that if they ever jam, it must be in very particular or unusual circumstances.
Conversely, the ease with which rock hard jamming was produced with Ashley's bend and Hunter's bend, suggests that these bends will often be hard to untie when the cord or rope in which they have been made has been heavily loaded, but not heavily enough to jam them rock hard.
Thanks to roo for suggesting this experiment.
Back to context Home.