Knot testing, part 2

Knot testing, part 2

Tracy&James | Thursday, 2 November 2017

Continuing from last week’s FP on tippet knot testing I have some more data to share. First up, some myth busting – we’ve all read BS about 100% knot strength from certain quarters, however based on last week’s numbers, where a simple overhand ‘wind knot’ loses 40% of the nylon’s strength, then 100% is surely an impossibility?

The most widely talked about ‘perfect strength’ knot is the Bimini twist and it was suggested that I should test two of these loop-to-looped. I must admit to being more than a bit rusty when it came to tying these loops, in fact my first 4 or 5 efforts went straight in the bin as I didn’t think they were worthy of a test. Once I’d figured out to clamp the spool between my knees for tensioning, thus giving me two hands free to hold the loop, compress the twists and feed the tag end back in, then I started to produce some reasonably good looking knots. They did still have the odd stray twist in the loops but they were, in my mind, acceptable to test.

So on to the Instron results – as I write this I have my notes in front of me.  Immediately next to each of the Bimini loop-to-loop tests I have scrawled an asterix, and off to the side of the page I’ve written the following – “* - did not fail at the knot”!

So perhaps this is the wonder knot as others have stated - especially if my tyings, after not much practice, all held firm.  As such, the failure strengths recorded against the Bimini twists is the point at which the nylon material failed and not the knot.  Compared to the control group the average was over 97.3% of the UTS - this 2.7% reduction could just be a statistical aberration arising from not enough data or it could be because of the residual twists (that I mentioned above) being forced from the knot into the main line during the test and weakening it very subtly.  Note – in a fishing scenario these twists would not be confined as they are on my tester, I’d therefore have to agree that the Bimini twist is very likely a 100% strength knot. (I also suspect that given more experience in tying the knot I’d eliminate the residual twists).

On to some other knots.  I’ve added the 2-turn water knot (or Surgeons knot) for comparison with the 3-turn version that I tested last week.  The results were remarkably similar, both in terms of the range of UTS figures recorded and the standard deviation of the results.  Both these figures concern me, mainly because this is my go to knot for adding a tippet to a bonefish leader, as my results indicate there is a pretty good chance that I will end up with a below 50% strength connection.  Now this could be down to how I tie these knots, but I’d prefer to think that I’d get such a simple procedure right every time.  As such I’m planning on changing to use something else.

An obvious candidate for replacing the water knot is the figure-of-eight knot as it’s very simple and quick to tie (incidentally during this study I’ve come to realise how many different names the same knot has across the world).  Something interesting happened during the testing of this knot; the initial tests were all quite good, typically 67% plus strength, but then I got a low result which didn’t seem to fit in.  As such, I decided to run some more samples, after a couple that seemed to confirm the good results I again got a low result, and then another.  I ended testing sufficient samples to convince me that there was two populations of results – ‘good’ ones with an average UTS of ~70% and ‘bad’ ones with a UTS of ~ 56%, with both these populations having a fairly small standard deviation.  Now the bad news for me was that I absolutely could not predict which group (i.e. good or bad) any individual knot would be in – to me I tied, wetted and pulled them tight in exactly the same manner, and bear in mind that this is an extremely easy knot to tie.  The figure-of-8 knot therefore has the potential to meet my needs but not before I understand the cause of the bad examples.

Next up was the good, old fashioned blood knot.  This was probably the first ‘proper’ fishing knot I learnt to tie.  The results were amongst the best I’ve seen so far (excluding the Bimini twist) with an average UTS of just over 70% and an acceptably low standard deviation.  A feature of these tests was that they all contained ‘glitches’ in the load/extension response, this was accompanied by a visible ‘twitch’ of the knot whilst under stress.  I strongly suspect that as the nylon is drawn it naturally thins and the barrel of the knot is then able to pull tighter once friction is overcome.  This momentary ‘slip’ is recorded as a drop and recovery in the force applied.


So where am I with regard to conclusions at this point?  Well, despite its 100% strength, I won’t be using the Bimini twist whilst out on the flats.  This is purely for practical reasons – I don’t think I could tie one stood in thigh deep water with nothing else to tension the tippet other than my two hands.  So I’ll save this one for parts of my leader set-up that won’t need replacing (perhaps my shark leaders).  The next best option to date is the blood knot, I intend adding to the data with this knot plus trying it in fluorocarbon.  I’d also like to try and understand the behaviour of the figure-of-8 knot, as if the ‘bad’ results can be weeded out this offers a good compromise between strength and ease/speed of tying.


I shall also be moving on to knots that I can use via a tippet ring in due course, no doubt this will further muddy the waters.


All the best, James.