Fly Fishing - Comparison of the Trilene Knot with the Improved Clinch Knot

AvidDavid · Location: Long Island, NY

Problem: Several knots have been developed to attach a fly to the end of a length of tippet material. For several years, I have been using the improved clinch knot; but, recently, having read Lefty Kreh’s book (Kreh, Lefty (1992). Advanced Fly Fishing: Secrets of an Avid Fisherman, Bantam Doubleday Dell Publishing Group, Inc, New York), I became aware of another knot purported to be stronger than the improved clinch knot. Since Lefty’s book predates my use of fluorocarbon monofilament, I thought that I would test the knot strength of these two knots tied using fluorocarbon material, “head to head” as described in Lefty’s book.

Purpose: the purpose of my study was to determine which knot created a stronger, more durable junction between the eye of the hook and fluorocarbon tippet material, by comparing knot strength for the Trilene Knot as compared with the Improved Clinch Knot. This was done using one piece of leader material between two “identical” hooks, determining which knot failed more frequently when tension was applied to the fluorocarbon and knots between the two hooks

Methods. In order to perform a fair (unbiased) comparison, two “identical” size 6 Mustad Wet Fly Hooks with a turned down eye were obtained from a rather old stock in my old fly tying materials (from over 20 years ago). The hooks, while old, were clean, unrusted and still maintained in the plastic packaging in which they were purchased.

The fluorocarbon monofilament leader material (Berkley VANISH 8# 0.25 mm diameter) was chosen rather than tippet material of smaller diameter because I believed that it would be less prone to damage during knot tying and less prone to being nicked during the procedures. This material was purchased last summer so I am guessing that it is probably between 1 and 2 years old (since production). (In searching for materials to use for this study, I found some nylon monofilament line greater than 20 years old that had been maintained in its original metal container. It was extremely easy to break. It clearly was not of the same strength as its original labeling indicated…therefore, I believe that nylon monofilament deteriorates within decades; becoming much, much weaker.)

The procedure was as follows. Each hook was grasped by a pair of hemostats in order to apply firm tension upon tying knots. To one hook, the fluorocarbon leader material was tied using a Trilene knot. The leader material was shortened to roughly 1- 1.5 feet in length. To the other hook, the free end of the leader material was tied using an improved clinch knot. Great care was used to assure that a specified number of turns of tag end around the standing part of the leader material in forming the knots. Tension was applied to snug each knot as close as possible to the eye of the hook for each of the two hooks. Tightened coils of monofilament were forced along the leader material, as close as possible to the eye of the hook. Once both knots had been tied and snugged, the hemostats grasping the hooks were held, one in each hand, and the two hooks were pulled apart, maintaining the shank of each hook nearly parallel to the shank of the other hook. Tension was slowly increased until one knot gave way (broke), leaving one knot intact and the other knot damaged or missing. The knot that remained intact was considered the winner or survivor of that match. Once that trial was completed with one piece of fluorocarbon leader material, the winner was noted, and all fluorocarbon leader material was removed from both hooks. New knots were tied using the same hooks and new leader material. This entire sequence was repeated 10 or 20 times to gain more information concerning knot strength and so that statistical tests could be applied to determine whether one knot was significantly stronger than the other.

Results
Qualitative information: The improved clinch knot, as the knot that I have used most in the past was constructed quite easily by holding the hook with a pair of hemostats and twirling the hemostats to create the desired number of turns prior to inserting the tag end of the tippet through the loops created during tying. The improved clinch knot was snugged up to the eye of the hook by tightening the knot loosely (to about a length of 3-5 mm along the leader material. The untightened knot was wetted with saliva and then the knot was snugged by applying tension to the leader end (the tag end was allowed to be untouched). The improved clinch knot snugged very nicely each time creating tightened coils looping around the standing part of the leader. The improve clinch knot showed a propensity toward loops of monofilament jumping back around the eye of the hook (encircling the hook shank rather than the leader). This could be remedied prior to snugging the knot.

In contrast, the Trilene knot was more difficult to tie. This knot required that the leader material be fed through the eye of the hook, twice = two times. Once the knot was grossly constructed, it was tightened slowly until the knot was about 3-5 mm long. At this point, the knot was wetted with saliva and the knot was snugged. Great tension was necessary on the tag end to achieve a compact knot with the coils accumulating adjacent to the eye of the hook. Tightening this knot is more difficult and requires forcing the coils down the standing part of the leader near to the eye of the hook. Attempts were made to force the coils down toward the eye of the hook by pushing them with my fingernails.

Based upon these qualitative judgments, I believe that the improved clinch knot is more attractive and is more compact (it does not extend as far out from the eye of the hook).

Quantitative information: Knots tied with 5 turns - In the first test, both knots were constructed with 5 full turns of leader material. One hook had the Trilene knot and the other hook had the improved clinch knot. 20 complete trials were performed. Of the 20 trials, the Trilene knot survived (outlasting the improved clinch knot) in 13 of the 20 trials. In the remaining 7 trials the Improved Clinch knot survived. Therefore, the Trilene knot’s strength was greater in a ratio of 13/7 or nearly two times as often as the Improved Clinch knot. (If neither knot provided an advantage, the ratio should be 10/10.) Therefore, in determining if the Trilene knot provides significantly more strength than the Improved Clinch knot we compare the observed results (13/7) to the results expected if the knots are equal (10/10). A Chi-squared test yields a value of 1.8 with 1 degree of freedom. The probability that this occurred by chance (with the two knots being of equal strength) is 0.1 < p < 0.25. These results indicate that there is no significant difference in knot strength between the two knots using size 6 hooks and 8# Berkley VANISH (1-2 years old).

Knots tied with 4 turns - In a second test, both knots were constructed with 4 full turns of leader material since this seemed most likely to be the optimal number of turns. One hook had the Trilene knot and the other hook had the improved clinch knot. Twenty complete trials were completed. Of the 20 trials, the Trilene knot survived (outlasting the improved clinch knot) in 13 of the 20 trials. In the remaining 7 trials the Improved Clinch knot survived. Therefore, the Trilene knot’s strength was greater in a ratio of 13/7 or nearly two times as often as the Improved Clinch knot. Therefore, in determining if the Trilene knot provides significantly more strength than the Improved Clinch knot, we compare the observed results (13/7) to the results expected if the knots are equal (10/10). A Chi-squared test yields a value of 1.8 with 1 degree of freedom. The probability that this occurred by chance (with the two knots being of equal strength) is 0.1 < p < 0.25. These results indicate that there is no significant difference in knot strength between the two knots using size 6 hooks and 8# Berkley VANISH (1-2 years old).

If both comparisons (knots tied with 5 turns and 4 turns) of the Trilene knot to the Improved Clinch knot are combined, the Trilene knot survived in 26 of the 40 trials; whereas, the Improved Clinch knot survived in 14 of the 40 trials. Therefore, the Trilene knot survived in a ratio of 26/14. Combined statistical analysis either as two groups or considering the original four groups yields the same Chi squared value of 3.6 with 1 degree of freedom (0.05 < p < 0.1). Therefore, without more testing there is no reason to suspect that the Trilene knot is significantly stronger than the Improved Clinch knot when smooth tension is applied.

The break point of all but a few trials was at the knot. On 2-3 occasions the line within ½ inch of the knot broke. I believe that these breaks were in regions of the fluorocarbon that had been stressed during knot tying. Therefore, I attribute the weakness to the nearby knot. Note also that breakage at or very near the knot indicates that the knot was associated with the weakening and that tying the knot had indeed weakened the line. The line always survived better than the knots.

Knots tied with 5 vs. 4 turns - In a third test, knot strength was compared between two Trilene knots, one with 5 full turns of leader material and the other with 4 full turns of leader material. Ten trials were performed, with both knots retied for each trial. The Trilene knot tied with 4 turns of material outlasted the Trilene knot tied with 5 turns of material in 6 of the ten trials. In the remaining 4 trials the Trilene knot with 5 full turns of material survived. Once again, there is no significant difference between these two different knots (Chi squared = 0.4 with 1 degree of freedom, p>0.5).

Knots tied with 4 vs. 3 turns - In a fourth test, knot strength was compared between two Trilene knots, one with 4 full turns of leader material and the other with 3 full turns of leader material. Ten trials were performed, with both knots retied for each trial. The Trilene knot tied with 4 turns of material outlasted the Trilene knot tied with 3 turns of material in 7 of the ten trials. In the remaining 3 trials the Trilene knot with 3 full turns of material survived. Once again, there is no significant difference between these two different knots (Chi squared = 1.6 with 1 degree of freedom, 0.1 < p < 0.25).

Discussion. In head to head comparison of knot strength or durability, the Trilene knot survived in 13 of 20 attempts in two separate tests. However, many times the Trilene knot did not outmatch the improved clinch knot (~1/3 of the time). Therefore, the determination of which knot is better is not so clear. Apparently, the better knot may depend on the specifics of how it is tied with each use. It is unclear to me what differences there were that could have resulted in one knot being stronger than the other. I believe that there is some weakening of the line that occurs during snugging the knot that is not directly dependent upon the knot type. Such weakening during the snugging process may be the culprit that leads to knot weakness. Without more standardization of the snugging process, knot strength may be unpredictable. The use of fingernails to push the coils of the Trilene Knot toward the eye of the hook may have been detrimental to knot strength. However, I was unable to minimize the expanse of the knot without this technique.

I realize that these hooks are slightly larger than flies used routinely to catch fish of the size that I usually catch. In addition, the diameter and strength of the fluorocarbon monofilament material used here was greater than most tippet material that I use for trout. Therefore, some may say that the study does not reflect what we might see with smaller hooks and finer fluorocarbon. However, these materials were chosen for several reasons. They were large enough to facilitate tying many knots using the geriatric eyesight that I presently am relying upon. They require a substantial tension to break the leader and will hopefully be less influenced by small nicks and inadvertent scratches to the leader material.

Conclusion: The durability of the Trilene Knot and the Improved Clinch Knot were statistically indistinguishable Evil or Very Mad

in a direct comparison of which knot survived smoothly applied tension.