Testing of Stitched and Seized Eyes

11-15-2013, rev. 12-31-2013, rev. 8-5-2024

A few weeks ago I posted that stitching and seizing an eye was a viable alternative. I've had stitched eyes outlast the line and no failure, but given the high load on the genoa sheets, I though some back-up calculations were in order. I broke out and old rock climbing gear testing rig and did some breaking.

All testing with used 1/2-inch Sta-Set polyester double braid. All tension figures refer to a sewn or stitched eye. Stitching was hand work, round stitching with sail makers needle and doubled 50-pound waxed whipping twine, about 3/16" apart and about 25% of the distance to the rope centerline, staggered slightly. All seizings were hand tight (about 1-pound, allowing for slippage) and side-by-side, about 40 per inch. Three trials of each fabrication, the minimum of which was used for all calculations. The whipping twine was tested to confirm the 50-pound rating. The angle formed by each eye where it was slipped over a beefy eye was 20-30 degrees; certainly realistic. If for some reason a wider angle is needed, a sturdy throat whipping is in order, with sufficient thread count to keep the throat closed. On 1/2-inch line this should amount to about 25 turns, or a 1/2-inch whipping, just to be safe. However, we did not use a throat whipping during the stitched eye trials.

Rounds Stitching on the Outside Quarter. At 4500 pounds tension a maximum of 9 round stitches on each side (total of 18) of doubled 250 pounds per round stitch (2 passes of doubled twine) . At 2000 pounds 4 stitches on each side (total of 8) also carried 250 pounds per round stitch. This is with the end of the thread unsecured (I cut the last stitch with a razor knife) to simulate wear. Over 8 trials, the variation per stitch was less than 10 pounds. Each time the failure was abrupt, the line only slightly distorted around the stitches, and no line damage resulted even from repeated stitching and testing to failure. Thus, I have no reason to believe that the the eye should be significantly less than 100% line strength; probably 90% like most splices. I will test some smaller line to failure to confirm this.

This works out to 62.5 pounds per strand or about 125% of thread strength. Some of the load is carried by friction around the eye and some by line to line friction, while at the same time the angle of the stitches increases stress. Anyway, it all works out to 62.5 pounds per strand. Thus, 20 round stitches on each side of doubled 50 pound line should give a failure strength = 80 passes *2* 62.5 = 10,000 pounds. A nice safety factor. Hard to believe on a gut level, but watching those few stitches holding 4500 pounds set the mind at ease and proves the engineering.

Stitching Through the Core. This time I simply ran  a plain stitch through the core about every 1/4-inch. Over the course of 3 trials up to 2500 pounds, the result was the same; 50-60 pounds per strand. Advantages or disadvantages? On smaller line, 5/16-inch and smaller, this is easier and perhaps less likely to damage the line. However, the problem is the distance needed for the required number of passes. Assuming we sew two lines, rotating the rope 90 degrees, the maximum stitch count is 16 per inch of line or 800 pounds of holding per inch. The second pass can be difficult (the line is compressed) and could be damaging (the fibers are more likely to be cut by the needle, as they cannot move out of the way). Neglecting that factor and ignoring the last 1/2-inch or line....

• 1/4-inch, 2000 pounds:    3 inches
• 5/16-inch, 3000 pounds:  4.25 inches
• 3/8-inch, 4000 pounds:    5.5 inches
• 1/2-inch 8000 pounds:   11.1 inches

So for larger line this gets a bit impractical; my Warp Speed sheets I would need 28 inches of stitching! Of course, they can be combined. Sew the center line first, as the line will be compressed if the edges are sewn first and the needle very difficult to pass through.  A positive is that the stitches pull down into the weave and are thus partially protected from abrasion. If the tail is to be tapered to keep it smooth, a few round stitches will be need to hold the cover down.

Myths:

• The failure of one stitch due to wear can cause zippering. Not true. In fact, in some cases, aft the eye would not break at the full capacity of the test rig (5ooo pounds) I would cut stitiches with a razor until it did fail.While this reduce the average load per stitch to some extent (20-30%), they were still immensely strong. the first example (18 stitches at 4500 pounds) was obtained after cutting excess stitches away under load!
• Pre-compression with clamps or basting down the center helps. Nope, not what we found. As stitching progresses the rope gets quite compressed (round stitches do that). Pre-compression makes it very difficult to get the needle through and results in more fiber damage.
• Even stitch tension is vital. Nope, I'm not a machine and I don't do particularly neat work. While we didn't go around leaving slack in the stitches,  an obvious corollary of the stitch cutting experiment is that moderate tension variation doesn't matter. The waxed twine merely slides a little, like a lashing, and the stress is still carried evenly.The zig-zag shape of a round stitch probably promotes this, one of the reasons a zig-zag is always used in sails, even back in the hand stitching era.

Seizing.
Plain Seizing. Surprisingly, seizing was very hard to quantify. It seems to depend very strongly on the surface of the rope and it seems clear to me that it made more sense on the 3-strand rope of old, where the strands could lock together.

Since the seizing were drawn up about 1 pound with 40 passes per inch, about 80 pounds per inch of

Pre-load. On the other hand, the line shrinks under tension, so there is probably no actual preload.

Instead of counting strands I measured the length of the line-touching-line seizing. The failure was not The force to start sliding varied from 800 pounds per inch of seizing, so a 2 inch seizing should hold about 1600 pounds as a loop. Apparently not much additional clamping force is generated by the strain as the load comes on. Not much value here, compared to stitching. Perhaps the value of seizing was different back in the day of laid natural fiber ropes, but now it is more a matter of protecting the stitching and perhaps reducing fatigue; it would keep stress off the stitching at stress below the point of sliding.
by breaking but by sliding, so there is no point in using heavier thread for strength (UV and abrasion are another matter).

Would a second layer of seizing help? Mostly a second lay prevents abrasion and UV damage. That is the way the old-time sailors saw it.

What about shrinkage with wet dry cycles? That is the greater concern, as a loose seizing is undependable.

And yet, I have used seized-only eyes a few places for many years. Why no failures? I seriously doubt they ever saw stresses above the sliding point. In fact, one of the lines I tested with was an old seized bridle line and it held without apparent strain to 4500 pounds. There were 3 inches of seizing,which brought the capacity to 2400 pounds, plus there some stitching under the seizing, used to hold things tight while working. The 2 acting together shared the load, most likely never passing 50% capacity, and the construction probably matched the breaking strength of the line. Knowing what I do now, I would have added more stitches!

An historical note: 1/2-inch hemp rope had a breaking strength of 2350 pounds. 3-4 seizing totaling 3 inches would have constituted a 90% efficient termination and could have been used make exactly shaped eyes, and tarred and made with linen thread, would have lasted as long as the line. They knew what they were doing. It just does not apply to double braid.

Wracked Seizing. 
I went back a few weeks later and tried a wracked seizing. In this case, instead of simply wrapping the line , the wraps for a figure-8 around the twin strands. I tried #4 whipping twine, #8 whipping twine, and mason's twine (equivalent to #16 unwaxed whipping twine; 1/12 of line diameter is generally recommended (1/2-inch Staset), in this case, #8 whipping twine. No change, it slid at 800 pounds/inch. Perhaps wracking turns made a difference with laid rope, they don't with double braid.

Note on photos, to right: The line broke right at it's rated strength, not at the knot (modified fugure-8), not at the splice, but in between. Both the splice and knot are very near line strength. Just 10 stitches and a small throat whipping held 2100 pounds. The knot is a modified figure-8.

 

Nylon rope is even more difficult to seize than polyester. The problem is that nylon shrinks when it stretches and the seizing comes loose. In fact, even seizing a nylon rope o protect stitching is difficult in high-load applications, because it comes loose when the rope is highly loaded. You could seize when the rope is under high tension, but it's easier to cover it with webbing  or even heavy duty heat shrink.

---------

The moral of the story? Stitched eyes can be quite strong and reliable. They are even quite forgiving of poor technique. They are far more common in industry than spliced eyes. However, they do not have the same abrasion and UV resistance and should be protected with chafe gear. Seizings look strong but are best thought of as protection for the stitching underneath.

Because of variability in stitching method and rope, and the greater exposure of stitched eyes to UV and wear, it is certainly good practice to apply at least 3x the calculated number of stitches. It's easy to and gives a margin for wear.

My stitched and whipped eye (100-pound Kevlar thread in this case) should have a strength near that of 1/2-inch Warp Speed (stitching and whipping = 25*4*(100/50)*62.5 + 3,000 = 15,500 pounds), although the winch would fail, clew pull out and forestay collapse long before that!