Documenting my hand-sewn eye process

"Brocky: Interesting concept with the hitch cord. Your modded pulley isn’t a HH but a RW being used in the Australian mode/ configuration. Michoacán works nice, as the legs don’t touch the rope, and lets it rotate fully."
Question I have about the reversed RW setup shown on the right - when tending then, could the rope have a tendency to rotate the slick pin in the RW a lot and could the bird then possibly cause excessive wear on the RW, where it wouldn't normally? Slick idea though. Is this setup just because they're in the southern hemisphere?
 

Brocky

Well-Known Member
Location
Michigan
The Slick Pin and shoulder bolt on the pulley don’t seem to rotate when loaded while descending, only when tending the hitch upward.

Correct, possibly reinforced by recently hearing the Men At Work song, the name is due to the devices being down under, your Reversed RW sounds better!
 
The Slick Pin and shoulder bolt on the pulley don’t seem to rotate when loaded while descending, only when tending the hitch upward.
That's kinda what I meant - the use of the slick pin for tending means it's now acting "like a pulley", whereas in the normal intended design use as Kevin foresaw it, the RW was in the upper position above the prussik and hence didn't really rotate the pin. Using equipment differently than the original design intent can sometimes have unforeseen consequences. This rotation could potentially cause wear from the rotation of the pin's spring lock (sorry I don't know the name of this thing) against the body of the RW? Which maybe also gives me pause about a question someone asked a few years ago now about their gear experiencing corrosion - they worked near a salt water environment. Once the annodized coating is worn away, the metal might be more susceptible to this also (but I'm fussin here now).
 

ductaper

Member
Break Testing the Hand-Sewn Eyes (long post warning):

Back story: I bought a big box of mostly arborist rope from Atwood Rope Co (which I was very pleased with, BTW; their static 11mm climbing line is absolutely perfect for SRS) as I wanted some shorts for projects. Some of the rope I got was non-spliceable 12 strand with a parallel core and some was kernmantle static. I didn't really want to learn to sew eyes so I tried to find a local shop to stitch them for me. I tried marinas, canvas repair shops, shoe repair guys, etc. No one had a machine with enough lift to sew the rope and I understand the big tree supply outfits don't do it anymore. So, I found this thread (pun) and ordered a speedy stitcher.

Luckily, in the box from Atwood was a bunch of their micro and nano cord which is great thread and is already the right material, i.e. it matches the material the rope is made of. A quick call to a beekeeper netted me some beeswax and I was off. I met @moss on a rec climb and came away thinking that he is total guru in this sport of ours so I wanted to follow his directions exactly.

I really love climbing trees but it's not my profession. I'm a mechanical engineer by education so I couldn't help but make a science project out of this. I also wanted to help the readers of this thread to know the limits of what we are doing and try to increase the information available to us going forward. I don't think fear is nearly as good of a decision metric as data is. With that said, please let this post inform you but continue to make your own life-safety decisions.

First I broke the thread I was using so I'd have accurate numbers for the calculations:

IMG-7369x.jpg

I got numbers for each type of thread. The thin nano cord (1/32" / .75mm) broke between 22 and 38 lbs depending on color. The bigger micro cord (3/64" / 1.18mm) was in the 70-78 lb range, again depending on color (more likely the tightness of the braid). Then I stitched up a prototype of what I'll make as a lanyard adjuster in the future:

IMG-7370.jpg

IMG-7369.jpg

No pre-compression on this one. This had a theoretical stitch strength of 3212/14kN lbs per eye (although the thread isn't actually in line with the load there is a component of friction too. This cordage happens to have a very slippery cover so I expect it to break below the aggregate strength. We can discuss this in another post. Others have found break strengths in the neighborhood of aggregate thread strength) Note: This wasn't designed as a life-support component, just a test prototype. I believe it was mentioned in another post that the standard is 3360lb/15kN per eye minimum on hitch cord when both eyes are used (can anyone confirm this?) and we all know the Z133 standard says 5400lbs/24kN for all components in the climbing system. Time to test. First, a basket pull as it would be used:

IMG-7372.jpg

I have access to overhead cranes and forklifts at work but our hanging scale only goes to 2000lbs. The hoist I was using is a 4000 lb unit. The basket pull to that weight showed no deformation in the stitching. Ok, time to double the stress:

IMG-7377.jpg

No problem going to 2000lbs in a straight pull. You can see the thread which started perpendicular to the cordage is starting to look diagonal. This seems to be a good indicator of heavy loading, in my testing, and I'll use it during gear inspection going forward.

I took the scale off so as not to damage it and continued pulling. The piece I was lifting was a little over 5000 lbs of steel and I picked up the half I was connected to before the end (and the pin in the shackle) broke. That made quite a racket and brought some employees out to see even though it was break time. Oops. So, since I wasn't picking it up from the outside edge a (very) conservative estimate of the break strength would be 2500lbs or 78% of theoretical. See my note at the end of the last post about better instrumentation. Here's the carnage:

IMG-7378.jpg

So, for my actual life-support version of this I'll use a stitch count that equals at least 130% of the desired strength because of the slippery cover. Likely more because I don't have the exact number it broke at. It was very encouraging, though, to see the fruit of my labor holding that much weight off the ground.

Now time for some bigger rope. This is nice, supple 12-strand from Atwood which holds a lovely knot but isn't spliceable (but I bet stronger) because of the parallel core. This is their 1/2" product but measures closer to 5/8". It's listed at 6000 break strength but their other products I've broken have broken above rated strength.

This time I used much thicker thread (3/64" / 1.18mm) as I wanted to test fewer, stronger strands. It was a challenge with a 190R-25 size needle but I got it done, this time with a little pre-compression:

IMG-7386.jpg
Starting second pass:
IMG-7387.jpg
Third pass complete:
IMG-7390.jpg

I had a thought when this was done: Why are we stitching eyes? To create an eye in the rope. What are our other options to create an eye? Splicing and knots. If you can't splice the rope you're left with a knot. So, on non-spliceable rope my hypothesis is that our stitched eyes don't necessarily need to exceed the break strength of the rope but should exceed the break strength of a knot (typically 75% of the rope). This could make it a better option than a knot, as long as the stitching stays protected. With that in mind I tested my next sample to see if I could exceed the strength of a knot. I tied a bowline with a Yosemite finish in the other end of this test sample:

IMG-7399.jpg

I'm at my picture limit so the next post will feature the test data. Stay tuned.
 

ductaper

Member
Testing the larger thread/fewer stitches: This eye had 17 stitches/pass/side for a total of 102 stitches with 70lb actual strength thread. That's a theoretical strength of 7140 lbs and the jacket of this rope has less edit: more friction than the last test piece. I only have cranes at work up to 4000 lbs so I had to get creative. I used a forklift to pick up the back end (counterweight end) of a 10,000 lb rated forklift (8990 lbs on the rear axle per manufacturer data sheet) Let's see how we do:

IMG-7400.jpg

I had to stop because I heard the fibers in the rope breaking at the knot and didn't want to add another satellite to space. I wish I knew the actual pull amount but we can get in the ballpark. The truck doing the lifting is rated to 5000 lbs but they can pick up more if the load center is closer to the front tires like it is above. The rear tires on the test dummy were able to be moved by hand so they were significantly unweighted. Let's call this a (very) conservative 6000 lbs. Here's another pic of it under load:

IMG-7404.jpg

And here's what the test piece looked like afterwards:

IMG-7406.jpg

Note the fibers at the knot beginning to break and the stitching that has gone a little diagonal. Again, seems to be a good indicator of overloading. I'd say we achieved our goal of having the stitching be stronger than the knot. Also, side note, look how the knot rolled under load!

I'd rather have hard data vs. the close estimates I have now. I'm in the process of welding up a break test stand and will report back when it's done. So far I have the I-beam, cylinder and power unit but still need a load cell and a readout and they're pricey. I'll let you guys know when it's done and then do some more break testing.

Takeaway: @moss 's technique is one I'll continue using for myself. Nice job, Andrew!

I'm curious to hear what you guys think so bring it on.
 
Last edited:

SomethingWitty

Arkansawyer
Location
LR
Testing the larger thread/fewer stitches: This eye had 17 stitches/pass/side for a total of 102 stitches with 70lb actual strength thread. That's a theoretical strength of 7140 lbs and the jacket of this rope has less friction than the last test piece. I only have cranes at work up to 4000 lbs so I had to get creative. I used a forklift to pick up the back end (counterweight end) of a 10,000 lb rated forklift (8990 lbs on the rear axle per manufacturer data sheet) Let's see how we do:

View attachment 67531

I had to stop because I heard the fibers in the rope breaking at the knot and didn't want to add another satellite to space. I wish I knew the actual pull amount but we can get in the ballpark. The truck doing the lifting is rated to 5000 lbs but they can pick up more if the load center is closer to the front tires like it is above. The rear tires on the test dummy were able to be moved by hand so they were significantly unweighted. Let's call this a (very) conservative 6000 lbs. Here's another pic of it under load:

View attachment 67532

And here's what the test piece looked like afterwards:

View attachment 67533

Note the fibers at the knot beginning to break and the stitching that has gone a little diagonal. Again, seems to be a good indicator of overloading. I'd say we achieved our goal of having the stitching be stronger than the knot. Also, side note, look how the knot rolled under load!

I'd rather have hard data vs. the close estimates I have now. I'm in the process of welding up a break test stand and will report back when it's done. So far I have the I-beam, cylinder and power unit but still need a load cell and a readout and they're pricey. I'll let you guys know when it's done and then do some more break testing.

Takeaway: @moss 's technique is one I'll continue using for myself. Nice job, Andrew!

I'm curious to hear what you guys think so bring it on.
The only thing that I have to say is that engineers are the only guys as cool as tree climbers.
 

flushcut

Well-Known Member
Location
Delavan, WI
Only if they are not the ones designing modern trucks or machinery.


I'd like to see actual break test data. But good test nonetheless.
 

Winchman

Well-Known Member
I'm curious. How exactly are you sewing?

Traditional sewing uses two threads. One is carried through the material by the needle, and a loop is formed on the far side. A second thread from a bobbin is passed through the loop, and it's pulled into the material as the needle goes pack up. The interlocked loops will hold the stitching in place if the thread breaks. A short back stitch will lock the treads in place at the end so a knot isn't needed.

You can also use the needle to push the thread through the material, then grab the thread and pull all of it through. That process is then used from the other side, and repeated over and over until all the thread is used up. The thread needs to be secured with a knot at the end. This method results in more wear on the thread from being pulled through the material over and over.

I expect the special sewing machine used to make professional eye splices is using the traditional method with two threads.

From the picture of the failure of the sewn eye in the black rope in post 746, it looks like friction is doing a pretty good job of holding the threads in place.

It's good to see there's another engineer on the forum, and one who is also a welder/fabricator. This is one of my more challenging projects after retiring from engineering to take up a hobby of welding, fabricating, and machining. It's a saw table for the local steel supply company, complete with rollers on the bed and a roll-out support extension. I also built the table behind the welder in the background.
2020-04-23_cutting_table.png
 

SomethingWitty

Arkansawyer
Location
LR
I'm curious. How exactly are you sewing?

Traditional sewing uses two threads. One is carried through the material by the needle, and a loop is formed on the far side. A second thread from a bobbin is passed through the loop, and it's pulled into the material as the needle goes pack up. The interlocked loops will hold the stitching in place if the thread breaks. A short back stitch will lock the treads in place at the end so a knot isn't needed.

You can also use the needle to push the thread through the material, then grab the thread and pull all of it through. That process is then used from the other side, and repeated over and over until all the thread is used up. The thread needs to be secured with a knot at the end. This method results in more wear on the thread from being pulled through the material over and over.

I expect the special sewing machine used to make professional eye splices is using the traditional method with two threads.

From the picture of the failure of the sewn eye in the black rope in post 746, it looks like friction is doing a pretty good job of holding the threads in place.

It's good to see there's another engineer on the forum, and one who is also a welder/fabricator. This is one of my more challenging projects after retiring from engineering to take up a hobby of welding, fabricating, and machining. It's a saw table for the local steel supply company, complete with rollers on the bed and a roll-out support extension. I also built the table behind the welder in the background.
View attachment 67554
Everything about that picture is beautiful.
 

moss

Well-Known Member
I'm curious. How exactly are you sewing?
Ductaper is using the technique I've detailed in the beginning of this thread and have also updated in the thread as I learned more and refined my technique. It's a long thread ;-)

Interesting to see the two legs of the sewn eye shift during the break test. Typically when my sewn eyes have been pull tested there is no movement in the sewn area or of the two legs of the eye. Since I moved to what I call my "second generation" technique (thorough pre-compression of the two legs before sewing) no one has been able to break any stitching or cause it to shift. The rope always fails first and away from the stitched area.

Here's my most current technique, always use ball-tipped industrial needles intended for textiles, do not use leather cutting chisel needles that come with many sewing awls:
https://flickr.com/photos/naturejournal/albums/72157712585698993

I use a lock-stitching technique, you can see it in any "Speedy Stitcher' sewing awl tutorial video. It is effectively like a manual sewing machine stitch. Instead of a bobbin with a second thread spool on the bottom side, to start the sewer takes a single length of thread and pulls it through the cordage join to the halfway point of the thread length. The length of the thread on the "bottom" side functions essentially the same as a separate bobbin thread. Others have gotten good break results using a "baseball stitch", basically a continuous wrapping stitch through the cordage with no locks. I like lock stitching because it greatly reduces or eliminates a zipper failure effect on extreme loading or say when a few stitches are abraded or broken from contact with tree parts or hand saw. With the lock stitches any thread failures are isolated. Even so on inspection time to retire "the eye" but no one will get dropped out of a tree from a few stitches being broken. This is the same as machine eye stitching. My 3-pass stitching pattern is derived from machine stitching which I studied by dissecting them and by looking at videos of cordage sewing machines operating.
-AJ
 

moss

Well-Known Member
Oh, yeah, standard disclaimer, it takes a long time, is nowhere near as fast as traditional splicing. I got into hand-stitching before custom machine stitching was widely available. Nowadays you can have your cordage stitched to your specifications for a very reasonable cost. I use hand stitching primarily for prototyping these days and special cases like for example sewing a rope bridge to my harness anchors.

Like so:
49265400858_ccac330ef3_b.jpg


-AJ
 

ductaper

Member
Interesting to see the two legs of the sewn eye shift during the break test. Typically when my sewn eyes have been pull tested there is no movement in the sewn area or of the two legs of the eye. Since I moved to what I call my "second generation" technique (thorough pre-compression of the two legs before sewing) no one has been able to break any stitching or cause it to shift. The rope always fails first and away from the stitched area.
Totally agree. I was doing this test to find what would be better than a knot vs. trying to make the joint as strong as possible. Your stitch density seems higher than what I used in the red rope. I was coming at it from a different angle: if all I can use in this rope is a knot, what would be better than it?
 

moss

Well-Known Member
Your stitch density seems higher than what I used in the red rope.

You probably noticed that as you stitch the stitches tend to "roll" towards the center of the join as the cordage compresses for each pass I move the needle entry point further away from the center, this keeps the stitching engaged with more of the cordage fibers and allows me to fit the stitcches more densely, especially th e last pass which has the most stitch density.
-AJ
 

Jonny

Well-Known Member
Location
Buffalo
@moss or anyone that knows. Got a link to a source for some ball point needles compatible with a Speedy Stitcher, that can accommodate fat thread? I think the thread was called Tec400 or something like that. Also think it said .9mm but that sounds massive, Idk.
I lost mine... I’ll check the haystack but I’m not hopeful.
I had found a link to get them before on this thread but can’t find it now.
 

ductaper

Member
@moss or anyone that knows. Got a link to a source for some ball point needles compatible with a Speedy Stitcher, that can accommodate fat thread? I think the thread was called Tec400 or something like that. Also think it said .9mm but that sounds massive, Idk.
I lost mine... I’ll check the haystack but I’m not hopeful.
I had found a link to get them before on this thread but can’t find it now.
Here’s the one I used for my big thread, size 25/200: https://www.ebay.com/itm/261513034459
It fit the SS with no modification.
 

Brocky

Well-Known Member
Location
Michigan
72142A97-9F19-4F70-BD43-854FCB102F0E.png
This is the next type of stitching I’m going to try next, it just takes a lot more thread and cord, and time!
There are a few different methods that I know of, the above type, the multiply passes rock hard kind, the single pass flexible slides some with use known as the Petzl splice mostly, and the Round Perimeter stitching that moss described above.
 

Birdyman88

Well-Known Member
Location
Arlington
Somebody recommended these ball point 135x17 needles in one of the previous posts. They work well with no mods on SpeedyStitcher. I've done 8mm cord using .61mm thread. I also used them with .71mm thread repairing some throwline cubes. I don't think the eye with accommodate any larger thread without some patience.
 

New threads New posts

Kask Stihl NORTHEASTERN Arborists Wesspur TreeStuff.com Kask Teufelberger Westminster X-Rigging Teufelberger Tracked Lifts Arbor Expo
Top Bottom