Stretch Test!

On my “Good Ideas To Do At Some Point In The Future Maybe When Things Quiet Down” (GITDASPITFMWTQ) list there has been an item that’s been hovering near the top of the list (above “Right grate American novel” but below “Close the door you’re letting all the heat out” ) has been a bench test to see how the different materials, treatments and brands of rope we use here at CYR perform in one of the most important metrics we have: stretch.

As cordage evolves, it almost always trends in 2 positive directions: lower stretch and higher strength.  In the last couple years strength has become less relevant as the finished size of the rope becomes the limiting factor.  If we really wanted to, we could have 4mm cores on 35′ racing boats and not be worried about it blowing up.  Stretch has become more important. If you look at the “flavors” that Dyneema generations come in,  the most recent iterations have become more and more stretch focused.

Dyneema SK60 (most similar to Spectra 750) was strong and light, but stretched, crept and wasn’t significantly stronger than the aramid fibers it was compared to. SK75  (Spectra 1000) was lower creep and much stronger. SK78 was lower creep, but not stronger (this is now the “standard” Dyneema from most marques). SK90 was stronger but no better on creep than 75, so has been supplanted by SK99 (better in creep and stronger).  DM20 is… weird and I’ve never used it.

In the past year,  I’ve made halyards with each of the ropes shown below.  All of these are 3/16 or 5mm so comparable in size and would yield a ~5/16″ halyard.  From the top:

MARLOW SK78 MAX The MAX line of Marlows Dyneema cores is their version of heat setting.

Defining characteristics: It’s black!  The black comes off on your hands. Second stiffest rope in test.  Looks cool. The braid angle is exactly average in this sample set.

Sizing:  The 5mm core measured 4.4mm after being loaded.  Round? Not really,  unloaded it measures 5.3 one way and 3.69mm the other

Dumb observation:  If this were a car, it would be an Audi A3.


Defining characteristics: Silver. Not as stiff as HSR or Max, but stiffer than STS78.  Longest braid angle.  Smells nice.

Sizing:  The 5mm measures 4.6mm loaded. Round? Meh.  It measures 4.73mm one way and 4.2 the other.

Dumb observation: If this were a football player, it would be Odell Beckham Jr.  Never heard of XTM either, but it’s good.


Defining Characteristics It’s white (comes in 7 colors too) it’s floppy compared to the Vectran and the Heat Sets. Very shallow braid angle. Gets super stiff once loaded.

Sizing 4.9mm under load Round? Round!

Dumb Oberservation While typing this up, I dropped all the samples on the floor. I could tell the STS78 just by feel. I bet if I were to total up all the STS78 (aka STS75 aka Endura 12) I’ve spliced, it could go to Belmont harbor from the shop and back.  Lets put that on the GITDASPITFMWTQ list!


Defining characteristics Gold! Much more abrasive than any of the Dyneemas.

Sizing 4.2 under load. Round? NO.  2.9mm one way, 4.6 the other.

Dumb Observation I totally forgot I had any Vectran in stock.  It’s gone from about 40% of my high tech cordage to exactly .8% ( I have sold one Vectran rope all year) If this were a car it would be a Pontiac, because in a few years people will be all “Pon-Tee-Yak? What’s that?”


Defining Characteristics It’s stiff! Really stiff. When you put a cover on it properly the crew all hates you for the first race because it only gets stiffer when covered.

Sizing: 5.1mm under load Round WOAH, they make round rope! 5.2mm one way, 5.3 the other. About as round as it gets.

Dumb Observation If this were a car it would be a WRX.  Not exactly pretty, some people don’t like how they feel but does everything really well. And I sometimes cover it with orange cover, and WRX’s are often covered in tacky aftermarket gear as well.


So how did they do?  Well, nothing broke and no one got hurt, which is good.  Before I share any results though, lets talk methodology.  Before we talk methodology, lets make a bunch of disclaimers for how rudimentary my testing methods are. Before we do that,  have you listened to Astronautalis? He’s great, and I think it’s a damn shame that the yanswer to the previous question is always “who?”

METHODOLOGY I spliced all the samples into 10′ lengths. Then loaded everything to 1000lbs with a dwell of 30 seconds.  The samples were length checked, and respliced to 10′.  Why respliced?  The construction stretch on new rope is huge.  Even heat set ropes lose some of their “set” when coiled and handled.  Splicing takes up a long length of rope, but then releases some of that once it’s under tension. Everything was loaded again and checked again.

To test stretch,  I loaded all the samples to 200lbs, did the CYR Load Distribution Procedure (hit rope with mallet, kick hydraulic cylinder) then then loaded everything to 100lbs.  A clamp was made off at the dogbone holding one end of the sample. Then the tension was added slowly until 1000lbs was reached and stable. The distance between the clamp and dogbone was then measured using a digital mic. The samples were then tested in reverse order after being left flat on the bench to minimize bending the rope.  After that test, all the samples were taken to 2000lbs with a dwell of 1 minute, then retensioned to 1000lbs.

ROPE Stretch @1000lb Stretch @ 1000lb 2 Stretch @ 1000lb after 2000lb load
Yale Vectrus 55.25 31.37 26.99
Marlow MAX SK78 21.64 19.52 11.64
Alpha Ropes D Core XTM 78 24.32 17.53 9.78
New England STS78 29.9 22.93 14.33
New England HSR 20.35 14.42 8.84

There is lots of interesting information here.  All the ropes developed less stretch after being loaded, with the benefits increasing with the amount of load applied.  This fits with what conventional but often ignored wisdom, as most riggers recommend preloading your halyards before use.  For mains I like to load the main halyard around the boom at the mainsheet strop, then tension the mainsheet as hard as possible. For genoas, jibs and code sails I use the tack fitting, tension with a winch, then “banjo” the halyard from the foredeck by pulling aft on the line.  For spinnakers I go get coffee and think about the weather.

The vectran stretched more, significantly more, than all the flavors of Dyneema.  This is not a surprise, but keep in mind that Vectran should in theory have less creep than Dyneema SK75 and below over time, although SK78 and Vectran are supposed to be comparable.  I will be modeling this, but since creep takes a long time to develop, I’ll be loaded the samples over a weekend and checking on Mondays.

The non heat set STS78 showed the next most stretch, but comes with a caveat.  When setting the lengths to 10′ on the samples, I took nearly 100mm  of initial set out of the STS78, whereas every other sample was more like 5-60mm after the splices were set.  This means that if using regular Dyneema for length critical applications like strops or pennants, it needs to be set under tension. For best results I usually exceed the working load by 3x and have had good results with regular Dyneema after that.

The Heat Set Dyneemas all performed very well.  The Alpha XTM stretch a lot initially but settled down after hitting our target load, and did particularly well after getting tensioned to 2x the test load. The Marlow MAX started better than the XTM but was overtaken after it was tensioned to 2000lbs.  If you’ve got lots of experience with rope, this is very intuitive after handling the lines; the XTM starts out being much less stiff so it stands to reason that it would stretch more initially. The XTM braid angle however, is much more parallel to the rope, so once all the stretch is removed, it makes sense that it would show less load stretch.

The New England HSR did the best, especially once tensioned.  This again fits with the “feel” one gets from the rope, as the HSR is the stiffest running rope I have ever used.

This was a fun test and will be updated as the creep results come in.  The good news is that all the Dyneema lines performed really well compared to the state-of-the-art-circa-2005 Vectran.  If we accept these numbers, and assume a boat like a T10 has these ropes on the main halyard.  The T10 sailors, being good preppers and not tired from the party, preload their halyards before sailing.  It’s windy out, but the ropes only stretch: Vectrus 130mm,  MAX78 56mm, XTM78 47mm, STS78 69mm, and HSR 42mm.

As always, let me know if you have any rigging related bits for the GITDASPITFMWTQ list.


The overnight time on the bench has been surprisingly tied up, first with a steering cable and now with loops over the weekend, so I’ve decided to do some shorter term testing.  Figuring that on a bad day, a windward leg of a race takes 30 minutes,  I let each sample dwell on the bench for 45 minutes at 1000lbs after a 1500lb set.  As it turns out, not much at all happens in 45 minutes.

Yale Vectrus 0 978
Marlow MAX SK78 0 982
Alpha Ropes D Core XTM 78 0 982
New England STS78 0 980
New England HSR 0 986

The ropes didn’t move. Even a little. Testing was done with a metal square clamped to the bench even with the dogbone, and after the first sample I kind of knew the score, and that this  test wouldn’t be very exciting.  The most interesting thing is that the load numbers at 45minutes were down off of 1000lbs, but I’m willing to chalk this up to the bench itself, as there is always some movement and it’s not particularly exact as a dedicated testing rig would be.  An interesting proof to this is that once the ropes were taken to 1500lbs, and then tension released to below 1000, then brought back up to 1000lbs, the numbers actually went UP before they went down.  Could be the load cell itself, the loops, the winch, the tensioning rope (1/2″ vectran)

Here are some videos of the best and worst stretch from the test.  This was done after the actual testing, so the clamp zero is a bit off, and the camera isn’t in line, but even so you can see the huge difference in stretch

With no change in dimension and negligible change in load, it seems like creep in the context of something like a windward/leeward race is not a significant factor. My understanding on creep is that it takes a combination of load/time and temperature to occur, and that it happens on the order of days and not minutes or hours. If I ever end up going upwind with the same tension on a halyard for days, please send help, as not only MIGHT creep be a factor but that sounds like an absolutely boring race and I don’t want to do it.

The interesting observation from the setup for this test was how much all the ropes initially moved when being spliced and set to 10′ lengths. So, for kicks,  I also took a 6′ piece of STS78 and did the same test. This piece of rope only had a quick 1500 set, then was taken to 1000lbs.  This one did grow, by about 28mm, and dropped tension to 658lbs.  This fits my own experience in that prestretching any halyard on the bench makes it perform better once on the bench.  Heat Set Dyneemas only really needs the ends stretched after splicing, and Vectran doesn’t seem to either (Dr. Bam Miller of Oyster Bay Boat shop thinks this is due to the higher coefficient of friction with vectran, but he’s not a real doctor and I don’t think Bam is even his real name.) Regular Dyneema does grow by quite a lot,  the splices set as well as the rope itself elongating with load. The good news is that this settles down with a decent prestretch or lots of use.

My news for customers is this:  get a Dyneema halyard. If you want more performance get a Heat Set Dyneema halyard.  If you really want to come in and talk about creep for a boat that does buoy racing, please excuse me while I bang my head on the test bench.

Creep Update 1 HSR heat set Dyneema

The HSR sample was left on the bench at 1000lbs from Friday afternoon through Monday morning,  the length was unchanged and the load was at 960lbs.  Over 60hrs with no dimension change is pretty good!  Assuming you were on a J105, and this was your main halyard, and the ENTIRE Mac race was upwind on the same tack in the same amount of breeze, you still don’t have to worry about creep in a halyard.  So far this testing just reinforces my recommendation for heat set Dyneema halyards, aft standing rigging and critical control lines.

Creep Update 2 STS78 Dyneema

On Monday I loaded the STS78 sample to 1000lbs and planned to leave it there through Wednesday. As it turned out,  it was left until this morning, so about 60 hours just like the HSR.  In theory, this would be the most creep prone sample.  Heat setting as a process removes the constructional stretch, but in theory also reduces creep since it accelerates it during production and aligns both the fibers and the molecuslar structure of the rope.  Regular Dyneema should show considerably more constructional stretch, as well as more dynamic strech and definitely more creep.  The constructional stretch on this sample should have been mitigated by all the cycling (multiple times to 1000, 1500lbs, once to 2000lbs) but I was still expecting to see some creep as compared to the Heat Set.

Instead,  the sample is exactly where it was left on Monday, showing no dimension change.  We’re at about 16% of break load.  Although normal working loads for racing/cruising running rigging are about 20%, I feel that 1000lbs in this case is pretty indicative of a typical max-normal halyard load on a boat that would use 5mm Dyneema core.


Over the weekend (OH THE BEARS…) I decided to continue the creep testing by going with a higher-than-recommended working load on the STS78 sample.  So far, the loads  tested for creep have been loads you would likely see on the water with a 5mm piece of rope. No creep has been observed in any of the Dyneema lines, even in STS78 which is non heat set Dyneema and should in theory show creep.  The fact that this piece has been prestretched by being cycled to 1000 (many times with long dwell), 1500 and 2000lbs has made stretch pretty minimal, but I was a bit surprised we couldn’t generate any creep (OR ANY PASS RUSH, BEARS).

To try and generate creep, I loaded the line to 40% of the breaking load (2440lbs). This is well above any SWL for running rigging, and would not likely be encountered on the water. After round 55 hrs the rope had elongated by just under 1mm (UNLIKE THE BEARS SECONDARY WHICH ELONGATED ENOUGH FOR 48 POINTS).

This is still really good!  There was no prestretch beyond getting the rope to tension, so you’re likely seeing construction stretch as much as anything.  Getting <1mm over 3048mm at this % of break load is incredible. You could debate whether we’re seeing construction stretch (probably) or creep (unlikely) but it’s nice to have this data (AND THE INEVITABLE TOP FIVE DRAFT PICK IN 2016).

If you’re curious about the angry parentheses…. 

T10 Backstay Update

The Tartan Ten is a staple of life at CYR. At this moment I have a T10 boom on the bench, and another strapped to the ceiling; there are 2 T10 backstays in boxes on my desk, a forestay just came off the swager, and there are 2 boxes of halyards and sheets on their way out the door UPS. T10s are the alpha and the omega around here! For every cool big boat project, there’s a dozen T10 bits, so I take them pretty seriously even when their owners come in looking for innovative blender solutions, or a shade of rope that matches their favorite beer can (both true)

Anyway, one of the specialty T10 products has been the fiber backstay. Proud to say that if you see a fiber backstay on a Ten it’s probably one of ours. One of the original pre-2010 backstays came in from the cold the other day, so I took the chance to see how it’s held up. This has 5 seasons of use including regattas, beercans and distance races. Pull test results will post shortly, but for now lets look at the condition.

The top eye is covered in chafe sleeve to keep it safe near the masthead crane. You can see this thimble is too large, as the eye has been crushed a bit by the top of the masthead crane. We started using smaller thimbles or no thimbles after 2011. The loads on this are light enough that a smaller thimble won’t distort under load, and even the pin alone would offer plenty of bend radius.

Here you can see what 5 seasons of the backstay flicker ring have done to the chafe sleeve on the backstay: not a lot This is a 3/16″ stock stainless ring, and it hasn’t made a dent in the backstay itself.

Here you can see where the top batten hits the backstay. The very first time we tried Dynex Dux as a backstay, the T10 top batten was fuzzing up the backstay on the first sail, so the chafe sleeve was added. The chafe sleeve is quite slick and tough, and the tight weave and stretching during splicing make it snag free. Very pleased to see how fresh this looks entering year 6.

We’ll break this soon and see how much of the rated 5T strength is left.

We broke this recently, and saw how much of the rated 5T strength was left…

The backstay broke at 5,640lbs, or approximately 53% of original rated strength after 5 years of normal/heavy use.  What’s funny is that that number is still stronger than the 1×19 wire it replaced.

Dyneema durability

Pulled one of our Shields traveler lines out for a quick test to answer a customers question: but how will it be in 4 years?

Our traveler lines actually have 5 seasons and change on them, but figure it would still be interesting to break test one and see how it did.

They are 1/8″ Endura 12, originally rated for 2100lbs break. They see pretty mild loads, but to break it down, if we assume we rarely race over 25kts (last weekend excepted!) and that the mainsheet load in that condition is 1180lbs, and we assume traveler load is 20% of sheet load, we get 236lbs. That seems about right, as we’ve got 4:1 controls and it’s rarely very hard to pull. I’d say most of the time we get more like 70-100lbs on the car. Obviously shock loads from gybes would exceed these numbers, but the line was still pretty safely specced. Condition was pretty good since all the leads were fair and the end terminations were over smooth pins. There was some chafe in the middle from where it would run over the non skid deck, but it was minor.

Rope broke at 1384lbs, so retained 65% of it’s strenght after 5+ years. Pretty good I’d say, as thats about the time I’d recommend replacing it anyway.

Great stuff Dyneema!

1/8″ Dyneema Break Test Final: Bury Splice vs Brummel Splice

For the final it was no surprise to see the bury splice and the brummel splice.  The previous rounds saw knots, specialty splices and things done purposefully wrong, so these should be the top end.

The bury splice, again, is the tapered tail of 72x rope diameter, inserted into the rope to form an eye, and stitched to keep the splice intact under no/low loads. The brummel is the same, but with an interlocking weave instead of the stitch.

Well, as you can see it was the bury splice that walked away the winner, besting the brummel which lost at 2958lbs. This is 134% of rated strength for 1/8″ Endura 12, so pretty happy with the result. If you read the semi finals, you’ll get some explanation of why it’s so far above rated, but it’s nice to know CYR splicing is beating the numbers we use for specifying rope.

Based on the result,  you might ask why we don’t use the bury splice as standard, instead the brummel is the default for sheets and halyards.  The reason is the brummel is faster, and has a locking mechanism which can be verified and can’t possible wear out or be removed.  At the numbers seen in this and other tests, it’s always broken above rated for good quality lines, so I can use rated strength when speccing with no concerns about strength.

For the full series of tests:

1/8″ Dyneema Break Test Bracket: Quarterfinal Rd 1 Skiff Knot vs Bury Splice

1/8″ Dyneema Break Test Bracket: Quarterfinal Rd 2: Loop vs Short Bury Splice

1/8″ Dyneema Break Test Bracket: Quarterfinal Rd 3: Bowline vs Sliding Sling Splice

1/8″ Dyneema Break Test Bracket: Quarterfinal Rd 4 Soft Shackle vs Brummel Splice

1/8″ Dyneema Break Test Bracket: Semi Final 1 Full Bury Splice vs Short Bury

1/8″ Dyneema break test bracket Semi Final 2: Brummel Splice vs Sliding Splice

That concludes the break testing for now, if you have suggestions or requests for similar tests feel free to get in touch with me at

1/8″ Dyneema break test bracket Semi Final 2: Brummel Splice vs Sliding Splice

This was a neat one, on the left we have a sliding splice. The sliding splice (aka whoopie sling, aka adjuster, aka hiking strap splice) is a splice in which the tail is extra long, and left outside the body of the rope.  This makes a handle you can use to change the dimensions of the eye.  Quite handy for adjustable length lines, and for making something adjustable in the field. We use this for things like hiking straps, backstay gross tunes, topping lifts etc.  It’s quite handy, but I really had no idea how strong it would be. Since theres a large disruption in the braid where the tail exits, the rope should experience going from half load to full load all in one area, making it weaker. I’ve always treated this as a 50% loss to be safe.

On the right side is the most sold splice at CYR, the brummel splice with full bury. This is a brummel splice to lock the eye closed at low loads, with a 72 diameter bury.  We assume the brummel weakens the line a bit, but treat this as about 85% of rated strength when speccing line. Since we use a 5:1 safety factor for most things, it’s still a very solid choice.

Neat! The sliding splice turned out to be stronger than expected, still breaking above rated for the line.  The big caveat here is that this was new line, which had been loaded up only once before. I expect the strength loss will accelerate as time goes on.


1/8″ Dyneema Break Test Bracket: Semi Final 1 Full Bury Splice vs Short Bury

A belated resumption of the break test competition (Can’t call it March Mildness anymore!) sees 2 similar splices go head to head.

The most basic splice in 12 strand is when a tail of rope is buried back inside the body of the rope.  In this case, 72 diameters (9″) of 1/8″ Dyneema was tapered and buried with only a simple hidden stitch to keep it together. The hidden stitch only works at really low load, as this splice will hold all the way up to break under just friction. The reason riggers use a lock stitch is to keep the splice secure under low loads, or when the splice gets snagged on something that tries to pull the eye in two directions.  The most likely culprit here is someone playing with the splice!

For the right side of the screen, I did the very same splice but with a crucial difference; the bury is only 5″ long.

We’d expect the short bury to be weaker, since the splice has less space for loads to distribute from the buried tail into the body, but how much weaker will it be?

As you can see, the short bury was weaker, but still broke at a surprisingly high 3066lbs.  For rope that’s rated at 2200lbs, this is quite good.  A few things account for this:

The rated strength is typically an average minimum strength, in which the lowest numbers of a sample are used.  Additionally, New England tends to be quite conservative with their numbers, which are usually high enough (read, better than similar products from competition)

Also Dyneema is a bit odd in that as it’s loaded, it actually gains strength as the fibers align.  It’s the same principle behind heat set lines like Dynex Dux. Over a 4 year working life, the rope typically gains strength over the first year, and then slowly loses it over time. Since this splice had been loaded up in the preceding rounds, it’s actualyl gained strangth! Neat, huh?


1/8″ Dyneema Break Test Bracket: Quarterfinal Rd 4 Soft Shackle vs Brummel Splice

Here’s the last result from the quarter finals.  That’s an 1/8″ soft shackle with diamond knot on the left, and a pair of brummel splices with full buries on the right. They are in basket configuration to even things up with the soft shackle.

3718lbs break, and it was the soft shackle that failed.


1/8″ Dyneema Break Test Bracket: Quarterfinal Rd 2: Loop vs Short Bury Splice

This one was a bit trickier for the oddsmakers in Las Vegas.  We have a plain loop, formed by doing 72 diameter buries together, and the short bury splice.  This splice is pretty common;  it has a brummel to lock the throat, then a short (in this case 5″) bury.  It’s a common fault in 12 strand splicing to make the buries too short, either by not having enough room, or plain laziness.  Theres not much time saved, and I’m not sure that saving 4″ of rope is worth it!

The test has the loop in a single/vertical configuration, and the short splice in basket. This is to try and even the playing field, as the loop has 2 passes of rope and the short splice was obviously 2 single passes.

I was a bit surprised the loop broke first, although the result may have a bit of an * next to it, as the tope broke around the stainless steel quicklink, and not at the join or at the end of the splice.  Looking closer at the quicklink, there was a tiny steel burr there that may have given the rope a rough surface to bear on.  The 1/8″ Endura 12 has a rated strength of 2100lbs, so this isn’t out of line, but I’ve found NER Dyneema usually breaks well over the rated strength.

If the test was for anything other than fun value, I’d probably retest the loop after smoothing out the burr, but since this is just for kicks I let the result stand.  Please don’t tell the International Riggers Casual Break Testing Organization, the last thing I need is an IRCBTO cease and desist letter.


Winner: The Short Bury Splice*


Studio Gang exhibit at the Art Institute of Chicago

In the middle of CYR’s busiest season ever, an opportunity came along that was just too cool to pass up!  

Chicago-based architecture firm Studio Gang was presenting an exhibition of their work at Chicago’s Art Institute, and they came to CYR for a very rigging themed part of the project.  They’d created four “rope rooms” made from steel rings and rope. The designs were taking cues from past designs and buildings, and needed to be secured to the ceiling of the gallery.  Initially they wanted to go with steel cables, as was traditional for this sort of thing.  We had a meeting to sort out costs and dimensions of turnbuckles and cable, and I happened to bring a piece of New England STS-75, just to provide an alternative.  Once they saw the dyneema sling next to the clunky swaged cable (or-gasp-rope clips per the drawing) their eyes lit up!  The dyneema was lighter, easier to install and cheaper. Most importantly the slim white rope would be far lower profile compared to steel cable and turnbuckle, and where it was visible it would blend in with the rope rooms themselves.

Once we’d decided on going with cordage, CYR had to find a way to make the assemblies to a tight length tolerance, ensure no stretch and still provide a way for the architects to level the rooms to the floor once installed.  Enter the lashing thimble!

CYR has been using thimbles for adjusting fixed assemblies and low travel/high load purchases for years.  Initially it was custom machined and anodized rings, and now Antal and Tylaska are making these in volume for great prices.  So long as these are applied properly (please no more spin sheet turning blocks!) they’re fantastic products.  The right length was critical, as well as having zero movement once installed, so they were made and prestretched on CYR’s testing bench.  Each sling was taken to 1500lbs to remove the initial stretch. As part of the package we prooftested one assembly, and provided an install demo and photo instructions.  The installers loved it, and were blown away at how fast and easy it was to work with, getting our parts in took less than a full day!405418_10151187498952502_1706797292_n.jpg (720×960)

After a quick inspection during installation, it took a few weeks before I was able to see the finished product.  It’s a real pleasure to see CYR rigging in such a unique location, and the rooms look great and fit with the rest of the exhibit well.  The architects seemed very interested in the capabilities of Chicago Yacht Rigging, and are going to visit the shop and North Sails loft soon, hopefully we get more exciting projects from outside the the boating world!  Here are a few more images, and be sure to check out the greater exhibit now through February!