Load Testing Results

[estarzinger]

Originally Posted by nigel1

Was the Palomar knot easy enough to remove after the load test, or after being loaded up.
How would a halyard hitch compare, I know they are a bugger to remove after being tensioned.

I had to cut it off. It was pretty much welded together.

Just a quick test, 5 pulls each, in 80lb test dyneema fishing line, figure 9 loop was 30% stronger (statistically significant) than figure 8. May be the right solution palomar can't be used (because you can bring a loop over the attachment point). I will have to do some further tests in the amsteel.

[estarzinger]

Figure 9 in amsteel . . . 1184lbs average vs 1144lbs for figure 8 . . . but the variability was twice as much 79lbs stdev vs 30lbs . . . .so i believe if you dressed the figure 9 very carefully you could get 1240lbs (the average of the three highest pulls) which is statistically higher.

[JPA Cate]

Evans

I'm glad you're having fun with this and posting the results.

Ann

[B-Baysailer]

Thanks for trying the palomar knot . ya you cannot untie it after you load it up .but its a knot you can tie to a shackle pretty fast if needed

[Cotemar]

Originally Posted by Ann T. Cate

Evans

I'm glad you're having fun with this and posting the results.

Ann

He should charge admission for all the entertainment and data he is providing.

[StuM]

> Here is a summary chart of the knots. I could not get any of the bends to hold/ not slip on the dyneema single braid

How about testing the zeppelin bend?

[estarzinger]

Yes, I will do zeppelin and some other bends.

But I am just starting to think about bend radius.

I am a bit surprised the below broke at the figure 8 rather than at the 1:1 bend.

Before - two loops (figure 8 and figure 9), joined just simple loop to loop with 1:1 bend radius at join


After - broke at figure 8, at 44% of line strength, which is consistent with what I found in the figure 8 test series


And I just did one with a figure 9 loop and a spliced loop . . .broke at the figure 9 . . .so seems pretty much any knot is weaker than a 1:1 bend radius.

[estarzinger]

Splice to splice test

Before: two pieces, each with two splices, joined loop to loop, with 1:1 bend radius



After: broke at the end of the taper of one of the splices and not the 1:1 radius. Broke at 98% of rated strength (rated average strength with two splices). Big bang



Surprising . . .tentative conclusion . . .1:1 bend radius is just fine and anything bigger is gravy.

[wsmurdoch]

Does the rate of extension change the results? I remember shopkeepers in my youth breaking cotton string by moving fast (with a jerk) when they were tying up packages.

[estarzinger]

Interesting thought.

Faster would create a higher (peak shock) load.

I am not sure what other dynamics faster could cause.

Later, I am going to do some faster/shock loading tests and I came come back to bend radius then and see.

But I am guessing, just from watching the piece explode, that these sort of double spliced test pieces will still break at the weakest point of the splice rather than at the 1:1 bend radius.

[thinwater]

Originally Posted by estarzinger

Interesting thought.

Faster would create a higher (peak shock) load.

I am not sure what other dynamics faster could cause.

Later, I am going to do some faster/shock loading tests and I came come back to bend radius then and see.

But I am guessing, just from watching the piece explode, that these sort of double spliced test pieces will still break at the weakest point of the splice rather than at the 1:1 bend radius.

So, is the corollary that large thimbles are probably not needed on Amsteel lifelines?

I have seen testing that suggested that any thimble that causes an excessive throat angle (sailmakers thimbles) can actually cause weakening of the splice, though I recall the loss was very small.

The 1:1 thing is not too surprising, since rock climbers never rip a carabiner through the loop, not even in drop testing. The loop is carrying only 50% of the line stress. A line running around a block is completely different; the line is carrying 100% of the load then. Similarly, pure loops (climber's slings) generally break at the stitching, but sometimes at the biner.

But I'm loving your thoroughness! I honestly didn't think there was this much to learn pulling knots apart, but by testing odd cases, new information is revealed.

[estarzinger]

Done some research and traded notes with Allen, and have the answer to the puzzle.

First here is a graph from Hampidjan (makers of Dux):



It says for a static bend, a 1:1 bend radius will reduce the line strength by 50%. BUT in a splice the line is 'doubled', so the line inside the splice is only carrying half the load. That means a splice with a 1:1 bend inside is at 100% - the same as the line pulling on the splice. The whole system is at the same strength with a 1:1 bend inside the splice. Which means in a slightly sloppy splice the line will break at the end of the taper (as in my test), or with a perfect splice it might break anywhere. That is all completely consistent with my test results.

The 3 and 4:1 'minimum bend radius' conventional wisdom comes not from splices loops but from applications like bollards where the line around the bend is 'one ended' and carrying 100% of the load (and like the 'package string' mentioned in a post above). In that case, looking at this graph, you see that a 4:1 bend will reduce the strength of the system by 25% (and 1:1 by 50%).

But it seems that some in the yachting business (I will not name names) have mis-interpreted this to apply to 'fixed loops' (splices or knotted). Which it does not, as those give you 'full strength' at 1:1.

[estarzinger]

Regarding the soft shackles, I was thinking about equalizing tension between the two strands and I went back and looked and all my samples first broke on the 'buried' strand rather than the cover strand. That makes some sense to me that the buried strand will be "straighter" and take higher tension than the cover. That would seem to suggest to me that the soft shackles constructions without any bury at all (Kohlhoff style?) might be potentially stronger, if the two strands are tied evenly into the diamond knot.

[thinwater]

Other knot styles? On stropes it is common to use a double fisherman's, for example. I know that some soft shackles users have used other knots, with apparent success. On a flogging jib, perhaps some security is lost, but on and anchor rode I doubt that is so. It is much easier to tie these other knots equalized.

[estarzinger]

I am not going to have time to break much today . . .but a quick one . . .

I was seeing whether I could get the double fishermen bend to stop slipping in the bare dyneema by adding a little resistance/friction with some (relatively light) sewing down the tails.



As you can see in the pic, the sewing just ripped out, and it did it very easily, so obviously near full load was being transmitted to the slipping tails.

That led me to do some quick stitch strength tests, which are interesting . . .

10 stitches of doubled V69 thread has a tensile of 212 lbes (10 x 2 x 10.6lbs), while loops stitched with 10 doubled V69 breaks (all the stitching breaks) at 440lbs = 2x tensile

15 doubled stitches have a tensile of 318, and the loop breaks at 620lbs = 2x tensile

5 stitches of single 80lbs test fishing line has a tensile of 400lbs, and the loop breaks at 600lbs = 1.5x tensile

So, there is a significant "clamping effect" increasing the stitching strength above the pure tensile, and using the tensile to determine the number of stitches would be 'safe/conservative'.

Note: ThinWater also has some excellent sewing/stitching results on his blog