Prohaska Hitch Loop (approaching) 100%???

[James Petersen]

Here is a simple friction hitch I have been using for a belt for the past few years. I don't know if it has been described/claimed before, but I find that it is very simple to tie/remember and in my (limited) experience seems every bit as reliable as a sailor or icicle hitch when used on lines/cylinders of the same diameter or larger. For obvious reasons, I simply call it the belt hitch, and will be using that name in this thread unless someone corrects me. See the photos in this post.

My reason for presenting this hitch is twofold, the first being to introduce what I think is an excellent friction hitch that is easily remembered, effective, and easy to tie and untie. My second reason relates to my ongoing tests of the Lazy Dog and other bends, loops, and interlocked loop bends. That thread can be found at http://igkt.net/sm/index.php?topic=4150.0 .

That is a rather long thread and in which I went through (re?)discovering what I call the lazy dog family of knots which are simply turns around turns. At the urging of X1, I began testing the loop and bend forms of these knots against some other "golden standard" knots for purposes of comparison against what might be termed more or less "known" quantities. In these tests, I noticed that in knots that have both a bend and loop forms, the loop forms tend to be stronger. From here, I posited that a stronger form of bend should be that of two interlocked loops, after which I started testing some examples of loop-to-loop bends. I had planned to move to larger line after I had finished with the admittedly small line that I was using.

When I was nearing the last set of tests on "size b" Chinese knotting string, it occured to me that since many pepole feel that gradual turns/corners going into a bend/loop tend to make it stronger, perhaps a loop or bend made from a friction hitch(es) would be strong also. So I did some tests using my belt hitch to make loops. The knot was tied by making 7 full turns (2520 degrees) rather than the three shown in the photos. This is when I entered the twilight zone...

I am posting the raw test results, since I don't quite know how to analyze them. (Again, the string used was the same (from the same roll) as that used in the Lazy Dog tests mentioned earlier.) The results are as follows:

Line	Belt Hitch Loop	Distance From Knot
22.1 kg	21.7 kg	cm.
21.9 kg	21.5 kg	cm.
22.1 kg	23.1 kg	4.5 cm.
23.1 kg	21.3 kg	2.7 cm.
21.5 kg	21.0 kg	cm.
22.0 kg	23.2 kg	10.0 cm.
23.2 kg	23.2 kg	5.6 cm.
21.1 kg	21.5 kg	1.0 cm.
21.4 kg
22.8 kg	23.2 kg	6.0 cm.
23.3 kg	* 19.8 kg	cm.
	* 21.9 kg	cm.
	21.5 kg	0.5 cm.
22.0 kg	22.4 kg	0.7 cm.
23.7 kg	20.0 kg	cm.
24.6 kg	21.8 kg	cm.
24.7 kg	**    23.4 kg	cm.
21.5 kg	22.0 kg	4.2 cm.
23.7 kg	21.8 kg	cm.
22.4 kg	21.3 kg	1.8 cm.
22.6 kg	21.3 kg	cm.
22.0 kg	21.0 kg	0.8 cm.
23.0 kg	22.0 kg	cm.

Cells with empty cm. values indicate a value of 0.
Cells left blank indicate no testing was done.
* -- The knot slipped until tight against the anchor, then failed.
** -- Knot tied somewhat loosely -- slipped to anchor, then failed.

I have since tested the Belt Hitch loop several times on several different kinds of cord, including a larger Chinese knotting string and two types of "poor man's paracord, all with strengths between 70 and 100 kg. The incidence of "not-adjacent-to-knot" failures was present in all kinds of cord I have yet experimented with, with varying degrees of frequency. I made videos of some of these tests and posted them at : http://archive.org/details/Belt_loop_hitch_loop_test .

[SS369]

Hi James and thanks for the great work!!!

The hitch does have a name if I may. It is called a Blake's hitch or Prohaska hitch. It is usually tied with the addition of a stopper knot on the tail because the tail can be consumed when using it as ascender slide and grip knot.

The videos and .gifs are terrific, thank you.

SS

[James Petersen]

Hi James and thanks for the great work!!!

The hitch does have a name if I may. It is called a Blake's hitch or Prohaska hitch.

OK. I'll try to change the name of the thread. Thanks for the info.

The videos and .gifs are terrific, thank you.

SS

I am having some issues with viewing the videos using the flash player in my browser, but everything seems to be fine when I download the files directly (the links are below the flash video player) and play them with a media player. Sorry for that. This is my first time uploading videos anywhere other than facebook.

[Dan_Lehman]

The hitch does have a name if I may. It is called a Blake's hitch or Prohaska hitch.

Hmmm, by me it's often called the "Prohgrip", as a way
to both recognize Heinz's discovery and give a good-sounding
English name (which will be heard as "pro"(fessional))!
Heinz, himself, once named it something like "Gesteker Wickliknoten"
--German (Austrian)--, which is less handy to my ears.   

It is just one of many friction / gripping hitches devised by Heinz,
btw.  And Franz Bachmann is another fellow who was keen about
such knots.

It is usually tied with the addition of a stopper knot on the tail
because the tail can be consumed when using it as ascender slide and grip knot.

That's a safety check that is in some dispute,
where though the one-time popular arborist rolling hitch
needed it, but the Prohgrip arguably does not.

Note that the Progrip in standard form has the tail
tucked under two wraps rather than one.  Heinz remarks
that stiff rope benefits from this wrap-enlargement and so
more such tucks might be needed for stiff ropes; in slippery
non-stiff ropes, more wraps beyond the tucks might help.

It wasn't clear from your descriptive name "belt loop" how
you were using this knot : it seemed by that that you were
joining ends of a belt!?  But then you seem to be testing it
as an eye knot, or are you?!

It would be nice to see a heavily loaded but unbroken specimen,
and esp. in light of

The knot was tied by making 7 full turns (2520 degrees) rather than
the three shown in the photos. This is when I entered the twilight zone...

which gets a "wow!" from me --and wondering about tucks
(just one, with the other 6 wraps around a single diameter?).

Thanks,
--dl*
====

[James Petersen]

The hitch does have a name if I may. It is called a Blake's hitch or Prohaska hitch.

Hmmm, by me it's often called the "Prohgrip", as a way
to both recognize Heinz's discovery and give a good-sounding
English name (which will be heard as "pro"(fessional))!
Heinz, himself, once named it something like "Gesteker Wickliknoten"
--German (Austrian)--, which is less handy to my ears.   
.....
It wasn't clear from your descriptive name "belt loop" how
you were using this knot : it seemed by that that you were
joining ends of a belt!?  But then you seem to be testing it
as an eye knot, or are you?!

I really wasn't clear about the "belt" part. I have been using the knot (slipped)  in a short piece of rope tied around my coat (at the waist) to keep cold air out during cold weather. It could be used just as effectively to hold your pants up. I have, indeed, been testing it as a loop. I wanted to get an idea of its strength when used as a loop knot and tested to failure.

It would be nice to see a heavily loaded but unbroken specimen,
and esp. in light of

The knot was tied by making 7 full turns (2520 degrees) rather than
the three shown in the photos. This is when I entered the twilight zone...

which gets a "wow!" from me --and wondering about tucks
(just one, with the other 6 wraps around a single diameter?).

I could do that, since on many occasions, it doesn't break at the knot and I still have all the knots. The line from the tests, again is small and fluorescent orange, so taking pictures can be a challenge.  Pictures while loaded are another matter, since both my hands and my eyes are busy (and the test device has no ratchet mechanism.) I will attach the photo of the 17th test of the loop (broke 4.2 cm from the knot.) You can also view the videos at http://archive.org/details/Belt_loop_hitch_loop_test . (If you have issues viewing the files from within your browser, you can download them and view them on a media player program -- simply pause the video where the knots are heavily loaded.)

And yes, it was wit six wraps around a single diameter -- I actually tied it by counting the number of wraps after after the one that would take the (single) tuck.

[James Petersen]

Am I the only one that is perturbed/excited/perplexed/surprised by the results of using this hitch as a loop? In my previous tests using this same line(http://igkt.net/sm/index.php?topic=4150.msg26174#msg26174), the line strength averaged around 22.6 kg., and the strongest knots had average strengths of 20-21 kg. and never did the line fail anywhere but adjacent to the knot. In this group of tests using the same line, with the same average strength (22.6 kg.), the line failed somewhere other than adjacent to the knot in eleven out of the twenty-two tests. And in seven of those cases, the distance from the knot exceeded 1cm. Does that count as one hundred percent knot strength in those cases? Should they be considered part of the line group? both?  Can I calculate an average load at failure if the line didn't fail at the knot? How? Where are you X1? Another poor, lonely, tortured soul seeks help.

Since this  is a friction hitch being used to form a loop, there was always some slippage. I wonder if taking measures to stop or limit that slippage (an overhand knot in the standing leg of the loop perhaps?) would be effective and to what extent. That will have to be part of some more experimenting.

Actually I am perplexed with the brown "poor man's paracord" shown in some of the videos. In my excitement, I tested the prohaska loop with it a number of times before I thought to test the line without the knot. In all of these  tests (of the loop)  save one, the line failed near the windlass, at approximately 100 kg. Afterwards, when testing the line alone, with the end wrapped six times around the carabiner, it failed next to the carabiner each time with a load never exceeding 85 kg! I guess if I want to use that line for more formal testing, I will have to change to a different anchor or method of anchoring that can at least approach 100 kg.

BTW, I believe I have ironed out the issues I was having with the videos that I uploaded. Sorry if anyone was unable to view them previously.

[X1]

   Traveller, there are no paths. Paths are made by walking.

there was always some slippage. I wonder if taking measures to stop or limit that slippage (an overhand knot in the standing leg of the loop perhaps ?) would be effective and to what extent.

  My advice is that, even if something, as a whole, will not work, there would be some parts of it that will. So, do not add more complex things on top of what you already have - simplify the initial thing. I believe you should stick to the no-slippage-at-all condition, either on bends or on loops. One way is to add more and more round turns on your friction hitch used as an end-to-end knot, until you reach a point where, under a load that will be certain fraction of the maximum load this main line can bear, there will be no slippage at all. The number of those turns will be an objectively defined measure of the efficiency of the particular friction hitch, tied on the particular material. You can start loading your friction hitch by, say, 50% of the ultimum strength of the main line, then 55%, 60%, and so on. By finding the minimum number of turns this friction hitch would need in order to avoid any slippage under each of those loadings, I believe that, at the end, you will have a pretty clear picture of the efficiency of this knot as an end-to-end knot - tied on this particular material, of course.
  So, first try to get a clear, complete picture of the friction hitch used as a end-to-end knot (bend), and only afterwards proceed to test it as an eye knot (loop). When you will have some experience with the loadings which puts the friction hitch under the maximum of loading before it slips, you will handle the eye knots where the loading would be only 50% much easier.
  I had tried to measure the relative efficiency of two friction gripping hitches tied around the same pole - the 6-wrap TackleClamp hitch, and the 6-wrap Double Cow hitch. I had thought that the tug-of-war kind of measurements, where one measures only relative values, would be very easy, in relation to the more difficult measurements of absolute values. Needless to say, I have failed miserably, and I was not able to get any decent results ! So, you can be sure that you are not the only one that can not get the clear answers he wishes at the first go... Rome was not built in one day ! 

[Dan_Lehman]

Am I the only one that is perturbed/excited/perplexed/surprised by the results of using this hitch as a loop?

Sorry to be a slow reader of the details,
but glad for your attention, and I expect that I'll join
you in this puzzlement when I can attend more closely
to it!

Actually I am perplexed with the brown "poor man's paracord" shown in some of the videos. In my excitement, I tested the prohaska loop with it a number of times before I thought to test the line without the knot.  In all of these  tests (of the loop)  save one, the line failed near the windlass, at approximately 100 kg.  Afterwards, when testing the line alone, with the end wrapped six times around the carabiner, it failed next to the carabiner each time with a load never exceeding 85 kg! I guess if I want to use that line for more formal testing, I will have to change to a different anchor or method of anchoring that can at least approach 100 kg.

I'll need reminding of the nature of the "windlass",
et cetera --the both ends of the test specimen.
But can you examine a just-heavily-loaded-not-broken
state of the cord at the 'biner?  I'm thinking 'D'-shaped
'biner, and then how the SIX wraps must pile up on
the axis corner of it?  And I'm wondering if you might
find there some torsion evident in the cord, there?!
(E.g., I've often noticed in braided rope tied in a fig.8
knot that some of the fine strands going Z-direction, say,
are arc'd upwards obviously devoid of tension, while the
opposite-direction strands are pulled tight --and think
"gee, there's half the material idle at this point! (at
least on the surface)" !?

Otherwise, I'd think that a roughly 10mm diameter
smoothly round (? maybe not round?) cross section
metal 'biner would be amply kind for something so
small AND compressible as paracord.  E.g., IIRC even
8mm kernmantle tied in a strangle noose around such
a 'biner broke not there but at the noose-SPart where
it entered the strangle knot --in one fellow's report,
anyway.  (Could it be an issue of so little frictional
resistance that great force flows deep into the wraps
and tickles some ugliness of form?  --but you have the
break near the entry point and not deep within the
wraps (evident I suppose, possibly, by retained coiling)?! )


--dl*
====

[Dan_Lehman]

Okay, I got to look at some of the videos --quite interesting!

One can wonder what harmful effects the slippage
of the friction hitch has on the hitched-to line,
such that that might explain the ruptures a little
outside of the knot; it won't explain rupture down
at the broad wooden winch dowel, though!

--dl*
====

[James Petersen]

I'll need reminding of the nature of the "windlass",
et cetera --the both ends of the test specimen.

The description of my testing equipment is at: http://igkt.net/sm/index.php?topic=4150.msg25952#msg25952.

... I'm thinking 'D'-shaped
'biner, and then how the SIX wraps must pile up on
the axis corner of it?  And I'm wondering if you might
find there some torsion evident in the cord, there?!
(E.g., I've often noticed in braided rope tied in a fig.8
knot that some of the fine strands going Z-direction, say,
are arc'd upwards obviously devoid of tension, while the
opposite-direction strands are pulled tight --and think
"gee, there's half the material idle at this point! (at
least on the surface)" !?

Otherwise, I'd think that a roughly 10mm diameter
smoothly round (? maybe not round?) cross section
metal 'biner would be amply kind for something so
small AND compressible as paracord.  E.g., IIRC even
8mm kernmantle tied in a strangle noose around such
a 'biner broke not there but at the noose-SPart where
it entered the strangle knot --in one fellow's report,
anyway.  (Could it be an issue of so little frictional
resistance that great force flows deep into the wraps
and tickles some ugliness of form?  --but you have the
break near the entry point and not deep within the
wraps (evident I suppose, possibly, by retained coiling)?! )

--dl*
====

There is a picture of the carabiner in the description linked to above. Or here's a direct link to the picture -- http://igkt.net/sm/index.php?action=dlattach;topic=4150.0;attach=9205 -- an "0" shaped device with a round cross section. At one point I became concerned that the turns all squeezing into the belly of the carabiner was affecting things, so I ran the turns up the back of the carabiner and secured the end opposite the gate -- making sure that the loaded end ran straight down from the center of the belly -- with no change in results. Go and figure.

[jimmyh]

Am I the only one that is perturbed/excited/perplexed/surprised by the results of using this hitch as a loop?

Count me in on the excitement!

I suspected this hitch would make a strong loop, and have been meaning to test it.

I also played around a bit with variations to spread the friction over more distance.

This one is tied with 8 turns and the tail fed under 5 of them. The interesting difference is that it is left "loose" so that when loaded, it twists around itself like shown in the picture.

I'll get back with test results one of these days

[X1]

have been meaning to test it.

 Good ! However, this cord I see in the picture ( probably a hollow braid one ) had been flattened so much it almost became a webbing ! Can you use a more "round" rope, so we have a more representative test of this hitch ?

[Dan_Lehman]

Am I the only one that is perturbed/excited/perplexed/surprised by the results of using this hitch as a loop? In my previous tests using this same line(http://igkt.net/sm/index.php?topic=4150.msg26174#msg26174), the line strength averaged around 22.6 kg., and the strongest knots had average strengths of 20-21 kg. and never did the line fail anywhere but adjacent to the knot.

Is this absolutely correct : i.p., that the line never failed
within the knot (which is some cases will leave
one with no-knot, sometimes bits of it) ?!  It's certainly
an easy-ish judgement to make, in that the knot remains,
intact.  And this assertion is made for the prior knots,
not (merely) the Prohgrip eye noose, yes?!

I have seen some cases where 3-strand rope broke and
might have (mis)led one to the conclusion that the break
came outside of the knot but where scrutiny would suggest
that maybe the initial rupture occurred within --i.e., that
one strand showed breakage within.  Except by conjecturing
that damage done to the line when it is being pulled from
within the knot will ultimately see it break (if even when
it now lies outside), I find it hard to explain how material
clearly removed from the knot will nevertheless be where
the weakness lies (but for some material defect).  !?

In this group of tests using the same line, with the same average strength (22.6 kg.), the line failed somewhere other than adjacent to the knot in eleven out of the twenty-two tests. And in seven of those cases, the distance from the knot exceeded 1cm. Does that count as one hundred percent knot strength in those cases? Should they be considered part of the line group? both?  Can I calculate an average load at failure if the line didn't fail at the knot? How? Where are you X1? Another poor, lonely, tortured soul seeks help.

If the only aspect is the line breaking away from the knot
--and esp. if the break force lies below tensile strength--,
no, that doesn't ipso facto count as 100% (as is obvious
when going to figure that, if one's confident in line strength).
But you have four results >23kg  which is one of the higher
material strengths and for which the break was remote;
there are a couple others of remote breakage also with
considerable strength though <23kg..  In these cases,
one should regard the knot as 100%, IMO.  --closEnuff !!
(If not quite, so what : what finicky bit of % is left to claim?)
((One might say : theory isn't convinced, but also that
our material isn't theoretic but actual!))

Since this  is a friction hitch being used to form a loop, there was always some slippage.

I will admittedly optimistically call this a mild non sequitur
--i.e., I'll hope that in some cases we might not have such
(or much) slippage.  (There will be some in adjustment, surely,
but maybe in some practical cases slippage won't figure much.)
Yes, per the video'd testing, there was slippage.  (One can
wonder about more dynamic ("shock") loading and this.)

I wonder if taking measures to stop or limit that slippage
(an overhand knot in the standing leg of the loop perhaps?)
would be effective and to what extent.

Perhaps, but one would want to beware the effect of
frustrating/impeding increased frictional gripping by
having in place some such slippage-prevention structure.
(In some of the recently posted bowlines there is some
hope of some degree of frictional effect in the wrapping
around the SPart, which flows into the central nipping loop.)

Actually I am perplexed with the brown "poor man's paracord" shown in some of the videos. In my excitement, I tested the prohaska loop with it a number of times before I thought to test the line without the knot. In all of these  tests (of the loop)  save one, the line failed near the windlass, at approximately 100 kg. Afterwards, when testing the line alone, with the end wrapped six times around the carabiner, it failed next to the carabiner each time with a load never exceeding 85 kg!

Wow, that looks like a 120%+ knot !!      
(Of the quartet of videos I saw, I don't think there was this?)


--dl*
====

ps:  Ditto ("+1") to X1's observation about that blue cord
used by JimmyH (and to some degree, paracord as well)
--i.e., that it's more like tubular webbing than rope, so
behavior will likely differ (esp., perhaps, in the ability of it
to be gripped by something wrapping it).

[jimmyh]

The blue cord in the picture is paracord (from these guys http://www.supplycaptain.com) and is fairly round until squeezed on. No quantitative testing yet, but a 4/2 prohgrip will hold more than my weight without slipping.

I can try to find some rounder solid stuff to test on, but I'm limited to 1000lb

[X1]

In these cases, one should regard the knot as 100%, IMO.  --closEnuff !!
(If not quite, so what : what finicky bit of % is left to claim?)

  "This" knot might be 100%, but with this particular number of wraps, when tied on this particular material. How one can be sure that, given this particular material, this is the optimum number, indeed, if he has not tested it with one less wrap ? My point, at the previous post, was that we will acquire a much more useful, more complete picture of the knot ( the knot in general, not "this" knot tied with a particular number of wraps, on a particular material ) only if we know the minimum number of wraps where there would be no slippage, at various percentages of the maximum loading which the main line can bear. I believe that the Blake or any other friction hitch with n wraps, is a different knot from the "similar"/of the same name friction hitch with n+1 or n-1 wraps ! The Blake hitch with 8 wraps may be able to withstand a 80% of the maximum loading, and another hitch, with 4 wraps ( 50% less material), a 60% of the same loading ( 25% less load ). Which one is more efficient ? Which one should we chose ?