Friction hitch without access to ends?

[shadowjack]

Consider following scenario: I have rope under tension (1), and want to transfer weight to other rope (2). I need to attach second rope to first, but do not have access to either end of rope 2. How should I proceed? What if I do not have any other gear (Prusik loop, carabiners)?

[xarax]

   If the first / main line is tensioned, you are not supposed to be able to form a 360-degrees nipping loop on it - and if it is tensioned a lot, you are not supposed to be able to even bend it locally, to any degree : so your situation is difficult, and I do not see how it can improve : As there is no geometrical "obstacle" formed on this first / main tensioned line, to firmly attach a second line on it, is already a challenge. Fortunately, there is one way you can achieve this, which is used by sailors to secure the tensioned anchor or mooring line of the ship on board, by attaching a second line on it : the rat-tail stopper (1). All the other multi-wrap gripping hitches which are better known and more used by the knot tyers ( and, especially, by the climbers ) are either inferior, or they require a not-very-tensioned first / main line, which can be forced to curve, locally, to a less or more noticeable degree - that is, a first / main line which can be forced to provide a local "bump", on which we can attach a hitch tied with a second line.
   Now, you even want to be able to do this with a second line of which you do not have access of any end !  You want a midline-to-midline bend, where the first / main line is tensioned...
   I do not believe that there is ANY way you can do this with a practical knot - if you do not wish a ridiculously bulky and ugly tangle tied around the first line, of course  ! It is my favourite example I use to offer, of a situation which proves that knots are wonderful things, but they can not be the universal answer to each and every task involving attachments of tensioned ropes.
   If the second line is of a much smaller size than the first, you can perhaps try to tie a double-line rat-tail stopper ( a first challenge, as this doubled second line, while it forms the two crossed helices of the rat-tail stopper around the first line, should be tucked "over" and "under" itself many times, AND it should be dressed very carefully, so its two lines revolve around the first line, but not around themselves ! ), and then join, somehow, the two-line Standing and Tail ends of this double-line rat-tail stopper ( a second challenge ! ). I would nt suggest such a knotting exercise to anybody ! 

1. http://www.animatedknots.com/rattailstopper/index.php?LogoImage=LogoGrog.j..

[roo]

Consider following scenario: I have rope under tension (1), and want to transfer weight to other rope (2). I need to attach second rope to first, but do not have access to either end of rope 2. How should I proceed? What if I do not have any other gear (Prusik loop, carabiners)?

My initial thought is that you could fold rope 2 to make a pseudo-end from a bight and then proceed to tie a hitch that can take lengthwise pull such as:

http://notableknotindex.webs.com/sailorhitches.html
http://notableknotindex.webs.com/blakeshitch.html

You'd have to experiment with the number of coils, and things would depend heavily on the characteristics of the two ropes, such as size ratios, stiffness, slickness, tension level in rope 1, how many ends of rope 2 see tension, etc.

[xarax]

   The so-called "Sailor s hitch", suggested in the previous post, in not ANY sailor s hitch ( it was never, and it will never be, I am afraid... ) - it is only a mediocre gripping hitch, which, if tied with a doubled line, becomes even more mediocre...
   The Blake s hitch is a very good climbing hitch, indeed - but all climbing gripping hitches are designed so they will be able to "open" and "close" repeatedly, while they are transported along the mail line, so they involve elements which are not required in the general case of a mid-first-line to mid-second-line attachment. In the case of the rat-tail stopper, we have helical segments that work as riding turns on other helical segments, squeezing them on the surface of the main line, and forcing them to "bite" it : this does not happen, or does not happen in such a degree, in the case of climbing gripping hitches.
   Moreover, a double-line climbing gripping hitch is an even bulkier, and an even more ugly, thing, than a double-line rat-tail stopper ! It is difficult to dress all the parallel coils of a double-line climbing hitch properly, so they do not revolve around, and do not cross each other. If they do cross each other, they will lose their continuous contact with the main line, but they will "bite" it at some points, and this may be beneficial to their gripping power, indeed - but then, why not just tie a rat-tail stopper in the first place, where those crossings between the ( crossing helical ) coils are orderly and evenly positioned along the main line, right from the beginning ?
   However, I have to say that I can not be sure that, in the vast KnotLand, there is no double-line gripping hitch that will be able to work in a satisfactory degree, without being too bulky, ugly, and difficult to dress. I am just not aware of any such hitch right now, and, based on my repeatedly failed attempts, I do not believe that such a thing can exist. 
     

[shadowjack]

Thank you for your replies.
I know Blake's Hitch, as I'm recreational climber. In fact, I'm fan of it - my ascending contraption has 5 of them (and one progress capture pulley).
I thought about tying it with bight - but I was sure it wouldn't grip. As it was suggested here, I tied it (see photo). I can't test it on properly tensioned line (any suggestions how?), but while pulling line with one hand (~20 kgf) it grips repeatedly. Rope is stiff 9 mm kermantle, almost new (Tendon 9 mm Speleo).
Thanks about suggestion of rat-tail stopper. Is it specific hitch or any hitch used to fix end of mooring line? If the first, can you point me to image of it?
UPDATE:
Regarding friction hitch not gripping on tensioned rope. As I stated, load will be transferred from first rope to second. So first rope after tying the hitch will be unweighted, and hitch has a chance to grip.

[xarax]

( any suggestions how? )

   To tension a line by hand, the easiest way is to use a knot mechanism utilizing a mechanical advantage, that is, a rope-made block & tackle simple machine : You fasten the one end of the line at the first anchor. Near the other end, you tie a slipped overhand knot, or any other mid-line loop. You pass the other end of the line around the other anchor, then through the eye of the loop, then around the anchor again, and finally through the eye of the loop for the second and last time. By pulling this end, you tension and "lock" the line at the same time. ( This is what I was doing when I was searching for gripping hitches around tensioned lines, and was lead to the rat-tail-stopper, of which I was not aware at the time (1)).( See the attached picture). Another way is to attach the main line to a Spanish windlass rope mechanism tied on another line - which way is also able to deliver a quite strong pull, and it is also able to be "locked" easily.

  Why do knot tyers have not yet learned the rat-tail-stopper, while commercial and navy ship sailors throughout the world use it on daily basis ? I believe that this is, in a large degree, due to an Ashley s sentence, which established another enduring knotting myth... Ashley wrote that :
  "...straps are easier on spars and rigging than corded rope, besides being less liable to slip"( ABoK#1758 )
  and, by this, he misled generations of mid-19th and 20th century knot tyers, who think of "rigging" only as "wire rigging" = steel cables ! ... and many of whom still use it only around poles, where we have so many much better tight hitches ! (2)

1. http://igkt.net/sm/index.php?topic=2849
2. http://igkt.net/sm/index.php?topic=1889.msg13604#msg13604

...load will be transferred from first rope to second. So first rope after tying the hitch will be unweighted, and hitch has a chance to grip.

  The hitch should grip in advance, otherwise the load will not be transferred, it will escape !  I mean, you have to suppose that one of the lines will be carrying the total load at any moment, every instant, continuously ( that is, there will not be any time window during which there will be no load, or the existing load will be supported by other means ). The hitch should be tied, dressed and pre-tensioned while the main line is still under the 100% of the load -otherwise it will start sliding and it will not grip, or it will grip only after some time, during which it will be transported along the main line ( or the main line will slip through it ).
   If you will be able to carry the load by any other means, even for some seconds during the "transfer", then you will also be able to insert a nipping loop, or just bend the main line a little bid - and this will make things MUCH easier ! We have many hitches which are able to withstand even a lengthwise pull when tied around a rope, and which can be firmly attached on it, if it will provide them a "handle" to grip on, however slight that may be - a small straight angle deflexion, or a locally curved segment, for example. 

[shadowjack]

Thanks, I'll try rat-tail stopper.

The hitch should grip in advance, otherwise the load will not be transferred, it will escape !  I mean, you have to suppose that one of the lines will be carrying the total load at any moment, every instant, continuously ( that is, there will not be any time window during which there will be no load, or the existing load will be supported by other means ).

Not totally true - at any moment load should be supported by both ropes in some proportion. For example, first line can support 20% of load, and second 80%.
Suppose load is 100 kg. Hitch slips at 20 kg on tensioned line, but by that time load on first line is reduced to 80 kg, allowing hitch to hold 30 kg, and so on.
I think there is problem like achiles and tortoise paradox.
It might not be the case in particular material/load, but such mechanism does exists.

[xarax]

   The requirement " without access to the ends" may be interpreted as a requirement for just a TIB hitch. However, that is not true : it is a much more strict requirement, which excludes many TIB hitches - and most of the multi-wrap, single line hitches which would had been able to do the job.  : if we can not use any end of the second line, and, also, if we can not release the tension on the main line even for a brief moment ( which means that we can not form a loop into this main line ), we can not tie any multi-wrap single-line /  single "nipping/gripping tube" hitch, even if it is TIB. ( We can tie a twin "nipping/gripping tube" hitch, that is, a hitch where one right-handed and one left-handed coils are placed in line, the one after the other - but this leads to bulky and ugly solutions ).
   See a TIB Pile-hitch-like hitch in the attached picture. It is TIB, but we can not form the single "nipping / gripping tube " around the main line, "without access to the ends" of the second line.
     
   

[xarax]

...at any moment load should be supported by both ropes in some proportion.

....ONLY IFF the hitch tied on the second line has already been "locked" around the main line completely ! If it is not "locked", it will not be able to support any significant portion of the load - and even then, after it will be locked, we will only hope that it will support the load ! Lengthwise pull is a difficult thing to handle !
   A not-locked, just partially tightened hitch, which will be loaded lengthwise, will start sliding along the main line, and, when this will happen, you will need some strong veeery-instant glue to stop it ! 
   The "locking" of the hitch on the surface of the main line is NOT a continuous process, where one hitch tightened by, say, 50% of the load, will be able to apply 50% of its gripping power ! The ropes are deformed locally, and a strong gripping is possible only if the friction between them is enhanced by those deformations of their cylindrical surfaces. When one segment of a rope is squeezed on another, the contact areas become saddle-shaped, and this provides a micro-obstacle, a local micro-bump, which inhibits a "free" slide.
   If the hitch has not been "locked" at a certain area of the main line, and has not placed its "dents" hard and deep into its surface, it can not support a significant portion of the load - and to be "locked", it has to be tensioned almost as much as it can be, otherwise it will start sliding, and it will grip ( if it will ever grip...) only after some time, and some rope length...
   Here we can remember another thing of the mechanism of friction : "Static friction is friction between two or more solid objects that are not moving relative to each other. The coefficient of static friction, typically denoted as μs, is usually higher than the coefficient of kinetic friction. Kinetic (or dynamic) friction occurs when two objects are moving relative to each other and rub together (like a sled on the ground). The coefficient of kinetic friction is typically denoted as μk, and is usually less than the coefficient of static friction for the same materials. "

   Do NOT let your hitch be partially tightened, and wait/hope that it will be able to support the percentage of the total load which corresponds to the percentage of its final tightening ! It will start sliding, and then you will have to be very quick with this glue...

[DerekSmith]

   A not-locked, just partially tightened hitch, which will be loaded lengthwise, will start sliding along the main line, and, when this will happen, you will need some strong veeery-instant glue to stop it ! 
 

I have to disagree with this statement, because in some (most) grip hitches, it is the process of partial tightening which causes grip to start and once started, through a feedback mechanism it continues a positive feedback to further build the grip.

The VT (Valdotain Tresse) hitch is a perfect exmple of this process and you might consider utilising a version of the VT hitch made double line from an inline bight.

[xarax]

in some (most) grip hitches, it is the process of partial tightening which causes grip to start and once started, through a feedback mechanism it continues a positive feedback to further build the grip.

   IMHO, this does not happen in "some", but only in a "few" gripping hitches - and, of course, it does not happen in "most" of them ! 
   In particular, it does happen in the rat-and-tail stopper, indeed, as at the VT hitch - which VT hitch is nothing but a shortened ( minimum length ) rat-tail-stopper, with a certain "closing". ( The two "cross gartered" helices of the rat-tail-stopper can be joined together with many different ways in their "lower" part, so the pull from the one, only, end can be transferred evenly to the other, too : It is paramount that the two crossing helices can work in tandem, and are evenly tensioned, otherwise such a hitch runs the danger to be distorted badly, with detrimental consequences in its gripping power ). However, as far as I know, it does NOT happen in any other of the many climbing hitches I am aware, because they are different ! ( Of course, it does happen also in the very similar hitches around poles shown by Ashley, but let us not mix apples with oranges here : the hitches around soft, compressible materials, as the ropes are, and the hitches around poles or wire rigging ). So, if the "most" of the above sentence is changed to "few", I agree. I, too, had in mind to point out this difference, but as it concerns a few only hitches, I decided not to - but it is good that Derek Smith made this comment, and it offers me the opportunity to explain what I mean.
   Why it does happen ? Because, in those two hitches ( and possibly in a few, only, others, that I do not know...) the wraps are TWO crossed helices, and the angle of the coiling, relatively to the axis of the helix, is large. I do not know the optimum angle ( I would be glad to learn it, because I guess there should be one optimum angle...), but let us say that this angle is / should be almost 45 degrees - while in "most" other gripping hitches, the helical coils are adjacent, so almost parallel to each other ( and the corresponding angle is almost 0 degrees...).
   Why this large angle ? Because this angle makes the particular gripping mechanism of those hitches more effective : the friction between the two helices and the surface of the main line makes those helices to be elongated more, and the added length means a subtracted diameter, and the less diameter of each helical nipping/gripping tube means that the helices are squeezed around the main line and around each other more, offering even more friction, etc. Here the positive feedback mechanism mentioned by shadowjack works, indeed - but only here ! 
   In "most" climbing ripping hitches, the coils are meant to be and to remain parallel and adjacent to each other, and  the "nipping/gripping" tube works as a whole : it is not as long and it is not elongated so much, so its diameter does not shrink so much, so the friction between the coils and the surface of the main line is not becoming greater GRADUALLY, as claimed.
   I have to stress that those observations concern a really tightened main line : a free-hanging main line, like the main lines used in climbing, behaves differently, because it can be bent, locally, by the hitch. A tensioned anchor/mooring line of a ship can not be bent by the rat-tail-stopper, of course !
   
   ( A double-line VT hitch is not less bulky, ugly and difficult to dress than a double-line rat-tail-stopper. I do NOT recommend those knots, they are NOT practical ! )

[Ruby]

tumble hitch?

[xarax]

tumble hitch?

   It is nice you mentioned this hitch, Ruby ! Of course, the "common" tumble hitch can not withstand any serious lengthwise pull - it is a very weak hitch, because its two wraps are too few, and can not be pre-tensioned.
   However, imagine a multi-wrap tumble hitch, where the wraps could somehow be pre-tensioned by the pulling of the "locking" last bight ( the bight which, by its stiffness, acts as a toggle, and prevents the straightening of the Standing Part s first curve ). That would possibly work, indeed - but I, for one, had not been able to devise a tight "Tumble-like" hitch, able to withstand a lengthwise pull, till now... 

[xarax]

   I would like to repeat that the friction hitches around solid objects work differently than the friction hitches around ropes - and that we should not suppose that, if a hitch works on the former, it will also work on the latter. See the attached picture : A simple ( but mediocre, IMHO ) double-line slipped(*) Pile hitch, if endowed with a sufficiently large number of wraps, will be able to withstand any lengthwise pull - but the "same" hitch, around a rope ( the diameter of which can not be so big, relatively to the hitching line ), will not.

(*) "Slipped", in a certain sense : the one of the two strands of the double line is slipped, and works as a toggle / a "key" of the mechanism - the other is not. This is a general method to "lock' double-line knots.

[SS369]

Consider following scenario: I have rope under tension (1), and want to transfer weight to other rope (2). I need to attach second rope to first, but do not have access to either end of rope 2. How should I proceed? What if I do not have any other gear (Prusik loop, carabiners)?

Good day shadowjack.

You posed a very interesting scenario. A few more details are needed for your particular situation. I don't think we should address this just in generalities. But, we probably will....

The weight to be transferred?
The actual ropes to be used, dynamic or static (their construction materials).
Life support activity?
No carabiners, accessory cord or other equipment? Not very good planning! ;-)

The assumed scenario is that you have similarly sized ropes of similar materials and construction. With this info, it should be brought to mind that in a parallel pull scenario very few hitches work with similarly sized ropes. And even less if the main rope is severely tensioned.
If the rope is dynamic, then it will further elongate and reduce in diameter when the added load is applied, so the hitch has to handle this, being able to reduce to below one diameter where it is trying to nip the main line. This is going to be a difficult task as the main rope is more rod-like now, and potentially a slippery pipe. And if it does not deflect some, as xarax has point to, then the gripping structure will have to cover a lot of area. imo

If there is any deflection at all, the bight tied Blake's is a good possibility. Then after the load is exchanged to the second rope, the first rope can be allowed to make a better, more appropriate connection. I am assuming that the first rope will remain anchored (to hold the second line at least?)

Can/will you share the potential scenario where this may have to be employed?

Most good gripping hitches that could cover this application rely on a difference of rope diameters (prusiks, 60% of main line size approx.) and type of material used (soft second rope for hard first rope and vice versa).

The photograph of the bight tied Blake's you've shown is the idea to try, in your "test" scenario, in my opinion. Maybe set up a tensioned rope between two anchor points (trees perhaps) using a come-along or a versatackle set up.
You will have to dress it (the hitch of the moment) so very tightly, play with the number of coils and do some hard pulls and jerking to the second line.

Other hitches to try are, the Taught line (midshipman's, rolling, adjustable grip, etc....),Valdotain Tresse, Icicle, and many others that could be tried "in-the-bight" or slipped (Make sure to secured the slip very well!).

Better to be prepared with additional gear!

SS