UKC

I recently learned about not using a figure 8 to for joining two ropes for an abseil. I understand that a fig8 can infinately roll due to opposing tension on the 2 strands.

​​​​Similarly I take it that using a fig8 to shorten a sling and clipping only the shelf (not the loop) also may cause the knot to roll?

so my question is...

we tie in using a fig 8 (and a stopper, I guess this may be the important part) to the rope and then belay from this rope loop.

Could this cause the knot to roll?

I think the critical difference between a figure-of-8 to join ropes and to tie-in to a harness is the angle between the two strands exiting the knot.

Tie-ing into a harness, the angle might be, say, 45-60 degrees (depending on how snug it is to your harness).

If you use it to join two ropes, the angle will be 180 degrees.  That will create a much larger force which is trying to open up the loops that form the knot.

If you want to join two ropes and load the exit strands in opposition, either use a double-fishermans or an alpine butterfly.

People also use a single overhand knot to join ropes for abseiling (including me).  As long as you leave long tails, it works fine.  I've never seen the knot roll.  I suspect it needs very particular circumstances to occur.

Yeah I know to use an overhand with long tails to join two ropes.

With your response in mind, then using a fig 8 to shorten a sling and clipping above the knot (the shelf) would also be ok? 

Also, if the knot is going to roll, it will roll towards the harness, at which point it will get stuck.

I've never seen anyone with their knot bunched up against their harness from this happening.

Why do you say it would roll towards the harness? Surely it would roll away from the applied opposing force. Towards the stopper?

I don't think the OP's concern is that there might be a problem of cross-loading the knot if you weight the system by falling or hanging on the rope. I think he's concerned about using the rope loop(s) to belay from, rather than using the belay loop.

There's been debate on here before about this. Some instruction manuals recommend belaying off the rope loop. I think that's because it makes it easier to escape the system and/or because it adds a little more dynamic "give" to the belay. Neither of these arguments have convinced me and I put my belay karabiner through both the belay loop and the rope loop. If you use the rope loop alone  - tied with a Fig 8 - then you do create a 180 degree angle at the knot when you hold a fall, which could potentially cause it to unroll. In practice it doesn't. I'm not sure why, but it could be because the stopper knot prevents it happening (as long as you've remembered to tie one of course).

Martin

Yeah this is what I was thinking.

I was always taught my more experienced partners to belay through rope loop (rather than harness loop) for dynamic addition but also (if on a multipitch or belaying from above) that the force transfers to the belay more directly.

But noone ever warned of rolling fig 8.

You always hear it's the strongest knot, bomb proof.

So when I learned of the possibility of rolling on abseil, made me start thinking about rolling on belaying from it.

The stopper knot on a figure of 8 does mitigate the affects of crossloading the tie knot and causing rolling. However the real reason its not a real world issue is that when belaying off the rope loop the sustained load when crossloaded is always pretty low e.g. 2Kn at most. If it gets that high to roll someone is hurting. The highest loads generated are in a high fall factor leader fall but even then there is a brief peak load and not enough to compromise the knot. Rolling fig 8 are more of an issue in rigging where if used where it can be crossloaded and tensioned up e.g. In a rescue haul with high load or a tyrolean/highline the knot can roll. The real advantages of belaying through the rope loop is that it gives the be layer a bit more wiggle/comfort room when loaded if the belay is built properly (the forces bypass the harness. It also reduces wear on your harness though that is rarely a real issue either. 

In a slow pull test the fig8 knot doesn't roll until 6kN. Then it is tight and doesn't roll anymore and the rope snaps at the knot at 14 kN.
youtube.com/watch?v=QGqGlFc3oFs&

The impact force at the belayer will very rarely be higher than 2 kN. Not even for at fall factor 1.
https://www.petzl.com/INT/en/Sport/Forces-at-work-in-a-real-fall?ActivityNa...

A stopper knot is not needed.

A few points.

(1) The purpose of the rope-loop belay is to transfer the load directly to the anchor with no twisting or compression of the harness.  This requires that the belayer has a snug tie-in so that the load transmission is immediate.  There is no significant additional dynamic advantage and belay escape isn't affected one way or the other.  There is no point in using the rope loop rather than the belay loop if the belayer is unanchored, say on the ground for the first pitch.  This is one case when the belay loop should be preferred.

(2)  If a stopper knot is tied snugly behind the figure-eight, the figure-8 can't roll.

(3) All the rolling failures in testing are caused by steady slow pulls over at least a few seconds; we really know nothing about what happens with the load is high but of very short duration.

(4) A ring-loaded figure-8 without a stopper knot withstands at least 1500 lbs in slow-pull testing; see vimeo.com/40767916 . Some of the figure-8's with so-called "Yosemite finishes" do not fare as well...

(5) I think is very difficult with an anchored belayer to get the figure-8 in position to roll, as the load from the tie-in will orient the loop with the belay carabiner opposite the knot.  In this case the two sides of the knot bearing the load are parallel and no rolling is possible.  To get ring loading, you'd need significant load from the harness belay loop and very little from the tie-in, and this shouldn't happen if the tie-in is rigged snugly.  It is conceivable that the ring-loading configuration might occur momentarily when catching a leader fall, before the belayer is jerked up against a high the tie-in. Finally, if ring-loading does occur for some mysterious reason, then because there is a loop, the load to the knot is half the load applied to the belay.

All told, I think there is nothing to worry about.  If you put a backup overhand or barrel knot behind a figure-8, there is less than nothing to worry about.

Thanks for the reply. Interesting to watch the overhand roll completely off!

Personally I think that while theoretically it could happen, the reality is it's a not issue.

I did a video a while ago about why "we" often belay off the rope loop:  youtube.com/watch?v=rpiPEsyKJ08&

I think what you say in your video agrees almost point-for-point with what I said.  And your conclusion about comfort is not to be sniffed at.  I've had a second who fell off an overhang and swung out into space prusik up the rope, which took a while, and I held the belay with no discomfort at all with the rope-loop set-up.  Things would have been far worse with a harness belay loop belay. 

You do make two additional points I'm not so sure about though.

(1) You mention the belay loop as a possible option to protect a dicey anchor.  I think we both agree that if you have an anchor that won't take a second's weight, you're in pretty deep doo-doo.  If you are truly well-braced, ok, but if the belay pulls you over, then you've got your weight and the second's weight on the anchor and you would have been better off with the direct connection. 

(2) We agreed on the belay escape, but I don't agree with the way you've got it rigged.  The problem is that your belay loop is an intrinsic part of the anchor rigging, because all the anchor strands are clipped back to it. If you were to untie the belay loop---as you would ultimately have to do to escape the belay---the whole anchor rigging goes with it.  So to escape the belay with your set-up, you'd basically have to rig a separate anchor with slings and transfer the load to that before you can untie.  This could be quite a problem if the anchor pieces are out of reach.

The solution is to tie a small overhand loop in the rope just beyond the tie-in knot and clip the anchor strands back to that.  This way the anchor rigging retains all of its integrity if you have to untie your connection.  This has another use as well, if you're one of those people who fancy using a plaquette for belaying.  The overhand loop gives you a "power point" to hang the plaquette from, and with this system you can tie the overhand loop exactly where you would want the power point to be, say up at face level.

David Coley mentions this method on his Multipitch Climbing website, where he calls it the DIL loop.

(3) One other thing.  Your system uses four strands for two anchor points; presumably six strands for three anchor points.  You can cut the number of strands used almost in half.  In the two-anchor case, rope goes from belayer to first anchor, clove to first anchor, leave a tad of slack and clove to the second anchor, and then bring the strand back and clove to the belayer.  So instead of four strands used up in the belay you're only using two strands, plus the little bit of slack running from anchor 1 to anchor 2.  If the pitch is long relative to your rope, this might make a difference...

This question - can the fig 8 capsize if you belay from it, and would a stopper help? - comes up quite often. Here's the last time:

https://www.ukclimbing.com/forums/rock_talk/re-tied_figure_8_with_yosemite_...

As rgold and jezb1 have said, the main reason to belay from the rope loop is simply comfort.  There are other minor advantages, such as reducing wear on the harness (I don't care about wear to the end of the rope because it can be trimmed) and making it very slightly easier to escape the system (which you'll probably do once every 20 years or so..!) but mostly it's just more comfortable. 

I'm unaware of any reports of a fig 8 ever having capsized and rolled whilst used to belay.  That rather heroic last video in Stefan's link to the Petzl website shows that, even in a ridiculous FF1 situation, you barely get 2kN at the belay plate.

I'm also unaware of any data that shows that adding the stopper actually makes the fig 8 any stronger or safer. It's frequently stated as gospel, at this point in the discussion, that even if the fig 8 did start to roll (and we know it won't in this situation) then the stopper would help - but no-one seems to have done the testing on this.  One day I will get access to a pull testing system and I will find out for sure!

What the stopper does do is introduce a potentially fatal false loop; there's been more than one death caused by this.

Put overhands in slings or use the right length sling, it's neater.

In theory yes under high load, in reality it works fine.

jk

To address the other point raised about joining two ropes together with figure eight knots:

• A figure eight bend is fine to use to tie two rope of similar diameters together - (https://www.animatedknots.com/figure-8-bend-knot)

• A offset figure eight bend or flat figure eight bend is NOT SAFE and has resulted in several deaths when it has been used to tie two ropes together - (oppelter_Achknoten_Tropfen.jpg" target="_blank" rel="noreferrer noopener nofollow" id="guid-5f29369887807" class="counterlink">https://en.wikipedia.org/wiki/Offset_figure-eight_bend#/media/Fileoppelte...)

Safe, but I wouldn't choose it.  As it's not an offset knot it's a bugger for getting caught on things.  Much better with a double overhand: https://www.needlesports.com/Information/Need-Advice/Abseil-Knots

Interesting. I normally now (ie for at least 30 years) tie in with the rope loop the same size as the belay loop. That's a lot tighter than most people I see. Somewhere I saw it suggested that this is best because it aligns the harness correctly if you are hanging on the rope. Doing this the way I do, I don't think belaying through the rope loop rather than the belay loop (or both) will give me any more wiggle/comfort room. I may try it out with a larger loop though and see if it makes holding a fall more comfortable as you and others have suggested.

Martin

All good points that you raise, the video was belay loop v rope loop, I've done loads of others including the normal ways of creating a direct belay out of a rope system, and other videos with other belay systems etc.

Thanks. I usually would use an overhand, but only really learned of the knots rolling recently, hence me thinking about other places that you may use a fig 8

Thanks everyone. Much appreciated.

I am going right back to the original draft of the question .
The first part of your question is hard to answer because you're not clear on the precise loading profile of what you describe as a "figure 8" (eg loading profile can also include; 'cross-loading' and 'circumferential loading').

The term "Figure 8" can be interpreted in different ways.
For example, there is #1047 Figure 8 (which is a fixed eye knot) that is commonly employed as a climbing harness tie-in knot.
But, there is also an 'Offset' form of this knot.
The term offset refers to the knot core being displaced from the axis of tension.
For example, the poorly named 'EDK' is an example of an offset knot.
An offset Figure 8 bend (ie end-to-end joining knot) can be unstable - and there are certain dressing states that exacerbate this.
It is possible to achieve a dressing state of an offset Figure 8 bend that resists capsizing - but it requires significant practice and deep understanding of knot geometry. Because of this, it is generally not recommended to use an offset Figure 8 bend in any life critical application.
If you want to learn more about the technical details of offset knots, I recommend this link:
Link: http://www.paci.com.au/knots.php (at #5 in the table).

Just a note to anyone reading this. This thread gave me the tools to find more historical threads.   For example this

https://www.ukclimbing.com/forums/rock_talk/ringloading_a_figure_8-511621

Some important/interesting info in this vid...

https://go.skimresources.com/?id=51389X1256615&isjs=1&jv=14.2.0-sta...

And another useful quote from that thread that didn't occur to me until now was this...

"Just to note, the video doesn't show the fig8 being pulled in the directions it would be at a normal belay, as one of the lines would be taking the force. It shows what would happen if you put a cow's tail on the rope loop.

It would be hard to really ringload a knot at the belay." 

This makes more sense to me now

The only time I've ever dropped a belay device was a direct result of belaying through the rope loop (it was winter and for maybe the first time ever I clipped it into the rope as it was easier to see/clip). When I untied to ab off it disappeared into the gloom. Never again!