Terminology

Doug Turner wrote:

"I have never seen any reference to "unravelling" of unloaded knots
nor any standard way to measure the propensity for this to occur."

"I would be grateful for any advice or pointers on this."

Doug, what follows is just a simple "thought experiment" to try to get
an understanding of the situation you describe.

First, what is holding a knot together? I would suggest that the
answer is friction. To have friction there must be some forces or
loads in the knot and the material must have some coefficient of
friction. (I assume there is no "glue" and the material does not bond
together at a molecular level)

Where can the load, that creates friction, on the knot originate? I
look at the knot used by a surgeon as a binding knot. Further, I see
it as having three parts. The first would be the loop holding or
pinching the tissue together. The second could be the knot itself. The
third is the loose ends. The load must come from one (or more) of
these parts.

First, the "loop". We can look at an old rule of thumb. If you are
binding hard materials together, e.g. steel pipes, use a rope that
stretches. The stretch in the rope creates and maintains a force that
helps create the friction in the knot that holds the knot together.
If you are binding a "soft" material together, e.g. a rug, you can use
a rope with little stretch. The force, to help create friction in the
knot, is created by the rug wanting to expand. I assume with tissue
(soft?), there is an initial force caused by the tissue wanting to
separate. As the tissue heals, I would guess that initial force
decreases, perhaps to zero. I would also guess the "stretchiness" of
the thread might contribute a force. But, depending on the tissue, as
the tissue heals that force might also be lost. I would guess that the
"loop" does not contribute to knot's friction and stability after a
"short" intial period.

Second, the knot itself. The knot is tied, dressed and set. It is the
setting or tightening of the knot that will stretch the "rope" within
the knot. That stretch, within the knot, will create and maintain a
force, that will create and maintain the friction to help hold the
knot together.

Third, the loose ends. No force, no help.

If the above simple model has any validity, I might say the following.
You want a binding knot that can be tightened to pull tissue together.
That "pulling of the tissue together" creates a force that helps stop
the knot from unravelling. However, that "pulling of the tissue
together" will decrease over time to zero. As a result, you also want
a knot that will not unravel when shaken that has loose ends (standing
parts). We might call it a bend. Now I understand some of the multiple
"knot knots" I have seen in pictures of knots used by surgeons!

Elsewhere you mention a "knots-in-a-sponge - warm rinse cycle". I like
it! As a "rough and tumble" first look at the question of unravelling,
I would tie a knot through the sponge and over a rod so that the rod
can be removed after the knot is set. The rod's removal should set the
force on the knot from the loop to zero. I would tie a particular knot
several times for each material. I would also try to tie each knot
with the same tension. I would try to obtain the "stretch" and
coefficent of friction of the materals and see if there is any
correlation between them and the results of the "knot-in-a-sponge -
warm rinse cycle".

In a specific material, it would not surprise me that the resistance
of the knot to unravelling was related to the tension used to set the
knot. However, one would need to measure the tension used. With a
range of tensions with a specific knot and specific material, one
could repeat the "knot-in-a-sponge - warm rinse cycle" and see if
there is a threshold. Similarily, different materials may have a
requirement for a much higher tension when setting a knot to obtain
resistance to unravelling than other materials.

As a "knot tyer", I now have another criteria to evaluate knots - a
"secure and stable" bend that can be tied as a binding knot. I will
leave the "tied in thread using forceps to you! :-)

I hope the above thoughts and the simple model is a help and has some
validity! I am looking forward to your comments.

All the best,
Brian.