Knots in industrial slings

In a blogpost somebody asked:
"Roundsling question: The datasheets specifically forbid knots of any type (i.e. to strop two slings, or shorten a sling), under any circumstances, regardless of the load, don't even de-rate it, just no.... why?"

My answer:

Here is a video I took.

The slings have a WLL of 2 tons. 
 
By the way ... here is a video where I tested industrial slings in a girth hitch. It were 2 destroyed slings from a previous test, each knotted together.
Again ... the knot had no influence.

Rules (according to my tests, not statistically approved):

1. If you can open a knot after a sling being loaded ... relax, it was safe.
2. If the force in the sling is divided in two branches, a knot is no problem, because it has about 50% reduction.

Only if the knot is directly loaded, it reduces the breaking load.

Is Friction in a knot dangerous? No. Its a modern urban legend

Recently there was a discussion on facebook about this knot:

One reason was mentioned why to put a thimble where you could just do this:

"Rope on rope equals friction. That friction generates heat and that can be devastating to the rope integrity."

If you load this knot then the rope may cut/melt through the loop. 

Ok, My approach is: Test it.

I built a Cyclic load  machine:

It looks very complicated but it loads the knot with 1 kN.

After 4.000 times loading I did the breaking load test.

 

 It broke at 20,58 kN.

As a comparison I tested an unused rope. It broke at 22,24 kN:

I decided to do the ultimate "fatigue" test. I cyclic loaded the knot between10 kN and 1 kN.

I checked the rope after 1200 cycles:

They look damaged, but the core is not visible.

At 1500 cycles I stopped. 

Although the core was not visible I think everybody would have discarded the rope.

Then I tested the breaking load:

 21 kN.

So there was no difference between a new rope and this rope.

It is a modern urban legend that the rope will become damaged in this application only by loading it in rope access.
 

Fatigue test with cable clamps axial loaded

In a previous post I tested cable clamps (EN 13411-5-1) in axial load.
Now I wanted to find out what happens when cyclic loaded. 
I choose a load of 15 kN because it is 2,5 times the allowed force that may occur in a safety system of a ropes course when a climber falls. 
In fact, there are 3 kN maximum, so we have a safety factor of 10.

Very roughly speaking:
The fatigue resistance of steel is about 50% of the breaking load, that means, if a clamp breaks at 30 kN it should be resistant to fatige if loaded with 15 kN.

Here is my assembly. I tested 3 clamps. As a side effect I also tested some steel wire terminations and aluminum carabiners.

The two carabiners failed after 8.200 and 8.700 cycles. The MBL is 35 kN. After

 
The cable clamps did not react.

I did the breaking load test:

Breaking  load was
32,84 kN, 37,38 kN, and 44 kN.

This is like uncycled clamps.