Hand-sewing webbing;  strength and number of stitches.

Danger Will Robinson! This site discusses sewing of webbing. Sewing can be very dangerous, if you are unable to handle certain concepts, like multiplication and personal responsibility. DANGER DANGER DANGER!

It is well known that the strength of sewn loop in webbing depends proportionally on the number of stitches and the strength of the thread, and has little to do with the pattern of stitching (e.g., sailrite); that is

Strength of sewn material = (number of stitches)*(tensile strength thread)*(GF=geometric factor) ,

Where no loop can be stronger than the material sewn.

For machine sewing using a lock stitch (thread on top and bottom), and when the material and thread have similar strength per yarn, the GF is ~1.4 to 1.5. Each thread is responsible for half the strength, and the strength of each thread is reduced from 1 by a factor ~ sqrt(2)/2 because the threads dive roughly at a 45 degree angle from the outside of the fabric to where they cross.

I have sewn webbing for break tests by machine, using hundreds of stitches; the strength of a sewn loop is as great as the inherent strength of the webbing, if an adequate number of stitches is used.

However, I often sew soft goods by hand, with very strong UHMWPE  “thread” (HMPE, like Spectra or Dyneema). This "thread" is actually braided fishing line. I do this when the area for stitching is very restricted and not easily accessed by a sewing machine. At issue is the fact that the UHMWPE is much stronger than the polyester or nylon being sewn, and can actually cut through the weaker fibers under stress; UHMWPE is also weakened by tight bends, which invariably form as the stitched webbing pieces start to come apart. Unlike machine stitching, hand-stitching involes just one thread per penetration of both pieces of webbing, so is ~half as strong per "stitch."

For this test I hand-sewed loops on each end of pieces of 5/8” (~16mm) Bluewater  climbspec nylon webbing; with a breaking strength variously given as 2023 (really 9kN) and 2300 lbs. I did two samples each of 5 different stitch numbers: 16, 20, 25, 35, and 50 stitches.  I supplied the pulling force with my Jeep in 4low, and measured the force at 640 Hz with a linescale 3 load cell. This type of pull supplies most of the impulse over less than a second, and is extreme compared to the slow pull one might get on a lab bench.

The results:

 The 50 stitch samples recover about 100% of the webbing strength. I normally would use about 70 stitches for this application. The GF appears to be about 0.43, rather than the 0.7 one would expect for nylon-on-nylon (stitching on webbing).  When I have sewn loops in the ends of webbing with polyester thread, I used at least 540 stitches for each end.

Here is a picture of the broken loop ends of the two 50-stitch samples. The Spectra thread did break, but appears to have ripped apart the nylon webbing in the process.