Rigging expert has more on rope-to-chain splice
I enjoyed the recent article on rope-to-chain splicing ("Splicing rope to chain," Issue No. 93, Sept./Oct. 1998). It's one of my favorite subjects, and it's good to see it covered so well.
Working with the very kind folks at New England Rope, I've done some testing on these splices, too. First, the four-strand splice, though simple, might be relatively weak. I only had five broken, but the average efficiency (the percentage of the parent rope's breaking strength) was only 75.6%. The rope was 5/8-inch MegaBraid, spliced through the last 13 links of the chain and seized with constrictors. It seems likely that the rope moves through the links under extremes of load causing distortions, and thus weakness.
You didn't have the crown splice tested for your article, but we got 87.3% average efficiency for the untapered version, and 80.9% for the tapered version. I was surprised to see that the taper resulted in a weaker splice, but this has been borne out, at least in three-strand nylon, in other tests. The crown splice numbers are a little higher than the ones you got for the three-strand direct eye splice, and my experience in the field is that the latter is much more likely to chafe than the former.
As for the two-strand direct splice, we only broke two untapered ones, but both gave way at close to 100% of the rope's strength, a figure far higher than the 87% that you obtained. Even our tapered samples averaged 91% efficiency. The reason for the disparity is, I believe, that the two-strand splice to chain is simply very difficult to do well, especially in nylon, which is poor at holding its lay. It takes a lot of practice to get it right, so it's important either to practice accordingly or to hire someone who has. Either way it's worth it, both for strength and smoothest transition to chain.
As for the thimbled eye splice, not only is it no stronger than a proper two-strand splice, it also has a couple of significant drawbacks: the shackle and thimble are a major headache at rollers and gypsies, and the thimble can get sideways in the eye under heavy loads and chew its way through the rope it was meant to protect. The only sure way to prevent this is to use a bronze "keeper" thimble, with straps that prevent rotation. But it is an expensive, even clunkier alternative. Nylon keepers are too easy to break.
I'll be doing a great deal more testing on a wide variety of knots in the near future, including more chain splices, and would be happy to pass along the results. The tests are for my next book, Working Knots, which, among other things, will address odd splices in nylon, Dacron, and exotic materials.