We have been using the same symmetrical splice technique on our 2L bottles for a couple of years now, but I have never quite been happy with it. We have had quite a few failures along the way, and the no-leak yield hasn't been as high as we would like. This meant a lot of wasted bottles and wasted time. The other issue has been the relatively low pressures the spliced pairs could hold. 130psi has been the limit we could practically push them to.
We used to use 2.25L bottles as the sleeve since they fit well over the 2L bottles and no shrinking was required. But recently we have found that some of the 2.25L bottles don't fit as well as they used to so it has been difficult to use them as well. To try to stop the leaks we have been using a different glue to do the sealing in the joint and PL to hold the splice together but this made the process more complex and we still ended up with leaks although less often.
So we are currently trialing a different technique that will hopefully yield better results. The technique is not new but a combination of several different techniques used together. We are using just PL glue for this splice and only the one type of bottle. I'll post the full technique on the main site, but here is a quick run down:
We first heat shrink about 2-3 cm of one of the bottles using hot water so that it fits inside another bottle of the same type. Then curl the edge on a pan. This has the effect of giving a tight seal where the heat shrink section ends and meets the other bottle since they are the same diameter at this point. We sand and glue these sections together. We then make up a sleeve about 4 cm wide made out of the same type of bottle. Because the diameter is the same we simply cut the sleeve to turn it into a strip. We then use another section of sleeve about 4 x 4cm and glue half of it to the sleeve strip to cover the gap. We let the glue dry for a couple of days. Then we glue the sleeve strip over the splice to provide further strength to the splice.
The drawback is that inserting one bottle into the other means that the volume is smaller, but not by too much.
We have had to previously reinforce the necks of the bottles with rings made from other bottles, but it only strengthened the necks not the rest of the bottle. This time we are going to use Richard Wayman's bottle-on-bottle technique to not only strengthen the neck but a significant portion of the rest of the bottle.
Hopefully this technique will yield better sealing results and will be able to take higher pressures.
The glue is curing at the moment so tests will be carried out at the next opportunity.