Quote:
Originally Posted by Alan A
I can speculate with the best of them…
It failed because it wasn’t overbuilt sufficiently and with enough redundancy to withstand whatever happened to it between the design studio and the ocean floor.
Engineers tend to be cautious. Salesmen tend to look for quick, imperfect results.
|
Alan, when I'm designing race car parts my approach is to try and work out all the way it can 'f*@k up and find a design/manufacturing spec. that avoids this
From the brief video I can see quite a few ways it could go all wrong
Titanium end ring bond surfaces don't appear to have been either shot blasted or have an etching primer, the metallic side of a metallic/carbon bond joint usually fails 1st and more often that not due to poor bond preparation
Titanium end ring side bond surfaces don't appear to have any draft, this would minimise/avoid smearing/wiping off the glue when fitting into position the titanium part onto the carbon tube
Titanium end ring doesn't appear to have any design features to maintain a consistent glue gap, a bond joint requires a specific bond joint thickness for maximum strength
Titanium end ring is being wiped down with an old rag and not clinically clean lint free clothe, cleanliness is everything to avoid bond joint surface contamination
Titanium end ring is being lowered using a garage chain hoist, a concentrically accurate sliding bong jig assembly should have been used to minimise smearing/wiping off glue
Stiffness transition from titanium only thickness to carbon only thickness appears very abrupt, a gentle stiffness transition would minimise local stress hot spots that could start a bond joint failure propagating
Titanium ring doesn't appear to have any provision of air holes, this prevents prevent aircpockets hydraulicing the glue out ,
Titanium ring doesn't appear to have any glue injection holes, these feature makes sure the whole bond gap has been filled in all areas by injecting extra glue until it can bee seen squeezing out around the bond joint edges
Glue appears to be mixed in a tub by hand and not de-gassed in a vacuum chamber, this removes any included air bubbles which will weaken the glue strength
The carbon bond joint surface appears to be uni-directional fibers ie all in one direction, instead of a more stress friendly woven fiber +/-45 diagonal direction
Resin weight % to fibre weight appears very high as it's possible to scrape off excess resin, the excess resin wouldn't lend the carbon to being cured under pressure in an autoclave, at best cured under vacuum which makes it far less structurally sound than it could be if autoclaved cured
I'm not a composite submersible structural specialist but have design many titanium/carbon bonded structures 100x smaller in scale and I've been down to 150ft on a tourist submarine
