FWIW plenty of aerospace components are welded using gas-shielded processes these days.
Totally agree, but I was referring to critically stressed components and those materials subject to inclusions such titanium, hence my reference to titanium.
there are all kinds of joining processes used on all kinds of 'critical' parts these days. 'Inclusions' are not usually the main issue with Ti weldments, and they certainly are not the main reason that you might choose to use a vacuum-based process over a gas-shielded one. The shield quality is critical in gas-shielded Ti weldments simply because oxygen and nitrogen dissolve and then react strongly with Ti. In some processes this is obvious because the appearance of the weldment is altered. In others it isn't.
In bike frames you (the customer) don't get a chance to see because the back beads are not visible and the top beads are always cleaned using abrasive or similar.
In the early days of Ti welding it was thought that a hardness increase in the weld might be a good indicator of contamination; however careful tests (which IIRC are summarised in Lancaster's book on welding metallurgy) showed that a hardness increase of about 10-15HV (which is only just detectable amongst the scatter in the data) in Ti 6,4 could correspond to a reduction in ductility of about 30%, i.e. the weldment could be seriously compromised without there being much sign, hardness-wise.
There are many different ways of making unsound Ti weldments and after the job is done telling the difference is very difficult. When buying a Ti frame it is a leap of faith, hopefully backed up by a good track record and a good warranty.
cheers