wow thank you for all the feedback chaps, and also apologies for digging up that old fillet chestnut.

Interestingly, I've not used the 2-rail sweep method before (usually I just loft two curves).


Avoiding segmentation seems to be the key, but it happens to easily it's a bit frustrating. If you look at the original model attached and start diagnosing it by selecting naked edges, individual edges etc it all looks, well, pristine. (I did say "looks" ;) )

It's not until we start adding fillets, or doing a shell operation that problems become apparent as it all goes to Sh!!@

Is there any other methodology for diagnosing problem areas before taking a surface further towards the end result ?