Another Keel Failure

[flaming]

Interestingly It's not the first fairly new Ker designed yacht to have had a replacement keel. I am aware that Software Mistress a Ker 36 went through at least one keel iteration. When it was bought by a new owner it came with a spare keel of different design! I understand the replacement was intended to increase performance, nothing to do with structural issues as far as I know.

Software mistress, now Skermisher, has never been the most competitive boat. Doesn’t surprise me at all that they’ve tried a few mods in search of some speed. Shame, I’ve always thought it was a very good looking boat.

 

[JumbleDuck]

A Ker 40 is not an extreme race boat. And it was on a delivery at the time.

I accept that racing people go out in weather which keeps the rest of us snugly tucked up, but cannot help wondering if it was really necessary to do a delivery in 48kt (F9/F10 border)? Or did it blow up unexpectedly?

It is a huge assumption that this is a design issue rather than a fabrication / manufacturing issue.

The designed of the keel also made it, so can't really dodge responsibility either way. Also, as was posed before, it is the designer's responsibility to specify exactly how fabrication is to happen and to what standard. The designer should anticipate potential manufacturing problems with, for example, a test regime.

 

[Heckler]

It always seems to me to be generally bad engineering to design something so crucial as a keel and then to rely on a welded joint at exactly the point in the structure that is under the most load.

Putting a type of join that it is very difficult to be entirely sure about in such a highly loaded and safety critical part of the structure simply looks like poor design.

Structural engineer?

 

[JumbleDuck]

It always seems to me to be generally bad engineering to design something so crucial as a keel and then to rely on a welded joint at exactly the point in the structure that is under the most load.

Putting a type of join that it is very difficult to be entirely sure about in such a highly loaded and safety critical part of the structure simply looks like poor design.

There is nothing wrong with having a weld in a critical place as long as the weld is properly designed, properly made and properly tested. I read an MAIB recently in which a weld failed ... it would have been fine but as I recall it had been made about half the design size and on one side only rather than both.

 

[Birdseye]

My point is more that putting a weld in that position puts the whole security of the keel to hull join onto a type of fixing that it is possible to get wrong without it necessarily being obvious that you have done so.

Its possible that you are both wrong and right. You are wrong in that its quite possible to weld, make it as strong as the parent metal and be 100% sure. In my working days we supplied welded parts to nuclear reactors and aircraft engines in steel, aluminum alloys and titanium and with zero failures. We used to weld big ship crankshafts together throw by throw.

However its not cheap or easy to work to that standard and I would be very surprised if boat builders or their keel makers did so. To put it bluntly, the cost of failure in a nuclear reactor and a yacht are a bit different.

 

[Laser310]

In spite of what Flaming says above, the Ker 40 is a pretty high performance race boat.., and it's not unusual for race boats to go through several keels. I raced on two different boats that went through three keels each.., and they weren't as high performance as the Ker 40.

This situation is a little bit unusual: the owner apparently rejected the original designer's (Ker's) design for a new keel and went for something else... Also, as a race boat.., there is not really any issue with compromising the below-decks layout to get the best possible keel attachment done.

While in this case, it looks like it wasn't the attachment that failed.., rather it was a weld in the keel.., i think it certainly could be argued that the keel attachment method was a factor in the loss - had they not been using a "bolt-on" method.., there wouldn't have been a weld like the one that failed.

I think it is pretty clear that for very high aspect keels, with narrow struts and big bulbs.., the best attachment method is the plug and socket arrangement secured by transverse bolts or rods, where that socket is pretty tall, and is integrated into the beams reinforcing the hull. Yes - this compromises the interior, but on a race boat that is acceptable.

Boats that can not make this compromise, should have lower aspect keels.., with smaller bulbs.

what's scary is that even moderate keels on moderated performance cruiser/racers are failing because of poor design.., or designs that are not tolerant of any mistreatment.., and which are entirely inappropriate for that kind of boat. Ceeki Raffiki, for example.

Those 40.7's are getting quite old now, and are often operated on a shoestring budget. I follow a few FB groups where boat owners look for crew to help with deliveries - just yesterday i saw another one of the pay-to-race operations looking for volunteer crew to get a 40.7 from the Caribbean back to Britain - as these boats age, we are going to see more of this type of "accident".

 

[flaming]

Structural engineer?

Compulsory structural modules in the first years of my degree, but no.

My point, as better explained by Birdseye below is that whilst a weld can be just as good as the metal it is joining, it can also be bad, and undetected. Thus it doesn't really matter if your FEA analysis comes back saying it's fine to put a weld here if you always have a little bit of doubt that the weld was actually good.

The person who is actually a structural engineer specialising in keels here, Jason Ker, had the following to say on the subject.

The keel, designed by the owner’s local Naval Architect, was configured as a vertical hollow strut welded to a horizontal mounting plate, which in turn was sitting in a shallow recess in the boat’s hull and bolted to the boat’s structure. This is not a design approach we would ever endorse as the high stress point at the junction is coincident with the horizontal welded joint.

Showtime's keel not designed by Jason Ker - MySailing.com.au

 

[Keen_Ed]

Jason Ker wrote an interesting article about the whole thing in the current seahorse. The main problem with the welded-plate-to-keel-fin structure is that the weld is at the exact point of the stress concentration. He notes that he's happy to design a welded keel structure - and that the "bolting a plate to a recess in the hull" is a perfectly valid system, but one with proper radiuses to minimise stress concentrations.

Time to take stock - Seahorse Magazine