Scuttlebutt Website SCUTTLEBUTT
Who's Online WHO'S
Log in LOG IN         

Forum Index: .: Dock Talk:
Canting keel failures
Team McLube


The Publisher

Dec 23, 2008, 7:49 AM

Post #1 of 3 (13561 views)
Canting keel failures Log-In to Post/Reply

On December 18, 2008 during Leg 3 of the Volvo Ocean Race, Delta Lloyd experienced a massive failure to the port side bulkhead that attaches the keel canting hydraulic ram to the boat. They were able to center the keel and complete the leg, and while it was the first significant keel failure of the 2008-2009 VOR, it did bring back memories of the many failures from the 2005-2006 event.

From Andrew Troup, New Zealand:

Once again keel canting rams - or the structures mounting them - are failing offshore. The most recent example is the wonderfully cosmopolitan Dutch/Irish VOR entry with the highly respected Argentinean designer. This is actually a 'much-modified' ABN Amro One, so some of what I have to say may not be applicable in this specific instance, depending on the extent of the modifications.

If a pattern of failures emerges in a new generation of VOR competitors, this may be a cause for deeper concern. Ahead of this possibility, while there's still potentially time to do something about it, I care to provide a few observations, hopefully in sufficient depth to at least be meaningful, possibly useful. It goes without saying that this is the last area one would *want* to see failures. It is also the last area I would *expect* to see failures, for reasons I hope to explain:

1) Hydraulic rams, and their support structures, hardly ever fail in other well-resourced disciplines

2) For relatively new installations, overloads which could cause catastrophic failures are likely to be up around the nominal breaking load for the ram.

I say this because there shouldn't have been enough load cycles already to trigger fatigue failure, which can happen at much lower, and much less predictable, load levels.

Furthermore, the structure around the ram should be at least as strong as the ram.

3) Externally applied loads which exceed internally generated forces can be difficult to handle in engineering mechanisms. Such loads are notoriously high and unpredictable for canting keels.

Hydraulic rams are almost uniquely well suited to routinely shedding the energy content of external overloads, protecting themselves and their surroundings, hence point 1 above.

In this mode, the keel would "relax" away from the set angle, in much the same way a suspension strut retracts when a vehicle hits a severe bump.

Having shed suddenly applied, high-energy loads, adding the capability to return automatically to the desired angle is straightforward.

4) The devices and circuit design to achieve all this - in spite of equipment failure - are relatively well understood by industrial engineers with experience in fields where such load situations are common.

Individual hydraulic components and peripherals are not immune to failure, but the natures of these individual failures tend to fall within a fairly small set of possibilities, and hence 'fail-safe' design, IN RESPECT OF SETTING AN UPPER LIMIT TO EXTERNAL OVERLOADS, can be implemented with such rigour as to be (almost) true to label.

Such circuits and devices need not entail a serious weight penalty.

The fact that ram and structural failures have been so common in this application suggests to me that substantial pockets of people have been engaged in this field despite unperceived deficiencies in knowledge and/or experience relating to design of these mechanisms and circuits. This might arise because the latter are relatively newly applied to sailing, although not at all new elsewhere.

No individual should take offence at this suggestion: taken on its own, any one failure could easily arise from hidden inherent material defects, rather than faulty design.
However there have been so many instances that this simply cannot be true in all cases.

Where the failure is structural, arising at loads within the nominal working load for the rams, the problem could hypothetically be with the engineering of the structure itself.
However I am confident that this is unlikely to be the case, because the people concerned are very experienced and skilled at this sort of structural challenge, and are well accustomed to working with the materials concerned.

To me, as a mechanical engineer, the other structural challenges routinely solved in design of these yachts are very difficult, because the loads are so hard to quantify.

In stark contrast, the challenges around the canting loads of the keel should be relatively straightforward, because a strict ceiling can be set for the loads which can be developed or applied.

Hence my frustration and perplexity that the problems have not been laid comprehensively to rest, even at these elite levels of practice.

The Publisher

Feb 3, 2009, 9:39 AM

Post #2 of 3 (12519 views)
Re: [The Publisher] Canting keel failures [In reply to] Log-In to Post/Reply

By Kenny Robertson, Associate Member of the Institution of Mechanical Engineers (AMIMechE), Comrie, Scotland:

With all the talk about failing keels recently, I remember my reaction midsummer this year, when Yachting life had pictures of a fabricated canting keel being manufactured. Absolute shock. Whoever designs these things needs to dig out BS 7608 [Code of practice for Fatigue design and assessment of steel structures] or other related standards such as PD5500 Annex C [pressure vessels - my thing] for advice on fatigue resistant design and fabrication.

Without getting too involved, the current advice is that all steels in a fabricated [i.e. welded] structure, no matter how 'strong' they are, or where the welds are, have pretty much the same fatigue life, and that fatigue life is related solely to how they are fabricated and inspected. The most fatigue resistant design is one where the welds can be fully inspected by NDT for external and internal flaws from both sides of the weld and have a smooth profile. The worst ones contain welds that cannot be fully inspected and have 'crack like' features.

Alas the pictures I saw fell into the latter category. The keel was full of fillet welds [triangular section welds] laid down by the MIG welding process. MIG welding is fine for many things and can be used easily by relatively unskilled personnel, but it is not used on critical things like pressure vessels for one simple reason. The weld is of relatively poor quality and does not penetrate far into the joint. As for fillet welds, they are triangular in section cannot be meaningfully inspected in anyway and always contain crack like features at the welds edges unless they are ground smooth afterwards and always a crack feature at the weld root.

The designers of these keels need to specify full penetration, preferably double sided welds that can be properly inspected, using properly qualified and tested welders and welding procedures, using base materials that have been inspected for internal flaws and laminations - otherwise the keels will keep on falling off.

Making some huge assumptions and playing with some figures based on an imaginary keel I reckon the difference in fatigue life between the two models are: - full pen weld from both side, full NDT etc and smooth profile = 9188hrs; fillet welded construction = 1187hrs. That is a huge difference in working life. This ain't rocket science - it is common engineering pracitices and there is no need for these common failures.

Bruce Thompson

Aug 19, 2011, 6:32 AM

Post #3 of 3 (10157 views)
Re: [The Publisher] Canting keel failures [In reply to] Log-In to Post/Reply

Once again a canting keel has failed on a boat going to windward in moderate conditions that are likely to create wavelengths near the length of the boat thereby maximizing the coupling between the wave period and the natural pitch frequency of the hull. The resultant harmonic resonance is what is destroying the keels. Of course it is when going to windward that a canting keel has the largest benefit, so designing a keel so as to not fail is antithetical to going fastest. Expect the keel failures to continue.

Viewing the Forums: No members and guests

Search for (options) Contact Forum Forum FAQS Markup Tags Forum Rules