CHEEKI RAFIKI: Hull Found Again, Post Mortem

26 May

Cheeki hull again

The fate of the four crew members aboard Cheeki Rafiki was confirmed on Friday when the U.S. Navy again found the overturned keel-less hull and inspected it closely enough to determine that its liferaft was still onboard. So with much drama and angst and effort we have at least confirmed what the U.S. Coast Guard initially surmised when it first suspended its search for survivors. I don’t think the effort was wasted or useless. Given the enormous interest in the fate of these four men, I think it was well worth it to achieve closure on that point.

I would hope some people who criticized the Coast Guard rather harshly for suspending the search might now express some regret (I noted, for example, that Brian Hancock, a well-known racing sailor, accused the Coasties of abandoning the search “without really trying”), but I’m not holding my breath on that. What’s more important is to focus on what we can take away from this tragedy to make sailing safer.

Capsized boat without keel

Time to wake up! This happens all the time

I’ve seen people discussing liferafts and such, but for me this big issue here is keels. The four crew on Cheeki died because the boat’s keel fell off, probably very suddenly, and this is not, as some have suggested, an unusual occurrence. It is frighteningly common. Modern fin keels fall off cutting-edge high-end race boats all the time (e.g., keel loss is a common reason for Vendee Globe withdrawals) and off less exotic race boats (e.g., I have one good friend who lost a keel off a TP52 while racing and know of many other similar incidents) and off common production boats, both while racing and cruising.

Cheeki damage

The underside of Cheeki Rafiki, showing the area where the keel ripped off. Note the large swath of damaged laminate below the keel’s footprint

On production boats like Cheeki, a Beneteau First 40.7, it is probably true that most keel failures are the result of damage sustained in groundings. This is a tricky business, as grounding damage can be very hard to assess accurately, and damage can be cumulative over several groundings. Even worse, with charter boats like Cheeki, there may be one or more groundings that take place and are never reported to the boat’s owner or those responsible for maintaining it.

For an excellent discussion of the damage sustained on Cheeki, I recommend you dive into this Sailing Anarchy thread here, from whence I pilfered these photos:

Cheeki damage detail

Enhanced out-take of the keel’s footprint from the image above. Questions raised: 1) are those bolt-heads and washers we see on the two forward keel bolts? Or are they broken off? 2) the aft bolt clearly seems to have been corroded, so is this where the trouble started? 3) the central bolts seem to have been the last to let go and took with them a big chunk of laminate, but was the laminate under the keel cored?

Barracuda without keel

Another First 40.7, Barracuda, that lost its keel. Note the similarity in the damage to the underbody

First 40.7 keel bolts

Keel-bolt pattern on a stock First 40.7, as seen from inside. Note that the keel’s attachment points are not tied directly into the structural bilge grid. Also, this is an exceptionally shallow bilge!

Interestingly, on page 7 of the SA thread you’ll find one participant,, who claims to have been on two different supposedly undamaged never-grounded First 40.7s where the aft keel bolts started weeping when you honked down hard on the backstay. Another commenter, axobotl, claims to have been on a First 40.7 that grounded at hull speed without sustaining any detectable damage.

Thinking of that rusty aft bolt on Cheeki, I have to wonder if this is a weak spot on all First 40.7s that have been raced hard. (And there are a lot of them. They have an active one-design thing going on.) If you trap moisture against that bolt every time you crank down hard on the hydraulic backstay adjuster, corrosion seems inevitable.

In perusing the online commentary, I’ve seen that some people don’t believe it is possible to engineer a bolted-on fin keel that is not vulnerable. That this is a risk you have to take when sailing on boats like this.

Personally, I don’t accept this. I’m not an engineer, but I have to believe it is possible to design a keel attachment that spreads loads over a much wider area of the hull. After all, we never (or at least almost never) hear of wings shearing off of airplanes. Yes, I am sure “over-engineered” keel attachments would be heavier (and thus would decrease performance) and more expensive (thus less economically attractive), but they must be feasible. On page 6 of the SA thread, for example, you’ll find links to a patented Swedish system for attaching a fairly aggressive fin keel that looks incredibly strong.

As a starting point, I would say a “properly” engineered fin-keel attachment should spread loads over such a large area that you should need to effectively destroy the hull to remove the keel (like on a full-keel boat). Also, there should be some mechanism or “fuse” that lets you know when the assembly has been critically damaged.

I can only hope that all the energy that went into browbeating the Coast Guard to continue looking for Cheeki might now be channeled into this purpose. Then the crew of Cheeki would not have died in vain.

How do we create this new standard of construction in what is effectively an unregulated industry? It would help a lot, I think, if race organizers and rule mavens started the ball rolling. If the high-end race boats whose keels fail most often were forced to be safer in this regard, a lot would follow from that.


  1. Sean

    It’s funny to hear the thought of engineering a stronger keel/sump joint will add unwanted weight. I hear this often and I can only shake my head at the thought of how stupid it sounds. Increasing glass thickness and stronger structural beams in this sump area will add lots of weight…unwanted weight? Isn’t this area the same spot they’re bolting 1000’s of pounds to? I think the increased weight of the improved structure will be low enough to be incorporated into the keel displacement. There I have said it and now it will be passed onto engineers/ builders…I have saved the industry with 5 minutes of my time…you’re welcome

  2. Frank Riddle

    Dear Reader, There are two serious issues here both of which require the EC directive to be upgraded so that safety is adequate for Ocean rated boats.
    issue 1.PRIMARY SAFETY; The keel should never fail but if it does the hull should remain buoyant and not invert. (see separate drawing)
    2. SECONDARY SAFETY Liferaft deployment is very poor in many boats.
    Frankly the designer or those certifying the craft as A rated should think about what they have done. Sailors need systems to work.
    Pr0popsed changes for improvement of the directive.
    In view of the unnecessary loss of life on Cheeki Rafiki Sailing Boat First 40.7 when the keel broke away from the hull and the boat inverted making the life-raft inaccessible. We hereby propose that the following measures are considered,taken forward and developed for amendment of the EC directive so as to ensure Ocean rated sailing boats are of adequate design, fit for the purpose and equipped with satisfactory secondary safety equipment.
    1. DEFINITIONS. I presume abnormal conditions means something not regularly found in the world’s oceans.
    This needs revising to be more specific about what is abnormal and exempted because you can’t run for cover mid Atlantic. Even submerged containers are normal in today’s world and this requires adequate secondary safety of floatation after collision and life-raft deployment.
    2. –
    3.1. Structure. Add after “strong enough in all respects”. “to cope with all conditions frequently found in the world’s Oceans” .
    Add For Cat A Yachts.
    i) An impact test standard should be established for the hull skin.(too many flimsy single skins about)
    ii) A force test shall be applied to the keel when the boat hull is on its side to ensure the fixing is of sufficient strength to cope with all conditions found in the world’s oceans.
    This force N can be established from experts in the industry and experience with earlier failed designs like Volvo, Drum etc. e.g. Weight of hull fully laden x 1.5. In theory this can be done by calculation/virtual testing but an actual test is better proof.
    i) A fatigue test shall be performed when the hull is held as shown in my sketch (out of the water in free air) the fixing of the keel to the hull shall not separate or degrade after the cyclic test has been performed by the maker . This test should involve a force when N x times the weight of a fully laden hull in running order is applied at set or variable intervals over a long period.
    E.g. A cyclic test of 100,000 forces of 1.5 x hull weight applied to the keel end and repeated every 1 second at 7 second intervals in a test rig where the hull or section of it is held firm. This is to simulate maximum wave force x 1.5, or whatever the experts agree.
    ii) Keel tether retention; There shall be at least 2 tethering straps or cables, capable of standing continuous flexing, that will hold the keel to the boat should it break away from the hull so as to provide a righting moment to assist hold the boat in a horizontal knockdown position.

    3.3. Buoyancy and floatation;
    Delete the word “multihull” in the second sentence.
    Add “all fin keeled boats where the keel is not part of the hull manufacture but is added and fixed later by bolts or other such retaining method and where those fixings are, the main method of retention shall be fitted with a foam filled mast and an automatic masthead floatation device sufficient to prevent the boat inverting”.
    The latter should be retrofitted and perhaps applied to all A category.
    Add; All hulls of A category shall contain buoyancy in their hull so that they remain afloat after a knockdown to provide survival space in the main cabin.
    Add; Any A or B cat. Yacht Suffering a knockdown shall float for more than 1 hour so that the mast and 50% of the hull shall not sink below the waterline. Test on flat water with full fuel, crew load etc.
    The EPIRB must transmit in this condition.

    3.7. Life-raft stowage.
    Add; “Either i) a manual unrestricted launch of the liferaft using one hand (the other one is holding you stable) shall be possible without lifting 0r ii) Automatic deployment of the life-raft sufficient to clear the hull shall be installed with a short painter and a long one so that lifelines can be transferred from the jackstay to the short painter”. This needs testing and development.
    E.g. a spring loaded or pyro technic powered launch system from an envelope at the transom. Outside the rail.

    3.8 Escape
    Delete the word “multihull” .

  3. Pingback: Another yacht abandoned, 450 NM SE of Cape Cod - Page 2 - SailNet Community

  4. Simon

    The keel-bolts in the last photo are clearly stainless steel. An engineer will tell you that the fatigue properties of stainless are not as good as those of carbon steel. Stainless ‘work-hardens’ earlier and fails due to fatigue. Given the clean shear of the fore and aft bolts on Cheeki this looks to be the issue. Carbon steel bolts will corrode – and hence will need regular replacement, but that is better than them looking all shiny and snapping mid-ocean. can Beneteau produce the fatigue calculations for these bolts in partially corroded condition? What is their recommended change-out interval?

  5. Martin

    These boats are under constructed for ocean cruising! it’s about time designers wound back the clock and stopped building yachts that are meant to be “cutting edge ” and over classing them !! the results have just been shown and people have lost their lives. Cruiser racers are fine in coastal waters but it is blatantly obvious they are not capable of crossing oceans and the design specs should be changed before more lives are lost !

  6. Steve

    Two forward bolts and rear bolt snapped. The 3 double center bolts tore off bottom of boat. All at one time? Hard to say. In any case, the hull was not strong enough. Is the “structural grid” in photo just a liner? It looks like it. If so, it only supports the floorboards and adds little to strength of hull. Beneteau needs to beef-up their hull laminate. It looks pretty thin in the photos of broken, awash hull. Some older production boats had either a “pocket” of built-up laminate into which the keel was inserted and subsequently bolted, or an area of built-up laminate onto which the keel was bolted. I don’t see either here.

  7. Patrick

    Very sad loss. Engineering design of robust keelbolts is easy. Cost efficiency makes it less strong. Regular checks required. Especially if about to cross Atlantic.

  8. Henry

    I am sorry for the loss of the crew.. RIP.. But I wonder, did the Coast Guard or Navy dispose of the vessel or just let it float off and still be a hazard to other vessels??

  9. Brian Wernham

    A simple solution would be a sensor wire along the bottom of every keel. This could be easily tested for a continuous circuit before every race and/or charter. If it is broken, then a diver would have to go down to replace it and report on what the cause of damage was…

  10. Sea Wind

    My condolences to family and friends who have lost their love ones too soon. This tragedy should not turn into a finger pointing session. I agree 100% with the author, in that we should examine what led up to the keel’s ultimate failure, and look into minimum design parameters to create a boat that is can do open ocean travel safely.

  11. Tim

    The last photo clearly shows the keel bolts passing through thick metal backing plates (load spreaders) inside the structural grid. The bolts are separated by thick ribs that make the structural grid. So the load is spread by the backing plates into the thick structural grid which goes all the way up to chain plates and is laminated to the hull. It’s a strong structure.

    Note in the Cheeki Rafiki photo, the forward two bolt and aft-oat bolt holes are perfectly round and intact – showing no signs of stress. Were the bolts loose? Did the keel get revolted without the backing plates? WIth the fore & aft bolts gone, that would concentrate the keel stress on the center bolts which clearly ripped through several thick layers of laminate.

  12. Andy

    This is the type of accident that terrifies me the most, and also the kind I hear about the most. These keel attachments are clearly not built heavily enough on these production fin keel boats. I once saw the “How Its Made” episode where they show how Catalinas are built and I was shocked, reminded me of a preschool papier-mache project. What criteria does the USCG use when they certify these boats anyway?

  13. Rick

    I worked in and taught FRP for many years. Unless the close-up is deceiving, I see brittle fracture on the port side off the damage. I see no evidence of woven roving or any use of advanced composites such as Kevlar or carbon fiber.

    My condolences to the family and friends of the lost crew. To Beneteau, if you are using only mat or chop in this area, you deserve all of the bad press this tragedy will bring your way.

Leave a Reply

Your email address will not be published. Required fields are marked*. Comments are moderated and generally will be posted if they are on-topic and not abusive. For more information, please see our Comments Policy.

More from the AIM Marine Group