Cynthia Woods verdict

[Al - s/v Persephone]

December 19, 2008
Report: Poor Repairs Caused Texas Boat to Capsize
By THE ASSOCIATED PRESS
Filed at 9:43 p.m. ET

TEXAS CITY, Texas (AP) -- The deadly capsizing of a sailboat in the Gulf of Mexico in June probably resulted from improper repairs to damage caused by previous groundings, U.S. Coast Guard officials announced Friday.

The sailboat Cynthia Woods capsized June 6 after its keel broke off, killing one crew member and stranding five others for 26 hours in the Gulf of Mexico. The Texas A&M University racing team was competing in a regatta at the time.

The Coast Guard investigation found the sailboat, despite previous groundings, had not undergone any major repairs, examinations or marine surveys.

Coast Guard Cmdr. Jim Elliott said Payco Marine, a Galveston boatyard, reattached the keel early this year after the most recent grounding, but was not asked to do any other work on the boat. Investigators did not fault Payco Marine in the accident.

Additional work on the damaged area was reportedly made by students who were not specifically qualified or trained to make those repairs, according to the investigation. ''At no time did the sailboat manager who was overseeing the operation and repairs to the vessel consult with the vessel designer, manufacturer or a third party surveyor regarding this damage,'' Elliot said.

The sailboat manager, Jim Atchley, was fired about a month after the accident for failing to disclose his criminal history when he applied for the job., Texas A&M officials said. His firing wasn't related to the sailing team.

Atchley said Friday that after the boat came back from Payco Marine, he thought no other repair work needed to be done. He said that some students did do cosmetic work on the keel but he stressed that all repairs were done by Payco Marine.

''There was no reason to doubt Payco Marine didn't do the repair work correctly,'' he said. ''They have always done repair work very satisfactorily.''

Investigators concluded the design and manufacture of the keel were within industry standards and guidelines, Elliott added.

 

[sailaway78]

And let the lawsuits begin

From the get go, this has sounded like a CYA deal for those involved. Looks to me like Payco Marine and Texas A&M had better get ready for a bunch of lawyers knocking on the door. And anyone else who has been close enough to touch the boat for that matter.

 

[Ross]

I wonder if boats of this design come with a notice that states the need for structrual inspection following a grounding.
I am never concerned about grounding my boat on anything less than rocks because of the way she is designed.

 

[kloudie1]

Cynthia Woods Keel

Interesting, I saw the article this morning as well.. When I saw the pictures on the net of the keel after recovery form the Gulf, there was a chunk of glass/composite attached to the top and the bolts/nuts were still intact.. Pretty plain that the keel pulled out the bottom of the boat at the attachment. This blade keel arrangement just doesn’t look to have enough bearing surface at the top to take care of the immense loads generated. (OK, I am a mechanical and not a marine Injun-Ear, but loads are loads!). Probably OK when in perfect condition, and the repair was probably as good as a “normal” boat would require… but not OK to put a “normal” repair on a boat that has an abnormal appendage !! That attachment is less than a foot across and it has to handle the load of the bulb which is 6 feet or so away..

 

[Ross]

Kloudie, try to design a cantilevered shelf like that to serve as a work bench in the shop.

 

[kloudie1]

Ed Zachary Correct

Kloudie, try to design a cantilevered shelf like that to serve as a work bench in the shop.

You are exactly correct, Ross.. Wouldn't even hold up a six pack of beverages!

 

[Roger Long]

As one who has seen photos of the keel, talked to the builder of the prototype boat about subsequent changes to the design, and designed a fair number of keels and other marine structures myself, I am very surprised by this verdict.

As one who has been involved in marine accident investigations, forensic analysis, and participated in or followed numerous inquiries, both in the US and the UK, I am not the least bit surprised.

For the record, I have every confidence that the USCG finding that the keel met all industry standards is correct. What I can not see however, from the limited information available to me, is how the keel and bottom structure could have had the reserve strength to continue safely in service after the slightest bottom contact without total disassembly and reconstruction of the supporting hull structure.

For the record, I am sure that the builders of the vessel had adequate safety margins for the strains of foreseeable grounding events. However, they must have been using methods and materials unfamiliar to me to have achieved this within the keel footprint and interior clearances of the hull because I’m damned if I can figure out how they did it.

Formal inquiries are funny things. I had a chance to talk for an hour with the USCG officer who presided at a similar inquiry that came to a clearly erroneous conclusion and he told me basically that he knew I was right but inquiries are a processing of public record and information and you can’t just inject your gut level common sense into them.

In another inquiry, I proved conclusively that it only took a 22% increase in wind speed to capsize the casualty vessel when she was sailing at her normal maximum heel angle. Since gusts 50% over nominal wind velocity are common, this made her a death trap. The inquiry found that, since there was no way to know the speed of the wind gust that actually sank her, it could have been one greater than any vessel could be expected to survive and my finding therefore did not speak to cause. This vessel was a film prop and didn’t even have any real ballast. They still needed to have a full inquiry to determine why she sank. The owner was exonerated.

BTW if you have seen the two History Channel shows, 2006 and 2007, on the sinking of the Titanic, you’ve seen me with my accident analysis hat on. The book that came out in October, “Titanic’s Final Secrets”, by Brad Matsen, is in most bookstores now and discusses my trip down to the see the ship and subsequent analysis in detail. It's a great read and there are still three more shopping days until Christmas

Later note: Subsequent revalations by other have considerable changed my views on some of the things I said in this post but read the whole thread as a dialogue.

 

[kloudie1]

Great stuff, Roger ! I have done a good amount of accident investigation/failure analysis for the oil guys, but nothing so notable and historic as what you are talking about ! Wow! That must have been a real adventure!! Congrats on that! Yes, I have seen the same problems with committee led investigation boards.. Lots of times, the politically expedient answer is the posted one!

 

[Franklin]

strainght up fact....the boat hit ground several times before race....the boat was repair....keel bolts weren't tight enough....boat was beating against 14' waves (I was there, I know) and keel wore out fiberglass so keel fell off.

My boat has the same amout of backing on those keel bolts as it does but my boat made it just fine.

As the USGC said, faulty repair, but boat manager should have had it inspected by a survaver. This is a lesson learned for all of us. Keels are the most imporant thing on the boat and the one and only thing that can sink a sailboat in seconds. This means that ANYTIME a boat goes offshore, it should have a "pull on this line to release safety gear" for all lifejackets, liferafts, handheld radio, mirror, whissle, flairs and so on.

The boat went over very quickly. Three crew members were down below including a safety officer who I met at the safety meeting for the race. He noticed the water inside and woke up and got the other two out the compainionway but he didn't make it in time. The last guy couldn't get through the compainionway with his lifejacket on so he took it off. 5 people shared 4 jackets (reason for having 5 jackets on the quick-release).

GPS tracker stopped responding so CG was called. It took CG 26 hours to find them with coordinates. Only way they found them is because night came and one crew member had a flashlight with him. 26 hours in the water with planes flying overhead and no rescue.....talk about demoralizing.

So check your keel bolts!!!!!!! and have a quick release of all necessary safety equipment for all on board.

I learned my lesson form this....have you?

 

[Roger Long]

Thank you

strainght up fact....

Good to have information straight from the horse. Would you mind adding to our understanding of this incident by answering some additional questions?

I see in the picture posted above that the backing blocks remained on the keel and the keel bolts are intact. As the inquiry found, the keel was clearly strong enough. There must be remnants of the fiberglass hull attachement left between the backing plates and the top of keel. I think I can even see them and their sheered edges.

Do you think this was a failure of the hull supporting structure?
How did loose keel bolts contribute to this?

If they were loose, all I can think is that the motion of the slight slack created impact loads well in excess of the design conditions.

The two pictures below were taken after the grounding and published shortly after the accident. Note the extensive crushing of the fairing block at the trailing edge and the tension separation evident at the front. Do you think this can be explained other than by significant deformation of the bottom structure?

Do you think simple retorquing of the bolts would have been sufficient to restore a fiberglass structure deformed to this extent to original as-built strength?

You say your backing blocks are the same size. All forces on a structure like this are translated into vertical strains. The backing blocks appear to be just about the same size as the keel so the fiberglass hull is sandwiched between them and the two clamping surfaces put the fiberglass skin in sheer. I can't see any evidence in the picture previously posted of the bolts and backing blocks being above any floors or other structure to spread these forces out into the hull instead of concentrating them at the edges of the backing plates.

We now know, courtsey of the USCG, that the entire structure was adequately designed so can you explain how the hull bottom was reinforced to stiffen it and carry keel strains out from the keel?

You report your keel backing blocks being the same size. Are they in essentially a flat bottom panel as seen in the photos below or is there a molded stub so that the fiberglass skin quickly turns 90% close to the backing blocks stiffening the area locally, bringing the major stress in line with the skin, and minimizing the secondary panel flexing that can take place right around the backing plates?

BTW do you note in the pictures the outline of an original, larger footprint, keel evidently contemplated at the time the hull mold was constructed?

 

[Franklin]

Roger...I'm sure you know that a loose keel will rock side to side and each time generating much greater force on the fiberglass it's bolted to then it would if it was secured tightly.

 

[Roger Long]

Roger...I'm sure you know that a loose keel will rock side to side and each time generating much greater force on the fiberglass it's bolted to then it would if it was secured tightly.

Yes. And, I said so in my post.

So, that's really all there was to it? Did anyone notice or feel anything before the keel came off? Was water seen coming in around the working keel bolts?

I guess then an hour with a big socket wrench and a pipe slipped over the handle would have prevented this accident. That really makes it a tragedy.

 

[John]

I still cannot see how this design is not inherently unsafe, or at least less safe. As others have pointed out, the small area of attachment to the hull means a greater stress per square inch. Maybe there were other problems - loose keel bolts, repairs not made properly, etc. - but it seems to me that part of the problem must be in the very design itself. As for the boat having been built to "industry standards" - who sets those standards?

Some time ago, I read an article about exactly this issue for boats with this design keel. The article commented that there were several cases of exactly this happening. I thought I'd read the article in Practical Sailor, but I can't find it there.

 

[kloudie1]

Roger, here is another shot of the keel top.. Pretty high res.. you can zoom in pretty tight to look at it. The pics are from the Texas A&M website.. Look at how the aft backing plate is dished.. Assuming it was straight when built, this pretty well shows that the keel had been doing a pretty good transverse wobble.. I don’t doubt the Coastie’s statement that it was designed properly.. I did a thumbnail calc with some rough assumptions and it has plenty of glass to hold it IF everything is in great condition.. The first picture I posted shows the forward plate is dished too. Man there is a lot of 5200 on those center bolts!

 

[Roger Long]

Wow. Talk about putting your dirty laundry out in public. Those are great and detailed shots:

www.tamus.edu/.../cynthiawoods/keelphotos.html


This one in particular:

http://www.tamus.edu/offices/communications/cynthiawoods/images/keel2.JPG

I can see the very slight dishing but I'm not sure it conclusively shows wobble. It's similar to the dishing you might see if the material under the plates was crushing from the initial bolt tensioning. More indicitive are the apparent attempts to stop weeping and leaking. I can't see any other explanation for the white stuff smeared all over every bolt. If someone torqued these bolts after the grounding, and it's hard to imagine they didn't, I find the dishing much more suggestive in terms of that not stopping the leaking than as an indicator of wobble.

This is worth reading:

www.sailinganarchy.com/article.php?get=1738

You will note that, at least according to the press release, the USCG is careful to say that the KEEL met industry standards. This doesn't necessarily mean anything but the omission of AND HULL is the kind of careful hair splitting often found in such reports. There are some fairly clear and simple standards for keel bolts that the keel could easily be checked against. Supporting structure is a much fuzzier and difficult to area to get into and there are lots of reasons why the USCG might just not go there. They like to close cases and move on.

In the theoretical perfect engineering design, everything lets go at once. In a weight critical structure like a racing boat, that tells you that you weren't carrying around anything extra. That isn't achievable in the real world and second best is having the failure mode be least catastrophic.

If a keel had come off of one of my designs, wobble induced forces not withstanding, I would like to find broken keel bolts. I would then know that the supporting hull structure had been compatible in strength with the keel. The bottom of the hull is the least weight critical portion of the boat from the performance standpoint so a lot more strength could have been obtained here with little penalty. If this result had been achieved, the hull bottom would have remained watertight. This might have only extended the floating time slightly due to hatches and other topside openings but only a very short time period could have greatly increased the chances of survival for Roger Stone.

 

[Ross]

There is one rather simple premise. Wings don't fall off airplanes and keels don't fall off boats. When you tear the wing off an airplane the damage to both the wing and the airframe is severe. When the keel gets torn off a boat the damage to the keel and to the hull should be severe. Somebody skimped on supporting structure in the hull.

 

[sailingdog]

I think Ross may be right. The damage to the hull doesn't seem to be in line with what should occur when the keel tears out of the hull. I'd expect to see some significant tearing along the hull where the keel tore free. There isn't any. Whether this is due to the design being too weak initially, or due to the design being sufficient, but having been damaged in the previous groundings, which could have sheered layers of laminate, not being properly repaired can't be determined from the photos I've seen.

I'd really like to see a photo of the underside of the Cynthia Woods taken from directly below.

 

[Franklin]

Yes. And, I said so in my post.

So, that's really all there was to it? Did anyone notice or feel anything before the keel came off? Was water seen coming in around the working keel bolts?

The safety officer who was down below (the guy who didn't make it out) noticed it was taking on water and informed the captain. He then woke up the other two who were down below with him. It was said they had about 30 seconds before it broke off but I think during that time they were trying to figure out where the water was coming from.

 

[Ross]

A transverse timber at each bolt would have spread the load of the keel across a large segment of the hull and would have prevented the tearout of such a small area of the hull.

 

[Maine Sail]

It sure.

It sure looks like just about every keel bolt was leaking or had leaked (lends credence to Franklin's theory). Multiple groundings and a bad band-aid job probably did her in. Someone obviously knew something was amiss due to the obvious attempt at minimizing or stopping the leaking via marine sealant? IMO that keel should have been totally dropped, and the laminations fully examined and repaired, before a re-set... Note the forward keel bolt it is about three threads short of the top of the nut...

Even if they did need to repair the keel stub, and or add more glass, the builder left them NO room to do so as the keel bolts are cut flush with the nuts???