Pride and Stability

[Roger Long]

I have decided to finally tell the story of my involvement with the Pride of Baltimore and the aftermath of her capsizing.

I first want to make one thing clear to avoid conclusion jumping that has sometimes resulted from statements I’ve made about the affair in the past: I take no issue with the design of the vessel. I think Thomas Gillmer fulfilled his design brief brilliantly. I am not aware of any error or omission that, if corrected, would have led to her hull shape, sail plan, or arrangement being the least bit different than it was on that day when her keel first touched the water.

This is not, therefore, a smoking gun story. It does however fill in a part of the story that is important to understanding the dynamics of how decisions were made about the operation of the vessel. I’ve told the story to three different authors of books about the ship but none seemed to be able to get a handle on it. Perhaps they felt it wasn’t important to understanding the events. You will now be able to make up your own mind about that.

I’m currently cruising around Narragansett Bay and there is spell of poor weather coming up so I will post this story in short chapters over the next few days. Please feel free with questions and comments although I may ignore them if answering them means jumping ahead of the story.

[gilberj]

I am all ears

[johnw]

Thanks, I look forward to it.

[varadero]

I have bought the breath bait and the popcorn!

[Roger Long]

The evaluation of a vessel’s stability and other characteristics effecting seaworthiness is an important part of the design process in order to insure that the client gets a vessel that meets their needs. The desired characteristics will vary depending on mission. There’s a big difference between designing a coastal cruiser for fair weather summer sailing and a vessel intended for a circumnavigation.

The comparison of vessel capability with mission requirements doesn’t end with the design. All boat owners do it when they look out at strong winds, high seas, and a poor weather forecast and make the decision to set out or remain in port. Institutionally owned or operated vessels do this more formally. Decisions such as sending a vessel on a far ranging trip around the North Atlantic are made by boards and committees. The captains of the vessel usually have a very significant role in this process.

Perhaps the most important lesson that has come out of all the studies and attention given to the subject of large sailing vessel stability in the last half century is the difficulty of evaluating the ability of a vessel to survive extreme wind events. Even the most experienced masters are put at a terrible disadvantage by the physics.

It is possible for a vessel to be so tender and “crank” that she will put her rail in the water and become virtually unmanageable before enough sail can be carried to drive her to windward. That same vessel however, could be virtually immune to being capsized by wind forces. Conversely, a vessel could be enormously fast and powerful sailing to windward and able to carry a great spread of canvas yet vulnerable to sudden capsize if allowed to heel only to a little over 30 degrees.

An experienced master with a good understanding of stability principles could probably discern the difference between these two hypothetical extremes but not the difference in ships of more normal proportions. Even an experienced sailing vessel designer needs the aid of calculations and measurement.

The question of surviving extreme wind events becomes especially difficult when changes are made to a vessel’s stability. For further reading on this subject, see: http://www.cruisingonstrider.us/Treacherous.htm. (You may want to start at the beginning of this web site: http://www.cruisingonstrider.us/Stability.htm.

The responsible operators of an institutional vessel will often have someone in their ship operations committee to provide advice on the vessel’s abilities to guide decision about routing and mission. This is often dispensed with in the case of vessel certificated by the U.S. Coast Guard or similar agency since stability analysis is almost always part of regulated operation. The Pride of Baltimore was operating as an unregulated vessel without the benefit of mandated stability analysis performed to established standards.

I served in the technical advisory role for many sail training organizations back when I was active in the field. This usually consisted of my simply conducting stability tests and getting the vessel through the process of obtaining USCG certification. I have had occasion more than once to tell operators of an uninspected vessel, “There is a high probability of your losing this vessel by capsizing or downflooding if you operate it long enough.”

Thomas Gillmer took on this technical advisory role by default as her mission was gradually extended from being a dockside attraction and experiment in “active archeology”, as either her designer or builder was once quoted. He was well qualified for the task having written a standard text book on stability for naval officers and others on the subject of naval architecture used by generations of naval and Coast Guard officer cadets.

Evan Wilson, in his book, Epitaph for a Beautiful Ship, http://www.lulu.com/shop/evan-wilson...-18177914.html explores the history of the vessel’s operation and makes the case for a defective management culture and decision making leading to the accident as the vessel took on ever more ambitious voyages. The single most important person in this decision making process would have been the technical consultant. The nature and quality of the advice and guidance he gave is an essential part of the story.

I was not privy to the nature of the advice Mr. Gillmer gave to the organization but I was in a position to evaluate the technical work behind it and have a window into the attitudes that shaped it. That part of the story is essential to a full understanding of the events leading the ship to 23 N 67 W and has never been fully told.

[Thad]

Thanks for the announcement and introduction, carry on.

[Roger Long]

The path to my involvement with the Pride of Baltimore began in 1973 when I was walking around the Boston waterfront in need of a job. I was a young man who wanted to be a yacht designer but my life was about to change.

Despite prior employment at Philip L. Rhodes in New York, I had just been politely shown the door at C. Raymond Hunt’s office. At their suggestion, I was headed for John W. Gilbert’s commercial vessel design firm on the next wharf where my drafting skills got me hired. It took about three days for me to realize that working boats with real jobs to do were far more interesting than toy boats, even if they didn’t have sails.

An energetic fellow, named Cory Cramer, went into the front office about a year later with a roll of plans and I heard him discussing a large steel schooner with my boss. Later, the chief engineer went in and I heard him say, “Roger likes sailboats. Why don’t we give him this project?”

The vessel was the Schooner Westward, recently purchased by the young Sea Education Association. As I did stability analysis and designed a new deckhouse for the vessel, I learned about the program and became enamored. Here was a working craft with sails and an important mission. Within a few months, I had worked out an arrangement with S.E.A. to go to work for them designing and planning a second ship to accommodate their expected growth. As a first step, I would leave Gilbert’s and be S.E.A.’s superintendent in the shipyard overseeing the work I had designed. I would then sail to Bermuda aboard Westward for my first deep sea experience.

I sailed to Bermuda that October, a trip in which we blew out a sail every night and I saw the ship completely underwater after driving through a rogue wave in the Gulf Stream. While I was at sea on the Westward, repairing sails for hours in the deckhouse I had designed, Cory Cramer was engaged in a struggle with his board. Strong willed founders of programs usually have problems as young organizations mature from one-man rule. The board was looking for an issue with which to slap his wrists about taking independent action and I was handy. I arrived back from Bermuda unemployed.

I decided to go back to my original plan, becoming a wooden boat designer and builder. During the few years I tried to make that work, I couldn’t shake the feeling that designing Westward’s replacement was my destiny. I designed a few yachts, built a couple of wooden boats, and spent a winter working as a consultant to S.E.A. in their offices, preparing the conceptual design for a new vessel although considerably larger than they eventually built.

I then spent a couple years working in the Port Office at the Woods Hole Oceanographic Institution where I developed what would become my primary professional interest, oceanographic research vessels. Jack Gilbert took me back into his office in 1980 and I worked there for a couple more years during which one of my projects was to a proposal to design S.E.A.’s next ship.

In the fall of 1982, I was invited to join two other Gilbert alumni as Vice President of Woodin and Marean in Boothbay Harbor. I had just learned that John W. Gilbert Associates and Woodin and Marean comprised the short list for designing S.E.A.’s next ship. I bundled up the cat, the litter box, my few other possessions, and moved to Maine. Two weeks after I arrived, Woodin and Marean were selected to design the vessel which would become the Corwith Cramer.



http://www.cruisingonstrider.us/Boats.htm#Cramer

[Canoeyawl]

Excellent, thank you for this.
This type of knowledge can take years to aquire (if ever).

Standing by...

[Paul Pless]

Thanks for this topic Roger. You've already hooked me. . .

[landlocked sailor]

I'm on board for the voyage. Rick

[Andrew Craig-Bennett]

This is jolly good; thank you Roger.

I went aboard the Pride when she was in the Pool of London and I remember being very impressed by two things - the enthusiasm and knowledge of her crew and her seeming authenticity.

[Roger Long]

The Sea Education Association recognized that it could not build a suitable ship to operate under the U.S. Coast Guard passenger vessel rules. These stability requirements restrict most vessels in ocean service to sailplans which are basically suitable only for putting on a show for the passengers while the vessel proceeds under power.

A Council of Educational Ship Owners (later renamed the Sailing School Vessel’s Council and folded into ASTA), was formed. The efforts of this group resulted in Congress requiring the Coast Guard to set up a joint USCG / Industry task force to develop appropriate regulations for sail training vessels. Woodin and Marean were selected to be the naval architects and technical consultants for the project.

We divided up the work. I would tackle stability since I knew how to program the new machine on the desk that wasn’t even known as a “PC” yet. Parker Marean, with his incredible capacity for attention to detail, would take on everything else.

The stability project began by sending out a worldwide request to every sail training organization and large sailing vessel operator for stability information. Packages of vessel plans and stability calculation books began to arrive in the mail and pile up on my drawing board.

As designers of traditional wooden vessels like the Rockland windjammer Heritage, and having conducted the USCG required stability tests for the many other vessels in the Maine fleet, Woodin and Marean already had a significant body of data on the stability characteristics of large sailing vessels. The data that was flooding in added considerably to this.

Fairly late in this process, a thin envelope with Gillmer’s return address arrived and we knew that we finally had the data on the Pride of Baltimore. We already had a good idea what to expect, based on comparison with vessels in our data base, and were eager to see how close we were. The ship was well known as a “hot” vessel so anticipation was high when we gathered around to open the envelope. We expected very accurate calculations in view of her designer’s background and authorship of a standard textbook on stability.

There was a shocked silence followed by laughter. Inside were just a couple sheets of paper showing righting arm curves with no other calculations or supporting data. The righting arm curves showed the kind of stability that could only be expected in a light modern yacht with high freeboard and a deep keel. I can’t remember Parker Marean’s exact words but they were along the lines of, I don’t know what they smoke down there in Annapolis but this is useless. We put the envelope aside along with any intention to include the ship in our analysis.

Months of analysis and meetings with the Coast Guard followed. During the meeting where we presented our preliminary findings, the Coast Guard asked, “Where is the Pride of Baltimore on these graphs?” I explained that we had been unable to get usable stability data from the designer and had therefore not included her. They asked that the ship be included in the database because they felt that she was not suitable for sail training and wanted to clearly understand how her stability would relate to the new regulations. We were asked to try again to obtain stability data.

[Hughman]

I'm listening...

[Ben Fuller]

Fascinating. I spent some time with the stats compiled by Garitee in the Republic's Private Navy when I did an exhibit down at Fells Point some years ago. Despite the literature about how unseaworthy the Baltimore clipper schooners were, I think only something like 5 out of the almost 200 were lost due to unknown causes ( presumably weather related). Of course they went to sea with crews of close to 10O. Somewhere I ran across a fascinating account of riding out a Bay of Biscay buster with nothing but a deeply reefed foresail. One of the things I think we forget is that these craft were handled by deeply experienced people. And the management of sail training programs forces the crews to be as small as they can. These masters run ships with fractions of the professional crews that they once had.

[Roger Long]

Of course they went to sea with crews of close to 10O.

And that is a very important and overlooked aspect of the whole story. Remember the question of whether the ship should cross the Atlantic with her topmasts up? I believe that Gillmer, to his credit, recommended that she carry them on deck. I don't know why she had them up at the time of the capsize but it perhaps had something to do with wanting to arrive able to set the full sail plan for photographs. Whatever the reason, the significant issue is how fast 100 highly trained and very tough seamen can get a set of topmasts down and lashed to the deck. They could also probably have all the guns off their mounts and swayed into the hold so fast our heads would spin.

These ships were not sailed in their time in remotely the manner that they are sailed today.

[gilberj]

Thanks Roger I am following closely.

[htom]

I'm hooked, too, Roger. Thank you.

[Capt Zatarra]

I too am reading this very carefully. Capt. Z.

[Breakaway]

Thanks Roger. Eagerly reading

Kevin

[gilberj]

Thanks Roger I am following closely.

[Gerarddm]

Dialed in too. Next?

[P.L.Lenihan]

Thank you for starting this Sir.I could eat stuff like this all day long,every day.Carry on at your leisure Mr.Long. Most fascinating!



Cheers!




Peter

[zauberberg]

Most interesting stuff, Roger. Thank you for posting.

[Roger Long]

We renewed our search for believable stability data on the vessel at the Coast Guard’s specific request. First, we purchased prints of Gillmer’s plans for the vessel from a company that made them available to model builders. Either a phone call or letter to Gillmer resulted in his sending additional stability calculations. This was cause for more amusement and consternation. Stability calculations for ships in operation begin with an inclining experiment, often called a “stability test”. This a complex operation that results in a thick report covering every aspect of the procedure and recording a complete audit trail of what is done. I had participated in many during my years at Gilbert’s, performed some on my own (including two on the Westward), and they were bread and butter work for Woodin and Marean. Gilmer’s report was just a few sheets of hand written notes recording almost nothing.

An inclining experiment is delicate as the effects of a single person walking around on a one hundred foot ship can be seen. The ship is heeled no more than a couple of degrees by moving weights around on deck. The amount of heel is measured with pendulums hung in buckets of oil to dampen the motion and the amount of movement is marked on sticks mounted above the buckets. Three pendulums are used and the deck weights are moved to three or more positions on each side of the ship. All of the results are then graphed in order to discard any erroneous results and the remainder of the data points averaged. Gilmer used just one pendulum and only moved the weights to a single position each side.

The objective of the inclining experiment is not to determine the vessel’s stability but to take its known stability characteristics, as calculated from the hull lines plan, and work backwards to find the center of gravity at the time of the experiment. It is vital therefore to know exactly where the vessel is floating so as to determine the exact shape and volume of the underwater hull at the time of the experiment. To do this, measurements are taken at frequent intervals from the rail to the water all along the hull and then plotted and averaged carefully to determine the exact flotation line as well as verifying that the hull has not sagged or that the ends were not built in a slightly different shape than as shown on the lines plan. The latter is very common as a vessel’s sheer is often adjusted on the stocks to suit the eye of the builder. Any error in determining the vessel’s actual displacement at the time of the inclining will be directly reflected as a proportionately equal error in all subsequent calculations to predict stability in various conditions so this is a vital part of the procedure. Gilmer’s report provided no data on where the vessel was floating, no drawing of the freeboard measurements, or even any indication that they had been taken. The results appeared to have been calculated using the displacement listed on the lines plan.

Nevertheless, we did have a single pendulum reading, which was better than nothing. I made a working assumption that it was correct and set about trying to figure out how he had gotten righting arm curves showing about twice as much stability at high angles of heel as our database of many vessels indicated was possible.

I digitized the hull data for entry into our computer. After entering the center of gravity from Gilmer’s inclining experiment, I ran righting arm curves and got almost exactly the curve we expected to see. It agreed very closely with Gilmer’s up to about 20 degrees. Thereafter however, the curve diverged dramatically from Gilmer’s. His showed about twice the overall stability (represented by the area under the curve) as ours. It wasn’t just an error in center of gravity calculation that made the Pride look like the safest ship in our data base.

The deck level was not clearly shown on the lines plan and I had to look at the construction plan to find the bulwark height and measure back down on the lines to find the deck. I then noticed a note in the stability data indicating that the hull characteristics had been calculated not by Gilmer but by an outside computer consultant. On a hunch, I re-digitized the hull as if the deck were at the level of the bulwark rail cap. This considerably increased the hull volume immersed when the vessel heeled, inside the computer anyway. When I re-ran the calculations, I got a close match with the righting arm curves that Gilmer had supplied.

It was now pretty clear what had happened. The subcontractor Gilmer had hired to digitize the lines plan into a computer had simply assumed that the deck was at the bulwark top as it would be if the ship had stanchion and wire lifelines. This wasn’t an unreasonable assumption since the actual deck line was not shown.

The amount of stability the vessel had up until it heeled enough for the deckedge to reach the water level was same on both righting arm curves. Beyond that, the real righting arm curve diverged rapidly from the one Gilmer had provided us. The properly digitized hull showed an ability to recover only from 92 degrees of heel where as Gilmer’s curve ran far out past the 100 degrees limit of his graph, indicating a vessel virtually impossible to capsize. A high performance vessel with a very large sail plan was out deepwater voyaging with bogus stability data. I wasted no time in reaching for the phone.

I thought of that phone call years later when one of my boys asked, “Dad, what does “condescension” mean?” A recording would have been worth 10,000 words in this case. The famous voice on the phone stuck me as a master of politely sliding home a velvet sheathed blade. He was, of course, perfectly aware of how the hull had been modeled in the computer. He could quite understand how a young and inexperienced naval architect could be upset about it. However, he went on to point out, the stability of a sailing vessel at angles of heel beyond the point the deckedge goes under water is highly theoretical and does not provide any useful information. If I wished to understand why, he suggested that I read the textbook he had written and which he was careful to point out was used at both the Navy and Coast Guard Academies.

I was on the cusp of maturity at this point in my life so I refrained from asking him why, if the righting arm curves at large angles didn’t mean anything, had he bothered to draw them? As the naval architect and principle investigator conducting the research for the joint industry / USCG task force considering modifications to regulations based on measuring the areas under vessel righting arm curves, I felt that my belief in the significance of the full curves was not naiveté. The analysis of the full curves was, in fact, the very basis of the regulatory method then in use. The methods we were using were described in an 1870’s English book on naval architecture. The Coast Guard naval architects who learned stability from Gilmer’s books evidently did not share his conviction, nor have I ever heard another naval architect express a similar opinion.

Here is the copy of Gillmer’s righting arm curves for the vessel which I sent back to him in a subsequent letter with our notations.



The dotted line shows the curve I got with our computer model modified to show the hull incorrectly extending up to the bulwark rail cap. The lowest curve shows the righting arm curve obtained using Gillmer’s reported center of gravity with the actual hull form. Even this curve was later found to be optimistic.

To this day, I do not know whether Gillmer was actually aware of the computer error before my phone call or whether his explanation was just the bluster it sounded like. I find a simple oversight of then new computer calculations more forgivable than knowingly providing false data to a survey of sailing vessel stability being conducted by a formal industry/USCG task force. If he had even a passing familiarity with the regulatory regime, or the history of the subject going back to the 1870's, he would have known the significance of the high angle righting arm values. Even if he had a different opinion about them, I think a note about the computer model would have then been appropriate when submitting the data.

[Roger Long]

I sent Gillmer a letter for the record describing our findings. The graph in the previous post is from that correspondence. He responded shortly after with a much more restrained letter repeating his assertion that the difference in the righting arm curves had no safety significance and suggesting again that I read his book. He did say that he would mention the matter to the board so that they and the operators would fully understand the stability characteristics. I presume that this briefing included his opinion that stability characteristics beyond twenty or thirty degrees of heel were irrelevant to assessing the safety of the vessel.

We then set about trying to determine for ourselves, with the little good data available, what the vessel’s true stability characteristics were likely to be. I prepared an estimate of the vessel’s center of gravity completely independent of Gillmer’s test. This didn’t involve as much guessing as you might think. Jack Gilbert built his career on the science of parametric analysis which enables the designer to predict characteristics like center of gravity location very early in the design process. He passed these methods and his strong belief in them on to a generation of naval architects who worked for him.

It is very expensive for the designer to finish the plans for a vessel, complete the incredibly time consuming task of listing the weight and location of every item in the vessel’s structure and equipment, perform the stability calculations, and then discover that the center of gravity is too high for the performance or regulatory certificate required. It’s then necessary to go back and change the hull shape which may require redrawing most of the plans for the boat. Back in the days of hand drawn plans, this would be a huge effort.

Survival in the not particularly lucrative business of designing USCG regulated vessels requires being able to predict the weight and center of gravity outcome of a design at a very early stage. Woodin and Marean had taken Gilbert’s methods even further and had an extensive database of information on traditional wooden vessels to which we added the data on the vessels recently sent to us. Treating the plans for the Pride just as if it were the preliminary design for a new vessel, we had the tools to predict her center of gravity with a high degree of accuracy. Woodin and Marean did this all the time with vessels like the Rockland schooner Heritage. I helped perform her stability test and she came up a little heavier and with a slightly higher center of gravity than expected. Parker called up the owner and asked, “Did you add something when you built the boat that you didn’t tell us about?”

“Oh, yes” the owner replied, “I put a band of steel plate around inside the frame to help her hold her shape over the years.” Parker took the dimensions, performed some calculations, and determined that the weight of this plate almost exactly matched the difference between the predicted and measured center of gravity. We were working within the very fine limits forced by the Coast Guard stability requirements and an inch or two difference in center of gravity location could be a very big and expensive deal.

Our independent prediction of Pride’s vertical center of gravity indicated that the height which would provide only about a 75 degree range of stability was right at the optimistic limit of probability. We continued to be uneasy about the accuracy of the data.

Sometime in the next year, I encountered Armin Elsaesser at an ASTA event and asked him about the Pride organization’s reaction to Gillmer’s disclosure of the hull digitizing error. I got a blank look and said, “We’ve got to find a quiet corner and have a talk.”

Later that summer, the Pride visited the Maine Marine Museum and Armin came to our office to further discuss the stability. I explained the importance of knowing exactly where the vessel was floating. He said that there was a cocktail party that evening. He would put me on the guest list and instruct his crew not to say anything if they saw me dropping a measuring tape over the side.

Everything I knew about the vessel’s stability was spread out on the drawing table in front of us as I walked Armin through it and the methods we were using in our analysis. This was when I first heard about her upcoming transatlantic voyage. The visit is etched in my memory because it was the last time I spoke to Armin.

That evening, I attended the cocktail reception with a tape measure and small notebook in my pocket. As others sipped and talked I quietly went around and slipped the tape measure over the side at several points and wrote down the measurements along with a head count of the people on board. The next day, I plotted the measurement as I would have in a stability test and then made a correction for the weight of the guests.

I then reworked the calculations of Gilmer’s inclining by incorporating my freeboard measurements, some additional stability data of Gilmer’s that Armin had supplied (I believe without knowledge of the Pride organization) and the estimates predicted by our independent methods of prediction. From this analysis, we determined that a 76 degree range of stability was the most probable figure and incorporated that in our analysis for the Coast Guard. Shortly after, the thick package went off into the long Coast Guard review process and I turned my attention to designing the Corwith Cramer.

During the design process for the Corwith Cramer, my tenure at Woodin and Marean was becoming increasingly uncomfortable because of disagreements about the financial management of the business. I wanted to see the design through to the end but I finally decided it was time to start my own company. As soon as the design was down to the detail stages, I left to open my Portland design office with an agreement to finish up the rigging drawings for the Corwith Cramer as a sub contractor.

[Ian McColgin]

That last sheds a very disturbing light on Gilmer, even taking it at it's best that he was blindsided by not having realized that the computer modeling was so wrong. He may not have even grasped the full import of Roger's call. I can understand Roger's reticence on this topic. Even so, I cannot understand any sailor not noticing what happens to such a boat as it goes past about 30 degrees no matter what the curve said. Regardless of the curve and of Gilmer's apparent dismissal, my memory of the time and at some great distance is that Gilmer and most sailors knew that it was a little late for speculation if Pride put her rail under.

Wonderful story Roger. This is an important aspect, disturbing as it seems at first glance. I am hoping that you have stabity data on Pride II so we can see what Gilmer did when the mission was described a bit differently.

[Roger Long]

Gilmer and most sailors knew that it was a little late for speculation if Pride put her rail under.

In a way, this is true of any vessel. Job One for the large sailing ship master is keeping the deck edge out of the water. At larger angles, people get hurt on board, things go to hell in the galley, steering is effected, and lines the crew may need to handle to regain control of the vessel will be underwater. If the master is successful in keeping the deck edge above water, stability at greater heel angles will be, as Gillmer said, irrelevant. The strong incentive to keep the vessel comfortable and manageable is why a truly horrible accident waiting to happen like the Marques could complete a circumnavigation with the stability that eventually sank her.

Mistakes happen and the forces of nature sometimes overpower even the best of ship handling. For the same reasons ship often have double bottoms, which they don't need if the master is successful in his near shore "Job One", sailing vessels need the ability to recover from unusual heel angles. The amount of this safety margin is what the whole science of large sailing vessel stability, of which Gillmer was so dismissive, has been about since the 1870's.

It's strange how things come back to you. While all this was going on, I remembered a conversation I had with John Atkin when I visited him during high school. We were discussing various boats and he said to me in a conspiratorial whisper, "Just between you and me, I think Tom Gillmer is a blowhard." I was shocked. Yacht designers were gods. Gillmer designed..... I was from a very proper private school where we wore coats and ties and said, "Mr." and "Sir", a lot. Did adults really talk about each other like that? I was young enough at the time of these events that it was a comfort to remember this assessment by one of Gillmer's peers.

Evan Wilson's book is largely about the management and culture of the Pride organization and its role in the loss. Management culture is heavily influenced by personality and it is this window into the mind of one of the most significant individuals in the organization's decision making that is the untold part of the story.

[Roger Long]

Just a few weeks after I had moved to Portland and opened up my new office, I was invited to give a talk to the Society of Professional Sailing Ship Masters on the topic of stability and the research I had carried out at Woodin and Marean. I was pointing to the graphs with symbols showing the stability characteristics for all the vessels we had studied. Different methods of displaying the data all showed the same pattern. The vessels which had capsized were all clustered far off by themselves down in the lower corner of the graphs. Up in the other corner was the larger cluster of “successful” vessels. Just out on the edge of this group, slightly off by herself and towards the capsized group, was the Pride of Baltimore. I said this indicated that, of all the vessels in the study, the Pride was the one most likely to suffer a stability related accident.

None of us in the room, in fact no one outside the liferaft drifting far away in the lonely ocean, yet knew that the ship had already capsized and four people, including Armin, were dead.

The shocking phone call came either the next or the following morning. I soon after received a call from the Pride’s senior captain, who had just returned from interviewing the survivors. He was about to meet with the board and was calling me for my thoughts on the stability issues before talking with them. He told me that I was the first person he was speaking to about the event since his arrival back in the states. My notes of this conversation, typed up immediately after I hung up the phone, end with this sentence in the last paragraph:

“….does not remember discussing my stability assessment of the vessel with Armin or seeing the stability numeral plot.”

I was quite struck by this statement while reviewing my files to write this account. I had just finished reading Evan Wilson’ book and found the fact that the senior captain was unaware of my conversations with Armin a disturbing affirmation of his assertions about the organization’s management. Armin was quite concerned when he left my office in Boothbay Harbor. I can only conclude that someone convinced him not to raise the issue.

The phone rang pretty constantly for the next few days. One call was from the Sea Education Association. They had been contacted by a reporter for the Boston Globe who had asked them to comment on the event and whether it was an accident that could happen to their vessel. S.E.A. had given the reporter my name as someone who could explain the stability issues and asked me to please try to clarify the differences between the two vessels when he called.

I was the target of a lot of criticism later for “running” to the media and this eventually had a direct effect on the inquiry into the accident. The media was sent to me however. Since stories were going to be published anyway, it seemed best to me that they be informed by accurate technical facts. The reporter called and an article appeared in the newspaper shortly after.

The media situation then descended into a nightmare. I received a call from a high official in Baltimore, the city treasurer I believe, who asked that I come down immediately to talk with them. A reservation and plane ticket were waiting for me. I flew to Baltimore and was taken directly to a room full of reporters with TV cameras. The official, I later learned, was a notorious loose cannon who had always been opposed to the Pride project. He introduced me by saying that I was a naval architect there to tell them how the city had killed four people. I realized instantly that I had been ambushed, mumbled my way through the event as best I could, and got out of town just as fast as I could find a cab.

I believe that this event was the primary reason for my later being excluded from the Coast Guard inquiry on the basis that my participation would be prejudicial. As the lead researcher on a joint industry / Coast Guard analysis of sailing vessel stability, probably the largest ever conducted at the time, my exclusion was a bit strange. However, as I have quipped many times over the years, nothing is more prejudicial to the proper outcome of such an inquiry than facts.

Soon after, I was asked to provide video taped depositions and written analysis on the accident to the British government for which I was continuing to work as an expert witness and consultant in their inquiry into the capsizing and loss of the sail training vessel Marques. The British Government along with many of Woodin and Marean’s other sail training industry clients had followed me to my new company, a point which created some bad feeling. Woodin and Marean were in frequent contact with Coast Guard headquarters and this may have also have been a factor in my exclusion from the inquiry.

[Roger Long]

My energies became consumed by getting a new company up and running as well as beginning the design of a 170 foot sail training vessel that unfortunately lost its funding just as I completed 270 feet of hand inked drawings.



I was busy enough that it was actually a relief to be excluded from the inquiry and I paid little attention to it until I received a call from a reporter.

The reporter wanted to ask me about a letter that had been offered into evidence, but not found relevant (according to him). He told me that he had been sitting behind someone who was about to testify and read the letter over their shoulder. It was a letter sent from Malaga by Armin to Pride of Baltimore, Inc.

In the letter, which concerned long range maintenance planning, Armin said (according to the reporter in this phone call) that he was still very concerned about the ship’s stability and suggested that I should conduct a new stability test when the ship was in Maine the following summer. The reporter said there was also a reference to this being politically expedient, being out of sight of Baltimore, as well as convenient. Did I have a comment?

I told the reporter that I had no knowledge of such a letter but that it was consistent with my last discussions Armin. I told him that I wasn’t following the inquiry but thought that a letter from the captain of a ship which had capsized expressing reservations about its stability was something I would consider relevant to any wreck inquiry. Evan Wilson’s book indicates that the letter might have eventually been accepted into evidence. I do not know as I never obtained a transcript of the proceeding.

The Coast Guard report was released in February of 1987. I noted with interest that their independent calculations and research indicated that the vessel had a vertical center of gravity in August 1984 of 9.75 feet and contained a righting arm showing a range of stability of 76 degrees at a center of 9.80 feet. This was very close to the Woodin and Marean estimates before the accident.

However, the report went on to state that the stability had subsequently been improved by lowering the center of gravity to 9.42 feet as well as lightening the vessel by 7.55 tons, giving her a range of stability just over 85 degrees. This 10 degree increase in the angle to which the vessel could heel and recover doesn’t sound like much but Gillmer is the only naval architect I have ever heard express the opinion that it is not of enormous importance.

Here is why it is important:

The blue line on this graph is Pride’s 1984 righting arm curve as traced from the USCG report.



The red line is a heeling arm curve which shows how the force of the wind drops off as the vessel heels. There would be a different curve for each sail plan variation and wind speed. The red curve here is for any combination of wind force and sail plan that would create 10 degrees of heel as indicated by where the two lines cross.

With the same sail plan, stronger winds will produce a higher heeling curve like this:



This wind heel curve just touches the righting arm curve at about 58 degrees which means that is a wind force which would heel the vessel to that angle. Beyond this point however, the curve is again above the righting arm curve. This means that just a small increase in wind force would be sufficient to capsize the vessel. Not only that, the ship would be committed to capsizing not at 76 degrees, as indicated by the range of stability, but at the point where the red heeling arm curve rises above the blue righting arm curve. In other words, if the wind increases to heel the vessel to 59 degrees, it will capsize with no further increase in wind force unless the pressure is immediately relieved by easing of sheets or changing course.

The difference in upright values of the two curves is proportionate to the difference in wind force. The upright value of this curve is three times higher than the first so, if Pride was being sailed at a 10 degree angle of heel, a three fold increase in wind pressure would capsize her. Wind pressure varies as the square of the speed so, if the first curve represented the ship sailing in a 16 mph wind with sails adjusted to create 10 degrees of heel, it would only take an increase in winds speed to just over 27 mph to capsize her if prompt action was not taken.

Now let’s look at Pride as Gilmer claimed to have improved her and as accepted by the Coast Guard to be her condition at the time of the capsizing.



The point at which a heeling arm curve can be entirely above the righting arm curve and the vessel capsize has been moved out to about 83 degrees, nearly as far as the positive stability range of 86 degrees. The wind force increase required to cause capsize is now nearly 17 times greater than that which would heel the vessel to a normal sailing angle of 10 degrees. This would be a 65 mph wind if rising from 16. This change in Pride’s center of gravity would have represented an enormous improvement in her stability. Even though she would still not have been able to lie flat in the water and recover, she would have been essentially uncapsizable in most foreseeable circumstances.

[Roger Long]

Did the improvement in the Pride of Baltimore’s stability accepted by the U.S. Coast Guard inquiry actually take place?

Gillmer is quoted from the inquiry in Wilson’s book describing how vessels tend to accumulate weight.

Q: Had there been any significant weights added during that time, say from 1977 to 1985?

A: (By Gillmer) Apparently there had, but this is – quite normal. It shifts as they grow older, like people, they gain weight. People bring things aboard and never take them off. I would say a large part of this was material that became worn and had to be removed like old anchor ropes and old sails and things of that sort. They can be fairly heavy.

Q: From your first inclining until, say, the inclining in March of ’85, was there much change in the height of the center of gravity?

A: Between August of ’84, before she went into the yard and March of ’85, the center of gravity was lowered by approximately eleven inches.

The Coast Guard determined in their report that the improvement in center of gravity location was actually a more believable 4 inches which corresponds to the curves above. An 11 inch change in center of gravity in a ship of this size is of the magnitude that can only be achieved by removing the entire rig.

Wilson talks in his book about an attitude of the Coast Guard officers conducting the inquiry which he describes as “incurious”. I think that term might apply to their failure to do a reality check on Gilmer’s reported improvement to the boat or even their scaled back estimate.

Here are her reported displacements and vertical center of gravity from the USCG report with the weights multiplied by the centers of gravity to produce the vertical moments.



It’s easy to subtract the 1985 weight from 1984 to see that the vessel was lightened by 7.55 tons. You can also subtract the moments to find difference of 110.76. If you divide this difference by the change in weight, you come up with the common center of the weight change, in this case 14.67 feet.

Pride’s deck was about 12 1/2 feet above the baseline used in these calculations. If the weight change had therefore been one single item removed and these figures were accurate, the item removed would have need to be two feet above the deck. If two items had been removed, for instance one item 3.76 tons that was six feet lower, down where all that un-needed gear might have accumulated, then a corresponding amount would have need to be removed from the rig 8 feet above the deck. Where would tons of material hiding above the deck and in the rig? No matter how you play with various scenarios, this simple calculation proves the conclusions of the Coast Guard report to be based on erroneous data.

What could account for the improvement found between Gillmer’s two stability tests? An inclining experiment is like a sextant sight. Delicate measurements are taken and then subjected to a complex series of calculations. There are many points at which errors can creep in which is why the procedures are so carefully spelled out in various USCG documents and naval architecture textbooks. Gillmer’s variance from these procedures itself should have been a subject of comment in any inquiry into a capsizing.

Let’s say you are on a vessel using celestial navigation and you see the navigator using a wind up pocket watch instead of a chronometer, taking a single snap shot, and then just scratching out the calculations on the corner of the chart. If two subsequent sights show the vessel to be making 40 knots, you don’t assume that it suddenly bypassed the laws of physics. No, it’s pretty clear that one, or both, of the sights are in error.

In this case, the Coast Guard, whose officers have quibbled with me about the most inconsequential of minutia while supervising numerous stability tests, could clearly see the slapdash way in which the stability tests on Pride were conducted. This did not prompt them to subject the results to the simple test above. As Wilson says in his book, there was a notable lack of curiosity in the inquiry.

So which set of results is correct? Could Pride have been the better vessel shown above and the first stability tests erroneously pessimistic? For that answer, you have to go back to the independent calculations performed by Woodin and Marean. The notes on Gillmer’s righting arm curves above are Parker’s and not mine because Parker and I both did an analysis independently and then compared results. We used data from dozens of vessels available for parametric comparison.

The only way Pride differed significantly from ships we had very definitive data on, having designed and conducted USCG supervised stability tests on some of them ourselves, was that she had a lot of tropical timber in her construction. These woods are generally heavier than domestic woods which means more weight in structure and less in ballast which decreases stability.

There no doubt in my mind that the 9.75 foot center of gravity and 76 degree range of stability represents the best stability that the ship could have achieved. She probably had slightly less. If her stability had been truly understood by her management and operators, I think there is a chance she would still be sailing today. I gave Armin a full briefing on her stability but he later talked with the designer who, to my knowledge, is the only naval architect to have ever stated that righting arms at large angles of heel are of no significance. The fact that another captain of the vessel had no knowledge of my discussions with Armin and had not seen the material I gave him is a disturbing management issue that Wilson discusses in his book.

[Roger Long]

A couple years after the Coast Guard inquiry, two senior captains of the “Pride” and I found a quiet bar corner at an ASTA conference and put everything we knew about the capsizing on the table along with the beers. We traced the event practically minute by minute and I doubt that a more informed or objective analysis of the tragedy has ever been undertaken. This discussion has caused me to say often over the years that it was as much a control failure as a stability deficiency. However, the stability characteristics precluded recovery.

The “Pride” was certainly not an inevitable accident waiting to happen as were the other capsizes I have studied. I think the “Pride” could well have been still sailing today. The probability of that outcome would have been significantly increased by a better understanding of her actual stability characteristics and how it affected her mission planning and operation.

The Coast Guard inquiry left the Pride organization comfortable enough about Gillmer’s handling of the stability issues on the first vessel that he was selected to design the replacement. A press release stated that the new ship would meet the strictest Coast Guard requirements, including carrying paying passengers on ocean routes. I remember thinking, This I’ve got to see. The whole sailing school vessel regulatory effort had been based on the impossibility of getting sufficient sail on ocean route passenger vessel for all except a very narrow range of types. Baltimore Clippers were certainly not in this group and the first sail plans I saw released had me wondering how much all the carbon fiber necessary to make it work was going to cost.

I became a board member of the American Sail Training Association and was the founding chairman of their technical committee. My company was selected to design a full rigged ship billed as “America’s Tall Ship” and to be named Discovery. A few months after the plans for the Pride of Baltimore II were announced, I was sitting with the executive director of the “Discovery” organization in the outer office of the Coast Guard officer in charge of the group that reviews and approves vessel plans. We were waiting to discuss our new project with him when another USCG officer went into his office through a side door.

I could hear him say, “We’ve got these plans for the new Pride of Baltimore here and there are no stability calculations. The designer says he wants us to do them. We never do that, do we?”

The officer at the desk looked out and saw us sitting in the outer office and held up his hand. The other officer closed the door. I’m sure the quiet murmurs we heard after that concerned the fact that submittal of stability calculations is always required. The Coast Guard prepares their own, but to insure that the calculations are independently done twice. I don’t know if Gillmer ever submitted his own calculations but the ship never received certification remotely close to what was promised at the beginning of the project.

I have no basis for reservations about the stability of the new ship and would be glad to sail on her myself. However, the way in which everything about the stability of these vessels was handled will always bother me.

Not long after this meeting, I received a subpoena from the IRS and the Discovery tall ship project quickly folded up into trial and punishment for people who had been my friends. The unpaid bills forced the closing of my design office. I turned my back on sail and entered the oceanographic research vessel design phase of my career which has been very productive and satisfying.

The Pride has kept resurfacing in my life as books have been published. Many years later, I saw the designer’s beautiful book about the ship at a friend’s house. Leafing through it, I read:

However, another and most troublesome attack on the Pride's credibility was deliberately introduced through the news media by an employee of a ship design firm in New England. This report charged that in a study prepared for the Coast Guard for certifying the requisite stability of sail training vessels according to certain criteria, the criteria of the Pride of Baltimore showed her to be the least stable all sample sail training vessel selected for comparison, with a stability range of only 76 degrees. This was purely a contrived criticism, factually and technically incorrect, put forward for inexplicable reasons. The study was wisely marked by the federal investigators as not to be admitted into evidence. Its use as a reference of comparison was prejudicial.

I remember thinking it was merciful that an old man wasn’t carrying guilt to his grave but that the story needed to be told someday. A mutual friend later told me, “Tom wakes up in the night hearing screams.” This conversation was largely the reason I waited so long to tell this story.

I have told this story to the authors of three books on the sinking but none really grappled with it. The pieces of the story are there spread out through three different books and some magazine articles but it would take more time and knowledge than most readers posses to assemble them. I hope this thread will clarify some technical aspects of the tragedy and how the handling of the technical issues may have affected decision about her operation. The failure of the management and operators of the ill fated vessel to better understand her stability characteristics will always be central to the story, regardless of what conclusions may be drawn about its contribution to the loss of a ship and four lives.

[Gerarddm]

Wow.

[Andrew Craig-Bennett]

A very interesting account.

[tigerregis]

Thank you, Sir. Very enlightening through its clarity.

[zauberberg]

Not only in banana republic: There is on the one hand the formal world of laws, regulations, organisation, processes, and then there is on the other hand the world of human beeings, acting in this formal world, connected to each others by often invisible bands of interests, personal goals, dependencies. Individual interests are bundling up, often silently, and togther drive a process into a specific direction, sometimes along the edge of legality, sometimes clearly over the limits. And as you remarked it's the personality / integrity of the key personal (together with the cultural background of all players) and the example they give by living and acting that makes all the difference. Designing the formal world in a way so that it becomes independent from it's players is impossible

The question of course now goes to each of us: How do we act, what are the standards that we apply to ourselves, how do we resist the temptation to make use of the always existing "opportunities" granted by circumstancies and connections. The world seems to be full of examples how people, initially maybe of good intent, are gradually drawn into situtations that turn out to be just too demanding.

[Paul Pless]

seems almost criminally negligent

[Uncle Duke]

Thank you very, very much, Roger, for the clean and clear explanations. Absolutely fascinating.

[Thad]

Is that all Roger? Is more known about her handling before the capsize?

[Roger Long]

Is that all Roger? Is more known about her handling before the capsize?

Isn't that enough?

Since there were survivors, the handling and events just before the capsize are very well understood and covered in the three books on the subject.

[Thad]

Not a complaint, your history of the PoB stability issues is much appreciated. Might have to read the books. Thanks.

[Paul Pless]

Isn't that enough?

LOL

Can we have a question and answer session at this point?

I'd like to know if Gilmer's 'attitude' towards the stability of Pride, extended to any of his other designs?

[Roger Long]

Can we have a question and answer session at this point?

That's one of the reasons I decided to release this on a forum like this.

I'd like to know if Gilmer's 'attitude' towards the stability of Pride, extended to any of his other designs?

I have no way of knowing but I've looked at his designs for years and all of them, including the Pride of Baltimore appear to be fine craft for their purpose. He left a great legacy as a yacht designer. In the case of the Pride, the mission of the vessel changed and he was part of that process. The vessel's stability was actually superior in some significant ways to some of the vessels operating in the Maine windjammer fleet. However, they mostly operate in the same conditions and waters that people sail around in open daysailers.

[Peerie Maa]

The vessel's stability was actually superior in some significant ways to some of the vessels operating in the Maine windjammer fleet. However, they mostly operate in the same conditions and waters that people sail around in open daysailers.

This is a key issue. I have read, but cannot remember where, of an American fishing schooner that crossed to seek fish off the Irish coast. She took such a pasting that she was forced to retire home hurt. This image of the Irish coast indicates the power of the sea after building up energy in the Atlantic fetch.

[johngsandusky]

A compelling article, thank you.
I didn't know Gillmer, or any other principle in the Pride's construction. But I will suggest that it was more blind optimism and enthusiasm than criminal negligence. I have occasionally been too optimistic about my position (on the water), or the capability of myself or my vessel. This has led to minor damage or embarassment, no worse. It's how accidents happen, like tailgating or cutting a corner while driving. No room for error. We all should try to learn from this and every sailing accident.
It is ironic that her name was "Pride".

[Ian McColgin]

Roger, what a clear and helpful explanation.

As applied to small boats, it might help folk see why a boat like the H-28 or the Rozonante are fine boats despite the lack of self-bailing cockpit. Or why a boat like Marmalade is a fine boat despite the fact that even if the cockpit were more than pretend self-bailing it's still so large that a fill-up from a pooping sea would sink her just as surely as her lack of return stability from a 70 degree knockdown (that's a guess and it could be less) would sink her. Or the wonderfully seaworthy Friendship sloop. All great boats for their purposes, which do not include a North Atlantic winter crossing or life at high southern latitudes.

Thank you again.

[rbgarr]

Thank you for the effort you put into making all that clear to the layman, especially the description of how stability tests are done.

My memory is probably faulty but I thought that when the Pride I project was first proposed the vessel was intended to be an 'ambassador' for the city, which implied more than mere display or harbor tours and perhaps more than coastal passages to other East Coast ports. Am I mistaken?

A year or so before you became involved with SEA and Westward, I sailed transatlantic aboard her. I recall the captain being very mindful of the levels of the water tanks as they related to her stability. If there's a relatively simple answer, do you recall how much the fresh water tanks contributed to her stability and what water levels may have created concern in that area?

Enjoy Narragansett Bay!

[Roger Long]

Am I mistaken?

I don't know. I would refer you to Tom Waldron or Evan Wilson's books for the early history. I just know the she was lost on a voyage considerably more ambitious than anything originally contemplated.

I recall the captain )of the Westwardd being very mindful of the levels of the water tanks as they related to her stability.

One of the first things I did working for S.E.A. was supervise (and smash my finger doing it) the installation of 12 long tons of lead ballast indicated by the stability analysis I worked on at John Gilbert's. This completed the progressive addition of ballast over the years finally bringing here nearly 18" below here designed lines. Some of this final addition was because of the added weight of the new deckhouse but she was also always tender. Her stability on her first circumnavigation was probably worse than the Marques, this discovered by the shipyard after her departure. There is a frantic letter from the builders that chased the owner around the world recommending that she be immediately re-ballasted. The haste was evident in the translation from German. I just remember the last line, "After that she don't come up any more."

I thought for years that they put too much steel in her but ran across the designer's original calculations. A critical number, the baseline for almost everything else, was off by a foot. If Eldredge-McInnis had detected the error before finishing the plans, she certainly would have been 1 - 2 feet wider.

[Roger Long]

As applied to small boats...

Yes, I'm sure the worst wind and sea conditions I have ever experienced were in an 18 foot Marshall Cat, probably the second and third worst as well. Of course, I was young and stupid but it is amazing what vessels can survive.

I did end up sitting on the bottom of that boat and then her transom as all the ballast fell into the forepeak after being flipped by a rogue wave on the shoal by The Cuckholds near Boothbay Harbor.

[Gerarddm]

Not having read the books, were there ever any negligence lawsuits?

[gilberj]

Thank-you Roger for your detailed description. Mr. Gilmer's attitude at the revelation of apparent shortcomings of his work is sadly very human. I have seen it often before, and perhaps been guilty of it once or twice, though without the magnetude of the consequences. We can only learn. Marine safety is a goal, not a destination. The loss of the Concordia off Brazil is an echo from the past.
I have seen a number of boats which have made offshore passages and are thus credited as 'bluewater boats' that had significant flaws, which the operators had clearly not realized or ignored. I spoke to a man last year who had put a large deck house on an an older wooden west coast trawler and taken his family to Mexico and back. When I stepped on the rail there was a significant dip as the 25 tonnes of boat adjusted to the additional 180 odd pounds. The boat was a good seawothy boat in its original form, but substantial modifications made it a completely different vessel This man had been a commercial fisherman and one might think his experience would help him. Apparently not. He was lucky though and brought his family and boat home without incident. The trip and the boat obviously a complete success in his view. I tried to speak to him regarding my concerns and was brushed off, much as Roger apparently was.