The Offshore Voyaging Reference Site

How To Do Amateur Engineering¹

From time to time we get a question in the comments, or via email, asking us to opine on whether a planned modification to a boat or rig will be strong enough.

Let’s take a look at a typical example that just came in:


Member Denny asked:

I am building a hard dodger for an O’Day 35, I want to move the traveler from the deck forward of the companionway to the aft end of the dodger–out of the way.

The main arch at the very rear which will be the load bearing part, has a base (entire top ) of 1/2″ cell modulus (NICORE) with one layer of fiberglass matting on the bottom and three on top, laminated. In addition, across the top inside will be two beams two feet apart and 65″ wide made of six 2″ strips of 1/4″ plywood laminated into a curve with an apogee of 6″.

On the sides in the area of the traveler there will be a 1/2″ aluminum plate bolted to the side just above the deck with a straight piece of the laminated plywood connecting to the top beam with bolts, a 10″ x 10″ gusset and a aluminum “T” bar where the “T” is 1 full inch as is the cross part.

This will be screwed to the wood piece and then the outer part. will accept the side structure on top of the aluminum plate and inserted into the notch in the aluminum upright that is connected to the wood upright.

The sides will be a layer of fiberglass with under a 1/2 ” piece of Nicor and two or more layers of fiberglass inside. I’m going to make a deck casting out of fiberglass to mount the sides in and the whole thing will be smoothed and faired with more fiberglass, resin, etc. the casting will ne bolted to the deck.

I don’t need to go further since the question is: Do you think this construction will be sufficient to handle the load of the traveler and its sheet?


Login to continue reading (scroll down)

Inline Feedbacks
View all comments
Colin Speedie

Hi John

I agree with every word you say. I see the damage caused by ‘DIY’ modifications and repairs all the time, often having caused serious and expensive damage way beyond the initial problem.

Getting objective, professional advice for significant repairs or upgrades is vital and can rarely be found on YouTube. Someone experienced needs to inspect and appraise the ideas you have – that’s 50% of the deal towards a successful outcome.

As Lou and I look around for the right old boat to refit and see our days out, I stay vigilant for home repairs in every boat that I see.

Rob Gill

Maybe also think about weight, especially how high that weight will be positioned, and how close to the stern it will be. So how is the yacht stability and trim before the start?

On a 10 metre yacht this build can be more tricky than say a 15 metre yacht, since people who need to fit under the dodger are basically the same size. A nice looking, hand-laid dodger, built strongly in fibreglass and epoxy could weigh 80 kgs. That’s an adult crew member standing on deck doing nothing all day – doesn’t even bother to change sides when we tack.

But then there is the aesthetic, and on a pretty yacht like the O35, it is easy to affect the look (and so resale) with a hard angled dodger – we’ve all seen them. But even slight curves are hard to achieve and add heaps of weight. A friendly yacht designer can help with stability questions AND looks, but also sight lines for visibility from the helm, from the cockpit seating and from under the dodger, for the tallest and the shortest crew members. Could be worth the investment in their time.

Then perhaps also think about unintended consequences – there must be a reason O’day put the traveller forward. It keeps the mainsheet, and the mainsheet car, clear of heads and hands for families that will sail her. Place the mainsheet and traveller at the back of the dodger, just where people love to lean to lookout, and you could have an accident waiting to happen?

Lastly, on a cruising boat, having the leech open up as a wild gust hits thanks to the forward sheeting position, is a good thing surely? Race boats not so.

Drew Frye

Well said. I could add that engineers often follow steps 1 through 8 even when they understand the calculations, because knowing the worst case forces in the ocean, with any certainty, is often quite impossible. So you scale from things that have worked.

I typically take years before making what later seem like obvious upgrades to boats I have owned, not because I am afraid of drilling holes and laying glass, but because it takes me that long to understand the design, to learn what it needs, and to truly believe that I can improve upon a boat built by a smart team of engineers and craftsmen. Just adding stuff doesn’t always make it better. As John says, when in doubt, leave it off.

And the disappointing part of a job well done is often a result that is nearly invisible, that looks factory. A good thing, overall.

Dan Manchester

It’s not always about about calculations as a professional engineer, especially on an existing ‘thing’ that you want to modify, as you usually don’t know the full details of what you have already. It’s often more about being able to see load paths, understand material behaviours, and see where failure points are going to manifest. I’ve reviewed plenty of designs from young engineers that I knew weren’t going to work despite them having done the numbers. Without a full FEA model, it’s almost impossible to completely understand how a multi-material complex design subjected to multiple dynamic forces is going to behave.

George L

Agree in general.

just to add, the proposed change is massive, so the overbuild, test, etc option isn’t viable.

moving the main sheet forward will likely require a stronger boom as well ..

in home construction no builder or architect would touch something like this without having the stamp of a structural engineer, so why would anyone do that on a vital part of a boat.

for a very good trail on the Asteria, see

George L

I stand corrected, I thought he was moving it forward.

Seen in their own right, of course his ideas are compelling. The problems and expenses come if (or perhaps when) things don’t quite work as expected and when upon resale all the money spent is a write-off.

Indeed, Peter is clear in his text that this is his thesis, subject to analysis, verification, etc. So he would agree with you that there are other possibilities.

Concerning the “ax”: Peter was concerned with the modifications of the boat and the way the vane has been mounted. Nothing what he writes has to do with his general criticism on DM’s dealings with wind vanes (which I believe is spot on). Obviously Peter thinks his vanes are superior to the competition, he spent 50 years optimizing every bit of it, both, in terms of materials and design and if he were aware of a potential improvement he would implement it. Only because the messenger has a vested interest doesn’t diminish the validity of the message, especially since quite often he is the lone voice who is willing to talk about the elephant in the room.

Eric Klem

Since the person asking the question mentioned the boat model which has a marconi main, this doesn’t apply but is a great example of the non-intuitive things you find in engineering.

For a main with a well distributed load along the foot such as most gaff sails, end boom sheeting is actually not the lowest stress for the boom. Yes, it is the lowest stress for the mainsheet and attachments. However, the boom’s ideal attachment point for the mainsheet is something like 3/4 of the way from the tack to the clew. The reason is that it shortens the span distance and which is extremely important to stress in beams. The result is that a correctly placed sheet attachment allows for a lighter boom construction but your boom bail, sheet, traveler, etc. will get heavier.

The reason that this doesn’t apply to most modern rigs is that the foot of the sail does not distribute loads evenly on the boom at all and the tack and clew take the vast majority of the load. The most extreme case is a loose footed sail and then ideally you have the mainsheet attach right where the clew does leaving the boom only in compression. Of course it is really hard to actually line these up in practice and most boats have a boom vang so you rarely get the nice compression only use case and if you did, we would all have round cross section booms.


Eric Klem

Hi John,

Your 2 options make a lot of sense to me even though they might not be what people hope for.  I assume it is obvious but if people do search out an engineer, they need to find one in the right field as it is a pretty specialized discipline.

One of the things that I find hardest when doing stuff on boats is actually figuring out what realistic loads are. For example, if I want to install a cleat for a dockline, I need to know what a reasonable worst case load is on that line which is not easy to look up or figure out. So even for mechanical engineers, if they don’t work directly on this in their day job and know where/how to get the load, you end up having to just be really conservative and do stuff like design around the breaking strength of the dock line.

Whenever the strength of a design is not obvious to me, I find it very helpful to draw the part/assembly and then draw the forces on it. Just having it drawn can be really helpful, at least to me.

The thing that I find myself cautioning my non-engineer friends against the most when they diy is making abrupt geometry changes in their parts. Any geometry change like a corner, step in a shaft, etc concentrates stress around it. You can never fully eliminate this concentration but you can lower it a lot by having nice smooth, large radius transitions with material concentrated in that area. Of course parts do have to have ends and such but if these can be kept outside of the load path, they are usually not a big concern.


Eric Klem

Hi John,

I am in agreement on thinking through failure modes but I think in practice, trying to design in a single fuse doing amateur engineering is probably a bit optimistic in a lot of cases. Mechanical fusing is actually remarkably tricky to do well without significantly overbuilding everything but the fuse, especially when amateur levels of fabrication are involved. Maybe it is more realistic to say that your amateur engineering project should attach to something that is way stronger than it so that you can only break what you added?

One place I have run into the topic of fusing a lot is in regards to rudders. I have seen lots of proposals to do things like run a router around the rudder half way down with the hope to create a breakaway section for big impacts. But to actually fuse a rudder, you really need to know exactly what the construction is and then run a bunch of numbers. It would be quite easy to either cut the groove in a place that wouldn’t actually serve as a fuse or to do it in a way that it broke off under an undesirably low load which is also bad. I can’t see how you could do this correctly without first destructively taking things apart and carefully measuring then running a bit of analysis, probably with expensive software like Ansys. Just like I don’t like belt and suspenders design in many cases, fusing should be done properly or not at all in most cases.

In my line of work, we generally try to design most things not to ever fail due to stress. If your loads and number of cycles are known with some degree of certainty, this is possible. We do still look hard at single fault analysis such as an overload and make sure we understand and are okay with the way in which it fails. When loads are not well known or you have problematic failure modes like wear, we then look at mechanical fusing. Since I work on expensive capital equipment, in the rare cases where we do fuse, we try to make it non-catastrophic meaning that the system will keep operating but that takes even more time to do correctly.

Your lazyjack example is a good one. Using normally sized components, even with say 1/4″ polyester double braid at ~2500lb for the lazyjack lines, there is not a natural fuse in there as that is still way too much load to put on the spreader. That leaves you with adding a new piece in series that is a real fuse with a breaking strength of a few hundred pounds at most if you go the fusing route. This may even have usability issues as you may break the fuse with regularity in normal use. As you point out in the article, the right solution is to move the blocks back to the mast where it is so strong that it kind of doesn’t matter and you don’t have to carefully design a single fuse as you know the mast section is stronger than many of the components attached to it and something there will go first.

One final fusing thought. There are companies that specialize in them and whenever you can buy a fuse, that is best. There are obviously all types of electrical fuses including ones that automatically reset but there are also all sorts of mechanical fuses such as shear couplings, slip couplings, etc. The companies that make these have done the design, analysis and testing on them and you don’t need to re-invent them, especially as an amateur.

Sorry for the long post on a slight tangent.


Dick Stevenson

Hi Eric, No apologies necessary. Quite the contrary: it was interesting and helpful in thinking more clearly and accurately about a wide angle of challenges.
Thanks, Dick Stevenson, s/v Alchemy

Brian Russell

Let us not forget the value of the good ol’ reference book. Dave Gerr’s “Boat Strength” is an excellent reference on how the various parts of boats, in various materials, are designed and built. It is a great addition to any sailor’s library.

Dick Stevenson

Hi Brian,
Yes, agree completely about Dave Gerr’s “Boat Strength”. In fact, there is not a Gerr book I would not recommend. He is not only a Naval Architect, designer, educator, writer and journalist but has the distinction of being all those things and enjoying the ability to communicate effectively with the average boater: a very rare combination. He is also very very funny in some of his writing.
A Dave Gerr anecdote:
I knew Dave slightly through mutual friends and he also did a very helpful prop/drive-train consult decades ago on an earlier boat.
From my memory and boatyard scuttlebutt, perhaps not accurate in places: but, I hope,  enjoyable nonetheless:
Dave used to write an annual April Fool’s article for Ocean Navigator magazine. One April, ON published an article of Gerr’s where he wrote a report citing the research of a professor (Dr. Loof Lirpa) at the Univ. of Mich. Essentially, Dr Lirpa, had done many lab tests and field testing to document that drilling holes on the leading edge of your propellor blade increased efficiency, saved fuel and increased speed. (This may seem far-fetched, but some may remember when there was work to get small air bubbles released along the hull to improve slipperiness: similar idea.)
I came to my boatyard early one morning to find a go-fast powerboat in the slings and the boatyard owner and the skipper screaming at each other: “I will not drill holes in the leading edge of your prop before launching: no way. And, who the hell is this Gerr fellow? I came in this morning to find 16 feet of fax paper on the floor asking questions and ordering holes drilled in the boats here!”
I believe that there were a few boats launched that spring with holes on the leading edge of the prop.
I also believe it was the last April Fool’s article from the pen of Dave Gerr.
My best to all, Dick Stevenson, s/v Alchemy
Ps. Spell the Drs. name backwards.

Michael Fournier

I agree with everything you said except one addition. A engineering degree is just knowledge gained from schooling. And anyone with a brain can study the same books the engineer has. Anyone can get access to the same data and if they paid attention in math class can also follow the engineering and come up with the same data as a engineer. And many of the modifications people want to do will entail engineering someone else has already done on a similar boat. And even then boats designed by well established navel architects have had design flaws. Many of the older fiberglass boats were over built simply because the engineering of the material was still being developed at the time. Early fiberglass boats were designed by wooden boat designers and many had a structure of a wooden boat and the glass hull simply replaced the planking. So my point is instead of asking someone on a forum people need to learn about boat building study the designs of known navel designers. (Learn about engineering by actually reading the knowledge of the world is in books yes’m I know all that math and formulas can be boring to most but there ARE many people out there with brains capable of understanding the math and the strength data and engineering principles require to focus on a narrow application of that knowledge. Just because someone has no degree does not mean they can’t learn or apply expertise in another field of design toward boat design. I build houses my background was a trade school education then military helicopter crew chief and then later art school education. BUT I studied design and engineering beyond that on my own. So I feel comfortable enough to apply other knowledge to a problem. And your right COPY is always a great idea but sometimes the armature engineer fails to see the pass on effects of a design change. But many get away with changes because many older boats were over built so applied forces to a area not really designed for it works out because of fools luck. And the fact the boat was over built at a time even the designers were not completely sure just how strong was strong enough so chose to air on the side of safety. On later race boats they are on the edge of engineering and often lighten until failure as speed is priority and use risk analysis. But also consider that a boats in a race USUALLY are being monitored and it may be at most a few hours before someone notices the boat is lost or behind scheduled based on their previously known position. In other word someone is likely coming along to rescue the racers if they can hold the boat together a few hours or so. On a cruising boat it could be days or more if at all before anyone comes along to rescue you so OVER engineering is always a good thing when your life’s is on the line but not all modifications if they fail will result in a dangerous situation. It could simply mean the solar arch starts to wobble a bit at 40 knot winds. And may need some added security in a blow. Risk assessment think what could happen if this fails and how catastrophic will that failure be. If it’s a bad idea this type of risk assessment will usually dissuade any Amateur engineer from ether. Attempting modification or leaving it as it is.