Parma

carbon/carbon rudder post failures?

Recommended Posts

I was under the impression that carbon rudder posts were as strong as stainless steel, but I guess not? Do steel posts fail as often or are the carbon rudder posts that are failing simply created with an unseen or unknown flaw ? Simply under engineered? Is carbon as a building material still undergoing research/understanding and "field testing"?

 

What's the truth?

 

 

post-15858-0-86457900-1492471085.jpg

Share this post


Link to post
Share on other sites

What are we looking at there??? A bungled mess designed and built by exactly whom and for what??

Share this post


Link to post
Share on other sites

I know right?

I'm seeing three different materials there making up the rudder post

Share this post


Link to post
Share on other sites

Well designed and built carbon post should be more durable than metal. Was post in OP some kind of backyard cockup?

Share this post


Link to post
Share on other sites

What are we looking at there??? A bungled mess designed and built by exactly whom and for what??

A really small picture? I'm guessing it's the bottom bearing surface. Looks like it has a nasty crimp in the side. Crushed bearing surface and shaft failed between bearings? WAG: something horrific happened here... Or not. Kinda hard to tell.

Share this post


Link to post
Share on other sites

That's the weirdest rudder post construction I've ever seen. Who exactly designed and then built it?

What bit is weird? I can't make out any details of the layup from the picture. Writ large, it looks like a composite shaft - insulating layer - bearing surface. That's not unusual.

Share this post


Link to post
Share on other sites

J-111 Rudder, Hull #3, 2010. Broke hard reaching in breeze.

 

J-111 Rudder, Hull #4, 2011. Broke hard reaching in breeze.

 

Stump didn't look as messed up as that pic though.

Share this post


Link to post
Share on other sites

Did not want to mention the boat type/mfgr that raised this issue for me because I'm really only interested in the reliability of CF rudder posts (since I have one)

 

Rumor has it it is a carbon fiber post with a foam core but for the life of me I can't imagine anybody putting a foam core into a carbon fiber post since that could lead to all kinds of separation and compression ( as somebody noted there is a kink in it) issues.

 

How do we know CF is stronger than stainless? Failure rates or laboratory testing? I have a sneaking suspicion that the type/kind of carbon construction layup/laminate will determine the result

Share this post


Link to post
Share on other sites

I always thought this video was a pretty good representation of the relative strengths of carbon vs steel.

 

Share this post


Link to post
Share on other sites

That video tests in torsion though, and that rudder shaft failed in shear.

Share this post


Link to post
Share on other sites

It's really just an engineering and fabrication issue. You can engineer them both with the same breaking load. But carbon is likely a bit trickier to fabricate.

Share this post


Link to post
Share on other sites

Well, I have a 19 year old carbon shaft and rudder that has been trouble free and weighs only 17 pounds. Stiffer, way lighter and stronger than any stainless rudder I've ever seen.

 

Of course, it was laid up by Cookson. So there is that.

Share this post


Link to post
Share on other sites

 

J-111 Rudder, Hull #3, 2010. Broke hard reaching in breeze.

J-111 Rudder, Hull #4, 2011. Broke hard reaching in breeze.

 

Stump didn't look as messed up as that pic though.

 

J/111 Hull #5, Broke hard running in breeze. Not kidding.

Share this post


Link to post
Share on other sites

That's where you go back and go over the calculations. Check the safety factor. Check the speed used. Carbon is so light for its strength I have always favored going generous with the laminate.

 

I worked with one builder and he objected to my CF chainplate specs. He thought they were too light. He showed me some chainplates on one of his 70'ers. I asked him if he built the to the design spec. He said no, he doubled the spec.

Share this post


Link to post
Share on other sites

Well, if all the early J/111 rudders were built as the photo above suggests, it was not a problem of the amount of carbon so much as the fact that it looks like a 5th grade "Laminated" them by putting all the plies in a pile and then pouring resin on the top.

 

So, to the insurance guy in the crowd, there's your answer. A trend!

Share this post


Link to post
Share on other sites

A carbon or stainless post is as strong as you design and build it to be. Properly laid up carbon is stronger pound for pound and square inch for square inch compared to stainless steel.

 

One advantage of stainless steel is that it is made in a factory and of relatively uniform and known quality, compared to CF laminate which is shop made, and can be badly bungled. Another advantage is the material is ductile and tends to fail by bending, rather than brittle like CF which fails by shattering. Still, I'd take a properly built CF rudder post any day.

Share this post


Link to post
Share on other sites

Except one thing...

 

Severity of failure mode.

 

When it does fail, with a SS rudder post it bends and you still have a usable (albeit bent) rudder. With CF it will fracture and you are left with NO RUDDER.

 

One is a pain in the ass. the other a life threatening situation.

Share this post


Link to post
Share on other sites

Except one thing...

 

Severity of failure mode.

 

When it does fail, with a SS rudder post it bends and you still have a usable (albeit bent) rudder. With CF it will fracture and you are left with NO RUDDER.

 

One is a pain in the ass. the other a life threatening situation.

 

A SS rudder post that bends may not leave you with a usable rudder but with a hole in the hull and/or damaged rudder drum. Whereas a a rudder post that snaps leave you without rudder - but undamaged hull.

 

It is possible to sail without a rudder, but not as easy with a big hole in the hull.

 

Just to demonstrate that any such scenario can be viewed from another angle. Jefa (Rudders) comments this on their homepage. They are experts on rudders and rudder systems.

 

//J

Share this post


Link to post
Share on other sites

//J

 

Point taken, However given that the stiffer section of the carbon post would transmit more energy into the hull potentially making your scenario more likely on a CF design.

 

Also most fatal CF post failures have not been impact damage. They have been cross current load changes that I (at least) think SS is a better material.

Share this post


Link to post
Share on other sites

As a curious observer, could someone address the foam core question?

 

I was under the impression that what is in the middle doesn't matter if the the tube is engineered properly. (IE: no point in filling core of tube up with carbon unless maybe you aren't bothering to do any maths at all)

Share this post


Link to post
Share on other sites

As a curious observer, could someone address the foam core question?

I was under the impression that what is in the middle doesn't matter if the the tube is engineered properly. (IE: no point in filling core of tube up with carbon unless maybe you aren't bothering to do any maths at all)

Yup. As someone said above. If the material does not stretch to share stress with inner layers then the entire load is on the very outermost fibers...until they fail. Then the next layer, which is not as capable to handle the load as it is at a smaller radius. And so on. Making a big bang sound and lotsa running around on deck.

 

So material inside is of no use whatsoever in resisting bending failure. Quite unlike steel. It is useful for unorganized stresses near the bearings and such. And maybe for durability in resisting scratches. I would guess a heavy woven outer layer is what someone who knows what they are doing would specify.

 

Than all said, carbon is far far superior to iron age materials...done right.

Share this post


Link to post
Share on other sites

I'll take a shot at the foam core question and say it makes a nice form to lay up some laminate. Certainly not structural at low densities. It might even keep water leaks from accumulating in the post and cracking during a winter freeze.

Share this post


Link to post
Share on other sites

Years ago we had the rudder break off in 8' seas power reaching under spinnaker. It was a Butler boat. Turned out the rudder post was not a single tube, but two tubes welded together; broke at the weld. Even though the boat was a decade out of warranty, Frank replaced the rudder for free, saying that the post was designed as a single tube, and that the contractor who made the rudder screwed up.

 

And yes, we made it to port sans rudder, using sail trim to steer.

Share this post


Link to post
Share on other sites

Years ago we had the rudder break off in 8' seas power reaching under spinnaker. It was a Butler boat. Turned out the rudder post was not a single tube, but two tubes welded together; broke at the weld. Even though the boat was a decade out of warranty, Frank replaced the rudder for free, saying that the post was designed as a single tube, and that the contractor who made the rudder screwed up.

 

And yes, we made it to port sans rudder, using sail trim to steer.

 

That is an example of why Frank Butler was such a successful boat builder. He stood behind his product. That is why I bought 2 new boats from him.

Share this post


Link to post
Share on other sites

A properly designed and built carbon shaft and rudder will not break under regular use. All of this starts with the structural engineer having the right numbers to start. This is pretty straightforward believe it or not. As previously mentioned, having that rudder fabricated by someone that is familiar with these types of structures is importantly. Having access to particular tools like an autoclave for example is also helpful and something that you should look for when seeking someone to fabricate your rudder and shaft.

Share this post


Link to post
Share on other sites

Had a meeting with Frank Butler years ago. His big office was lined with broken parts and gear. He took care of each claim himself.

Share this post


Link to post
Share on other sites

 

 

J-111 Rudder, Hull #3, 2010. Broke hard reaching in breeze.

 

J-111 Rudder, Hull #4, 2011. Broke hard reaching in breeze.

Stump didn't look as messed up as that pic though.

J/111 Hull #5, Broke hard running in breeze. Not kidding.

J/111 hull #1... Django? Lost at sea after shaft broke

Share this post


Link to post
Share on other sites

Rudders need to be designed for tip deflection when reaching. Probably around 15% of length. Got an endoscope screenshot of a maxi with what appears to be around 2ft deflection while going 18 to 20 broad reaching with an A3 up.

Share this post


Link to post
Share on other sites

Who built the original J/111 rudders? Who builds them now? I know a company is claiming to build them now, but not clear if they built from the beginning or not.

Share this post


Link to post
Share on other sites

 

Yup. As someone said above. If the material does not stretch to share stress with inner layers then the entire load is on the very outermost fibers...until they fail. Then the next layer, which is not as capable to handle the load as it is at a smaller radius. And so on. Making a big bang sound and lotsa running around on deck.

 

So material inside is of no use whatsoever in resisting bending failure. Quite unlike steel. It is useful for unorganized stresses near the bearings and such. And maybe for durability in resisting scratches. I would guess a heavy woven outer layer is what someone who knows what they are doing would specify.

 

 

 

This isn't quite right technically speaking. Every material - including carbon fiber - has flexibility, allowing the outer fibers to stretch and the inner fibers to take some of the load. Metals (as opposed to laminates) also have ductility, allowing the outermost fibers to permanently yield without rupturing, while the inner fibers continue to carry a load. The result is a damaged (bent) but not broken rudder post. This is a more graceful failure than rupture, but still a failure. The closer to you get to the center of the post, the less stress there is on the fibers, the ones at the very center will have no tension or compression on them at all (though they will have shear).

 

Standard mechanics of materials formulas account for all this by giving the maximum fiber stress at the outer, most stressed layers, and that is what should be used as the design criteria.

Share this post


Link to post
Share on other sites

+1 ^

Designing a layup isn't rocket surgery, finding the correct load case is where the hard part​ is. Balancing bending, shear, torsion loads will decide what fibre orientations​ you use and how much of each.

I wonder sometimes if shafts designed for a Max Von Mises stress might fail under conditions​ with lower combined stress but a higher directional stress

Share this post


Link to post
Share on other sites

My approach to my own post was to calculate max stress for the presumed worst case, then multiply by 4, and keep that below the fatigue limit of the carbon, not just the ultimate. This was much stronger than calculated by the NA and steering gear suppliers.

 

Like keel attachments, the rudder post is a bad place to save money and weight.

Share this post


Link to post
Share on other sites

I looked at bending and torsion, and tried to do combined bending and torsion. The worst case as defined by the NA, PYI, and Edsen was rudder at stall with boat at hull speed. I took worst case to be rudder at stall at 2x hull speed, (4 times the stress). We have never done 2x hull speed, but had the occasional surf to 1.5x hull speed and have run for hours at 1.25x hull speed. The calcs where done prior to easy access to FEA, I would do that now (carefully - garbage in = garbage out).

Share this post


Link to post
Share on other sites

We broke the rudder shaft on our SC37 not because there was anything wrong with the shaft but because the hole was too small and under hard load the shaft deflected enough to rub on the edge of the hull and this caused the break.

Share this post


Link to post
Share on other sites

Dont blame the material. Failure is usually an engineering problem. I had 2 SS rudder shafts snap off under the hull, a day apart. Poorly engineered.

Share this post


Link to post
Share on other sites

does the small area of the keel placing heavy lateral resistance loads to the rudder of a boat like a J-111 have anything to do with the failures?

Share this post


Link to post
Share on other sites

The post should be engineered for the area and depth of the rudder. It should be able to withstand maximum force (just before the stalling angle) at the maximum expected speed with a safety factor. You don't need to know anything else about the operating conditions.

 

With carbon, I think some designers become enamored with the ultimate strength specs, forgetting that in a laminate the fatigue limit is well below that. Less true of metals. Real engineers should know this.

 

What you say is true at Airbus apparently, where the carbon rudders break off because the designers didn't expect them to be used much.....maybe the J boat designers went to that school.

Share this post


Link to post
Share on other sites

The post should be engineered for the area and depth of the rudder. It should be able to withstand maximum force (just before the stalling angle) at the maximum expected speed with a safety factor. You don't need to know anything else about the operating conditions.

 

With carbon, I think some designers become enamored with the ultimate strength specs, forgetting that in a laminate the fatigue limit is well below that. Less true of metals. Real engineers should know this.

 

What you say is true at Airbus apparently, where the carbon rudders break off because the designers didn't expect them to be used much.....maybe the J boat designers went to that school.

 

Ummm

 

Not sure this is true, fatigue is typically more of an issue in metals than composites. (especially if there is a welded joint)

More of an issue is that you need to use higher safety factors in composites to account for manufacturing variability and hidden damage.

 

It may be true that carob epoxy has higher ultimate strength than the yield strength of steel, but personally I would tend to use a lower allowable strength for a composite laminate because of this variability.

Share this post


Link to post
Share on other sites

The fatigue strength curves for metals and composites are quite different (and both quite well studied). Metals typically have a well defined knee, composites not so much. Ordinary mild steel for example, has an infinite fatigue life cycled below it's critical stress value, which is why steel is such a valuable engineering material. Composites don't have an infinite fatigue life at any stress level - you pick your point in the curve and live with it. Carbon composites also have a free edge effect that is somewhat unique to them and often should be considered.

 

The yield point of 316SS is quite low, lower than a good aluminum alloy.

 

You should use these materials with knowledge of their proven, tested characteristics. I'll agree with you to the extent that a composite layup - and a steel weld - is a shop made product, if the people in the shop don't know what they are doing you better not expect too much of the product. Done correctly, both are quite predicable.

Share this post


Link to post
Share on other sites

 

 

 

J-111 Rudder, Hull #3, 2010. Broke hard reaching in breeze.

J-111 Rudder, Hull #4, 2011. Broke hard reaching in breeze.

Stump didn't look as messed up as that pic though.

J/111 Hull #5, Broke hard running in breeze. Not kidding.

J/111 hull #1... Django? Lost at sea after shaft broke

 

Calling hull #2. Hull #2 where are you?

Share this post


Link to post
Share on other sites

The free edge effect is unique to composites, and worse with stiffer materials. The layers of carbon are much stronger and stiffer than the epoxy between layers. When layers with high bending stress encounter an edge (or a hole) the highly stressed outer layers try to slide relative to the less stressed inner layers, causing delamination and failure propagating from the edge. If you take a stack of paper and bend it, you can see this happen at the edge. Metal (being homogenous and isotropic) and does not suffer this problem. The solutions seem to be: avoiding the situation in design, lower the stress magnitude with extra laminations near the edge, or wrap the edge with laminations which provides a continuing path for the stresses. If you believe some of the research, near a substantial hole you need an increase in thickness of nearly 50% to reduce the edge stresses to match the field material.

 

On something like a typical rudder post the bending stress diminishes to zero at the free edges (ends) and it will not be a problem - unless there is a hole, or steps in the section, or taper.

Share this post


Link to post
Share on other sites

The fatigue strength curves for metals and composites are quite different (and both quite well studied). Metals typically have a well defined knee, composites not so much. Ordinary mild steel for example, has an infinite fatigue life cycled below it's critical stress value, which is why steel is such a valuable engineering material. Composites don't have an infinite fatigue life at any stress level - you pick your point in the curve and live with it. Carbon composites also have a free edge effect that is somewhat unique to them and often should be considered.

 

The yield point of 316SS is quite low, lower than a good aluminum alloy.

 

You should use these materials with knowledge of their proven, tested characteristics. I'll agree with you to the extent that a composite layup - and a steel weld - is a shop made product, if the people in the shop don't know what they are doing you better not expect too much of the product. Done correctly, both are quite predicable.

 

scatter in carbon fiber epoxy test data is MUCH higher than scatter in steel test data. Better manufacturing shops have lower scatter but its always worse than metals.

its predictable, but you use a very different way of calculating the design allowables (B-Basis anyone)

anyone using a headline number for the ultimate strength without consideration for the degree of scatter is taking a big risk. Unless you have information about the actual data sets used to generate the ultimate strength number its a bad idea to trust it.

Share this post


Link to post
Share on other sites

 

 

 

 

J-111 Rudder, Hull #3, 2010. Broke hard reaching in breeze.

J-111 Rudder, Hull #4, 2011. Broke hard reaching in breeze.

Stump didn't look as messed up as that pic though.

J/111 Hull #5, Broke hard running in breeze. Not kidding.

J/111 hull #1... Django? Lost at sea after shaft broke

 

Calling hull #2. Hull #2 where are you?

 

Dang. Serious ouchie if all true.

Share this post


Link to post
Share on other sites

The free edge effect is unique to composites, and worse with stiffer materials. The layers of carbon are much stronger and stiffer than the epoxy between layers. When layers with high bending stress encounter an edge (or a hole) the highly stressed outer layers try to slide relative to the less stressed inner layers, causing delamination and failure propagating from the edge. If you take a stack of paper and bend it, you can see this happen at the edge. Metal (being homogenous and isotropic) and does not suffer this problem. The solutions seem to be: avoiding the situation in design, lower the stress magnitude with extra laminations near the edge, or wrap the edge with laminations which provides a continuing path for the stresses. If you believe some of the research, near a substantial hole you need an increase in thickness of nearly 50% to reduce the edge stresses to match the field material.

 

On something like a typical rudder post the bending stress diminishes to zero at the free edges (ends) and it will not be a problem - unless there is a hole, or steps in the section, or taper.

Yes. Well said. Better than my hack attempt above. Combine all the issues mentioned in this thread with the unfortunate quality and consistency issues in the marine industry (compare aircraft) and there will be unexpected failures. What may be carefully designed as a carbon rudder post can easily become no more that a Black Tube by the time Gilligan in the boatyard gets the build project.

 

Surely the laminates in a proper carbon layup are more intimately joined than by a layer of relatively rubbery epoxy? Right?

Share this post


Link to post
Share on other sites

Now, how about a definition of "scatter" please John.

I think it's a reference to data...whether the test data is in a tight target, or spread out which is less reliable. The tightness of the data is more or less the margin of error.

Share this post


Link to post
Share on other sites

Now, how about a definition of "scatter" please John.

By scatter I mean the variability of the data, my preference (for composites) is to understand the actual strength data for each test, but more commonly you will get a measure of standard deviation.

The standard deviation can be misleading unless you know what the distribution of results is (e.g. does it fit a normal distribute, weibull distribution or some other model.) Ultimately the objective is to estimate the probability of failure under the worst case loading condition.

Share this post


Link to post
Share on other sites

 

Now, how about a definition of "scatter" please John.

By scatter I mean the variability of the data, my preference (for composites) is to understand the actual strength data for each test, but more commonly you will get a measure of standard deviation.

The standard deviation can be misleading unless you know what the distribution of results is (e.g. does it fit a normal distribute, weibull distribution or some other model.) Ultimately the objective is to estimate the probability of failure under the worst case loading condition.

 

From the way you wrote it above it almost sounds like you mean the "variability of the data"... as in the measurement error. Or accuracy and repeatability of the test. But I don't think its the test method variability you mean, do you? I think (?) you mean the variability of the test subjects (in this case carbon tubes) built to supposedly the same standard.

 

In other words the scatter indicates the variability of the build of the test subject, not the variability of the test method.

 

Not intending to put words in your mouth and its an odd nuance because data scatter in this instance can be caused by either (test subject or test method variability or non-reproducibility). In this instance I think the test itself is highly reproducible but not so for the material/build.

Share this post


Link to post
Share on other sites

 

scatter in carbon fiber epoxy test data is MUCH higher than scatter in steel test data. Better manufacturing shops have lower scatter but its always worse than metals.

its predictable, but you use a very different way of calculating the design allowables (B-Basis anyone)

anyone using a headline number for the ultimate strength without consideration for the degree of scatter is taking a big risk. Unless you have information about the actual data sets used to generate the ultimate strength number its a bad idea to trust it.

 

 

Obviously you use the strength value in which you have confidence. This is also true of metals. Aluminum varies 20% between heat lots in the same grade. 316 commonly varies from 30Ksi to 80Ksi, depending on hardness condition. Same with 17-4PH. The scatter on carbon composite from a good shop will not be that wide. In either case you still use a single number for ultimate - one that you are sure you will get.

 

Thanks DDW. I knew holes had to be reinforced, intuitively, just didn't know the science.

 

Holes in any material will cause a stress riser: you have removed material and caused an interruption in load paths. But composite has the additional problem of free edge effects.

 

 

Surely the laminates in a proper carbon layup are more intimately joined than by a layer of relatively rubbery epoxy? Right?

 

 

No. But a properly engineered composite structure does ask for more than the rubbery epoxy can give.

Share this post


Link to post
Share on other sites

From another thread:

 

fastyacht, on 05 Apr 2017 - 7:11 PM, said:snapback.png

When I make a hole in new composite, if at all possible, I do it without cutting. Instead, I push a waxed pin through the weave during lamination.

Assuming there is room for the displaced laminate, how does this technique sound to an engineers ear?

Share this post


Link to post
Share on other sites

From another thread:

 

fastyacht, on 05 Apr 2017 - 7:11 PM, said:snapback.png

When I make a hole in new composite, if at all possible, I do it without cutting. Instead, I push a waxed pin through the weave during lamination.

 

Assuming there is room for the displaced laminate, how does this technique sound to an engineers ear?

 

Not perfect, but much better than a hole.

Share this post


Link to post
Share on other sites

 

Now, how about a definition of "scatter" please John.

By scatter I mean the variability of the data, my preference (for composites) is to understand the actual strength data for each test, but more commonly you will get a measure of standard deviation.

The standard deviation can be misleading unless you know what the distribution of results is (e.g. does it fit a normal distribute, weibull distribution or some other model.) Ultimately the objective is to estimate the probability of failure under the worst case loading condition.

 

I'd think the Standard Deviation of a significant number of test samples would probably do for most workshop fabrications, given the significant number of potential variables in such operations.

By far the easiest number (plus average) to calculate and try and control to.

(From my maths stats days doesn't the Central Limit Theorem say that in such multi-variable situations the data will tend towards a normal distribution? Stats was never my strong suit tho'.....)

Share this post


Link to post
Share on other sites

 

 

Now, how about a definition of "scatter" please John.

By scatter I mean the variability of the data, my preference (for composites) is to understand the actual strength data for each test, but more commonly you will get a measure of standard deviation.

The standard deviation can be misleading unless you know what the distribution of results is (e.g. does it fit a normal distribute, weibull distribution or some other model.) Ultimately the objective is to estimate the probability of failure under the worst case loading condition.

 

I'd think the Standard Deviation of a significant number of test samples would probably do for most workshop fabrications, given the significant number of potential variables in such operations.

By far the easiest number (plus average) to calculate and try and control to.

(From my maths stats days doesn't the Central Limit Theorem say that in such multi-variable situations the data will tend towards a normal distribution? Stats was never my strong suit tho'.....)

 

 

When you take a probabilistic approach, you're really getting into the realm of Limit States Design (I forget what it is called in the U.S.) rather than Allowable Stress. The concept of Safety Factor goes out the window.

 

IIRC, the Limit States process is more rigorous, but the two methods yield not too dissimilar results for the most part.

Share this post


Link to post
Share on other sites

I think the central limit theorum says that if you take a series of samples, the sampling distribution will be normally distributed whether or not the original distribution is normal -back in the day i used to teach this stuff...but that was way back in the day

Share this post


Link to post
Share on other sites

That video tests in torsion though, and that rudder shaft failed in shear.

Actually the rudder shaft failed in bending - no shear

Share this post


Link to post
Share on other sites

 

 

Yup. As someone said above. If the material does not stretch to share stress with inner layers then the entire load is on the very outermost fibers...until they fail. Then the next layer, which is not as capable to handle the load as it is at a smaller radius. And so on. Making a big bang sound and lotsa running around on deck.

 

So material inside is of no use whatsoever in resisting bending failure. Quite unlike steel. It is useful for unorganized stresses near the bearings and such. And maybe for durability in resisting scratches. I would guess a heavy woven outer layer is what someone who knows what they are doing would specify.

 

 

 

This isn't quite right technically speaking. Every material - including carbon fiber - has flexibility, allowing the outer fibers to stretch and the inner fibers to take some of the load. Metals (as opposed to laminates) also have ductility, allowing the outermost fibers to permanently yield without rupturing, while the inner fibers continue to carry a load. The result is a damaged (bent) but not broken rudder post. This is a more graceful failure than rupture, but still a failure. The closer to you get to the center of the post, the less stress there is on the fibers, the ones at the very center will have no tension or compression on them at all (though they will have shear).

 

Standard mechanics of materials formulas account for all this by giving the maximum fiber stress at the outer, most stressed layers, and that is what should be used as the design criteria.

 

Except that in compression the resin probably takes more load than the CF. CF is really poor in compression unless you can keep it perfectly in column (which is what the resin tried to do)

Share this post


Link to post
Share on other sites

Watching our guys wind prepreg shafts the first thing you noticed was how strong their hands were. High tension when winding. Multiple days to allow time for overnight vacuum debulking and cured in the oven under vacuum. Never a failure of the shaft did I hear about in the history of the company. Bearing buildups were also composite and machined on the shaft (Orkot). Works of art.

 

Properly engineered and built with the proper prepreg carbon and the shaft wont be the problem unless someone puts a blade on it that its not designed for :)

Share this post


Link to post
Share on other sites

All this talk about shafts and strong hands is making me wet.

Share this post


Link to post
Share on other sites

Hi ,

i am finishing a small cruising mirage30 extended to 31 feet sailing boat and one of the tasks was the rudder
Originally i had a stainless steel stock 1 1/2 inch and i have seen the same type of rudders on the same boat for 30 years no problems Apart all have absorbed  water and leaking

A boat designer and naval architect suggested to build the rudder with carbon stock

Reduced weight , and no problems with water

Initially would not have done it but the carbon stock was not that expensive , i would not advise laying it your self ever (time, and material is more or less the same by the time you do it )

But also when the design was done ,  most companies are not using tube but square or rectangular shape as it is easier for bearings and melding into the foam, due to the size restrictions and additional advise in regards of strength i used normal tube with extra strengthened carbon layed in 3 directions of 27 layers in total

I think that was the original problem of Parma post

Have done extremely long research on the subject, my rudder has bearing on the bottom middle and top skeg  so the layers was mainly on +- 45 and  few 0 degrees  but that was calculated in advance

One of the problems when you have rectangular shape for the sleeve and bearing can be a week point and if not supported on the bottom as mosr rudders are not and not produced correctly or sleeve is offset,  under sized or just any other manufacturing issue can break easily

For the above reason i decided to go ahead with plain tube and worry about the bearing afterwords

 

I have ordered the carbon stock with about 20% extra on what was calculated ( the calculation already have some reserve ) 

 

My view

In building the rudder i think was much easier in comparison to Stainless stock , i actually started first with the stainless

I have added extra layers around the tube to about 1/4 to 1/3 of the rudder for extra stiffening and more fraction ,which were not needed by design but to ensure that no slippage and make the rudder stock over 100 % more stiff in the working area with out producing hard spots  extra 1KG, 1.5Kg  who cares  From memory around the tube i have in total 3 for the ribs on 0 degrees, 6 layers   overlapping mainly between the ribs and 5 ;layers   overall most of the layers are 400g only the top 2 layers are 600g

All the ribs are also fiber glass instead of metal , note that ribs for carbon are probably optional  as per most composite designs but gives real strong shape . In my design ribs were not optional

Original weight of  the Stainless stock with the metal blades/ribs was close to 30KG final product with Carbon fiber, foam and glass is 9Kg well it is not painted yet

As i am doing in the garage all of it apart from the stock and bearings i think over all the final product is better in comparison what i was doing with the stainless and happy that switched over 

The Problems seen

- have to be careful not to put resign on the stock where the bearings are - after initial  fiberglass ribs   i used tape and foam cover the bearing parts

- Matching the bearings as the Carbon stock do not like Metal , preferred to use Wear sleeves from type of plastic and original  plain vesonite bearings . Note therer much better bearings but i already had some and also 3 point bearing is forgiving.

- If i start again i would do the foam differently easier way , instead of cutting to shape and gluing them with chpmat each 30mm a bit better stronger extra 500g of weight would just bugger peaces between the ribs and sand it do to the rib level

- With out a mold is difficult but nothing too hard a bit more sanding and fearing

The main advantages :

- No water

- weight - Especially when you have to put it on and off around 30 or more times to do the adjustments and measurements and tests

- as all the surface of the stock is glued a small mistakes probably will be more forgiving , but still need to be careful for the air especially around the stock 

- better bonding

- temperature and stress is equal across as the same type of materials

 

Lessons learned::

If you have to lay multiple layers and event for the first layer . Do not use machine for the final preparation sanding ever.

It takes only 10 to 30 minutes to hand sand 0.6 Sqm , but makes a lot of difference for the gluing . That is after the machine sanding

The machine sanding even with 40grid do not proved nice surface for bonding and can skip spots especially uneven places or around the stock

Total hours spend  around 50 hours may be 60 hours  but the calendar time with the research and only using a few weekends  was 3 Months, well i still have another boat in the water can not spend every weekend on this one

 

Finally:

I agree if the Carbon/Composite fiberglass/ Eglass is done properly it is difficult  to fail

One of the main argument against carbon for cruising boat is what if you hit rocks , well in stainless it bents and you can not sail the boat in carbon it breaks the sami. Which is better no ruuder or stuck rudder at angle . In both cases spare rudder is a good idea . If you have skeg better as less chance to break it

I hope that helps some one as in my researches got help from a lot of people , any questions or advises are most welcome

 

 

Insert other media

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now