Rudder Failures

[UNCIVILIZED]

There are plenty of AP's out there which can "learn", quite literally. However, the more sophisticated you want the AP, the more it & the instruments are going to cost.
You wont find many, if any basic ones which will do this.

A "smart" system does things like the following. We all know that wind changes can be sensed at the masthead before we feel them at the deck level. And when we're driving, if we're paying attention, then we see such, and can alter the boat's course to compensate for them before such changes fully impact the sails.

Such as if we notice a gust coming, we can bear off or come up, depending on what course/destination point we're trying to steer the boat to. And when we do that, the forces on everything onboard go down compared to if we had just stayed on course. Including the sails, spar, & steering system components.

On top of which, a good helmsman is using this little bit of extra wind to push the boat towards it's destination faster, such as using a puff to come up a couple of degrees due to the shift in apparent wind, thus increasing CMG & VMG. This as opposed to simply muscling the wheel, just blindly trying to follow a compass course.

Now with good wind instruments, & one of the smarter brains which both reads the instruments input, as well as providing input to the AP. The instruments will pick up on this change in the wind early, before it fully impacts the sails & boat (as the wind instruments are at the masthead), just like a good helmsman would. And it'll send a message to the AP to come up, or bear off a few degrees, just as a human helmsman would. Again, lessening the forces on the steering components (rudder included).

It does the same exact thing when a wave strikes the boat. As waves too have the effect of altering the boat's course. Though again, the smarter the pilot, the better it'll react.
It'll read off of a gyroscope onboard to see in what direction & at what speed the wave strike is trying to push the boat (how much acceleration's involved), & respond proportionally. And if it's "smart" enough (good enough software), it'll wait until the full impact of the wave strike has passed, prior to bring the boat back onto course (these functions are tuneable).
Where as a "dumb" pilot will simply try & muscle the boat back to the proper heading, it being unknowing of what forces are acting on the boat, or how & when is the best manner to bring it back on to/keep it on course. So it just pushes harder on the rudder... until the boat's back on course, or eventually, over the course of X number of cycles like this, something in the steering system breaks. Be it the pilot, the rudder, various mounts, etc.

The higher end the "brain", the better it'll drive. As between the crew/navigator uploading the polars into it, & the system learning how the boat handles in differing wind & sea conditions, it corrects for changes further & further in advance. The better ones steering to a far better standard than all but a couple of percent of the top helmsmen in the world.

This is important because, as it's anticipating & correcting for changes in sea & wind changes farther & farther in advance, it needs to use less & less rudder to keep on course. Thus causing less wear on everything on the boat, including and or especially the rudder.
Plus given time, good software, & an integrated instruments package, it learns how a boat handles, just as a human helmsman, & as it learns uses less & less rudder to reach point X. Thus also increasing VMG & CMG.

And of course some of the perks from this, are a smoother ride, less power draw from the pilot, less wear & tear on gear, & faster sailing, as less rudder is being used in order to keep the boat on course.

[colemj]

Uncivilized,

Exactly. This sums up why AP integration with the rest of the instrumentation is actually a very good thing to do, as well as why saving a couple hundred dollars on a system that does not have these brains and capabilities is false savings.

I am surprised in the threads where people go on about the evils of integrating the AP with instrumentation, or extoll the virtues of 30yr old systems or the necessity of sitting at the control head diddling trim pots around as conditions change. I can only surmise that those people have never sat watching the helm in rough conditions while a current-model NKE or Simrad/B&G AP with a fast-response drive unit is in command.

A good autopilot hardly uses any power at all. Even in rough conditions, our pilot rarely moves the rudder much, and when it does it is hardly ever in response to being off course - it is almost always in anticipation of how the boat will react to a wave/wind change before the boat starts to actually react. Our AP is rated for 30A - I have yet to see it exceed 5A.

I have used too many older systems that only react when the boat is already under the control of a wave or wind change, and then burns up amps trying to muscle control back while yawing around and over-compensating, to ever go back or recommend old or cheap technology. Our previous old generation AP used 17-20A in rough conditions and was rarely ever in control of things - always reacting to them.

Mark

[a64pilot]

Originally Posted by colemj

Uncivilized,


A good autopilot hardly uses any power at all. Even in rough conditions, our pilot rarely moves the rudder much, and when it does it is hardly ever in response to being off course - it is almost always in anticipation of how the boat will react to a wave/wind change before the boat starts to actually react. Our AP is rated for 30A - I have yet to see it exceed 5A.
Mark

How do they anticipate? What kind of sensor do they have that does this?
I think they can only react to a course change, I don't believe they can anticipate anything, but I'm willing to learn, and if there are some that can, I want one.
I know there are autopilots that can "learn" and adjust how much rudder is needed for a desired result, but that's not reacting to something that hasn't happened yet.

[neilpride]

Originally Posted by UNCIVILIZED

I don't doubt you in the slightest, but I'm definitely curious to see pictures. And this is FAR from the first time I've heard of bulkheads in Beneteaus being attached with something out of a caulking gun. Some friends who'd done a lot of cruising mentioned it to me 20yrs ago.
That said, it's still a jaw dropper to hear. I mean, I know that they use wonder adhesives to glue the window-walls into skyscrapers, but methinks that it just doesn't attach to enough surface area to properly affix a bulkhead to a hull.

Here are some pictures... include the cleat washers under the deck, and i really i dont know why they open a opening in the bow where looks like a crash box ?? this frenchy guys are crazy...

[neilpride]

More...

[neilpride]

Smack you can add this pics to your production boats and the limits SN ...

[a64pilot]

On the short arm thing, as it get beyond center the effective length gets shorter and the ratio of arm movement to rudder displacement changes.
I can see how a significant of lateral force can be imparted into a rudder stock from an autopilot arm, if the designer doesn't take this lateral force into account in the design, it could exceed design loads, and how can you hold a designer responsible for a modification to their design?
I can see ideally where an autopilot ought to use the same load paths for steering as the factory installed steering system, I believe most factory installed steering systems impart a rotary motion to the rudder stock, not a lateral one, I know my rack and pinion one does, in mine I won't use a hydraulic autopilot and an arm on the rudder stock, I'll drive the pinion gear with a rotary drive.

[robert sailor]

Well they certainly are building them different these days. Our boat, a Moody has all the bulkheads and furniture heavily tabbed into the hull/deck and then they thru bolt all these at about 8" centers, sort of a belt and suspenders sort of thing. Its pretty easy to see where the $$$$ is taken out of the construction. Its also interesting to see how thin these hulls are as well. Obviously they are up for the job they were designed for as most of them are hanging together but they sure don't give me the sense they have been built for the long run.

[colemj]

Originally Posted by a64pilot

How do they anticipate? What kind of sensor do they have that does this?
I think they can only react to a course change, I don't believe they can anticipate anything, but I'm willing to learn, and if there are some that can, I want one.
I know there are autopilots that can "learn" and adjust how much rudder is needed for a desired result, but that's not reacting to something that hasn't happened yet.

They have algorithms that take sensor data and use both stored past boat response data and current response data. The boat itself doesn't start to react to a change in condition until a long time after the rate and yaw sensors, compass, wind sensor, etc detect a change in conditions. So while anticipating a condition that hasn't happened yet is not possible, it is very possible to anticipate a boat's reaction, that hasn't happened yet, to a condition that is currently changing. It is primarily how a human steers a boat.

Those sleds racing in the Southern Ocean in 60kt winds and 40' seas are using currently available AP's whose software has been tweaked to closely match the characteristics of those squirrely boats in those conditions. Otherwise, you can buy these AP's off-the-shelf for your boat (almost all of them are NKE and B&G), and even the less expensive versions have the same types of algorithms and advances.

These things don't cost an arm and leg - they are just modern autopilots.

Mark

[colemj]

Originally Posted by a64pilot

On the short arm thing, as it get beyond center the effective length gets shorter and the ratio of arm movement to rudder displacement changes.
I can see how a significant of lateral force can be imparted into a rudder stock from an autopilot arm, if the designer doesn't take this lateral force into account in the design, it could exceed design loads, and how can you hold a designer responsible for a modification to their design?
I can see ideally where an autopilot ought to use the same load paths for steering as the factory installed steering system, I believe most factory installed steering systems impart a rotary motion to the rudder stock, not a lateral one, I know my rack and pinion one does, in mine I won't use a hydraulic autopilot and an arm on the rudder stock, I'll drive the pinion gear with a rotary drive.

The lateral force thing is a red herring. Even cabled quadrant systems supply a lateral load when turning the rudder. Hydraulic steering systems are also rams hooked to short arms. In rotation, that lateral force is spread around 90-180* of the bearing, then changes to the other 90-180* on the return side.

It doesn't take much bearing engineering or mounting to account for this. Even in the pics above, that system design is just fine for these loads, with the exception that it isn't tabbed into place.

There are a lot of advantages of having your autopilot drive a separate system from your normal drive - both in redundancy and in removing excess play and drag that just hurts an AP performance.

Mark

[Polux]

Originally Posted by neilpride

Here are some pictures... include the cleat washers under the deck, and i really i dont know why they open a opening in the bow where looks like a crash box ?? this frenchy guys are crazy...

What surprised me more in your pictures, taken from a boat that was grounded by an hurricane and suffered therefore impact forces that he was not built to resist is how well that green glue resisted; Not a fissure, not a crack:
amazing!

[a64pilot]

The green glue is phenomenally strong, it's not tabbed so the load is concentrated, not spread out. Failure mode that I have seen is the parent material breaks right at the glue joint.
Now that was on powerboats as I'm new to sailing but I don't see the difference, that stuff is very widely used in high production smaller powerboats, like my old Trophy Pro center console.

[Polux]

Originally Posted by colemj

They have algorithms that take sensor data and use both stored past boat response data and current response data. The boat itself doesn't start to react to a change in condition until a long time after the rate and yaw sensors, compass, wind sensor, etc detect a change in conditions. So while anticipating a condition that hasn't happened yet is not possible, it is very possible to anticipate a boat's reaction, that hasn't happened yet, to a condition that is currently changing. It is primarily how a human steers a boat.

Those sleds racing in the Southern Ocean in 60kt winds and 40' seas are using currently available AP's whose software has been tweaked to closely match the characteristics of those squirrely boats in those conditions. Otherwise, you can buy these AP's off-the-shelf for your boat (almost all of them are NKE and B&G), and even the less expensive versions have the same types of algorithms and advances.

These things don't cost an arm and leg - they are just modern autopilots.

Mark

To do that the autopilot has to be linked to the rudder through a sensor that in many cases is an option to the autopilot and deserves all the money it costs.

This season when they dismounted the rudder they did not connect rightly the sensor and I was not able to reconnect it the right way (the damn picture on the manual was misleading!). I thought I had a software problem and made the season without the sensor: What a difference!!!! for worst off course.
The thing was working in what she called "virtual mode" and the ability to anticipate was greatly reduced and in difficult sailing conditions I had to hand steer, or take some sail out. I was so happy when they found out that the sensor was just mounted backwards!!!

[transmitterdan]

The scariest thing I noticed in Neil's pictures is that it seems the laminate plywood is now a structural part of the boat. Instead of a platform for the fiberglass as in days of old the actual wood and glue in the laminating process is taking stress loads. Laminates are strong but only in certain axis. I think they are least strong when stress is trying to pull the laminate apart as could happen in a boat. The green goo is almost certainly stronger than the plywood panels.

The fender washer cleat is a joke, right?

[neilpride]

Originally Posted by Polux

What surprised me more in your pictures, taken from a boat that was grounded by an hurricane and suffered therefore impact forces that he was not built to resist is how well that green glue resisted; Not a fissure, not a crack:
amazing!

Maybe you want the pictures with the plexus cracked?? and 2 dislodged bulkheads?? really?

The boat suffer damaged in a marina, they broke free from the dock and end in the rocks surrounded the marina, now go ahead Pólux and tell your Green snot theory to the owners of Blue Pearl, another oceanis 50 lost in the atlantic...