Anti-Capsize Technology and Systems

[Manateeman]

Dear Mr. Catsketcher. Perhaps you feel thread drift is a good thing. We do not.
Why not start your own thread on catamaran flips.
Chris White and I are from the same small town. I’ve built catamarans professionally. You have your opinions on how to sail. That’s nice. The thread very politely asked posters to discuss the technology of devices.
You have continued to make your point again and again that you believe sailing skills are far more important.
But that is not the subject of this thread.
Monohull vessels and catamarans are different but anti capsize device technology might be very useful to monohulls. The discussion is how to improve safety with a device. If you want to discuss catamaran flipping, please begin your own thread.
For that reason I’m going to ask the moderators to remove your post.
Captain Mark and his manatee crew of professional mariners.

[Manateeman]

Load cell systems. Well we will never know if these systems might or might not have made any difference in the capsize of the two Chris White boats.
So does that mean all load cell based systems will not work or only that the currently available systems won’t do the job.
A load cell is a really simple device and it’s what it’s connected to that matters.
The speed and capability of computers / processors is what drives what we would call a system because you can have more than one sensor. Just because the current equipment is based on one sensor that does not mean multiple sensors and a far better computer cannot react fast enough.
So we see three parts of the system. The sensors, the computer and the action.
Same in humans. Our eyes, ears, skin..just sensors. Our computer, our brain.the action is both what we do with our body and change on the boat.
Isn’t it reasonable to suggest a non human system could out perform our reaction speed...because it seems the speed with which the wind changes is one key factor. The second is what devices will speed reducing the sail power.
We know just how fast these gusts / microbursts happen. We are not ready to say not much can be done.
Mark

[catsketcher]

Some of my posts show that I think that devices that do not reflect the need to change sailing angle are not going to be as effective as they could be. As shown by the capsize of Anna and others. I have also posted on the inneffectiveness of pendulums, as they are very simple devices that can be triggered by waves. My posting in a thread doesn't stop others from doing so but if the moderator deletes or removes posts, I am okay with that. They are the unsung heroes of this forum.

I was talking to my son last night. He is building my cats new PyPilot autopilot. He says that the 9 axis combination Inertial Measurement Units (IMUs), can be easily cross reference to sort sway from roll, to detect real changes in the cat's attitude or heel. So simple single axis accelerometers (like pendulums) can be bettered by the simple and cheap technology found in our phones. But he left again before I could tie him down on more. Again someone like Sean Depaigner, who wrote PyPilot, is probably the gun to talk about with getting real heel angle from a simple IMU, getting that heel angle to be user selectable and for the IMU to couple to an Ardunio and for it to produce a voltage that could flip a relay. The resolution of an IMU is high with a rate up to 200 Hz.

Paul McWorter has a whole series on Youtube about using IMUs. As they cost about $50 and the Arduino about the same, you could possibly get a really lovely heel (and pitch) sensor for much less than Up side up. You would still need a relay to release a cleat or snap a snap shackle.

https://www.youtube.com/watch?v=2AO_Gmh5K3Q

I am pretty good with glass and epoxy and can wire up 12 volt systems, but the IMUs are a bit much for me. Maybe someone out there is expert and interested in this stuff.

cheers

Phil

[Manateeman]

I apologize. It’s obvious you are sincerely interested in the technology.
Your latest post is exactly what I had hoped would surface in the thread.
I got to ask my doctor to adjust my meds.
Mark

[catsketcher]

No worries, no harm done.

[thinwater]

Originally Posted by Franziska

... By sheeting in before the gust hits one may actually reduce the exposed main area on a downwind course.

Unless a bow digs in, the rudder stalls, and the boat broaches. Then you have made it much worse.


All of this "bear away" talk assumes a responsive helmsman (the AP will not be fast enough) and that the rudders are biting well. I've sailed a lot of multies, and many require the bear away to be slightly preemptive for it to work. Once the boat starts to load up, the bow will be pressed down and it won't bear away. I recall a few AC cats getting in trouble that way. Some cruising cats have pretty undersized rudders. My F-24 won't bear away if you wait too long and an ama buries; it will round up. You need to blow all the sheets, not just the main. On the other hand, if you bear away just a little earlier, she takes off on a plane, smooth as anything.

[Franziska]

Originally Posted by thinwater

Unless a bow digs in, the rudder stalls, and the boat broaches. Then you have made it much worse.


All of this "bear away" talk assumes a responsive helmsman (the AP will not be fast enough) and that the rudders are biting well. I've sailed a lot of multies, and many require the bear away to be slightly preemptive for it to work. Once the boat starts to load up, the bow will be pressed down and it won't bear away. I recall a few AC cats getting in trouble that way. Some cruising cats have pretty undersized rudders. My F-24 won't bear away if you wait too long and an ama buries; it will round up. You need to blow all the sheets, not just the main. On the other hand, if you bear away just a little earlier, she takes off on a plane, smooth as anything.

I agree there is a point against it.

It's when you realise that the wind speed is going up for a longer time than initially thought.
Than you better luff up while it's still possible.

We actually did have to do such an "emergency" luffing today and I am very glad it worked.
We wanted to reef the main and the idea was to furl the genoa away quickly to do it more easily.

Stupid enough the furler jammed half way through the furl and the wind was building quickly. We managed but our genoa now needs some stitches...

Sorry about the tread drift Mark.
Won't happen again.
Interesting systems discussed so far. Still following.

[catsketcher]

Originally Posted by thinwater

Unless a bow digs in, the rudder stalls, and the boat broaches. Then you have made it much worse.


All of this "bear away" talk assumes a responsive helmsman (the AP will not be fast enough) and that the rudders are biting well. I've sailed a lot of multies, and many require the bear away to be slightly preemptive for it to work. Once the boat starts to load up, the bow will be pressed down and it won't bear away. I recall a few AC cats getting in trouble that way. Some cruising cats have pretty undersized rudders. My F-24 won't bear away if you wait too long and an ama buries; it will round up. You need to blow all the sheets, not just the main. On the other hand, if you bear away just a little earlier, she takes off on a plane, smooth as anything.

That is an excellent point. I totally agree that easing sheet is vital for bearing away in breeze. I will get a little misty eyed with my cat and say that she is beautifully well behaved with her big rudders and medium sized rig. But your point is well made.

So if I was to make an anti - capsize control for my cat. It would need two inputs. One for wind speed and one for heel, and two actuations.

If the wind speed increases to my set limit, the main gets dumped say 1 metre and the boat bears away (if below death zone angles). If above death zone angles it just heads up. I can build something like that - a 1 metre dump made from a relay pulling a snap shackle trigger at the top of the mainsheet which has a dacron rope 1m metre safety and an IMU on an Arduino (actually I need someone to do that part but I have someone in mind).

If the boat heels too far over it will trigger a similar response, ease of sheet and steering action. Maybe heading up is the best solution unless easing sheet can be incorporated into the system, or the boat is below about broad reach conditions where the main is already eased. So possibly the "death zone" is very different when on autopilot compared to hand steering.

Good points Thinwater

[CatLove]

It looks like trying to reinvent the wheel is the substance of this thread. Systems that measure rig loads, runner loads, foil loads ,foil angles, 9axis IMU's etc exist and are being used both by racers and cruisers .


Here's a link to one of the IMOCA60 dashboards as it was used thru the last Vendee Globe race:
https://exocet.cloud/grafana/d/bsbc_...osk&refresh=5m
Inactive now but still shows all the parameters that were measured and monitored.


All Vendee Globe autopilots were able to not capsize boats in the last race. Stuff broke but boats did not capsize due to autopilots being too slow as being hypothesized here. Yes, IMOCA60 are monohulls in name but given the last few editions with foils and before that canting keels they have similar speeds with fast multihulls. And they sail in the hardest weather conditions in the Southern Ocean.

Here's a link to the NKE autopilots used on the IMOCA60's :
https://nke-marine-electronics.com/ocean-race/imoca/
They state 3D calculations at 25Hz .



Another reference point for computers being used , measuring and controlling stuff on boats is the flying catamarans of America's Cup.



Even the lowly Raymarine EVO autopilot EV-1 sensors have 3 axis accelerometer, 3 axis gyro and 3 axis magneto plus GPS .
Quote from the EVO specifications :"Precision monitoring of heading, pitch, roll, and yaw allowing the autopilot to evolve instantly as sea conditions and vessel dynamics change".


Even the mini transat 6.50 are now flying on foils or scow type and have perfect stability under autopilot:
https://nke-marine-electronics.com/o...ace/mini-6-50/


At this moment wondering whether good autopilots could steer under all wind and sea state conditions without capsizing the boat is ludicrous. It may not be doable with an Arduino uno r3 and a simple $5 9DoF IMU without having deep pockets for software development, testing ,protyping, rinse and repeat. Or at the very least just a lot of work if not deep pockets. But professional level equipment doing it already exists .

More details about real product implementations (clue - not by using steel weights and pendulums) can be had in great detail by searching the USPTO for the patents associated with the above autopilots.

[CatLove]

Also I recall maybe in the 1990's I saw a blurb about a system to prevent cruising catamaran capsizes that was purely mechanical. It was based on some floats that were free to move vertically in tubes placed vertically in each hull bow and stern . Their bottoms were open and allowed water in and out freely.

There was a mainsheet release that reacted if the floats inside the tubes dropped too low, meaning the bow or stern was coming out of the water.I believe the system had settings to release the mainsheet either when both floats on one hull were down, indicating a flying hull, or just one (bow or stern) , being a safer setting.
This may have been installed on a Catana . Memory is foggy though about these details.

[catsketcher]

Ah yes - the cool tubes by Tristan Jones and Leo Surtees.

https://patents.google.com/patent/US4541356A/en

I really like this forum. I have had some excellent arguments (in the proper sense of the word) that have helped me work out what I want.

Personally, I won't or can't buy the ready made kit, both for a top line autopilot and Up side Up kit. The cost for a good autopilot is more than $5000 in Australia. Some people then buy a spare drive unit. Of course I could just buy the stuff, but Sean has PyPilot for sale for about $100 and I can get a nice powerful motor for another $100. (The motor in my Evo Raymarine autopilot is a total joke - I have replaced it a couple of times already). So a good (hopefully) autopilot for about $200 instead of $5000. Open source is amazing. The Open source community have been very good at developing autopilots for drones using the same IMUs we can buy for $50. There is an awful lot if intellect being thrown at the issue.

As well as that, I can carry a whole spare kit for not much when I go cruising again in 2 years. Also, because the thing is not proprietal, any motor can be used, even a wiper motor if needed. I am not tied into spending huge money because I have already spent much.

I would like the same open source idea for an anti capsize system. For a second $50 IMU, an Arduino and relays you could get lots of the function of a top quality system at again, vastly reduced costs. Also it does seem an easier problem, getting heel data from the IMU and connecting to relays and the autopilot. I won't have the spare cash to buy Up side Up, but I certainly do have the money for an open source system. It is trivial compared to the off the shelf systems and so there is a good chance I could get one, substituting some time for lots of money.

Anyway - I will bow out now about my own ends. Thanks for the well thought out points. I will ask my son about my diagram I have drawn about inputs and outputs. Maybe he can make me something worth writing and I can install it.

Cheers

Phil

[Franziska]

Originally Posted by catsketcher

Ah yes - the cool tubes by Tristan Jones and Leo Surtees.

https://patents.google.com/patent/US4541356A/en

I really like this forum. I have had some excellent arguments (in the proper sense of the word) that have helped me work out what I want.

Personally, I won't or can't buy the ready made kit, both for a top line autopilot and Up side Up kit. The cost for a good autopilot is more than $5000 in Australia. Some people then buy a spare drive unit. Of course I could just buy the stuff, but Sean has PyPilot for sale for about $100 and I can get a nice powerful motor for another $100. (The motor in my Evo Raymarine autopilot is a total joke - I have replaced it a couple of times already). So a good (hopefully) autopilot for about $200 instead of $5000. Open source is amazing. The Open source community have been very good at developing autopilots for drones using the same IMUs we can buy for $50. There is an awful lot if intellect being thrown at the issue.

As well as that, I can carry a whole spare kit for not much when I go cruising again in 2 years. Also, because the thing is not proprietal, any motor can be used, even a wiper motor if needed. I am not tied into spending huge money because I have already spent much.

I would like the same open source idea for an anti capsize system. For a second $50 IMU, an Arduino and relays you could get lots of the function of a top quality system at again, vastly reduced costs. Also it does seem an easier problem, getting heel data from the IMU and connecting to relays and the autopilot. I won't have the spare cash to buy Up side Up, but I certainly do have the money for an open source system. It is trivial compared to the off the shelf systems and so there is a good chance I could get one, substituting some time for lots of money.

Anyway - I will bow out now about my own ends. Thanks for the well thought out points. I will ask my son about my diagram I have drawn about inputs and outputs. Maybe he can make me something worth writing and I can install it.

Cheers

Phil

Very right about costs and affordability. If you come up with some open source solution that would be awesome.

[CatLove]

The system I was recalling was not Tristan Jones' .
It was sold under the name Aeromax and there is a patent for it:
https://patents.google.com/patent/WO1998024687A1/en

[catsketcher]

Interesting idea. I think that actuating a relay AFTER a hull is flying is going to cause the sheet to be released too late. You are on the wrong part of the stability curve when the sensor says to release and that can't be good. You want any release to occur before you hit the peak stability and this occurs on all cats just as the windward hull lifts.

Also I can't work out how the sensor would work and release properly in a world of chop and the somewhat adhesive and turbulent properties of water in a tube.

But searching patents is a great idea.

cheers

Phil

[CatLove]

There seems to be support for sailboats in Ardupilot:


https://ardupilot.org/rover/docs/sai...lboat-hardware


https://ardupilot.org/rover/docs/sailboat-home.html


Here's a link to a completed project (using Pixhawk) of an autonomous sailboat that describes the system architecture, software etc:
https://www.frontiersin.org/articles...21.630081/full


However ,DIY'ing something like this won't be as cheap as people might think. Pixhawk , IMU, GPS, actuators all add up and in the end it won't be much cheaper than buying a Raymarine EV-1 sensor core and ACU , which can be had for around $1k . Orders of magnitude simpler too. Then add whatever motors, pumps, actuators you desire to the Raymarine core.