Aha, okay, I got it.
Sharing some wing experiments with material, design and builds.
The Laser has ~950 sq.inches of sail per default (~2400 cm2). Keeping that same area and using the proportions from the reference boat in earlier post, then height will be 110cm and width 22cm using the aspect ration ~5:1.
The area of the trim tab should be ~12% of wing area and using same 5:1 aspect ratio gives height of 37cm, width 7cm. The control flap will have a lever arm with a length of 50% the wing span, so 110/2 = 55cm.
The wing profile will be a symmetrical airfoil, and trying to build it close to dimensions of NACA 0018 (nice and fat max thickness 18% at 30% chord) but I am just using a simple step design with no smooth curves, like Kline-Fogelman style.
I have also assumed the aerodynamic center of pressure will be at 25% of chord and will place rotation axes there (mast).
I have not used the water resistant foam board discussed previously but a 0.5mm think more simple one, coated with paper and thin plastic layer. The surface needs to be painted to protect from sun and water and paint seem to stick and stay on the surface and also keep water away. However, time will tell if it will fall a part or not. Good enough for now I think.
I made two Wing prototypes with the length of only 20cm not the full 110, to figure out how to use the hot glue etc etc. Both were designed to have a mast fixed to the boat and not turning as the Wing turns. Reason is to be able to mount the new wind instrument at mast top not at front deck. The mast is a 10mm diameter and 1mm think carbon fiber tube (same dimension as Laser default).
The first one has a PVC-U tube glued inside the wing with an inner diameter of 15mm. The idea is to have the fixed mast move inside the PVC tube using 3-5 bearings so the wing can turn without any binding.
However, having completed that I found it a bit heavy.
So I decided to make another one without any tubing. Idea is to drill a long hole for the mast after the wing is made. It could maybe work as good as the first one. The PVC tube will increases the weight ~80% and if completed the wing would weigh 0.55kg compared to 0.3kg without.
I am not sure I trust myself drilling a 110cm long hole in that wing. It seems hard to keep a straight angle for such a long hole. Just when drilling the 37 cm hole in the trim tab, it did not turn out straight. It was just barely acceptable.
So I ended up with the Wing in the first picture - that mast is glued in to the wing and is fixed there. It will not work to mount the wind instrument in mast top.
Maybe it can work as plan B if the other two wing designs fails.
I spend some time thinking how to design and attach the trim/control flap to the wing, with two rods or will one be enough? I decided to first try make it with one and if it fails then use two. For the mounting I came up with making a little housing with the PVC tube, just 3cm long, for two bearings. A carbon fiber tube running through the leading edge of the trim flap will then move inside the bearings without binding.
The housing will then be attached to just one rod that will be fixed to the Wing.
Thats next.
Also need attach a servo and a weight balance so the whole rig will be neutral.
2 Likes
looks really good!, Iâm excited to see how it works
Latest bit of testing. NMEA wind vane works well, significantly better than my potentiomiter windvane. High mounted antennas have made big difference to the range, at-least double couldnât test and further and still see where I was driving.
Motor sailing is working fairly well, switches nicely between motoring and sailing. Still one or two things to iron out, but should be useful for getting out of tight spots or if the wind drops.
3 Likes
Great to hear the new NMEA guy works noticeably better.
The motor mode seem giving great boat speed.
How do you control the thrust of motor? or maybe you donât, just set it at some value and then on or off?
In manual mode you setup a aux RC option to control the sail and use the main throttle for the motor. In modes above manual it does it automaticaly based on a 3pos switch.
- Always sail, never use the motor
- Allow motoring, this uses a wind speed threshold to trigger the motor
- Always motor, this tuns off sailboat navigation and uses the motor
1 Like
Hi, sharing some more design stuff.
Had to make some changes to the tail design to make it lighter.
Reason is that every single gram put out there will have 0.55 m leverage and needs to be balanced off on the front of the wing. So, made another design with lots less weight but also lots less stability. Still needs serious wight to balance it.
Since I just hate to put weight on that is just ballast, I prefer to redistribute the items thatâs really needed in such a way it creates balance. Like motor, ESC, LiPi, Servo, Antenna.
Most of those guys are just put in a plastic box for now, inactive, but still
apply their weight.
I simply put the the angle control on a nob of the transmitter.
Deflection angle is around ± 30 degrees of the control tab.
Wing Sail test with rig 1 out of 3.
(+) can fit the wing sail in the blue bag for transportation
(+) if wind over 2 m/s then the control tab have have control over Wing angle
(-) light winds seems not fit for this type of wing/sail?
All in all - I am happy with the test. Itâs working as expected for the control tab to be able to control angle of attack of the Wing.
I am sorry I need to go on a trip for 4 weeks getting back 23 July.
So cant do more testing until after that date.
2 Likes
looks really good, Thanks for building it! no worries about the delay, it will give me a chance to get the code done!
Thank God 
and thanks a million for all the coding you already done to the sailboat!
When I get back my NMEA instrument should also have arrived.
Hi folks.
Wow, reading some of these comments has been mind blowing! We are working on a boat with a group of young people as a tech challenge. It is a 1M Hurricane RC boat. I am a sailor but not a great techy and we are a bit stuck. Wondered if any of you could shed some light?
We have taken a slightly simplistic approach and have sort of committed to seeing the plan through given that the idea came from the group, should just about work, and - provided we can learn how to set some of the mission planner parameters and modes to the right values/settings - should get us round a triangle course which will be the challenge.
Our build consists of a Pixhawk 2.4.8 connected to mission planner. We then have an Arduino Uno which we have coded and connected to the wind vane (situated on top of the mast) and the two are designed to switch between them using a relay.
The basic principle we are applying is that the pixhawk will attempt to steer the boat around the way points using fixed sheeting (a massive compromise in know) Should the wind become unfavourable the arduino will kick in and take over control from the pixhawk. It will run a loop which essentially bears the boat away from the wind and sails for 5 seconds until then switching back to pixhawk and allowing it to try again for the way point.
I get how rudimentary this approach is, and that it is massively limiting in terms of sail trim, obstacles etc etc but⊠We are pressing ahead for now to try and realise the groups ideas.
Our problem is that as things stand we have absolutely no success in getting pixhawk to try and control the boat at all. Reading this thread tells me we need to look at getting it MP in the right flight mode. we are about to try Acro, auto and steering modes to see if that helps.
Also aware that we might not have steering/throttle servo control set up correctly in the parameters. As we only want to steer the boat at this stage and donât want to control the sheets particularly, what would any of you suggest?
Thanks in advance! we are in the workshop all week this week trying to get the problem sorted. If anyone can chat on the phone Iâm happy to share my number.
Ed
Welcome here Ed 
Lets await the gurus answers here,
but I just wonder why you complicate things with a pixhawk plus an Arduino Uno?
Why not just stick to one pixhawk controller?
Itâs basically a case of the plan being formed prior to a lot of other understanding about how things work and what is possible and that fact that we still have some investment in the original idea. We were also handed the pixhawk and the arduino in the kit we were given at the start of the challenge and have only really discovered the potential of the pixhawk to âdo it allâ in more recent times. plus the young people knew enough C++ to code the arduino so they have invested something in the coding and wiring of the whole thing now. There is a keen will to try and see our idea into reality but Wednesday is our deadline to make it work. Regardless we donât have any success getting the pixhawk to control the boat at all with or without arduino plugged in. Just trying to get the pixhawk to control the servo! Something in settings isnât right for us.
The rover wiki should have all in info you need to get the steering working in acro mode.
http://ardupilot.org/rover/index.html
I would recommend removing the Arduino if posible. You wind vane may be of a type that is already supported by the code directly.
http://ardupilot.org/rover/docs/wind-vane.html
You can even use a fixed wind direction or take it from a RC input. This might be a better place to start.
Fixed sheeting angle is not ideal but should work I think, just set your tacking angle quite large. SAIL_NO_GO_ANGLE of 60 or so.
Hmmm⊠Yes, in principle that sounds like a good plan butâŠ
At the moment we have issues when it comes to arming to the pixhawk at all. We get repeated error codes. Error - command rejected by Mav. these are âpre-arm 3D accel calibration neededâ and âpre arm hardware safety switchâ.
We donât have a safety switch plugged in. in fact we donât have one at all. We thought weâd got round this by setting the BRD_SAFETYENABLE value to 0. Do I need a safety switch?
We are trying this on acro mode. We are also unsure whether we should have servo 3 set to throttle or disabled given the fact we arenât aiming to use the sheeting servo.
Can you think of any other reasons that i canât arm the pixhawk?
no need to use a safety switch
You need to do a aceleromiter cal, this can be done using mission planner, Accelerometer Calibration â Rover documentation
You will have to have either throttle or a sail configured as one of the outputs, the code knows it needs some way of making progress. I suggest you configure the sail output, you donât have to actually plug anything in, just have it setup.
1 Like
So,
Found a safety switch, 3D calibration success. succesfully armed, mission started. No servo response.
In the servo output page, the ground steering value was suggesting it was trying to make a turn however the servo was not moving.
Wiring issue? Could it be anything else?
If the green bar is moving and you have the servo plugged into that channel then it is most likely a wiring issue, are you powering the servos from a separate bec?
no. all coming from the same power source. Currently as we are dry testing we are just using a portable power pack with separate 5v usb outputs to power the pixhawk and arduino units. We know that the arduino gets sufficient power from this to power the servo no problem.
We are just going through and disabling a load of fail safe actions atm as well as we have had a failsafe warning showing.
Also have directly wired the servo into the pixhawk to test. Arduino and relay no longer in the picture (for now!)