Very cool, I already built a prop-based autonomous sailboat, but it was very small (only tested it in ponds, because the wind sensor topmast made it very susceptible to waves, and I couldnt afford a bigger hull!)
Would you find any use for my documentation and/or code? We later used it for a motorized autonomous survey boat.
I'd think the whole sail assembly would work wonderfully as a wind direction vane. Of course, corrections would need to be made for flap deployment and the position of the tail. www.usdigital.com/products/encoders/absolute/rotary/shaft/ma3/ would be an appropriate encoder to use.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Lunch cures all problems! have you had lunch?
Lawson said...
I'd think the whole sail assembly would work wonderfully as a wind direction vane. Of course, corrections would need to be made for flap deployment and the position of the tail. www.usdigital.com/products/encoders/absolute/rotary/shaft/ma3/ would be an appropriate encoder to use.
totally right!· Thank you!· That super simplifies the mast head fly design and gives us a very acurate reading of wind direction at the same time.· Since we will probably only be using 3 different positions for the sail (port tack, starboard tack and centered)· we can measure the three offsets and get wind direction easily.
Great addition!· Oh yeah, you·KNOW we will be using that encoder!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ ·- Ouch, thats not suppose to be hot!··
Michael King
Application Engineer
R&D
Digital Technology Group
I looked into using strain gages to measure the direction and magnitude of stress at the base of the mast in order to figure out wind, abandoned it because it was hard to watrproof on my small vehicle, if it helps -- would work nicely with a bigger vehicle, plus no moving parts.
Another "power" sensor for a wing or sail a differential pressure sensor. If the two pressure taps are placed opposite each other on the 'top' and 'bottom' of the wing, the pressure reading should be proportional to the lift the wing is producing. Also, 'tufting' or taping lengths of yarn to the surface of interest, is a great way to visualize boundary layer air flows. (www.ez.org/tufting.htm is a good example of this)
Marty
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Lunch cures all problems! have you had lunch?
I have admit, that really is a good idea.· If we can accurately measure the efficiency of how the wing is "flying", we can trim automatically.·
I think this will go into the sensor list.· This is not a huge addition but provides some vital data.· Question is where to put the sensors for the most accurate readings.· I suppose that test runs gauging forward movement verses the pressure readings will give us the best pressure/speed ratio.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ ·- Ouch, thats not suppose to be hot!··
Michael King
Application Engineer
R&D
Digital Technology Group
You want cheap, here·is a sailboat that you can make with (1) 4'x10'x3/8" marine plywood and less than a sheet of 1/2 or 3/4" plywood and some mohagany molding. It's a "Stich together" hull with some fiberglass tape reinforcements over the stiching. Piece of cake
Post Edited (Capt. Quirk) : 9/4/2008 4:29:16 AM GMT
Fluid flows analysis always uses Dimensionless Numbers because dimensionless numbers can fully describe a system while using the lowest number of variables. en.wikipedia.org/wiki/Buckingham_%CF%80_theoremen.wikipedia.org/wiki/Dimensionless_number Damned powerful analysis technique. Makes everyone's life difficult by hiding intuitive relationships though.
Likely the most interesting number for sail power estimation is the Pressure coefficient. Modifying it to be the differential pressure across the sail over the velocity pressure of the air is likely most useful for power control. Measuring 1/2 * density * velocity^2 is easy, a Pitot Tube measures this directly.
Marty
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Lunch cures all problems! have you had lunch?
Another way to measure pressure and increase the performance of the sail, is to use a turbulator setup that allows air to flow from several holes on the pressure side of the airfoil to many small holes on the top side of the airfoil. Using a differential pressure sensor, you can measure the·- /·+ pressures on the top and bottom sides. I was planning on using that same setup on one of my airplanes to determine when the plane is flying, and not on the ground.
You can take that 1 step further and measure other points on the wing in between the high point of the airfoil and the trailling edge, to determine a flow seperation and a loss of wing efficiency.
Whether you make a model or a small boat, I would recomend building a foam wing covered with balsa or·thin ply. Instead of a built up symetrical wing, that will require a jig or detachable legs built·into the ribs.
Another thing to consider, Arizona lakes and the Colorado river act like they·are at a high altitude, depending on the temp, humidity and especially when there is·strong high pressure in the upper atmosphere. Testing for density altitude would be a good idea for comparing results.
I uploaded this before somewhere, you may want to look at it. It is a application for a ocean buoy that uses ultracaps to offset the battery use when solar charging is used.
Monday is wood day.· I am buying all the lumber necesary to start work on the fiberglass plug.· I want this to be reproducable.· So figured a good way to do this is create a fiberglass mold.
The UAV company i worked for has donated a Solar panel and a wire foam cutter to the project so a wind will soon follow.· I still need to get some vacuum bag stuff.· Also,· I am starting as website dedicated to this project and should be up sometime next week.·
This project have officially been named Persistant Sailing Autonomous Vehicle (PSAV).
This is finally on its way.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ ·- Ouch, thats not suppose to be hot!··
Michael King
Application Engineer
R&D
Digital Technology Group
This site will have all the images and construction details for the project as it happens.· I have got alot of support on this furum for this project and i wan to thank you all.
A website name will be bought and the appropriate DNS assigned soon.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ ·- Ouch, thats not suppose to be hot!··
Michael King
Application Engineer
R&D
Digital Technology Group
I just thought of another really good reason to have an auxiliary motor. An auxiliary motor could generate backup power during long cloudy stretches. (besides improving harbor maneuverability and getting some use out of excess solar power)
What we did for ours was use a minn-kota motor, they're not very energy efficient, but they are VERY tough and can take corrosion, being left in the water for weeks at a time, etc.
I assume the vessel will be designed so that it is under human control when entering/exiting the harbor? I don't think you've mentioned anything about obstacle avoidance, and the only thing that would really work on the water is a radar setup, which would certainly drain the power quickly.
SRLM said...
I assume the vessel will be designed so that it is under human control when entering/exiting the harbor? I don't think you've mentioned anything about obstacle avoidance, and the only thing that would really work on the water is a radar setup, which would certainly drain the power quickly.
Correct!· We will be using the DIGI Xtend modules.· 900Mhz Freq hopping spread spectrum data modems for manual control.· We will control rudder and reefing of the wing (altering the lift with changes to the flap and the angle of attack).· As far as obstacle avoidance.· We are very limited in this respect.· Keeping the craft out of the shipping lanes and a radar reflector are our only real options at this point.
Lawson said...
I just thought of another really good reason to have an auxiliary motor. An auxiliary motor could generate backup power during long cloudy stretches. (besides improving harbor maneuverability and getting some use out of excess solar power)
I am using a bit of overkill on the solar panel.· It is a 150w cell.· Because I will have complete control of all of the electronics,·we will be able to switch off non-essential items in the event of the batteries getting low.· I estimate we will have over 15000maH available to us.· The propeller driving 1 servo often and two servos on each tack of the vessel should get us pretty far.·The math says that 3.5 days of sun while under load should charge the batteries fully.· On the other hand we should get about 10 days of life in pitch black.· Even on a cloudy day the panel will still produce.· I think we will be good.·
Now in the event of the batteries failing, and night falls.· the autopilot will of course shut down.· However,· the autopilots memory will be an SD card.· All of the waypoints will be stored there.· In the morning, the PSAV should continue mission.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ ·- Ouch, thats not suppose to be hot!··
Michael King
Application Engineer
R&D
Digital Technology Group
Check out Project: PSAV Persistant Sailing Autonomous Vehicle
What do you plan to do if there is a problem? Prossible problems include getting hit, the bilge filling up, low battery for an extended amout of time, no wind for an extended amout of time, failure of critical parts, and getting lost. My guess is that you'll have it send out a distress signal over the radio (perhaps with a code to describe the problem). Then what? Do you abandon the ship? Hope somebody picks it up? Or are you going to go out and pick it up from the middle of the ocean?
If you do have to pick it up, then you'll have to charter your own vessel (going away from shipping lanes has it's downside...) Perhaps one of those large floatplanes would work if you could take off the airfoil. Of course, the expense would be enormous, possibly more than the value of the boat.
That leads me to the last point: an auxilary engine would allow you to motor into the nearest port if there is a problem. You could have it either autonomous or manual at this stage.
IF we get to the point where it is in teh middle of the ocean and we have a mojor failure...· we are doomed.· Granted the cost of the· craft may be high but·the recovery would be astronomical!
Our initial design here for possible deployment is 80"· this would not accomodate a long rang motoring engine.· i know the duldrums can suck but i can't remember be stuck in them for more than a couple of days.· I am making this as redoable as possible.· the hull will be from a mold I have started to build.· I have a foam cutter for building the wing.· Servos are cheap.· the autopilot shouldnt cost more than a few hundred to make including GPS and gyros.· i may be out about $1500USD by the time I am done but just the adventure is worth that!
If it sinks... make it float better next time!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ ·- Ouch, thats not suppose to be hot!··
Michael King
Application Engineer
R&D
Digital Technology Group
Check out Project: PSAV Persistant Sailing Autonomous Vehicle
I have been working on the code alot lately.· I finally figured out how to implement tacking into the navigation routine.·
The line between two waypoints is called the Line of postion (LOP).· The boat always is trying to stay on this LOP while going between two waypoints by way of calculating our distance from this LOP. However there is 32 degrees we cant navigate into that is 16 degrees each side of the direction the wind is coming from.
When our bearing to the waypoint falls inside that 32 degrees we cant sail in, we have to tack or, for lack of a better word, zig-zag toward our waypoint.·
This example is assuming our next waypoint is directly into the wind.
When we reach a waypoint, we capture the reverse bearing from the new waypoint to the old waypoint.
Start to tack 16 degrees off the wind (with a wing we can tack as close as 15 degrees of the wind)
Calculate our position and distance from the·current waypoint
Using our distance from the·current waypoint and the reverse bearing, calculate where on the LOP we should be (coordinates).
Calculate the distance from our current position to the position calculated in step 4.
If that distance is greater than the Max_Dist_LOP, go to opposite tack (turn to 16 degrees off the wind in the other direction).
check to see if we can navigate directly to the waypoint with this heading. No: Step 3
This should keep us navigating correctly into the wind.· When is goes into tack mode, it stays there until it can nav directly to the waypoint.· If the wind changes and the boat can navigate directly to the waypoint, the boat will then try to get on to the LOP again.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ ·- Ouch, thats not suppose to be hot!··
Michael King
Application Engineer
R&D
Check out Project: PSAV Persistent Sailing Autonomous Vehicle
Comments
Would you find any use for my documentation and/or code? We later used it for a motorized autonomous survey boat.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
http://forums.parallax.com/showthread.php?p=650217
meow, i have my own topic now? (sorta)
Post Edited (M. K. Borri) : 9/2/2008 8:02:30 PM GMT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Lunch cures all problems! have you had lunch?
Great addition!· Oh yeah, you·KNOW we will be using that encoder!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Ouch, thats not suppose to be hot!··
Michael King
Application Engineer
R&D
Digital Technology Group
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
http://forums.parallax.com/showthread.php?p=650217
meow, i have my own topic now? (sorta)
Marty
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Lunch cures all problems! have you had lunch?
I think this will go into the sensor list.· This is not a huge addition but provides some vital data.· Question is where to put the sensors for the most accurate readings.· I suppose that test runs gauging forward movement verses the pressure readings will give us the best pressure/speed ratio.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Ouch, thats not suppose to be hot!··
Michael King
Application Engineer
R&D
Digital Technology Group
Post Edited (Capt. Quirk) : 9/4/2008 4:29:16 AM GMT
Likely the most interesting number for sail power estimation is the Pressure coefficient. Modifying it to be the differential pressure across the sail over the velocity pressure of the air is likely most useful for power control. Measuring 1/2 * density * velocity^2 is easy, a Pitot Tube measures this directly.
Marty
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Lunch cures all problems! have you had lunch?
You can take that 1 step further and measure other points on the wing in between the high point of the airfoil and the trailling edge, to determine a flow seperation and a loss of wing efficiency.
Whether you make a model or a small boat, I would recomend building a foam wing covered with balsa or·thin ply. Instead of a built up symetrical wing, that will require a jig or detachable legs built·into the ribs.
Another thing to consider, Arizona lakes and the Colorado river act like they·are at a high altitude, depending on the temp, humidity and especially when there is·strong high pressure in the upper atmosphere. Testing for density altitude would be a good idea for comparing results.
Hull length 80"
Sail heght 90"
wing chord 14"
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Ouch, thats not suppose to be hot!··
Michael King
Application Engineer
R&D
Digital Technology Group
The UAV company i worked for has donated a Solar panel and a wire foam cutter to the project so a wind will soon follow.· I still need to get some vacuum bag stuff.· Also,· I am starting as website dedicated to this project and should be up sometime next week.·
This project have officially been named Persistant Sailing Autonomous Vehicle (PSAV).
This is finally on its way.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Ouch, thats not suppose to be hot!··
Michael King
Application Engineer
R&D
Digital Technology Group
http://67.135.195.32
This site will have all the images and construction details for the project as it happens.· I have got alot of support on this furum for this project and i wan to thank you all.
A website name will be bought and the appropriate DNS assigned soon.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Ouch, thats not suppose to be hot!··
Michael King
Application Engineer
R&D
Digital Technology Group
www.instructables.com/id/Solar-Powered-Trike/ Source of the power-tracker used in the link. www.solarseller.com/ Looks like a power-tracker can be used with wind or hydro generators too with some precautions.
Marty
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Lunch cures all problems! have you had lunch?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
http://forums.parallax.com/showthread.php?p=650217
meow, i have my own topic now? (sorta)
I am using a bit of overkill on the solar panel.· It is a 150w cell.· Because I will have complete control of all of the electronics,·we will be able to switch off non-essential items in the event of the batteries getting low.· I estimate we will have over 15000maH available to us.· The propeller driving 1 servo often and two servos on each tack of the vessel should get us pretty far.·The math says that 3.5 days of sun while under load should charge the batteries fully.· On the other hand we should get about 10 days of life in pitch black.· Even on a cloudy day the panel will still produce.· I think we will be good.·
Now in the event of the batteries failing, and night falls.· the autopilot will of course shut down.· However,· the autopilots memory will be an SD card.· All of the waypoints will be stored there.· In the morning, the PSAV should continue mission.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Ouch, thats not suppose to be hot!··
Michael King
Application Engineer
R&D
Digital Technology Group
Check out Project: PSAV Persistant Sailing Autonomous Vehicle
If you do have to pick it up, then you'll have to charter your own vessel (going away from shipping lanes has it's downside...) Perhaps one of those large floatplanes would work if you could take off the airfoil. Of course, the expense would be enormous, possibly more than the value of the boat.
That leads me to the last point: an auxilary engine would allow you to motor into the nearest port if there is a problem. You could have it either autonomous or manual at this stage.
Our initial design here for possible deployment is 80"· this would not accomodate a long rang motoring engine.· i know the duldrums can suck but i can't remember be stuck in them for more than a couple of days.· I am making this as redoable as possible.· the hull will be from a mold I have started to build.· I have a foam cutter for building the wing.· Servos are cheap.· the autopilot shouldnt cost more than a few hundred to make including GPS and gyros.· i may be out about $1500USD by the time I am done but just the adventure is worth that!
If it sinks... make it float better next time!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Ouch, thats not suppose to be hot!··
Michael King
Application Engineer
R&D
Digital Technology Group
Check out Project: PSAV Persistant Sailing Autonomous Vehicle
The line between two waypoints is called the Line of postion (LOP).· The boat always is trying to stay on this LOP while going between two waypoints by way of calculating our distance from this LOP. However there is 32 degrees we cant navigate into that is 16 degrees each side of the direction the wind is coming from.
When our bearing to the waypoint falls inside that 32 degrees we cant sail in, we have to tack or, for lack of a better word, zig-zag toward our waypoint.·
This example is assuming our next waypoint is directly into the wind.
- When we reach a waypoint, we capture the reverse bearing from the new waypoint to the old waypoint.
- Start to tack 16 degrees off the wind (with a wing we can tack as close as 15 degrees of the wind)
- Calculate our position and distance from the·current waypoint
- Using our distance from the·current waypoint and the reverse bearing, calculate where on the LOP we should be (coordinates).
- Calculate the distance from our current position to the position calculated in step 4.
- If that distance is greater than the Max_Dist_LOP, go to opposite tack (turn to 16 degrees off the wind in the other direction).
- check to see if we can navigate directly to the waypoint with this heading. No: Step 3
This should keep us navigating correctly into the wind.· When is goes into tack mode, it stays there until it can nav directly to the waypoint.· If the wind changes and the boat can navigate directly to the waypoint, the boat will then try to get on to the LOP again.▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Ouch, thats not suppose to be hot!··
Michael King
Application Engineer
R&D
Check out Project: PSAV Persistent Sailing Autonomous Vehicle
Lawson
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Lunch cures all problems! have you had lunch?