Radio controled torpedo experience

[jlcrl]

New video big pool maneuverability test

long video
https://www.youtube.com/watch?v=IlGAJr0OdNQ
short video
https://www.youtube.com/watch?v=pZF_BPKKS84
https://www.youtube.com/watch?v=vHQ3vDcRdzw

[trout]

Outstanding! Fun to watch them run!

[jlcrl]

Nick,
obviously the stern must becomes heavier and the bow alleviate.
After removing the back foam the torpedo balances like this
https://youtu.be/I3fywe7hE5k
but once the propulsion is on, it is balanced horizontal with tail slightly under water.
the back ring acts as a stabilizer, making the stern up when speed increases.
But stern must be hold slightly under water.
Otherwise bubbles appears and makes propulsion and direction inefficient with poor speed.

After 7-8 runs I measured the battery voltage at 4V at home.
As the 5:1 gearing dramatically decreases consumption, a smaller battery should be considered.
I got a 70 maH 14C 1A max and a 80 mah 15C 1.2A max but the best is probably 100 120 maH 15C difficult to find actually.

For the test I was using the 1 magnet rudder.
I observed a trend to deviate on the right, possibly the rudder is not exactly vertical but with a slight angle ?
At 50% throttle the torpedo is easily maneuverable but near full speed it is limit to keep it straight.
For these reasons I will put back the 2 magnets rudder and remove some sand from the bow tank.

The good news: the radio control works perfect.

The ultimate state would be to have a torpedo running just some cm under water
see this:

https://youtu.be/balkc5xeAoI

But with the risk is to lose it as the balancing would be near neutral buoyancy.
I think to make a non RC model (timer controlled) to study this.
Also a small tank at the bow: tube, piston inside and magnet inside piston.
At neutral buoyancy the tank is full; when the timer stops the motor, an electromagnet just in front off the magnet push the piston for
some seconds and the torpedo surfaces.

[He Who Shall Not Be Named]

Most impressive!

[Monahan Steam Models]

Very good progress! Nice work!

Seems it runs much better without the polystyrene. Is the polystyrene absolutely needed for the turns?

[jlcrl]

The video links (first tuning, to be improved)
Straight direction (the 3 first) and turning low speed (torpedo unbalanced there)

keep straight, full power after tuning (back polystyrene removed)

https://youtu.be/lwJhXmxOSjs

other 'keep it straight' test

https://youtu.be/9Pu2hfqX9s8

https://youtu.be/k4R6-TUDAg8

turn test low speed ((back polystyrene not removed, bubbles avoid to reach full speed)

https://youtu.be/3xg1GJwoRyA

https://youtu.be/vPsSInrKydM

[jlcrl]

Torpedo is now balanced.
Moving the lipo back may increase sinking tail angle.
Under HF zone, a tank with aquarium sand inside.

Ejection system is a Teflon sheet inside 3D printed tube, holes to prevent suction cup effect.
( Torpedo body is coated and sanded 4 times until it looks like a mirror. )
Or (better) just 4 steel rails. This minimize frictions.
Then the torpedo gently ejects at 1/2 throttle by radio control without anything else.

[jlcrl]

So I think to solve like this to achieve the first prototype.
Magnet servo is not inaccurate but differs from conventional one. There is no neutral return. PWM into the coil wire change the magnetic flux,
making less or more pressure on air or water. The first tests into water made with 72 mhz torpedo version were positive.





About the BLDC version:
When I was playing with RC micro planes I used to make my radio control, ESC and motors myself.
You can see a range of motorizations usable for the torpedo and the esc 11.5x12.75x0.5 mm
This one has the ability to start any micro motor as the start up technology is not a ramp but electric signals monitoring.
The last 1s 2s 3s version was used by 3D indoor aerobatic pilots and they won 4 times french championship and a podium on world championship with it.
I made a modified circuit for ESC version to be tested. As I own the source code I can adapt as I wish.
The 2g motor I made is 13mm OD so very limit for the torpedo (inside tube 13.60mm)
Efficiency is 5-6g thrust per volt near 20g thrust for less than 1A in the air.
The chinese motors are 9 and 10 mm OD but low efficiency (2 times less) so drawing very high current for 20g thrust.
They need to be rewind in Y with thinner wire and more turns before use but I have a doubt not getting more than a conventional brushed motor
as the way they are built don't give really good torque. To be tested.
JL

[Bob Gato]

Or a variation of that: the outrunner housing turns one screw, the motor mount turns the other (pretty much how the MK 37 torpedo works). One motor providing perfect torque matching between the two counter-rotating propellers. However, this negates the walking-propeller steer unless you outfit the motor case and motor mount with breaks. This, of course, necessitates the use of slip-rings to get power to the rotating motor mount.

David TM2 (SS/DV), retired

Good idea-you'd get 1/2 the rpm from each prop- but there's plenty to go around with an outrunner, it would make a great straight electric runner without the gears...I threw out the prop walk steer idea in homage to the KISS method-and the KILL (keep it light Lucy) method-because the thread is for RC fish-that's -a tough one... If one could push the shaft out of a 10mm outrunner (without destroying it) and replace it with a tube, the coaxial thing would be easy -you might be able get away with using one ESC- being that small-timing might not come into play that much-if you wanted to steer it at that point, just electrically cut one motor or the other with a nano RC switch of sorts.

[Monahan Steam Models]

The run time is adjustable. For testing I chose 7 seconds as I didn’t want to be chasing them around. Yes there is a limit to the battery’s discharge rate for LiPo’s and many manufacturers post incorrect data.

Battery capacity was oversized in my tests to provide multiple runs at the preset time before charging or overload limits were experienced. Motor speed vs current draw and time was solved by changing the friction in the transmission and pitch of the propellers. This works well for the design of a non rc torpedo.

RC torpedoes are a different challenge but same ballasting is required if you want to reuse them. I use computational fluid dynamics often in my day job as an engineer to test efficiencies of new designs.

Early tests I did in the prototypes I was working on had the same characteristics in regards to floating warhead up out of the water with tail sunk. What I was failing to grasp was the difference between CG and BC.

As slick as the coil based rudder control is, you’re right, it’s too heavy. I’ll add not precise enough. I see a set up like this being very twitchy in real world use but I like were you’re going with it.

Nick

[He Who Shall Not Be Named]

Sorry-I am late to this party ..they were moving a yacht on the expressway...OK- what if you dispensed with the gearbox and ran axial contra- rotating brushless motors now all you have to do is control the sync or un sync of the motors to steer with side force/prop walk. Eliminating the gearing should lighten the load to where you now can control buoyancy with ballast.

Or a variation of that: the outrunner housing turns one screw, the motor mount turns the other (pretty much how the MK 37 torpedo works). One motor providing perfect torque matching between the two counter-rotating propellers. However, this negates the walking-propeller steer unless you outfit the motor case and motor mount with breaks. This, of course, necessitates the use of slip-rings to get power to the rotating motor mount.

David TM2 (SS/DV), retired

[Bob Gato]

Sorry-I am late to this party ..they were moving a yacht on the expressway...OK- what if you dispensed with the gearbox and ran axial contra- rotating brushless motors now all you have to do is control the sync or un sync of the motors to steer with side force/prop walk. Eliminating the gearing should lighten the load to where you now can control buoyancy with ballast.

[jlcrl]

The problem is that the torpedo is vertical noose up except removing the 0.2g rudder and then it is as you describe horizontal 2-3% tail down but no more an RC torpedo without direction control...
These 0.2g are just at the end of the tail, not easy to remove.
Now for the motor / battery / gearbox the problem is different for an RC torpedo. The torpedo must be like a little submarine, capable to follow its target and to come back to pilot by radio control.
This means runs duration much more than 4-5 seconds. Of course the non RC torpedo may run direct drive with 1C 2C battery, but they are into a battery and motor overloaded situation.

the picture shows what happens when you overload the battery. In my sense it may works some seconds but doing it 2-3 times without recharging the battery and the protection circuit cuts.
So you see it is a "25C" 150 mah battery discharge test but the seller lies because the battery is in reality 10-15C.
So at 25C discharge (blue line) the battery works some seconds before to reach the forbidden voltage. But at 10 C it is quite good. For the 1C-2C battery when changing from 60 to 90 mah you just
give some more seconds before cutting.






Now for the motor / gearbox you have the full simulation of the torpedo and different propellers on this link

https://www.ladsm.fr/viewtopic.php?f=7&t=22

Simscale offers a large amount of free server CPU time and their engineers gently helped me to make the (difficult) setup
Templates are free for copy to anybody wanting to simulate the propellers
the speed was set to 7 km/h (green zone) and other colors is the drag to be removed.
from the torque and thrust computed I finally draw the efficiency bell curves.
It should be around 7000 rpm but the 15mm length 7mm diameter drone pager motor is running possibly 50000 - 60000 rpm no load.
So before the top of efficiency curve the gearbox ratio is too much and the motor "mills" with low speed and few current drawn.
Passed the top the motor force as it cannot supply sufficient torque. Then the electric input power P=U*I is not converted into mechanical power
but into Joule effect. The motor heats and the current drawn is high and the battery voltage falls to compensate too high current.
These results as given "as it" and I may be mistaken bus I suspect the best efficiency not to be direct drive but perhaps 2:1 or 2.5:1 gear ratio.
To be sure an eddy current dynamometer should be used to verify the propeller torque but this is not the purpose here.
Just to explain what I am doing.
JL

[Monahan Steam Models]

The work I did last year with the 1/48” scale G7e electric torpedoes resulted in the final trimming of them exactly as David has described. They’re just slightly positive buoyant at an angle of a degree or two nose up tail down. When launched they track extremely straight for the duration of the time they are set to run.

The gearboxes for the counter rotating shafts can be tricky. Although the first one I designed worked, it did take many small revisions to get a design that had barely any friction. I took a similar design path that you are exploring currently. The final design I arrived at was 6mm x 15mm coreless motor directly driving the counter rotating gear box.

I started work on some 1/35” G7e scale electric torpedoes that will be using a 10mm od brushless outrunner motor. These are intended for a type XXIII coastal submarine project.

Nick

[He Who Shall Not Be Named]

I made the coil, mounted all and put the torpedo into water.
Despite using light materials, the stern has a negative buoyancy and sink. Removing the 0.2g rudder + magnets the stern buoyancy is slightly positive.
The bow has 4 - 6g positive buoyancy.Changing lipo position or adding weight under the bow don't solve.
Does anyone have the experience of sinking tail and the behavior when the torpedo running ? I found a post about this but cannot find it back.
Gearbox tests at different RPM shows some weakness in the concept, referenced by the numbers on the picture
1) the 681 bearing is only 1.2 mm width and probably ABEC-1 and the crown gear base can take some angle with horizontal, making friction with the spur gears.
2) alu stern bushing grease is required as alu on alu friction coefficient is bad. This grease goes away fast, making frictions and more current drawn.
3) inserting / removing propellers is a tricky problem. Gluing them may solve but then I must break propellers to unmount the gearbox




I printed a new gearbox, changing the pinion system to solve.
Machined Teflon bushings have larger wall, solving the mechanical problem of thin tube and does not require grease.
Removing the bearings save some weight.

You want a near-neutral buoyancy overall (slightly buoyant) with the trim having the tail pointed down. You want the weapon to skim the surface during the run. Sounds like you're already there. Nice graphics.

David