Adding pitch control to a model sub.

[cstranc]

Lets face it. Submarines are big and powerfull, they are elegant in the water.

Last year I built my first model sub. While on the surface it was ok, it could have a hard time keeping a nice gracefull dive angle. So I thought about using the same APC that controls the dive planes to move a weight within the hull correcting the pitch automatically.

It seemed like a good idea, so I did some experiments and submitted an article to the SCR.

Of course a few experiments do not really answer the question "will it work" they just point the way. So since I finished that article I have been working on a 1/60 Sierra I. This model will be my real live test bed for implementing pitch control.

I have a seperate thread covering the building of the Sierra. In this thread I would like to go beyond what was in the SCR article.

[cstranc]

First a quick recap of

First a quick recap of what was in the SCR article...

My first prototype was nothing more than a small tub floating in a big tub. Inside the small tub a lead weight was on a trolley, and that trolly would move from sided to side based on the signal from a APC (Automatic Pitch Control) See http://www.mikessubworks.com/subs_only.html for a sample APC.



This simple prototype worked as expected. If I moved the stick on my RC transmitter forward the APC would sense it was not at the right angle, and start moving the servo. The servo would move the trolly, and pretty soon the tub had adjusted to the desired angle. Cool enough.

But let's face it this little tub looks like a "surface vessel" and not one of our elegant guardians of the deep. So it was time for prototype 2.



The idea was to get closer to a sub, that meant everything was going under water. The reciever, battery, APC and servo moved into a tupperware container. The container was mounted on some aluminum bars with variable bouyance devices (bottles), and there was the trusty lead weight on a trolley.

Let the testing begin. My first tests the bottles were full of air making the prototype positively bouyant. And it reacted pretty much the same as the tub. The prototype would gracefully adjust to the requested dive angle.



But then came the really important test. What happens when we remove air from the bouyancy tanks and make it neutrally bouyant...


Hmm, not so good. Instead of going to the desired angle it would overshoot in one direction, then overcorrect in the other, ... It would rock back and forth. Not quite the desired effect. Why was it happening?

Well it turns out that a boat on the surface is very different that a sub under water. Let's start with a few simple points.

The "center of mass" is like the midpoint for all the mass / weight in the boat. There is the same amount of mass to the left of the center of mass as to the right. If you held something by it's center of mass it would be nicely balanced, and not rock to one side or the other.

The "center of bouyancy" is the midpoint for all the floaty stuff. For us that would mean the ballast tank, any floatation foam, and closed container with air, etc. When you are thinking about the center of bouyancy only the floaty stuff that is underwater actually counts.

Now any boat will always position itself so that the center of mass is directly below the center of bouyancy. If they do not align, the boat will rotate to align them.

So now let us look at what caused our submerged prototype to be so twitchy...



We start off with a nicely ballanced sub on the surface. Then we shift a little weight towards the front of the sub. As expected the front of the sub dips down into the water and a new ballance is achieved. Looking more carefully...

When we shifted the weigth forward this moved the center of mass a little forward.

When the front of the sub went down more floatation was submerged creating more floatation at the front. Also the stern rose up a little removing floatation / bouyancy from the rear. The result it the center of bouyancy moved forward also.

The sub came to a new balance point when the new center of bouyancy caught up to (was directly above) the new center of mass.

It is important to think about that change in volume of floatation at the front and stern as we move our weight around...

So now what happens when the sub is submerged. Well first, all the floataion is under water already. So there is no change in the location of the center of bouyancy. It is fixed in one spot. Period. End of story.

When we move our weigth forward the center of mass still shifts forward, but the center of bouyancy cannot move so the sub simply rotates until the center of mass is again below the center of bouyancy. This time the center of bouyancy does not shift forward because no more foam is being submerged as the sub rotates.

I hope that's clear.

And that's basically where the article ended. Maybe with a 66" hull there will be enough water friction and inertia to prevent the rocking back and forth. Hard to tell without building one

[cstranc]

Now enter the Sierra I.

Now enter the Sierra I. Here is the chance to see if we can make a really smooth dive even when stationary (no help from the dive planes).

During the article I always talked about moving a lead weight around to adjust things. It seemed easier then trying to move a floaty thing (like the ballast tank). Hmm but I was anticipating the problem that I would have to move the weight a long way to change the angle while I am on the surface, then when I am submerged it get's all sensitive and I don't want to move it much at all. It felt like I had to either get it nice for the surface operation, or submerged, but not both.

Then I thought about moving the ballast tank instead. On the surface the ballast tank is full of air so when it moves it moves a lot of floatation around, and that can give me the desired control. When submerged the ballast tank is mostly empty. So it does not move much floataion around...

Hey that sounds good. So in the Sierra I planned on moving the ballast tank and getting good control on the surface and while submerged. I like it.

Well first thing is to build the ballast tank. This is a Russian sub, so it has lots of reserve bouyancy. It floats just shy of the centerline. Now how big should the ballast tank be? Time to test.

I knew the outline of the ballast tank (slightly smaller that the hull, and does not rise above the water line). So I cut a few foam blanks out of 3" and 1" foam. Then I figured the ballast tank would need to lift about 1/2 the hull out of the water, plus some control surfaces, plus this and that...

So I made a little lake in the basement (when everyone else was asleep), dropped 1/2 the hull and a bunch of other weight in, then I started pushing the foam blanks under the hull until there was enough foam in the water to lift the hull (and other weights) clear. I figure this is about the volume I need.



Then I dried of the foam peices, and glued them together to form the blank for the ballast tank.



You can get the idea how the model will be layed out in the following shot...



Working front to back...
- Peice of test hull (not real hull) where I have been playing with retractable dive planes.
- RC water tight container. Made from plexiglass (thanks again Dad).
- Ballast tank. This is mounted on a track to allow it to move.
- Copper air / propell tank.
- Battery container
- Motor

I thought there was going to be a huge space inside the hull, go figure I have filled it already.

Now in the article I move the lead weights around using string wrapped around a pulley. Yuck, That just did not have the "I'm well engineered" feeling. I also noted that you should design for the day when your weight / ballast tank hits the end of the track. With the string I thought it would just start slipping around the pulley.

But I still was not happy about it.

It's funny how your perspective can change. I see Home Depot, Canadian Tire, MEC, all my favorite stores as places that just don't realize how much potential submarine merchanise they sell. For example. There I was walking around in Home Depot when I spy a threaded 1/4" stainless steel rod. Now the fact that it was stainless steel is a dead giveaway that it really should be placed into a submarine someplace. What they had was a in-expensive lead screw that I could use to move my ballast tank!

This is what it looks like when it has found it's true purpose in life.



The track will fit into the hull. I have it coupled to a 12VDC motor (from Princess Auto, another store that does not appreciate how many sub supplies they sell).

Woah. My mouth is getting ahead of the technical stuff. The pitch control has evolved a little since the article. It lays out kinda like this...

- Signal from the APC is split with one end going to the dive plane servo and the other going to a speed controller.
- The speed controller drives the motor that turns the lead screw that moves the ballast tank.

Now you have noticed that the 12VDC motor is sitting out there in the wet area. I tested it for a week with 1/2 day submerged, 1/2 day in air to see how it survived. It seems fine. and if it has a problem it's $6 and I have spares.

You can see a nut on the lead screw. The ballast tank will engage this nut and it will move the tanks back and forth.

This is a good time to review that "what happens when it hits the end of the track" scenario. I though of having micro switches and all sorts of things to prevent the ballast tank from plowing into the end of the track. In the end I decided to go for simplicity. I shall file away the tread on the lead screw at the end points. This way when the nut gets to the end point it will stop pulling. I will use a little spring to push the nut back onto the thread so it can re-engage and go in the opposite direction.

At least that's the plan now.



Here you can see the ballast tank mounted on the track and in position.
The track is 1/4" aluminum C channel. The ballast tank has brass plates on either end, and these brass plates have 1/8" rod that engages the C channel to hold the ballast tank in place.



Here you can see the front of the ballast tank. If you look carefully you can see the nut sandwiched between two brass plates.

Inside the ballast tank you can see the ballast blow assembly.

Because the ballast tank moves I cannot have mechanical linkages to blow the ballast tank. Instead I need to build a valve assembly with a soleniod. Then the wires from the solenoid can sit in the hull as the tank moves back and forth.

I beleive it is normal to locate the propell storage tank in the center of the hull. Unfortunately my ballast tank will be wandering though that location so I shifteed the propell tank further back. I am assuming that the pitch control system can compensate for changes in the weight of the propell tank...

One other little side note. When the sub is submerged the ballast tank must still be slightly positively bouyant so that it can be used to adjust the pitch. If the ballast tank was negatively bouyant it would not work at all.

[himszy]

That's a good setup but

That's a good setup but IMO there's a certain degree of overkill there...

Let's say I'm got the same setup as you and I've finished my sub and it's in the lake awaiting to be trimmed. I pump some water into the tank so it's decks awash and then look at my level. I noticed the bow is a little high and so move the ballast tank forward slightly, surely as I'm moving the tank forward the water in there is going to be sloshing around inside the tank being thrown against the walls and creating forces in all sorts of directions affecting my trim?

Therefore, would you not be better to move the battery instead of the ballast tank? It would also make connections easier as wires (for the battery) can be moved easier than tubes (containing air/water for the ballast tank)

Also I doubt you need 30cm+ of travel and a motor with an ESC when it seems some folks in Germany do it with a high torque servo and 5cm of travel IIRC.

Not wanting to blow your idea out the water....I just think you may have exaggerated the trim issues people have and over engineered it slightly

Michael

[subshop]

pitch control

I read your article too and i found it quite interesting. my question is : could you wire the apc to a pump that is connected to two trim tanks at either end of the sub? start off by filling one tank completely then even out the water between the two tanks. when the pitch control needs to make an adjustment the pump transfers the water fore or aft. Just a thought! neil

[JWLaRue]

That's a good setup but

That's a good setup but IMO there's a certain degree of overkill there...

Since Chris is taking a nice methodical approach to this, I for one would like to see actual results before making any claims about overkill.

-Jeff

[mylo]

I really like what you're

I really like what you're doing here Chris. You're thinking "out of the box", designing, testing, applying, ...and could very well come up with something innovative. It's easy to tell that you have a lot of interest in what you are doing. Show us how your designs can make an r/c sub even better. I look forward to seeing what you're up to.

Mylo.

[JWLaRue]

I definitely agree with you,

I definitely agree with you, Mylo! Chris is building something and not afraid to try new ideas!

-Jeff

[salt6]

Add some baffles to the

Add some baffles to the tank to reduce sloshing.

[cstranc]

Wow, thanks for all the

Wow, thanks for all the feed back. Lets see...

Michael,
I have to agree with many of your points. I think that moving that ballast tank (especially that far) is overkill. I have been pondering this for some time now. As noted above the sub gets more sensitive when all the floatation is submerged. Hmm. And Dave Merriman's video tells us to ensure everything is below the waterline (when surfaced, except the hull of course). So all my components are under the waterline. That means this model should be quite sensitive while running on the surface too (as long as it does not start lifting floatation out of the water).

So with this tought in mind moving that ballast tank sure seems like overkill. I am guessing that I will set it up. Test it for a while. Write up the results here, and then... Decouple the ballast tank from the track, fix it to the hull and add a weight to the lead screw instead. Play with that for a while... See what happens.

All this is one of the reasons I chose the Sierra. At 1/60 scale I get lots of space to play with things like this.

The idea of using a servo to rotate the lead screw sounds pretty good too. Who needs the cost & space of another APC. When I get a better idea how fast the ballast tank / weight needs to move maybe it will be time to do a conversion.

When I mentioned that the ballast tank must be positively bouyant I should have been a little more specific. The tank is meant to be run either full of air or water. To make it bouyant I was thinking of adding foam so that it was stable.

Neil,
I like the idea of a pump and forward and aft tanks. I beleive that's what a real sub does. Two things spring to mind. The pump should not be one of those impeller pumps, or else when the pump stops the water will just go back where you don't want it. Unless you use a check valve. But then that gets complicated. I think a small gear pump would work it should prevent the flow when it is not running.



I have to thank you all again for your feed back. I was really worried that people might read the SCR article, think I knew what I was talking about, and then put it in their model only to find it does not work quite as expected.

That's why I really wanted to build the Sierra and start this thread. It's an idea in progress. I have to agree that I tend to over engineer things. Hey, I'm an engineer by schooling. It comes naturally for me

It's not the destination, I love the journey.

[Guest]

I read your article in

I read your article in the report with some interest.

I've seen a couple of boats with built in ballast shifting.

The vast majority set-up the main battery as the 'weight', with the shifting to be actuated via a servo. Nothing fancy there.

This works very well, and only a small range of movement is evident in the boats I have seen fitted with this. Maybe +/- 3/4".

Clearly the metacentric height of the boat in question will have a large bearing on how effective ballast shifting is.

Using an APC for automatic ballast shifting seems like it could cause problems when used in conjunction with an APC on the hydrovanes. I think a relay to switch out the APC for ballast shifting when under way would be advisable.

Andy

[bob the builder]

I didn't read through in

I didn't read through in great detail, but I did get the overall gist of what you're trying to do. One though that I've had in this regard is that the threaded rod approach may be too slow to effectively manage trim control.

From watching my APC work in my Nautilus, the planes are working constantly, especially when the sub is running fairly level. I don't think the threaded rod will be able to move the center of buoyancy fast enough to effect a useful degree of control.

I could be wrong. I've never tried doing what you are, but I have thought about it, and wondered at the speed of the weight...

Good work! Keep up the out-of-the-box thinking!

[Rogue Sub]

If your curious how this

If your curious how this is done in the real thing.

The boat as a main trim pump that is an impeller tip with a smaller impeller booster pump to maintain the header filled with water. Then water can be shifted by a valve manipulation to aligh the to and from tanks.

I geuss you could achieve this in a boat as large as yours if you wanted to get way over kill. Im geussing you could us a bow thruster and a few solenoid valves working in opposites to switch the flow path. Of course youll need more channels. Maybe something for those of you that like over engineer.

Great idea you have. I check in on this thread everyday just to sea how its gonna work out. Peaked my curiosity.

[subshop]

PITCH CONTROL

CHRIS, I was thinking about using a fuel transfer pump made by Slimline. they make products for r/c cars and planes. it is a peristaltic pump and works on 12v. see here http://www.slimlineproducts.com/online_ ... r_pump.htm
neil

[Guest]

If transfer of ballast is

If transfer of ballast is what you are after, it's very difficult to beat twin piston tank and servo feedback control of each tank.

Using a pump it is difficult to accuartely access how much water is in each tank, plus I would think if using an APC, a wider window of hysteresis would be required to prevent the pump constantly hunting back and forwards.

The way I see it, dynamic control of the boat is best left to hydrovanes or thrusters, unless the boat is a very slow moving vehicle indeed (i.e. a Brandtaucher).

However weight trimming is very valuable for static trimming of a boat.

Andy