Adding pitch control to a model sub.

[cstranc]

Ok, So I have realised

Ok, So I have realised the first thing about trimming a sub. It takes a while, and it may well be part art, and part science.

So I took a little time off the Automatic Picth Control thing and went back to look at the 23 pages of topics on the R/C modeller area for information on trimming a sub. I really expected to find lot's of questions and details how to do this, because everyone has to do it for every sub.

Deep down inside I was annoyed that my sub did this little listing thing as it went through it's dive. It started level, it ended level, but half way through it did this:



So I figured this time was the best time to experiment with trimming a sub.

Now my theory went like this. There was foam above the waterline, and there must be a little more foam on one side then the other so as it dives you get a little more bouyancy on one side for a while, ... Sounded semi-logical.

So I marked the waterline on the foam.


Then I cut off the offending foam and moved it down below the waterline.


Then it was back into the tank. No change. It still had this miserable list to one side half way through the roll.

Ok, so the other thing that was really annoying me was that towed array pod. When I was testing earlier it was definately throwing things off. While diving it would not fill with water, so it would float the back end. Then when it had finally flooded it would not drain quickly so the sub tried to lift the weight of the water and that caused even more listing.

Off to the drill press for some strategic holes...

Enlaged air vent.

and a place on either side for the water to enter / exit. I shall try and make these less obvious later...

Now the modifications to the towed array had a visible benefit to the submerging and surfacing. But they certainly did not remove the listing.

So what to do next. Well that was obvious. Throw 3lb more weight on the keel, and lift the foam higher in the hull. That's gotta fix things.

To raise the foam I decided to fit a plexiglass plate into the hull. I would attach the foam to the plate and the plate just below the centerline of the hull. Now I get foam higher in the hull.

This is the plate:

and here you can see it installed in the rear of the model with a sheet of 1" foam mounted on top. The additional weight is on the hull beneath the foam.


Did this fix things? not really. Yes, they were a little better, but now "wow I'm happy with that".

So what next...


oooh that's ugly. Front and back I modified the foam to reach the under surface of the top of the hull.

So I had always thought that foam above the waterline is counter productive. It is above the water line, so it does not provide any bouyancy when you are on the surface, AND you need a bigger ballast tank to take on extra water just to sink the foam too. Counter productive.

But it worked perfectly. A couple of days later I found a post recommending a chunk of foam fore and aft above the waterline. It is worth increasing the size of the ballast tank.

So what's going on (I think). Well it raises the metacentric height. That's always a good thing, because it means there is a stronger force keeping the hull vertical. But something else was happening, and it starts with the air leaving the ballast tank...

If we released a cubic inch of air, this would translate into about a cubic inch of foam needing to submerge so we got a nice predicatble dive. Because the foam was located on the centerline of the hull it also generated a stronger force to keep the hull vertical.

So the things I will be pondering as I finally trim the hull.
- How can I get as much foam on the center line of the hull and as high as possible?
- I will have two columns of foam touching the underside of the top hull. These will be located the same distance from the center of gravity / bouyancy to give an flat dive.
- It is safer to significantly oversize your ballast tank, because it means you can have more of this "over the waterline" foam. Besides you can always reduce the effective size of a ballast tank by putting foam into it.

One last thing about this foam. When the prop starts spinning it will generate a torque on the hull. The greater the distance between the center of mass (lotsa lead on the keel) and center of bouyancy (move foam up) the less the sub will roll on it's side. That assumes you have a single prop. If you have two counter rotating props forget this paragraph.

I believe that foam rising vertically above the centerline is far more effective than the same volume of foam lying horizontally. I am not sure why. I am sure the answer is in a text book some place. Or probably half the people that read this post already know it...

Well, I'm off to the cottage this weekend. That will give me the opportunity to make the weights for the later part of testing. Also before I do too much more testing I need to make a way to switch the radio gear on and off. I think that will be critical when I hook up the APC and the ballast tank gets a mind of it's own.

[Guest]

Increasing the metacentric height will

Increasing the metacentric height will make your boat more stable, but be careful of going too far.

An overly stable boat will be a pig to keep dynamically level under water, especially if you only have stern mounted hydrovanes, as the static stability will fight the hydrovanes attempt to tilt the boat.

If you have a boat with thrusters, instead of hydrovanes, this isn't an issue and you can make your boat as statically stable as you want.

Andy

[anonymous]

Hi Chris,

You raise some very

Hi Chris,

You raise some very interesting and thought provoking ideas.

I have been having a discussion recently with Andy. We were looking to minimise the volume of the ballast tank and we had decided that the best way to do this was to minimise the displacement of the superstructure.

We saw the way to do this was to use the thinnest, lightest and strongest material possible for the superstructure.

Keeping it light would also help lower the metacentric height.

Your point that foam is important ABOVE the waterline is very interesting and may mean that it is worth trading in some of the ballast tank volume to lose some metacentric height. Ballast tank minimisation is not everything.

I suppose it depends at what point in the dive you need the stability. I think most submarines at neutral buoyancy are rather unstable - hence the need for some forward speed - as per Andy's post.

You make the point too that vertical foam seems to work better than horizontal. Could this be a difference in turning moment (i.e horizontal distance from the centre of gravity (perhaps centre of buoyancy is best term))

I look forward to the next observation when you come back from your cabin.

Davy

[anonymous]

Sorry I said cabin I

Sorry I said cabin I meant cottage

I noticed that you were from Canada and word association did the rest.

No offence meant!

Davy (A Brit)

[cstranc]

Davy,
Of course no

Davy,
Of course no offence taken, cottage, cabin, what's in a word. The important thing is a place closer to nature, and farther from the stresses of work. We had a wonderfull time at the cottage. I shall go into the details of the sub related construction a little later.

I am really glad you mentioned

...at what point in the dive you need stability.

I am using a length of fuel line to blow and vent the ballast tank. This means it discharges very slowly. That is good to look at what is really happening. As I watch the dive the model will slowly ease the hull down (this is with foam up to the upper hull). Then it get's too the sail. There is no foam in the bottom 80% of the sail and it quite litterally drops until the foam at the top of the sail engages.

So I am really tempted to say]http://www.subcommittee.com/forum/icon_smile.gif[/img] I shall try it out soon.

Andy,
You raise an excellent point about creating a craft that is so stable that is does not respond to dive planes. I am used to thinking about the metacentric height when considering roll, but pretty much ignoring it while looking at pitch... Hmmm.

I will have stern hydrovanes under APC control, and the forward will respond directly to the stick (no APC).

It's kinda scary to think that after all this time in a 8' tank trying to get the model to work nicely she could behave like a pig while under power in the open water. But I have to admit there are many things I am not considering due to lack of experience. Please keep the comment coming.

[cstranc]

So before I went off

So before I went off to the cottage I did get a little time to play with the model. You may recall I have my ballast tank travelling 7" allong the hull. My first trial this 7" motion cause the hull to rotate about 5 degrees (while submerged). I expected it to rotate/pitch more. When I thought about it I decided that the small motion was due to the relatively small amount of bouyancy in the flooded ballast tank.

So this time round I trippled the amount of foam in the ballast tank, trimmed the model again and when all was said and done I got about 5deg total deflection while on the surface ( according to the protractor stuck on the sail), and about 8 deg while submerged.



Now 5deg may not sound like a lot, but on a 66" hull this translates to either end rotating through 4". On the surface, at the stern it would go from having the propeller shaft above the waterline to the waterline just below the towed array pod. The change on the bow was not quite as pronounced, but it was still quite visible.

You would think I would be happy with this. But I was not. The cost of having almost 1/3 of the hull allocated to shifting the ballast tank around is too high for my taste. So I am aborting the "how does it respond to the APC" and any other testing.

It's time to move on and convert to a "let's move a ballast weight allong the keel and see what happens".

[anonymous]

Hi Chris,

I think your phrase

Hi Chris,

I think your phrase *pause point" describes it very well.

Following an excellent article in the magazine by a German contributor some years ago, I always include a small piece of foam at the top of the conning tower to do just this. Your idea of having fore and aft pieces of foam for more stability is, I think, new.

I take it that you are dropping the idea of using the moving ballast tank ( not enough buoyancy when dived?) When you look at a moving weight don't forget the battery! (one bit of weight you must have.)

BTW I've just got the latest issue of the mag and it was very good to see your article in what is a really good issue.

Davy

[cstranc]

Davy,
Thank you for the

Davy,
Thank you for the feed back. I would love to read the article you mention. Is it in an old edition of the SCR? I'm a relatively new member (2 years) and I am very aware there is a lot of knowledge out there. So I try and scan old posts on the message forum etc, but I bet I get 1/10th of the good info.

Yes, the mobile ballast tank is dropped. Not for a lack of bouyancy. Even after adding foam in my last experiment I still had excess bouyancy for the target waterline. I just could not handle allocating that much space in the hull to moving the tank around. It seemed so wastefull.

So I have been spending the last few days converting to mobile ballast weights on the keel. Hence no posts. But I have been taking pictures and should be able to do an update soon.

Chris

[cstranc]

Finally past the conversion from

Finally past the conversion from moving the ballast tank to moving ballast weight. This is how it went...

Well first stop we needed some convenient weights, so it was off to the cottage where my father has all you need to pour lead weights (he says for fishing, but I think he was just planning ahead for this day).


500watts of lead melting power. and a little aluminum mold to form the lead in. Our first weight was the runt of the litter:

You know molten lead has to be hotter if it ends up looking like that. Actually speaking as a carbon based life form, it's hard to imagine wanting anything much hotter than molten lead near you. We were very careful with that stuff. In the end (after many hours of industrious fun) we had converted 9lb worth of diving weights into:

The rectangular weights are my "mainstay" for getting a weight distribution in the hull. Each rectangular weight weighs just over 8oz. I use the small fishing weights to adjust it while trimming (they weigh 3/8oz).

So the plan was to use five of the rectangular weights as the mobile ballast weight. Below you can see me start to piece the moving weight together.

After sanding the weights so I had a parallel sides I had to drill a hole through the center of each weight. A threaded rod went through these holed to hold all five weights together.

This is the jig I used to ensure the hole in each weight was in the same location.


I did not want the weights to slide around on the hull, so I designer two little roller assemblies. One set of rollers for each end. One of the two assemblies would also hold the nut that engages the lead screw that moves the ballast weight.

This is the bottom of the roller assembly un-assembled.

There are two brass rods. The larger diameter one is attached to the aluminum support (with CA), the smaller one is slightly longer so that when I tighten the nut / screw it does not grab the two collets that I use as wheels.

Here it is assembled.

Grumble grumble. It's just the angle of the picture the collets do extend below the aluminum frame and roll quite nicely. But I found when it was all assembled in the hull the edges of the lead touch the hull and prevent my little rollers from doing their work

Having made these wheels and been lugging all that lead around I thought I would add some more resin to the bottom of the hull. This would serve a few purposes:
- Make it stronger.
- Give a smoother surface for the ballast weight to roll allong.
- Increase chance of surviving torpedo attack.
The area prepared for the additional resin (and two layers of cloth).

And after the resin is applied.


It certainly is a lot stronger now. It's not quite as smooth as I would like. I will probably go back after finishing the testing and clean it up...

So we have the lead weight. Now it's time to create the lead screw assembly to move it. Now it was clear that the lead would need to move more slowly so I designed this with a 3:1 belt reduction.

This assembly is just for testing. It sits in two sets of flood vents so I can secure it to the hull without needing any glue, etc. If you follow the lead screw to the right you can see a brass tube glued into the aluminum assembly. Then there is a brass tube located on the lead screw. This smaller tube will just slide into the fixed tube and is there to reduce the friction. There is a collet to hold the lead screw on one side and the pulley on the other side.

And here it is all assembled:

Missing the green tie wrap material I use for holding things in place while testing...
Looking end on you can get a better view in the hull.


Now I nkow many of you are thinking "so why didn't he setup to move the batteries???". Well I could have, but the thing I did not like with moving the ballast tank was all the space wasted so the ballast tank could travel around. The batteries are in a large case too, and I did not want it taking up space as it moved around.

My little lead weight on the other hand can live at the bottom of the keel and...

... I can put in a "false floor" on the sub. This frees up all that space above the false floor to do neat things.

Time for a little info while I remember. As it stands (for testing) the lead weight is about 3lb. The lead screw allows the weight to travel 20" allong the keel of the hull. It takes 46seconds to go from one end to the other. When tested the hull weighed 21lb (I had weights installed to simulate the components that were not in place).

Of course allll of that will change before the hull is complete, but now you know what the testing will be based on.

In the next shot you can see what it looks like before I close the hull up for a test.

The bottom half of the hull is in the background. The ballast tank and electronics are attached to the false floor.

The front half of the hull is in the foreground. Note the triangular pieces of foam attached. These were to give me my "pause point" when the hull was awash. Based on Davy's warning I did play with the hull with these installed. There was significant resistance, when it was hull awash, to pushing the nose down below the surface.

I think I shall play with the shape of the foam so there is less foam right under the hull... Ahh the endless hours of fun.

Speaking of fun. It's time to trim the hull after all the changes.

Ohh look it's too high, what to do....

I really like those little weights and the tie wrap stuff. Its easy to add a little, shift it around until you get the desired effect, then pull the model out and make the changes permanent.

Soon enough it was trimmed and ready to start the next round of testing.

For this round I will not be using the APC again. I shall control the ballast weight directly. Now I started measuring angles again, but bleah. A picture or 10 is worth a hundred words.

Here it is on the surface, trimmed and neutral, with the ballast weight centered.


Move the weight 10" forward...


and then 20" back...

The waterline is about 1/2" below the midline of the hull.

Looking from the back...

We are about half way through the towed array pod.

Now we bring it back to neutral and down to the hull awash...


Neutral position from the stern...


Ballast weight goes all the way forward...



and all the way back...


Finally we vent the ballast tank fully (with the weight fully back)...


Beleive me the stern has grounded out.

And forward...


Ta da. Testing done for the night.

So what did we find?

As expected, when submerged moving the weight around had a much larger effect then when it was surfaced.

That towed array pod was up to it' evil deeds again holding an air bubble and then letting it out. I hoped I had fixed that.

The distribution of weight was excellent. There was no hint of it pitching over as it went through the dive. In fact I would push and prod it and there was a strong force keeping it vertical. This will be important then the prop starts spinning....

Obvoiusly when the boat is submerged you can move the weight and tip the model more than you want. But the model did respond very predictably and, dare I say it, gracefully to the motion of the ballast weight. If you dont' move the weight as much while submerged this could work quite nicely.

Many people may say that 46 seconds to go from one end of the track to the other is time for the coffee to get cold. I'm not really too worried about that now. Currently I am using a cheap 12V motor out in the water. I suspect I may shift to a high torque 7.2V motor (still with the 3:1 reduction). I would locate this inside a WTC. I think that will have a significant impact on the traversal time. And if I don't like what I get, in terms of performance, I can simply add weight (effectively speeding things up) or remove it.

I love having options.

But the real big unknown is how the APC will work when it is in the system. Is it going to make the model rock back and forth? When the weight is way far forward at the start of a dive will it pull the weight back fast enough to maintain the pitch angle?

So that's the next test. But before I get there I need a disable switch...

I can say now though:
- I am sold on moving the ballast weight around.
- If you moved a much larger weight it would not need to go as far, but for the distances I am looking at I am happy to be running a motor and a lead screw not a servo. But I would really really like to know more about other people's experiences with moving batteries using a servo.

The waterline on my Sierra is not parallel to the midline of the hull. The stern is sunk lower into the water (for obvious reasons). Of course when submerged I am sure the model would like to sit horizontally. I am planning on mounting the APC on a servo so I can simulate this effect. Trivial, I know. What can I say "too many channels, too little time."

Chris

[Rogue Sub]

Chris,

You lucky dog I want

Chris,

You lucky dog I want to melt some lead. It just sounds so manly!
Seems the weight is working 20x better then the ballast tank was. Your boat is gonna look real when you final go out for alfa tests.
I cant imagine how much money youve clicked away in your experimentaion. You know if you get this thing nailed you could produce smaller kits for sale to recoup some of your costs.

[cstranc]

Nuke,
It has not been

Nuke,
It has not been expensive (at least not the testing). Dad had the lead melting pot (I think it retails for about $80cdn). I got 9lb of lead for $9, and the rest is the type of stuff you just have on hand (or else you have to run out to the store all the time).

It has not been the money, its the time. Wow. I'm not as brave as Mylo. I do not have an "elapsed hours" counter. That would be too scary.

It's night and day between moving the ballast tanks (seemed like a good idea at the time...) and the ballast weight (look it really works )

I sure hope this sub behaves well when it goes for it's first run. I wonder how many months away that is....

[Guest]

If you got yourself a

If you got yourself a copy of 'model submarine technology' by Norbert Bruggen, you'll save yourself a lot of time and effort (this is all covered in the book).

Regarding the array pod, if you can't get it to vent reliably, fill the blighter up with resin.

Andy

[cstranc]

Andy,
I found a reference

Andy,
I found a reference to the book "model submarine technology" last week when I was looking back through the old posts. It's on order, and I hope to recieve it any time now.

Chris

[cstranc]

Well last night I decided

Well last night I decided it was time to prepare for the final tests: How will this work with an APC (automatic pitch control) guiding the weight.

While setting up for the test I switched out the cheap 12V motor for a slightly less in-expensive 55 turn (high torque, low RPM) motor. With the old motor it took 46 seconds to travel from end to end on the lead screw. The new motor covers the distance in 16 seconds. It seems like a nice speed.

Then I connected the APC between the reciever and the Model Control Devices ESC.

And that's when things kinda fell apart. With the motor detached from the lead screw I verified that minor changes in the APC angle will start the motor turning. The MCD controller has a really nice low speed response. But then I attached the motor to the lead screw and tilted the sub (it's still on the bench). Nothing. So I tilt some more... nothing. It turns out I have to pitch the APC over 5 deg from level before it sends a strong enough signal to the ESC to actually get the lead screw to start moving the weight.

Well that's no good. I did this whole thing so that I would have "fine" control over the pitch angle. If it has to nose dive that much to start correcting I think it's pointless.

I stared at the APC ( tried using its different sensitivity settings actually) and the ESC. Thinking this may have been a little bit of a waste to time.

I have a few notes about converting the model so that it does not use an ESC. Or try a different ESC, ... But I don't really have the time / heart to try a new approach now. There are too many other things to do.

But in truth I has not been a waste. I have spent so much time playing around trimming my sub and seeing how different locations of lead and foam work. It now submerges beautifully. If that was my final goal, and I don't need anything else to achieve it. Then so be it.

[anonymous]

Hi Chris,

Don't give up -

Hi Chris,

Don't give up - it sounds like the response of the ESC is the problem - they often don't have a linear response (e.g Small stick movement from neutral may have little effect on motor speed) The accelerometer chip in the APC4 should be easily sensitive enough to sense 5 degrees IMHO.

You may need a different speed controller or have to have to simpler one made up. (Let me know if you want to go along this route.)

On another matter, the article in the SCR I was thinking about is "Trimming a sub" by Volker Klein, pp 16-18, SCR Issue 53, June 2003. email me and I can send you a scanned copy.

BTW I think you will find Model Submarine Technology a very useful but challenging read!

Davy