Moebius Skipjack and proposed ballast system

[DanLew]

Here's a conceptual sketch - an overall ballast mechanicals system concept and an inner-tube constraint concept.
Re heat on the dry pump - ah, I see your point. So, add a water feed line - minor flow thru small orifice and one-way valve - from exterior to keep pump wet when in reverse bag-empty mode?
Pics attached of pump source etc. Pump does 7/8 cup (7oz.)/min. at 6.4V with spray nozzle on fine stream. Exceeds 2cups/min with open discharge.

[Subculture]

The check valve needs to be servo or solenoid controlled. With the pump on in either direction the valve is open, once you stop the pump the valve reseals.

A simple and effective solution is to use a servo to pinch off the hose. This is well proven to quite high pressure (e.g. in excess of 40psi) so you've no worries about being a pioneer. As someone once once said, pioneers often get arrows in their backside.

[DanLew]

Based on feedback, an updated conceptual sketch....

First step will be to use the 2.75 x 22" WTC to approximate the amount of water to be pumped to go from surface bouyant to just fully submerged.
Plan to load the tube with weight to reach half submerged. Will assume for now the component weight outside the WTC will take the finished hull down to approx. desired waterline. Will add weight til she sinks. That should be near the delta water weight range for the bag. Need to understand the critical weight difference range between just afloat and just submerged.
It's a start. Reco's welcome.
Thanks

[DanLew]

Hey, what happened? Yesterday there were two posts here that are now gone. One had a link to a Skipjack build and the other was from M.
Gone forever?

[redboat219]

Whistling while looking up...

Anyway, here's a nice system you could try and emulate...

http://www.rcgroups.com/forums/showthread.php?t=917946

[He Who Shall Not Be Named]

Hey, what happened? Yesterday there were two posts here that are now gone. One had a link to a Skipjack build and the other was from M.
Gone forever?

In my case I'm suffering a double-posting issue at this site. I removed the redundant posting to make things here a bit less cluttered.

M

[Kazzer]

Whistling while looking up...

Anyway, here's a nice system you could try and emulate...
http://www.rcgroups.com/forums/showthread.php?t=917946

Ark Model! Caveat Emptor!

[redboat219]

I didn't say buy, I said emulate...

[DanLew]

Bookmarked the references. Tim's Sculpin approach is great. I've a lot to learn.
So, given all the info on available commercial stuff, excellent past work on every ballast approach imaginable, why should I go forward? I guess I'm doing it for the challenge to make this approach work and to learn along the way.
What advantages might it have? Relatively simple pitch (and depth?) control, uses some cheap readily available parts, ability to surface bow-in-the-air, maybe something new and useful.

Anyway - some (likely boring to the experienced) test results:
Testbed WTC: L,internal=59cm, ID=6.35cm, V,internal=1868cc
WTC weight = 503g
Ballast added for approx. 1/2 immersion = 875g (total W=1378g)
Ballast added for 99% immersion = 480g (total W=1858g)
Ballast added for very slow immersion = 8g (total W=1866g)

Calc's and test agree. So, simply assuming roughly 500g of water to go from fully surfaced model to sinking, the bike tube needs 500cc internal volume. The tube is roughly 38mm ID, so total tube length required would be 44cm (17.5in). Two tubes, side-by-side, will be used, so length of the ballast bag assembly will be about 22cm (8.7in).
Good to go on fit.
The pump rate, minimal head P, was about 500cc/min, and about 220cc/min against significant head. That was with a 10" small dia. intake tube. Rates are higher with free intake. I'll assume about 375cc/min as a ballast pump. So dive time, from full surface to full dive would be 500/375, or about 80 seconds. Seems slow.....
Any thoughts on the pump rate/dive time?
To go from submerged at surface to slow sink is only about an 8g increment. A pump burst of about 1.3 sec would deliver that increment. That seems like the pump would offer manageable buoyancy control, even if total dive time is slow.
Next step - mock-up ballast system in WTC based on above, just to see how it might work.

[Subculture]

Pump sounds a bit slow to me. Personally I find the sweet spot for dive time is about 10-20 seconds, much more than it starts getting a bit wearing.

A kavan pump operates at about 1.8 litres per second, which would give you a dive time of under 20 seconds.

[DanLew]

Made a test-version of the innertube ballast system to determine fill rate, size, etc.
Got info from tsenecal that on his water ballasted Skipjack it takes about 0-400ml of water ballast to operate his sub. That's less than the approx 500 I estimated. I made up a 3-tube ballast assembly that will fit inside a 2.75" or larger diameter WTC. Overall ballast tube length is 10.5 inches. If the final version needs to be shorter, I'll go up in diameter on the bike inner tubes.
Pic of the test system with the cheapie pump. Performance: At 6.3V, the pump filled the bags, on average, with 420g of water in 40 seconds. In reverse, it emptied them in about the same time. So far, so good. A bit slow, but the pump can always be upgraded if necessary.
To address the comment that a flow check valve would be needed to prevent leakage flow through the pump, I plan to use the servo setup in the sketch. At center, the servo would activate neither of the attached microswitches. The switches are wired as a reversing circuit for the motor. In the center postion, a roller on the servo arm compresses the water feed tube, acting as an "off" valve. When the servo is moved either direction - to fill or to empty the ballast bags - the roller simultaneously opens the tubing to allow water flow as either pump switche is activated. One servo covers it all.
Next step is to rig these tubes in a closed cylinder with a pump control servo, bag compression servo, Rx and batt pack to see how it functions. Maybe by late this coming week.

[redboat219]

Go with a single bag. Each bag may not fill exactly like the one next to it causing roll instability. BTW, you're all ready good with your present set up. If I were you I'd ditch the bag compression/leveling mechanism. Keep it simple.

[Subculture]

Don't think you'll get much of an issue with the bags filling unevenly. The pinch valve looks fine, it's recommended to use a servo with a ballraced output, although by no means essential, just make sure the tubing you are pressing against is silicone and quite easily squished. Some modellers like to set the valve up so it leaks very slightly, This gives a built in mechanical failsafe, should you lose electric the pressure inside the cylinder should gradually push the water back through the pump and your boat will come to the surface.

[redboat219]

Some modellers like to set the valve up so it leaks very slightly, This gives a built in mechanical failsafe, should you lose electric the pressure inside the cylinder should gradually push the water back through the pump and your boat will come to the surface.

What a great idea. Thanks Subculture.

[DanLew]

Go with a single bag. Each bag may not fill exactly like the one next to it causing roll instability. BTW, you're all ready good with your present set up. If I were you I'd ditch the bag compression/leveling mechanism. Keep it simple.

Thanks for the continued interest - feedback keeps me thinking...
When I first RC'd the Trumpeter Seawolf - long ago - the single airbag never seemed to fill the same way, constantly causing a different roll and pitch with every fill/empty cycle. After trying many bag materials and approaches, I found that two long, stiff-walled bags (the innertubes) worked great.
To tell the truth, I don't see at all how a balloon bag fills with water in a repeatable way. I don't doubt that it does, I just don't see how. A soft bag of water would seem to slosh around, causing instability.
My main purpose at this point isn't to just build a Scorpion conversion. It's to play with the idea to see what advantages (and disadvantages) it might have. In the testing so far, the bags look to fill and empty very evenly many times over. And pitching/rolling the bag assembly doesn't allow the water to freely shift. Touching against each other, the three stiff-walled bags look to have a baffle effect.
I did test the system without a check valve. Leakage flow thru the pump was drop-by-drop. So maybe it would be good to eliminate the check valve and count on the leakage as the slow failsafe.
Next step will answer many questions.
Best,
Dan