Can a jet drive stator stator fully counteract the motor torque reaction?

[Mermaid]

:confused: I am a model submariner but have subscribed to this forum because the the basic principles governing the design of model vs full scale vessels are almost identical.
Model subs need model armament so I am currently designing a 1/20 scale (1.05" x 14.5") RC heavy torpedo. Being RC, it is most important that the unit not rotate due to torque reaction while under power. Most torpedoes have a pair of counter rotating props to negate the torque reaction.
At this small scale, a counter rotating system will be difficult if not impossible to implement. Consequently, I am exploring the possibility of using a pump jet propulsor (jet drive) in lieu of the props as is the case with the US MK 48 torpedo.
I am considering a high volume low pressure mixed flow design with a single 3 bladed impeller ahead of a 4 bladed stator. The question is: Is it possible to design a stator that will both a) straighten the water low behind the impeller and b) generate sufficient counter torque to balance the motor's torque reaction?

[Rick Willoughby]

It is possible. A bit of a fiddle to get it just right.

Why not just ballast the torpedo so it has enough righting moment to counter the torque on a conventional shrouded prop. All the heavy bits down low and the light bits or space up high. Spin the prop at high rpm to keep the torque down while delivering the required power.

Rick W.

[Mermaid]

Dear Rick: The problem is that at 1/20 scale, the torpedo will be only 1.05-1.1"D x 14.5-15"L and displace APR 170g H2O. Within this small package, I will have to accommodate the drive motor, speed control, 4ch micro receiver, 350mAh battery pack, 2 nano servos and micro antenna.:!: All this (including all of the torpedo's structure) must weigh less than APR 170g or the thing will become negatively buoyant and sink (permanently)!! I think that you may see that I have very little latitude in terms of ballasting.:eek:

[Rick Willoughby]

Dear Rick: The problem is that at 1/20 scale, the torpedo will be only 1.05-1.1"D x 14.5-15"L and displace APR 170g H2O. Within this small package, I will have to accommodate the drive motor, speed control, 4ch micro receiver, 350mAh battery pack, 2 nano servos and micro antenna.:!: All this (including all of the torpedo's structure) must weigh less than APR 170g or the thing will become negatively buoyant and sink (permanently)!! I think that you may see that I have very little latitude in terms of ballasting.:eek:

Where is the problem. Like I said heavy bits on the bottom air at the top. If they can make this thing fly:
http://www.wirelesshut.com/307772/12616610.html
at 10g and with remote control I cannot see that you will have a problem with 170g. Monstrous by comparison.

(Not that I would like to work with such tiny parts. I have one of these. I wonder if the optical control is easier to do than radio.)

Rick W

[Mermaid]

Dear Rick W.: ::( The problem is that the heaviest components (motor and battery pack) will have to be mounted on the C.L. of the torpedo. Remember that I am having to fit rectangular components (motor excepted) inside a 1" I.D. tube. I'll have to cram everything else into the remaining space. The weight of the wires will also be significant. I will certainly consider what you have suggested. The proof will be in the pudding so to speak so I'll have to build it to work out the final bugs.:D

[Rick Willoughby]

The motor in the model plane referenced is about 1/2" in diameter and drives the prop through a 6:1 gear ratio. No need to mount the motor on the centreline. Even AA batteries could be mounted well below the centreline in an inch tube. Tube could be balsa on top and lead on the bottom. Lots of ways to get weight distribution in the best place.

Rick W.

[Mermaid]

Dear Rick: What suffices for micro RC aircraft will not work for a high performance model torpedo. Water is 800 times as dense as air and therefor drag/unit speed is about 800 times as great. For a given drag coefficient, for equivalent top speed, you would have a power requirement which is orders of magnitude greater. Of course, the greater power requirement is concomitant with greater battery capacity. (more later)
I HAVE considered using the Feigao 1208436X motor. This is a 12mm, 17g motor rated for 4A. It performs at 4133rpm/V. I would have to mount it off axis and use a 2.8:1 gear reduction to get reasonable torque and rpm at the impeller. Max power output would be about 32W @ 11.1V at 72% overall efficiency. However, one of the main points of this exercise is to keep things simple as possible. I want to avoid having to building a contra rotating prop drive and am exploring my options. Last thing I want to do is to add a gear train if at all possible.
My motor of choice is the Feigao 13084141S. This is a 20mm, 43.4g motor rated for 6A. it operates at 2283 rpm/V and could be connected directly to a suitable impeller. Max power @ 11.1V is about 57W @ 85% overall efficiency.
Lithium polymer cells are the only type of battery with sufficient power density for this project. I have chosen the 11.1V 350mAh Flightpower Evo Lite power pack. This is capable of 18C discharge (6.3A). NiMH's are too heavy, have lower power density and have problems maintaining voltage at high discharge rates. Same (only more so) applies to NiCADs.

[Rick Willoughby]

.....] However, one of the main points of this exercise is to keep things simple as possible. I want to avoid having to building a contra rotating prop drive and am exploring my options......

You said it. I am suggesting the simplest possible way to avoid complexity of vanes and tuning them is to just put the weight in the right place. Seems the 12mm motor is about what I was thinking. That little plane I have flies for about 15 minutes. Not certain if it uses a capacitor or battery but it has all the bits to do remote control flight. It climbs at about 30 degrees.

Typically flying is more energy intensive than sliding through a medium supported by buoyancy - the plane is quite light but also your torpedo will have a low drag coefficient. I can do the drag and power calculations if you want.

Rick W.

[marshmat]

Hi mermaid,

My first choice, as Rick has suggested, would be to rely on ballasting as much as possible- anything that possibly can go below centreline, should.

But I agree with you that in a 1" tube, getting the weight low enough to make a difference will be tricky.

2283 rpm/V at 11.1 V is a hair over 25300 rpm. I can find a 13084xxS in Feigao's catalogue, http://www.feigao.com/sdp/85838/4/pd-102732/1336473-51426.html , but you quote 3 digits in the gap where Feigao lists 2. How many turns on the one you're considering? What torque are you expecting it to produce?

I have a hard time believing you could get the overall CG off centre by more than 25% of the radius- or about 7-8 mm. For a 170 g cylindrical craft that's a righting moment- at 90 degrees heel- of about 0.013 newton-metres. A model aircraft can be made to have some degree of inherent stability; a torpedo is a symmetrical cylinder and has no form stability to speak of.

Try to get the weight off-centre, for sure. You need the static stability as well as the torque counteraction underway. But I'm not sure that'll be enough.

The whole idea of an internal jet pump at that scale strikes me as just as complex as a counter-rotation gear would be. Are you sure a conventional shrouded prop, with S-shaped stator vanes downstream of it to counter the torque, is impossible?

[Rick Willoughby]

Matt
Did you do the torque calculation at say 15000rpm? I figure a 1" shrouded prop would do wonders with 20W behind it at this rpm. I determine around 50% efficiency.

Still think just weight distribution can achieve the required result in the simplest fashion.

Rick W.

[marshmat]

Agreed that a shrouded external prop could do very well in this application.

Not convinced weight distribution alone can handle the torque reaction without some help in straightening the flow, but this could come in the form of vanes aft of the prop.

How close to upright do you want to keep it?

[Capt-Ron]

What diameter is the motor shaft? If you could replace the shaft with a hollow one of the same OD then mount two motors end to end and extend the front shaft through the rear hollow shaft. Wire the motors to counter rotate.

CaptRon

[Rick Willoughby]

What is your target speed. I am getting low drag numbers for the torpedo. Looks like it will need a fraction of the power you have proposed unless of course you want to do more than scale speed.

Rick W.

[Mermaid]

Thanks again for all the feedback folks!

:D After careful consideration of your suggestions, I have decided to go with a 12mm diameter Feigao brush-less motor (1.5mm shaft, 32W @ 4A and 11.1V = 72.1% efficiency with gear reduction). I am aiming for a top speed of 10mph. To date, model electric torpedo performance has been a bit anemic. I am hoping to change that with this design. Scale speed would be a rather unimpressive 1.75 kt. I agree that incorporating as much static stability as possible would be very beneficial. Offsetting the motor will certainly help. I also agree that I shouldn't rely on static stability alone to counteract the torque reaction. The stator is going to have to generate enough counter torque to handle the balance. I am planning to use an axial mixed flow pump jet (jet drive) with a 0.75" 3 bladed impeller with a pitch of 1.05" with a 4 bladed stator placed just aft of the impeller. Outlet area is to be APR 66% of the inlet area. I estimate that 16,761rpm will be required at 40% slip (if prop design concepts even apply to a pump jet) for a top speed of 10mph. Torque will be 2.582 oz-in. The length/beam ratio is 13.2 to 1. The total surface area is approximately 50 sq in. Rick: I have Apr 32W available with the above mentioned motor. I don't know if this will suffice. It would be a great help if you could estimate the power requirement. I have some of the equations used in designing full sized surface craft but their usefulness in designing model submarine craft is debatable.

[Rick Willoughby]

I have attached the prop design data for the conditions I calculate at 10mph. I determine the drag to be only 1.1N for a proper streamline shape. My area is 36sq.in and weight is only 131g. This is not quite the shape of a torpedo but not too far off. The power requirement for this is 6W at the shaft.

If I go up to to 1.15" diameter at the same fineness then the displacement is 175g and area is 44sq.in. Drag goes up to 1.3N and power goes to 7.2W.

Going to 1.23" diameter gets me to 50sq.in. This gives drag of 1.5N. Corresponding power is 8.4W.

The prop efficiency is determined to be quite high. This particular analysis uses Re# of 50,000 (lowest standard value) and it could be a bit lower than this depending on the blade chord length. I get only 18% increase in flow through the jet in the lowest drag case and 23% in the highest drag case so proposed nozzle necking to 66% is a bit much. I suggest around 80%.

To put the power into perspective - If I had 32W on my 24ft 100kg boat I would expect to do 5mph. So it is reasonable that something weighing only 1/600th should easily achieve twice the speed with 25% of that power; 32W would be way more than necessary.

Rick W.

[Mermaid]

Thanks Rick: You beat me to it! I was not happy with my most recent post and chose to edit it this morning. Your reply just came in! Just out of curiosity, what would my top speed be at 32W if I can achieve 10mph at 7.2W?

[Rick Willoughby]

Thanks Rick: You beat me to it! I was not happy with my most recent post and chose to edit it this morning. Your reply just came in! Just out of curiosity, what would my top speed be at 32W if I can achieve 10mph at 7.2W?

If the motor will spin up to 18,000rpm then you will get 15.6mph at 32W on the shaft. The velocity ratio is 1.23. This is for the 50sq.in case. Power very close to cube relationship with speed.

Rick W.

[Mermaid]

Sounds great to me!
:?: There are a couple of questions I'd like to ask about the impeller design though. What clearance would you recommend between the rotor tips and the I.D. of the shroud? Logic indicates that reducing this to a practical minimum on the order of 0.005" would be beneficial. However, at this small size, I expect that the hydrodynamics of the setup may have a few surprises in store. Also. I'm wondering about projected area. What would you recommend? I chose a three bladed design to avoid setting up harmonics with the stator so a large projected area (APR 70%) means a wide blade chord given only three blades. How would a large projected area effect efficiency?
By the way, How does one go about inserting images into one's posts? I've tried but the site requests an image URL. I don' have a web site but I DO have quite an extensive collection of documentation that I would like to be able to share with members from time to time. The current situation is a prime example. A few pix of the intended torpedo/impeller designs would sure help.:D

[Rick Willoughby]

Clearance - I don't think the water molecules will mind being squeezed. On the other hand the induced drag is a global effect of the order of a couple of blade chords at least so there is no need to have minute clearances.

Blade Aspect or EAR - I used JavaProp to do the analysis. I selected a 6% asymmetric foil as this one had the lowest Re#. The blade maximum chord for the high power case was 4.3mm. I have not worked out the EAR but it will be quite low. A thicker foil section would have even smaller chord. The efficiency of the prop is related to blade aspect. The higher the aspect ratio the more efficient for an unshrouded prop. With shrouding it is not as critical but using more blade area than needed means you have to drag more material through the water at high speed.

The blade size will more depend on the required strength. Altering size a bit will affect efficiency but not that much. If you are buying a prop then I would go for the lowest EAR available. The prop loading is only medium. Most boat props are heavily loaded. Going to 2 blades the chord goes up to 6.4mm but I agree that the vane passing could be an issue. When you think about it the loss of area created by the vanes probably adjusts for the velocity gain so necking is of doubtful benefit.

One other aspect I should point out is that my drag calculation is for a submerged body. That means it needs to be at least 3D under the surface to avoid wave drag. I was wondering how you will control the level. I expect you would lose radio contact if it goes too deep.

Rick W.

[Mermaid]

:) Thanks Rick: I am quite surprised at the efficiency figure (82%+)you obtained for the projected 1:20 scale prop (impeller). How confident are you in it? I only question the results because small propellers are significantly less efficient than larger ones all things being equal and an efficiency of 82%+ is good even for the prototype prop!

I calculate the maximum projected area to be only 21.86% of the disc (hub = 9.97%, blades = 11.89%). At a diameter of only 19mm with a max chord of 4.3mm and a foil thickness of 6%, the hydrodynamics are significantly altered by the so called "scale effect". Relative to the hydrofoil, the water molecules are 20x larger that they would be for a full sized foil. I wonder whether the program that you used takes this scale effect into consideration.

The torpedo will be launched at a minimum depth of 4". It will be ballasted to be just BARELY positively buoyant so I would expect that it will take a quite a few seconds to drift up to the surface especially at high speed even if I did not have RC. However, I will have control of both the direction and depth. :idea: If the depth control becomes too much to handle manually, I have the option of installing an APC-4 unit. This is a micro sized solid state gyro that automatically controls the pitch angle of model submarines under water. It can be adapted to do the same for the torpedo.:D

[Rick Willoughby]

The little prop is just that - little. But how many big props do you see doing 16,000rpm. The end result is that the Reynolds number is up around 50,000. The foil data I used was based on this.

The actual foil modelled was a cambered 6% foil - see attached. Given the short chord this is quite thin. So to some degree it depends on how well the foil is shaped. A deburred piece cut from a tin can would drop about 10% in efficiency. So might do say 1mph slower. There is benefit from the ducting as well. It is a while ago now and I cannot remember the detail.

I expect a bit of deburred tinned can twisted and soldered on to a hub would get about 60% if it was unducted. Ducting and a nice cambered foil as shown will get over 80%. Depends on how accurate the prop is manufactured. If you buy one you can try as is and then see if you can improve performance by shaping.

If you have not shaped a blade before, remember the lifting face is toward the front.

Going up to a 10% foil with a flat back side will not alter things much and will be noticeably stronger. I think working with anything under 0.5mm would be difficult. On the other hand you can get really carried away with perfection here for very little benefit. The power requirement goes up with the cube of speed so dropping from 80% efficiency to 60% will reduce top speed by 10%.

I am interested in your results as always but I would be surprised if my estimates are not in the ball park. I would go looking for what you have wrong before I question the numbers.

Rick

[Rick Willoughby]

The props I run on pedal boats and electric drive have an EAR of around 12%. The blades are lightly loaded so very efficient. Having more area than needed is just more area to drag through the water.

The constraint with you blades will be to get adequate strength at only 6%. I do not think this is practical so you can go to 10% and say 5mm chord. These would have a curved front face and flat back face. Somewhere between 4 to 5% camber. Something more like the Eppler attached. The differences with these things are no more than 1 to 2% in efficiency.

Rick W

[Mermaid]

:) Sounds good. Love the under-cambered foil. I had expected a flat pressure face. Obviously, the foil thickness will have to increase from the tip to the root to maintain the structural integrity of the blade. :!: I am certainly not going to fabricate the blades from bits of a tin can!. If I decide to make the blades out of brass, the best way would be to fabricate an accurate pattern in resin. A rubber mould taken from this would allow me to cast up a bunch of wax blade patterns. I might even be able to skip the wax! There is a special two part resin called Duralay which is used to make direct patterns for dental inlays. Duralay burns out just like wax. The blades could then be cast in brass via a lost wax process. I just took a look at JavaProp. Looks like a great applet though it's going to take me some time to learn to use it.:D

[Rick Willoughby]

I can make a prop out of steel flatbar that will get better than 85% efficiency. A milled prop gets at best 1% better efficiency. By comparison most outboard props are lucky to get 60% but then they are usually more heavily loaded.

If you use JavaProp you can compare a flat plate with the other foils. It will surprise you how well it performs. So the tin can is not too bad. Might need a rib of solder to get the strength near the hub but it is a fast means to test the whole thing out.

Rick W

[Mermaid]

I am VERY surprised! If by a flat plate, you mean that you are dispensing with the helicoidal axial blade twist as well as the cambered suction surface.
That would mean that you would have a variable pitch going from 27mm at the blade tip to 8.53mm at the blade root. If my assumptions are correct, the effective pitch would be about 19.1mm at 0.707r. With a flat suction surface the thrust would be developed purely by the backward acceleration of H2O by virtue of the blade attack angle according to Newton's third law. You are also dispensing with the hydrodynamic pressure differential between the suction and pressure surfaces generated by the camber according to Bernoulli's principle. This generates most of the thrust at higher speeds! :eek: This CAN'T be over 85% efficient!! It violates just about everything I know about propeller design.:D

[Rick Willoughby]

I am VERY surprised! If by a flat plate, you mean that you are dispensing with the helicoidal axial blade twist as well as the cambered suction surface.
That would mean that you would have a variable pitch going from 27mm at the blade tip to 8.53mm at the blade root. If my assumptions are correct, the effective pitch would be about 19.1mm at 0.707r. With a flat suction surface the thrust would be developed purely by the backward acceleration of H2O by virtue of the blade attack angle according to Newton's third law. You are also dispensing with the hydrodynamic pressure differential between the suction and pressure surfaces generated by the camber according to Bernoulli's principle. This generates most of the thrust at higher speeds! :eek: This CAN'T be over 85% efficient!! It violates just about everything I know about propeller design.:D

The flat shape refers to the chord. It is one of the options available in JavaProp. You have misread my post. What I said is that you would lose about 10% going to a flat plate. You can try it yourself as it is the default profile in JavaProp. The flat plate is either very thin or has a rounded nose and tapered tail.

Even a flat plate will fly it just has a narrow operating range.

The steel flatbar props I make have about 2 hours of hand grinding to profile them:
http://boatdesign.net/forums/attachment.php?attachmentid=23371&d=1215983491

Rick W

[Mermaid]

Thanks Rick, I think I understand what you mean. Sounds a lot like the flat X-sect blades that were used in the earliest propeller designs. I notice that the helicoidal blade twist is featured in the example that you attached. It has occurred to me that cancellation of the torque reaction with a single stator will, under the best of circumstances, be very much like walking a tightrope! :idea: Even though two contra-rotating props will be a pain to set up at this small size, I'm beginning to think that this may be the most viable solution after all! The designers of full sized torpedoes knew what they were doing. I know that the aft prop has to have an slightly smaller diameter, opposite hand and have a higher pitch than the fwd prop. :?: Is there a rule of thumb relating the aft to forward prop pitch? The torpedo would have the aft prop set just behind the forward one. I have the specs for a system used on the USS Albacore. In her case, the spacing between propellers was about half the forward prop diameter. The aft propeller pitch was approximately 1.2 x the fwd propeller pitch.:D

[Rick Willoughby]

You need to recalculate the prop using JavaProp at half the thrust. The flow plot will give the velocity ratio. Set the velocity for the second prop at the design velocity times the velocity ratio for the first prop. From memory the ratio was quite low. Much smaller than 1.2 but I would need to go back over what I did. If you select a pitch that is too high only one prop will work anyhow.

I expect it will be unnecessary complexity given the low torque. As originally stated I expect it would be quite easy to ballast the hull to counter the moderate torque.

I noticed in recent scanning of human powered subs that most prefer high efficiency unducted single props. Just use ballast and control surfaces to counter the torque.

Rick W

[Mermaid]

Thanks Rick: I agree that the best and most efficient drive arrangement will probably be the single prop and stator. I think that the contra-rotating props could be made to work but the system is inherently inefficient as power is wasted in the reversing gear and the coaxial shaft bearings and seals. I asked a retired submariner friend why WWII torpedo designers used contra-rotating props as opposed to a single prop and stator. He thinks that they never thought of applying axial turbine technology (stator) to a marine final drive system.
:idea: Each of the AGSS-569 Albacore's contra-rotating props was driven by a separate motor as opposed to a reversing gear. As such, speed of each shaft could be varied. This might account for the seemingly great disparity in pitch ratios (APR 1.2) between the fwd and aft.props.:D http://www.ussalbacore.org/img/img_virtual_tour/05-lg.jpg

[Rick Willoughby]

Those props indicate very high velocity ratio compared with what I like to achieve.

If you think about your torpedo you are running a prop similar in diameter to the actual hull. So by comparison with Albacore's prop it is a monster.

If the Albacore prop was the same diameter as the hull then it would have been more efficient but it would have been difficult to make the blade root strong enough.

It is a great photo. It just shows how clean that whole design is back aft. only compromise on that boat was the sail.

When are we going to see the first test run of the torpedo?

Rick W

[Mermaid]

Yes Rick, the velocity ratio IS quite high. I should mention that the Albacore used variable pitch props with the pitch being specified at 0.7 r. The Albacore had STUNNING performance being able to achieve 30+ knots underwater on battery power!! I only included the picture to illustrate what I was talking about regarding relative propeller pitches. I would not consider using this particular configuration on a model torpedo although I am adapting this design for a 1:32 scale model of the SSN-605 Jack. She was the ONLY US nuclear submarine adapted to employ a contra-rotating propeller drive. Jack's drive design was spurred by the exceptional performance of the phase IV variant of the system developed by the AGSS-569 Albacore research team. (below, bottom and the photo) Unfortunately, the Jack's designers strayed too far from the Albacore design and her performance was rather disappointing. Eventually, the ship was retro-fitted with a conventional single propeller drive. With the model, I plan to correct the mistake made on the Jack. The propeller configuration will be almost identical to that of the phase IV Albacore except that they will be about 14% over-sized (to improve efficiency) relative to the scale Jack props. The model will have a max design speed of APR 5-6 knots, surface area of 3,647.24 sq in, (25.328 sq ft) dived displacement of 336.1 lbs and measure 11.925" x 112.454" I have adopted a fixed pitch propeller setup rather than the variable pitch used on Albacore. Fwd prop: 5.3437"D x 6.424"P. Aft Prop: 4.44044"D x 7.695"P spaced 2.3379" apart. I'm not sure how much power it will take to get to 5 knots. my ball-park estimate is about 350W/shaft. The torpedo should be operational within a year if it is built at all. The project is more of an exploration of design options than anything else.
http://www.ussalbacore.org/img/img_albacore_story/6-1albac.gif

[Mermaid]

Dear Rick: :?: I'm having a couple of problems with this site so I need to ask a couple of seemingly dumb questions. (1) How does one mark a reply as read? (2) How does one reply to a private message? I have one private message so far but try as I might, I cannot open the message on the site nor find any means to reply.:D

[Rick Willoughby]

Dear Rick: :?: I'm having a couple of problems with this site so I need to ask a couple of seemingly dumb questions. (1) How does one mark a reply as read? (2) How does one reply to a private message? I have one private message so far but try as I might, I cannot open the message on the site nor find any means to reply.:D

It may depend on how you have selected your default screen but the one I use has the Private Message link hypertexed to the slightly light blue "Private Messages" text under my welcome in the upper right hand corner of the window.

If you left click with the mouse over this hypertext link it will open a new page. You then left click on the message in the list and it will open the new message. From there you can choose to ignore or reply.

Rick W

[Mermaid]

Thanks Rick: I've finally managed to reply to my personal messages. :confused: I STILL don't know how to mark a reply or message as read. The site indicates that I have about eight unread replies yet I have read every single one.

Best,

Mermaid

[Rick Willoughby]

It might reset if you log off and come back in.
Rick

[Mermaid]

Thanks for the advice Rick. I've logged off and back on again and my unread relpys show empty.:D