In-Water, Real Time Thrus Measurement for FE Boats

[questtek]

In this series of tests I plan to evaluate the total thrust generated by a variety of brushless motors, ESC’s, propellers, and battery voltage combinations. The video briefly describes my in-water, real-time test set up. It is based on a simple see-saw principal with motor prop on one end submerged in a lake and a digital scale on the other end to measure thrust.

http://www.youtube.com/watch?v=tTSXH0VC5_k&feature=youtube_gdata

I am using a large lake as my “test tank” so there is not container of wall effects. Water depth is about 4 to 5 feet so I have no bottom interaction effects. Force or overall thrust is measured by a digital scale and motor / battery data is logged by an Eagle Tree V3 data logger with brushless motor RPM sensor.

In the video you will note the immediate force or thrust values as measured by the scale in response to changes in speed via an Astro Flight servo eliminator. One interesting thing is the ability to see the start of cavitation bubbles beginning to form on the prop as speed is increased. The clear water of the lake provides an excellent view of propeller dynamics with changes in RPM.

I have designed the test set up where I can run 1, 2, 3 and even 4 brushless motors at one time to see how thrust varies with hardware dollars expended. I can also do some in-water load evaluations with one ESC powering several brushless motors. The motor loading can be easily unbalanced by varying the props on each motor.

Although my application is more than just FE boats, I hope interested forum members can gain some useful information from my testing.

http://www.youtube.com/watch?v=tTSXH0VC5_k&feature=youtube_gdata

[johnson22456]

Look's good, anxious to see some more results.

Are you using a surface prop during the test?

[m4a1usr]

I'm going to assume you know the weight on both ends of the lever needs to be at least statically balanced for the measurement to be relative in value? And in distance they must be equal in length? Otherwise an increasing value of error will occur as acceleration forces increase? Its not going to be linear either.

John

[jac4412]

Hmm interesting... I look forward to see what comes of this.. great idea.

My only concern is that props and motors that are very similar (but still have a great difference) won't show a significant difference in your test.

[questtek]

I'm going to assume you know the weight on both ends of the lever needs to be at least statically balanced for the measurement to be relative in value? And in distance they must be equal in length? Otherwise an increasing value of error will occur as acceleration forces increase? Its not going to be linear either.

John

Yes, that is why the cup at the end opposite the motor. Lead shot is put into it to balance it exactly before strting the test. Having the ability to balance this static weight also permits me to use different motor configurations and even multiple motors. I can also move the battery to adjust the balance in addition to the lead shot.

The aluminum channel has a pivot point that is exactly in the center. This is why there is a direct, 0ne-to-one relatonship between the Thrust of the motor and the force indicated on the digital scale.

[questtek]

Hmm interesting... I look forward to see what comes of this.. great idea.

My only concern is that props and motors that are very similar (but still have a great difference) won't show a significant difference in your test.

I do not understand how this can be. If the systems creat different thrust, the force indicated on the scale will be different. The accuracy of the scale is to the nearest gram which is close enought for my needs. I will start posting some results..first wiht the single Leopard and changing prop diameters and pitch. It will be interesting to see just what the differences really are. On value of interest to to me is thrust per watt.

[questtek]

Look's good, anxious to see some more results.

Are you using a surface prop during the test?

No, props are fully submerged. I know in FE boats the props are half out of the water but I am not testing for this. My main applications deal with fully sbmerged props only.

[graill]

I'm going to assume you know the weight on both ends of the lever needs to be at least statically balanced for the measurement to be relative in value? And in distance they must be equal in length? Otherwise an increasing value of error will occur as acceleration forces increase? Its not going to be linear either.

John

Well stated John. And John just chipped at that iceburg.

Before i gave up last year Just trying to work with full size racing companies to do figures for rudder force for servo(s) selection, the size of our boats and the power they can put out made the big guys with phd's heads hurt, and they have some trick equipment.

[LiPo Power]

Kool test!!!!!

[jac4412]

Well how sensitive is the setup/scale? Is it a micro/nano scale?

My concern is that a test like this you want to look for the thrust difference between two setups... now my point is that while 2 similar setups may have a HUGE difference when running your boat, it doesn't necessarily mean you'll see a significant difference in this test. A 2-3 or even 5 mph difference or a significant difference in real time acceleration may not show a big difference in this test... which is what we're looking for? Maybe I'm just nitpicking but regardless this is a great test and a good idea... I look forward to some results.

[questtek]

Well how sensitive is the setup/scale? Is it a micro/nano scale?

My concern is that a test like this you want to look for the thrust difference between two setups... now my point is that while 2 similar setups may have a HUGE difference when running your boat, it doesn't necessarily mean you'll see a significant difference in this test. A 2-3 or even 5 mph difference or a significant difference in real time acceleration may not show a big difference in this test... which is what we're looking for? Maybe I'm just nitpicking but regardless this is a great test and a good idea... I look forward to some results.

The current digital scale I am using reads to 11 lbs. The resolution is 1 gram. I checked the accuracy with known value weights and it is very, very good. By using the lead shot in the white cup to balance the weight of the motor I can get it within a few grams then just TARE it out. Thus the scale force represents the actual Thrust of the motor.

I agree with acceleration, boat set up, etc. However this test is ONLY to compare performance of submerged thrusters steady state output. The end result is for a new military underwater propulsion system. Small is key so I need to get the highest thrust...(that will eventurlly propell a diver hands-free)... with the max efficiency which, in this case amounts to the lowest watts consumed per pound of thrust.

This may not be directly applicable to predicting FE performance but it hopefully will show how thrust varies as various components such as motor, batteries, props, gear boxes, etc are changed. Hopefully I will have a chance to post some initial results today.

[Make-a-Wake]

Not sure about all these error values and inaccuracy in his testing since he will be using the exact same setup for each motor.............that is the "control" of the experiment. I bet it will show that a 4074 Leopard 2200kv creates more force than a 3674 of similar kv.

Only "tweak" i would offer is to get the prop down a bit further in the water.............once again though, using the same setup will offer some accurate results.

[questtek]

Not sure about all these error values and inaccuracy in his testing since he will be using the exact same setup for each motor.............that is the "control" of the experiment. I bet it will show that a 4074 Leopard 2200kv creates more force than a 3674 of similar kv.

Only "tweak" i would offer is to get the prop down a bit further in the water.............once again though, using the same setup will offer some accurate results.

Good tweak suggestion and I have included that possibility in the build. You will notice a HYCO black fitting that holds the shaft outer tube. By untightening this fitting the outer tube that supports the shaft can be removed. I can handle outer tubes from 1/8" up to 3/8" I can use teflon liners or not. I can test from a thin wire drive down to a full 1/4" drive shaft. I can aso make the tube and drive shaft any length I want.

I used the short shaft in the initital set-up because I wanted to view what was hppening in terms of cavitation on the prop. It was easy to see and I can nove my prop down more now. It is currently about 4 inches underwater I believe.

[Fluid]

Static thrust measurements are interesting when comparing different props or motors, but they may tell little about actual performance. Rpm and thrust can be significantly different once the boat (or plane) begins to move.

I tried a vaguely similar test apparatus years ago but found that prop cavitation was a major issue when static testing. Once the prop speeded up enough to began to cavitate (water vapor surrounding the blades) the thrust dropped a bunch. Keeping the prop deep enough and slow enough to prevent cavitation will be necessary.



.

[m4a1usr]

I still applaud you Joe. I didnt have sound available when I watched the video so I couldnt gather all your facts from any vocal commentary. Dont take my comments negatively. I dont mean them that way. My hats off to you in all respects.

John