On board dyno for boats?
 

Marc at Bam raised an interesting point on the speed calculator thread. "Wouldn't it be nice to have an on board dyno for testing." I've wondered about this too and did a web search for answers last year. I think it could be done.

You can have a dyno in your race car. It's a wireless strain gauge that mounts on the driveshaft yoke and transmits a signal to a receiver shaped like a doughnut, mounted around but not touching the shaft. The minute torsional twist in the shaft is directly proportional to torque, the receiver can also sense shaft RPM. HP is just a calculation away.

I think to mount this kind of system easily and temporarily on a boat I would try to measure strain in the hub of the prop instead of the propshaft. The receiver could fit inside the back of the bullet. Because the strain gauge must be calibrated, you would have to have a set of test props with strain gauges installed. That could become expensive real quick for anyone offering the service.

One of the things this kind of test equipment might shed some light on is the %efficiency (not % slip) of the prop and the ability of the prop to absorb torque at different RPM. This could help you select the correct combination of gear ratio and pitch. We all know of cases where one ratio and pitch perform better than another ratio and pitch even though they have the same mathematical final drive ratio. Maybe prop efficiency is part of the reason why.

Isn't technology wonderful?

Sounds like a great idea. I am not an engineer, but. What numbers could you use for a baseline of measuring torque in the hub? Is there such a transmitter that would hold up in front of a prop under water?

Maybe it would be easier to attempt to measure the torque in the upper shaft that goes into the coupler.

That would be a great tool. Maybe an aircraft mechanic could come up with something like they use on turbine powered airplanes.

mike

Tomcat, fill me in on the strain gauges. The only ones I remember had a long "S" shaped wire which changed resistance as it stretched.

Do these new ones work by a light beam or what? Also, what is the source, Jegs?

I would think that it has to be passive to avoid commutation problems.

Anxiously awaiting your reply.

I can't remember who sells the on board dyno for race cars, may be Land & Sea. If you do a search for strain gauges there are lots of manufacturers selling them for industrial applications. Sorry, I don't remember the principle of measurement. Its been a while since I did the search.

Land and Sea makes a unit that attaches to the yoke comming off the trans. I measures 'strain' or twist and thru previous calibration can figure the amount of torque being applied to the yoke. The premise is that any shaft having torque applied to it will twist or deflect a slight amount and if we can measure that amount and compare it to a known amount of applied torque and it's produced twist.
Having said all that, I have looked into this quite extensivly and can't come up with a way of mounting it with a stock bravo. But I am buying a stellings box which I believe I can make it work on. My IMCO box won't work either. Imagine a 2 line digital readout on the dash which will display HP and TQ at your current driving rpm- on the fly! This is only some R and D away. So how much would you pay for this?

Again, I am not an engineer. But, how do you get those much needed and relied on baseline numbers????? If those numbers are not accurate, the entire concept will not be accurate.

Good idea marty! You could send the extension box driveshaft to Land and Sea and have them install and calibrate the strain gauge. Then install the receiver inside the box and your done. Very smart.

From a theoretical point of view IIRC, you just need the material properties and physical dimensions to determine the deflection at some given torque. You'd have to assume a few things like homogenous material and the dimensions remain constant under the deformation but I think you'd be close enough for a baseline.

Tomcat, That's it but the L+S product is expensive, connects to computer, no dash digital display, decent accuracy. I have another company which has all the best features- reasinabky priced, dash digital display with memmory for maximum values, excellent accuracy and can input to a data logger or computer. Allways thinking.

I looked up the Land and Sea site.

They sell a torque transducer which is mounted to a U joint yoke that they have modified. It appears that this yoke will twist quite a bit more than an all steel one and in the process twists a piece of ferromagnetic material rotating as part of the drive shaft. The magnetic properties of the magnetic material changes as a function of the amount of twist, which is in proportion to the torque. The stationary half of the device reads the change in magnetism and gives out an electrical reading in foot lbs of torque. (foot lbs is the same as lb feet as this is a non vector and the dimensions are commutative. We discussed this long ago, just believe me.)

They don't say how to calibrate the device but I would say the easiest way is to lock the crankshaft in a fixed position and put a torque wrench on the prop shaft spline. Then for every reading of the torque wrench, you set the calibration data points on the electronics. Remember, that torque is torque whether the shaft is rotating or stopped. Other torque measuring machines, e.g.,dynomometers are calibrated with the shaft stopped.

I looked under other related patents and this is a well developed technology. There are others which work off of light beams.

In any case, measuring torque means measuring the amount of twist on a shaft and one of the ways to do it is to measrue the reaction forces on the load as the traditional Pony brake machine.

The other way is to measure torque is to do what this device does and measure the twist of a rotating shaft with a non contact transducer.

It is very clever, but I like the idea of the optical pickup better, as it may be cheaper and less prone to drift.

Note, the car people are looking at fast transients in the torque readings. Most boat people are looking for steady state torque, unless you are in a drag boat.

Horsepower is the multiplication product of the torque times the rpm with a constant factor.

The transducer is mounted at the factory and the yoke is locked down/placed in a vice and a set amount of torque is applied and measured and calibration is set.

So how much would you pay for this on-board real time digital readout dyno?

I have no idea how much this might cost, but $1000 for the dash unit and $500 per engine might fly, as long as the installation was relatively simple. I wouldn't want to buy an extension box as part of the required installation. But sending my props away to have the strain gauge installed wouldn't bother me. That's assuming that the prop location is possible. Pulling the drive and having the input shaft modifed wouldn't bother me either.

As an option on boats I think you would market this instrument through the builders of high dollar boats, and it might have appeal beyond the high performance boat arena. For example, it might be worth it to the owner of a fast 50+ foot sport fisherman. Any where you can brag to your buddies about a big number right? It could be a testing tool for high performance marine service centres. And it could help engine builders sell their product, as long as the "after" test met expectations. Boat tests for magazines and TV shows might be another outlet.

I wonder if it would be possible to modify the engine coupler with the sensor? Not only would you get a torque reading but you might also have a pre-indication of coupler failure(i.e. your horsepower/torque would suddenly increase due to excessive coupler movement). It would also then be easily adaptable to virtually any marine engine/drive combo.

It needs to be permanetly fixed to a shaft. The stelling box looks to be the easiest application. High end builders who use these boxes- Skater, Formula, Fountain, etc. Might be able to also go on boats with trans.

We have an on board "real world" dyno that we use during testing. With our own computer system, we measure through dual mass air flow sensors and two wide range oxygen sensors. So were measuring air by weight going in, then we measure what was burned. It's as close as most state of the art dyno cells. So it's very possible, but expensive. This doesn't drift or vary, it's very unique.

Thanks,
Dustin

Dustin, So how does that get you tq and hp numbers?

From a theoretical point of view IIRC, you just need the material properties and physical dimensions to determine the deflection at some given torque. You'd have to assume a few things like homogenous material and the dimensions remain constant under the deformation but I think you'd be close enough for a baseline.

ttt. I'm getting closer to the on board dyno. Today dropped off the pieces to get it made. Can't wait.

If it needs to be calibrated with a torque wrench why not incorperate the strain guage to the motor mount or make the guage part of the mount. Even though the mounts are ridgid they still see the load. If it is buffered you should be able to correct for it. Tell me if Im wrong.

Mine will be calibrated on the bench with a 3' lever arm and a known amount of hydraulic force applied to it.

tomcat,
i always enjoy reading your posts. you always have some creative ideas.
just my opinion here, but i think it would be a very tough challenge to measure propshaft torque accurately. i agree with insptech about reading the coupling or somewhere on the driveline before the verticle shaft in the drive. in my opinion the propshaft has too many variables that would make accurate calibration very tough. critical speed harmonics, propshaft bearing wear, gear wear, oil viscosity, prop cavitation, different gear cuts, 90* gear loading and trim angles just to name a few things. sorry for the negative input on a good idea:( i don't know if you were going to read millivolts but that is another can of worms underwater.

So if we can't measure propshaft TQ and HP then how about the next best thing and engine or drive input shaft TQ and HP. Loss thru the drive is a constant not changeable.

marty,
i agree with you about parasitic horsepower losses being constant. my thinking was just that calibration that far removed from the power source would make the calibration under those conditions extremely difficult. if the input shaft was read, the horsepower loss could be calculated into the formula and still get accurate propshaft horsepower data since the loss remains constant.

Dynos
 

Have been looking at your post and Know that I am not in your league as Knowing about building a onboard Dyno,However you might care to check with Boostpower In Newbury park Ca, He has a dyno that hooks up to the prop-shaft and he uses it for fine tuning in his shop, plus for R&D on their race boats. Like I said I know very Little about this subject, but If this little word will help, Great!!Otherwise, Please forget that I even Mention it!!;) ;) ;)

eliminator219

Why would the L&S concept work on a Stellings box, but not on an IMCO box? Both boxes use a driveshaft. Why can the magnetic strain be mounted to the driveshaft on the Stellings and not on the driveshaft on the IMCO?

I haven't priced the L&S, but if the unit requires having the transducers plugged into a computer, then has anybody checked to see if the unit could be run into a PALM device or a WindowsCE handheld unit?

This thread is very interesting to me.
(I do, however, see obstacles that may be difficult to overcome in measuring torque on any shaft inside the drive case - iron particles in the lube would have a definite impact on any noncontact magnetic sensing device).

M

It is not a magnetic strain guage. The strain guage needs to be bonded to the shaft.
The IMCO shaft is a long input shaft for the drive and it must slide thru the gimble to the coupler and if anything is mounted to the shaft then it won't be able to slide thru the gimble.
The Stellings box has a coupling flange inside the box where the standard length drive input shaft goes into and then a mini drive shaft connects to the engine crank. It is on this coupler that a strain guage can be permanantly mounted.
The signal from the strain guage then goes to a contral box and is processed and it's output is a 0 to 5 volt signal porportional to the TQ and HP and can be displayed on a digital readout, sweep guage or to a data logger. My device can meassure the TQ of any shaft provided that the shaft is accessable and at least 1" length of shaft to allow "organization of the TQ" thru the metal.
The L+S prop shaft dyno requires the boat to be out of the water and connected to water supply. It can't give real time numbers on the fly. Boats and propellers and really just a water brake- think about it.

Marty


The strain guage I saw was on a superflow dyno a few years back. It looked like an exaggerated "S" made out of aluminum and mounted between the frame and the pump. I suppose the best way to mount it would be to float one side of the motor and mount the strain guage on the opposite side. My question to all is if the engine is mounted in the conventional fashion can this guage be used as a motor mount/load cell even though the motor is mounted ridgid? My thought is the strain guage should be able to be calibrated for this

? So it measures the pressure on the motor mounts and by some corralation comes up with TQ of the motor?

Marty - I'm answering your question about shaft torque and prop efficiency in this thread.

Tieing the boat to the dock and taking the engine to WOT would cause a lot of slip. With the prop completely submerged the engine would likely come to an RPM less than Max ( like stall speed on an auto trans converter). Without any forward boat movement, the only water flow across the blades is what is drawn into the prop, so the slip calculation is high. This test might be useful if you were trying to set up a boat for a quick launch.

The shaft torque reading has as much to do with how well the prop hooks up as with how much torque is available from the engine at that RPM. So if the prop is not completely submerged, or starts sucking air, the RPM will take off and the readings will be meaningless.

Prop efficiency is a curve. Maximum efficiency occurs at a given angle of attack/slip % for a given prop. When you look at the equations, the biggest variable that affects slip and therefore efficiency is blade area. The simple explanation, (the only one I can understand) is that as blade area goes up too much, slip goes down, but too much power is wasted turning the blades through the water, so the net thrust is lower. Too little blade area, slip goes up, but less thrust is produced to begin with, never mind that it's easier to turn the blades through the water.

Your on board dyno can help explain why some props are faster than others. The "right" prop for speed is the one that generates the most thrust. That means it is just big enough to absorb all the engine power, with the least amount of blade friction. It often seems that spinning a bigger prop slower is the fastest setup, even though the Max RPM is reduced. A prop that converts 80% of 500 HP to thrust would be faster than a prop that converts 70% of 550 HP to thrust.

As an example, I can go 78 mph with 29 Mirage Plus props at 4300 RPM, and 78 mph with 26 Bravo four blades at 5000 RPM. My engines make a lot more power at 5000, but the extra blade soaks up that extra power. My dyno sheets says 413 HP @ 4300 and 483 @ 5000. Assuming the 29 Mirage is 80% efficient (maximum efficiency achievable in most props), the 26 Bravo is only 68% efficient. It is much better at getting on plane, but after that the extra blade is a liability. These results are prior to raising the drives 2". Now I run 82 mph at 5000 RPM with 28 Bravos, but I want to test a pair of 29s again. Since they are 15" diameter, bigger than any other Mirage prop, they may still bite well enough to prevent blowout at the higher drive position.

I know what you're going to say, it's torque, not power, that is important. But I say it's max prop efficiency (more thrust out for every HP in) that counts. Build the damn thing and settle this argument once and for all!;)

CMarty

Been looking at the L&S site. Appears that, as you say, they use a "normal" strain gauge buried in a tailshaft or something similar. The noncontact aspect of their sender works off of inductance and power source for transmitter same (via AC current from some sort of switching power supply). Cool idea.

I also see why you say the setup won't work on the IMCO box as it sits. No room for junk near the Ujoints. Could probably be made to work on the coupler end, though, with mods.

I am liking this whole realtime dyno stuff. Too cool.

About the only way I can see to get it on the stock brave is to use a 'mini' cutdown 'stern jack' and have it housed inside it. Actually L+S is impenging on my electrical engineers pattents. They don't have it quite right and they don't make any money so he is not going after them.

Tomcat

Speaking of props, what about a four blade with the two opposite blades cut short to help get the boat on plane, once its up, their out of the water an the larger blades take over?

JSV - Interesting idea. I'm not sure if it would work any better than reducing the diameter of all blades, but it sure would look different. The outboard guys have this all figured out. Adjustable blade area (hydraulic jackplate).