I think what I took away with me from both this thread and this experiment, is that the theoretical ideal reduction ratio is not always necessarily the best in reality. For one thing, even the best reduction gear robs a certain amount of power. It may or may not be significant in an installation, and it can be optimized to some degree but never made to go away entirely. A less than optimal (by calculation) setup can sometimes be more efficient than one that is perfect from an engineering standpoint. Modeling formulae and software don't lie, but OTOH they do not always reflect reality. So, experimentation has its place alongside sensible predictions based on physics.
The final analysis was that the motor was unharmed, and the noise was a coupling setscrew grating on a keyed shaft which came loose due to vibration. The key fell out, the shaft backed out, much noise was made, I stayed drunk for two days and then waited a day for the rain to stop. I put the system back together and saw a best ratio of .7w/rpm, even though the shaft was severely misaligned. There was not enough adjustment possible without major work and I had other pressing matters, so I put the 2:1 gearbox back in the drive train and saw about .8w/rpm. About like before, but better than with the 3:1 gears.
Before reinstalling the 2:1 gearbox, I compared the resistance to turning by hand of the two units, both at the input and output ends. The 3:1 offered more resistance to turning than the 2:1, even from the output end. I think the 2:1 gearbox was simply broken in from use, whereas the 3:1 was new. Possibly with a bit of wear, the 3:1 gearbox might have been as efficient or more efficient than the 2:1, as suggested by reasonable engineering predictions. BUT the fact remains that in this case it delivered the worst efficiency while the logical worst configuration was actually the best. Without a standardized way of quantifying and applying the mechanical losses of the gearbox or a belt drive or chain and sprocket or whatever, a piece of the puzzle is missing. We can't just assume equal efficiency inherent with a reduction system, because it simply doesn't work that way in the real world. Components are not always ideal, nor are construction and engineering methods.
I have other things I must to do to the boat before I return to work. When I get home from my next job I plan to completely reengineer the motor mounting system and ensure sufficient range of motion to adapt to any forseeable configuration. The thrust bearing will stand alone on a thrust plate. The shaft between thrust bearing and motor will be broken and coupled with a splined shaft arrangement so that a small amount of axial shaft movement will have no effect on the motor, and motor mounting will not be as critical. I will ensure that I can couple the motor direct, or use the reduction gearbox, or a belt drive, without disturbing the alignment of the prop shaft, which will be adjusted for minimum resistance to turning, and locked into its adjustment. Just the weight of the shaft within the shaft log is enough to deflect the hose mounted stuffing box and allow the shaft to drag on the fiberglass shaft log. When the stuffing box is held exactly right, the shaft turns smooth as silk, and will freewheel for a couple of turns after a quick spin with the fingers. I imagine that makes a few watts difference right there. I want to couple a small brushed motor with a resistor network speed control and an ammeter in series with the motor, and find the sweet spot for the thrust bearing position. I think thrust bearing placement and adjustment will prove to be a dealmaker or dealbreaker, and help me to achieve optimum efficiency from whichever drive configuration I end up with.
I may run the motor with the 3:1 gearing for a couple hundred hours and see if the resistance to turning diminishes. I suspect that it might. It is in no way difficult to turn, but even a small resistance can have a significant effect.
One other thing I might try, since I have two identical 5kw motors in addition to the big motor, is to rewind one of the small motors with fewer turns of bigger wire, for a lower RPM/Volt constant. I may end up destroying a perfectly good motor. It may not even work. That's okay, because I will know one more thing that doesn't work, if it doesn't work.
I heard very sound and logical argument to go with a lower voltage. Out of economic necessity I am sticking with these 220ah golf cart batteries, and I do not wish to wire batteries in parallel. If I were serious about switching to 24v or 12v or whatever, I would probably go with some very large single cells, at least 1000ah. As it is, I think 48v is either optimal, or at least a necessary compromise. But winding a motor specifically to deliver lower RPM/volt, while certainly reducing peak power output, might improve direct drive efficiency at lower speeds. I have to look into that before I start tearing up the spare motor.
I am also considering building a motor. Twice as many poles ought to deliver half the rotational speed for the same electrical rpm. A larger diameter motor ought to deliver greater torque at low speeds. Or, maybe not. One thing I have discovered is that even very knowledgeable and intelligent experts, and the conventional wisdom, OUGHT to be questioned and verified with real world experiments even if they "can't possibly" work. The real world likes to play little tricks on us.
| Reply via web post | • | Reply to sender | • | Reply to group | • | Start a New Topic | • | Messages in this topic (30) |
No comments:
Post a Comment