Oil filled or Water filled or both?

[kenl]

I've just about finished my first thruster, it will be an enclosed thruster with a seal where the shaft exits the housing.

The reason to go enclosed is that I couldn't be comfortable with the motors running open in the ocean, not just the salt that doesn't get cleaned out but also the sand going through the motors.

I was initially just going to go down the mineral oil filled path, and may well start out that way. But I think there is every chance that the motor housing will get at least some water in it through the seal. So as I was working at my lathe I thought why not use distilled water and and or cutting oil? My main concern is would it be abrasive, as it is used to aid cutting.

I'll post some photos soon of the thrusters, I'm quite pleased so far.

[a_shorething]

Preface: I have no professional training. :0

I'm going to go out on a limb here and say that any mixing between the oil in an oil-filled motor housing and the outside seawater is going to be minimal if you have a seal in place. Granted, the inside pressure will be less and depending on the properties of the fluid you use it will actually be 'technically' compressible but I'm guessing not enough to cause any concern.

The only forces acting on it would be a)pressure and b) the spinning motor and prop shaft where it goes through the seal.

The pressure will be one-way on the way down and the other way on the way back up if it equalizes at depth. You could probably figure out based on the properties of the fluid and the volume of your motor's container how much that will actually be. I'm guessing it's not even a measurable amount.

The spinning motor will circulate the fluid inside the motor housing but since the pressure will be greater from the outside pressing in, there will be two forces keeping the oil on the inside: 1. outside pressure>inside pressure and 2. Nothing to displace the fluid that would presumably be trying to leak out (vacuum). If you turn a bottle upside down the water can't pour out faster than air can get in. The same will be true of your motor housing. With nothing to displace the oil, there is no way the oil will 'seep' out, especially to a higher pressure area.

Just my .02. (looking forward to pics of your design)

also, I think cutting oil is only used to prolong the life of your drill bits, it shouldn't have any abrasives in it. http://en.wikipedia.org/wiki/Cutting_fluid

[kenl]

Ok so I've assembled my first thruster and even given it a dry run.

[kenl]

I don't know where extra comments would appear i the above post so I've Started a new one.

I did end up filling this one with oil. (46 grade Mineral oil)

My initial concern with oil filled thrusters is from comments by Rover3D that expanding oil (due to heat) will crack any fully enclosed cases, such as ones with Magnetic couplings.

With that in mind I set out to have a housing that a relief mechanism (the oil seal) and reasonable cooling surface. (the aluminium tube)

So I ran the unit dry for only 2 or 3 minutes and the temperature increased quite a bit, I only potted the wires with bees wax as it was meant to be temporary and oil started to seep out as soon as the temp increased. This would have normally exited through the seal if potted with epoxy. (I think?) That's OK but as it starts to cool again it would suck in water (or air if back on board) that was why I thought of water soluble oils.

Another possible solution is to have a compression spring on the housing bolts so the unit can expand and contract with temp(slide on the O rings) , I've had a quick look on line for stainless compression springs, but no luck yet, I don't think spring washers would give enough expansion?

The fan is from an old PC, it was the CPU cooling fan. I've got some no name props on order from Hong Kong, but really like this fan, certainly compared to the cheap airplane props on my test rig. I may go to the tip (dump) tomorrow and see if I can get a few old PC's to recycle.

The bearing is a standard run of the mill skate bearing with the seal removed.

[rossrov]

Hi Ken.

Nice-looking job. Curious how you made the ends. CNC lathe/mill? The oil issue can easily be solved, but usually adds bulk to the thruster assembly. Put a hose barb of say 15 to 20mm OD in where the beeswax is (or machine another end), solder longer wires to the existing ones, and slide on a thin-walled silicon tube that matches the barb. Seal the other end of the hose with another barb, with the wires potted in. Once you've filled the assembly with mineral oil (just slide the hose back) and removed as much air as possible, put some kind of spring clamp along a section of the tube. You now have a positive pressure bias on the inside of the thruster assembly.

Ross

[kenl]

Thanks Ross, I used a NC router to machine the basic shape then did the rest on a Lathe and Drill press.

NC1.jpg (348.66 KiB) Viewed 6477 times

Today I tested the thruster in water and the case didn't get at all warm. However, I did run it pretty hard for about 20min at 11amps to the point where the ESC overheated and cut out. (I measured a thrust of 1000g in fwd and 800g in rev) The wire going through the potting did get very hot and melted the bees wax, also the oil seemed to effect the insulation on the motor wiring, by making it very soft.

I have seen a few of the thruster posts with the hose and barb but was trying to avoid it due to the added complexity and the bulk, as you pointed out. My thought was to use very fine hose, 1-2mm id and link them all to a spring loaded compensation ram, open water and spring on one side, and oil on the the other.

When I reassemble it I will solder heavier wires as close as I can to the motors to the protect the potting epoxy from undue heat.

I will also fill the the thrusters with cutting oil and distilled water, just to see how it goes? I fill the the housing from the top before pressing the oil seal in, I'm pretty sure there is no air left in there. In the third picture you can see the small holes I have drilled, they are behind the seal. I insert a syringe through one hole and down through the bearing race so as to fill from the bottom up.

Ken

[a_shorething]

Ah yes... the other factor: heat.

There's the rub.

Well, I think one answer to some of the overheating problems that people are having is a software fix. You can work out what your specific configuration can handle and then dial things back from there in your control system so that you never feed it more than it can handle with a safe margin for error. Assuming of course that there is some number that doesn't push everything too hard but will also give you some amount of thrust that will work for you.

Depending on your control system this can also give you better control over the thrusters as well. If you 'shorten' the range of power that your software is using you give yourself more resolution within that range.

Here's an example:
Let's say your control input has a range of 0-100 in forward and 0-100 in reverse. 100 is way too much and 0-20 just isn't enough to make the motor turn.

If you find out the highest amount that is workable for prolonged periods of time without overheating your ESC is about what your current system tells you is 85 for example, your usable range is actually from 20-85. In your software you 'map' the range of motion in your control input to the new range and suddenly the minimum (non-zero) input becomes what used to be 20 and the maximum it can ever put out is what used to be at the 85 point, so your new range actually has two things: 1. An upper limit that is guaranteed not to fry anything and 2. More resolution because you don't have the 'dead zone' of 1-20 that used to be there at the low end, and the full range now covers a shorter range so the amount you move it will be more 'sensitive' and responsive.

[kenl]

I doubt I would need to run the ESCs that hard in a real dive situation. Most of the dive video I've seen has the thrusters running only 50% of the time. I am using code written by Nick (njs552)and that does have some mapping for the topside control that I think does what you are referring to. LSYVal = map(LSYVal, 0, 255, 60, 185)

If there is enough room in my hull I will try to squeeze a cooling fan in above the ESCs. I also plan to fit an Aluminium rear bulk head, with decks welded to it that will have the ESCs and batteries mounted on them. Hopefully that will dissipate a lot of the heat in the hull in general.

But that's all still in my head, I haven't even got the 125mm / 5" hull pipe yet.

[kenl]

Since early December I have had the same oil/water emulsion in my motor housings, in that time I have had about 15 dives with the ROV, but none of them real deep.

Yesterday I opened up the housings to modify them and to see how things are going in there.

In three of the 4 housings the emulsion had split, into water and oil (like oil and vinegar) in the other it was still a white milky liquid. In all 4 of them everything else was fine, the motors looked like new and even the black steel circlips were like new.

Alot of the running of these motors has been on the bench, and when run hard they do excrete some fluid through the seal, I can only assume that air is sucked back in through the seal?

I also assume that salt water has penetrated the housing by compressing the same air, and then causing the emulsion to split..

That assumption is based on some google searches referring to poor water quality and emulsions, though there was no mention of salt water.

Anyways, the upshot of it all is that I am happy with using (fresh) water soluble oil in my housings, as it seems to protect the internal components even after it has failed.

I am changing my housing so that where the shaft exits it I have 2 seals, one facing the open water and one facing the motor, hopefully this will help to prevent the vacuum forming in the housing in the first place. But I will need to provide some sort of pressure relief.

[scubersteve]

what did you end up using for propellers?
You have an impressive looking thruster housing there.