## Mechanical seal for ROV thruster

I am designing an ROV thruster to be used in either salt or fresh water at depths to 100 meters. I am trying to find an appropriate mechanical seal to seal the propeller shaft.
This is for my own personal use so I'd like to keep the costs somewhat reasonable. I have been trying to find a resource...online manual, sales brochure or textbook that can help me find a proper solution. So far..no good.
Any ideas for something that can help me solve this problem would be appreciated. Heck...as many design challenges as I have had with this project, any input on resources on ANY facet of ROV design will be appreciated.
Bill Black
landlocked engineering student
Going coastal in 2006

 PhysOrg.com science news on PhysOrg.com >> Front-row seats to climate change>> Attacking MRSA with metals from antibacterial clays>> New formula invented for microscope viewing, substitutes for federally controlled drug
 Recognitions: Science Advisor You may want to look into Chicago Rawhide's lip seals. The usage will depend greatly on the shaft diameter and surface speed. Is there a chance that the pressure is going to be the same on both sides of the seal? I would worry about having a huge delta P across the seal. However, they do manufacture lip seals with spring reinforcements that may be able to do what you are looking for. Take a look here: http://www2.chicago-rawhide.com/catalog_pdf.htm Let us know what you think and talk to a rep. If lip seals won't do it we can see if we can come up with something else.
 Blog Entries: 7 Recognitions: Gold Member Homework Help Science Advisor Seals are most easily defined by providing the following: 1) Type of seal (static or dynamic) (seal on stationary/rotating shaft or face, or reciprocating) (I'll assume this is a rotating shaft?) 2) Pressure differential to be sealed. (Assume 100 meters water pressure?) 3) If dynamic, aproximate rotational or reciprocating velocity (What is RPM?) 4) Aproximate envelope dimensions available (Please provide) 5) Fluids to be sealed (Assuming water/air?) If you can provide this type of information it will be much easier to determine the type of seal and potential manufacturers. I could also give you a simple seal design if you're interested in doing it yourself.

## Mechanical seal for ROV thruster

I'm assuming that you want: dripless seals, shaft spins less than 5000 rpm, 100m depth.

There are a few companies that make graphite faced mechanical seals, "stuffing boxes". The most common is at www.shaft.com. Look at this site for a visual. I've seen many of these. They use the elastic nature of a rubber bellows to seal the two graphite seals under light pressure. These seals surround the shaft like two large washers placed face to face. I have seen some designs with spring reinforced bellows.

Your design would have to deal with variable pressure inside the bellows (which is open to the water). The initial seal pressure at surface is a fixed value. It is relatively light. Your design would have to handle 100m of head, thats roughly 150 psi. Applied over the area of a 1 inch shaft seal (assume 2 square inches of seal area) gives a maximum of 75 pound force on the seals. That's too much...I would expect seal failure.

You would have to design a variable loading system on the bellows. As the ROV descends to greater depths the force on seal should remain constant therefore the bellows/spring force would need to decrease/incrfease (depending on how you orientate the bellows/spring). You could engineer a device that changes the loading as the ROV descends. That is the way I would first approach it. I am certain that there are other ways to go about it. Perhaps there are materials that could handle a high constant loading, especially for limited use.

 Hi there and thanks for the replies! I am proposing using a .375 polished 303 stainless steel shaft. The end cap on the thruster will be a minimum of 2.25 inches O.D. The casing will be aluminum tubing. I am planning to use Delrin (acetal) to construct the endcap that houses the bearings and seal. Shaft will have a maximum RPM of 1000. Fuilds are sea water and air. So..dynamic seal- pressure differential of 150 psi..if that's possible. I understand that some companies use an oil filled thruster for depths in excess of 1000 meters, but I don't feel like that is something I should attempt on my first iteration. Fred..I'm looking at the Teflon seals running on a polished stainless 303 shaft of .375 dia. but I fear the Teflon seals might be a bit more pricey than I am prepared for. Any idea on costs? Pete-When I click on that link- I get a porn site. Could you repost that, please? More later--I have thinking to do!
 Sorry about that Bill and to anyone else who hit that site. The actual site is www.shaftseal.com. I suppose if shaft is a porno site then shaft seal would be a contraceptive site A week never goes by in my high school physics classes where physics terminology gets some giggles and laughter from some students. Lip seals are common on prop shafts that are very short or have a larger diameter versus length ratio (rigid). Lip seals are very vulnerable to radial play or flex in a shaft. In inproper applications or with bent prop shafts they can act as a concentric pump and perform a function that is totally opposite to what is desirable. A 0.375 inch shaft is small, but this all depends on the loads/forces on the propeller? Also, how close is the bearing to the prop? The moment created by the prop loading should cause less than a few thousandth's of flexure on the shaft. If you do can use lip seals make sure you use a covered seal, they call them "fish-line cuttters in the marine industry. They protect theseal lips from fouling.
 Blog Entries: 7 Recognitions: Gold Member Homework Help Science Advisor I think Pete has pointed out some very important considerations, and I'd reiterate his comment on putting the seal as close to the bearing as possible and ensuring the two are centered. The only comment I'd have is on the suggested seal. I suspect the shaft seals pointed out are only for boats, and not submarines that need to seal at 150 psi. The problem that particular seal will have is it has a rubber bellows that will collapse onto the shaft and wear. If it were a metal bellows or was reinforced in some way, then it would be ok. Supporting it on a tube might work, but 150 psi on that bellows may still be too much. On the good side, the seal is a face seal as opposed to a shaft seal (ie: it seals on a face as opposed to a shaft). This means you can get lots of wear out of your wearing component before it goes out on you. Also, they use a graphite filled material which wears well. Graphite is an outstanding additive for seals. The only other comment would be that it looks rather expensive, but that depends on what you want to pay for something. I've got the impression you're only going to be using this ROV on occasion, not 24/7 and not without pulling it out and inspecting it every few hours. If you want a real, leak tight seal that's dirt cheap, use a conventional O-ring. Use the thickest one you can get for that size, a -204 should work best. I'd suggest polyurethane if you can find it, if not use Nitrile. Lube it with grease and I'd guess you can get 10 to 100 hours out of it. Make the seal easy to pull out and replace, that way you just toss out a 10 cent O-ring and replace it ever so often. If you need help designing an O-ring gland, Parker has something called "inPHorm" here: http://www.parker.com/ead/cm2.asp?cmid=2550 O-rings generally aren't used to seal a shaft, but if you're only going to run this a few hours a day for a few weeks at a time, they're the cheapest way to go. If this is to be used for an extended time, such as a commercial exploration vehicle that may be underwater for 12+ hours a day for months at a time, then a version of the face seal as suggested by Pete would work best. It should be pressure energized and probably use a graphite filled Teflon. Other filled Teflon's are also worth trying. I wouldn't use a lip seal (ie: U-cup shaped seal). I've never had much luck with them primarily because of the need for a very clean and highly polished surface to seal on.
 Recognitions: Gold Member I can't help with your original question, but I was wondering if you can clarify what you mean by "end cap" housing the bearings and seal? There's a potential problem with Delrin if there's any side-load on it.
 Hi Danger...I plan on making the body out of 6061 Al tube, with a piece of Delrin machined to the same OD with a lip that slides inside the tube.An O-Ring should provide a good seal between the thruster body(tube) and the Delrin end cap (bearing and seal housing). I am not married to the idea of using Delrin, I just happen to have quite a bit of it. Until I get this bearing/seal part hashed out, everything else is subject to change. Spinning a propeller with a DC motor is no great trick, but keeping it functioning @ 100 m. is making it a LOT more challenging. Thanks for your reply..and please...any discussion helps the process!

Recognitions:
Gold Member
 Quote by BillBLack any discussion helps the process!
Okay, I'm still a little shaky on understanding the setup, mostly because this isn't an area that I'm familiar with. If you could possibly post a diagram, it would help immensely. It sounds as if my questioning of Delrin might be appropriate, though. My experience with it is as a locksmith. The 'A' series Weiser/Falcon/Norlock doorknob sets use a Delrin bushing between the shaft that the knob is attached to and the collar that it passes through in the mounting plate. With almost any doorknob, there is a repeated pressure in a roughly 45° downward direction (to which side is determined by which hand is used) when it's turned. Delrin 'flows'. When pressure is applied in one direction like that over an extended period, it gets squished away so that the high-pressure side gets thinner and the mass migrates to make the low-pressure side thicker. Then the stupid thing rotates and gets the fat side wedged in the part with the smallest clearance. If you're machining it, though, the material is probably quite a bit thicker than the few thousandths of an inch that I'm referring to. Its self-lubricating quality is very attractive.

 Recognitions: Science Advisor Bill, Not too much can be done on the material side and cost, although, I would think that there's not much need to go into teflon for your application. You'll have great cooling of the seal and you're only sealing water. A general Nitrile type of elastomer will be fine for this. I would agree with Q's suggestion for using an o-ring if you don't want the lip seal. BTW, proper installation of a lip seal will prevent any shaft deflections from being an issue. Parker has a section on the proper gland design and selection for rotary seal applications. You are lucky in that the surface speed of your shaft is pretty low, only about 98 ft/sec. It is recommended to actually use the smallest cross sectional area o-ring possible. However, in your speed range it doesn't really matter. You would have to put a step down on the shaft OD to accomodate an o-ring though. They do not have an o-ring with a Ø.375 ID. You could go with a -012, -110 or a -204. It would all depend on the amount of room you have for the gland. You'll want below a 10% squeeze. Also, don't forget a healthy chamfer lead in on the shaft for assembly.
 Recognitions: Gold Member This might be a kinda dumb question since I really don't know what's available out there, but have you considered just buying a waterproof motor and leaving the thing unsealed? You could then isolate you control/instrument stuff in a sealed housing, and use pre-sealed cable or maybe magnetic connections between your servos and control surfaces. Just a thought, since you'll have to seal those anyhow. Alright, I'm off to work now. See you in about 11 hours.

Recognitions:
 Quote by Q_Goest Hey Fred, just a few thoughts, - The -204 is actually meant for a .375" shaft. There are "industrial" and "military" gland sizes, but the exact dimensions are less important than the amount of squeeze. The -204 ring is slightly smaller in ID than the .375" shaft which is what you want to ensure some squeeze on it.
I agree with what you are saying except one caveat: the -204 is actually meant for a Ø.375 shaft in STATIC applications. The -204 for rotary sealing is specified for a Ø.360-.361 shaft. The worry I have here is an even more premature failure of the o-ring due to stretch and an effect Parker refers to as the Gow-Joule effect. It says that if an elastomer is stressed (due to stretch) the resultant heat load from friction will cause the o-ring to contract even further. This supposedly happens only when the rubber is under tension. That is why I would suggest a step in the shaft at the end to accomodate the o-ring and get the squeeze from the gland design.

 Quote by Q_Goest - I know I've seen recommendations such as "use the thinnest one" so I thought I'd explain why I suggested the fattest one. In this case, there can be some significant missalignment (ie: a few thousandths of an inch or so) and a thicker O-ring can accomodate that missalignment slightly better than a thin ring. Second, this ring will be wearing out. Some material will be rubbed away as the shaft rotates. Having a thicker O-ring allows more material to be scrubbed off before the ring looses contact with the surface. Pressure helps to force the O-ring against the shaft, but when it wears so much that it doesn't touch the shaft in the unpressurized condition, it will begin to leak.
I agree with your reasoning on this one too. I think the surface speed of the shaft is low enough to not really worry about it, so it might as well help the most in covering the misalignment that will arise. I have no real experience with this aspect, so I am bowing to Parker's recommendations and yours too.

 Quote by Q_Goest The use of an O-ring here is subject to proper gland design. Consider where on the periphery it will be sealing. On the ID against the shaft is one spot, the other spot could be on the OD or it could be on the flat face depending on how you design the gland.
Admittedly, the only thought that came into my mind for this set up was an o-ring on the OD of the shaft and the gland on the ID of the housing. I guess looking into other options would be a good thing for me to do here.

 Recognitions: Gold Member Hi, I'm back; A few minutes of beer-encouraged thought have brought a totally different approach to mind, but of course it would have some problems of its own and I don't really know what I'm talking about. If you have already started putting stuff together, it's already too late to even consider this, but I'm going to lay out the idea anyhow just in case. If torque/speed requirements aren't out of line with it, what about using a hydraulic motor outside of the case with the propellor directly attached to it? If you also use hydraulics for your control surfaces, then all you have to do is seal the 'pod' containing the pump and electrical bits, which should just involve grommets and sealant where the lines and remote-control/feedback cables exit the housing. The hydraulic cylinders and motors are already sealed up to about 5,000psi (depends upon type).
 Recognitions: Science Advisor Most likely a weight issue...
 Recognitions: Gold Member Aha. Right you are. I never thought of that. Hydraulics are pretty dense critters. (And, of course, more complex and expensive than electrical systems.)