Oil seal between rotating and stationary surface?

[RodRico]

All,

I'm designing an engine that employs a rotor spinning between two stationary side plates. Oil is fed into the rotor near the center of rotation then exits near the rotor periphery. The oil return port has 0.14" diameter centered 2.9 inches from the rotor's axis of rotation. The maximum rotor RPM is 2626, and the oil flow rate is high but its pressure is low.

My original thought was to use o-rings, one with an outside radius of 2.83" and another with an inside radius of 3.6" to form a channel between the rotor and side plate with several return ports in the side plate allowing oil to return to the center axis of the rotor. Unfortunately, I calculate surface speeds that I believe to be incompatible with the use of o-rings that are compatible with hot (220F max) engine oil.

Radial space is at a premium.

Does anyone have any suggestions as to how best this interface should be sealed?

Thanks for any guidance you can provide!

Rod

 

[berkshire]

Do you have room to fit a knife edge oil seal?
B.E.

You are judged not by what you know, but by what you can do.

 

[BrianPetersen]

Don't seal it. Encase the whole deal in an enclosure and let the oil fly around everywhere inside and drain down to a sump.

 

[RodRico]

berkshire,

Thin O-rings are already driving rotor and side plate radial stack-up dimensions. Looking at the current o-ring gland designs, the widest is only 7.8mm at present. I'd really like to find a thin seal that doesn't drive the rotor diameter. Sadly, searching for "knife edge seal" returns articles on commando knives! I'll keep trying.

BrianPeterson,

No can do as there are reed values serving an air pump piston (for scavenge and charge) on the outter preiphery of the rotor. Beside's there are currently a series of o-rings dispersed radially along the rotor to allow passage of fuel and oil (one oil inlet near the center shaft running at 900 fpm, a fuel inlet pair running at 1650 fpm, and oil outlet pair near the periphery running at 4300 fpm) between the rotor and the side plates, and I can't let them all bleed through, so I have to solve this probem.

All,

I'm currently looking at creating cylindrical protrusions from the rotor into the side plates to allow use of shaft lip seals good to 5,000 fps as well as low profile labyrinth face seals.

Rod

 

[TheTick]

http://www.skf.com/us/products/seals/index.html

 

[BrianPetersen]

Feet per second? I sure hope not! Feet per minute, perhaps.

Your surface speeds are within the range of a normal crankshaft end seal, which is a lip seal. Normally there is no significant pressure across them but they will handle a little bit. They take up a fair bit more space than what I suspect you are looking for. I don't think O-rings will survive (or seal properly).

 

[RodRico]

BrianPeterson,

You're correct; I meant feet-per-minute not feet-per-second and have fixed my earlier response accordingly.

I think I may have found a solution using pairs of concentric lip seals such as the Parker FlexiCase CMN Series seals which are good to 6,000 fpm. These type seals are, however, intended for use sealing a shaft not one rotating plate adjacent to a stationary plate, so I may have to be clever and use concentric pairs of seals riding on surrogate shaft surfaces as shown in the concept drawing below. I'm still working my way through it to confirm it can be made to work and, if so, how much space it will consume.

Rod

EDIT: I already see the first problem... the top seal as shown would need to be an outside seal, and I don't think they're available in the FlexiCase Series. Hopefully they're available in *some* series.

.......

 

[RodRico]

The Turcon Varilip PDR Series may work better. Note they go out to 170mm (6.693") in the metric table but only 4.125" in the inch sizes. I particularly like the "Fluid transfer" arrangement as it may use less radial space (I have plenty of axial space). They sell these seals for both inside and outside sealing, so they also support the arrangement I showed in my prior comment.

 

[berkshire]

Rod when I said knife edge seal , I was thinking of this kind of rotary shaft seal,

https://www.netseals.it/EN/Catalogue/Rotary_shaft_seals

However I do not know if you can find one small enough for your application.
B.E.

You are judged not by what you know, but by what you can do.

 

[drawoh]

RodRico,

Try searching "gland seal". There are magnetic couplings out there, used to eliminate seals on small pumps.

--
JHG

 

[RodRico]

All,

This solution is simple and low cost yet meets all requirements and consumes less space than the o-ring solution.

In this approach, standard seals which are normal stationary and sealing a rotating shaft are instead rotating around stationary circular manifolds extending into the rotor from the stationary side plate. I don't believe the approach violates any of the functional requirements or capabilities of the seals. Though I flip one seal relative to the other, I see no mention of sensitivity to rotation direction in the selected seal's literature.

The center illustration shows the preferred arrangement in which the side plate manifold is most rigid, identical seals are used, and the rotor cavity is simple. This configuration, however, requires additional radial space to route oil between the seals. The arrangement shown in the rightmost illustration is less desirable due to lower rigidity, use of different seals, and complexity of the rotor cavities, but it incorporates space for axial routing of fluid between the seals and is thus the preferred for the outermost oil return seal.

The seals are from Parker's Flexicase CFN Series. The seals are a nice fit to my needs because they're fast (6,000 fpm), operate at high temperature in the presence of oil (up to 600F), and have a small profile (0.250 high x 0.175 wide). The largest risk in this approach is the requirement for 0.005 dynamic run-out on the "shaft." To mitigate this risk, I will switch from the 6202 main bearing in the rotor to a 7202 precision bearing and match-bore (stack and clamp the parts then bore all together) the central axis of the rotor and side plates. Another risk is in my interpretation of their specifications in terms of profile height and width and my assumption I can get these seals in any arbitrary bore and shaft diameter insofar as they're under 6.00 diameter and comply with the 0.250 cross-section limit. To address this risk, I'm contacting my local distributor to find out what parts are actually available for folks like me who need only a handful.

Does anyone see any issues with this approach that I've missed?

Thanks for your help!

Rod