Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Cyclonic Separator Design

Status
Not open for further replies.

jimbod20

Aerospace
Sep 8, 2010
75
I have some questions regarding cyclonic separator design and performance. I need to separate air from an air-oil mixture (~10 gpm air and ~2 gpm oil) vent the air externally and retain the oil for recirculation. I want to maximum vent air and minimize oil loss. I recirculate the oil for cooling/lubrication and need/want to minimize oil aeration. Dwell prior to recirculation is short. I believe a cyclonic separator would satisfy my design needs. Long story short, the cyclonic separator is an add on and as you can envision I have a very small space in which to install the separator.

My Questions.

I'm trying to design and evaluate a cyclonic separator without the conical section due to physical envelope constraints. Does anyone know of predictive methods for cyclonic separator design without the conical section? Can the forum members elaborate on issues problems I will encounter without the conical section?
 
Replies continue below

Recommended for you

Not much of an expert here other than you WILL get aeration. Can you install some sort of "settling" tank to allow it to bubble out??

Any chance you could condense it out??
 
Just have to ask, what is wrong with a can full of coarse steel wool?
 
Design a separator. W/o the cyclonic action, its just a separator. You don't give much clue to how small it must be.
Has to have a diameter large enough to slow the gas velocity down so it doesn't carry away the oil.
Can you use a ready-made coalesing filter?

What would you be doing, if you knew that you could not fail?
 
BigInch. Good point.

I have a cylindrical can, with a vortex finder and the option to create flow geometry from the exit of the cylinder. I have been testing this design.

The can is ~ 2 inches in diameter and length of the body is ~ 2.2 inches. I need a custom design to fit the installation.

MiketheEnginer.
The oil drains from the bottom of the cyclonic separator, dwells for ~ 5 seconds and then recirculates via a pump. I do get some separation via buoyancy.

1Gibson.
I have tried steel wool and various sized screens in the bottom of the cyclonic separator.

I wish I had some predictive methods to help guide design options.
 
Put the screens at the top in the gas flow area to catch the upward bound oil dropletts. Screens or steel wool, whatever, will give the smallest size than without. Experiment. Anything you'll find in regard to modeling will probably be for a much larger scale and probably won't scale down well.

What would you be doing, if you knew that you could not fail?
 
I should offer more information regarding the design I'm currently testing.

The inlet to the cyclonic separator body is ~ .375 by .750. The inlet is tangential to the inside of the cylinder. The separator body id is 1.8 inches (correct previous 2 inch number) and the length is 2.15 inches. I have a cover plate on the bottom of the separator body with a pedastle to reflect air back up through vortex finder. I have an annulus between the separator body and the pedastle to allow oil to drain. The vortex finder air exit diameter is ~ .25 inches.
 
BigInch,

I'm considering CFD modeling in attempt to gain insight into performance based upon constraints. Comments/recommendations/ideas?

Unfortunately, I am beyond my 20+ years of experience with a tight schedule trying to sort through separator design.
 
That's
Tiny .... even for tiny.

What would you be doing, if you knew that you could not fail?
 
You call it a "cyclonic separator". That confuses me a bit. Is it a seperator with a cyclone internal or is it a cyclone?

However, i would think that the conical wall are what makes a cyclone a cyclone? Now i dont understand "physical envelope constraints" do you mean that theres not room for conical walls? Does the cyclone becomes to long?

Best regards

Morten
 
if the inlet is angled and directed almost tangentially on the inside wall, it will make a cyclone of some kind, perhaps not so pronounced.

What would you be doing, if you knew that you could not fail?
 
I suspect safe to say the design is a 'cyclone of some kind, perhaps not so pronounced.' To clarify, I provide the attached picture. The unit I am testing is identical without the cone. I don't have the length in the installation for the conical walls.

 
 http://files.engineering.com/getfile.aspx?folder=ab3cd483-b70d-420a-b57d-950f6f84e3f8&file=AOS_schematic.png
Or search "catch can" used for automotive PCV systems, size may be appropriate but I don't think it would keep up with the flowrate at all.

Ballpark, 2gpm oil * (60 sec/min) * (128 oz/gallon) = 4.26 oz/sec of oil... Not sure this can be done with a typical screen or steel wool type (which most of the catch cans are.)

For air compressor applications, I'm sure it can keep up with the airflow but still the level of oil saturation is high.
 
4.26 oz./sec of oil through a "cyclonic separator" that tiny is overwhelming it. Don't put any sort of "pedestal reflector" in it, the density difference and resultant difference in normal force due to angular velocity causes the separation. You can't push air in 2 different directions. The suction / pressure difference has to do provide the motive force.
 
The only guess I have left would be to run it through a "turbine" of sorts, something like a sight glass meant to indicate liquid flow.

The air/oil hits the vanes of the impeller, hopefully oil sticks to the vanes and is flung outward due to centrifugal force (you would need to create a single "volute" that can gravity drain the oil) and a separate exit for air.

I could sketch it on a napkin but that's as far as it goes right now. Main point is, put something mechanical in the path that can't be bypassed unless it spins.
 
I think what I am testing now is more like a 'turbine of sorts'. The can or cylinder is a single volute. I bring the flow into the cylinder tangentially at the top and achieve some air-oil separation via density difference and angular velocity where the air exits at the center of the can upwards and the oil drains via gravity through the bottom.

I like the idea of a 'sight glass' test configuration where I might be able to improve or optimize air-oil separation via hardware modifications and visual test observation (ie diameter, vent tube size, length, etc). I can measure air-oil flow rate from the top of the can and reconcile air-oil out the bottom via continuity.

I was trying to use 'cyclonic separator' geometry/design as a starting point.
I might need to remove or modify the pedestal reflector at the bottom.

 
The only reason I referenced the sight glass was to illustrate the concept of putting something that spins inside, rather then just relying on cyclone action from the air current. Same reason I referenced a turbine, the idea was to put some sort of impeller inside, rather than just empty space.

But yes it certainly would help to see what is going on in there.

You may even want something that has multiple chambers, so there is no direct path from inlet to outlet. The seals/wipers at the edges of the dividers could possibly serve to deposit a flim of oil into the volute if they were flexible enough. The oil would collect on the non-working (back side) of the chamber wall. Imagine a very tiny revolving door...

You are going to have to do something very unique here, that much is certain.
 
Hi,
It is always nice to try to invent a solution. What I did in a comparable situation, was to contact a supplier in Germany, who has a good separation concept. I am not quite sure about the air/oil mixture you are referring to. It is merely meant for removing oil aerosols, but I didn't have the need to explore the working range of the oil concentration in the air. The separation method is based on evacuating the mixture through a micro fibre coalescing element, that collects the oil drops due to capillary forces. The air will be removed at the top of the tank, where the separation element has been built in. The collected oil will gather due to gravity and can be sucked out with a pump or by gravity. If you want I can supply you with the companies name and a contact person. Their scale of installation is far bigger, than what you want, but they might be interested anyway. Just to check it out.

Karel Postulart, The Netherlands
Nuon Power Generation
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor