Gear pump lobe design? 2

[salukikev]

Hi,
Since my 1st post worked out so well, here is a much simpler one (hopefully) for a completely unrelated project.

This item is a toy product, and I need to move a very small volume of fluid, very slowly, with minimal variation in flow.

Also, because it is a toy product, cost is king- and we're looking for some dirt cheap plastic solution.

In consideration of all the above, I've arrived at trying my hand at CNC'ing a gear pump. The only think I don't have is experience in the lobe design. I have created a simple cad file based on the concept illustrated here:

http://en.wikipedia.org/wiki/File:Gear_pump_2.png

Actually, here's an edrawing of my assembly so far. 4 parts is a pretty good deal, and if it works- it would seem to match up to the initial specs I outlined.

I've never built a custom gear pump before, and I imagine tolerances/materials and geometry all factor in pretty dramatically. Currently all lobes are tangent radii, not spline curves.
I don't need high pressure at all- so I was thinking I might eliminate some precision by using some flexible (silicone) materials- ie. the rotor

This pump will be flowing a substance very similar to mineral oil.

Thanks for any help!
-kevin

 

[salukikev]

Ok.. I think I did the attachment wrong- here's attempt #2, since I can't seem to edit my post.

http://files.engineering.com/getfil...8d-489c-8256-55fd5a87dea2&file=Minipump1.EASM

 

[MikeHalloran]

Take a look at a Jabsco or similar impeller pump.

Then consider that you can replace the molded impeller with sponge.

Tried it; it works.



Mike Halloran
Pembroke Pines, FL, USA

 

[hydtools]

Mike, I like it. A sponge pump element. Like a flexible tube pump, only simpler.

Ted

 

[MikeHalloran]

Credit is due my friend Dick Adams, who conceived the idea.
I merely built and tested it on a slow day.




Mike Halloran
Pembroke Pines, FL, USA

 

[salukikev]

Hi- Thanks for the creative & clever idea! Can you tell me any details about your specific geometry and/or materials to optimize it?

The concerns I have so far are:
1. Longevity- I'm assuming a urethane foam sponge rotating against a rigid (acetal) outer ring seems like it may wear out rapidly.
2. Bonding- I will have an axle driving this sponge element- if I bond to it, it will reduce the absorption to the core and may not like being torqued all the time.
3. Memory- when at rest this sponge element may take a set and then be ineffective at start up.
4. Pass through- This will be primarily regulating a gravity fed system operating at near 1atm, but I don't want my fluid to leak when its not running. This is an advantage I presume of a gerotor design, but I'm not sure that I can depend on surface tension to prevent leaking in a sponge based system.

The Jabsco impeller might simplify things though, and has some advantages over concerns listed above. This should be a very small pump- current design is about 1/2" diameter, and I anticipate the rotor moving at around 10rpm if that.

Would I need a shaped sponge element or just a blob of sponge?

All in all- it would be the same # of components as a Jabsco, as only the impeller material needs to change- is that correct?

Thanks again!
-k

 

[salukikev]

BTW- I somehow overlooked the /pumps section of this forum, which I suppose would be more appropriate to this question. Mods- please feel free to move this thread as you see fit.
Thanks! (& sorry!)
-k

 

[MikeHalloran]

My prototype comprised a cellulose sponge knife-cut ring epoxied to the remains of an actual Jabsco impeller that had chewed off all of its vanes. Urethane would have been a better choice, but I didn't have any, and didn't care. The sponge had to be wetted for assembly, and surely wouldn't have survived much dry running.

Bonding would certainly be a concern in any case. From that perspective, an impeller molded in a single piece would be better. (The Jabsco was not a single piece either; it comprised a thermoset hub, overmolded with a nitrile rubber shell and vanes). Polyurethane may be the only material that _could_, _maybe_, meet the conflicting requirements of a hard-ish hub that could be driven by a flatted shaft, and also some kind of flexible radial element with decent abrasion resistance and strength.

I have since had some experience with Jabsco bronze/nitrile seawater pumps, which are typically capable of lasting several hundred hours at 1000+ rpm at 10+ psid. At much lower speeds, pumping a compatible fluid with some lubricity, they should do very well. Material selection will still be a challenge.

Jabsco pumps typically stop the flow effectively when the shaft is stopped, unlike gear pumps, which have a relatively high internal leakage.

Gear pump leakage occurs in three places:
- tip to housing
- tooth to tooth
- both axial faces to housing
Tighten up any one of those clearances, and when you get it tight enough to minimize internal leakage, they get hard to rotate.

Back to sponge pumps: I'd expect an open-cell, or highly reticulated, sponge to exhibit a lot of internal leakage. A closed-cell sponge, much less so, but bonding will still be an issue.

The sponge pump is kind of a cute idea to play with, but I suspect you'll end up with a flatted shaft driving a hard hub with overmolded vanes of some flexible material. Maybe with lots of vanes; Jabsco pumps have a fairly strong noise signature at the vane pass frequency.

Or, if your head is low enough, maybe an Archimedes screw pump.

Resist any suggestion of using peristaltic pumps; they never fulfill their promise. Trust me on that.










Mike Halloran
Pembroke Pines, FL, USA

 

[salukikev]

great info, thanks! I like the sponge idea, but in the grand scheme of things, it sounds like this particular application is better served by a soft impeller. I'm adding that sponge idea to my cache of novel approaches for weird projects and I'm sure it'll get used someday! thanks!

back to soft vanes- I may prototype this with a urethane or silicone rubber part from a delrin mold. I guess that's best, but to deal with the clearances difficulties of a gearotor design, do soft or semi-soft rotors ever get used to deal with that? I guess its ultimately evolving into a vaned rotor anyway- but it seems like it would have a wiping/sealing action that would be desirable and maybe you could get a smoother flow due to spline curves/etc. My question about determining node geometry still remains on that one.

I'd guesstimate my head to be around 1-2" so none of my demands for this item are terribly challenging, really. I'm actually using a tiny peristaltic pump right now that I robbed from some bubble making toy. It's working, but not optimal.


*On an unrelated side note- I'm really starting to love this forum, just browsing through some of the posts and info I wish I'd started participating more years ago! Is there a chat room, or do members ever meet up?

 

[salukikev]

I'm starting to layout my micro-flexible pump impeller design, and so far I'm just creating my interpretation of materials I've seen. Any guide or tips in my arrangement or things to avoid/include would be appreciated. A link I've referenced recently is here:

http://www.liverani.com/ita/frame_girante_flessibile_uk.htm

I'm essentially scaling my pump down to a 1/2 diameter pump head, and pulling fluid from a reservoir the size of a thimble.

Thanks for any help!
-k

 

[MikeHalloran]

Chat room? Open to the general public? No.
Participate, share what you can, help when you can, and see what happens.

Meet up? In Real Life? Site management discourages exchange of contact information here, because the site is both public and global, and all of us get enough spam already. Communication happens nonetheless.

But seriously; would you want to hang out with a bunch of reprobates and swap stories about misadventures that most assuredly did not happen and were cosmically expensive to someone else? Yeah, me too.





Mike Halloran
Pembroke Pines, FL, USA

 

[btrueblood]

Star for the sponge pump idea, MH. If you ever decide to do the Charlie Sheen thing and take your show on the road, be sure to post it somewhere here.

 

[MikeHalloran]

Yeah, back to pumping with a sponge.
For lifting mineral oil an inch or two, I could see something as simple as a foam-covered horizontal axis drum rotating slowly against a vertical wall.

If it had to be a closed unit, then I might consider something resembling a gear pump, i.e. two foam-covered drums rotating against each other in a housing. That solves the problem of the foam rubbing against the fixed cam in a Jabsco style housing. Would one foam drum drive the other, or would they need to be synchronized with gears? I dunno.



Mike Halloran
Pembroke Pines, FL, USA

 

[salukikev]

I'm sticking with the Jabsco variation for the prototype at least. The advantages make it a clear win over the experimental sponge pump concept for the time being and the particulars of my application.

Disappointed to hear of the lack of chat/real life encounters- not since Battlebots have I had the privilege of hanging out with such a delightful bunch of reprobates- the stories were great too!

All the same, thanks for the help! I'm glad to have found this place, even if its only a virtual hangout!

 

[salukikev]

You know whats funny? I was messing with this system again last night, and I'm starting to think I could pull off the same trick I did for the water cooler (

http://www.eng-tips.com/viewthread.cfm?qid=297646&page=1)

(completely unrelated project) where I pump air into a reservoir instead to force out the fluid. Does that change anything? I guess it eliminates the sponge option and suggests a diaphragm, or maybe piston-pump, right?

 

[MikeHalloran]

Regenerative blowers can do many inches of head.

Mike Halloran
Pembroke Pines, FL, USA

 

[unclesyd]

You might be able to use this pump principal for your application.
I know the rotating disk will pump water in a larger size , 12"+, at low speed. This is from building a water heater that was a 12" dia disk inside a very low clearance pump type housing. By restricting the discharge were were able to heat and pump water into a tank with a 5" head in a circulation loop. The kicker on this arrangement was the power to run the pump/heater was a windmill.

http://www.discflo.com/home