Axial Swash Plate Feed Pump

[Nikhil Bhale]

I think performance of O ring seal will have different criteria for static seal and dynamic seal.
In static seal clearance can be more than dynamic type as there is less possibility of O ring squeezing through the gap.

regards
Nikhil

 

[Toymaker]

I think performance of O ring seal will have different criteria for static seal and dynamic seal.
In static seal clearance can be more than dynamic type as there is less possibility of O ring squeezing through the gap.

regards
Nikhil

Can you tell me what clearance values are safe for dynamic conditions ??

 

[The Ignoble Troll]

Dep

Can you tell me what clearance values are safe for dynamic conditions

Depends on durometer number and material. If at first you don't succeed, try a harder ring.

Single O rings, like you have, are common in hydraulic valve bodies and do not leak. The oil helps them slide and seal.

 

[The Ignoble Troll]

Check out

https://www.alliedmetrics.com/products/o-rings/metric/types/

 

[Nikhil Bhale]

I don't know about clearance values. But I have seen use of backup ring, generally made of Teflon to prevent squeezing.
I have seen O rings in pneumatic valve spool but never in hydraulic valve spool.

 

[Toymaker]

I used O-rings because I have them and because they're what I'm most familiar with,....BUT, this old tinkerer is certainly willing to learn some new Mechanical Engineering designs.

I briefly looked through the allied products pages Troll posted the link to, which left me a bit confused on what hydraulic seal would be best for my application. Because it will be quite challenging for me to mill a grove into the cylinder walls, my preference is use a seal that fits into a grove in the piston, same as my O-rings fit into.

It looks like a simple U-Cup might work ? If yes, should I use 2 seals, mounted back-to-back, one for pushing and one for suctioning in the water. The pumped water will never exceed 100 C,

 

[The Ignoble Troll]

I don't know about clearance values. But I have seen use of backup ring, generally made of Teflon to prevent squeezing.
I have seen O rings in pneumatic valve spool but never in hydraulic valve spool.

Get your hands on a BM series BLB valve.

3600psi with only one o-ring at each end. They probably also help to centre the spool as well as sealing it.


I don't actually know of a spool valve that I've dissessembled that lacked end o-rings

 

[The Ignoble Troll]

I used O-rings because I have them and because they're what I'm most familiar with,....BUT, this old tinkerer is certainly willing to learn some new Mechanical Engineering designs.

I briefly looked through the allied products pages Troll posted the link to, which left me a bit confused on what hydraulic seal would be best for my application. Because it will be quite challenging for me to mill a grove into the cylinder walls, my preference is use a seal that fits into a grove in the piston, same as my O-rings fit into.

It looks like a simple U-Cup might work ? If yes, should I use 2 seals, mounted back-to-back, one for pushing and one for suctioning in the water. The pumped water will never exceed 100 C,

I think you will be fine the way it is, but worst case, you might need to play around to find the right material for friction and hardness.

my 0.02$

 

[Toymaker]

And now I have 9 tiny brass slippers . So far, all is going smoothly.

Next up: make the end cap that holds one of the bearings and pins the swash plate in place, keeping it from rotating.

 

[Toymaker]

Yes!!! It pumps water !!! More testing & assembly still lies ahead, but for now I've verified that my 9 piston pump at least pumps water under no pressure.

I need to diss-assemble it to check for internal water leakage around the O-rings, and signs of wear or galling between the port plate and the valve plate as well as wear on the slippers and swash plate. If everything checks out, I'll mount it to the drive motor and proceed with pressure testing.

I need to polish out all those tool marks from my lathe chuck

Powered with a cordless drill,....pretty good flow rate, (which needs to be measured).

 

[Toymaker]

Blew the pressure side O-ring seal at about 60 psi; not at all what I expected would happen. The O-ring wasn't damaged, just pushed out of it's milled channel which is located on the bottom side of the valve plate, and seals against the bottom of the case/housing. The valve plate is not screwed to the housing, but is pressed against the housing by two strong springs (but maybe not strong enough).

Slowly closing the small valve reduces the pump's output hole size which causes pressure to build.

 

[peterl95124]

Just found this video on YouTube showing brass valve plates being lapped and polished. Not a very exciting or informative video beyond it confirming that at least some commercial valve plates are made from brass.

Lapping Brass Valve Plates

Of course this doesn't guarantee that brass valve plates will survive pumping water, meaning I will need to find out for myself.

you can't tell from the video 1) if this is really "brass" or is mis-named "bronze", and 2) what the valve plate is in contact with. what I'm trying to say is that brass-on-brass is not good (was not good for me at least), and maybe the plate in the video is intended to slide on steel (although I have additional evidence that brass does not make a usable bearing material for a steel crankshaft, I have to use bronze instead).

 

[Toymaker]

you can't tell from the video 1) if this is really "brass" or is mis-named "bronze", and 2) what the valve plate is in contact with. what I'm trying to say is that brass-on-brass is not good (was not good for me at least), and maybe the plate in the video is intended to slide on steel (although I have additional evidence that brass does not make a usable bearing material for a steel crankshaft, I have to use bronze instead).

Thanks for the "heads-up" caution.

At this point in my development I have perhaps 10 or 20 minutes total run-time on the pump, with that time being broken up into many short sections with tear-downs in between. So far, I'm not seeing any galling or other detrimental wear on either of my brass-on-brass surfaces, in fact, those surfaces keep getting shinier and shinier .

Edited addition: Truth is, I don't know if the parts I've made are brass or bronze; I purchased the stock from a local marine supply shop near the fishing docks, and I'm not sure the sales people at the shop know with certainty if their stock is brass or bronze.

I plan on posting pics of the valve plate's and port plate's contacting surfaces after I have more run-time, but for now I will post this pic showing the swash plate and slippers,....not the best view, but you can see how shiny those surfaces are.
You can see many little droplets of water on the oily surfaces.
First pic shows the piston's pulled out as far as the swash plate allows; the last pic shows the pistons pushed in to their greatest depth. Measured piston stroke is 0.270"; my design goal was 0.260",...close enough to do the intended task.

 

[Toymaker]

Progress update: I'm still trying to stop water leaking from what I know is past the outlet holes between the valve plate and the bottom of the case. Post#31 shows the valve plate with the inlet O-ring sitting where it should, and the outlet O-ring pushed out of it's channel. Below photo shows the inside to the housing where the valve plate sits. The two arc-shaped milled channels, right and left, in the bottom of the housing are the inlet and outlet ports. The valve plate is held in place only by spring pressure. Everything works good until pressure begins to build, which pushes the valve plate upwards far enough to allow the O-ring to be pushed out of it's channel. This problem is still unresolved.

In an effort to hold the valve plate in place more firmly, I increased spring strength, which had the unpleasant side effect of bending the piston lift plate,...that's the flat plate with 9 slots shown in the below pic. The new unbent plate is pictured below.

I've also learned a bit about the theory of operation of these pumps, such as, how do the port plate and valve plate stay together? Even on my tiny pump, the hydraulic pressure pushing the plates apart approaches 2000 lbs; surface area of the port plate is just under 2 sq in, so at 1000 psi, the force on the port plate is 2000 lbs. So, if pressurized water gets between the port plate and the valve plate, the pump's spring pushing those plates together would need something over 2000 lbs to keep them together, and that much force would be similar to what automotive disc brakes see,...meaning no electric motor could turn the pump.

The trick is to keep pressurized water from getting in between the plates, which is accomplished by polishing the plates to 0.00001", which is mirror-surface flat, which keeps a thin layer of hydraulic fluid between the plates which creates what's known as "hydraulic float", which is essentially surface tension of the hydraulic fluid between the plates. My big unknown,...will this hydraulic float phenonium occur with water?

 

[The Ignoble Troll]

Progress update: I'm still trying to stop water leaking from what I know is past the outlet holes between the valve plate and the bottom of the case. Post#31 shows the valve plate with the inlet O-ring sitting where it should, and the outlet O-ring pushed out of it's channel. Below photo shows the inside to the housing where the valve plate sits. The two arc-shaped milled channels, right and left, in the bottom of the housing are the inlet and outlet ports. The valve plate is held in place only by spring pressure. Everything works good until pressure begins to build, which pushes the valve plate upwards far enough to allow the O-ring to be pushed out of it's channel. This problem is still unresolved.

View attachment 157487

In an effort to hold the valve plate in place more firmly, I increased spring strength, which had the unpleasant side effect of bending the piston lift plate,...that's the flat plate with 9 slots shown in the below pic. The new unbent plate is pictured below.

View attachment 157488

I've also learned a bit about the theory of operation of these pumps, such as, how do the port plate and valve plate stay together? Even on my tiny pump, the hydraulic pressure pushing the plates apart approaches 2000 lbs; surface area of the port plate is just under 2 sq in, so at 1000 psi, the force on the port plate is 2000 lbs. So, if pressurized water gets between the port plate and the valve plate, the pump's spring pushing those plates together would need something over 2000 lbs to keep them together, and that much force would be similar to what automotive disc brakes see,...meaning no electric motor could turn the pump.

The trick is to keep pressurized water from getting in between the plates, which is accomplished by polishing the plates to 0.00001", which is mirror-surface flat, which keeps a thin layer of hydraulic fluid between the plates which creates what's known as "hydraulic float", which is essentially surface tension of the hydraulic fluid between the plates. My big unknown,...will this hydraulic float phenonium occur with water?

It does in dynamic seals.

I don't think it will in your pump because the film forming properties of oil are so much higher.


Maybe some high temp silicon grease?

 

[Toymaker]

It does in dynamic seals.

I don't think it will in your pump because the film forming properties of oil are so much higher.


Maybe some high temp silicon grease?

I tried using multi-purpose synthetic grease three or four days ago; works well at pressures below 40 psi, then appears to get washed away by the pumped water, and then leaks at 20 psi. I'm guessing that when pumping hydraulic oil, the thin oil film between the plates is constantly replaced, which of course doesn't happen with water.

After wet-sanding the brass plates with 2000 grit sand paper, I noticed that water beads up on the surface in the same manner as water on a freshly waxed car. Water has absolutely no adhesive filming properties on brass plates; with that in mind, I must reluctantly agree with you that water will never form a thin film layer between the two rotating plates, which is essential to the function of this design.

I believe I've made a very nice, small pump that will work quite nicely for pumping hydraulic oils, but a rather lousy feed pump that needs to pump 800 psi water

Hopefully, I can use most of the parts in a slightly different design,....I need to design a valve head that attaches to the block containing the pistons, then rotate the swash plate. Essentially a triplex water pump, but with 9 pistons instead of 3.

 

[The Ignoble Troll]

I tried using multi-purpose synthetic grease three or four days ago; works well at pressures below 40 psi, then appears to get washed away by the pumped water, and then leaks at 20 psi. I'm guessing that when pumping hydraulic oil, the thin oil film between the plates is constantly replaced, which of course doesn't happen with water.

After wet-sanding the brass plates with 2000 grit sand paper, I noticed that water beads up on the surface in the same manner as water on a freshly waxed car. Water has absolutely no adhesive filming properties on brass plates; with that in mind, I must reluctantly agree with you that water will never form a thin film layer between the two rotating plates, which is essential to the function of this design.

I believe I've made a very nice, small pump that will work quite nicely for pumping hydraulic oils, but a rather lousy feed pump that needs to pump 800 psi water

Hopefully, I can use most of the parts in a slightly different design,....I need to design a valve head that attaches to the block containing the pistons, then rotate the swash plate. Essentially a triplex water pump, but with 9 pistons instead of 3.

Can you slot in teflon slippers?


You did make a gorgeous pump.


What about adding steam oil and trying it with an emulsion?

 

[The Ignoble Troll]

I guess that at least this way you'll be able to correct the displacement error

Seriously too bad you're first go didn't work.

 

[Toymaker]

Can you slot in teflon slippers?


You did make a gorgeous pump.


What about adding steam oil and trying it with an emulsion?

The brass slippers & brass swash plate are holding up beautifully,...no sign of galling, scratches or wear; so yes, I could swap in Teflon slippers, but Teflon really isn't needed.

I might be able to come up with a soapy solution that would have the surface-wetting properties I need, but what impact would that have on the long-term corrosion to the boiler's copper tubing?

Even though it means more work and a fairly major re-design, I think the better solution is to make a pump that works with distilled water.

 

[Toymaker]

I guess that at least this way you'll be able to correct the displacement error

Seriously too bad you're first go didn't work.

If I were building this pump for a business, to sell for profit, this set-back would be aggravating,...but I'm retired and building this little pump is just a part of my hobby which I do because I enjoy building "toys" and learning new things along the way. Before I built this pump, I had no idea that the adhesive forces of a "wetted surface" were so large,...that's just incredible!!

And now,....well now I get to design and build a check-valve assembly that has 9 intake valves, 9 output valves, and is as small and compact as I can make it. I do enjoy a challenge