Old Hydraulic Gear Pump

[NJ Toolnut]

Hi All,

Since I've never ran a hydraulic log splitter and because I'm an absolute noob when it comes to hydraulics, please go easy on me (I learn fast)!

My neighbor gave me an old Commercial Shearing cast iron hydraulic gear pump from a dump truck. The only numbers on its data plate are 110356 and 2/86. The numbers GB1685 appear on the main casting. The inlet and outlet ports are located opposite each other. Inlet diameter appears to be 1.25" or so, outlet diameter is about 1" with 12 threads per inch. Splined shaft diameter is about 7/8", splines are ANSI 22-4/SAE B, 13 tooth, 16/32 diametrical pitch. The splined shaft sticks out from the 2-bolt mounting flange by about 1.25". I have not yet determined the mounting bolt circle size. Is anyone familiar with these gear pumps or able to suggest a resource from which I can information about them? I'd like to know displacement per revolution, max RPM, direction of rotation and max pressure. Regarding a potential application, I'm exploring the possibility of using this pump to power a log splitter from my Ford 1910 Compact Tractor PTO. I have about 20 hp available at the 540 RPM PTO shaft. I know I will need a gearbox to obtain required pump RPM, but I have a milling machine and a lathe, as well as access to a welder. Since this is not a two stage hydraulic pump cycle times may be relatively slow, but on the other hand this seems like a fairly large pump with a relatively high flow rate for a splitter application. Getting it free was huge for me, since I could never justify the price of a commercial splitter.

Thanks in advance for your comments!

Stan

 

[Farmerford]

Stan: Its Friday afternoon, I don’t want to start the next project (and don’t have too since I am both chief and sole worker), so here are some rambling thoughts on your project.

1. You can calculate the displacement per revolution of a gear pump with a formula on the Bailey’s hydraulics site once you pull the end plate and measure the gears. That formula is probably in many places on the internet.
2. RPM is harder to pin down. If the pump was driven by a typical truck transmission bolt on PTO it is still not narrowed much. It has been a while, but most of those PTO’s were stamped with a ratio on the side expressed as a percentage or multiple of engine rpm; i.e., an 80% or 0.80 means the PTO output shaft turns 80% of crankshaft speed. The problem is that the range of ratios I recall was very wide: perhaps from ¼ crankshaft speed to an overdrive of 2:1. If you have access to the parent truck you might find the ratio indicated on the manufacturer’s plates.
I would think a large tandem dump would need 15 gpm to move the ram at a reasonable speed. That number and the displacement of the pump would let you back into an rpm. I also think the 1.25” inlet is consistent with up to 20 gpm, but probably no more than that.
I think maximum rpms for gear pumps are limited primarily by the inflow of fluid; if the pump turns too fast the fluid will not enter fast enough to prevent cavitation, which of course is both inefficient and harmful. Cavitation will show up in vibration/noise at the pump and probably bubbles in the hydraulic fluid. I have two Prince PTO gear pumps; one for a 540 rpm PTO and one for a 1,000 rpm PTO; I note that they are exactly the same pump, only the pto coupling sizes differ. If it were me I would assume your pump could turn 2,000 rpm without cavitation.
I realize there are other limitations on rpm, including the bearings and side wear plates (or the side plates themselves if there are no replaceable wear plates). But my gut feeling is that 2,000 rpm is not too fast for them.
3. I think the most likely limitation on your pump’s output is pressure. Pressure affects the torque on the input shaft, leakage past the gear teeth and the end seals, and the load on the shaft bearings because the difference in pressure between the input and output creates the bearing load. It is probably an older pump and on a system that might have operated at a lower pressure to reduce leaks, hose blowouts, etc. I would certainly limit pressure to 2,000 psig, and perhaps even lower.

4. So here are some completely off the cuff, totally unsupportable, SWAG’s:
2 cubic inch per revolution pump
PTO on dump truck had 1:1 ratio, so the pump turned at crankshaft speed
Owner operator who is paid by the load would hold the engine at 2,000 rpm to speed up dump, so we assume the pump can tolerate 2,000 rpm
2,000 rpm at 2 ci/r= 4,000ci/min, or 4,000 divided by 231 = 17 gpm
7” diameter cylinder on dump truck takes 2 gallon per foot of extension
So it takes 10 gallons to extend cylinder 5’, which requires about 35 seconds for a full dump. Seems slow, but I won’t rethink at this point. Maybe pump is 3 ci/rev.
You drive pump 2,000 rpm with 4:1 gearbox from 540 rpm pto.
17 gpm at 2,000 psi is 19 hp pump output, so considering losses your 20hp pto will probably bog a bit.
4” logsplitter cylinder requires 2.5 quarts per foot to extend, so to move the typical 2’ stroke will take 5 quarts, which till take the pump 3.5 seconds to produce.

All this is worth exactly what it cost.

PS: I am rebuilding a 1931 Lodge & Shipley 14 x 30. It is in pretty good shape except for the bevel gear on the back of the apron that is turned by the carriage drive rod; somebody crashed it somehow and messed up the gear selector.

Send pictures.

 

[Leejohn]

I had a pump, just about what you guys are talking about, on my log splitter and run it at 540. The speed was good and it had good pressure. I would try it that way 1st, but run the pressure line the size it is now all the way to the valve then reduce. I only ran about 1500 to 2000 rpm on the tractor, the pump lasted around 15 year. The gears in mine were about 2" long and about 1 1/2" dia.

 

[NJ Toolnut]

Farmerford,

Thanks for your thoughts!

I'll look up the formula for calculating pump displacement from gear dimensions. I don't have access to the dump truck the pump came from, my neighbor told me the pump was attached to the front of the engine, not the transmission. I don't know if it ran the cylinder for the dump or only the plow that was attached to the front of the truck. Based on the diameters of the inlet and outlet ports and the research I've done, my best guess for flow is also somewhere between 15-20 GPM. I'm also going to assume the pump can tolerate 2000 psig and 2000 RPM unless I hear otherwise (I've emailed what is left of Commercial Shearing in S. Africa). I was considering a 4:1 chain drive, perhaps using #40 chain.

My lathe is a 1945 Monarch CY, 16.5" x 54" catalog size--actual swing is 18". It is in mint condition, and I've held <0.001 on diameter over 36 inches. Lodge & Shipley also made superb lathes. Is yours a Model B? You can probably have a replacement bevel gear machined, but you will need deep pockets. I'll post some images of the lathe and the pump.

Leejohn,

Thanks for your post! I think my pump gears are smaller than 2" long by 1.5" in diameter, but I will measure them. If I could run at PTO speed it would eliminate the chain drive design, machining and fabrication effort. One particular design concern is avoiding radial load on the pump shaft.

Stan

 

[oldnslo]

NJ,
A word of caution on taking the Commercial pump apart. Some if not all of them where pressure side loaded and the seals for the side loading feature can be a real PITA to reinstall especially when they fall out before you get a chance to look at them. I believe Commercial shearing was purchased by another company but don't remember who it was. Possibly Parker or called Commercial Intertech today?? They where used a lot on construction and commercial machinery. Possibly try calling places like Attica hydraulic exchange in Michigan or Hydraulic parts source in Michigan. Both of these companies rebuild hydraulic components and they may be able to provide you with some specs for your pump.

If your fluid level in the reservoir will be above the pump inlet you can run these at 540 RPM with no problem. If oil level is below the pump inlet at 540 RPM you may have trouble getting the pump to prime.

I would guess farmer ford is close on his sizing SWAG.

Mounting I would suspect is a standard SAE-B 2-bolt mount. Measure the mounting pilot diameter to determine which one.
SAE-A has 3 1/4" dia pilot
SAE-B has a 4" dia pilot,
SAE-C has a 5" dia pilot

 

[NJ Toolnut]

Oldnslo,

Thanks for sharing, and especially for the words of caution!

Regarding dissassembly (potentially avoiding it) someone on another forum suggested setting the pump up with some reducers and small diameter hoses for the inlet and outlet ports, priming it with fluid from a clean bucket and then measuring the volume delivered per revolution (while turning the shaft by hand) as a way to estimate displacement. The obvious advantage of this procedure would be that dissassembly would not be necessary. I wonder how accurate the result would be. In the absence of information from the manufacturer or elsewhere, I need to determine the relationship between RPM and flow rate.

Yes, Commercial Shearing became Commercial Intertech and was subsequently purchased by Parker. My experience has been that companies typically do not retain information about old models of equipment beyond ten years, but there are exceptions. When I called what is left of Monarch Machine Tool Company seeking information about my lathe, they were able to tell me who the original purchaser was, and were even able to provide the list of accessories ordered as part of the original purchase in May, 1945! I will try calling Attica and Hydraulic Parts to see what I can learn.

I currently plan on mounting the reservoir on the splitter above the pump, so no worries getting it to prime. I wonder whether flow would be adequate at 540 RPM to result in a reasonable cycle time. I won't be sure until I learn or estimate the displacement.

I measured the diameter of the mounting pilot and it is 4". SAE B confirmed.

Stan

 

[CNC Dan]

then measuring the volume delivered per revolution (while turning the shaft by hand) as a way to estimate displacement....
I wonder how accurate the result would be.

Make 10 turns and divide the result by 10. That should be more accurate.

 

[NJ Toolnut]

CNC Dan,

Thanks, good suggestion!

Stan

 

[NJ Toolnut]

Hi All,

I received a response Monday morning from Commercial Shearing in S. Africa, who forwarded my message to a US distributor (Swanson Industries in Morgantown West Virginia). The Swanson contact replied almost immediately to inform me that he could determine my pump model from a couple of external casting dimension measurements. I made the measurements last night and emailed them, then received his response this morning. The pump is identical to a Parker Hannifin Model P30. The nice guy at Swanson even provided a brochure for it. Rotation is CCW, displacement is 1.97 cubic inches per revolution, maximum pressure is 2500 psi and maximum RPM is 2400. The brochure has tables showing required HP as a function of RPM at 2500 psi and flow as a function of RPM. From these tables, required HP ranges from 14 at 900 RPM to 36 at 2400 RPM, and flow ranges from 6.5 GPM at 900 RPM to 19 GPM at 2400 RPM. From the data in these two tables, I was able to graph flow vs. HP. Basically, at my max PTO HP (20) the pump will flow about 9.8 GPM, and this occurs at around 1300 pump RPM. I extrapolated flow to PTO RPM (540--the tables did not go that low) and obtained about 3.5-4.0 GPM. My conclusion is that this pump could definitely be used to run a splitter if I gear up the pump from PTO RPM to 1300 RPM to obtain reasonable cycle times, but cycle times will not be fast even then.

I would appreciate any comments you may have in view of this new information.

Stan

 

[NJ Toolnut]

Hi All,

It gets even better: I checked the rotation direction of my PTO shaft and it is CW as viewed from the back of the tractor. This is the correct direction of rotation to mate up with the hydraulic pump through a chain drive, since the pump shaft turns CCW as viewed from the shaft end of the pump. In addition, I checked the owner's manual for my tractor and found out I have 28.5 PTO HP. I could use all of this available HP by driving the pump at about 1900 RPM, getting about 15 GPM flow at 2500 psi.

 

[Leejohn]

I had a feeling that pump would work just by the way you 2 were talking. Here is the way I did mine and the cycle time is good, I run my tractor from 1500 t0 1800 rpm. Just make sure you get the side load right.

 

[NJ Toolnut]

I had a feeling that pump would work just by the way you 2 were talking. Here is the way I did mine and the cycle time is good, I run my tractor from 1500 t0 1800 rpm. Just make sure you get the side load right.

Thanks Leejohn, that's an interesting design. Do you remember what size chain you used? How tight is it? This design will put some radial loading on the pump shaft even with some slack in the chain, have you used it a lot?

 

[Farmerford]

Stan:

Looks like you have about worked it out.

Regarding radial loads, in the gear pump the roughly half of each gear toward the outlet is under high pressure from the fluid being pumped. The other half of each gear toward the inlet side is under very low pressure; a low vacuum in most cases. If you draw a line between the centers of the two shafts, the parts of the gears on the outlet side of the line are under pressure from the fluid and the the parts on the inlet side are under very little, if any pressure. Of course, the pressure change around the gear from maximum near the outlet to zero near the inlet is on a gradient, but it is convenient to think of pressure on the outlet side and none on the inlet side.

The pressure on the outlet side tends to push the shafts apart and toward the inlet side. Therefore, the bearing load on the shafts is on the half of each bearing toward the inlet side, since the pressure is pushing the shafts back toward the inlet side.

It seems to me better to locate the pump so that the radial load of the drive chain does not add to the load on the bearing for the driving shaft nearest the sprocket. That means the drive side of the chain should be pointing away from the half of the bearing toward the inlet, perhaps even a bit angled in toward the outlet. Generally, the drive side of the chain should point toward the power sprocket in the same direction as the outflow of fluid from the pump case (assuming the ports are on the "outside" of the housing and not on the ends).

I admit that the effect of this arrangement is to add some radial load to the fluid pressure load on the other end of the driven shaft due to the moment of the chain force around the sprocket end bearing. But unless the sprocket is way out on the end of the shaft, that unfavorable force on the offside bearing should be less than the favorable force on the sprocket side bearing. Indeed, if the sprocket is placed very close to the sprocket side bearing, the unfavorable additive force on the offside bearing will be only a fraction of the reductive force on the sprocket side bearing.

At least that all made sense to me on the couple of occasions I have chain driven a gear pump that was probably not designed for a chain drive. And so far neither one failed, though I admit they don't get a lot of use. But since you cannot avoid a radial load in some direction, it makes sense to me to select the direction that logically does the most harm.

I am looking forward to pictures.

 

[Leejohn]

I used #50 chain and it is tighten just enough to take the slack out of it. Yes I have used it alot. That pump was put on I would guess about 5 years ago, the old pump I didn't have the radial load right and I builded it in the early 80's. That log splitter has split a lot of wood. The spec's you got should of gave you where the radial loading are. If you blow the picture up you can see the chain adjuster.

 

[NJ Toolnut]

...If you draw a line between the centers of the two shafts, the parts of the gears on the outlet side are under pressure from the fluid and the the parts on the inlet side are under very little, if any pressure. The pressure on the outlet side tends to push the shafts apart and toward the inlet side. Therefore, the bearing load on the shafts is on the half of each bearing toward the inlet side, since the pressure is pushing the shafts back toward the inlet side. It seems to me better to locate the pump so that the radial load of the drive chain does not add to the load on the bearing for the driving shaft nearest the sprocket. That means the drive side of the chain should be pointing away from the half of the bearing toward the inlet, perhaps even a bit angled in toward the outlet. Generally, the drive side of the chain should point toward the power sprocket in the same direction as the outflow of fluid from the pump case (assuming the ports are on the "outside" of the housing and not on the ends).

I admit that the effect of this arrangement is to add some radial load to the fluid pressure load on the other end of the driven shaft due to the moment of the chain force around the sprocket end bearing. But unless the sprocket is way out on the end of the shaft, that unfavorable force on the offside bearing should be less than the favorable force on the sprocket side bearing. Indeed, if the sprocket is placed very close to the sprocket side bearing, the unfavorable additive force on the offside bearing will be only a fraction of the reductive force on the sprocket side bearing.

At least that all made sense to me on the couple of occasions I have chain driven a gear pump that was probably not designed for a chain drive. And so far neither one failed, though I admit they don't get a lot of use. But since you cannot avoid a radial load in some direction, it makes sense to me to select the direction that logically does the most harm.

Farmerford,

Thanks for your interest, and especially for your analysis.

I fully agree regarding the forces at work on the pump interior, but I believe it may be possible to design a chain drive that eliminates external radial loading on the pump shaft. I think we can safely assume that Commercial Shearing took these interior forces into account when they designed the pump and specified the shaft bearings, but it seems unlikely they considered external radial loading on the pump shaft since pumps installed as intended would not be exposed to such loading.

If I can design a case for the chain drive that uses radial ball bearings to support both ends of the shaft attached to the small sprocket for the pump, and then add a frame bolted to the exterior rear of this case to which the pump is attached in turn, and then ensure that the two bores on each side of the case for the small sprocket shaft outer bearing races as well as the bore in the frame external to the chain drive case for the pump pilot are all concentric, would not such a design eliminate radial loading on the pump shaft? Such a design could use interior splines on the small sprocket shaft to couple to the external splines on pump shaft and transfer torque to the pump, and would feature oil seals in the bores on both sides of the sprocket case exterior to the small sprocket shaft bearings to prevent leaks. The side of the case closest to the tractor could be bolted directly to the back of the tractor using four existing large bolts, and could be bored to index to the pilot diameter of the PTO shaft boss. I think PTO shaft bearings are designed to be capable of withstanding the radial load imposed by the large sprocket and would not require outboard bearing support.

Leejohn,

Thanks for your response.

I agree #50 chain can easily deal with the torque involved, and its good to hear your second design was an unqualified success. I suspect #40 chain would also work, but it would probably wear faster. Another consideration around chain size is availability of sprockets with the correct numbers of teeth to obtain the RPM ratio I currently believe is optimum, as well as interior and exterior diameters that will work for the shaft diameters and case size. I will try to enlarge the image you provided to see the details of your chain tensioner.

I'm still not sure that it would be best to drive the pump at 1900 RPM to enable full advantage to be taken of the available 28.5 PTO HP. I don't want to labor the tractor too much, but I also want reasonable cycle times. Do you know how much flow you obtain from your pump at 1500-1800 RPM?

The cylinder I have (also from the same dump truck) has an external diameter of 4.5" and is 15" long. My woodstove takes rounds up to 19" long. I believe that this will work, but that for some hard to split chunks I may need to insert a block of wood in front of the end plate in order to allow the wedge to travel the full length of the chunk I'm trying to split. Does anyone have any thoughts about this situation?

This cylinder has either 1/2" or 3/4" female NPT connections for the hoses. It obviously worked with this pump on the dump truck, but I'm concerned about restriction. I also need to think about what valve makes the best sense, again considering minimizing restriction. Finally, I'm wondering whether it would make good sense to set relief at less than the rated pressure of 2500 psi in order to minimize wear. I really need to review all the informative design-related posts on this board.

Stan

 

[oldnslo]

NJ,
Any idea on what pressure the cylinder is rated for? Changing the pressure from 2250 to 2500 would make very little difference in the wear on your pump. You would only see this pressure when splitting hard wood.

If cylinder OD is 4.5" I would guess for operating at 2500- 3000 PSI that it is a 3 1/4" diameter bore. If 4" bore it would only be rated to 1500 PSI.

 

[NJ Toolnut]

NJ,
Any idea on what pressure the cylinder is rated for? Changing the pressure from 2250 to 2500 would make very little difference in the wear on your pump. You would only see this pressure when splitting hard wood.

If cylinder OD is 4.5" I would guess for operating at 2500- 3000 PSI that it is a 3 1/4" diameter bore. If 4" bore it would only be rated to 1500 PSI.

oldnslo,

Thanks for your response. It really got my attention and has given me more to consider!

I have no idea what pressure the cylinder is rated for, but it was used with this pump, which is rated for 2500 psi. Cylinder OD is definitely 4.5", but I have not had it apart to measure its bore diameter. I was not aware cylinders exist that are only rated for 1500 psi. How likely would it be that a 1500 psi rated cylinder would be combined with a 2500 psi rated pump? I suppose the dump truck's hydraulic system could have been limited by hydraulic relief to 1500 psi. Now that you've brought it to my attention, it seems more likely to me that the cylinder is rated for the higher pressure, with a bore of 3.25" and wall thickness of 3/4" but this is just a WAG based on possibly flawed logic. Obviously, bore diameter will profoundly affect tonnage the ram can exert--about 10 tons for a 3.25" bore vs. close to 16 for a 4" bore, assuming both are operating at 2500 psi. I previously (and ignorantly) assumed 1/2" cylinder wall thickness without considering/realizing it would need to be thicker to withstand 2500 psi.

I suppose I could measure the cylinder volume by filling it with oil and cycling it into a bucket, assuming I can exert enough force on the ram. I could then subtract this volume from a volume calculated using the cylinder OD and stroke length to determine cylinder wall thickness by difference. Alternatively, I could remove the shaft end cap and measure, but it looks like this would require a really large face spanner wrench. Finally, I could just use it and hope for the best. If it failed, would it be catastropic in nature or would it only involve seal failure? It certainly would not explode like a compressed air tank, but I guess it could rupture. Is it even possible to predict the failure mode? If it turns out the ram can only exert ten tons, is that sufficient force for a splitter?

You said something else that also got my attention, that I would only see this pressure splitting hard wood. I think (based on my limited knowledge of hydraulics) the cylinder itself would only see this pressure if it stalled while splitting hard wood, but the pressure side of the system upstream from the valve that controls flow to the cylinder will always see whatever the relief pressure is set at, correct?

Stan

 

[oldnslo]

NJ,
Cylinder failure: This would typically not be an explosive failure. What I would expect is either a seam splitting and leaking or possibly the cylinder wall ballooning. Most cylinders have a 3:1 safety factor so the odds of any failure are reduced.

"You said something else that also got my attention, that I would only see this pressure splitting hard wood. I think (based on my limited knowledge of hydraulics) the cylinder itself would only see this pressure if it stalled while splitting hard wood, but the pressure side of the system upstream from the valve that controls flow to the cylinder will always see whatever the relief pressure is set at, correct?"

NJ,
You will only see pressure when the flow is restricted so as long as the valve is held fully shifted the pressure to the cylinder will be very similar to the pressure before the valve. Only difference is line losses.

No, pressure line to the valve should not see system pressure unless there is a restriction to flow or you have a closed center valve. If you partially shift the valve or have to small of lines you will see build pressure in the inlet to the valve.

examples:
17 GPM in 3/4" ID hose is around 14 feet per second (FPS) velocity and will drop around 1 PSI per foot of hose or 10 PSI in ten feet
17 GPM in 5/8" ID hose is around 18 FPS velocity and will drop around 2 PSI per foot of hose or 20 PSI in ten feet
17 GPM in 1/2" ID hose is around 29 FPS velocity and will drop around 5 .75 PSI per foot of hose or 57.5 PSI in ten feet

This does NOT include and hose ends or fittings..

A good system design will always have some pressure on the pump to keep it lubricated. Gear pumps best guess is around 75 - 100 PSI. Piston pumps typically around 300 PSI

 

[NJ Toolnut]

Thanks for correcting my thinking, oldnslo!

Here are my take-aways from your response:

Cylinder failure: Low risk of failure, and low probability of harm to the operator (me) if it were to actually occur. Bottom line, go with the cylinder I have and hope for the best. The worst case would result in a ruined cylinder and spilled hydraulic oil.

Hydraulic circuit and system pressure: The only time the system is pressurized is when the open center control valve (the type required for hydraulic circuits that use positive displacement pumps like gear pumps) is actuated, either for forward ram movement or ram retraction. The rest of the time, the system pressure is governed by system frictional losses. This is because until the control valve is actuated, hydraulic oil continuously flows from the pump outlet through the open center and back through the filter to the tank. Pressure relief valves are usually included in open center control valves, but they function only when the system pressure rises after control valve actuation to a level high enough to overcome the force of the spring holding the relief valve closed.

Pressure drop as a function of hose size: Bigger diameter is better to limit restriction and frictional heating. Sharp bends should also be avoided when possible. Some pressure always present due to frictional losses but it is needed to ensure pump lubrication.

Do I now have this all correct?

Stan

 

[oldnslo]

Stan,
I would suspect that your cylinder is fine. If this is a welded cylinder Vs Tie rod style. Tie rod style has 4 rods with threaded ends that are used to hold the end caps onto the barrel. If welded style a possible way to guesstimate wall thickness is to look in one of the ports unless they welded onto the end caps? If welded onto the barrel you should be able to see the bottom of the fitting / ID of the barrel and measure that distance from the top of the fitting. Subtract the fitting extension height above the OD and this should provide a idea on the wall thickness. Once the wall thickness is known you could figure the cylinder bore.

Yes you have the system operation nailed down on how or when it would build pressure.

Roy the oldnslo one