Conversion from one to two steering pumps

[Zven]

Probably some of the members here are knowledgeable about hydraulics?
I seem to be out of my depth..!
Here is the story:

My motorboat has two inboard engines. One of them drives a hydraulic pump, which in turn drives the steering. It is a SeaStar “power steering” system, working approximately like in this illustration:



However, I believe there is a fundamental problem with this setup:
It works only when the starboard engine is running.

I like having full redundancy... and hence, wish to ensure the system can be powered also from the port side engine. (No need to ask why I have come to crave redundancy..!) It would be fine to have to turn some valves to "switch sides".

To this end, now I have had the same type of pump installed on the other engine too. The pumps are of this type - Bosch KS00001391:



Each pump has only one input and one output connection. The next question is, how exactly to connect the hoses?

The engineer who has done the work so far has simply connected the two pumps in parallel… using two “T” connectors. There are also two valves for each pump – one for its input and one for the output. These are simple ball valves which can be used to open or stop the flow (or perhaps leave it half open). Like this:



I have been told, I can now transfer control from one side to another by turning the valves...

But this is clearly not advisable. Since both pumps are powered by the engines via gears, they tend to be both running at the same time... and it would be unwise to let one of these pumps run while its input or output is blocked.

Is there any reason why I cannot just leave all the valves open at all times, and let the pumps run (mostly) in parallel? Or is it perhaps better to leave the valves for one ("duty") pump fully open, and for the other half-open?

Of course, sometimes the engines run at different speeds... so I worry that this approach could give unpredictable results. Or perhaps there is a risk of backflow through a pump which is not running.

Another person has suggested that more T-connectors and valves could be installed, to ensure both pumps always have some oil to pump... in a circle if nothing else. But I'm not convinced this is the way to go.

A fully automatic and more professional solution seems possible, by introducing an automatic combining valve at the output side of the pumps. I guess this would be a piece of kit which has two inputs and one output… bringing the flows together...

But the solution does not have to be fully automatic. And I wonder if it's wise to combine the maximum flow from the two pumps in the first place.

What would you do?

[EngineerRetired]

Honestly, I wonder why in the world boat manufactures are still using hydraulics for steering. There is one absolute rule for all hydraulics, they are going to leak...

With the electric rack and pinion systems available today (Even for some of the biggest field tractors on the market) I just don't understand sticking with hydraulics.

Better the devil you know sort of thinking?

[Nicholson58]

I have to agree with ‘Retired’. I am also a retired engineer and built a lot of machinery including custom hydraulics. The pump you have is marketed as a power steering pump. It is probably quite reliable. It has built in regulation and over pressure control. I am unable to find a flow schematic for it but I can see from the photo a number of appendages that look like typical circuit controls. These are necessarily there because the input speed and output demand vary a lot.

Our boat is 40 tons, 58 feet, (17 meter) and steering is manual. My autopilot is servo electric and connected to its own steering sector by a electro-mechanical clutch. I consider anything hydraulic on boats a point of future trouble and overly complex.

I presume your concern is more about loosing the engine that drives the pump. You could:

Drive it electrically, not attached to the engines. If you do this, keep the ‘extra’ pump as spare. You don’t care which engine runs. This may be the best option. SERIOUSLY, consider this. It’s hard to make a mess of an independent system.

To combine two pumps in one circuit automatically I think you can connect the output of each pump to the main circuit with a check valve so that neither pump can be back driven by the pressure out of the opposite pump. Both pumps would be attempting to supply the circuit and their respective controls should regulate what is going on.

The problems you might encounter:

Overheating of the oil. Since these things are designed to be dead headed then maybe this is not to worry.

There is a reservoir somewhere in the system. Once you combine the outputs and return lines there will be no control over which reservoir receives the return oil. Without seeing your total system and doing a bit of research I would suggest that you want only one reservoir central and equal distant from each pump with a good uncomplicated downhill run to each pump inlet.


To combine them not automatically you need two 3-way valves. Pipe one in the combined output and the other in the combined return. The pressure side will have the Pump outputs to the two side ports. The valve is operated to select which input is connected to the common output. The single return line is connected to the common port of the other three way valve. The two side ports direct the return flow back to the pump in use. You must always be careful that both valves are in the correct position or one pump will starve while the reservoir of the other pump overflows. Damage will happen to the starved pump. The unused pump will still be running always so both pumps see equal wear. You will not know if the unused pump is viable unless you periodically shift the use.

[Zven]

Thank you Nicholson58 for the input.

For information, there is only one reservoir in this system... as it's originally a single pump installation, and the two pumps are just connected in parallel. They are both working with the same reservoir. So I suppose, this part is not a problem.

Also, there is an oil cooler at the reservoir... so all of the oil from both pumps will be cooled.

Point noted about "equal distant from each pump".

I will probably aim to combine the two pumps into the one circuit automatically.

But a possible issue that some people have brought up is this:
When two pumps are pumping instead of one, there will be more total flow or pressure than the overall system is perhaps built for. Is this likely to be a problem?

[Nicholson58]

Originally Posted by Zven

Thank you Nicholson58 for the input.

For information, there is only one reservoir in this system... as it's originally a single pump installation, and the two pumps are just connected in parallel. They are both working with the same reservoir. So I suppose, this part is not a problem.

Also, there is an oil cooler at the reservoir... so all of the oil from both pumps will be cooled.

Point noted about "equal distant from each pump".

I will probably aim to combine the two pumps into the one circuit automatically.

But a possible issue that some people have brought up is this:
When two pumps are pumping instead of one, there will be more total flow or pressure than the overall system is perhaps built for. Is this likely to be a problem?

Please re read my post.

These are purpose built power steering pumps. They are designed to be self regulating and dead headed. I think this is not a problem. I am pretty sure you will need check valves.

You might try contacting the vendor. They might suggest a circuit diagram and specific components.

[RaymondR]

You just need a shuttle valve on the discharges of the pumps and between the shuttle valve and the pump in each pump a dump valve back to the reservoir.

The shuttle valve will direct the output of whichever pump has it's dump closed to the power steering cylinder and the flow from the pump with the open dump valve will just circulate back to the reservoir.

To switch pumps you just open the bypass on one and close it on the other and the shuttle valve will route the discharge from the closed bypass to the power steering.

One assumes that there is a relief valve in the output of the pumps in case you close both dump valves.

Or instead of a shuttle valve you could put a check valve in each discharge with bypass valve between each pump discharge and each check valve and then if you closed both bypasses nothing detrimental would happen except you'd have twice as much fluid going to the power steering cylinder.

[chris95040]

On tractors I’ve had, the hydraulics were always setup as ‘open center’- When the control valve to a gizmo is “off”, it dumps the oil directly back into the reservoir, no real load on the pump. It’s a little more complicated when multiple gizmos are off the same pump, but that behavior is maintained: when the gizmoes are off, the pump has it easy, pushing that oil right back into the reservoir.

There is always an automatic, high pressure bypass (relief) valve - for situations like where a ram has reached the end of its travel, etc, but you don’t want to be using that path except for short times. Pushing the fluid through the ~2500PSI relief means the pump is working it’s hardest, the oil is squealing and heating up.

On a cold day, seizing a ram and activating its hydraulic path is a popular way to trigger the bypass and get the oil warming up to operating temp.

In short, you certainly don’t want to dead-head (prevent ANY flow) out of a pump, you want a relief valve to make sure that’s *impossible*, and you want a way to direct the fluid directly back into the reservoir (no load, not even a relief valve’s load) when that pump is not in use.

[RaymondR]

Like this.

[Lepke]

I believe the pumps have internal pressure valves. I would run them as they are, valves open and monitor temps for awhile. They both need an oil flow to avoid heat damage.
On tugs and commercial fishing, I have run multiple pumps on the same supply line.
This pump is used on Volvos, so no economic replacement.

[RaymondR]

The potential problem you may have with runn ung the pumps in parallel is that if the system is designed around a single and you double the flow rates through it you may encounter heat build up problems.

The thing about pressure losses in flowing fluids is that they increase as the square of the velocity which means that if you double the flow rate the pressure losses will quadruple and pressure losses turn into heat.

So if you rig to run in parallel all the time you need to keep an eye on how hot things get until you are satisfied they are not getting too hot.

[Zven]

Thanks for the further tips here.

I have now learned that check valves are absolutely necessary when using two pumps...

Yesterday I tried paralleling them without check valves, just to see what happened (at low RPM).

While running only one engine, I found the steering to be unreliable with this setup.

Stupidly, I then tried starting the other engine at the same time. This caused some loud noises... and it broke the whole second pump!

Specifically, its little internal shaft broke right off.

I assume, what happened was that while running one engine, oil was being pumped at regular pressure backwards through the "inactive" pump on the other side. Hence, not much pressure was left for steering. Then, starting the other engine caused a force in the opposite direction, within the pump. The sudden change must have been enough to break the poor thing.

I can have a new shaft made at a nearby machine shop.
Will be more careful in future.

[C420sailor]

If it were me...

I’d designate one pump primary, one secondary. I would use biased shuttle valves, installed in such a way that they would favor the primary pump. In normal use, the primary pump would feed the steering unit, and in the event of a primary pump failure, the secondary pump would actuate the shuttle valve and feed the unit.

Make sure you install pressure relief in both sides of the system. When a pump isn’t feeding the steering unit, it needs to bleed the pressure it is creating back to the reservoir.

I think pressure regulation is going to be the tricky part...not sure how your current system manages that.

Just my $.02. I’ve operated systems like this, but never designed them.

[RaymondR]

Originally Posted by Zven

Thanks for the further tips here.

I have now learned that check valves are absolutely necessary when using two pumps...

Yesterday I tried paralleling them without check valves, just to see what happened (at low RPM).

While running only one engine, I found the steering to be unreliable with this setup.

Stupidly, I then tried starting the other engine at the same time. This caused some loud noises... and it broke the whole second pump!

Specifically, its little internal shaft broke right off.

I assume, what happened was that while running one engine, oil was being pumped at regular pressure backwards through the "inactive" pump on the other side. Hence, not much pressure was left for steering. Then, starting the other engine caused a force in the opposite direction, within the pump. The sudden change must have been enough to break the poor thing.

I can have a new shaft made at a nearby machine shop.
Will be more careful in future.

Vane pumps and vane motors are pretty well identical as is the case with most hydraulic rotating devices.