Ramping up an actuator on a load sensing system 1

[hydroman247]

Hi guys,

I have a problem whereby a rotary actuator will overspeed on startup causing issues. It has a flow control valve before it that can't handle the sudden influx of flow and will send 10-50% more oil to the actuator in the initial 2-3 seconds. This is a load sensing system with the LS feed coming back to the pump from just after the flow control valve. After the initial spike, it will settle down to its required speed setting and work as normal. Due to the nature of this application, overspeed on startup is not acceptable.

Does anyone have any experience with ramping up flow on an LS system?

Details:
45cc LS pump
5cc motor
The FCV is set to 15L/min.

Thanks

 

[kcj]

Is this the only LS device on that pump?
If others, do they exhibit overshoot also?
Is this sluggish on other speed and load changes, or just the first startup?
Is it different first start of the day when system was shut down, vs just turning the actuator on and off all day?
Long LS lines?
Temperature sensitive, first cold start or heatedoil?
All air purged from LS lines?

 

[hydroman247]

No it is now.
The others are non-rotary and do not seem to display this issue.
Just the first startup. After ~3 seconds, it settles to the correct speed and even under load it works fine.
LS line length is not long, 2m and others are longer anyway.
Temperature makes barely any difference.
Yes.

 

[Oldhydroman]

Hi there Hydroman247

If your rotary actuator is receiving its flow from a simple throttle valve then the only reason you are over-speeding sometimes is because there is too much pressure drop across the valve, i.e., you are getting too much flow. Your load sensing system is supposed to be keeping the valve pressure drop constant - but for some reason the pump pressure must be too high for a few seconds.

It's only a thought, but is it possible that when you first start the movement you experience a momentary high pressure because of the rapid acceleration of the rotating mass, and somehow this high pressure signal is being trapped in your load sensing line? It might take a few seconds for the signal to decay to the right level...and that's the root cause of your over-speed.

You mentioned that there were several actuators being supplied by this pump, so presumably you have some form of check valve system or shuttle valve system to enable the pump to resolve the highest value LS signal. These check valves or shuttle valves can allow a high pressure pulse to get trapped in the LS line. Often the pump controller has a bleed off orifice in the LS line which would allow the signal to decay, but the orifice is usually very small and might take a few second to rid the LS line of the excess pressure. The effect is exacerbated if you have a lot of hoses in your LS circuit. The jet could be blocked, or it could even be absent! Some versions of the pump controller have no jet as standard because the presence of the jet interferes with the accuracy of any external pressure control you might want to apply via the LS line.

If you do add an external orifice, to allow the LS line to reduce pressure more quickly, then you could do worse than put the orifice at the far end of the LS line rather than at the pump end. The presence of the orifice will mean that there will be a real flow down the LS line rather than just a pressure signal. If the orifice is at the pump end then the flow induced pressure drop in the line will be seen by the pump and the LS signal will be slightly too low. Put the orifice at the other end of the line then the flow will be towards the orifice and the pump will see none of the line pressure drop.

As I said, it's only a guess.

DOL

 

[HPost]

Is it not just the fact that the pump being biased to max flow, chucks too mush oil at the rotary actuator. Without a pressure compensated flow control valve, the actuator will of course leap into life. With controls line at the length stated, it stands to reason that the pump will take a second or two to come off stroke.

Is there a mechanical stop on the pump to limit the max flow?

Can you fit a small accumulator to supplement the control signal on start up?

Is it possible to change the pump controller to an electronic control to offset the start up issues?

Hpost CEng MIMechE

 

[hydroman247]

@HPost

There are several other functions that require higher flows so this is not an option.

I would also rather not fit an accumulator if possible.

Also, not really possible to change to electronic controlled pump at this point.

@Oldhydroman

This is what was suspected since LS is what is controlling flow. This particular system does have an LS leakage path but it is not in use in the lower flow functions. It has a cut off system which is normally closed unless certain funcions are being used and that is all built into a block so not easily changed.

There is also LS leakage within the pump itself but I do not know the size of that one. An additional one just for this function (behind a check valve so as not to affect other functions) could be tested to see whether it helps.

 

[Oldhydroman]

On systems with multiple LS signals where several actuators have to operate simultaneously it is common practice to install pressure compensated flow control valves for each service. The actuator which is working at the highest pressure “calls the tune” and dictates what the pump pressure will be. All other actuators are, by definition, working at a lower pressure and the compensators on their individual flow control valves come into play. You could [temporarily] fit such a system on your circuit to see if that solves the problem.

Install a direct operated reducing valve in series with the inlet to your flow control valve and connect the reducing valve’s drain line to the outlet of the flow control valve (the same place from which you pick up your LS signal). Make sure this LS signal is protected from high pressure signals from the LS connections to other actuators. Set the new reducing valve to, say, 10 bar, and have the LS differential pressure on the pump set to, say 14 bar. When everything is working properly the new compensator (the function the reducing valve gives you) will set a 10 bar pressure drop across the flow control valve and you can then re-adjust the flow control to give you the speed you want. The pump supply will be 14 bar more than the load pressure so there will be a 4 bar drop across the new compensator. If another service (or a trapped high pressure LS signal) calls the pump pressure to be higher than 14 bar above your load pressure then all that happens is an increase in the pressure drop across the new compensator. The new compensator will still be in circuit so your actuator speed will remain constant.

It doesn’t have to be a full blown reducing valve, you could use a fixed setting logic valve (such as SUN LPDC) and choose a relatively low spring rating. Alternatively, you could change your existing flow control valve for a pressure compensated flow control valve, but this takes away any control you would have over the pressure drop and forces you to raise the pump differential pressure setting which has a significant [negative] impact on efficiency and heat generation.

DOL

 

[hydroman247]

I forgot to mention, when this function is being used, no other functions are being used. The other issues is that the flow control valve for the problem actuator is part of a manifold block which contains everything for that unit so tee-ing into some places is not possible.

So far, looking at the overall picture, I think it is just the reaction time of the pump is the issue since the actutor settles down fairly quickly and runs fine. If there was a way of speeding up the LS response time for this function, that is what would be needed.

 

[Oldhydroman]

Not being able to break into the manifold makes it tricky. Can you send a circuit? Here's a few suggestions anyway, hopefully one of them has some potential.

If the actuator always works at the same sort of pressure (when running properly) then you could consider teeing a small, rapid response relief valve into the LS line from just this actuator. That way, when you started up, the full magnitude of the high pressure pulse wouldn't make it through to the pump controller and you wouldn't be stuck waiting for the pump pressure to back off because it never went high in the first place. [Don't forget to put an orifice upstream of the tee point otherwise the relief valve will be swamped and won't work the way you intended.]

A soft switching directional control valve might help, you haven't said what sort of on/off control you have. Some cartridge valves are readily interchangeable with their soft switching counterparts. The soft switching period is only a matter of tenths of a second but it can just take the sting out of the on/off pressure pulse.

You could fit a new [pressure compensated] flow control valve directly on the actuator port and create a new LS line from the actuator port itself (unless the manifold is mounted straight onto the back of the actuator). Abandon the original LS line and fully open the original flow control valve.

I'm pretty sure you will have done this one already, but, do make sure your LS lines are bled of all air - including any dead ends, and make sure the pipe routes don't allow the LS line to empty under gravity when the machine is shut down. If you have hoses in your LS lines then try to replace them with rigid tubes, smaller/stiffer hoses, or fit check valves so the bare minimum of volume suffers the high pressure pulse [this minimises the accumulator effect of all your LS line volumes].

If your pump (what make by the way?) does have a bleed off orifice then try changing it for one with a bigger hole. Or introduce a new bleed off orifice, maybe on a small solenoid valve so that the extra orifice was out of circuit for all other users.

Check that the LS compensator spool in the pump controller isn't gummed up with grease or sticky stuff, and that it is moving freely in its housing. It wouldn't be the first time a pump controller had the wrong spool in it. Some LS compensator spools differ from the pressure compensator spools only by virtue of an orifice down the middle of the spool - check they haven't got mixed up at some point. You could even ask the supplier to send you a spare compensator assembly off an old pump just to see if that eliminates the problem. [The compensator wears much less than the pump pistons and bearings - a compensator from an old pump might just be run-in enough to make all the difference.]

Kind Regards

DOL

 

[hydroman247]

Thanks for your advice. For commerical reasons, I cannot upload a schematic but all the ideas you have posted have been taken into account.

The unfortunate thing is the unit with the issue (actuator and manifold control valve) is a unit we buy in ready made and modofying it directly would void warranty etc. We have spoken to them and there is a possible solution but we are still waiting for a response.

Once again thanks for your time and knowledge sharing. There is a huge lack of hydraulic knowledge online. I search for information on specific hydraulic problems and I usually end up back here reading your posts.

 

[MikeHalloran]

Just for jollies, I'd try:

- Making the long LS line(s) a LOT smaller in diameter.
- Vacuum bleeding the whole system.

... and see what, if any, difference that makes.



Mike Halloran
Pembroke Pines, FL, USA

 

[kcj]

Making the long LS line(s) a LOT smaller in diameter.




And try a stiffer construction multi wire or microbore gauge line maybe.
Cold viscosity may be an issue, but try some things as tests.
Both goals are to take out capacitance and lag in the feedback.
An accumulator in the LS line would add capacitance and probably mek things worse.

k

 

[PNachtwey]

Are the lines full of oil when starting? If there is no oil in the outlet line of the rotary actuator there will be a lower back pressure which means the pressure across the rotary actuator will be higher at first than when running in steady state. The rotary actuator will accelerate until the sum of torques on the rotary actuator are 0.


Peter Nachtwey
Delta Computer Systems

http://www.deltamotion.com

 

[PNachtwey]

"- Making the long LS line(s) a LOT smaller in diameter"
It is the length of the line that makes the response slower. There are two reasons for this.



Peter Nachtwey
Delta Computer Systems

http://www.deltamotion.com

 

[hydroman247]

The LS line is short and very small.

We may have a possible solution, although not hydraulic, it should solve the actual issue and we can then live with the overspeeding actuator.