Circuit opinion 1

[RobertHasty]

Hello,

I would like forum members to give me some tips regarding my circuit.
I have two cylinders that need to be synchronized. Each of them has a counterbalances due to negative load.

Also, I need local (manual) and electrical actuation.

Thanks in advance.

 

[RobertHasty]

Sorry, I forgot to insert my file link.

http://files.engineering.com/getfil...5c-48c5-adf7-513046dc6bba&file=Schematics.pdf

 

[Oldhydroman]

Hi Robert

I'm sure there will be plenty of other opinions. Your circuit will work, after a fashion, but it may not be as good as you might like it to be.

Your full bore cylinder overcentre valves will react badly to the meter-out flow control effect of the flow divider when it is acting as a combiner. All the overcentre valves will also react badly to the meter-out flow controls on your DCV's. For a standard overcentre valve the required pilot pressure will increase by a factor of (pilot ratio + 1) times the back pressure. Try to arrange the flow controls to be meter-in so that the overcentre valves are allowed to do their work. To acheive this you could use two flow dividers (one for extend and one for retract) each with reverse free flow check valves. Then swap over your DCV flow control valves to make them meter-in as well.

Also note that the effect of the PO check valves means that the pressure in the pilot lines to the cylinder overcentre valves will not decay immediately you centre the DCV. The result is is means the overcentre valves may not close as promptly as you would like - leaving some of the load pressure trapped downstream of the overcentre valves.

The PO check valves also disable the relief valve function of the cylinder overcentre valves. You can re-establish this function by using externally vented (or atmospherically referenced) overcentre valves on the cylinder and a pair of port relief valves (A-T and B-T) fitted to one of your DCVs (as long as the flow dividers are doubled up and have reverse flow check valves).

All the best.

DOL

 

[RobertHasty]

Hi Robert

I'm sure there will be plenty of other opinions. Your circuit will work, after a fashion, but it may not be as good as you might like it to be.

Hello, thanks for the reply. Not so many of them after all .

Your full bore cylinder overcentre valves will react badly to the meter-out flow control effect of the flow divider when it is acting as a combiner. All the overcentre valves will also react badly to the meter-out flow controls on your DCV's. For a standard overcentre valve the required pilot pressure will increase by a factor of (pilot ratio + 1) times the back pressure. Try to arrange the flow controls to be meter-in so that the overcentre valves are allowed to do their work. To acheive this you could use two flow dividers (one for extend and one for retract) each with reverse free flow check valves. Then swap over your DCV flow control valves to make them meter-in as well.

I also thought that meter out and divider could be the problem due to backpressure. I have changed the flow controls as meter-in.
Unfortunantely, two flow dividers are not an option.

Also note that the effect of the PO check valves means that the pressure in the pilot lines to the cylinder overcentre valves will not decay immediately you centre the DCV. The result is is means the overcentre valves may not close as promptly as you would like - leaving some of the load pressure trapped downstream of the overcentre valves.

The PO check valves also disable the relief valve function of the cylinder overcentre valves. You can re-establish this function by using externally vented (or atmospherically referenced) overcentre valves on the cylinder and a pair of port relief valves (A-T and B-T) fitted to one of your DCVs (as long as the flow dividers are doubled up and have reverse flow check valves).

I have exchanged the counterbalances with vented ones. Also, I have changed the piston divider with gear one which has pressure relief valves incorporated.

Regarding check valves, I was thinking if they don't close, the check valves will hold the load.

I've changed my schematics a little bit with additional solenoid operated DCV to have a free path from divider to tank.

http://files.engineering.com/getfil...2-4e0e-b0da-5e54860e836d&file=Schematics2.pdf

 

[kcj]

quick look, but aren't your manually operated valves bleeding the entire circuit down when power is off?
PO checks at location 5 are defeated by manual valves
PO checks seem redundant to the counterbalances already at each cylinder. What is their purpose?

You can get electrically operated proportional valves with manual operators built on. Generallymore common on stack type mobile valves. D03/D05 will have manual operating pins on the ends for test, but not easily feathered or operated if you need to do it a lot. Some D03 have levers and solenoids.

 

[hydtools]

Get rid of Y1 and Y2 and use a cylinder spool in manual valve 1. A cylinder spool will block both port A and port B when the valve is centered.
How long must the load be held without leak down? Are the PO check valves necessary?

Ted

 

[Oldhydroman]

The purpose of the PO check valves is to allow there to be two (or more) DCV's in the circuit - they have nothing to do with load-holding. Even if the counterbalance valves didn't close you would find the PO check valves couldn't hold the load because of the internal leakage in the flow divider.

Unfortunately the PO check valves do interfere with the operation of the relief function of the counterbalance valve - but this has now been fixed by using a vented counterbalance valve and getting some extra relief valves in the circuit. The PO check valves at location 5 aren't defeated by the manual DCV because each check valve is only piloted open by a signal from the solenoid operated DCV.

The problem with a "cylinder spool" is that half system pressure will appear at the A & B ports when the DCV is centered with the P line still pressurized. The speed with which this occurs depends on the pressure, viscosity, valve leakage, circuit volume, hydraulic stiffness etc. and also the relative leakage of the manual DCV compared to the leakage of the rotary flow divider. With a cylinder circuit this half system pressure applied to both cylinder ports can cause the cylinder rod to creep.

There was no mention of needing any proportionality in the controls - and the fixed setting throttle-check valves support that idea. Maybe, for the purposes of redundancy, the manual DCV has to be a "full blown" second valve (rather than just a second way of operating the solenoid DCV). If not, then you could ask the solenoid DCV supplier to fit a more readily accessible manual override. Bosch-Rexroth, for example, have a version with a "palm button" instead of the normal concealed override or [painful] "thumb button".

Looking towards the world of mobile hydraulics, you could consider: Sauer-Danfoss PVG32, Walvoil, Brevini HPV, Hawe PSL/PSV etc. These proportional directional control valves have a manual lever operation, but also have options for solenoid operation (alongside the manual control).

Be aware that the rotary flow divider behaves differently from your spool type divider:

Firstly, when dividing, the pressure at the inlet port will be slightly higher than the average of the two outlet flows. If one of the cylinders stalls then there is a possibility of pressure intensification. The relief valves limit the extent of the intensification but if a relief valve opens then synchronization will be lost.

Secondly, when working as a combiner one stalled cylinder will be left behind as the gear sets are driven round by flow from the remaining cylinder. An anti-cavitation check valve will open, preventing any damage to the components, but synchronization will be lost.

It might be worth examining the full range of loads that the cylinders can experience in order to check that these particular contingencies don't occur.

The advantage of dual flow dividers each arranged to be meter-in only (sorry - I'm harping on about my previous suggestion), is that if one cylinder can't move (because of the load) then neither will the other one. If you could find a way of using two flow dividers then the extra solenoid valves that you've added means that your counterbalance valves wouldn't need to be the vented type and you wouldn't need the extra relief valves.

Finally, if you change the solenoid energization logic a little (energize both Y1 and Y2 whenever either Y3 or Y4 is energized) then this gives an added isolation between the solenoid DCV and the manual DCV. If either solenoid of the DCV is energized then the manual DCV becomes completely isolated from the circuit.

DOL

 

[RobertHasty]

Hello.

Unfortunately the PO check valves do interfere with the operation of the relief function of the counterbalance valve - but this has now been fixed by using a vented counterbalance valve and getting some extra relief valves in the circuit.

Wouldn't then the conventional counterbalance valves (non-compensated) in this changed circuit also be OK? I mean, by incorporating valves 2, the backpressure shouldn't be issue anymore. Or?

The problem with a "cylinder spool" is that half system pressure will appear at the A & B ports when the DCV is centered with the P line still pressurized.

Sorry, I'm having some difficulty understanding this. Could You please explain it a little bit more or point me to an explanation. Thanks.

There was no mention of needing any proportionality in the controls - and the fixed setting throttle-check valves support that idea. Maybe, for the purposes of redundancy, the manual DCV has to be a "full blown" second valve (rather than just a second way of operating the solenoid DCV). If not, then you could ask the solenoid DCV supplier to fit a more readily accessible manual override. Bosch-Rexroth, for example, have a version with a "palm button" instead of the normal concealed override or [painful] "thumb button"

The proportionality is not a demand. Manual operated DCV is a second way of operation. I know about Rexroth's "rubbered" override.

Looking towards the world of mobile hydraulics, you could consider: Sauer-Danfoss PVG32, Walvoil, Brevini HPV, Hawe PSL/PSV etc. These proportional directional control valves have a manual lever operation, but also have options for solenoid operation (alongside the manual control).

I have been looking for some of these valves. Actually, only Walvoil because of the price and delivery term, and I couldn't find a lever-solenoid configuration in their datasheet. I know there should be one, but I couldn't figure it out. Anyway, this configuration doesn't demand proportionality and should be simple.

The advantage of dual flow dividers each arranged to be meter-in only (sorry - I'm harping on about my previous suggestion), is that if one cylinder can't move (because of the load) then neither will the other one. If you could find a way of using two flow dividers then the extra solenoid valves that you've added means that your counterbalance valves wouldn't need to be the vented type and you wouldn't need the extra relief valves.

Excellent suggestion, I'll think of that next time.

Finally, if you change the solenoid energization logic a little (energize both Y1 and Y2 whenever either Y3 or Y4 is energized) then this gives an added isolation between the solenoid DCV and the manual DCV. If either solenoid of the DCV is energized then the manual DCV becomes completely isolated from the circuit.

I thought to energize only one of the solenoids (Y1 or Y2) so that through one of the non-energized valves return oil can flow when working with main solenoid DCV. The reason for this is to have less pressure drop in return line than through check valve 5, but in such organization return flow goes through one of the check valves and one of the DCV valves 2.

 

[Oldhydroman]

Hi Robert

I'm going to have to answer your many questions piecemeal today - if that's OK with you .

First of all - solenoid valves with a more convenient means of accessing the manual override facility:

The standard Bosch-Rexroth solenoid DCV comes with a concealed manual override (code: N9). The common alternatives shown in the datasheets are: the rubberised thumb button (code: N) or no override at all (code: {leave blank}).

What they don’t show on the datasheet are:
N7: as N9 but with a locking function (I think it’s a push and twist with a screw driver)
N2: Small mushroom button (palm button) – not lockable
N4: as N2 but lockable via a small spring clip which can be pulled over the top of the button
N6: as N2 but a bigger button (palm button for big hands)
N5: as N6 but lockable by a push-and-twist scheme

If you are going to use a lockable manual override then you must use a DC solenoid (even if you use an AC supply with a rectifier built into the plug). If you energise an AC coil with the opposite solenoid locked in (via its manual override) then the energised solenoid will continue to draw its in-rush current and will burn out after a few minutes. DC solenoids don’t have this problem.

Sun Hydraulics offer some interesting solenoid tube end caps that incorporate a twist to operate function (far more comfortable and convenient than pushing a probe into the normal concealed override). Sun also have a version which locks into place when you twist it, and an ingenious one which is momentary when you twist it one way and locks in place when you twist it the other. (See part numbers: 991-225, 991-226 and 991-227.)

Wandfluh offer some very convenient manual actuators on their solenoids as well – most of these can be retrofitted to the standard solenoid.

Walvoil do offer lots of their valves with solenoids on one end and the lever on the other - but the price/delivery may not be right for you, especially if you don't need any proportionality.

I'll get back to you later with some other answers.

DOL

 

[Oldhydroman]

Hi Robert

Here's a few words about what happens to the external circuit if you hold a cylinder spool DCV in the center condition for a long period with the pressure still on the P port.

Regards

DOL

 

[Oldhydroman]

Hi again

Wouldn't then the conventional counterbalance valves (non-compensated) in this changed circuit also be OK? I mean, by incorporating valves 2, the backpressure shouldn't be issue anymore. Or?

Your spool type flow divider on the full bore ports will act as a meter-out control when acting as a combiner. For this reason the full bore counterbalance valves should be the vented type. You could leave them as the conventional type but the counterbalance valve on the leading cylinder will then start to close when the flow divider increases the resistance to flow on that leg. The closing counterbalance valve would tend to slow down the leading cylinder - but this is a circuit interaction that you can manage without. It is often the case that circuits seem to work on commissioning but there's actually a lot of unknown [to the designer] interactions that only start to make themselves known later when something changes in the circuit (temperature, leakage rates, unusual loading conditions etc.) A vented counterbalance valve will be insensitive to changes in back pressure caused by the flow divider - thereby letting the flow divider do its job. Your cylinders will then have a meter-in flow control to the annulus and just the leading cylinder will also have a meter-out flow control on the full bore. [The counterbalance valves on the annulus ports can remain as conventional valves.]

I thought to energize only one of the solenoids (Y1 or Y2) so that through one of the non-energized valves return oil can flow when working with main solenoid DCV. The reason for this is to have less pressure drop in return line than through check valve 5, but in such organization return flow goes through one of the check valves and one of the DCV valves 2.

One of the problems here is that the performance limits of a spool type DCV depend on the nature of the flow through the valve. Flow forces [P to A (or B) and A (or B) to T] try to bring the spool back to a neutral position and this limits how much flow you can get through the valve. A modern DCV will often incorporate a type of flow force compensation where the flow from A (or B) to T is deflected by a conical end on the spool to deliberately create an additional [opening] force which partially balances the two closing forces. Massively different flow in the two paths through a 4 port DCV upsets the balance. So if you have a much greater flow from P than you have flow to T the forces try to push the spool back to centre, if you have a much greater flow to T than from P the flow force compensation is overdone and the spool will shift out of center when you didn't intend it to.

In this application there would probably be no harm in allowing the manual DCV to take a single flow path return flow, if the spool does shift it will apply a pressure to the energised, i.e., closed 2 way valve. So, as a matter of form, I always try to avoid making a 4 way valve take a single flow path. If you really have to do this (and the problem is worst with solenoid operated valves) then check the valve’s switching capacity very, very carefully. It is not uncommon for the switching performance to be reduced to about 10% when you plug one of the ports of a 4 way valve.

However, the particular reason that I suggested energising both of the 2 way valves was to avoid there being any effect of someone operating the manual DCV while someone else was operating the solenoid DCV. If you close both 2 way valves when either of the original solenoids is energised then the first person can do whatever they like with the manual DCV – it will have no effect because the solenoid DCV will take priority.

I’ve taken the liberty of drawing up some further suggestions (see attached file).

The first circuit is close to your original. You can use a cylinder spool on the solenoid DCV (this gets rid of the PO check valves) because when this DCV is centered the de-energised 2 way valves and the float spool of the manual DCV prevent any problems with “half system pressure”. The annulus counterbalance valves are conventional but the full bore counterbalance valves are atmospherically vented – and I’ve shown both counterbalance valves in a single cross-piloted block that can be mounted directly to the cylinder. I’ve gone back to your original valve type flow divider with a pair of direct operated relief valves to restore the relieving function of the full bore counterbalance valves. (A rotary flow divider without a meter-out control on its common port will provide no synchronising function at all when it is combining.)

The second circuit is an option that you might not have previously considered. A common [hydraulically actuated] DCV controls the extend/retract functions. This valve is piloted from either the solenoid DCV or the manual DCV – the signals from which are isolated from each other via shuttle valves. If you wanted one DCV to have a higher priority than the other then fit a reducing/relieving valve to the one with the lowest priority. The pilot signals from the DCV without the reducing/relieving valve will overcome the pilot signals from the other DCV. This second arrangement reduces the number of solenoid valves (and solenoids, leads, plugs, cable tray, power supplies etc.).

DOL

 

[RobertHasty]

Hi Robert

Here's a few words about what happens to the external circuit if you hold a cylinder spool DCV in the center condition for a long period with the pressure still on the P port.

Regards

DOL

http://files.engineering.com/getfile.aspx?folder=c40110c3-840f-4330-9e48-f5

Excellent explanation!

 

[RobertHasty]

Your spool type flow divider on the full bore ports will act as a meter-out control when acting as a combiner. For this reason the full bore counterbalance valves should be the vented type. You could leave them as the conventional type but the counterbalance valve on the leading cylinder will then start to close when the flow divider increases the resistance to flow on that leg. The closing counterbalance valve would tend to slow down the leading cylinder - but this is a circuit interaction that you can manage without. It is often the case that circuits seem to work on commissioning but there's actually a lot of unknown [to the designer] interactions that only start to make themselves known later when something changes in the circuit (temperature, leakage rates, unusual loading conditions etc.) A vented counterbalance valve will be insensitive to changes in back pressure caused by the flow divider - thereby letting the flow divider do its job. Your cylinders will then have a meter-in flow control to the annulus and just the leading cylinder will also have a meter-out flow control on the full bore. [The counterbalance valves on the annulus ports can remain as conventional valves.]

My flow divider is going to be gear one with pressure valves in each line, but the meter out effect is the same as on a spool one, I believe. I completely agree with You on a vented counterbalances, but I'm also interested in theoretical situation where I would leave full bore ones. In such a situation meter out effect of a flow divider would increase pilot pressure, i.e. pressure which is acting on an annulus area to open them full? Now is a question will pressure changes on a divider make cylinder to "jerk", so to say?
Also what's bugging me is the fact that in a situation when divider is working as a combiner, the flow goes straight to the tank, so this pressure drop increase on a combiner is confusing.

One of the problems here is that the performance limits of a spool type DCV depend on the nature of the flow through the valve. Flow forces [P to A (or B) and A (or B) to T] try to bring the spool back to a neutral position and this limits how much flow you can get through the valve. A modern DCV will often incorporate a type of flow force compensation where the flow from A (or B) to T is deflected by a conical end on the spool to deliberately create an additional [opening] force which partially balances the two closing forces. Massively different flow in the two paths through a 4 port DCV upsets the balance. So if you have a much greater flow from P than you have flow to T the forces try to push the spool back to centre, if you have a much greater flow to T than from P the flow force compensation is overdone and the spool will shift out of center when you didn't intend it to.

In this application there would probably be no harm in allowing the manual DCV to take a single flow path return flow, if the spool does shift it will apply a pressure to the energised, i.e., closed 2 way valve. So, as a matter of form, I always try to avoid making a 4 way valve take a single flow path. If you really have to do this (and the problem is worst with solenoid operated valves) then check the valve’s switching capacity very, very carefully. It is not uncommon for the switching performance to be reduced to about 10% when you plug one of the ports of a 4 way valve.

Yes, excellent point, this Rexroth datasheet on page 8 says exactly that.
But as You said it, I also believe that in this application that should not represent a problem because the flow coming from a cap side through a divider is almost double the flow that goes to the rod side, and therefore, some of that flow should also go through solenoid DCV. The second reason why I don't fear is the fact that the pressure applied to the annular side should be only that big to open the counterbalance mounted on a cap side, and the cylinder is mostly doing his work when extending. When retracting, some negative force also acts on it which "helps" opening cap side counterbalance valve.

However, the particular reason that I suggested energising both of the 2 way valves was to avoid there being any effect of someone operating the manual DCV while someone else was operating the solenoid DCV. If you close both 2 way valves when either of the original solenoids is energised then the first person can do whatever they like with the manual DCV – it will have no effect because the solenoid DCV will take priority.

Good point, I didn't think that far . But this is also a problem of monoblock valves which incorporate both solenoid and lever operated spool valves?

I’ve taken the liberty of drawing up some further suggestions (see attached file).

Thanks, I have also come up with one more possible solution which I'll post later.

The first circuit is close to your original. You can use a cylinder spool on the solenoid DCV (this gets rid of the PO check valves) because when this DCV is centered the de-energised 2 way valves and the float spool of the manual DCV prevent any problems with “half system pressure”. The annulus counterbalance valves are conventional but the full bore counterbalance valves are atmospherically vented – and I’ve shown both counterbalance valves in a single cross-piloted block that can be mounted directly to the cylinder. I’ve gone back to your original valve type flow divider with a pair of direct operated relief valves to restore the relieving function of the full bore counterbalance valves. (A rotary flow divider without a meter-out control on its common port will provide no synchronising function at all when it is combining.)

Not at all? So, solution would be to put a throttle check valve on it's main port.

The second circuit is an option that you might not have previously considered. A common [hydraulically actuated] DCV controls the extend/retract functions. This valve is piloted from either the solenoid DCV or the manual DCV – the signals from which are isolated from each other via shuttle valves. If you wanted one DCV to have a higher priority than the other then fit a reducing/relieving valve to the one with the lowest priority. The pilot signals from the DCV without the reducing/relieving valve will overcome the pilot signals from the other DCV. This second arrangement reduces the number of solenoid valves (and solenoids, leads, plugs, cable tray, power supplies etc.).

I have been considering some shuttle valves, but my creativity was suppressed at the time.

 

[RobertHasty]

Hello, I've made another schematics. The idea is to use sandwich plate counterbalance valves, and also hose break valves on cylinders.

http://files.engineering.com/getfil...9-4abe-8064-c9a6eb017e7c&file=Schematics3.pdf

 

[Oldhydroman]

Hi again Robert

One BIG word of warning about sandwich counterbalance valves: they are very convenient to install but they usually end up a long way from the cylinder and this is often a problem. Your use of the hose burst valves covers the increased risk of load release (the risk increases because of the greater number of connections between the load holding valve and the cylinder ports) but the distance between the valves and the cylinders massively increases the volume of fluid between the two. This causes the system to be "softer", i.e., to have a much lower natural frequency, and the liklihood of instability is very much increased.

Also note that the cross-port leakage across your flow divider may allow one cylinder to extend and the other retract when you think you've locked in the load by having both DCV's in neutral.

Don't forget to connect the tank line of your flow divider - I'm pretty sure you would in practice, but missing it off the drawing can trip you up because, in your mind, you skip over part of the analysis of the way the circuit might misbehave.

Your sandwich conterbalance valves, should you decide to stick with them, do not need to be the vented style. You're better off using the conventional (non-vented) versions whenever you can because they have a pressure override characteristic which is more suitable for successful speed control.

Oh, and by the way, you don't have a problem with the monoblock valves if the solenoids are trying to one thing and the guy on the handle is trying to do the other: whichever is the stronger will win. When you have two separate DCV's you could have one switched one way and the one switched the other way: you have to analyse the circuit to see what would happen in this instance.

DOL