Two Hydraulic Cylinders to be Synchronized

There are equal-flow divider (or "splitter") valves that can meter very nearly equal flow to your two cylinders. I think they can be as accurate as you are asking, but I'm not sure.

This sounds like a useful device. G.A.

Problem with splitter is that I think you will find that the lift with less load/resistance on it will rise more than the other. The fluid will flow to the least resistance. If you had a 2 same volume pumps driven from a single power source so that one pump doesnt pump more volume than the other, you could achieve this. A high tech method would be a feedback loop and servo control that monitors A vs B position, but that would be pricey.

An equal flow splitter is not merely a "Y" or "T" that would have this problem; it has internal gears or vanes that meter the flow equally. Over time, wear of clearances may degrade the equality, but there are ways to "reset" the rams at end of stroke that small differences do not accumulate.

http://stuartmdavisinc.com/blog/?p=7

http://www.fpweb.com/200/Issue/Article/False/21749/Issue

http://information-technology-articles.com/Automation/Study-on-Control-Methods-of-Doule-Servo-Valve-Controlled-Cylinders--Sy-1491.html

Hope these links help

Without flow dividers, you can also use your cylinders in tandem, when they have the same specs. like stroke and bore. The only thing you need to take care for is that you complete purge the line between the two cylinders.

This is how it works:

Your first cylinder (master) is connected to the pump on the bottom. The top has also a output connector and the area above the plunger is full with oil- up to the bottom of cylinder 2 (slave)

Master cannot move until slave moves with the same distance. This is a fool proof precision method as long you have no trapped air or leaks.

Enjoy.

In this scheme, the volume of oil going to the second cylinder will be less than that to the first, by the proportion of rod area to piston area.

I agree with you, the only way round it would be to make second cylinder smaller to allow for area of piston rod.

But then you would have a difference of load capacity as second cylinder would be smaller diameter than master giving a different lift capacity ?

Good point, though rods protruding from both ends are unusual.

correct, works only with slave cylinder upside down , sorry I forgot. Or cylinder with through- shaft.

"...you can also use your cylinders in tandem, when they have the same specs. like stroke and bore. The only thing you need to take care for is that you complete purge the line between the two cylinders". This will work. I did this on an AGV (Automatic Guided Vehicle) in 1983 to synchronize the movement of two load arms. Think about it - when the Master cylinder displaces say 1 liter of hydraulic oil, then that 1 liter enters the Slave.

what about the fact that the volume of the rod in the top half reduces that volume below a liter?

THOUGHT this was evident GA.

Other than involving regulators, pressure and viscosity fixes, all providing hysteresis and introducing supplementary control devices, the tandem still works fast and accurate.

It deals with everything that otherwise needs to be compensated.

Different weight, speed, temperature and with a simple fill pump (manual) exact height can be monitored and controlled on a daily or weekly basis.

Of course it may be too simple for a complicated mindset or too inexpensive for a unlimited budget.

TANDEM = 1 UPSIDE DOWN

You could syn then mechanically with a few cables and pulleys like an automotive lift does. To is cheap, simple and very reliable.

You can buy a cylinders where the input volume (smaller bore) matches the output of the master or use double ended cylinders.

Tony

Notwithstanding the comments made by others, the only accurate way to achieve your desired result is to use LVDT transducers inside the rod of both cylinders and connect to a PLC via a PID loop system. This in turn will control a proportional flow valve on one cylinder that will match the velocity of the other cylinder.

The LVDT will be accurate to 1024 (bits) ÷ 1500mm = 1.46mm.

You will then need to fit 'make up' check valve so the cylinders 'reset' at end of each stroke.

GA.

I was thinking along a similar line (i.e using position feedback and a controller). But my idea was driving both actuators with the same command and each having their own controllers.

I like your idea but don't have experience with that approach. Will the system be responsive enough to allow the slave cylinder to stay close enough to the master?

I suppose that, what I would do is to control the 'dumb' cylinder with a standard flow control (throttle) to obtain the desired overall speed and then control the 'active' cylinder with the PLC controlled system to match the velocity of the dumb cylinder.

If you try to control both cylinders, you run the risk of a 'hunting' effect between the two proportional flow control devices which would become very erratic!

As for the accuracy, the LVDTs would be accurate to around 1.5mm which is the incremental size detectable with a device of that length. I think that the PID loop would be quick enough (at around 800MHz) to be VERY responsive therefore very accurate. Also, it would be a fairly simple job for a PLC Techy to write the required algorithm (code) to run the system. The really good thing about it is that it is completely independent of outside influences such as heat, viscosity, load or seal sticktion.

I wasn't too worried about the PID (or even other controllers) being fast enough, but more of the command driving the slave not having a command until the master has actually started moving, therefore possibly introducing a lag between the two. This effect might be compensated for by using a lead term or if the speed of the cylinders are not that fast then the lag may not be enough to cause a problem.

I don't see a problem with hunting. I work with flight simulators. The motion base is a 6 axis Stewart platform of various payloads and stroke lengths depending on the design. Generally the hydraulic actuators are 4.5 in piston and a stroke of 60 inches (+/- 30 inches from mid stroke) with a payload near 32,000 pounds (total...not per cylinder). It's possible to drive all 6 actuators such that the platform stays completely level.

Assuming all the basic designed features are correct;fluid flow is the problem. Check the fittings and pipe sizes used. The volume of both the branches has to be the same. Then employ flow control valve on inlet and exit ports of the cylinders.

Adjust manually the flow by trial and error. After achieving the results remove the handles/ knobs of the valves or lock them, so that nobody tempers it.

Sometimes if the cylinders are single acting then while going down due to weight acting on them the downward stroke may remain different. In such cases use springs inside to provide retraction force.

kiran nawathe

I don't think that this will work as the characteristics of 'flow control valves' that are locked off will vary too much due to temperature variation affecting the viscosity of the fluid. The temperature variation will be due to natural daily fluctuations AND from heat induced within the hydraulic system.

I'm not an expert in hydraulics but my thinking is along Your lines. I always thought that two cylinders working together need to have the same length and diam. lines equal distance from the pump. Is this correct?

oilcan13

use a linked pair of cylinders, one for each side, used as pumps (one stroke)as the compressed fluid source. They are driven by a single cylinder from your control valve. So

the two will get squeezed and since they are directly connected to the two linked cylinders the volume pumped out of each will be the same, and will go to each end of the platform = equal rise, no servo or other link to fail.

Milindaur,

You have many good answers here, but that said you can still roll your own cheaply using modern components. Here is one overriding requirement-make it reliable and as simple as possible and then on top of that provide an automatic lock up to block the discharge of both cylinders in the event of impending doom using a completely redundant set of parts.

For example you can create excellent linear magnetic encoders with parts from Austriamicrosystems.com. Make sure the encoders and the magnets are solidly mounted. If you use a normal direction control valve to operate one cylinder, the identical magnetic encoders on the two cylinders can easily make the "slave" cylinder track the movement of the first cylinder electronically with PID precision. Just program an 8 or 16 bit microprocessor costing less than 2 US dollars. Follow the suggestion and provide a fail safe "lock-up" with normally closed valves, you might want to provide and "Emergency Stop" button/(s). If you require high speed movements, you will need a hydraulic servo valve controlling the slave cylinder, otherwise a solenoid operated spool valve with speed controls installed properly will track within your required tolerance.

Have fun,

Luther M

record each piston rod position and make sure both positions are equal by opening valves accordingly

Here is one thought,

First you can fix a T or Y divider that can do gross control of cylinders. beside this you can fix two Resistance scales corresponding to each cylinders, so the Resistance of each scale will increase as cylinders move forward, combining these two resistances with two more reference resistances in a Wheatstone bridge manner can give you a positive or negative output depending on which hydraulic cylinder is leading or lagging.

With the help of above said signal you can switch two different control valves connected to each cylinder, And this will allow you to fine tune the position of each cylinder.

A bit of electronics is involved but it should do the job.

Regards

Lots of post and many bring up different solutions. The original post however, lacks precise information.

And solutions can easily mount up to $5,000.00 if precision and speed are required. What platform is it on both sides? Are your specs complete?

What about failure or fool proof? Just when 1 cylinder starts to leak?

If you want precision and stability, consider trapezium or bearing rolls threaded rods that rotate synchronously.

This is the way platforms or machinery move, with mechanic coupling e.g chains or individually driven and controlled with incremental controlling devices, adapted to the drive to make them spin. (either DC, AC, or stepper) Poster gives no info about the frequency or speed of the movements - which can, be a critical factor to consider too.

If your cylinders are up stroking, single acting, lifting a load, then install a flow divider between the two cylinders. pump-relief valve-flow divider in-divider out-cyl 1-divider out cyl 2. If your cylinders are double acting, use the above and a counter balance valve common to both cylinder annular ports. I hope this helps Mike Chateau Base Carbons

You may want to look at an electrical solution with lead screws and motors. This might be easier to synchronise. Two synchronous motors could be feed from the same AC drive and would track each others' position from zero to full speed as long as you don't exceed the max torque. Two contactors could be used to run one motor at a time when independent positioning is needed.

Obviously, full flexibility would be obtained from two independent position (servo) controllers.

You may want to look at an electrical solution with lead screws and motors.

Have to keep in mind of the force requirements

I am not a mechanical engineer but the simple solution is for you to connect the two cylinders to same input lines thereby equalizing the pressure for both cylinders.

Also, the cylnders have to be identical in mechanical specs.

That would be a mistake. Its not that simple.

Its not the supply thats the issue, its the resistance of the cylinders. with (2) cylinders hooked up the way you suggest. A cylinder that would have more resistances will move less, that the one with less resistance will move more. Then it gets worse from there.

Understand?

p911

There are 2 types of flow dividers:

- one is based on a pressure balance of a piston controlling the flows so that the pressures on the 2 piston sides are equal. this type is viscosity sensitive and brings errors bigger than what you want to achieve

- the second type uses gear pumps/motor stages (up to 6 or 8) on the same shaft, which have a common inlet and with independent outlets. the gears are manufactured with a very small difference in displacement so that the flows are considered as equal for situations as yours.The system works in both directions and it is possible at every return to "zero" the piston position so that errors from successive cycles will not sum.

If you google the company Jahns Regulatoren you find all technical information you need.The product is since more than 20 years in usage in similar applications with great success since it does not need a complex servo-hydraulic approach if the precision requirement is not asking for. The differences in stroke can come only from the differences in the leakage and those are very small, the gears being pressure compensated. Pressures can be quite high and life expectancy with a clean oil is important. The cost of such a system is a lot lower than any feed back based system.