In order to make best use of pump capacity and reduce cost for valving on a recent project I coupled the blind end of cylinders at opposite ends of a transfer line. i.e. One cylinder retracts to accelerate the line, the oil from it's blind end goes through tubing to the blind end of the cylinder at the opposite end of the line to extend it in unison, at the end of line movement the cylinder at the opposite end decelerates it. The cylinders are controlled by a DO8 proportional valve. Into the intermediate line that couples the two cylinders together I included a "balance" circuit that a) adds oil through a metered poppet valve in the valve block when one cylinder is retracted but the other has not reached it's extended condition (Through the switches & PLC)., and b) has a relief valve that allows oil to bleed to tank if the "extend" cylinder is fully extended but the retracting cylinder is not fully collapsed.
The "balance" is necessary, as when the oil warms and expands I had to get rid of some. Likewise, overnight as the oil cools, in the morning there wasn't enough oil, so it adds some until the cylinder is extended.

The system does as designed for the most part, but now the issue... During the course of the morning as the system temperature comes up, they have to adjust values for speeds several times. I am using an analog card from a PLC to drive the proportional valve, all the speeds and ramps are done in the PLC math. I was think of modifying the math to factor in the oil temperature (I have a t/c on the tank to an analog input), or does anyone see an easier way to compensate to avoid the operator from having to intervene to keep the line moving the desired speed?

Visio-Drawing1.pdf (5 KB, 136 downloads) Circuit

Is this system mechanically linked through the transfer line/conveyor/machine mechanism?

If so, why the series circuit? Can you just put the cylinders in parallel and they will load share via the conveyor or linkages?

kcj

We have used a cable encoder attached to the cylinder clevis for an application somewhat similar. Already having the PLC control I believe you can incorporate a velocity loop to compensate for both load and temp change.

Maytag

Kevin,
There are latches at each end. one grabs a cart to pull the line, the other restrains the line to decelerate it, then unlatches for the return stroke. I also neglected to mention I'm moving about 60,000# in 5 seconds about 32" somewhat precisely, virtually all of the work is accel/decel.

Maytag,
The encoder idea I though about, but was trying to avoid scales )
I used a cable encoder once in a fairly clean environ, this application (iron foundry) has significant dust. How do you think it would handle it?

Thanks for the thoughts, keep them coming!
//jerry

Jerry,

This is a steel mill application(hot metal)-I am not at work and can't remember the brand/model-the reliability is good considering the environment.
Have had to watch the vendor and make sure the cable length matches the cylinder stroke.

Maytag

In the Vickers "CLOSED LOOP ELECTROHYDRAULIC SYSTEMS MANUAL" they discuss using a "Pressure Compensating Hydrastat" Sandwich Module under a Proportional Valve that effectvely makes the DCV a pressure compensated flow control.

However after seeing the cycle time the load moved and the travel distance it is unlikely that the hydrastat would be able to keep up with the demand.

It appears you can readjust and make the circuit perform after the oil heats up so why not install a Tank Heater to keep the oil at operating temperature before starting the machine. You may have to install some bypass loops to make sure the up to temperature oil is at least in the circuit out to the cylinder before the first cycle.

I know that is Low tech but if it works no one will complain.

BTW, I show the arrangement in the Basic book on page 12-2. http://www.hydraulicspneumatics.com/200/eBooks/

I think the circuit/design you made is too simple overlooking the rigidity at blind to blind...

My simple opinion is to try a simple "anti-cavitation check valve(teed to the blind), otherwise a charge pump has to be teed...

During restart of operation, bleeding has to be done on the blinds...

Jerry,

I once saw a piece of "Southern engineering" on a walking beam structure. Someone had rigged up an external temposonic and magnet setup. It was in a different section of the mill than I normally work-as far as I know its still in use.

Maytag

I have used a lot of PLC controls in mobile environments, and road use really creates some abuse. Sensing distance movement over short periods of time I have had great results with rod transducers (linear). My concern would be with debrie similar to weld spatter falling on the sensing unit. I'd stay away from proximity and other non-contact type sensors that utilize light, sound, or magnetics - considering the dust, debrie, and type of environment.

What about a tiny reservoir and a pre-fill valve?

Sounds like two separate issues: position control or feedback, and makeup oil between the twy cylinders.

Can you use double rod end cylinders to more approximate constant volume?

Or jsut have two independant valve and position controls, one on each end. more complexity but more flexability and better control?

quote:

Originally posted by Josh Cosford:
What about a tiny reservoir and a pre-fill valve?

The relief valve could drain to the mini reservoir.