Worms in hydraulic systems

[harryting]

Mind providing a link to the original thread where all these started?

 

[Peter Nachtwey]

Of course, if the load increases during lifting the cylinder will decelerate.

The load changes as the angles change.

How much is a function of the bulk modulus of the hydraulic fluid and the effective "spring rate" of the cylinders and hoses.

Yes, the hydraulic capacitance.

I'm sure some bright young PhD candidate could figure it out if anybody cared. However, for a system like this nobody cares!

No body cares on a system like this but when the same techniques are used to design motion control systems there is often trouble.

OK - I'm going to quit now. I'll never change Peter's mind

Not possible. You have picked a very simplistic case and have made many simplifying assuptions. You are assuming the pump is always pushing oil over the relief valve when not moving. That is not energy efficient. It does allow you to assume the acceleration is constant.

My point is that I can use my methods for designing a front loader or bucket system whereas your methods will not work on a servo system.

a) Demonstrating to the youngin's that not all hydraulic systems are about motion control and not all systems require precise operation.

I was never disputing that. What I object to is when your simplistic methods are applied to a real system there is a big chance the system will not work as desired.

b) Demonstrating that you can adjust your mathematical model of the system to suit the physical and commercial constraints of the system. Exercise of good engineering judgement and experience means doing an analysis that verifies you meet the functional requirements of the customer and the cost constraints of the supplier. There is no benefit in doing a sophisticated analysis that doesn't help one of these parties.

A simulation does help. For instance I don't assume a fixed displacement pump is being used or the pump is pushing oil through a relief valve. I can optimize/minimize the energy required.

c) Annoying Peter a little.

Hopefully many here will see there is a difference between barn yard hydraulic design and industrial design methods. The PLC programmers on this forum will be more likely to have to control an industrial motion control system than a barn yard tractor with bucket.

The annoyance was further aggravated by the implication that there is only one way to look at any hydraulic system and that any analysis that varies from the one true methodology is defective.

It is defective. Even the methods I use have limits. For instance I don't take into account the speed of sound in oil. It makes a big difference at times. When there is 40 ft of hose or pipe between the valve and the cylinder the pressure wave takes about 10 ms to affect the motion. That is dead time and it is eternity for a motion controller. If there is hose between the valve and the cylinder then the capacitance of the hose must be taken into account but the hose manufacturer don't provide a value for hydraulic capacitance. The greater the hydraulic capacitance the slower the response will be.

My methods are accurate enough given that the variabilty in the components contribute more to the modeling error than the model itself.

And Peter, don't take it too seriously.

I have seen too many poor designs not to. You need to come here and sit with our tech support guys for a while to understand.

I think the comment about navigating assuming the earth was flat was very good.

 

[Peter Nachtwey]

Well it happened again. Our tech support person just brougth me a hydraulic circuit for a press. The poor PLC/motion control guy couldn't control it because the press design is flawed. The hydraulic people blame the PLC/motion control guy but it is clear that these hydraulic 'designers' have never designed a hydraulic servo system before. Fixing this will be expensive. There are many 'logic' valves and other unnecessary junk in the system. The hydraulic 'designer' was definately trying to sell hardware and his engineering time. There is no thought as to how the system was going to be controlled.
Mean while the PLC/motion guy has wasted a lot of his time.
This is what I see over and over and over and over and over and over.................

 

[Tom Jenkins]

Really, Peter, these discussions remind me of a poem I read as a wee lad: http://www.noogenesis.com/pineapple/blind_men_elephant.html

The load changes as the angles change.

Well, that's true of course, but for a great many systems IT DOESN'T MATTER! Unlike the encounter with the flying cow, the change in angle and load are very gradual. For all practical purposes the load change will be compensated by a gradual increase in pressure. Technically there will be a slight slowing as the pressure increase causes expansion of the cylinder and hose, but that effect is going to be negligible and imperceptible in all but a very specialized system - perhaps motion control. Back in the day, when I did field service on constructiom machinery, if I had detectible softness in a hydraulic system I knew it was time to bleed system and get rid of the entrained air.

No body cares on a system like this but when the same techniques are used to design motion control systems there is often trouble.

And that, sir, is exactly my point! The fact that an engineer recognizes the difference between a servo system and construction machinery, for example, isn't a sign of incompetence. It means that the analyis is appropriate to the system. There isn't any point in doing an analysis of a motion control system when you are designing a crane, for example, or an elevator. Adding inconsequential terms to an equation or including consideration of negligible factors doesn't add value to the final product or indicate superior analysis. They just waste valuble engineering time.

You are assuming the pump is always pushing oil over the relief valve when not moving. That is not energy efficient.

But, of course, a gear pump IS always pushing oil. If the spool is in center position the flow is pumped right back to the sump at essentially zero pressure so there is no pressure drop or power. If the pump path to the sump is blocked, say if the spool is shifted to direct oil to the cylinder, then the oil must either flow into and move the cylinder or dump over relief!

The energy inefficiency may or may not be correct. It really is a function of duty cycle and system operation.

Pump hp = psi x gpm / 1714

A gear pump in an open center system dumping full flow at 0 psi draws 0 hp. A pressure compensated variable displacement pump in a closed center system at zero flow and max psi also draws 0 hp. During idle portions of the cycle, then, it's a wash. A lot of systems spend a lot of time at idle.

If an operator is pushing production he has the control valves at max position. A gear pump pushes its constant flow rate against the actual load pressure. A pressure compensated pump will stroke out and push max flow against actual load pressure. In this operation power draw in the two systems is equal.

The power advantage of a closed loop system only appears when the control valves are modulated and the system is operating at reduced flow. A gear pump will dump the balance of the flow to the sump at full pressure, wasting power. A pressure compensated pump will stroke back, reducing flow and saving power. If the hydraulic system being designed will spend a lot of time in this condition then the energy savings may offset the higher cost of the more sophisticated pump. A lot of construction machinery uses plain old gear pumps and open center valves. Many systems use multiple pumps so power isn't wasted when a particular circuit is idle.

And, for a lot of systems (like a backhoe) the flows are so high that sizing an accumulator to do more than attenuate pressure spikes just isn't feasible.

There are many 'logic' valves and other unnecessary junk in the system.

That doesn't sound like confusion over the role of pressure and flow to me - it sounds like the inability to incorporate the KISS principle and failure to understant the actual physical constraints of the system.

 

[Peter Nachtwey]

Well, that's true of course, but for a great many systems IT DOESN'T MATTER!

How does the common hydraulic designer know that it doesn't matter when all he has ever known is wrong.

That doesn't sound like confusion over the role of pressure and flow to me - it sounds like the inability to incorporate the KISS principle and failure to understant the actual physical constraints of the system.

Obviously the designer had some bad ideas on flow and pressure but the system he designed is uncontrollable.

 

[Tom Jenkins]

How does the common hydraulic designer know that it doesn't matter when all he has ever known is wrong.

Sigh!

 

[Tom Jenkins]

Mind providing a link to the original thread where all these started?

http://www.plctalk.net/qanda/showthread.php?t=42533&highlight=flow+makes

You asked for it!