Servo hydraulic with function generator for fluid flow

csiong

Member
Join Date
May 2008
Location
Selangor
Posts
2
We are woking on a servo hydraulic system for investigating the dynamic behaviour of fluid flow in a piping systems.

The system consist of a double acting electrohydraulic servoactuator attached to a stainless steel single acting low friction cylinder, which was attached to the water-filled piping system.

The electrohydraulic servoactuator consisted of a low friction double acting cylinder with a position transducer (LVDT) concentrically mounted in the rear of the rod and a servovalve manifold mounted to the actuator. The rather short stroke length required (variable) is 1mm to 40mm.

The controls included a function generator to generate sine, triangular waves from 1 Hz up to 10Hz, a servoamplifier for position control of the actuator. Pressure of the water in the piping needs not to be controlled. The equipment also included a hydraulic power unit.

E.g of the system running setting:
a) Stroke length = 11mm
b) Frequency = 3Hz
c) Waveform type = Sine wave

Our vendor has suggested some specialised servo hydraulic control software which is expensive. We therefore would like to explore the possibility to control it with PLC. Can anyone offer some advice how this can be done? Can someone recommended the components on the servo hydraulic part? Particularly, how do we integrate the PLC with the function generator?
 
From what you've posted so far I think your vendor is right.

If this can be done in a plc at all it is going to be a pretty expensive plc, probably more expensive than the motion control software you are being quoted. If your time doesn't mean much to you and you have the knowledge you may be able to do this with an embedded controller. But I think you are much farther ahead using a proven hydraulic motion controller for this.

Take a look here for some ideas.

Keith
 
I would recommend the RMC75E-AA1. It has a built in sine wave function and the square wave. The amplitude and frequency can be changes smoothly on the fly so you can to linear or logorithmic frequency sweeps. Sawtooth, and triangle waves can be generated using simple commands. There are also cam table that will all you to make any profile you want. The Ethernet interface makes it easy to get the motion profiles out of the controller real time. The controller can do this easily. The controller is only moving electrons around. The design of the mechanics and hydraulics are critical. The hard part is the design of the test system. Test systems that go over 5 Hz are tricky. First need a fast and linear servo valve. The valve must be mounted directly on the cylinder. There should be a small accumulator to keep the oil supply at a constant pressure. The can be no compromises on these points.

The LVDT will need to be amplified to 0 to 10 volts or 0 to 5 volts but 0 to 10 volts is better. The controller outputs -10 to + 10 volts. We sell amplifiers that convert -10 to 10 volts to current for servo valves. The valve manufacturers sell amplifiers cards that convert the -10 to +10 volts to current if the valve is a proportional valve.
 
Although I work for a Rexroth Distributor, I agree with Keith and Peter that the Delta RMC with built in function generation is the way to go here.

To recommend components as you request (I'm assuming you are looking for recommendaitons on the valve here) you need to fully define the systme such as what pressures you will be running at, bBore and rod of the cylinder, accuracies required, what forces the water side will resist the hydrauilcs with, what temperatures you need to work within, etc, etc, etc. For instance, some valves are have their flow rated at a 10 bar pressure drop and other a 100 bar pressure drop but what really matters is the pressure drop in your application. Do you konw how to calculate this? Hopefully, your vendor does. Getting this wrong usually leads to you having a valve that is too large for the applicaiton and you end up only using 20% of flow capacity so you spend more money to get only 20% of the performance you should have.

You will be best served by working with someone who undersands hydraulics.
 
The load for the water piping is rather low, it only pushes water volume of 10 liter inside a 3" pipe at atm pressure to creat turbulence flow.

Actuator:
Equal area 4kN
Working stoke: 40mm
0.314m/sec
Low friction seal design
LVDT displacement transducer

Valves:
High respond proportional CETOP3 servo valve
24V Safety slice for port CETOP3 isolation

attached also the schematic of the hydraulic system

http://www.plctalk.net/ganda/uploads/servo_hydraulic1.jpg[img]

I wonder why the valve must be mounted directly to the actuator/cylinder.
 
csiong said:
The load for the water piping is rather low, it only pushes water volume of 10 liter inside a 3" pipe at atm pressure to creat turbulence flow.
So the water isn't being compressed, just agitated?

Actuator:
Equal area 4kN
Working stoke: 40mm
What does the 4kN mean? Normally the cylinder diameter, rod size and length are given.

0.314m/sec
How did you compute that?
The formula for peak velocity for a sinusoid is
Vpeak=Amplitude*2*PI*Freq
Amplitude=half of 11 mm or 0.0055 m
Freq=10 Hz
PeakVelocity=.3455 m/sec
The formula for Peak acceleration is
PeakAcceleration=Amplitude*(2*PI*Freq)^2 or
PeakAcceleration=21.71 m/sec^2
That is over 2g. It is a good thing you are just agitating water and not really moving anything massive.

attached also the schematic of the hydraulic system

http://www.plctalk.net/ganda/uploads/servo_hydraulic1.jpg[img]
[/quote]
I couldn't see the drawing even it I cut and pasted the link in my browser.

[quote]
I wonder why the valve must be mounted directly to the actuator/cylinder.[/QUOTE]
Because oil compresses. Even water compresses. The more oil between the valve and the piston the more oil there is to compress. It takes time to compress the oil and that limits the rate of change in pressure or force.
When making a test system it is important that the hydraulics and mechanics have a natural frequency of at least 3 times the frequency of the test. In your case you need a natural frequency of about 50 Hz.
The valve needs to be at least as fast.

See these
[url]ftp://ftp.deltamotion.com/public/PDF/Natural%20Frequency%20and%20Ramps.pdf[/url]
[url]ftp://ftp.deltamotion.com/public/PDF/SpringEffectEffBulkModl.pdf[/url]
[url]ftp://ftp.deltamotion.com/public/PDF/Mathcad%20-%20Natural%20Frequency.pdf[/url]
 

Similar Topics

This webinar was held yesterday. I didn't post a link to it. We had over 200 people attend it anyway. There was more question and answers on...
Replies
1
Views
1,193
Since there are a lot of UK people here I thought I would ask. Has anybody taken classes at the NFPC ( national fluid power center ) in...
Replies
3
Views
1,708
Hello everyone, I've been reading through the forums on motion, servos, etc. I'm looking for some help in determining the cause of some...
Replies
45
Views
12,976
Hey Guys, Got an application where the Engineer purchased some Hydraulic Servo Valves and would like me to write a PID to control the position...
Replies
3
Views
1,639
Just the other day I saw a hydraulic forming press that didn't seem to have a pressure relief valve. It had a Servo motor connected to a pump that...
Replies
5
Views
2,065
Back
Top Bottom