GT Control: Servo Operation

N

Thread Starter

Neo

CSA:
I am a little busy previous time, but i would like to spend more time on it.
How to get it started?<pre>
Toolbox---hardware(firmware)---servo valve---hydraulic actuator---IGV/SRV/GCV
| |
| LVDT |
|-------------------------------------------------------|
</pre>
Above is the control loop i know about servo operation.

CSA, i would like to know your comment.What should a engineer should know about servo-operation?

Best regards!
Neo
 
I am now also waiting for CSA to give a description of how that new flanged digital valve position control provides the servo current.
 
I am wondering when the SRV/GCV/IGV valve is at desired position, will the servo-valve move back to middle position? At that time,the pressure exerted on the both spool valve are the same. how can the spool valve move back to middle place? I hope my statement is clear.

It is interesting that i find Chinese in JFB's document, and thanks again for your document, it helps!

I hope i can see CSA's comment on servo-operation, too.

Best regards!
Neo
 
The amount of movement from the middle (lets call it NULL) determines the percent of oil flow to the actuator. thus for example, a 10% current will only open the spool for 10% flow. this would have the valve opening rate equal change in volume under the piston as stroke changes equivalent to flow rate.

there can be some design leakage in the actuator so that 0 flow will result in the valve drifting closed, so in closed loop position control the LVDT will sense the change and thus the control loop will supply a small current (and flow) to keep the valve at very very near desired position.

if the position demand is ramped, say at 10% per second from 10% to 90% open, the position control error will generate a current in an attempt to have the valve travel match, based upon the servo flow design. this may mean the servo will have to open 50% to meet the ramp. as the actual position approaches the desired position the current will decay until it is just enough to provide the leakage with the servo spool slightly off from null.

On steam turbines, the servo valve spool at 0 current is off set so that it will result in a valve drift closing. So in addition to a current needed to maintain the leakage, there will be an additional small current need just to bring the spool to null.

The reason for the Chinese on the drawing, that PDF was part of a similar discussion I had previously on LinkedIn where the inquiring engineer provided me with his drawings.
 
M
I am not sure, but there might be overshoot, so the spool valve will return to its center position as a result of the overshoot. LVDT feedback is "higher" than demand so spool valve is moved back. This sequence can be repeated until the spool valve enters in its deadband. This happens so fast that cannot be noticed.

Some spool valves have a safe spring, so you need a null current bias continuously actuating. thus in case no electric input, the spool valve will relief the piston to close.

This is my idea, do not take it as an statement and wait for answer of other professionals.
 
Neo,

I've not forgotten about this thread. I'm traveling (outage season--so long hours and lots of schedule over-runs) and I'm trying to "draw" and it's time-consuming.

Bear with me. I think it will be worth the wait.
 
Hi CSA

There are a few control loops here. There is one implemented in physics that relates the torque in the top stage of the servo valve to the spool position via a 'bendy' pin.

This should help: http://lmgtfy.com/?q=moog+servo+valve+operating+principle
and look at the first result.

Then there is a control loop implemented in electronics that looks at the LVDT signal and compares it with the the demand signal. Then you'll use some proportional and often a little integral on the error signal to create the drive signal to the valve.

Hope this helps.
 
CSA:

I know!

I am waiting for your response and i am reading the application code at the same time.

Best regards!
Neo
 
Top