› PLCGurus.NET Live & Interactive Forum › PLC Questions and Answers › Rockwell, Allen-Bradley, Stratix › Rockwell – CASCADE PIDE CONTROL
Tagged: #Cascade #PIDE
For my project, I need to control a flow control valve (FCV). Where a pressure transmitter is located before FCV to check pressure and a flow transmitter is located after FCV to check flow rate. You may refer attachment for clearer image.
I believe this is a cascade control where,
I have started a draft of the Cascade PIDE, as attached. But I think something is wrong because it fails when I try to simulate using Logix Emulate. Can anyone please advise or help?
That is not a case for PIDE cascade control, you would use cascade control if the output of the first PIDE is the setpoint for the next. You need 2 seperate PIDE’s with a low select going to the valve.
Thank you for your advise! Did you mean this should be a multiloop control, scheme on page 7 as attached in pdf file?
Also i have attached a draft of a multiloop PIDE.
Yes, the 2 loops interact as per page 7 with it set for low select. Make sure to bring the output of the select block back to the CV Previous of both PIDE’s so the non- controlling one stays active and doesn’t windup.
I believe the primary concern is flow rate and having a nominal pressure as you would not be able to maintain flow if the pressure drops too low or your flow might be too much if pressure was too high (there might be other concerns such as pipe engineering limits etc.., your pressure will change as flow rate increases / decreases – your pressure is a limiting factor. So your desire is to maintain pressure first and then let the valve control the flow – the valve closing and opening effects the pressure – so your flow rate is what ultimately changes that pressure via opening and closing of the valve.
x flow at 8.5 pressure
What is the critical importance of the variable being controlled for the process – is it flow rate or is it pressure ? You only have one pipe and one valve so it can not be both.
The assumption here is that what you are really trying to achieve is that you want to maintain flow rate – and the input pressure to be maintained to ensure that flow rate is stable (otherwise who cares about the flow). If pressure drops too much you can not maintain flow, if it increases too much you would have spits and fits.
You could use flow rate to input into the secondary loop to control the pressure (your cascade is backwards in my opinion). Based on the criteria.
While you can use a select as to which PIDE is controlling the process <8.5 pressure PIDE,
> 8.5 flow PIDE , however this assumes that a given pressure results in a given flow rate and you have the engineering calculations done to know the proper test values – hopefully you do – because the cascade will work.
Remember you are using one valve which effects both variables .. so lets look at the process..
You can probably just as easily control the flow on an optimal setup – your pipes are only so big, the fluid is of a certain viscosity – it will only flow at a given rate per pressure – fluid dynamics. duckDuck some engineering examples of controlling flow based on pressure – because that seems to be the goal to control the flow and not allow the pressure drop below a certain point.
One of the things that comes to my mind is in winding applications where torque and velocity are controlled in order to maintain a smooth wind without dish-panning.
Same basic structure of process .. torque (pressure) , velocity (flow).
You are fine with the cascaded loop but you need to cascade the proper way.
1: What is it that is the process control – pressure or flow ? (Flow)
2: What are the limiting criteria on that flow ? Engineering – pipes, valves liquid(viscosity) and Pressure.
The pipes can not be dynamically changed .. your pressure can be.
FLOW is the the thing to output to the control valve being limited by pressure – therefore feed the flow rate into the pressure control and always maintain pressure at x rate to get n flow – pipes are only so big and with 8.5 bar (value of x) of pressure flow will only be n rate.
I know its wordy and said a couple different ways but hope that helps.