Help with Preact or Dribble Meaning....

[JoseM]

Can you guys help me understand what is the concept behind "Preact" and "Dribble" when controlling the supply of a product.

I have a mix tank that keeps overshooting the amount of fluid going into it by about 50 Gal. I was just wondering if I could used those two constant values, "Preact", and "Dribble" to reduce the overshoot rate.

 

[iant]

joseM youve been here long enough
more info please
as you say the over flow is 50 Gal's than it is OK to stop the flow and reduce it @ 50 Gal's less than the setpoint
- i bet your about to come back to us you only need 25 Gals....

 

[danw]

Typically the preact value is a setpoint lower than the final batch preset value. When the preact value is reached, the controller stops the delivery of material at that point to allow the material already "in-flight" to finish off or complete the batch. The assumption is that the amount "in-flight" is a constant from batch to batch.

Dribble can be a 2nd control output that controls a feed line that is a fraction the size of the main feed line, such that incoming material dribbles in at a slow rate compared to the main feed line.

You'll have to read the documentation to see exactly how each is implemented.

 

[JoseM]

Sorry for not providing enough information. This application is a batching process like "danw" mention. The situation is the following, if I set a recipe for a batch that calls for 1100 Gal of a fluid to be added to a tank, the system delivers more than what the set point of 1100 is calling for. Sometimes it delivers about 50 - 60 Gal more and it only happens on this tank and presumably on only one ingredient.

 

[mellis]

Simple preact works like this...

Setpoint - 1100 gal (the amount you want)
Spill - 50-60 gal (the amount you overrun)
Preact - 50 gal (about the same as spill, but constant)
Actual - (live reading from totalizer or scale)

The addition valve is open during the addition step when:
Actual <= Setpoint - Preact.

You should expect Actuals of 1100-1110 gal with a preset of 50 gal if you were getting 1150-1160 gal before. The preset can't do anything about the variation in the spill.

The key thing here is: Spill is the amount of material that gets through the addition valve between the time you detect the "Actual <= Setpoint - Preact" and the time the valve actually closes. This is true if you are using a flowmeter with the valve downstream. If you are using a scale, the amount held in the line that will drain into the tank also matters.

It can get more complicated by taking into account the flowrate during the addition, but 95% of the time it's not necessary.

P.S.
Since you mentioned that other materials don't have this problem, I recommend you take a look at the time that valve takes to close. If there is something wrong with the valve that is making it slower than it should be, fix the valve rather than change the program.

 

[JoseM]

All the ingredients used during the batching process go through the same valve but presumably only one of them is "Overrunning" the set point amount.

[ ]
(I1)--------------------------{Valve}---------------[ TANK ]
| | [ ]
| |
| |
(I2) (I3)

 

[mellis]

JoseM,

Your sketch didn't come through so well.

I'd still look at what is different about this addition. For instance, if it runs at a higher flowrate that the others, the spill will be greater too. If that's the case, the preact would be appropriate.

 

[micky]

JoseM,

Your sketch didn't come through so well.

I'd still look at what is different about this addition. For instance, if it runs at a higher flowrate that the others, the spill will be greater too. If that's the case, the preact would be appropriate.

i.e. viscosity of the fluid

 

[Lancie1]

Dribble can be a 2nd control output that controls a feed line that is a fraction the size of the main feed line, such that incoming material dribbles in at a slow rate compared to the main feed line.

If the valve controller does not have a built-in Dribble Mode (many do), then I have created my own dribble mode for standard solenoid valves using a timer. Stop the main feed before the setpoint (based on trial runs), wait for the amount in the pipe or chute to flow into the container, then open the solenoid again for a very short time period (pulsed output, and can be based proportionaly on difference between setpoint and measured amount) to add the last small material amount.

 

[Mickey]

Here is an example.

http://www.hardysolutions.com/produ...ltaneous_twospeed_container_filling_files.pdf

 

[JoseM]

I didn't mention it but the fluid in question for this problem is water. It does flow at a higher rate tan fructose syrup. Now, does the Preact has to be lower or increase.

 

[mellis]

In this case, I'd start with a preact of 50gal on the water addition. In general, I'd make all the additions use a preact and just set the other ingredients preact = 0. But that's just me.

You haven't confirmed if this is a flowmeter addition or a weigh scale addition, they behave slightly different. Since you were talking gallons and not pounds, I assumed flowmeter.

 

[JoseM]

This is what really baffles me, that the "Flowmeter" which is used to total the amount of water being put inside of this mix tank, when the process ends, the "Set Point" vs. "Actual" is only about 2 - 3 Gal. off.

But when I go to the tank, the amount of water I can see on a side glass and the amount measured by an "Anderson Level Sensor" exceeds the "Set Point" by about 50-60 Gal.

 

[Mickey]

Is the flowmeter total from the flowmeter or is it totaled in a PLC. If PLC how often is the total updated?

"Man with one watch knows what time it is, man with two watches is never quite sure."
Lee Segall

 

[Lancie1]

This is what really baffles me, that the "Flowmeter" which is used to total the amount of water being put inside of this mix tank, when the process ends, the "Set Point" vs. "Actual" is only about 2 - 3 Gal. off.

Then your PLC program is doing its job fairly accurately. Using Preact or Dribble functions can only help to eliminate the 2 or 3 gallon error, not the 50 gallon difference between flowmeter and sight glass. This difference error has to be in the flowmeter calibration or the PLC scaling for the flowmeter input. Is the flowmeter calibrated for room-temp water while the actual water is near boiling (less dense)?