level to flow

[asterof]

I need to determine flow based on loss of weight from a load cell. I know the volume of the tank.
tank is 6 x 6 x 6 which is 216 cu/ft
This means there is approx 1616 gallon of material
A gallon is 8.34 lbs so we have 13411 lbs of material
so there is 6 x 12" = 72" 13411 / 72 = is approx 182 lbs
per inch.
so for every loss of 182 lbs we have approx 21 gallons of flow

So what i need is a calculation that take the loss of weight
calculates the flow in lbs and if the flow does not change
within a given dead band sets an alarm
Anyone have a calculation
Thanks

 

[iant]

you need to take into attention with orifice size and hose I/D's etc. theae will limit your flow. unless you are using a free flowing chute or something.

 

[Steve Bailey]

Sample your weight once every 6 seconds. Calculate the difference between the current sample and the previous sample. The difference, multiplied by ten, is the rate of change in pounds (or kilograms) per minute (because six seconds is one tenth of a minute).

If the rate of change is less than whatever tolerance you establish, turn on the alarm.

 

[Ned_Flanders]

We have several loss-in-weight feeders and i would suggest this is not a simple task.
Depending on the flow rate,it might be an idea to sample every second - or even more frequently.
Is the system just an alarm or are you using it to control a variable speed drive?
Also, consider what happens when the tank refills (the weight will go up, not down.)

 

[Lancie1]

What I need is a calculation that takes the loss of weight, calculates the flow in lbs, and if the flow does not change within a given dead band, then sets an alarm.

If an alarm is all you need from LIW, then you do not really need to convert to pounds. Simply use a self-resetting timer for some appropriate time period. Each time that the timer DONE bit goes on, compare the New Weight raw counts to the Old Weight raw counts. If no significant change, then set the alarm. Also while the DONE bit is still on, save the New Weight to a memory location for the Old Weight.

However, if you need the raw counts from the load cell inputs to be converted to pounds or pounds/hour, then here is an example using RSLogix500. In this case, the cycle timer is a variable, so that adjustments can be easily made to get better data for different input ranges.

Rung 0007 is where the final flow rate in lbs/hour is calculated based on the loss-in-weight from a load cell analog input. Notice that during tank (test tube burrette) refills, that new calculations are inhibited, so that the process pump runs at the same old rate during the short refill time. Otherwise the process would be severely upset due to the weight gain instead of the expected normal weight loss.

 

[daba]

Exactly as Lancie1 has pointed out...

The flow out of the tank is proportional to the loss-in-weight of the tank, so you can directly use the load-cell data without needing to convert to flow.

So your original problem boils down to "have we had the correct loss from the weight of the tank, over the respective time period".

Just think of some value x that you want to reduce by a value y in a given time z, rather than trying to determine flow from x, y, and z, and it all becomes much easier.

 

[Lancie1]

Depending on the flow rate, it might be an idea to sample every second - or even more frequently.

The problem that I ran into for doing LIW control with a PLC program is that the more often the sample is made, the smaller the weight loss. For large tanks and short time periods, the number becomes so small that no change is seen at all, which makes process control difficult. On the other hand if the sample time is very long, the PID output oscillates, with a large change each time the weight signal is read. There is a trade-off for the sample time, so you have to pick the proper sample time for each LIW controller.