metering viscous fluids

[automatic2]

I'm looking at metering a PVA type liquid (wood glue) and was wondering if a pnuematic diaphram pump would give me any repeatability. Anyone try such an implementation?

 

[Latexman]

Never tried counting strokes on a AOD pump. I thought it would be too variable. We do have some success using a mag meter on such fluids. Just keep the electrodes submerged and wet at all times.

Good luck,
Latexman

 

[Latexman]

Viscosity range of the PVAs?

Good luck,
Latexman

 

[automatic2]

ranges from 1500 to 2500 cps, so I'm going with 2000 cps at ambient.

 

[automatic2]

only looking for about 20gpm batch dispensing. Thinking a 2" supply line to flood the input with about a 4' head. Outlet would be 1-1/2" with about a 12' horizontal run to atmosphere.

 

[Latexman]

Most likely it's shear thinning, so a mag meter should work. Again, keep the electrodes submerged and wet at all times.

Good luck,
Latexman

 

[jmw]

You don't say much in your first post and it is only in your third post you give us a flow rate, but just the flow rate even though you then say you are batch dispensing and you don't tell us the batch size.

To give any kind of sensible answer requires sufficient information.

If you are filling 100ml containers it makes a considerable difference to filling 10 litre drums or 50gallon drums (US or imperial by the way?).
In any batch application you have to look at the stability of the flow and the repeatability and duration of the end of batch sequence, valve closure times etc. whether you have multi-stage shut down or simply want to use the pump on off - do you have VFC control over the pump?

Also, what accuracy standards do you have to meet? Is this retail or wholesale or internal use?
In retail you must fill each container with at least the amount declared. That means that when you declare these to be 100ml containers, you must always deliver 100ml or greater.

You can do this with a 2% accuracy or a 0.1% accuracy. Obviously the 0.1% meter ensures that you give away far less product than a 2% accurate measurement (lets just keep referring to it as accuracy for convenience for the sticklers out there).
If you are not retailing then it may be enough that the mean value is 100ml. So long as systematic errors are calibrated out, the random errors will balance out within a very few batches. In such a case increased measurement accuracy may cost you a lot more but not deliver any significant benefits in the long run.
But how should you meter it? Pump cycles, adjustable stroke pistons, gear pumps or flow meters?
We really need to know batch size, flow rate, and purpose.

(Now if this were linked in someone would have already told us coriolis meters are the best solution without bothering to ask any questions .)



JMW

http://www.ViscoAnalyser.com