Metering Pump and Viscosity

Yes...the pump manufacturer.

Any pump of this type will need 'calibrating' according to the fluid pumped, back pressure, pipe dimensions etc.
Often they have an adjustable stroke to allow for adjustment.

And before some smart arse one says 'how can you measure the delivery in situ?'.
You can measure how much is draw up over a number of strokes (the ammount delivered will be virtually identical to that drawn up).

Del

The best the manufacture going to give you is the rate with a set viscosity @ temperature in which the pump was designed around.

Within the permissible vapour pressure, theoretically there should not be any difference in the quantities pumped for fluids with different viscosity.

But practically there is substantial difference due to: Slip. Viscous shear is a factor affects the slip, more the viscosity less is the slip.

Slip is possible at three areas: 1. Through suction valve (slips back to suction header during discharge stroke), 2. Through discharge valve (slips back to pump chamber from discharge header during suction stroke), 3. Through gland (slips back and forth into stuffing box).

As 'Del the cat' mentioned, "the amount delivered will be virtually identical to that drawn up", is true if there is no slip through glands.

Since these metering pumps are to handle very less quantity, as indicated, calibration/ensuring is a must. In my experience 8 out of 10 pumps works incorrectly. This more so with diaphragm pumps, that too at low stroke.

The following arrangement is the best I have come across for ensuring the performance.

Day tank is small container, which holds liquid for just a day such that the draw-up can closely be watched shift wise.

Further, if the level indicator is isolated, the draw-up is just from the level indicator only and pumping rate is well witnessed. It is interesting to see the existence/effect of slip. The level goes downs during suction stroke and rises a bit during discharge stroke, explaining the slip.