I am trying to pump liquids with viscosities ranging fom 200-1500 mPa.s. The NPSHa is higher than NPSHr. The flow is irregular although the discharge pressure is accurately measured and is constant. Could the discharge metering line be the problem if it was too wide? Could some backpressure be created if a ball valve is installed right after discharge?

First of all check your Releif Valve.Calibrate it to ensure its set pressure.
Even air ingress thro pump casing & excessive wear can cause irregular discharge

relief valve is perfectly set and pump is vented. The pump is brand new!

Are we to infer that your indication of irregular flow is the flow meter? Or do you have other observations?

Hi Gary. Actually there is no flow meter with this pump. The final product that is being manufactured using raw materials driven by this pump happens to be defected. This is why we are having doubts about the pump's irregular flow. Luckily we already have an existing functional pump that comes in parallel to the gear pump.
The new gear pump has no noise coming from the relief valve and its rotation seems quite smooth.

I would have first thought it might be irregular viscosity as I have seen that before. If your second pump begins to have issues in near distant future then I suggest looking for a solution of something else. Not all liquids are equal and if some type of solidification is occurring that makes a whole new ball of wax.

Anthony,
No one can solve this problem for you via internet.
You need to go out in the field and perform a test!
You did not furnish enough basic data to expect any sensible answer anyway.
Sorry.
Regards.

What is your process liquid?
It it sheer sensitive - sheer thinning / sheer thickening?
Are there solids or particulates?
Does it solidify or deposit out in the lines?
Are you using VFD to control speed?
If so, how is speed controlled?

May I assume you are using / want to use the same pump for the entire range of viscosity?

A gear pump (or virtually any other positive displacement pump such as screw or lobe) typically have internal tolerances are adjusted for viscosity. For thin fluids the tolerances are tight to limit internal slip (recirculation) within the pump. These tolerances increase with viscosity. A pump sized for a thin fluid will bind with more viscous fluid. It is suspected your pump was selected for the high end of your stated viscosity. If so, that pump used with a thinner fluid will have greater internal fluid slip and flow will decline. So you should expect the highest flow with the highest viscosity and lowest flow with thinnest viscosity. System friction losses also change with viscosity, increasing with viscosity. So at a constant flow the pressure will change - increasing with viscosity

What is your objective, constant flow with variable viscosity using one pump for all fluids / viscosities?

Hi Robert,
I think your suggestion and grabing the problem should be correct.
Shafiqur

I noticed you say 200 to 1500. I the liquid non newtonian. If so how much is the pump shearing it? That can cause a irregular flow. If this is the case, compare the shear of the pump to the liquid.

Leonard, thanks a lot for your comment.
Robert, the process liquid is a special type of alcohol based product with well dissolved solid particles that have a solidification point of -20 deg C, so basically nothing to cause any pump wear. Speed of the motor is controlled via VFD. The objective is to get a constant flow and a contant pressure.
Daniel, my main concern was to increase the backpressure on the discharge side in order to increase the internal friction.

Thanks a lot