Pulsafeeder Pulsar Shadow - Manual (Page 20)

15
6.5 Calibration
Figure 13
All metering pumps must be calibrated in order to accurately specify stroke length settings for required
flow rates. For pumps provided with DLC or DLCM electronic stoke length control, refer to separate
instructions as noted on page 8.
A typical calibration chart is shown in Figure 13. Although output is linear with respect to stroke
length setting, an increase in discharge pressure decreases output uniformly, describing a series of
parallel lines, one for each pressure (only two are shown).
The theoretical output flow rate at atmospheric output pressure is based on the displacement of the
hydraulic piston (the product of piston cross-sectional area and stroke length) and the stroking rate of
the pump. With increasing discharge pressure there is a corresponding decrease in output flow of
approximately 1% per 7 bar (100 psig) increase in output pressure. Whenever possible, calibration
should be performed under actual process conditions (i.e., the same or a similar process liquid at system
operating pressure).
To assure a sound hydraulic system, run the pump for 10-15 minutes prior to calibration. This will
allow the PTP (automatic bleed) valve to purge any air from the system.
To construct a calibration chart, measure the flow rate several times at three or more stroke settings (i.e.,
25, 50, 75, and 100), plot these values on linear graph paper, and draw a best-fit line through the points.
For stable conditions, this line should predict settings to attain required outputs.
Checking the actual flow rates is especially important in pumps producing low flow rates and operating
against high discharge pressures. In this type of system, normal losses of efficiency can result in lack of
measurable flow at shorter piston stroke lengths. This is a function of the system conditions and does
not indicate a problem with the pump. Careful measurement of actual pump flow at several test points
will allow for proper calibration over the complete flow range.