Venturi Flow Rate Sensor

This Venturi flow rate sensor is designed to be used with MEMS differential pressure sensors and a microcontroller to measure the flow rate of a gas mixture. It is intended to be printed on a high resolution 3D printer and has barbs on its inlet and outlet so that it can be placed inline with 3/4" ID tubing. The device is modeled in inches.

Step 1: Pressure Sensors and Setup

This Venturi is designed to be used with Honeywell 060MDSA-3 differential pressure sensors whose 6kPa sensing range allows flow rate measurements from 0 - 1.5 LPS with this Venturi design. The sensors communicate with SPI so they can be used with an Arduino or similar microcontroller to take flow rate measurements. This technical note from the sensor manufacturer details the communication protocol used by this sensor . In short, to take a measurement, 4 bytes are read from the sensor; the first 2 bits indicate the sensor status, the next 14 bits are the pressure reading and the following 11 bits are the temperature. The last 5 bits are ignored. To set up the Venturi, connect its pressure taps to the two ports of the differential pressure sensor using 3/32" ID tubing (e.g. this Tygon tubing from McMaster-Carr). The Venturi itself is placed inline in a 3/4" ID tube, with the arrow on the Venturi indicating the intended flow direction.

Step 2: Measuring Flow Rate

The above equation enables the calculation of flow rate through the Venturi from the differential pressure reading. In our calibration, we found that the discharge coefficient is 0.6. The diameter ratio is 0.3 and the throat area is 17.82 square millimeters. The density will depend on the gas mixture measured and its operating conditions.

Our test setup for calibration and the resulting calibration curves are shown as well. The discharge coefficient, C0, was determined experimentally using a rotameter flow meter as a calibration instrument (Cole Parmer item #: 03218-41). High pressure air from a compressor was regulated down to 25 PSI and passed through the rotameter, which throttled the flow rate down to a range of 0 - 58.5 LPM with an accuracy of +/- 1.17 LPM. Flow through the rotameter was correlated to a pressure differential between the Venturi ports for increments of 5 LPM between 0 and 58.5 LPM. The parameter C0 was adjusted from a theoretical value of 0.9 to 0.6 in the governing equation to fit the results of the calibration study.