We rely on rapid prototyping

Curved measuring channel

This concept follows the approach that the flow should be as independent as possible of the inflow conditions into the measuring device (Figure 2). The flow is determined with the aid of a differential pressure. The connection to the measuring device, whether realized with a right angle or straight connection, should not influence the pressure difference, since the flow is more strongly influenced by the curved geometry in the measuring device.

CFD simulation curved measuring channel

Using a CFD simulation (RANS, static), the hypothesis is tested whether an inflow into the gauge with a straight inlet produces an unchanged differential pressure compared to a right-angled inlet. The two configurations with 0° inlet and 90° inlet are shown on Figure 3.

Comparative measurements on the prototype

To verify the promising simulation results, a prototype is 3D-printed and a measurement setup is built (see Figure 4). With the help of a measuring device, the differential pressure is recorded at different flows. It is found that the measurement results and the simulation qualitatively agree quite well.

Learnings

Thanks to these investigations, the following insights are made for an improvement of the concept:

• CFD simulations can be used to predict trends for the relationship between flow and differential pressure.
• To obtain more accurate results from the simulation, an improvement in accuracy is needed. With the help of a time-dependent calculation, this can possibly be achieved.
• For fine tuning, measurements on physical prototypes are essential.

Thermal approach

In this concept, the flow is to be determined with the aid of a thermal mass flow meter. The advantage of this measuring method is that no inserts must be used, which get dirty and have to be replaced regularly. Furthermore, the resolution is very high, especially for small flows.

Figure 6: Schematic representation of the MATLAB Simulink simulation consisting of the electronics and physical model for flow and heat transport.

Various influences can now be investigated in the simulation. In this way, an understanding of the measurement principle is gained. Figure 7 shows the results of the simulation. The flow is increased uniformly. As expected, the measurement signal also increases. It can also be seen that good resolution can be achieved in the lower flow range. Further simulations are performed to check the influence of variable ambient temperature. Also, these results (not shown) show that with the chosen design a next step in a physical prototype is possible.

Comparative measurements on the prototype

In the next step, an electronics engineer creates a PCB for the driver and the sensor is mounted in a physical prototype. Figure 8 shows the measurement setup with a 3D-printed insert for the sensor, PCBs and flow meter for reference measurement of the flow.

In the experimental setup, the flow can be increased stepwise by 10 lpm at a time from 0 to 300 lpm. The flux curve and the measured signal in this experiment can be seen in Figure 9. As desired, the measured voltage gradually increases with an increase in flux. Comparing the simulation with measured results, a good qualitative agreement can be seen.

Learnings

The following insights can be made from this concept:

• The MATLAB-Simulink model allows us to test the measurement principle without the need for hardware.
• Furthermore, thanks to the simulation, a PCB for the sensor can be designed with increased certainty of success.
• The measurements show qualitatively good agreement with the simulation.

Perforated plate

In this concept, the gas flows through three in series connected perforated plates. The differential pressure (dP1) is measured at the inner perforated plate. This measurement is used for high flows. Additionally, the differential pressure (dP2) is measured between the first and the last perforated plate for measuring low flows.

This design has many influencing factors (e.g. hole diameter, open/closed area, number of holes, thickness of perforated plate, distance between perforated plates, ...), which also affect each other. In an initial design, elementary formulas are used to create a design that ensures laminar flow through the holes in the entire flow area.

Modular prototype

In order to check the many influencing factors, this concept is based on a modular prototype in which various perforated plates can be easily exchanged. Figure 11 shows the four-part measuring channel with three interchangeable perforated plates. The measuring channel and the inserts have been manufactured in two different processes to ensure that the parts are gas-tight and to obtain sufficiently accurate holes.

Measurements

Figure 12 shows the measurement results for the prototype with perforated plates. Since there are two differential pressure measurements per insert variant, a total of four curves can be seen. In the first series of measurements, three perforated plates manufactured using the 3D printing process are measured ("Process 1"). A second series of measurements with inserts produced in a different process ("Process 2"), is also recorded. Contrary to expectations, the difference is significant, especially in the full-range curves. A possible explanation for this difference is rounded edges. On closer examination of the inserts under the microscope (see Figure 13), the rounded or sharp edges can be seen very clearly. The measurement also shows very well that the required maximum pressure drop at 300 lpm is significantly smaller than 40 mbar.

Figure 13: Round edges on perforated plate produced using SLA printing process (left), sharp edges on perforated plate with drilled holes

Learnings

The following conclusions can be drawn from this concept:

• With the modular prototype, different variants can be quickly tested and measured within a short time.
• The measurements on the prototype show early on that certain manufacturing tolerance (e.g. radii/sharp edges of the holes) are important and have a great influence on the "performance".