Continuing our look at Accurate Calibration Gas Using MFC Based Dynamic Gas Blenders, Tom discusses how to select the right Mass Flow Controllers (MFC).

Qualifying and Selecting a Mass Flow Controller for a Dynamic Gas Blender

Just as the MFC is the heart of a high accuracy gas blender, the flow sensor is the heart of a mass flow controller. The most common and traditional technique for measuring and controlling mass flow is with a thermal mass flow controller. As noted earlier, this type of device measures the temperature shift (or differential) that occurs in a small tube as heat transfers to and from the gas.

A thermal mass flow sensor is shown in Figure 1. As gas flows through the sensor tube, it gathers heat when it enters the tube and transfers some heat back to the tube as it exits. The temperature differential in the tube is measured by two independent temperature sensors, TC1 and TC2. The electrical signal generated by the sensors, is amplified and transformed to a linearized signal that is commonly recognized by instrumentation.

When selecting an MFC for a gas blender, it is important to review sensor data provided by the MFC manufacturer. The data should include long-term stability and repeatability. Manufacturers of gas blenders need to have on hand all of the research data on the type of sensor employed to ensure that no inherent flaws may be present.

When flow controllers are calibrated, they should be calibrated in conjunction with the process gas that is to be analyzed. However, this is not always possible. When the process gas cannot be used for calibration, a surrogate gas can be used as a substitute. A conversion factor (K) is applied during this calibration to ensure that the flow through the MFC is accurate with respect to the process gas.

It is incumbent upon the manufacturer of the dynamic gas blender to confirm that his supplier of the MFC device has committed sufficient time to the proper evaluation and development of accurate K-factors for all of the gases that will be used with the blender. Ideally, K-factors should be derived empirically. Figure 2 is a partial list of one company’s K-factors and the methods used to derive them. (NOTE: A “K” factor is defined as the ratio of the actual gas flow rate to the equivalent nitrogen flow rate. To obtain the equivalent nitrogen flow rate, divide the actual gas flow rate by the K factor. Example: The K factor for argon = 1.45. QEQUIVALENT NITROGEN = (QACTUAL GAS / K) = 101.94 slpm / 1.45 = 70.3 slpm.)

The most important step that a gas blender manufacturer can take when selecting an MFC is to evaluate one set of MFCs, over a period of several months, from each qualifying manufacturer. The most important characteristics to analyze are repeatability, stability, and short and long term drift. Figure 3 illustrates an evaluation that spanned seven months. On the day of each test, the MFC was flow-tested, using a primary flow standard at four points, throughout the full-scale range; 10 repeatability points were taken at all four major flow rates to establish the average repeatability of each test.

To read more about the manufacturing and how to maintain a dynamic gas blender, stay tuned for the final installment, Part III!