Keys to Successful Gas Flow Measurement
Contributed by
Omron
Fluid dynamics is a very complex field of study. Commonly used calculations often take on assumptions of idealized conditions, when the actual case may be far from it. Even after running highly complex computer models, engineers frequently find themselves making adjustments to their designs after physical prototypes are built and tested.
The behavior of gas flow and subsequently achieving a stable, accurate mass flow measurement is affected by many factors. The composition of the gas will affect properties such as density and thermal conductivity. Environmental conditions like temperature , humidity and altitude play a part, as does the medium through which the flow is moving. The surface roughness of the tube/pipe/duct influences the resistance between the gas and the walls. Selecting a mass flow sensor that is suitable to the application can be a challenge, and mistakes can be costly if critical factors are overlooked. In the paragraphs that follow I'll illustrate a number of factors to be mindful of, but please note, this is not all inclusive. Some notes will be relevant to a variety of flow sensor types; others will focus on thermal based mass flow sensors.
Mass flow sensors are often preferred to volumetric sensors , as volume is dependent on pressure and temperature, while the mass can be thought of as "how much" (not to imply that mass flow sensors are completely immune to these changes, particularly temperature). This all goes back to basic thermodynamics and the Ideal Gas Law (PV=nRT) - a prime example of a formula assuming idealized conditions; at least this one advertises itself as such. Need a quick refresher?
Pressure x Volume = Amount of Gas (moles) x Universal Gas Constant x Absolute Temperature
Moles = Mass / Molar Mass. Molar mass is a gas-specific constant.
Rearranging and combining constants, Volume = Mass x Temperature/Pressure x Constant
So, why are "mass flow " sensors then rated in LPM (Liters per Minute)? If you examine the datasheets closely, you should find that the LPM readings are at a stated set of "standard" or "normalized" pressure and temperature conditions. These standards will vary depending on which organization is referenced (NIST, IUPAC...). 1 atm and 0 °C are commonly used.
Gas Composition
When selecting a flow sensor, one of the first things to think about is the appropriate flow range. Make sure to check which gas type the sensor is calibrated for, as gas composition will change the output of many sensors . Air is the most common, but not always the case. When using a gas, other than what is listed on the datasheet, always contact the manufacturer. Text book conversions don't necessarily apply across the board for different sensor types. Gas density is certainly a factor, but in temperature-based mass flow sensors, the thermal conductivity of the gas is also quite important. Omron , for instance, will provide alternate output curves for some gases (air vs. natural gas vs. argon), but recommends a custom factory calibration to achieve optimal results for other gases (ex: carbon dioxide and nitrogen oxides). If the gas composition will be changing significantly during use, look for a sensor that is designed specifically for that condition, as most cannot distinguish the difference between a change in density versus a change in the amount of flow.
Turbulence and Pulsation in the Gas Stream
Turbulence and/or pulsation in the gas stream can result in unstable and inaccurate flow measurement . These issues can be addressed in the design of the flow path around the sensor; however, some sensors are designed to be more tolerant of these conditions.
Figure 1 Several of Omron's inline sensors for example are designed with a series of inlet screens (see Figure 1) which smooth out turbulent flow and provide a very even flow across the diameter of the sensor. This results in a representative velocity moving over the MEMS chip. When using a sensor without a laminar flow structure, incorporating sections of straight pipe upstream and downstream of the sensor is usually recommended. The required length of this section can be calculated by using the Reynold's number, however, some suggest a ballpark of 10 diameters upstream and 5 diameters downstream. Additionally, when screens are not used, some manufacturers recommend a filter upstream of the sensor for particulate laden environments to protect the sensing element.
Pulsating flows, such as those from diaphragm pumps, can be addressed by a few methods. Outside of the sensor, a buffer tank upstream of the measurement device or an orifice down stream are two options. Within the sensor design, an orifice installed at the output side the sensor (ex. Omron's D6F-01/02A2 model Omron's D6F-AB type to be released in early 2012). A similar solution is to use a smaller sensor in a bypass (see Figure 2). More details on this type of set-up are discussed below.
Figure 2 The pulsation phenomenon is one of those that is very difficult to model, and sometimes trial and error is the only way to determine which is the best approach for a particular system. In exceptionally difficult situations, a combination of the above suggestions are sometime required.
Sensor Characteristics
It is also important to make sure the characteristics of the sensor are compatible with the application. The response time of the sensor is critical in a number of devices, and this characteristic can vary significantly among available sensors. Some manufacturers will modify their timing constant upon request. Note, however, that a faster response time may result in more noise in the output signal.
Some applications have a limit to the Pressure Drop they can handle (ex. breathing circuits in medical devices). A sensor with minimal internal restrictions, or using a small sensor in a bypass configuration (see Figure 2) are two options.
Bypass Set-up
The bypass set-up involves more design work, however, a smaller and often less expensive sensor can then be used. Omron's factory engineers offer design assistance utilizing their Computational Fluid Dynamics modeling software for customers interested in this option.
Flow sensors in a "bypass configuration" are similar to using a differential pressure sensor (an indirect flow measurement method); however, there are some differences in the technologies that should not be overlooked. In both instances, the sensor is connected to two ports, which are installed on either side of an orifice, to create a pressure drop.
In differential pressure sensing , the difference in upstream and downstream pressures is measured on either side of a diaphragm or a set of diaphragms. It is a static system, meaning the air does not pass through the diaphragm(s). A mass flow sensor is a dynamic system, as the airflow continually passes through the sensor. The major implication of this difference is that the pressure drop of the "bypass system" will influence the output of some sensors, meaning that the tube lengths of the bypass need to be controlled in the manufacturing process. In many applications this is not an issue; however, it can be in those such as HVAC control where tubes are installed in the field, not the factory. This problem can be avoided by the use of High Impedance Flow Sensors which have a very small, high velocity flow path (again, watch for a new Omron product release this fall, series D6F-PH).
When selecting a flow sensor, be mindful to avoid the common pitfalls listed above. Remember the sensor is just one piece of a fluidic system, and the other components of that system can influence the outcome. If fluid dynamics is not your area of expertise, don't be afraid to ask for some assistance prior to making your sensor selection. The Mouser and Omron teams are readily available to offer you additional support!
Hungary
