In Europe, the ISO6143 standard specifies how to determine and control the composition of the gas mixture used for analyzers calibration, and the norm ISO6145-1 defines the ways of preparation mixtures using dynamic volumetric methods. The table shows 8 different mixture preparation techniques.
|ISO 6145-2||Volumetrical pumps|
|ISO 6145-4||Continuous injection|
|ISO 6145-6||Critical nozzles|
|ISO 6145-7||Thermal Flow Controllers|
We look in greater detail below, only three of the possible techniques:
Sonic nozzles or capillaries: the nozzles generally operate under swirling, while the capillaries operate under laminar. Generally, the orifices are used in different section to get on each orifice flow values increasing in accordance with the 2^n progression, while the capillaries are generally selected to be "equal".
Flow Controllers: is often added the term "mass", but in fact are based on a measure of flow of type "thermal" (transfer of heat from a hot element to one or two sensors placed in the measuring cell) and on the control of the flow through a valve and a regulating circuit.
Other techniques (eg .: volumetric meter bubble or piston) are characterized by excellent accuracy values, but do not lend themselves to the production of diluters dynamic. They can also be used for quality checks on diluters
In spite of its "simplicity" both give excellent results in terms of precision, long term stability and reliability. The various dilutions are possible in discrete steps, corresponding to the different connections (to the gas to be diluted or to the diluting gas) from the different orifices or capillaries.
Calibrated nozzles (with increasing section) allow the selection of a number of "steps" of 2^n, where n is the number of installed nozzles. For example, seven orifices may provide 256 different dilution ratios between 0 and 100% of the gas to be diluted. For the metrological verification of the orifices in increasing section, measured flows change in a wide range and the testing bench is required for moreover excellent qualities in terms of linearity. Furthermore, only one flow meter can not obtain the required resolution over a wide band of values (orifices in different section). Using various flow meters the total uncertainty ido increase.
Capillaries (constant cross section) allow the selection of a number of dilution "steps" equal to the number of the installed capillaries. For example (BetaCAP30) with 30 capillaries 30 different dilution factors (+zero) may be realized. The "strong" advantages of this technique are simplicity and accuracy obtainable in the selection process of the capillaries with an "essential" test . All necessary measures correspond to flow values almost equal: the test conditions (eg .: the pressure applied) are adapted for the optimization of the resolution. All measurements are compared with the measurement repeated on a "reference" capillary of the lot: measuring values are important just to compare the capillaries one each other. The absolute flow value is not important and the traceability of the measuring system is not required. The sole required feature is the stability (applied pressure, temperature and flow rate measurement) in the time required to qualify a batch of capillaries to be installed in a single diluter. [More info]
They consist of a flow meter and a system of closed loop regulation: the contribution of uncertainty are linked to the uncertainty of the measurement and the dead band of the controller. The main advantage is constituted by the "continuity" of dilution factors that can be realized : the dilution factor can assume all the values included in a defined band.
Different "dilution ranges" may be obtained by coupling flow regulators characterized by different ranges. Are commercially available flow regulators with fields ranging from a few microliters / min to tens of liters / min. All manufacturers suggest a limited use between 100% and 10% of the range.
Iin fact, a significant part of the error is relative to the full scale of the regulator and it is evident that in relative terms the error is multiplied by 10 when operating at 10% of the working range.
Even the linearity of response of the flow sensors plays an important role : Some manufacturer claim the complete linearization by using a corrective function, but it is based on the the measurements along all the range and the low resolution in the lower side of the working range however also restricts the effectiveness of the polynomial linearization .
In portable applications, it must also be considered a lower reliability (mainly due to shocks sensitivity), the need for a long warming up time at power-up and ultimately a lower stability in the medium term (months). in case of regular use the required interval between two certifications may not exceed one year.
Finally it should be considered that the K factors (multipliers of the measured flux, which depend on the composition of the mixtures concerned) are influenced by many phisical parameters and also by the MFC model : simplified calculations based just on thermal conductivity give rough results.
I happened to hear a comment like ..Capillaries ? But I was looking for a solution a bit 'more "technological" !
Apart from the fact that Hagen and Poiseuille not belong to prehistoric times, what about the formula 1 cars wheels ? The circular shape has been applied since the Neolithic era, more than 7,000 years ago ': do we want to change it ?
The capillaries technique, is applied today with new solutions (which by the way see us as the main innovator) allowing the realization of functions that were unthinkable a few years ago '.
And it is the simplicity of the "rules" that govern the fluid flow in the capillaries that has enabled us to introduce our innovations in products and production processes, reaching the highest standards of uncertainty.