PSI - Issue 13

A.M. Polyanskiy et al. / Procedia Structural Integrity 13 (2018) 1408–1413 Author name / Structural Integrity Procedia 00 (2018) 000–000

1409

2

quality control of all kinds of metals and alloys. The influence of hydrogen on the mechanical properties of metals is studied in detail. Dozens of works devoted to this problem are published every year. An important feature of the industrial research is that the critical levels of hydrogen concentration or content are very low, only 4 million parts of the hydrogen mass fraction lead to the flocs formation. Therefore, vacuum methods of charging the sample Jordan and Eckman (1926); Brown (1945); Scafe (1945) were widely used, since all gases evolved from a heated in vacuum metal sample were pumped into a calibrated volume, then pressure was measured in this volume, and, finally, a component analysis of the gas mixture with hydrogen was perfomed. The main lack of vacuum methods was the long duration of the measurements. Manufacturers of industrial measur ing equipment proposed to use a fast melting method in a inert carrier gas flow for measuring hydrogen concentration. To measure the hydrogen concentration in the carrier gas, it was proposed to use the thermal conductivity cell Stewart and Squires (1955); Hulsberg (1954); Willenborg (1941) developed earlier for chromatographs Olof (1927); Willenborg (1936). The operationg principle of this detector is based on the measurement of the di ff erential di ff erence in the thermal conductivity of a pure inert carrier gas and a carrier gas mixed with hydrogen released from a metallic sample upon its heating. Reducing the duration of hydrogen analysis from one hour to ten minutes was crucial for the industry. Cheaper and faster “atmospheric” analyzers have displaced vacuum ones from factory laboratories. Nowadays about a dozen di ff erent “atmospheric” analyzers of hydrogen and the only one vacuum analyzer are produced serially. The lack of an alternative in the industry gradually led to the fact that metrological certifications, calibrations and scientific research started to be carried out mainly by applying “atmospheric” analyzers. It turns out that regular interlaboratory comparisons realized by certified laboratories are not carried out with regard to hydrogen. The measuring data of hydrogen concentration obtained by di ff erent laboratories diverge several times Nolan and Pitrun (2004); Konopelk´ o et al. (2018), especially for standard samples with a hydrogen concentration less than 1 ppm Hassel et al. (2013). A wide range of measurement results is also observed for aluminum alloys Andronov et al. (2017). Apparently, there is an unaccounted reason for the occurrence of significant errors in the measurement of small concentrations of hydrogen by using modern analyzers. One of the possible reasons is the hydrogen detector. In this regard in this paper we analyse the detector operation. At the present time, two basic principles of detection are widely used in “atmospheric” hydrogen analyzers. The older one is the classical thermal conductivity cell. Some analyzers use hydrogen oxidation on a cell obtained from cuprous oxide to water, followed by detection of water in a carrier gas by an infrared sensor. Infrared method has a large number of stages, problematic in terms of the occurrence of systematic measurement errors. Thus, standard argon or nitrogen of special purity, which is used as a carrier gas, contains about 10 ppm of water, hydrogen, hydrocarbons that can not be removed by industrial methods. In this regard, it is di ffi cult to expect a 100% removal of these gases inside the analyzer before hydrogen treatment on copper oxide, and the level of “useful” hydrogen concentration has the same order of 5-20 ppm. An additional experimental analysis of the most important stages of treatment of the carrier gas is required to take into account possible errors. Let us consider the operation of a simpler classical detector, i.e. the thermal conductivity cell. Its scheme is shown in Fig. 1. The thermal conductivity cell is a measuring bridge, exposed by constant reference voltage U . The current, the value of which depends on the precision resistors 1 and 2, heats the platinum sensitive resistive elements 3 and 4. Cooling of platinum elements is carried out by blowing by a clean carrier gas and a carrier gas with hydrogen released from the metal sample when it is heated. The crucible and a sample can be heated with either without contact by means of high-frequency current or by electric current by applying a special contact electrode. An approximate scheme of the extraction chamber is shown in Fig. 2. After a metal sample has been blown in the extraction chamber, the carrier gas is fed to a purification from various substances that are released along with hydrogen and can contaminate the sensitive platinum unit of the thermal conductivity cell. Heat exchange of the gas with the powdered sorbent takes place during the purification process. Glass tubes with a sorbent are placed on the front panel of the hydrogen analyzers. It is believed that the temperature 2. Analysis of the hydrogen detector operation