Several Analysis on Factors Affecting Thermocouple Measurement Accuracy

Thermocouples are widely used in industrial temperature measurement due to their simple structure, easy manufacturing, and convenient measurement. However, to obtain correct measurement results, it is necessary to understand the characteristics of thermocouples and use them correctly, otherwise there will be great errors in the measurement. The following describes the factors that affect the instability of thermocouple measurement from several aspects.

1. The influence of unstable thermoelectric characteristics of thermocouples

1.1 Effects of contamination and stress

(1) During the production and processing of thermocouples, the surface of the thermocouple wire is always contaminated after multiple shrinking and stretching. At the same time, from the internal structure of the thermocouple wire, there are inevitably stress and lattice inhomogeneity. These inconsistencies in the physical state affect the Seebeck coefficient of the thermocouple wire and the instability of the thermocouple indication. Therefore, a thermocouple that fails to meet annealing standards cannot be used for accurate temperature measurement. The stress introduced by quenching or cold working can be basically eliminated by annealing. The indication error caused by unqualified annealing can vary from a few tenths of a degree to several degrees. It is related to the temperature to be measured and the temperature gradient on the thermocouple electrode. Cheap metal thermocouple wires are usually delivered in an "annealed" state. If a high-temperature cheap metal thermocouple needs to be annealed, the annealing temperature should be higher than the upper limit of the use temperature, and the insertion depth should also be greater than the actual use depth.

(2) For thermocouples used as reference and standard, it is best to anneal them after each graduation point. For precious metal thermocouples, the annealing procedure is very important. The calibration procedure for standard thermocouples stipulates that standard thermocouples must be cleaned and annealed before calibration to eliminate contamination and stress of the thermocouple, thereby improving the metallographic structure of the thermocouple wire material and improving measurement stability.

1.2 Effect of inhomogeneity

(1) Thermocouple theory points out that the thermoelectric potential of a thermocouple made of a homogeneous conductor is related to the temperature at both ends, but has nothing to do with the temperature distribution along the length of the thermocouple. If the thermocouple material is not uniform and the two thermocouples are in a temperature gradient field, the thermocouple will generate an additional thermoelectric potential, which is called "non-uniform potential". The magnitude of the non-uniform potential depends on the temperature gradient distribution along the length of the thermocouple, the non-uniform form and degree of the material, and the position of the thermocouple in the temperature field.

(2) The reasons for the unevenness of thermocouples are mainly chemical composition and physical state. In terms of chemical composition, there are uneven distribution of impurities, segregation of components, local metal volatilization on the surface of thermocouples, oxidation or selective oxidation of certain metal elements, measurement of thermal diffusion at high temperature, and contamination and corrosion of thermocouples in harmful atmospheres. In terms of physical state, there are uneven stress distribution and uneven electrode structure. In the thermocouples used in industry, the additional error caused by the uneven potential is sometimes as high as 20°C. This will seriously affect the stability and interchangeability of thermocouples, so uniformity is one of the important indicators to measure the quality of thermocouples.

2 Influence of thermocouple’s own instability

Instability is the change in the thermocouple's scale value over time and under different conditions of use. In most cases, it may be the main cause of inaccuracy. If the change in the scale value is relatively slow and uniform, it is necessary to conduct regular supervisory calibration (such as standard thermocouples) or arrange periodic calibration according to actual use, which can reduce the error introduced by instability.

2.1 Factors affecting instability

(1) Contamination. It has been discussed above that the thermocouple wire will affect the Seebeck coefficient. The thermocouple wire material is often contaminated by the ambient atmosphere or impurities in the protective tube. Different degrees of contamination will generate different additional potentials. This additional potential will change the original graduation characteristics, which is a factor causing the instability of the thermocouple indication. For example, for a platinum-rhodium 10-platinum thermocouple, when the ceramic tube used contains iron impurities, the platinum-rhodium wire will be contaminated by iron, which will affect its thermoelectric characteristics; when used in a high-temperature reducing atmosphere containing silicon, the silicon is reduced to free silicon and combines with the platinum-rhodium wire to form a platinum-silicon compound, making the thermocouple wire brittle. The insulating porcelain tubes used to calibrate the standard thermocouple are required to be cleaned with aqua regia, baked at high temperature, and the positive and negative electrode perforation polarities are specified. If the positive and negative electrodes of the thermocouple are mistakenly inserted in the commonly used tube, the platinum in the original platinum-rhodium hole will penetrate into the platinum electrode and change the thermoelectric characteristics of the standard thermocouple. All of the above situations will affect the stability of the thermocouple.

(2) Thermocouples evaporate at high temperatures. Most thermocouple wire materials are alloy materials. Since the vapor pressures of the various component materials are different, the degree of volatilization is also different. After being used at high temperatures for a certain period of time, the proportion of the alloy components will change, which will lead to significant changes in the thermoelectric potential.

(3) Oxidation-reduction. The instability of many thermocouples is caused by oxidation of the thermocouple wire. Thermocouples such as copper-constantan, iron-constantan, nickel-chromium-nickel silicon, etc. can all react with oxidation. If the thermocouple is uniformly oxidized, the impact may be smaller; if it is preferentially oxidized, the impact is very serious. In low oxygen partial pressure (i.e., in the absence of oxygen), the chromium in the nickel-chromium electrode will undergo preferential oxidation and change the composition of the thermocouple wire.

(4) Embrittlement. Embrittlement is the most common factor in the failure of thermocouples. Contamination, grain growth, oxidation (such as nickel-chromium-nickel-aluminum) and long-term use at high temperatures and recrystallization (such as tungsten) can all lead to the embrittlement of the thermocouple wire.

(5) Radiation. When a thermocouple is working in an atomic reactor and is bombarded by neutrons, if one or more elements in the thermocouple wire material undergo a transformation (for example, rhodium transforms into palladium), the composition of the thermocouple wire will change, causing a significant change in the thermoelectric properties.

2.2 Thermocouple stability test

(1) To check the stability of the thermoelectric potential of a newly made standard thermocouple, the thermocouple is placed in an annealing furnace and annealed twice for 1h to 2h in a section 400mm from the measuring end at a certain required temperature. The first annealing can eliminate the internal stress of the thermocouple. After taking it out, measure its thermoelectric potential at the specified temperature. Then put it in the annealing furnace again and anneal it again according to the above method. After taking it out, measure its electromotive force again. The difference between the two thermoelectric potentials before and after the second annealing (at the same specified temperature point) is used as the evaluation standard. If it does not exceed the measurement point requirement, it is qualified.

(2) The stability of the thermoelectric potential of the standard thermocouple in use is determined by comparing the thermoelectric potential measured during calibration with the result of the previous calibration. If it does not exceed the allowable variation error value, it is qualified.

3 Effect of reference terminal temperature

The magnitude of the thermoelectric potential of a thermocouple is related to the temperature of the thermocouple material and the working end. The thermocouple graduation table and the temperature display instrument based on the graduation table are based on the condition that the temperature of the thermocouple reference end is equal to 0℃. Therefore, this condition must be followed when using it. If the reference end temperature tn is not equal to 0℃, although the measured temperature t is constant, the thermoelectric potential EAB(t,tn) will also change with the change of the reference end temperature tn. The magnitude of its change can be obtained according to the intermediate temperature law of the thermocouple. EAB(t,tn) =EAB(t,tn)+EAB(tn,to)When to>0℃, the thermoelectric potential of the thermocouple is reduced by EAB(tn,to), which will reduce the indication of the measuring instrument. Therefore, when the reference end temperature is not equal to 0℃, it has a very important impact on the accuracy of the measured temperature. When measuring temperature with a thermocouple, it is troublesome to keep the reference end temperature at 0℃, which is generally only necessary when making precise temperature measurements in the laboratory. Usually in engineering measurement, the reference end temperature is mostly at room temperature or in a fluctuating temperature zone. In this case, in order to measure the actual temperature, correction or compensation measures must be taken. There are several correction methods: thermoelectric potential correction method, temperature correction method, instrument starting point adjustment method, compensation wire method, reference end temperature compensator. All of them can achieve the correction of measurement results.

4 Heat transfer effect

When the thermocouple is inserted into the measured medium (such as airflow), it absorbs heat from the measured medium to increase its own temperature, and at the same time it dissipates heat to places with lower temperatures by means of thermal radiation and heat conduction. Due to the influence of the thermocouple's heat transfer, the thermocouple cannot reach the temperature that it should reach when absorbing heat. After a period of time, when the heat dissipated from the measuring end is equal to the heat absorbed from the airflow, a dynamic balance is reached, and the thermocouple reaches a stable indication. However, this indication does not represent the true temperature of the airflow, because the heat dissipated from the measuring end environment is compensated by the heating of the airflow, that is, the stronger the heat transfer between the measuring end and the environment, the greater the deviation of the temperature of the measuring end from the airflow temperature. In this case, if the indicated temperature Tcoupling of the thermocouple is still used to indicate the airflow temperature Tgas, it is bound to introduce (Tgas-Tcoupling) error, that is, heat transfer error.

5 Dynamic response error

After the thermocouple is inserted into the measured medium, it cannot immediately indicate the temperature of the measured airflow due to its own thermal inertia. Its indication Ta will gradually rise until the measuring end absorbs heat and releases heat to reach dynamic equilibrium before reaching a stable indication Ta∞. During the entire unstable process from the insertion of the thermocouple to the stabilization of the indication, there is a deviation between the instantaneous indication Tmeasured of the thermocouple and the indication T∞ after stabilization. At this time, in addition to various stable errors, the thermocouple also has a deviation introduced by the thermal inertia of the thermocouple (i.e., Tmeasured-T∞). This deviation is called the dynamic response error and is represented by △Tdynamic.

△ T動=T测-Ta∞=τ dT测/dt Where: τ is the time constant of the thermocouple. For the sake of simplicity, the following discussion assumes that there is no influence of other error factors, that is, T∞=T气, so τ dT测/dt+T测=T气. If the measured temperature is a steady-state measurement that does not change with time, as long as a sufficient amount of time passes after the thermocouple is inserted and the reading is taken after its indication stabilizes, the influence of dynamic errors can be avoided.

6. Impact of leakage current on the measurement system

Poor insulation is the main cause of current leakage, which has a great impact on the accuracy of thermocouple temperature measurement, can distort the measured thermoelectric potential, distort the instrument display, and even fail to work properly. Leakage causes errors in many ways. For example, the insulation resistance of the thermoelectrode insulation porcelain tube is poor, which can bypass the thermal current and reduce the terminal voltage of the thermocouple, but this effect is relatively small. If the electrical measuring equipment leaks, it can also bypass the working current and cause measurement errors.

From the above discussion, we can see that several unstable factors are issues that must be recognized when measuring temperature with thermocouples. With the development of industrial technology and the widespread application of thermocouples in the production field, the requirements for the use of thermocouples are becoming more and more precise. Therefore, only by understanding and mastering the characteristics and applications of thermocouples can we accurately achieve the various measurement purposes we need.