PSI - Issue 32

G. Gusev et al. / Procedia Structural Integrity 32 (2021) 49–55

52 4

Gusev G.N./ StructuralIntegrity Procedia 00 (2019) 000 – 000

diagram of the sensor in the measuring circuit. R1 and R4 - measuring arms of the bridge, R2 and R3 - temperature compensation. Vg is the measurement diagonal. Points A, B, C, D are circuit nodes on the bridge board. It was necessary to take into account the experience of previous work in the next step of creating a strain gauge for aggressive conditions. First, the sensor mounting schemewas changed. The sensor was produced in a laboratory environment. Active strain gauges were glued to a stainless steel plate using BF-2 glue. This glue was polymerized at a given pressure in three stages at different temperatures. The polymerization temperature range was from 60 to 180 degrees Celsius. The total polymerization time was 24 hours. Then the entire sensor was covered with a layer of a special Vixint compound - a sealing compound based on silicon rubbers. The thermo compensation part of the sensor was made on a steel plate using the same technology. Next, the sensor was assembled in a bridge circuit using wires. It should be noted that this sensor design has passed all the necessary mechanical tests on a special measuring stand. The mechanical design of the sensor has been simulated with ANSYS and has been verified. The total error in measuring the deformation was 1%. The strain gauge is welded to the structural column using spot welding. The measuring circuit is closed with an aluminum cover, as in the first case, to protect the circuit from mechanical damage. Figure 3a shows a new version of the strain gauge that is being field tested. Figure 3b shows a mathematical model of the sensor that has been numerically tested.

Fig. 3. (a) a new strain gauge; (b) a mathematical model of the sensor. New strain gauges were installed on the structure. After four months, the deformation values began to increase by analogy with the first experiment. The result of the investigation showed that this solution has a number of disadvantages. The Vixint compound was corroded by potash brine. BF-2 glue was also corroded. Despite the fact that the first time (two months) the readings of the sensors were stable and met the required measurement accuracy, then the sensor was completely destroyed. Figure 4 shows a strain gauge that has been dissolved with a potassium salt brine.

Fig. 4. (a, b) a new strain gauge that has been dissolved with a potassium salt brine.