PSI - Issue 78

Virginio Quaglini et al. / Procedia Structural Integrity 78 (2026) 105–112

109

Sliding Area t 9 t 10

Loading Direction

Loading Direction

t 1 t 2 t 5

t 4 t 6

t 3

t 8

t 7

Fig. 6. Sketch and views of the full-scale DCSS that show the position of the thermocouples.

3. Results Fig. 7 shows the experimental force-displacement loops obtained from the tests carried out at different loading conditions and at the temperatures of 20, 0 and -20 °C when the exposure duration to low temperatures is 24 hours. In the test at 20 °C friction shows the typical behavior of solid-solid contact, with a peak value at the initiation of motion (static or breakaway friction), followed by a decrease during continuous sliding (dynamic friction). On the contrary, in the tests performed after exposure to low temperatures, friction shows an irregular and unexpected behavior, consisting of a sequence of small increases and drops until a peak value is achieved; friction then drops to a constant value which is maintained during sliding. It is worth noting that friction achieves its peak value not at the starting of motion like in the test at 20 °C, but after a certain amount of sliding. Anyway, the static friction coefficient is recorded as the maximum value, regardless of whether it is activated at the breakaway or in a displaced configuration. The static friction coefficient calculated for each loading case of Table 1 is presented in Fig. 8 in a comparative way to highlight the effects of temperature, exposure duration, axial pressure and loading velocity. The static value after exposure to -20 ° C, μ St,-20 °C , is always smaller than the reference value at 20 °C, μ St,20 °C regardless of the duration of exposure, the axial pressure and the velocity. On the other hand, the static value after exposure to 0 ° C, μ St,0 °C , can be either greater or smaller than the reference value depending on the exposure duration and the axial pressure. For instance, after an exposure of 24 hours, μ St,0 °C is always smaller than μ St,20 °C , but after exposures of 3 and 12 hours, μ St,0 °C is greater than μ St,20 °C for axial pressures of 40 and 60 MPa, and smaller for pressure of 80 MPa. Regarding the dynamic friction coefficient, it is apparent from Fig. 7 that the forces of the DCSS at the first cycle recorded after exposure at -20 °C are significantly lower than the forces recorded in the test at ambient temperature, while in the second cycle and more in the third cycle, the forces recorded after exposure to low and very low temperature get closer to the reference ones. Such a behaviour is alleged to the change in the friction coefficient at the sliding interface during the cyclic motion. Fig. 9 shows the trend of the average dynamic friction coefficient, µ dyn , calculated over the three cycles of motion in each test for all the loading cases of Table 1. Regardless of the normal stress and loading velocity, the minimum value of µ dyn is always obtained after exposure to very low temperature (- 20 °C). On the other hand, for exposures at 0 °C, µ dyn can be either lower or higher compared to the value obtained in the reference test conducted at room temperature depending on the loading conditions (stress and velocity). As an example, µ dyn assessed in tests after exposure at 0 °C for 24 hours is consistently lower than µ dyn from the reference test; while, for exposures of 3 or 12 hours at 0 °C, µ dyn is greater than the reference value at 40 and 60 MPa, and lower at 80 MPa. The result seems counterintuitive, as one expects that increasing the normal stress would correspondingly increase the mechanical work, leading to higher dissipated energy and temperature rise. However, the increase in normal stress also triggers a decrease in the friction coefficient; consequently, there is an overall decrease in energy dissipation. In addition to these considerations, it should also be noted that after exposure at -20 °C, ice formation