PSI - Issue 65

M.A. Skotnikova et al. / Procedia Structural Integrity 65 (2024) 248–254 4 M.A. Skotnikova, A.Y. Ryabikin, A.D. Shestakov, L.D. Tuptei, A.D. Novokshenov / Structural Integrity Procedia 00 (2024) 000–000

251

50 minutes were constructed using machine learning in Python (Fig. 3, a, c, e), in the area of the steady-state friction and wear process.

Table 2. Tribotechnical properties of the studied steels

The average moment of friction, Nm

Coefficient of friction

Steel grade

Hardness, MPa

Wear, mg

Wear resistance, 

09G2S

1740 4360 4340

6.84 0.38 0.87

1.0

0.47 0.75 0.84

0.44 0.58 0.69

Quard 450 Hardox 450

18.1

7.9

As can be seen from the graphs (Fig. 3, c, e), wear-resistant steels of the martensitic class Quard 450 and Hardox 450 had a more uniform friction moment in amplitude compared to ferrite-pearlite steel 09G2S (Fig. 3, a), the amplitude of the moment values of which varied significantly over time.

а

b

c

d

e

f

Fig. 3. Change in the moment of friction during testing of steels: before smoothing a) 09G2S; c) Quard 450; e) Hardox 450 and after smoothing b) 09G2S; d) Quard 450; f) Hardox 450

At the same time, during the friction of the Quard 450 steel (Fig. 3, c), two stable amplitude modes of the friction moment change were observed throughout the time interval. In order to highlight these changes more clearly, the author's program averaged (smoothed) the test results to eliminate unnecessary (random) vibrations caused by additional external vibrations (Fig. 3, b, d, f).