PSI - Issue 17

Eduardo A. Lima et al. / Procedia Structural Integrity 17 (2019) 246–253 Eduardo A. Lima/ Structural Integrity Procedia 00 (2019) 000 – 000

252

7

Fig. 7. Von Mises stress (Pa) in the end of the first rolling pass: a) wheel-rail model without TTP; b) wheel model without TTP; c) wheel-rail model with TTP and d) wheel model with TTP.

Fig. 8. Von Mises stress (MPa) as a function of equivalent plastic strain in same element in the wheel model with TTP and without TTP.

For comparison purposes, the time to run each numerical simulation for each approach was recorded, considering a specific work station, with 24 cores and 48 GB RAM. The results are shown in Table (2). As expected, the models without TTP consume more time than the models with TTP. Moreover, the time for one rolling pass in the wheel-rail model is reduced by up to 87% if compared with the wheel model.

Table 2. Comparison of computational cost in the numerical simulations with different approaches. Rolling Models Time (h) Wheel-rail without TTP (one pass) 36.5 Wheel-rail with TTP (one pass) 31.0 Wheel without TTP (eight passes) 33.0 Wheel with TTP (eight passes) 30.6

3.3. Fatigue Model The Dang Van stress (τ DVmax ), the number of cycles until crack nucleation ( N ), and corresponding travelled distance ( d ) were calculated for all elements in the selected cross-section, Fig. (5). From that set of results, maximum values were extracted and presented in Table 3. From the table, it is possible to observe that the residual stresses generated