Corrosion Fatigue Cracking in Paper Machine Felt Guide Roll Shafts

Abstract

This work aims to determine the root cause of unexpected failures of felt guide roll shafts of a paper machine. This was executed by performing finite element analysis (FEA) and metallurgical failure analysis. For the FEA, computer simulations were performed to assess load-bearing behavior of the shafts during operation. For the failure analysis, various techniques were employed including visual inspection, chemical composition analysis by emission spectroscopy, fracture surface analysis, and microstructural analysis by optical microscopy (OM) and scanning electron microscopy (SEM). The FEA results show that the effective stress values in the shaft cracking regions are far less than the fatigue limit but cracking still occurred. This suggests that the material’s mechanical properties deteriorated due to corrosion. From the failure analysis results, it was found that corrosion fatigue was likely to be the root cause of the shaft failure incidents.

Appendix A Estimation of the Fatigue Limit

This section shows the details of the fatigue limit estimation for the original and newly-designed shafts. First, the fatigue limit of the shaft material (S'_e), i.e., the AISI 4340 steel, can be estimated by the relation (Shigley et al., 2003; Topaҫ et al., 2012; Young & Budynas, 2001)

$$S_{e}^{^{\prime}} = 0.504\sigma_{ult}$$

(A1)

where $\sigma_{ult}$ is the ultimate tensile strength. As $\sigma_{ult}$ of the AISI4340 steel is approximately 1,110 MPa (http:asm.matweb.com, search, SpecificMaterial.asp?bassnum=m434ae), S'_e${S}_{e}^{^{\prime}}$ was estimated to be 559.44 MPa by Eq. A1.

Next, the Marin factors (Marin, 1962) were employed to estimate the fatigue limit of the critical location of a shaft in the real operating conditions (${S}_{e}$) from the fatigue limit obtained from lab specimens under controlled conditions (${S}_{e}^{^{\prime}}$). (${S}_{e}^{^{\prime}}$) and (${S}_{e}^{^{\prime}}$) are related by (Shigley et al., (2003):

$$S_{e} = k_{a} k_{b} k_{c} k_{d} k_{e} k_{f} S^{\prime}_{e}$$

(A2)

where k_a is the surface conditions modification factor, k_b is the size modification factor, k_c is the load modification factor, k_d is the temperature modification factor, k_e is the reliability factor, and k$_{\text{f}}$ is the miscellaneous-effects modification factor. The estimation of k_a, k_b, k_c, k_d, k_e, and k_f for the felt guide roll shafts in this work is demonstrated below.

The surface factor, k_a: k_a can be determined by:

$$k_{a} = a\sigma_{ult}^{b}$$

(A3)

The parameters ${\text{a}}$ and ${\text{b}}$ for the ground surface finish are 1.58 and − 0.085, respectively (Noll & Lipson, 1976). For the shafts in this work, the surfaces were ground and $\sigma_{ult}$ is approximately 1,110 MPa. Thus, k_a is approximately 0.8706.

The size factor, k_b: k_b is equal to 1 for components subjected to axial loading. For components subjected to bending and/or torsion loading, k_b is calculated from Eq. A4.1 or A4.2 (Mischke, 1987).

$$k_{b} = 1.24d^{ - 0.107} \quad {\text{for}}\;{2}{\text{.79 < }}d \le {\text{51mm}}$$

(A4.1)

$$k_{b} = 1.51d^{ - 0.157} \quad {\text{for}}\;51{ < }d \le 254{\text{mm}}$$

(A4.2)

During the shaft operation, the critical locations were subjected to combined bending and torsion loading (Ugural, 2022). The effective dimensions (d) of the original and newly-designed shafts are both in the ranges 50–80 mm and 60–95 mm, respectively. Therefore, Eq. A4.2 was employed to estimate k_b. Thus, k_b of the original and newly-designed shafts are in the ranges 0.7589–0.8159 and 0.7387–0.7940, respectively.

The loading factor, k_c: During the shaft operation, the critical locations were subjected to combined bending and torsion loading. Therefore, k_c is 1 (Shigley et al., 2003).

The temperature factor, k_d: As the shafts were operated at the temperature of approximately 50 °C, k_d is equal to 1.01 (Shigley et al., 2003).

The reliability factor, k_e: The reliability of 90–95% was used. Thus, k_e is in the range 0.868–0.897 (Shigley et al., 2003).

The miscellaneous-effects modification factor, k_f: The miscellaneous effects were assumed insignificant. Therefore, $\text{k}_{\text{f}}$ is 1 (Shigley et al., 2003).

From the estimation of k_a,k_b, k_c, k_d ${k}_{e}$, and k_f above, the k_ak_bk_ck_d ${k}_{e}$ k_f values of the original and newly-designed shafts are in the ranges 0.5792–0.6435 and 0.5638–0.6262, respectively. Therefore, S_e of the original and newly-designed shafts are in the ranges 324.03–360.00 and 315.41–350.33 MPa, respectively.