International Journal of Fatigue 116 (2018) 48–67
Contents lists available at ScienceDirect
International Journal of Fatigue journal homepage: www.elsevier.com/locate/ijfatigue
Fatigue behavior prediction and analysis of shot peened mild carbon steels a,⁎
b
Erfan Maleki , Okan Unal , Kazem Reza Kashyzadeh a b
T
a
Mechanical Engineering Department, Sharif University of Technology-International Campus, Kish Island, Iran Mechanical Engineering Department, Karabuk University, Karabuk, Turkey
A R T I C LE I N FO
A B S T R A C T
Keywords: Shot peening Fatigue life Modeling Artificial neural network
One of the well-known methods of severe plastic deformation (SPD), shot peening is widely used to improve the mechanical properties and fatigue behavior of metallic materials. The present study investigated experimentally the effects of different shot peening treatments of conventional and severe on the fatigue behavior of various carbon steels. A novel alternative approach has been proposed that utilizes the concept of artificial neural network to predict the fatigue life of carbon steels which are subjected to different shot peening treatments. The experimental results were employed to develop the network. After the results were modeled, it was indicated that the used neural network-based approach is greatly consistent with the experimental ones. Next, a comprehensive parametric analysis was performed while considering the influence of the effective parameters of shot peening on fatigue behavior.
1. Introduction Most failures such as fatigue fracture in engineering components are very sensitive to surface properties, so that in most cases the cracks initiate from the surface and propagate to the interior. It is well-established that the fatigue strength of mechanical components can be enhanced by grain refinement and improving the mechanical properties of the surface layer and producing compressive residual stress (CRS) [1,2]. One of the widely used approaches in this regard is employing the shot peening (SP) process that makes the nucleation and propagation of fatigue cracks, especially in metallic materials, more difficult [3]. SP is a cold working process in which the surface of a component is bombarded with small shots under a controlled velocity that may deform the surface layer plastically and create considerable CRS [4]. Almen intensity and coverage are the most important parameters in controlling the SP process. In recent years, researchers have suggested that by increasing these two parameters to unconventionally high values, and therefore, enhancing SP severity and changing it from conventional shot peening (CSP) to severe shot peening (SSP), more beneficial effects can be obtained [5–10]. After SSP is applied, the grains in the surface layer turn into submicron (< 1000 nm) scale. The grains with a size of 1–100 nm and 100–500 nm are called nanostructured (NS) and ultrafinegrained (UFG), respectively [11]. NS and UFG materials contain in their microstructure a very high density of grain boundaries, which can play a significant role in the development of superior properties. The fatigue behaviors of the conventionally and severely shot peened components have been explored on different materials such as ⁎
Corresponding author. E-mail address:
[email protected] (E. Maleki).
https://doi.org/10.1016/j.ijfatigue.2018.06.004 Received 9 March 2018; Received in revised form 1 June 2018; Accepted 5 June 2018 Available online 06 June 2018 0142-1123/ © 2018 Elsevier Ltd. All rights reserved.
steel and aluminum [12–16]. On the other hand, modeling complex and ill-defined processes by using artificial neural networks (ANNs) in different scientific and engineering areas such as materials science [17–20], manufacturing process [21], bio-engineering [22], and energy systems [23] is frequently preferred to costly and time-consuming experiments. Numerous studies have been conducted on using ANN for modeling materials’ fatigue life. Different ANN simulations have been proposed for fatigue life prediction of different materials such as composites [24,25], steels [26–28], and aluminums [29,30]. Moreover, some studies have employed ANN to model other aspects of materials’ behavior such as fretting [31] and corrosion [32], and fatigue life prediction of coated materials [33]. However, few studies have addressed modeling SP process via ANNs. In this regard, Karatas et al. [34] employed ANN to predict residual stresses in the shot peened material C-1020. Delijaicov et al. [35] developed a model for peen forming process planning of Aluminium 7050 alloy to predict Almen displacement. More recently, from 2016, Maleki et al. [36–40] presented an ANN modeling of the effect of various SP processes on different materials including 18CrNiMo7-6 steel, AISI 1017 steel, cast iron, (TiB + TiC)/Ti–6Al–4 V composite, and AISI 1060. In this paper, metallurgical properties and fatigue behavior of two different carbon steels of AISI 1045 and 1050 are subjected to CSP and SSP and compared experimentally. Different SP treatments, with different Almen intensities and coverages were applied. Microstructural characterizations were carried out using optical microscopy (OP) and field emission scanning electron microscopy (FESEM) observations.
