On the need to decompose fatigue strain signals associated to fatigue life assessment of the AISI 1045 carbon steel

doi:10.1016/j.matdes.2013.12.043

Materials & Design

Volume 57, May 2014, Pages 405-415

https://doi.org/10.1016/j.matdes.2013.12.043 Get rights and content

Highlights

•
Comparing the fatigue behaviour under three types of constant loading.
•
It is a need to assess the fatigue life using a new statistical-based analysis.
•
Higher stresses induced higher statistical (I-kaz) coefficient values.
•
It is suggested that the I-kaz technique is capable to assess fatigue failure.

Abstract

The evaluation of a new statistical-based analysis is discussed in this paper, leading to the fatigue life assessment at three different applied stresses during the cyclic testing procedure. Fatigue tests were performed following the standard ASTM: E466-07. These tests involve a strain gauge being attached to the specimen. For this test, AISI 1045 carbon steel was used as material due to its wide application in automotive and machinery industries. Fatigue tests were performed at three different stress levels of 305 MPa, 325 MPa, and 345 MPa with the testing frequency of 8 Hz, and the strain signals were collected accordingly. The Integrated Kurtosis-based Algorithm for Z-filter (I-kaz) approach, which provides its coefficient and three dimensional graphics, was utilised to define a statistically-based fatigue behaviour pattern. The I-kaz technique was used to extract strain signal patterns at the respective low, medium, and high frequency ranges for each signal at specific applied testing stress level. Results showed that highest strain amplitude occur at the low frequency range, suggesting the capability of I-kaz to identify fatigue damage pattern of metallic materials using statistical representation.

Introduction

Fatigue is a phenomenon of failing a component under cyclic loading prior to its ultimate stress. Theories of failure describe the condition for failure, such as application of linear damage accumulation of the Palmgren–Miner rule, which was based on the hypothesis on total damage [1], [2]. In any related application, a failure in specimens occurs when applied or external stresses exceed a possible value of maximum tensile stress, maximum compressive stress, or maximum shear stress. In a case of fatigue, failure will occur earlier for the maximum value of design stress. Fatigue can occur in a metallic material because of repeated application of stress and strain, especially those changes that lead to cracking or failure [3], [4].

For numerous automotive components, the primary mode of failure can be attributed to fatigue damage resulting from the application of variable amplitude loading. Predicting the life of the part stressed above the endurance limit is, at best, a rough procedure, especially for components such as automobile engine, steering, and suspension parts [2], [5]. For these cases, strain-based approach is commonly used to predict fatigue life [3], [6]. Strain-life fatigue model relates plastic deformation, which occurs at a localised region where fatigue cracks begin to the durability of the structure. This model is often used for ductile materials at relatively short fatigue lives. This approach can also be used where there is little plasticity at long fatigue lives. Therefore, this comprehensive approach can be used in place of stress-based approach [7]. Various approaches and techniques can be used to analyse fatigue data, such as root mean square (r.m.s.), and power spectral density (PSD) [8]. Those techniques and approaches were applied in previous works [9], [10] to evaluate fatigue damage, predict fatigue life, and assess fatigue behaviour.

Consequently, this research uses backward analysis, which can split fatigue data into various specified frequencies. It focusses on the investigation of a new statistical-based data scattering observation, namely Integrated Kurtosis-based Algorithm for Z-filter (I-kaz), in assessing fatigue damaging behavioural pattern of the AISI 1045 Carbon Steel using the strain decomposition approach into three different frequency ranges, i.e. Low frequency (LF), Medium frequency (MF), and High frequency (HF). Strain signals were collected via fatigue test at three different applied stresses at 305 MPa, 325 MPa, and 345 MPa. The I-kaz technique was used to decompose signals into designated frequency ranges as well as to provide a three-dimensional (3D) diagram for determining the statistical fatigue damage pattern. In addition, significant coefficients were calculated to observe fatigue life representation.

Section snippets

Fatigue life assessment for metallic materials

Fatigue damage is one type of failure mode for engineering structural components. In the previous century, several accidents were caused by fatigue cracks [11], such as aircraft and railway axles. These problems were investigated by August Wöhler. Several catastrophic failures of aircraft structures were caused by fatigue, which occurred in the 20th century [12]. Failures occur even when the maximum value of the cyclic load is much lower compared with the static strength of the material.

Materials and specimen design

AISI 1045 carbon steel is a medium carbon steel with great strength and hardness, and this kind of steel series is used extensively by all industry sectors for applications requiring more strength and wear resistance compared with typical applications of low-carbon mild steel. The numerous applications of this AISI 1045 steel in industries cannot be overemphasized. It is a medium tensile steel with the nominal tensile strength in the range of 570–850 MPa. In real applications, this steel can be

Results and discussion

Fig. 8 shows that three applied stresses at 305 MPa, 325 MPa, and 345 MPa, showing similar patterns to the constant amplitude loadings. In addition, Fig. 9 shows strain signals collected from an initial point until the end of test for all applied stresses. Fig. 9(a) shows almost a constant strain range because no obvious changes are seen in strain values from early until failure. Fig 9(b) shows that the strain values slightly increase from initial to breakage point. However, in Fig 9(c), strain

Conclusions

This paper discussed the evaluation of the I-kaz technique produced from the procedure of signal decomposition and were then compared to the fatigue damage criterion. The strain signals were decomposed into three frequency ranges. It was found that LF signal produced the highest fatigue damage compared to MF and HF signal. This LF-produced damage was compared to the damage from original strain. The highest damage values for original strain and LF strain were found to be at 2.803 × 10⁻⁵ and 2.485 ×

Acknowledgement

The authors would like to express their gratitude to Universiti Kebangsaan Malaysia and Ministry of Higher Education, Malaysia [Funding Grant No.: UKM-KK-03-FRGS0118-2010].