2
IRUS TotalDownloads
Altmetric
A deterministic approach for prediction of surface initiated rolling contact fatigue crack propagation using experimental and numerical methods
File | Description | Size | Format | |
---|---|---|---|---|
Kunzelmann-B-2023-PhD-Thesis.pdf | Thesis | 21.23 MB | Adobe PDF | View/Open |
Title: | A deterministic approach for prediction of surface initiated rolling contact fatigue crack propagation using experimental and numerical methods |
Authors: | Kunzelmann, Bjoern |
Item Type: | Thesis or dissertation |
Abstract: | To achieve significant efficiency improvements for rotating machinery, measures like downsizing and friction reduction are widely used. As downsizing increases the power density and friction reduction frequently relies on the use of lower viscosity oil, machine elements like gears and bearings are exposed to higher loads and lower lubricant film thicknesses. Higher contact stresses and lower film thicknesses lead to more severe metal to metal contact of the functional surfaces, consequently surface initiated damage and failures are more likely to occur in operation. In particular a surface damage effect called rolling contact fatigue (RCF), which is more commonly termed “Pitting” often determines the lifetime of the machine element. Until today in engineering practice mainly empirical models are employed to predict a probabilistic lifetime until pitting. These models are widely used throughout the industry but rely however on a number of empirical constants derived from experimental data. As empirical models do not include the physical mechanisms of crack propagation a better understanding of basic mechanisms of RCF crack growth is necessary. RCF micro cracks appear on the running track a significant amount of time prior to the occurrence of pitting damage. Therefore, it appears crucial to predict the crack growth of RCF cracks to estimate the residual life of a component as major parts of the lifetime are consumed by crack propagation. On the one hand a large number of deterministic numerical modelling approaches are available in the literature, on the other hand qualitative and quantitative RCF crack growth data has been gathered by experimentalists. Both suggest common links between the growth of RCF cracks and crack growth described in structural fatigue. The study presented in this thesis combines for the first time a numerical approach with a large dataset of quantitative crack growth data gathered on a model test rig for AISI52100 bearing steel. Presented study was able to identify key influencing parameters like contact size, load, roughness, residual/hoop stresses and influence of the lubricant on observed RCF crack growth stages. By means of an FEA model, mixed mode stress intensity factors during over-rolling could be predicted for a range of observed RCF crack shapes and related to experimentally observed crack propagation rates. Smaller RCF cracks with surface crack lengths up to 100-150μm showed typical behaviour of shorts cracks as defined in fracture mechanics, growing near or below the threshold stress intensity factor of 521000 bearing steel. Those short RCF cracks showed a high sensitivity to contact conditions like the magnitude of asperity stresses, residual/hoop stresses and influence of the lubricating oil. This is comparable to behaviour which is known to dominate short crack propagation rates in fracture mechanics experiments. Short crack growth rates in fracture mechanic tests are mainly dependent on the applied stress amplitude, residual stresses and microstructure of the specimen. By contrary longer RCF crack (>100-150 μm) have been found to follow a Paris` law with a stress intensity exponent of 4, dominated by shear mode (mode II) propagation for all investigated contact conditions. The stress intensity exponent is well in agreement with exponents derived from fracture mechanics tests with compact tension specimen. Therefore, the conclusion could be drawn that the growth of longer RCF cracks is controlled by the K-field which implies a continuum dominated crack growth mechanism. Consequently, it was possible to provide distinct proof relating the growth of RCF cracks observed in an EHL contact with basic fracture mechanics crack growth mechanisms like short crack growth and a continuum dominated Paris’ law stage. By relating RCF growth to basic fracture mechanics principles it was possible to create a simplified analytical deterministic crack propagation prediction model which is able to estimate the residual life of the critical RCF crack to failure within a factor of 2. Furthermore, a more sophisticated numerical model based on a boundary element approach was used to reconstruct the typical growth pattern of RCF cracks. Due to the complex mixed mode loading at the crack tip which is known to cause a deflection of the crack growth direction only qualitative statements could be made. |
Content Version: | Open Access |
Issue Date: | Dec-2022 |
Date Awarded: | Jul-2023 |
URI: | http://hdl.handle.net/10044/1/113838 |
DOI: | https://doi.org/10.25560/113838 |
Copyright Statement: | Creative Commons Attribution NonCommercial NoDerivatives Licence |
Supervisor: | Kadiric, Amir Dini, Daniele |
Sponsor/Funder: | SKF (Firm) |
Department: | Mechanical Engineering |
Publisher: | Imperial College London |
Qualification Level: | Doctoral |
Qualification Name: | Doctor of Philosophy (PhD) |
Appears in Collections: | Mechanical Engineering PhD theses |
This item is licensed under a Creative Commons License