| Project Team | Institution | Role | ||
| Professor Yuri Estrin | Monash University | Associate Investigator | ||
| Dr Weizhong Han | Monash University | Research Fellow | ||
| Professor Peter Hodgson | Deakin University | Chief Investigator | ||
| Professor Mark Hoffman | The University of New South Wales | Chief Investigator | ||
| Dr Christopher Hutchinson | Monash University | Chief Investigator | ||
| Mrs Shokoufeh Malekjani | Deakin University | Postgraduate Student | ||
| Mr George Tang | The University of New South Wales | Postgraduate Student | ||
 |
|
|
|
|
Executive Summary
The fatigue program is well underway with an appropriate balance between grain size and precipitate effects, and experiment and modeling, and the work is coherently arranged across the three nodes involved in the work.
Scientific accomplishments include characterization of the effect of grain size on fatigue crack growth behaviour, the instability of cryo-rolled structures to cyclic deformation and the development of a new approach for characterizing the internal stress during cyclic loading in precipitate containing alloys.
The fatigue program is well underway with an appropriate balance between grain size and precipitate effects, and experiment and modeling, and the work is coherently arranged across the three nodes involved in the work.
Scientific accomplishments include characterization of the effect of grain size on fatigue crack growth behaviour, the instability of cryo-rolled structures to cyclic deformation and the development of a new approach for characterizing the internal stress during cyclic loading in precipitate containing alloys.
Project Aims/Targets
Fatigue loading is typically classified into two categories: high cycle fatigue (applied stress amplitudes at a small fraction of the nominal yield stress and fatigue lives are usually greater than 107 cycles) and low cycle fatigue (stress amplitudes are a significant fraction of the nominal yield stress or may exceed the yield stress and the fatigue lives are usually less than 104 cycles). High cycle fatigue and low cycle fatigue correspond to different engineering environments and strategies to design materials suitable for these are required.
The overall aim of this project is to contribute to the understanding of the fundamental principles governing fatigue of Al alloys. The ultimate objective is to extract sufficient understanding to formulate design strategies for HCF and LCF fatigue resistant Al alloys.
An pproach that isolates the different microstructural features that influence fatigue in Al alloys is the most desirable way to examine the effect of microstructure on HCF and LCF behaviour. The microstructural features of most interest are the material grain or cell size (conventional grain sized materials and UFG materials), grain boundary character (e.g. misorientation distribution), solutes present in solid solution and precipitates. The approach being adopted isolates the different effects so that a quantitative description relating the microstructural feature to the mechanical response can be developed and ultimately this can be used to design Al alloys with improved fatigue properties.
Fatigue loading is typically classified into two categories: high cycle fatigue (applied stress amplitudes at a small fraction of the nominal yield stress and fatigue lives are usually greater than 107 cycles) and low cycle fatigue (stress amplitudes are a significant fraction of the nominal yield stress or may exceed the yield stress and the fatigue lives are usually less than 104 cycles). High cycle fatigue and low cycle fatigue correspond to different engineering environments and strategies to design materials suitable for these are required.
The overall aim of this project is to contribute to the understanding of the fundamental principles governing fatigue of Al alloys. The ultimate objective is to extract sufficient understanding to formulate design strategies for HCF and LCF fatigue resistant Al alloys.
An pproach that isolates the different microstructural features that influence fatigue in Al alloys is the most desirable way to examine the effect of microstructure on HCF and LCF behaviour. The microstructural features of most interest are the material grain or cell size (conventional grain sized materials and UFG materials), grain boundary character (e.g. misorientation distribution), solutes present in solid solution and precipitates. The approach being adopted isolates the different effects so that a quantitative description relating the microstructural feature to the mechanical response can be developed and ultimately this can be used to design Al alloys with improved fatigue properties.
Project Progress: Technical Details and Research Outcomes
To accomplish our objectives the effects of microstructure on the fatigue response of Al alloys are naturally separated into two streams:
A2.1 Grain Size and Grain Boundary Character Effects on Fatigue of Light Alloys
The specific target for this work is the development of an experimental and theoretical understanding of the effect of grain size and grain boundary character on cyclic hardening response of Al alloys under both HCF and LCF loading conditions.
The main achievements of effect of grain size on fatigue project have been:
To accomplish our objectives the effects of microstructure on the fatigue response of Al alloys are naturally separated into two streams:
A2.1 Grain Size and Grain Boundary Character Effects on Fatigue of Light Alloys
The specific target for this work is the development of an experimental and theoretical understanding of the effect of grain size and grain boundary character on cyclic hardening response of Al alloys under both HCF and LCF loading conditions.
The main achievements of effect of grain size on fatigue project have been:
Characterization of the effect of cryo-rolling on the cyclic deformation characteristics of Al
The fatigue response of cryo-rolled CP Al and 2xxx materials has been performed and the fatigue response is inferior to the as-received material. Cryo-rolled pure Al is undoubtedly unstable in terms of microstructure under the condition tested while cryorolled 2024 Al alloy behaves more stable under the same condition of testing. In addition, cryo-rolling deteriorates the low cycle fatigue performance of the materials tested. Performing annealing heat treatments on the cryo-rolled pure Al improved its fatigue performance in terms of fatigue life and cyclic softening ratio, while in case of 2024 Al alloy no dramatic change of behaviour was observed.
The fatigue response of cryo-rolled CP Al and 2xxx materials has been performed and the fatigue response is inferior to the as-received material. Cryo-rolled pure Al is undoubtedly unstable in terms of microstructure under the condition tested while cryorolled 2024 Al alloy behaves more stable under the same condition of testing. In addition, cryo-rolling deteriorates the low cycle fatigue performance of the materials tested. Performing annealing heat treatments on the cryo-rolled pure Al improved its fatigue performance in terms of fatigue life and cyclic softening ratio, while in case of 2024 Al alloy no dramatic change of behaviour was observed.
Modelling the cyclic plasticity of ultra-fine grained materials
A new model for the mechanical behaviour in UFG grains subjected to cyclic loading has been developed. Emphasis is understandably placed on trying to properly account for the internal stresses observed experimentally and an initial model based on internal stresses developed from the plastic incompatibility between the walls and interiors of the cellular structure has been developed. It is shown that such a single source of backstresses is insufficient to account for the experimentally observed values unless unrealistically high volume fractions of cell walls are assumed.
A new model for the mechanical behaviour in UFG grains subjected to cyclic loading has been developed. Emphasis is understandably placed on trying to properly account for the internal stresses observed experimentally and an initial model based on internal stresses developed from the plastic incompatibility between the walls and interiors of the cellular structure has been developed. It is shown that such a single source of backstresses is insufficient to account for the experimentally observed values unless unrealistically high volume fractions of cell walls are assumed.
3.
Characterization of the effect of grain size on the fatigue crack growth behaviour in 6061
The cyclic fatigue behavior of ultra fine grained (UFG) 6061 aluminium alloy produced by equal channel angular processing (ECAP) has been investigated. UFG materials are commonly made through severe plastic deformation (SPD) materials processing techniques such as equal angular channel pressing (ECAP). The high strains introduced by ECAP lead to large changes in microstructure and hence the mechanical properties of a material. Most of these mechanical properties have been extensively studied and documented in literature with the exception of cyclic fatigue crack growth.
Initial fatigue testing was conducted on the overaged UFG 6061 aluminium alloy samples to investigate its crack propagation behaviour. The crack almost always propagated at an approximately 45 degree angle to the far field stress axis inducing mixed mode loading which complicated the determination of stress intensity factor ranges. Finite element modeling (FEM) was used for this purpose and stress intensity factor ranges were determined at different crack growth lengths. More results need to be collected for a larger range of stress intensity factor ranges, especially at low SIF ranges to determine whether a threshold exists. Current results are shown below in Figure A4:
The cyclic fatigue behavior of ultra fine grained (UFG) 6061 aluminium alloy produced by equal channel angular processing (ECAP) has been investigated. UFG materials are commonly made through severe plastic deformation (SPD) materials processing techniques such as equal angular channel pressing (ECAP). The high strains introduced by ECAP lead to large changes in microstructure and hence the mechanical properties of a material. Most of these mechanical properties have been extensively studied and documented in literature with the exception of cyclic fatigue crack growth.
Initial fatigue testing was conducted on the overaged UFG 6061 aluminium alloy samples to investigate its crack propagation behaviour. The crack almost always propagated at an approximately 45 degree angle to the far field stress axis inducing mixed mode loading which complicated the determination of stress intensity factor ranges. Finite element modeling (FEM) was used for this purpose and stress intensity factor ranges were determined at different crack growth lengths. More results need to be collected for a larger range of stress intensity factor ranges, especially at low SIF ranges to determine whether a threshold exists. Current results are shown below in Figure A4:
Microstructural investigation is under way, performed using transmission electron microscopy (TEM) of the bulk material with samples being prepared by twin jet electropolishing, and of a region 100um near the crack with samples being prepared by focused ion beam (FIB) milling. Early results indicate that grain size values decrease with the number of passes and levels off at 4 passes. Further investigations will involve EBSD analysis of microstructure, more fatigue testing, and the study of how heat treatment parameters can be changed to influence fatigue properties.
A2.2 Solute and Precipitate Effects on Fatigue of Light Alloys
The specific target for this work is the development of a physically-based model describing the effect of solutes and precipitates on the cyclic hardening response of Al alloys as a function of the applied plastic strain amplitude. For this work a model Al-4Cu alloy has been chosen.
The main achievements of the Al-4Cu fatigue project have been:
A2.2 Solute and Precipitate Effects on Fatigue of Light Alloys
The specific target for this work is the development of a physically-based model describing the effect of solutes and precipitates on the cyclic hardening response of Al alloys as a function of the applied plastic strain amplitude. For this work a model Al-4Cu alloy has been chosen.
The main achievements of the Al-4Cu fatigue project have been:
General cyclic stress-strain behavior of Al-4Cu alloy as a function of precipitate state
The general cyclic stress-strain behavior of five aging states of Al-4Cu alloy, tested under five constant plastic strain amplitudes, were acquired. For the under aged alloy (aging for 10min and 30min at 200ºC), the cyclic stress-strain curves show a very large cyclic hardening behavior. In the case of Al-4Cu-10min alloy the hardening is ~80 MPa; however, for the peak aged (1.5h) and over aged alloy (8h and 4 days), the cyclic stress-strain curves demonstrate a slight softening behavior. Surface observations of the fatigue samples show that slip bands were formed inside the grains in the underaged materials and cracks can be nucleated both at the grain boundaries and slip band; while for the peak aged and over aged alloys, the cracks are formed only at the grain boundaries, and there is little slip localization inside the grains.
The general cyclic stress-strain behavior of five aging states of Al-4Cu alloy, tested under five constant plastic strain amplitudes, were acquired. For the under aged alloy (aging for 10min and 30min at 200ºC), the cyclic stress-strain curves show a very large cyclic hardening behavior. In the case of Al-4Cu-10min alloy the hardening is ~80 MPa; however, for the peak aged (1.5h) and over aged alloy (8h and 4 days), the cyclic stress-strain curves demonstrate a slight softening behavior. Surface observations of the fatigue samples show that slip bands were formed inside the grains in the underaged materials and cracks can be nucleated both at the grain boundaries and slip band; while for the peak aged and over aged alloys, the cracks are formed only at the grain boundaries, and there is little slip localization inside the grains.
Dynamic GP zone precipitation during room temperature cyclic deformation of under aged Al-Cu
TEM and HRTEM observations illustrate that the cyclic hardening behavior of the Al-4Cu-10min alloy is due to the formation of a large number of small GP zones. Hence, one can say that dynamic precipitation occurs for the Al-4Cu-10min alloy during the constant plastic strain controlled fatigue tests at room temperature. Surprisingly, tests conducted at different cyclic frequencies demonstrate that the dynamic precipitation behavior is relatively independent of test frequency (i.e. strain-rate). Both the size and number density of GP zones continually increase with the number of cycles until reaching the peak stress level. After passing the peak stress, the cyclic stress-strain curves show a rapid softening, which appears to be due to slip localization behavior. In order to identify the relationship between the dynamic precipitation process and the plastic strain amplitude, a number of tests using an Al-4Cu-10min alloy have been performed at different imposed plastic strain amplitudes but for the same imposed strain-rate. In agreement with a preliminary model to describe the dynamic precipitation process the results demonstrate that the dynamic precipitation process is plastic strain amplitude dependent. Based on these results, a model for the cyclic hardening due to dynamic room temperature precipitation is being developed.
TEM and HRTEM observations illustrate that the cyclic hardening behavior of the Al-4Cu-10min alloy is due to the formation of a large number of small GP zones. Hence, one can say that dynamic precipitation occurs for the Al-4Cu-10min alloy during the constant plastic strain controlled fatigue tests at room temperature. Surprisingly, tests conducted at different cyclic frequencies demonstrate that the dynamic precipitation behavior is relatively independent of test frequency (i.e. strain-rate). Both the size and number density of GP zones continually increase with the number of cycles until reaching the peak stress level. After passing the peak stress, the cyclic stress-strain curves show a rapid softening, which appears to be due to slip localization behavior. In order to identify the relationship between the dynamic precipitation process and the plastic strain amplitude, a number of tests using an Al-4Cu-10min alloy have been performed at different imposed plastic strain amplitudes but for the same imposed strain-rate. In agreement with a preliminary model to describe the dynamic precipitation process the results demonstrate that the dynamic precipitation process is plastic strain amplitude dependent. Based on these results, a model for the cyclic hardening due to dynamic room temperature precipitation is being developed.
3.
Internal stress development during cyclic deformation of Al-Cu containing shear-resistant precipitates
Traditionally the Bauschinger energy parameter is used to describe the changing shape of a hysteresis loop during a fatigue test. This has some relation with the magnitude of internal stress development as a function of cyclic deformation. However, the exact value of internal stress during cyclic deformation cannot be quantified using the Bauschinger energy parameter. We have developed, a new method, based on the idea of using three lines to best-fit the hysteresis loop, to quantify the internal stress for cyclic deformation. According to this method, the internal stress of an Al-4Cu alloy containing θ¢ precipitates was measured. It is demonstrated that the internal stress depends strongly on the imposed plastic strain amplitude and precipitate state. At low plastic strain amplitude level (ξpi/2≤0.001), the internal stress increases quickly with the plastic strain amplitude, and it will reaches a stable state at larger plastic strain amplitude (ξpi/2>0.001). However, the Bauschinger strain continually increases with plastic strain amplitude. The Al-4Cu-1.5h (peak aged) and Al-4Cu-8h alloy (over aged) conditions have similar internal stresses, which are higher than Al-4Cu-30min (under aged) and Al-4Cu-4D (over aged) states. The Al-4Cu-4D alloy has the lowest internal stress of the four states considered in detail. A model for the development of the internal stress and its dependence on plastic strain amplitude and precipitate state is under development. It is based on the introduction of plastic slip irreversibility into an extension of a model recently developed within the Centre for monotonic deformation of the same alloys.
Traditionally the Bauschinger energy parameter is used to describe the changing shape of a hysteresis loop during a fatigue test. This has some relation with the magnitude of internal stress development as a function of cyclic deformation. However, the exact value of internal stress during cyclic deformation cannot be quantified using the Bauschinger energy parameter. We have developed, a new method, based on the idea of using three lines to best-fit the hysteresis loop, to quantify the internal stress for cyclic deformation. According to this method, the internal stress of an Al-4Cu alloy containing θ¢ precipitates was measured. It is demonstrated that the internal stress depends strongly on the imposed plastic strain amplitude and precipitate state. At low plastic strain amplitude level (ξpi/2≤0.001), the internal stress increases quickly with the plastic strain amplitude, and it will reaches a stable state at larger plastic strain amplitude (ξpi/2>0.001). However, the Bauschinger strain continually increases with plastic strain amplitude. The Al-4Cu-1.5h (peak aged) and Al-4Cu-8h alloy (over aged) conditions have similar internal stresses, which are higher than Al-4Cu-30min (under aged) and Al-4Cu-4D (over aged) states. The Al-4Cu-4D alloy has the lowest internal stress of the four states considered in detail. A model for the development of the internal stress and its dependence on plastic strain amplitude and precipitate state is under development. It is based on the introduction of plastic slip irreversibility into an extension of a model recently developed within the Centre for monotonic deformation of the same alloys.
The major activities planned for this project are:
To couple the cyclic deformation and fatigue crack growth experiments being performed at UNSW and Deakin so that comparisons of the effect of grain size on fatigue can be compared in both ECAP and Cryo-rolled specimens of the same grain size but with different grain boundary character.
To incorporate the line tension stress of dislocations bowing out from cell boundaries in UFG materials subject to cyclic loading to test if this is sufficient to account for the back stresses observed experimentally.
To use the model developed for the cyclic deformation of UFG materials to help interpret the experimental behaviour observed in ECAP’ed and Cryo-rolled materials
To develop the theoretical descriptions of the effect of precipitates and plastic strain amplitudes on the cyclic deformation of Al-4Cu, which has now been characterised experimentally.
To develop a model for the room temperature dynamic precipitation observed in underaged Al-4Cu and to test if such dynamically formed structures have better HCF properties than statically formed structures with the same UTS or LCF properties of statically generated microstructures with the same elongation.