INfluence of casting Defects in the low cycle thermomechanical fatigue damage of lost foam casting AlumiNum Alloys – INDiANA

The main steps of the project are thus (i) to obtain a component microstructure representative sound casting of a fatigue specimen with controlled defects either in location and size and to characterize this cast specimens in 3D in order to validate the casting process and obtain 3D data for numeric models, (ii) to understand the low cycle (thermo)mechanical fatigue micromechanisms by using either in-situ tests under X-ray microtomography and conventional LCF/TMF tests with coupled periodic Digital Image/Volume Correlation, (iii) to propose a predictive fatigue model for heterogeneous and defective materials based on numerical and analytical non-linear homogenization techniques and (iv) to integrate these results in an industrial probabilistic based fatigue design strategy.

1. representative specimens : simulation was used to determine cooling parameters for obtaining representative samples. The development process has progressed. Meanwhile, fatigue specimens were taken from the cylinder heads and a selection (involving 3D characterization and calculations microstructure) was used to determine the most appropriate representative specimens for testing.
2. In-situ testing under CT: The high temperature fatigue tests were conducted on line ID19 of the ESRF in order to see the initiation of cracks and damage evolution in the AS7Cu3 cyclic mechanical stress. A specially designed oven to perform tests with a synchrotron were used.To date, these are the first fatigue tests which have been observed in tomography at a temperature different from room temperature.
3. microstructure Calculations: The FE calculations on 3D microstructures has been realized. 3D tetrahedral mesh (Avizo) is generated from the thresholded matrix without pores. Abaqus elastoplastic calculation is first performed for simplified loading conditions. After the test, a new calculation, on the outcome of the microstructure synchrotron images with the actual boundary conditions compares strain field calculated and field measured by image correlation to understand the influence of the pores.
3. Homogenization: The work done is based on the development of an analytical homogenization model for porous materials with pore shape effect and combined hardening: isotropic-kinematic. This model has been tested and implemented in software using a finite element UMAT on Abaqus. It is in the finishing phase and is compared to a digital progressively homogenization for monotonic loadings, non-proportional or cyclic performed on periodic elementary cells mono / multi-pores.

This research has also potential impacts in the medium term on environmental consideration by a process optimization to enhance structures reliability and weight reduction.

Limodin, N., El Bartali, A., Wang, L., Lachambre, J., Buffiere, J.-Y., and Charkaluk, E., 2014, «Application of X-ray microtomography to study the influence of the casting microstructure upon the tensile behaviour of an Al-Si alloy,« Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms, 324, pp. 57-62.

Wang, L., Limodin, N., El Bartali, A., Réthoré, J., Witz, J.-F., Seghir, R., Charkaluk, É., and Buffiere, J.-Y., June 29–July 2, “Influence of the Casting Microstructure upon the Tensile Behaviour in A319 Al-Si Alloy Investigated by X-Ray Tomography and Digital Volume Correlation,” 2nd International Congress on 3D Materials Science, Wiley - TMS, L’Impérial Palace, Annecy, France, pp. 73–78.

Dahdah, N., Limodin, N., El Bartali, A., Witz, J.-F., Seghir, R., Wang, L., Charkaluk, É., and Buffiere, J.-Y., June 29-July 2, «Influence of the Lost Foam Casting Microstructure on Low Cycle Fatigue Damage of A319 Aluminum Alloy,« 2nd International Congress on 3D Materials Science, Wiley - TMS, L'Impérial Palace, Annecy, France, pp. 97-102.

F Szmytka, N Limodin, L Wang, P Osmond, J Adrien, E Charkaluk , J-Y Buffiere, Probabilistic thermal-mechanical fatigue criterion for lost foam casting aluminium alloys based on 2D/3D porosities distribution, Fatigue Design and Material Defects, Paris 2014

S. Dézecot, V. Maurel, A. Köster, F. Szmytka, JY Buffières, X ray tomographic characterisation of damage in a cast Al alloy during thermomechanical fatigue tests, 3DMS conference, Annecy 2014

Since decades, due to environmental considerations and cost reduction objectives, automotive manufacturers conduct a downsizing strategy, particularly for engine parts as cylinder blocks or cylinder heads. In the later case, aluminum alloys were first chosen to replace cast-iron and, since a few years, Lost Foam Casting (LFC) is introduced in order to replace Die Casting (DC) process as a matter of geometry optimization, cost reduction and consumption control. One of the consequences of this downsizing strategy is an increase of the engine specific power and of the combustion fluxes received by the different components. The cylinder head becomes therefore one of the most critical part of the automotive engine, submitted to high thermomechanical loadings during the start-stop operation.
A major specificity of LFC process is the fact that the cooling rate is relatively slow compared to DC process, which creates a coarser microstructure when measured in terms of Dendrite Arm Spacing (DAS). Besides, the number of voids and inclusions (intermetallics, oxides) increase and form clusters. Recently, it has been demonstrated by some of the partners of the present project that Low Cycle Thermomechanical Fatigue (LCF/TMF) damage of LFC components results from a competition between porosity, intermetallic and eutectic phases which can reduce drastically, in some cases, lifetimes and, consequently, components reliability.
The main objective of INDiANA project is a fine understanding of the thermomechanical fatigue process of such alloys in order to improve fatigue criteria through a multiscale modeling approach and, consequently, the target reliability of the engine components. This requires 3D metallurgical investigation, isothermal and non isothermal fatigue testing with 2D and 3D state-of-the-art full-field measurements, microstructure computations and non-linear homogenization techniques that will bridge the gap between the current phenomenological LCF modeling and the full-field micromechanical simulations.
The main steps of this proposal are thus (i) to obtain a sound casting of a fatigue specimen representative of the component microstructure with controlled defects both in location and size and to characterize this cast specimen in 3D in order to validate the casting process and to obtain 3D data for numerical models, (ii) to understand the low cycle (thermo)mechanical fatigue micromechanisms by using both in-situ tests under X-ray microtomography and conventional LCF/TMF tests with coupled periodic Digital Image/Volume Correlation, (iii) to propose a predictive fatigue model for heterogeneous and defective materials based on numerical and analytical non-linear homogenization techniques and micromechanical analysis and (iv) to integrate these results in an industrial probabilistic based fatigue design strategy. This industrial integration may lead to the careful adaptation of the laboratory methods to propose a macroscopic design protocol from physical microscopic observations and mechanisms.
Success of INDiANA relies upon the combination of the existing expertise of the different partners in X-ray microtomography, fatigue micromechanisms analyses, full-field measurements, casting process, numerical models and micromechanics. In addition, a major strength of the proposed consortium is the acquisition of a strong industrial expertise of components fatigue design over several years of active collaboration via common PhD thesis, postdoctoral and internships supervisions.
This research project has also a potential medium-term impact upon environmental aspects by the process optimization of structural reliability and weight reduction.