In situ analysis of cast irons mechanical behaviour using synchrotron x-ray tomography and 3DXRD

(1)

IOP Conference Series: Materials Science and Engineering

PAPER • OPEN ACCESS

In situ analysis of cast irons mechanical behaviour using synchrotron

x-ray tomography and 3DXRD

To cite this article: T Sjögren et al 2020 IOP Conf. Ser.: Mater. Sci. Eng. 861 012039

View the article online for updates and enhancements.

(2)

MCWASP XV 2020

IOP Conf. Series: Materials Science and Engineering 861 (2020) 012039

IOP Publishing doi:10.1088/1757-899X/861/1/012039

In situ analysis of cast irons mechanical behaviour using

synchrotron x-ray tomography and 3DXRD

T Sjögren1,_{S Hall}2_{, L Elmquist}3_{, E Dartfeldt}1_{, E Larsson}1_{, M Majkut}4_, J Elfsberg5_{, P Skoglund}5_{and J Engqvist}2

1_{RISE Research Institutes of Sweden, Borås, Sweden}

2 _{Division of Solid Mechanics, Lund University, Lund, Sweden} 3_{Xylem Water Solutions Manufacturing AB, Emmaboda, Sweden} 4_{ESRF European Synchrotron Radiation Facility, Grenoble, France} 5_{Scania CV AB, Södertälje, Sweden}

E-mail: torsten.sjogren@ri.se

Abstract. When subjecting cast irons to mechanical loading the deformation and damage

mechanisms occur on a microstructural level and are dependent on the inherent microstructure. A deeper understanding of the relation between the different microstructural constituents and the macroscopic mechanical behaviour would be beneficial in material development efforts and for the ability to design and cast components with tailored properties. Traditionally, microscopy examinations on sectioned cast iron samples have been used when analysing the microstructure in cast irons. Since all microstructural heterogeneity is in three-dimensions (3D), methods that provide a three-dimensional characterisation are essential for a deeper understanding of, both the microstructural features as well as the deformation and damage of cast irons. Therefore, different cast iron grades have been studied using synchrotron X-ray tomography and 3D x-ray diffraction (3DXRD) at ESRF in Grenoble, France. The samples were stepwise loaded and unloaded in-situ at in the tomography/3DXRD set-up to study the deformation with regard to microstructural constituents and the microstructural evolution in 3D. Based on the 3D tomography image sequences, digital volume correlation (DVC) was used for full strain field analysis and for the analysis of damage and deformation mechanisms. In addition, 3DXRD data were analysed to provide details on the lattice parameters and lattice strain of individual ferrite grains. This work shows the possibilities of such synchrotron experiments for advanced study of the mechanical behaviour of cast iron.

1. Introduction

The development of highly sophisticated combustion engines is driven by the desire to improve efficiency and power density as well as to reduce emissions. Such developments will inevitably lead to an increase of the combustion pressure and temperature, which has the consequence of an increased loading on the materials. Today, spheroidal graphite iron (SGI) alloyed with silicon and molybdenum, SiMo-alloys, are commonly used in components such as manifolds in exhaust systems, where materials that can withstand the harsh conditions are needed. The mechanical behaviour of these materials has been studied by others [1-4]. However, the microstructural analysis used to correlate the mechanical behaviour to microstructural properties is performed on polished 2D surfaces taken from the position of final fracture of the specimen. For a detailed understanding of the microstructure and micromechanics of the material in-situ 3D analysis is needed. Also, a better understanding of the mechanical behaviour related to the 3D microstructure is essential to develop an understanding and models that can predict

(3)

MCWASP XV 2020

2

how the material will behave under load. In this work x-ray tomography, together with Digital Volume Correlation (DVC) and 3D x-ray diffraction (3DXRD) during in-situ loading (i.e., loading within the measurement set-up) has been used to identify and characterize the deformation mechanisms and corresponding phase interactions. The results will provide a better understanding of the influence of the different microstructural length scales on the macroscopic mechanical response and explain the observed crack propagation paths. In this paper an overview of possible analyses based on tomography and 3DXRD data are presented.

2. Experimental procedure

The experiment was performed at the European Synchrotron Radiation Facility (ESRF) in Grenoble, France, beamline ID11. For the experiment, tensile test samples were machined from 25 mm thick cast plates of SiMo material. A tensile test stage developed at the division of Solid Mechanics, Lund University, for the specific synchrotron measurement setup was used to apply and control tensile load on the specimen. Samples were monotonically loaded in steps, between which x-ray tomography and 3DXRD measurements were performed with the sample displacement held constant. The x-ray beam had a photon energy of 78.4 keV and a beam size of approximately 1500x1500 m2. Figure 1 shows the sample geometry and the setup inside the beamline. Force and displacement data were acquired from which the macroscopic stress and strain were derived. However, the deformation measurement includes deformations taking place in all parts of the tensile stage, i.e. strains are not representative of the actual strain of the SiMo material. The 3DXRD data includes nine, 100 m high slices over the sample height and covering the full gauge diameter width. Analysis of the 3DXRD data using the ImageD11 software (Wright, J., 2005, https://github.com/FABLE-3DXRD/ImageD11/), provides grain-by-grain unit cell parameters from which grain-strains are derived relative to “ideal” unit cell parameters.

Figure 1. Loading stage as mounted in the beamline hutch. The small pictures in the right lower corner show the mounted tensile test specimen and the specimen

geometry (total length≈10 mm, gauge diameter≈1.4 mm).

2D and 3D microstructure investigations, using light optical microscopy (LOM) and scanning electron microscopy (SEM), enabled correlation and verification of the different phases observed in the tomography images. Samples for the microscopy analysis were produced by grinding and polishing according to standard procedures. DVC analysis was performed using LaVision’s StrainMaster software.

(4)

MCWASP XV 2020

3. Results and discussion

The studied ductile iron is alloyed with silicon and molybdenum for improved properties at high temperatures. The standard denomination of the material is EN-GJS-SiMo5-1 and the chemical composition is given in table 1. The main microstructural constituents of the studied SiMo material are graphite in a ferritic matrix. Due to the high levels of Si and Mo, the microstructure also contains molybdenum carbides and spheroidized pearlite [5].

Table 1. The chemical composition in weight percent of the studied EN-GJS-SiMo5-1. The iron content is implicit.

C Si Mn S P Ni Mo Cu Sn Ti

3.16 4.33 0.41 0.008 0.014 <0.050 0.91 0.073 <0.01 0.017

These constituents are found together with shrinkage defects in regions corresponding to the last to freeze areas. Figures 2 and 3 show the microstructure as seen in LOM and SEM, respectively. In figure 2 the ferrite grains are clearly seen. Similar structures are observed in tomography images, as shown in figure 4, where the light grey/white areas correspond to the molybdenum carbides seen in figures 2-3.

Figure 2. LOM image of the SiMo material. Figure 3. SEM image of the SiMo material.

Through segmentation of the tomography images, using suitable grey level thresholds, it is possible to differentiate between the graphite, carbide and shrinkage porosities as shown in figure 5. In the figure, the graphite phase is shown in light grey, the carbides in light blue and the porosities in green. The volume percentage of each constituent (within the analysed domain) is; 9.4% Graphite, 5.3% Carbide and 0.5%shrinkage porosities. The volume analysed in figure 5 is a cubic sub-volume of extracted from the full image before deformation.

(5)

MCWASP XV 2020

4

Figure 4. X-ray tomography picture of SiMo sample in unloaded state.

Figure 5. 3D analysis of microstructure.

In the loading experiments, displacement controlled tensile loading was applied, as shown in figure 6. For this specific sample (DCI1) seven load steps were analysed (including the unloaded state) at which tomography and 3DXRD was done, with the displacement held constant, which resulted in some measured relaxation.

Figure 6. Stress-strain curve for sample DCI1.

By comparing the tomography images acquired at unloaded and loaded states, it is possible to see where in the microstructure that deformation and damage occurred. Comparison of the unloaded state in figure 4 and the deformed state in figure 7 (load step 4), it is observed that cracks occurred in the carbide areas (corresponding to the white areas). It is also observed that debonding occurred at the graphite particles where a gap forms preferentially above and below the graphite particle along the loading axis.

(6)

MCWASP XV 2020

Figure 7. Tomography section at 620 MPa (Step 4).

Figure 8. Axial strain DVC data at 620 MPa (Step 4).

DVC analyses of the tomography images allows quantification of the strains that arise within the sample and to further understand the severity of the damages and deformations observed qualitatively in figure 7. Figure 8 shows example DVC results for the SiMo material corresponding to the structure in figure 7 (i.e., the strain field between the unloaded state and step 4). Figure 9 shows a visualisation of the DVC-derived axial strain field. Central planes along the x-, y- and z-axis show the strain distribution. In addition, an ISO-surface, for which the chosen strain is 10%, reveals the structure of the distribution within the material in 3D. In both figure 8 and 9, the strain distribution reveals a shear behaviour that depends on the size and dimension of the sample. It also shows the strains in the vicinity of the discussed cracks and graphite debonding.

Figure 9. DVC data showing the 3D distribution of axial strain at 620 MPa (Step 4).

Crystallographic unit cell parameters were derived from the 3DXRD data for the individual ferrite grains. The input data for this analysis were 2D diffraction patterns acquired over a 180° rotation of the sample, examples of these are shown in figure 10. Note the smearing of spots at higher loads, indicating an increase in the orientation spread in the grains (inferred as being related to plastic deformation); this

(7)

MCWASP XV 2020

6

leads to reduced number of indexed grains. The number of indexed grains at the centre region drops from 3500 at load step 1 to 250 at load step 4.

Figure 10. Example diffraction patterns as function of load and position (load steps 0-4). The depicted diffraction patterns are taken from the centre 100-micron slice.

The grain-strains determined from the evolution of the unit cell parameters are shown in figure 11 for the axial (11) and horizontal (22, 33) directions. The grain-strains along the loading axis, for the indexed grains, are around 0.2% at load step 4. Since the grains subjected to high strains are not indexed, due to the smearing of the spots, this value reflects only the grains that are elastically deformed.

Figure 11. Average grain-strain evolution (load steps 0-4) for the axial (11) and horizontal (22, 33) components averaged across 100 m high slices over the gauge diameter. Data are

based on the upper, mid and lower 100-micron slice.

The centre of mass of each ferrite grain was also determined from the 3DXRD data, based on which an estimate of the ferrite grain structure can be made. Results from an initial analysis using an adapted distance transform and watershed algorithm is shown in figure 12. Further work is needed to

(8)

MCWASP XV 2020

improve this approach, e.g. to take into account the regions of carbides and porosities, however the grain structure in figure 12 is in good agreement with that in figure 2.

Figure 12. Tesselation of grain centres from 3DXRD overlain on the graphite distribution from the tomography images. The grains

are coloured by index 4. Conclusions

An overview of 3D analyses of the microstructure and micromechanics of cast irons using 3D x-ray tomography and 3DXRD has been presented and shows the usefulness of such synchrotron experiments, with in-situ mechanical tests, for the study of microstructural influence on mechanical behaviour in such materials. Combining 3DXRD, x-ray tomography, 3D image analysis and DVC enables analysis of deformation across multiple scales. With full-field tomography, the microstructure can be analysed at high resolution in 3D in combination with 4D evolution. In addition, DVC gives a continuum strain fields based on the tomography images. 3DXRD data provides additional information for modelling (e.g. initial grain positions and parameters) although further work is required to produce grain maps taking into account structural information as well as crystallographic data. For the analysis of texture evolution alternative approaches are needed (high grain deformation).

Acknowledgments

The authors acknowledge European Synchrotron Radiation Facility (ESRF) in Grenoble, France, for hosting the experiment which was facilitated through an application for academic beam time (experiment ref.no. MA3615).

References

[1] Norman V, Skoglund P and Moverare J 2016 Damage evolution in compacted graphite iron during thermomechanical fatigue testing, Int. J. Cast Metal Res., 29 (1-2), 26-33

[2] Norman V, Skoglund P, Leidermark D and Moverare J 2016 The effect of superimposed

high-cycle fatigue on thermo-mechanical fatigue in cast iron, Int. J. Fatigue, 88, 121-31

[3] Nadot Y, Mendez J and Ranganathan N 2004 Influence of casting defects on the fatigue limit of nodular cast iron, Int. J. Fatigue, 26 (3), 311-19

[4] Hervas I, Thuault A, and Hug E 2015 Damage Analysis of a Ferritic SiMo Ductile Cast Iron Submitted to Tension and Compression Loadings in Temperature Metals 5 (4) 2351-69

[5] Ekström M 2013 Development of a ferritic ductile cast iron for increased life in exhaust