Fracture Micromechanisms Evaluation of High-Strength Cast Irons Under Thermomechanical Fatigue Conditions

Abstract

The truck industry has constantly searched to increase the performance of heavy-duty diesel vehicles, either by reducing weight and the size of the engines, or by increasing their power. In this sense, higher strength grades of cast iron have been developed and tested under thermomechanical conditions, and the results have been positive, since life under such conditions has increased significantly. The purpose of this paper was to shed some light on the fracture micromechanisms acting during crack nucleation and growth under thermomechanical loading conditions, in two cast iron types for use in cylinder head manufacturing, namely gray iron grade 300 (GI 300) and compacted graphite iron grade 500 (CGI 500). The results were compared with those from the standard grades GI 250 and CGI 450. In both gray and compacted graphite irons, fractographic examination showed that the crack starts at graphite tips, grows through the graphite skeleton inside a eutectic cell, and progresses by the coalescence of multiple fatigue cracks from one eutectic cell to another, fracturing the matrix at eutectic cell boundaries. In CGI 500, the graphite in eutectic cells ended with a change in the graphite shape, from vermicular to a round shape end. This brings additional difficulties for the crack propagation process, and, together with the rough interface graphite/matrix and with the thick eutectic cell boundaries, it explains the outstanding thermomechanical results with the CGI 500.

Introduction

Cast irons are the most common materials used to make components of complex geometries and thin sections and for application that requires the combination of thermal and mechanical properties, simultaneously, to achieve a desired performance. The graphite provides good thermal conductivity, while the matrix is mostly responsible for the mechanical properties. As an example, diesel engine cylinder heads are subject to sudden temperature and loading variations during the operation period. There are regions, very close together, that heat rapidly by the ignition valves and, at the same time, cooled down by the cooling system, restricting the material thermal expansion. This fact results in a high compressive stress state at high temperature, inducing plastic deformations in places where the yield limit has been reached.1 After the engine shutdown, the temperature drops and tensile stresses are generated as a result of the previous compressive plastic deformations. These internal stresses generated by the regular engine-working period (start-up, operation, and shutdown) reveal a stress–strain hysteresis cycle, called out-of-phase thermomechanical fatigue, OP TMF. The result can be a crack formation, between the inlet and outlet valves of the cylinder head (Figure 1).

Figure 1
figure1

Cylinder head with cracks between inlet and outlet valves.

Seeking emission reduction as well as higher engine performance, the truck industry is pushing for the development of new diesel engines aiming for the enhancement of power density and operation under high peak firing pressure.2 New grades of cast irons, with higher strength, have been proposed, opening new alternatives to design engineers. For cylinder heads, the strength requirements are continuously growing in association with good thermal conductivity. In gray irons, the change from the traditional grade GI 250 to the grade GI 300 made it possible to increase the power density of engines. Besides that, the introduction of compacted graphite iron for cylinder heads, starting with grade CGI 450, allowed the use of higher peak firing pressures. The advent of CGI 500, still in restricted use, should bring new design opportunities for both cylinder blocks and heads.

The design of new diesel engines must consider the complex interaction between microstructure and the mechanical and thermal cycles. Understanding the micromechanisms of damage during OP TMF still leaves many open questions.3 According to Löhe et al.,4 multiple mechanisms occur during the OP TMF: plastic deformation, creep, oxidation, strain hardening, nucleation and crack propagation. The cast iron response to these mechanisms depends on another very important factor that determines the components useful life: the graphite morphology and the size of the graphite cells present in the microstructure. During the tensile and compressive loading cycle steps, the microstructure undergoes deformations, and these graphite particles act as stress concentrators, representing points for crack initiation and consequent propagation over time. This behavior is much more severe during the tractive part of the loading cycle. As mentioned by Seifert and Riedel,5 the graphite in cast iron weakens the material tensile properties by decreasing the stiffness, as there is partial delamination of the graphite particle from the matrix. This generates an asymmetric tensile–compressive behavior in cast iron. Ghodrat et al.6 showed that Young’s modulus in compression is 5% greater than in tension.

The aim of this paper is to discuss in detail the effect of the graphite morphology on the lifetime of gray and compacted graphite cast irons under OP TMF. The fracture micromechanisms from the standard cast iron grades, GI 250 and CGI 450, are discussed and compared to the ones from higher strength grades, GI 300 and CGI 500. All cast irons tested here present a pearlitic matrix structure, with very similar hardness. The final goal is to provide further information for the use of higher strength grades in cylinder head design and fabrication.

Experimental Procedures

Materials

Two gray irons (GI) and two compacted graphite irons (CGI) were tested. The average properties and microstructures of the cast irons are presented in Table 1. Gray iron samples were obtained from the fire face of an I4 cylinder head, while CGI samples were cut from the fire face of an I6 cylinder head. For each family (gray and compacted), the matrix is 100% pearlite, with similar hardness; the main differences are the graphite size and shape. Figure 2 presents the general aspect of the graphite in gray iron and in CGI. CGI 500 was Mo alloyed (0.15–0.20%). The purpose is to reduce the stress relaxation rate during TMF conditions.7 Figure 3 shows the presence of intercellular Mo-rich carbide in CGI 500. The presence of small shrinkage microporosities in CGI is illustrated in Figure 4, with size smaller than 200 μm.

Table 1 Microstructure and Mechanical Properties at RT (Mechanical Properties, Cell Size, and Graphite Particles: Mean Values of 3 Samples; Nodularity: Mean Values of 8 Samples)
Figure 2
figure2

Microstructure of GI 250 (a), GI 300 (b), CGI 450 (c) and CGI 500 (d). Without etching. 100 X.

Figure 3
figure3

Intercellular Mo-rich carbide in CGI 500, with size lower than 40 μm. 2% Nital, 500 X.

Figure 4
figure4

Shrinkage microporosity in CGI 450, with size lower than 200 μm. 200 X.

OP TMF Test Procedure

The specimen dimensions followed ASTM E2368-10 (2017), with 6 mm diameter and 15 mm gauge length. After machining, they were ground until #1000 grit sand paper. Figure 5 shows the applied thermal cycles, from 50 to 420 °C, with a dwell time of 180 s, under strain control and total constraint. The test apparatus used an extensometer model MTS 632.54F14, with ceramic stems and size of 12 mm.

Figure 5
figure5

Temperature and stresses evolution during TMF testing. Example for CGI 450.

The heating of the specimens was done using a 7.5-kW, 200-kHz induction furnace. The sample temperature was controlled with an infrared pyrometer, and additionally two contact thermocouples (type K) were used to monitor the temperature in the gage length. The cooling of the specimens was done by blowing air and by using water-cooled grips. Detailed experimental procedures can be found elsewhere.9

For evaluation of the mechanical behavior, the half-life hysteresis was considered, with life being determined as the number of cycles where the maximum stress drops by 10% or a complete fracture occurred. Some tests were interrupted during the lifetime, to study crack initiation and propagation. After testing the specimens, their fracture surfaces were analyzed.

Results and Discussion

Figures 5 and 6 present the stress development and a typical hysteresis curve formed from the OP TMF first and ½ Nf cycles. It is observed that in the 1st cycle, during the heating process, compressive stresses are formed due to the total strain restriction. The yield limit was reached at the dwell, and plastic deformation and stress relieving took place. On cooling, tensile stresses are developed and again some yielding is reached causing stress relieving. As the damage progresses, the maximum stress decreases and the test stops when failure criterion is attained (Figure 7).

Figure 6
figure6

Hysteresis of the first cycle and half-life. CGI 450.

Figure 7
figure7

Typical stresses evolution during TMF testing between 50 and 420 °C. CGI 450.

Table 2 and Figures 8 and 9 show the lifetime results from the OP TMF, for the four grades of cast irons. CGI presents much higher TMF lifetime than gray irons, as also reported by Langmayr et al.10 A strength increase in gray iron, from 250 to 300 MPa (YS increase from 227 to 285 MPa), results in 30% increase on OP TMF lifetime, and this is attributed to the refinement and distribution of the graphite. On the other hand, the effect of increasing strength in CGI, from 450 to 500 MPa (YS increase from 333 to 360 MPa), leads to a large increase in OP TMF lifetime (150%). It must be remembered that this is due mainly to changes in the graphite structure, as the pearlitic matrices are very similar (Table 1). The effect of graphite size refinement was also reported by Skoglund et al.11 for gray irons and by Ghodrat12 for CGI. From the half-life hysteresis, the maximum and minimum stresses values are evaluated. It is noteworthy that during the tests, for attaining the same test condition, gray cast irons sustain much lower tensile strength values (220 and 214 MPa) than those of CGI 450 (334 MPa) and CGI 500 (358 MPa); this is a function of their mechanical strengths and defects generated during the fatigue life. It must be clarified that the lifetime results in Figures 8 and 9 are valid only for the imposed thermal cycle and 100% constraint (Figure 5), which is much more severe than the cycles supported by real parts, like cylinder heads.

Table 2 OP TMF Lifetime (Nf) and Stresses at ½ life
Figure 8
figure8

Results of OP TMF between 50 and 420 °C.

Figure 9
figure9

Correlation of yield strength and TMF life, for gray irons and compacted graphite irons.

The process of crack initiation and propagation is similar for gray iron and CGI. Graphite particles play important roles in both fracture steps. The crack starts at the graphite tips and grows mainly through the graphite/matrix interface inside a eutectic cell (In gray iron and in CGI, graphite is a continuous phase inside a eutectic cell). To move further, the crack must fracture the matrix between the eutectic cell boundaries. In the process of cracks linking, the crack can change direction to grow further into the graphite skeleton, until complete fracture of the sample occurs. CGI and gray iron can present particular aspects during this process, as will be seen below.

In gray iron, given the high stress concentration of the graphite tips, the nucleation of the crack occurs at graphite tips, inside the cells or near the surface, depending on the local surface conditions. From Figure 10, it is observed that the cracks are always associated with graphite tips, in some cases connecting with other graphite particles. The crack growth process occurs mainly at the graphite/matrix interface. Figure 11 shows the crack propagation inside the matrix, with steps, typical of striation formation. Near the end of a TMF test (Figure 12), one can see a crack, open to the surface, connecting many eutectic cells. Oxidation is associated with cracks open to the surface; according to Norman et al.,14 since microcracks nucleate independently in the bulk, CGI is considered fairly insensitive to environmental assisted cracking, and oxidation did not play any significant effect on the crack propagation. The process of crack linking is mainly the fracture of the matrix between two eutectic cells (at a cell boundary or dendrite), so that the crack may grow further into the graphite structure (Figure 12). Figure 13 shows that, in gray iron, small shrinkage pores, located on eutectic cell boundaries, do not play any important role in the crack propagation.

Figure 10
figure10

Crack initiation in gray iron.

Figure 11
figure11

Crack growth in the matrix, between graphite particles. GI 300, 30 cycles.

Figure 12
figure12

Opposite sides of a test sample, near the lifetime of GI 250. Crack linking during propagation.

Figure 13
figure13

Detail of Figure 12b. Small shrinkage pore (inside the circle) do not participate on the crack propagation (see arrows). GI 250.

The fracture surface of TMF gray iron samples show many characteristics of those of High Cycle Fatigue (HCF), at room temperature13: very limited deformation on the fracture, most of the fracture area is graphite and graphite/matrix interface (Figure 14). In small areas, one can see the fracture of the matrix (Figure 15a). Also here, similarly to HCF, the fracture of the pearlitic matrix must be differentiated from the fracture in continuous loading, as in the tensile test, where the result of plastic deformation of ferrite lamellae can be clearly seen, resulting in rougher surface (Figure 15b).

Figure 14
figure14

Fracture surface of GI 250, 80 cycles. Most of the fracture area is graphite (white arrow) and interface graphite/matrix (black arrows).

Figure 15
figure15

Pearlitic matrix fracture under TMF (GI 250, 80 cycles) (a) and tensile testing (GI 250) (b).

These observations of crack initiation and propagation are in accordance with the development of tensile stresses at low temperature (Figure 6).

In CGI, the cracks also started near graphite particles (Figures 16 and 17), as also reported by Normal et al.14 Similar to gray iron, the cracks in CGI grow preferentially through the interface graphite/matrix (Figure 18), even if here this interface is not smooth, as it is in gray iron. The first step of the growing process takes place inside a eutectic cell (where graphite is a continuous phase) (Figure 19), and cell boundaries (and dendrites) represent a barrier to crack propagation. However, even though multiple cracks are formed inside a eutectic cell in CGI, their propagation is more difficult than in gray iron, due to the shape of the graphite, the roughness of the graphite/matrix interface and the space between cells. This results in a higher OP TM life than in gray iron. It was observed that Mo carbides and small shrinkage porosities (Figures 3 and 4) are not preferable sites for crack initiation or propagation path, confirming the findings of Norman et al.15

Figure 16
figure16

Crack initiation in CGI, near the surface of the sample.

Figure 17
figure17

Crack initiation in CGI, at graphite particles. Crack on the matrix, propagating with steps (see arrow).

Figure 18
figure18

The crack propagated preferably through the graphite/matrix interface. CGI 500, 380 cycles.

Figure 19
figure19

Crack inside a eutectic cell. CGI 450, 277 cycles.

Figure 20 shows a crack at the life end during TMF testing, propagating through the graphite structure inside the cells and fracturing the matrix at cell boundaries. According to Pirgazi et al.16, in CGI samples fractured in thermomechanical test, the area with graphite at the fracture surface is three times the area of graphite in any section, confirming that the fracture occurs preferably along the graphite skeleton.

Figure 20
figure20

From one eutectic cell to another, the crack had to go through the matrix (eutectic cell boundary or dendrite).

The fracture surface of CGI TMF samples show, at low magnification, cleavage areas, mainly in CGI 500. At high magnification, those facets show the pattern of fatigue, with striations, revealing that, in this case, the fatigue crack process occurs in the cleavage planes (Figure 21).

Figure 21
figure21

Cleavage fracture of the pearlitic matrix. On the surface, fatigue striations (perpendicular to the arrows) are observed.

CGI 500 also shows that the compacted graphite particles ended, at cell boundaries, in a very special shape, with a round end (Figure 22). In Figure 23, it is observed that the graphite morphology from the CGI 450 is quite different from the ones from the CGI 500 (Figure 24). This compacted graphite morphology of CGI 500, together with the higher nodularity and with the rough graphite/matrix interface, can explain the great difficulty to grow a crack in this microstructure, resulting in very high TMF life (Figure 8).

Figure 22
figure22

Fracture of CGI 500. Most of graphite worms ended, at eutectic cell boundaries, changing the shape to a nodule.

Figure 23
figure23

Graphite structure of CGI 450. Deep etching.

Figure 24
figure24

Graphite structure of CGI 500. Deep etching.

In summary, the fracture path presented by gray iron and by CGI shows the crack initiation at the tips of graphite particles. The crack propagates through the graphite skeleton in the eutectic cell, mainly through the graphite/matrix interface. The linking of the cracks can locally change the crack direction, since there is an eutectic cell in the neighborhood that can be reached very easily. Eutectic cell boundaries represent barriers to crack propagation for both groups of cast irons, so an increase in the cell count can improve the resistance to OP TMF. The graphite morphology of CGI, together with the rough graphite/matrix interface and with a thick cell boundary, resulted in higher TMF lifetime for CGI, compared to gray iron.

Conclusions

The experimental results for pearlitic gray iron and CGI, tested under out-of-phase thermomechanical conditions in temperatures between 50 and 420 °C, allow the following conclusions:

  • Crack initiation was always associated with graphite particles, in gray iron and in compacted graphite iron, for all grades. Small shrinkage porosities, located at cell boundaries, do not act as crack initiation sites.

  • Crack growth happens through the graphite structure, in gray and in compacted graphite irons, mainly through the interface graphite/matrix. From one eutectic cell to another, the fracture goes through the matrix cell boundaries and in CGI also through the matrix between the graphite worms.

  • In gray irons, the preferred path for crack growth is also through the graphite/matrix interface. Small areas of fractured pearlite are observed.

  • In CGI, the fracture shows many areas of fractured pearlite and graphite. In the matrix of CGI 500, the crack grows by fatigue, also along cleavage planes.

  • An increase of strength in gray iron, from 250 to 300 MPa, results in 30% increase in TMF lifetime, and this is attributed to the refinement of the graphite. On the other hand, the effect of increasing strength in CGI, from 450 to 500 MPa, leads to a big increase in TMF lifetime (150%).

  • CGI 500 shows a very particular graphite structure in the eutectic cell, with a change from worms to a round shape at the graphite ends, near the eutectic boundaries. This imposes a difficulty for crack propagation.

  • GI 300 and CGI 500 are very promising alternatives for use in cylinder heads.

References

  1. 1.

    S. Trampert, T.G. Gocmez, S. Pischinger, J. Eng. Gas Turbines Power 130(1), 1–10 (2008)

    Article  Google Scholar 

  2. 2.

    W. L. Guesser, L.C Guedes. The new generation of engines and the challenges for the foundry industry. AEA SIMEA, São Paulo, Brasil, 2016. http://www.proceedings.blucher.com.br/article-details/the-new-generationof-engines-and-the-challenges-for-the-foundry-industry-23721. Accessed 26 June 2019

  3. 3.

    A. Bignonnet, E. Charkaluk, Thermomechanical fatigue in the automotive industry. Eur. Struct. Integr. Soc. 29, 319–330 (2002)

    CAS  Article  Google Scholar 

  4. 4.

    D. Löhe, B. Tilmann, K.H. Lang. 5th International Conference on Low Cycle Fatigue. (Berlin, 2003), pp 161–176

  5. 5.

    T. Seifert, H. Riedel, Int. J. Fatigue 32(8), 1358–1367 (2010)

    CAS  Article  Google Scholar 

  6. 6.

    S. Ghodrat, A.C. Riemslag, L.A.I. Kestens, R.H. Petrov, M. Janssen, J. Sietsma, Metall. Mater. Trans. A 44(5), 2121–2213 (2013)

    Article  Google Scholar 

  7. 7.

    V.L. Diaconu, T. Sjögren, P. Skoglund, A. Diószegi, Stress relaxation of compacted graphite iron alloyed with molybdenum. Int. J. Cast Met. Res. 26(1), 51–57 (2013)

    CAS  Article  Google Scholar 

  8. 8.

    R.T. Wimber, AFS Trans. 88, 717–726 (1980)

    CAS  Google Scholar 

  9. 9.

    M. H. Ferreira. Análise da vida em fadiga termomecânica de ferros fundidos cinzento e vermicular. Master Thesis, Universidade de São Paulo, São Carlos, 2017. http://www.teses.usp.br/teses/disponiveis/18/18158/tde-05092017-081625/pt-br.php

  10. 10.

    F. Langmayr, F. Zieher, F.M. Lampic, MTZ 4, 298–304 (2004)

    Google Scholar 

  11. 11.

    P. Skoglund, J. Elfsberg, N. Ipek, L.V. Diaconu, M. Larsson, P. Schmidt, Mater. Sci. Forum 925, 377–384 (2018)

    Article  Google Scholar 

  12. 12.

    S. Ghodrat. Thermo-Mechanical Fatigue of Compacted Graphite Iron in Diesel Engine Components. Ph.D. Thesis, Delf University of Technology (2013)

  13. 13.

    W.L. Guesser, L.C. Guedes, A.L. Müller, V.B. Demetrio, A. Rabelo, Mater. Sci. Forum 925, 296–303 (2018)

    Article  Google Scholar 

  14. 14.

    V. Norman, P. Skoglund, J. Moverare, Int. J. Cast Met. Res. 29(1–2), 26–33 (2015)

    Google Scholar 

  15. 15.

    V. Norman, P. Skoglund, D. Leidemark, J. Moverare, Int. J. Fatigue 88, 121–131 (2016)

    CAS  Article  Google Scholar 

  16. 16.

    H. Pirgazi, S. Ghodrat, L.A.I. Kestens, Mater. Charact. 90, 13–20 (2014)

    CAS  Article  Google Scholar 

Download references

Acknowledgements

This paper is an invited submission to IJMC selected from presentations at the 2nd Carl Loper 2019 Cast Iron Symposium held from September 30 to October 1, 2019, in Bilbao, Spain. The authors would like to thank Tupy Foundry for the financial support and for providing the materials. Two of the authors give thanks to Conselho Nacional de Desenvolvimento Científico e Tecnológico, CNPq, Brasil (Proc. 306388/2017-0 and Proc. 314941/2018-5).

Author information

Affiliations

Authors

Corresponding author

Correspondence to Wilson Luiz Guesser.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Bon, D.G., Ferreira, M.H., Bose Filho, W.W. et al. Fracture Micromechanisms Evaluation of High-Strength Cast Irons Under Thermomechanical Fatigue Conditions. Inter Metalcast 14, 696–705 (2020). https://doi.org/10.1007/s40962-019-00399-w

Download citation

Keywords

  • compacted graphite iron
  • gray iron
  • thermomechanical fatigue
  • cylinder head