|
2002 World Conference on ADI
Environmentally Assisted Embrittlement of ADI - Current
Understanding
R.A. Martínez, S.N.Simison and R.E.Boeri
INTEMA - Faculty of Engineering
National University of Mar del Plata – CONICET
Mar del Plata – ARGENTINA
ABSTRACT
An unusual environmentally assisted embrittlement effect has been reported to affect ADI when it is tested in tension with its surface in contact with water and other fluids. This effect has been verified by several laboratories, but the embrittlement mechanism has not been fully explained yet. A thorough understanding of the causes and preventive actions is necessary to ensure that the steadily increasing number of applications of ADI continue to be safe.
This work gives an updated view of the current understanding of the environmentally assisted embrittlement of ADI. The influence of the environment (different liquids) and of different ADI grades is discussed. The features of the fracture surfaces are also detailed. Recent results of testing under applied potential are presented. An explanation of the fracture mechanism is proposed, based on the recent identification of cracks developed at the last to freeze portions as a result of plastic deformation.
INTRODUCTION
Effects on Mechanical Properties Austempered ductile iron (ADI) is being increasingly used in the fabrication of cast parts for a number of industries, such as railroad, automotive, agricultural and others. ADI combines excellent strength and good ductility with low cost and the ability to produce nearly finished parts through casting. It has been used to replace parts traditionally made of cast, forged or machined steels of different grades. ADI can be produced in different grades [1], reaching minimum properties ranging from 850 MPa to 1600 MPa of tensile strength, and minimum elongation between 10 and 0% respectively.
Shibutani et al [2], and Komatsu et al [3] reported that ADI suffers an uncommon environmentally assisted embrittlement (EAE) effect when it is tested in tension with the sample’s surface in contact with water. Later,
Martínez et al. [4], obtained similar results in an independent laboratory, using different base ductile iron and several ADI grades. These investigations showed that ADI suffers significant reductions in UTS and elongation, that can reach up to 30% and 70% respectively, when tested in tension, as shown in Figure 1. This EAE takes place almost instantaneously, and it reverses immediately when the surface of the sample is dried [2]. On the other hand, impact properties are not affected by contact with water. This suggests that the EAE does not act under high loading rate. The effect shows no dependency with the time of exposition to water, acting almost instantaneously.
Figure 1: Change in UTS and elongation as a result of tensile testing of ADI grades 1, 2 and 3 in contact with water.
Martínez et al. [4] found that EAE of ADI is also caused by contact with other liquids, such as isopropyl alcohol and SAE 30 mineral lubricant oil, as shown in Figure 2. The effect of these liquid fluids is not as marked as that caused by water. They also reported that the effect of water is independent from its pH, and remains unchanged when water based solutions of pH ranging from 5.5 to 11.9 are used.
Figure 2: Change of UTS and elongation caused by contact with lubricant oil and isopropyl alcohol.
Role of Microstructure
The EAE not only affects ADI, but also ductile irons (DI) of other microstructures, such as those having martensitic and pearlitic matrices. All investigations showed that the higher the strength of the DI, the greater the embrittlement effect. Only ferritic matrix ductile iron
has been found to be immune to the contact with water
[2,3]. It is known that a given ductile iron alloy can be
heat treated to show ausferritic, martensitic, pearlitic or
ferritic matrix. Only the first three matrices will suffer
EAE. Therefore, the embrittlement mechanism has to be
primarily related to the microstructure, and not to the
chemical composition of the alloy. Martínez et al.[4] have
discussed the role of the different microconstituents of
the matrix on the EAE effect. They pointed out that EAE
has so far been identified on DI of matrices composed of
ferrite and austenite (ADI), ferrite and cementite
(pearlite) and tempered martensite (fine dispersion of
carbides in a ferritic matrix). Furthermore, in high silicon
steels having a bulk composition similar to that of the
chemical composition of the ductile iron matrix (0.55-
0.65C; 1.8-2.2Si; 0.7-1Mn), embrittlement is also present
in the bainitic microstructure, but only causes a reduction
in elongation [3]. It is clear that the microstructure affects
EAE, but the nature of the effect is uncertain.
The presence of graphite spheroids does not
seem to be responsible for the effect, since EAE is also
detected in steels of similar composition that are free
from graphite, and it is not detected in ferritic ductile iron,
in which plenty of graphite exists. The presence of
austenite in the matrix cannot be the cause of
embrittlement, since it is only present in ADI matrix, and
other microstructures not showing austenite are also
affected.
Cementite and carbides are present in pearlite
and tempered martensite. Some investigations show that
very small transition carbides are present in the
microstructure of ADI. Carbides are certainly not present
in the ferritic matrix. Therefore, there is no definite base
to disregard a possible effect of the presence of carbides
on EAE.
The presence of ferrite, on the other hand,
cannot be considered to be able to prevent EAE, since
even though ferritic matrices have been found not to be
susceptible to embrittlement, all other embrittled
matrices have large amounts of ferrite.
Other factors, not just the phases forming the
matrix, could affect EAE, such as the presence of
interphases and chemical or microstructural
inhomogeneity. Interfaces are favorable paths for
diffusion and are also highly reactive sites. Ausferritic
and pearlitic matrices, that are very susceptible to
embrittlement, both have a large amount of interfacial
area. Low tempering temperature martensites show
incomplete recrystallization of ferrite and carbides, and
also have a large amount of interface area. On the other
hand, ferritic matrix microstructures only show the ferrite-ferrite grain boundary and ferrite-graphite interfaces. The smaller contribution of preferential diffusion paths, or the lower matrix reactivity could then
be preventing the embrittlement of the ferritic matrix. [4].
A new insight into the effect of the matrix
inhomogeneity on EAE has been given recently by Laine [5], who aimed to identify the fracture initiation site in ADI embrittled by contact with water. He examined the surface of the tensile samples during the fracture
process. Square section tensile samples were used.
Sample faces were polished metallographicaly before
testing. It has been shown that microsegregated regions
of the ADI microstructure, usually referred to as Last To
Freeze (LTF), cracked during testing after plastic
deformation starts, as shown in Figure 3. The microstructure of LTF regions after austempering may show, depending on the chemical composition of the DI and the heat treatment practice, different amounts of unreacted austenite, martensite and even small carbides precipitated during solidification or heat treatment. With this microstructure, LTF regions are usually harder and more brittle than the matrix. This may account for the cracking of LTF when the plastic deformation of the
surrounding austempered matrix imposes relatively large stresses on them, as observed by Laine [5]. It is known that the degree of inhomogeneity at the LTF and its extent is affected by the chemical composition of the iron, by the solidification rate, and possibly by the nodule count. Additionally, austempering heat treatment
variables also influence the microstructure of the LTF. This suggests that the quality of ADI may affect the EAE intensity. The validity of this speculation has not been verified yet.
Figure 3: Typical cracking of LTF regions of ADI samples after plastic deformation.
Examination of the Fracture Surface
The fracture surface of the embrittled samples has been examined by Komatsu et al.[3] and by Martinez et al.[4]. Small regions of cleavage fracture are present in most ADI fracture surfaces, conforming a fracture mechanism called quasi-cleavage. The fracture surfaces of tensile test samples tested in contact with embrittling liquids show a larger proportion of cleavage facets than the fracture surface of samples tested in air. Both Komatsu and Martínez identified flat bright portions covering a fraction of the fracture surface of the tensile samples tested in water. These bright portions were characterized primarily by cleavage fracture. Masud et al. [6]
investigated the link between the flat fracture regions on the fracture surface of tensile testing samples, and the fracture initiation site. In an attempt to localize the initiation of the fracture, the tensile sample surface was put in contact with water only at a very small location, by using a hyssop previously wet with water. This was done during the test, at a constant stress level higher than the tensile strength in contact with an aqueous solution, but lower than the dry tensile strength. Under this stress condition, contact of the wet hyssop with the sample surface caused instantaneous fracture. Figure 4 shows a macrography of the fracture surface. The arrow points
the location at which the sample surface has been wet. A nearly round, bright and flat fracture area, of approximately 1mm diameter, originates from the point of contact of the hyssop. The examination of the fracture surface by scanning electron microscopy showed that the fracture mechanism characteristic of the flat region is cleavage, as shown in Figure 5. The fracture surface out from the flat portion, Figure 6, shows a predominantly ductile fracture, characterized by dimples. The marked difference in the brittleness of both fracture types is also emphasized by the extent of plastic deformation around the graphite nodules, which is much greater on the predominantly
ductile fracture. The results proved that water induces brittle cleavage fracture of ADI. It is also suggested that when flat brittle portions are observed on the fracture surface of tensile samples tested in contact with water, they show the location of the fracture initiation
The work of Laine [5] also gives some very recent views on the fracture initiation and propagation. Laine used similar methodology to that of Masud et al., but worked on square section tensile samples of polished surfaces. The surface metallography of the whole sample was photographed before testing. After the sample is broken in contact with water, the surface fracture path can be observed. Laine found that fracture initiates at the LTF regions of the sample, which became cracked during the test, at stress levels surpassing the yield strength of this ADI. The observation of the fracture surface by Scanning Electron Microscopy showed that LTF regions fractured by
cleavage.
Figure 4: fracture surface of a tensile test specimen. The arrow points the location at which a water impregnated hyssop has been put in contact with the sample surface. Note the approximately round, bright area of about 1.5 mm diameter that originates from the contact point. Total sample diameter is 6.5mm.
Figure 5: Fracture surface of ADI embrittled by water (500x).
Figure 6: Predominantly ductile fracture surface of ADI broke in contact with water, shown at a location far from the fracture initiation site (200x).
Cause of Embrittlement
The cause of embrittlement remains unexplained. Shibutani [2] and Komatsu [3] found similarities between the embrittlement of ADI and the
behavior of hydrogen embrittled materials, and concluded that this effect is induced by the generation of hydrogen atoms from water on the ductile iron surface under plastic deformation. Hydrogen atoms would then diffuse into the ductile iron matrix, causing the embrittlement. This affirmation was further supported by their results of tension testing in H2 atmosphere, which caused an embrittlement similar to that caused by water. Nevertheless, the mechanism by which the protons will reduce at the water/ADI interface and cause an almost instantaneous effect has not been explained by the authors. Furthermore, the little dependency of the phenomena with the time of exposition
to water, and its fast reversibility, do not precisely suit the usual characteristics of the hydrogen embrittlement effect. The results of Martínez et al.[4] and Masud et al. [6] do not support the role of hydrogen as proposed by Shibutani and Komatsu, since the concentration of protons in the embrittlement media does not affect the degree of embrittlement. Furthermore, the fact that other liquids, such as mineral oil and alcohol, cause embrittlement, suggest that the explanation of the effect should be based on the influence of some other factors. Hydrogen embrittlement (HE), together with stress corrosion cracking (SCC) and liquid metal embrittlement (LME) are the most
extensively studied environmentally induced cracking (EIC) processes [7]. EIC failures are characterized by brittle failures in which cracks propagate at stress intensity (K) levels lower than the critical values in air or vacuum, as a result of the combined effect of a tensile stress field and the presence of a corrosive media. Corrosion rates are usually quite low. The mechanisms involved in this type of failure are very complex and remain under discussion. In consequence the occurrence of EIC failures in service is still difficult to predict.
HE involves brittle fracture caused by penetration and diffusion of atomic hydrogen into the crystal structure of an alloy. The kinetics of hydrogen generation is accelerated by an increase in the cathodic polarization. Thus, cathodic polarization should enhance HE, while anodic polarization should have the opposite effect.
Recent work by Masud et al [6] investigated the behavior of ADI on tensile testing in aqueous media under controlled electrochemical conditions, aiming to identify whether the loss of ductility can be either inhibited or enhanced by stimulating or avoiding the reduction of protons on the sample’s surface. They used ADI grade 2 (ASTM A 897M–90) tensile test samples, and tested them in contact with aqueous solutions at controlled potential The values of controlled potential applied during tensile testing were chosen based on the results of polarization curves. A potential of –1.45 V (SCE) was used to induce cathodic conditions, in which the generation of hydrogen on
the surface is stimulated, and a potential of –0.55 V (SCE) was used to inhibit hydrogen generation. Their results showed that the magnitude of the embrittlement effect caused by water was not affected by the application of potential. The results suggest that the EIC of ADI is not an electrochemical phenomenon, since neither cathodic nor anodic applied potentials have been able to inhibit or enhance embrittlement.
The characteristics of the embrittling effect caused by water on contact with ADI are quite unique, showing no complete similarity with any other EIC reported for metals. The velocity of the process and its fast reversibility resemble LME. 2002 World Conference on ADI LME causes the catastrophic brittle failure of normally ductile metal alloys when coated by liquid metal and stressed in tension. The fracture mode changes from a ductile to a brittle intergranular or brittle transgranular (cleavage) mode. It has been shown that the stress needed to propagate a sharp crack or a flaw in liquid is significantly lower than that necessary to initiate a crack in the liquid metal
environment. In most cases, the initiation of the propagation of cracks appears to occur instantaneously, with the fracture propagating through the entire test specimen. The velocity of crack or fracture propagation has been estimated to be 10 to 100 cm/s. LME is not a corrosion, dissolution or diffusion–controlled intergranular penetration process. The embrittlement is severe, and the propagation of fracture in the case of LME is very fast as compared to that in stress corrosion cracking [7].
PROPOSED OF FRACTURE MECHANISM
On the basis of the existing knowledge, and given the similarities with LME, the authors propose that the fracture of ductile iron in contact with water proceeds as follows: upon stressing ADI at a certain level above its yield strength, it develops cracks at the LTF regions, as shown by Laine [5]. When this takes place in contact with water or other liquids, the liquid penetrates the crack, the A-A atomic bonds at the crack tip are weakened by the chemisorption of an atom or molecule B, as shown schematically in Figure 7
Figure 7: schematic representation of the weakening of A-A atom bonds at a surface crack tip as a result of the interaction with an atom or molecule B supplied by the surrounding liquid.
The chemisorption process presumably takes place spontaneously or only after the A-A bonds have been strained to some critical value. In any event, electronic rearrangement takes place because of adsorption, and weakens the bonds at the crack tip. When the applied remote stress is increased so that local stress at the crack tip exceeds the reduced breaking strength of A-A bonds, then the crack becomes unstable and grows rapidly. The crack grows initially in a brittle manner, by cleavage, but changes to a ductile mechanism, as it grows far from the fracture initiation site. Taking into account the load conditions in tensile testing, the stress levels, and the sample
dimensions, and assuming a semieliptical surface defect and a K1C value of 90 MPa m1/2, the size of the critical defect can be estimated to be 0.8 mm. This would indicate that if the presence of water activates the rapid growth of a crack, and such crack extends beyond the critical defect size, then, even when the fracture mode changes to a higher energy consuming mechanism, the remaining ligament will not be able to stop fracture, and the sample will collapse. The size of the cleavage fracture surface observed in tensile specimens fractured in contact with water, as shown in Figure 4, is usually of approximately 1.5mm. This size is greater than the critical defect size,
supporting the proposed mechanism.
SUMMARY AND CONCLUDING REMARKS
Water and other liquids cause the embrittlement of ADI. The characteristics of this environmentally assisted cracking effect are quite unique, and share some of the features of liquid metal embrittlement. The effect of the environment is not understood. Recent experiments do not support hydrogen embrittlement as the cause of fracture. An explanation of the fracture mechanism has been proposed by the authors, based on the recent observation of cracking of last to freeze portions of the ADI microstructure upon straining. Future work should be aimed to clarify the mechanism of EIC responsible for the embrittlement of ADI. In particular, it is necessary to identify the role
of the different liquids on fracture. Additionally, the influence of the ADI microstructure on EIC should be investigated.
ACKNOWLEDGMENT
This research was supported by grant 15/G065 of Universidad Nacional de Mar del Plata.
REFERENCES
1. ASTM A 897M - 90. “Standard specification for austempered ductile iron castings”, Annual book of ASTM standards. Philadelphia. (1995)
2. Shibutani et. Al, “Embrittlement of austempered spheroidal graphite cast Iron”, International Journal of Cast Metals Research, 1999, 579-585.
3. Komatsu et al., “Embrittlement characteristics of facture toughness in ductile iron by contact with water”, International Journal of Cast Metals
Research, 1999. 539-544
4. R.A. Martínez, R. Boeri, and J.A. Sikora, “Embrittlement of ADI caused by contact with water and other liquids”. International Journal of Cast Metals Research. 2000, vol 13, .9-15.
5. Laine, B.“Estudio del Fenómeno de Fragilización de Fundiciones Esferoidales Austemperizadas”, Thesis to obtain Materials Engineer degree”. Department of Materials Engineering. Faculty of Engineering. National University of Mar del Plata. Argentina. December 2001.
6. Masud et al. “Embrittlement of Austempered Ductile Iron on contact with water - Testing under applied potential” Submitted to Journal of Materials Science, December 2001.
7. "Metals Handbook", Vol 13: Corrosion. ASM International 9th. Edition.
+ rimarti@fi.mdp.edu.ar
++ www.intema.fi.mdp.edu.ar
Back to top
|
