Additional elements / wt %
4. Hydrogen embrittlement cracking
The deleterious effects resulting from hydrogen in ferrous materials have been known since 1926, when it was reported that ductility losses in steels were attributed to the presence of hydrogen in the materials (Fast 1965, Cotterill 1961). Considerable experimental evidence exists indicating that hydrogen can degrade the properties of a wide variety of materials, ranging from high strength steels to soft iron (Bernstein and Thompson 1974). Recently, the magnitude of the hydrogen problem has come to be appreciated primarily because of the increasing demands for strength and toughness required of modern materials exposed to hydrogen environments.
A particularly important observation is the fact that hydrogen can be introduced in a component at any time during its fabrication (casting, welding, surface treatment, heat treatment, etc.) or when used in various applications such as pipelines, containers, gas wells, nuclear reactors, and ships (Folkhard et al. 1972, Glikman and Orlov 1968, Kudryavtsev et al. 1972, Smialowski 1962, Sheinker and Wood 1971).
26 H. NAGAI
The most recent type of hydrogen embrittlement to be investigated results from the direct exposure of a metal surface to a gaseous hydrogen environment (Hofmann and Rauls 1961). This form of hydrogen embrittlement has been regarded with increasing concern because of the predicted future widespread use of hydrogen as a fuel (Smialowski 1962). Absorption of hydrogen gas in metals is potentially a serious problem for electric current generating fuel cells and propulsion systems which utilize the hydrogen oxygen reaction as a source of energy or for systems being considered for the storage of high pressure gaseous hydrogen fuel (Schwartz and Ward 1968, Moeckel 1969).
Although an appreciable understanding of the effects of hydrogen on steels has been achieved, and some progress has been made in mitigating the embrittlement problem, the fact still remains that design engineers must incorporate considerable safety factors to insure the prevention of catastrophic failure in structural steel components.
A number of methods of inhibiting hydrogen embrittlement in high strength steels have been under study. These techniques include changes in microstructure (Cain and Troiano 1965, McCoy 1974, Bernstein and Thompson 1976), changes in alloy composition (Lagneborg 1969, Beck et al. 1971), baking (Johnson et al. 1958, Sims 1959, Troiano 1959), surface prestressing (Bates 1970, Carter 1972), plating, cathodic protection (McEowan and Elsea 1965), non-metallic coating (Speller 1951, Brewer 1974), selective changes in surface composition by heat treatment, and modification of the embrittling environment (Baker and Singleterry 1972, Zecher 1976).
In general, however, these techniques cannot always be applied, because a number of serious limitations exist. For example, for microstructural and base alloy composition modifications, overall mechanical properties, fabricability and economic considerations control the applicability of such methods to the extent that little flexibility exists for wide variations without other accompanying problems. Numerous methods have been studied to prevent hydrogen entry into a high strength steel component by the formation of a barrier between the steel and the service environment. Cathodic protection of steels can be limited by the absorption of hydrogen generated at the cathodic surfaces if high local current densities are applied (Uhlig 1963, McEowan and Elsea 1965). Metallic platings have been developed for the protection of steels, but there has been an accompanying embrittling action resulting from the plating process itself (Williams et al. 1960, Beck and Jankowsky 1960).
Because limitations exist with all of these techniques, research is being conducted to improve the current methods and to develop new ones to inhibit hydrogen embrittlement in high strength steels. One new method is the use of rare earth additions in these alloys.
4.1. Effect of rare earths on hydrogen-induced delayed failure
Several studies indicate that rare earth additions to steels offer potential to minimize hydrogen embrittlement without degrading the baseline properties of the alloys them- selves. The important consideration here is the mechanism by which rare earths minimize embrittlement. In one instance, improvement was attributed to the absorption of hydrogen to form stable, non-embrittling hydrides. In another instance, improvement was attributed
RARE EARTHS IN STEELS 27
co
"O
< c~
0.1
I
upper critical stress
"\11
I X < fracture time incubation i ~
time - - > 1 ~
\ \ iower criticat stress
*
I r I i i r i
1 10 100
Fracture time / hr Fig. 17. Schematic illustration of delayed failure behavior.
to a change in sulfide morphology. An additional important consideration is the amount of rare earth additions required to inhibit hydrogen embrittlement. It is well established that embrittlement results from hydrogen contents as low as 10 ppm (Vennett and Ansell 1969). Enough rare earth must be in the alloy system to provide improvement in resistance to hydrogen embrittlement without causing a significant alteration in the mechanical properties of the base alloy itself. Fabricability and workability of alloy systems containing these rare earths are also important factors in determining their potential use.
Ce additions at the 0.2 w/o level in 4340 type steels have resulted in lower susceptibility to the blister or flake formation type of hydrogen embrittlement by forming stable hydrides below 1010°C (Kortovich 1977). Kortovich (1977) investigated the hydrogen embrittlement cracking resistance of vacuum-induction-melted AIS14340 steel containing Ce or La in the 0.03-0.17 w/o range. As a basis for studying the hydrogen embrittlement resistance of rare-earth-modified steel, delayed failure tests were conducted on specimens cathodically charged in sulfuric acid and plated with cadmium. Delayed failure tests, which employ a series of varying static loads, represent the most sensitive method for studying hydrogen embrittlement. The essential characteristics of classical delayed failure are summarized schematically in fig. 17 (Troiano 1960). The most significant characteristic of delayed failure behavior is the fact that there is a minimum critical value of stress (the lower critical stress) below which failure does not occur. Studies performed on hydrogen-induced delayed failure of sharply notched high strength steel specimens indicate that an incubation time precedes crack initiation. Once a critical amount of hydrogen has reached the area in front of the crack tip, cracking proceeds discontinuously until a critical length is attained and rapid failure occurs (Johnson et al. 1958, Steigerwald et al. 1959).
The delayed failure curves for precracked specimens of hydrogenated and Cd plated (a) AISI 4340 steel without rare earths, (b) with 0.03% Ce, (c) with 0.09 and 0.17 w/o Ce, (d) with 0.08 and 0.16w/o La are shown in fig. 18a-d. The Ce and La additions showed a dramatic improvement in the delayed failure of 4340 steel both in fracture times and lower critical stress intensity. The lower critical stress intensity represents a three-fold
28 H. NAGAI
1 0 0
rl
.2 c -
• g 5 c e'l
" 0 .ll,l
<
(a)
"T I '
4340 steel
o C r a c k .___> ,,
initiation ~O •
o ~ . . . _o__',o._.
Lower critical no fracture
=24.8 MPa
i I ~ I , i
10 100 1000 10000
Time / rain
100 al O_
, J .=_
q~ 50
.~_
< e-J
Co)
0.03% Ce
• " ' o ~ o
\
C r a c k ) ', initiation ~,
1
i
. . . ~ : _ _ _ o e e ~ • . • L o w e r critical=2OMPa
1 O- 1 O0 1000 10000
Time / min
100
Q .
y:
69 50 P
.e
13.
< 0
10
(c)
I " ' 1 - , , - - - ' T "
. n _ ~ 2_ - - o l 3 : t - ~ o : i = - ' - - " - - - - m " " ~ n "
L.C,=t~[.5 ~
• . , - ~ q ~ • x . ~ Crack
o,x - ~ initiation
C r a c k > " o . . e . ~
initiation " o \ •
- o - • - - O.09%Ce ",
- m - n - - 0,17%Ce ~ 0 I
L.c.=23.1 MPa - -
I i
1oo moo 10000
Time / min
10(1
O.
t- 50 ,& £ .~_
(d)
t
L.c.=77.6 0.16% La
"°x~ ~ ~crack
no fractureCrack ""~ ~° 2 / i n i t i a t i o n
initiation ~1, 0.08%
. . . 2_o.. ` . . . . oo_8._"2"-- .. . .
L,G.=25.gMPa no fracture
lo loo looo 10o00
Time / min
Fig. 18. Effect of rare earth additions on the delayed failure behavior of hydrogenated 4340 steel: (a) without rare earth, (b) with 0.03% Ce, (c) with 0.09 and 0.17% Ce, (d) with 0.08 and 0.16% La.
improvement over the non-rare-earth baseline material• Crack initiation times were also much higher in materials with a higher Ce or La content. More important, however, was the shape of the delayed failure curve itself. The standard type curve exhibited by the baseline in fig. 18a was characterized by a rather sudden decrease from the upper critical to the lower critical stress intensity over a fairly constant range of failure times. The significance of this behavior lies in the fact that design engineers must insure that service conditions do not exceed the lower critical stress intensity. If this value is exceeded, delayed failure from hydrogen embrittlement can be expected. For the high Ce and La content materials, however, the difference between the upper and lower critical values was substantially smaller than for the baseline, and the decrease was much more gradual.
These results suggest a significant improvement in resistance to hydrogen embrittlement compared to the baseline 4340. In particular, the high Ce and La content materials can be used to a substantially higher percentage of its upper critical stress intensity without danger of delayed failure from hydrogen embrittlement. This is particularly significant for high strength steel components such as landing gears, cables and so on. Their production includes such operations as acid pickling cleaning and electroplating which can introduce hydrogen into the component•
RARE EARTHS IN STEELS 29 Crack growth kinetics measured with the use of compliance gages which recorded crack opening displacements during each of the delayed failure tests revealed that the crack growth rate in the baseline 4340 and low rare earth steels was an order of magnitude faster than that exhibited by the high rare earth steels. As a first approximation, these results indicate that the safe operating life for high rare earth steel components would be considerably greater than for the baseline or low rare earth steels. This crack growth behavior can be explained on the basis of hydrogen diffusing to the crack tip, resulting in discontinuous crack growth. In the baseline material and the low rare earth steel, the transport of hydrogen was not seriously affected and thus crack propagation was not retarded.
The delayed failure results clearly indicated that a substantial improvement could be obtained in the hydrogen embrittlement resistance of 4340 steel through additions of Cc and La. This improvement was manifested by longer times to crack initiation (incubation time), longer failure times and higher values of lower critical stress intensity.
Maximum improvement, however, was obtained only at the high rare earth levels (e.g.
0.16-0.17 w/o). The microstructures for these rare-earth-modified steels exhibited almost continuous grain boundary inclusion formation along prior austenite boundaries. In the high rare earth content material, the continuous grain boundary inclusions acted to entrap hydrogen, inhibiting its transport to the crack tip. Crack propagation was subsequently retarded. These observations suggest a possible mechanism to rationalize the enhanced resistance to hydrogen embrittlement exhibited by rare-earth-modified steel.
Room temperature tensile results for both uncharged and hydrogen charged material were comparative and indicated little difference between the elements. A maximum occurred in the ultimate and 0.2% yield strengths at approximately 0.1 w/o for both rare earth additions. This strengthening behavior was attributed to the deoxidizing and desulfurizing action of the rare earths.
The results shown above suggest that the ability of rare earth elements to getter or otherwise entrap hydrogen was responsible for the improved resistance to hydrogen embrittlement. That maximum improvement was obtained only at the higher rare earth levels can be rationalized by the fact that the larger amounts of Ce and La were required to effectively delay the critical value of hydrogen concentration from being reached at the crack tip. At lower rare earth levels, it was apparent that the critical hydrogen concentration was achieved in spite of the fact that certain amounts were probably entrapped en route to the crack tip.
Delayed failure test showed that, at higher rare earth levels, the threshold stress intensity (i.e., the stress intensity level below which failure did not occur) increased by a factor of about four, and the crack growth rate decreased by about an order of magnitude compared with 4340 steel without rare earth additions. This improvement was attributed to the ability of the rare earth elements to interact with hydrogen, thereby reducing the supply of hydrogen available for embrittlement and impeding the diffusion of hydrogen to the crack tip where it would accumulate and cause crack growth by local embrittlement.
The room temperature elongation, reduction of area and Charpy impact energy all, however, decreased with the addition of both Ce and La. The ductility and impact losses
30 14, NAGAI
in these steels were attributed to the formation of massive and continuous grain boundary inclusions which offered ideal paths for crack propagation.
4.2. Stress corrosion cracking
Hydrogen embrittlement has also been observed in steel structural components exposed to aqueous environments (Sheinker and Wood 1971). Termed stress corrosion cracking, this natural process can result in the failure of a component from the combined action of stress and chemical attack. It is now fairly well established that stress corrosion cracking of steels in aqueous solutions is governed, at least to some extent, by a series of electrochemical reactions at the surface which permit the entry of hydrogen into the metal (Parkins 1964).
Sheinker (1978) extended the concept of rare earth additions to high strength steels to stress corrosion cracking behavior, because this type of failure in high strength steels is believed to be a form of hydrogen embrittlement. The resistance to stress corrosion cracking for 4340 steels containing 0, 0.20 and 0.30w/o Ce in 3.5% sodium chloride solution at room temperature was evaluated. The Ce addition had a much smaller effect on the stress corrosion cracking resistance than the Ce and La additions had on the hydrogen embrittlement cracking resistance described above. The stress corrosion cracking threshold (K~scc) was about the same for all three steels. The higher Ce (0.30%) material, however, had longer failure times and lower average crack growth rates than the lower Ce (0.20%) material. It was found that the failure times for non-Ce steel would be shorter and the average crack rates higher than those for the lower Ce steel. The difference between the effects of the rare earth additions on stress corrosion cracking and hydrogen embrittlement cracking was attributed to the difference in the source of hydrogen in the two cracking phenomena, which affects the amounts of hydrogen available for embrittlement and the processes of hydrogen transport to the tip of the crack.
4.3. Hydrogen permeability in steels
Sheinker (1978) conducted permeability measurements to determine whether hydrogen transport through the steel was affected by the presence of the rare earth additions, which increased the resistance to hydrogen embrittlement as well as stress corrosion cracking.
Hydrogen permeability measurements were made on 4340 steels containing 0% and 0.21% Ce. The permeability of hydrogen through membranes of these materials was determined using a cell developed by Devanathan and Stachurski (1962) containing a charging solution of 1N sulfuric acid with 20ppm arsenic added to promote hydrogen entry. The steady state hydrogen permeation flux was measured galvanostatically (Chatterjee et al. 1978).
The half-time to reach the steady state hydrogen permeation flux, which corresponds to the apparent hydrogen diffusivity in the metal, was four times longer in the Ce-bearing steel than in the non-Ce steel in the 480 K temper condition and 2.5 times longer in the Ce-bearing steel than in the non-Ce steel tempered at 670 K. These results indicate that
RARE EARTHS IN STEELS 31 the apparent hydrogen diffusivity is lower in the steel containing Ce at both tempering temperatures. However, because hydrogen permeation transients are affected by hydrogen trapping and surface reactions (Chatterjee et al. 1978, Oriani 1970), it cannot be deduced whether the Ce reduced the true (lattice) hydrogen diffusivity. Ce compounds in the steel could be potent traps for hydrogen because the rare earth elements are known to combine readily with hydrogen (Gschneidner 1961). In addition, because most of the Ce in the steel was present as oxide inclusions and solid-solid interfaces are believed to be important sites for hydrogen trapping (Oriani 1970), the inclusion-matrix interface may be a potent hydrogen trap in Ce-bearing 4340 steel. Although the surfaces of the Ce-bearing and non-Ce steel membranes were identically prepared, the electrochemical reactions at the surface of the former could be affected by the presence of Ce. Thus, hydrogen trapping and changes in surface chemistry may have been responsible for reducing the apparent hydrogen diffusivity of 4340 steel when Ce was added to this alloy.
Their results also show that the steady state hydrogen permeation flux was three to four times lower in the Ce-bearing steel than in the non-Ce steel at both tempering temperatures. Because permeability is the product of solubility and diffusivity, the effect of Ce on the steady state permeation flux could be due to its effect on either or both the hydrogen solubility and the hydrogen diffusivity in the steel. The separate effects of Ce on the solubility and the diffusivity of hydrogen in steels are not presently known, but in view of the strong affinity of the rare earth elements for hydrogen, the solubility would be expected to be increased and the diffusivity decreased by the addition of Ce to steels. Thus, the reduced hydrogen permeability in the Ce-bearing 4340 steel is probably associated with a lower hydrogen diffusivity. Hydrogen permeability has been reported to decrease slightly and hydrogen diffusivity to remain constant for a similar high strength steel when the tempering temperature was increased from 480 to 670 K (Radhakrishnan and Schreir 1967). In general, it can be concluded that the presence of Ce retarded the permeability of hydrogen through 4340 steel at both tempering temperatures. This would reduce the rate of flow of hydrogen from the bulk of the metal to the tip of the crack.
Therefore, the presence of Ce would be expected to decrease the rate of hydrogen-induced crack growth and increase the delayed failure time in a high strength steel.
4.4. Rare-earth-modified powder metallurgy steel
As described above, it has been demonstrated that the resistance of AISI 4340 steel wrought plate to hydrogen embrittlement could be substantially improved by La and Ce additions of approximately 0.2 w/o (Kortovich 1977). The main limitation, however, in these rare-earth-modified high strength steels was the degradation of mechanical properties due to the presence of larger rare earth oxide inclusions in the microstructure.
Additions of 0.16/0.17w/o of La or Ce were desired for resistance to hydrogen embrittlement, but such high levels were undesirable due to reduced Charpy impact strength. Clearly, these studies showed that a homogeneous distribution of the rare earths and a fine size of the dispersoids are desired to optimize the improvement in resistance to hydrogen embrittlement without degradation of mechanical properties.
32 H. NAGAI
The powder metallurgy approach offers a means of obtaining a more uniform distribution of the rare earth elements in the steel and minimizing the formation of large rare earth oxide inclusions. In order to minimize the problems associated with a non- uniform distribution of rare earth compounds so that the benefit of rare earths could be optimized, Sheinker and Ferguson (1982) produced rare-earth-metal-treated 4340 steel bars from hydrogen-gas-atomized powder and from attrited powder by both HIP and hot extrusion. A master alloy (75%Ce-25%Ni) was added prior to atomization or during attrition. Oxygen levels of both the atomized powders and the attrited powders were higher than the desired maximum limit of 300 ppm.
The steel powders were consolidated by hot extrusion and by hot isostatic press- ing (HIP), and the tensile properties and toughnesses were determined. These properties were compared with C/W properties for both 4340 baseline and rare-earth-treated 4340 steels.
The mechanical properties of the HIP consolidated powders were unacceptable, with ductility and toughness being extremely poor because of fracture along prior particle boundaries (PPB). The high oxygen levels and the low amount of deformation involved in HIP are the root of this property problem because PPB cracking provides easy crack paths.
Extruded bars for both atomized and attrited 4340 + rare earth powders have signifi- cantly superior properties in comparison to HIPed bars. Extrusion, which involves metal flow and particle deformation, effectively breaks up PPB films to enhance metallurgical bonding between particles. The result is a ductility and toughness improvement for both atomized and attrited powders. Strength levels higher than those of C/M plates are achieved, but the ductility, although improved, still falls below that of C/W values.
Extruded bars show a beneficial effect of metal flow during particle deformation.
Ductility and toughness are improved as inclusions and embrittling films are broken up and strung out so that metallurgical bonding between powder particles is enhanced.
It has been demonstrated that rare-earth-treated steel can be produced by P/M tech- niques, but that the final oxygen level of the powder should be lowered, because rare earths are extremely reactive with oxygen, and the P/M process permits a high surface area/volume ratio to exist during processing.
5. Effect o f rare earth addition on creep b e h a v i o r
In the previous section, it was demonstrated that rare earth additions show excellent improving effects on the toughness, bend formability, and ductility of steels mainly by sulfide shape control and by elimination of impurities, such as S, P and hydrogen (Eyring 1964, Anderson and Spreadborough 1967). However, there is no obvious evidence that the tensile properties of steels, such as yield strength, ultimate tensile strength, and elongation and reduction of area, are significantly improved by rare earth additions. The small variation in these properties is considered to be attributed to minor processing and chemistry variations and not to rare earth additions. However, it has been reported by