Residual, Diffusible and Total Hydrogen Concentration and Embrittlement After Baking in Cd-Plated AerMet 100.

Transcription

1 Residual, Diffusible and Total Hydrogen Concentration and Embrittlement After Baking in Cd-Plated AerMet 100. Dogan, H., Gangloff, R.P., and Scully, J.R. Department of Material Science and Engineering University of Virginia 116 Engineer s way Charlottesville, VA AerMet 100 alloy is a secondary hardening martensitic steel that combines high hardness and strength with exceptional ductility and toughness 1. This modern steel is used in ultra-high strength applications and must be protected against corrosion. This is done by electroplating a 5-25 µm thick protective Cd-layer on the steel surface. Unfortunately, during the electroplating process, hydrogen co-deposits on the steel surface and diffuses into the steel. Even in very low concentrations in the bulk lattice, H can migrate to areas of high triaxial stresses, thereby magnifying its effect and causing time dependent hydrogen embrittlement 2. The AerMet 100 microstructure contains many features that are known to trap H in steel; such as dislocations, carbides, martensite lath interfaces, solutes and prior-γ grain boundaries. Trapping at such features can significantly affect the internal hydrogen embrittlement (IHE) behavior by complex mechanisms. For example, a distribution of strong irreversible H traps may prevent H from partitioning to lower interaction energy sites, such as prior-γ grain boundaries and lath martensite interfaces, that are susceptible to low-stress intensity H-induced fracture and increase the resistance to HE 3,4,5. In contrast, a distribution of weak reversible H traps may provide a reservoir of diffusible (mobile) H that can repartition to and within the fracture process zone to sites susceptible to H-induced cracking, thus worsening the fracture process 6,7. Earlier studies have shown that plane strain elastic-plastic toughness decreased drastically from 140 MPa m to less than 20 MPa m due to hydrogen precharging in alkaline Ca(OH) 2 and slow rate loading (Figure 1). The fracture mode changed from microvoid coalescence for the uncharged case to brittle cleavage-like transgranular fracture for diffusible hydrogen concentrations between 2 to 8 wppm (Figure 2) 8. This H-induced cracking of AerMet 100 at K TH values less than 20 MPa m is unique in comparison to HE in similar strength steels. Low K TH values are typically associated with intergranular fracture, while transgranular H-induced cracking generally accompanies much higher K TH. This is ascribed to a synergistic effect of H and impurities segregated to grain boundaries, decreasing the grain boundary strength 9. The low impurity levels of AerMet 100 have essentially eliminated H-induced intergranular cracking. Only limited areas of prior austenite grain boundary cracking occurred in AerMet 100 for the higher total H concentration levels. Instead severe

2 brittle transgranular fracture due to IHE in AerMet 100 results from stress-induced hydrogen repartitioning to potential fracture sites during loading 8. Current research has shown that the hydrogen embrittlement of H charged, un-plated AerMet 100 may be reversed and the recovery of the original fracture toughness value is possible if baked at elevated temperatures such as 190 C if the enough time is given. Baking greatly effects the diffusible H concentrations of charged specimens. For example; baking the H charged (C H Diff =7.5 wppm) specimens for three days at room temperature reduced the C H Diff, value to 0.34 wppm, by 95 percent, and elevated temperature baking at 200 C for 2.3 hours eliminated the diffusible H in the steel. Despite the great reduction of the lattice and the weakly trapped hydrogen, aprox. 40 percent of the total H remained in the steel which resides in the strong H trap sites 12. In order to find out whether such trapped hydrogen effects IHE or not, K IC was determined for a charged and for a charged/baked specimens. Figure 3 shows the load vs CMOD data for the charged and charged/baked CT specimens. H charged (η chg =-0.3 V [vs. Hg/Hg 2 SO 4 ] and C H Diff =4.0 wppm) specimen exhibited severe hydrogen embrittlement with threshold stress intensity, K TH =15.9 MPa m. Baking the identically charged specimen at 190 C for 24 hours was sufficient to eliminate all the diffusible hydrogen which reversed the HE and restored the fracture toughness of K jic = 123 MPa m which is close to or at the level of the fracture resistance of uncharged AerMet 100. However, these measurements must be repeated for Cd-Plated specimens. Cd-plated AerMet 100, like other ultra-high strength steels, is baked at 190 C for 5 to 24 hours to remove the dissolved hydrogen after electroplating. The effectiveness of this standard post-plating thermal treatment is questionable, particularly for AerMet 100. High hydrogen solubility in Cd, relative to that in Fe, allows the electroplated Cd-layer to act as a H source during the initial stage of baking, and consequently, the H concentration in the steel substrate may increase 10. Furthermore, because H diffusivity in Cd is much slower than that in steel, the Cd-layer act as a diffusion barrier to H loss during baking and significant concentrations of dissolved hydrogen may remain in the steel, even after heating times of 100 hours 11. Moreover the diffusivity in AerMet 100 is slower than in many quenched and tempered steels such as AISI 4130 and the H trap capacity is greater. These uncertainties in the postplating heat treatment process causes IHE to be a concern in Cd-plated AerMet 100 components. The objectives of this research are to quantitatively characterize and understand the effects of hydrogen that is introduced during the cadmium plating of AerMet 100 on the mechanical properties and investigate the effectiveness of post-plating baking treatment on the removal of diffusible and trapped hydrogen content. For these purposes, Barnacle Electrode method and LECO/ASTM 1447 hot extraction method are being used to quantify the diffusible H concentrations and the total H concentrations respectively. Investigation of the retained hydrogen as a function of trap state in the Cd-plated AerMet 100 specimens is being accomplished using the Thermal Desorption Spectroscopy. As a future work, experiments are needed using

3 the elastic-plastic fracture mechanics test method to determine the effect of baking on the fracture toughness of Cd-plated specimens as a function of baking time and temperature and to correlate to the remaining diffusible and trapped hydrogen. ACKNOWLEDGMENTS Financial support from the Office of Naval Research (Grant No. N ) under the direction of contract monitor Dr. A. John Sedriks, is gratefully acknowledged. We also acknowledge contributions of the electrochemical instrumentation by Perkin Elmer Corporation and Scribner Associates Inc. REFERENCES 1. Dahl, J.M, Ferrous-base Aerospace Alloys, Advanced Materials & Processes, May 2000, pp De Lucia, J.J and Berman, D.A., An Electrochemical Technique to measure Diffusible Hydrogen in Metals (Barnacle Electrode), Electrochemical Corrosion Testing ASTM STP 727, Mansfeld, F. and Bertocci, U., editors, ASTM, Philadelphia, PA, 1981, pp Pressouyre, G.M., Current Solutions to Hydrogen Problem in Steels, Interante, C.G. and Pressouyre, G.M., editors, American Society for Materials International, Materials Park, OH, 1982, pp Pressouyre, G.M., Trap Theory of Hydrogen Embrittlement: Experimental Investigations, in Hydrogen Effects in Metals, Bernstein, I.M. and Thompson, A.W., editors, The Minerals, Metals, and Materials Society, Warrendale PA, 1981, pp Scully, J.R., Van Den Avyre, J.A., Cieslak, M.J., Romig, A.D., and Hills, C.R., The Influence of Palladium on the Hydrogen-Assisted Cracking Resistance of PH 13-8 Mo Stainless Steel, Metallurgical Transactions A, Vol. 22A, 1991, pp Yamaguchi, Y., Nonaka, H. and Yamakawa, K., Effect of Hydrogen Content on Threshold Stress Intensity Factor in Carbon Steel in Hydrogen-Assisted Cracking Environments, Corrosion, Vol. 53, 1997, pp Yamakawa, K., Yonezawa, S. and Yoshizawa, S., Critical Hydrogen Content of Hydrogen Embrittlement of High-Strength Steel, in International Congress on Metallic Corrosion, National Research Counsil, Toronto, Canada, 1984, pp Thomas, Richard L.S., Internal Hydrogen Embrittlement of a Trap-Rich Ultrahigh-Strength Steel, AerMet 100, A Master Thesis Presented to School of Engineering and Applied Science, University of Virginia, May Yoshino, K. and Mc Mahon Jr. C.J., Cooperative Relation Between Temper Embrittlement and Hydrogen Embrittlement in a High Strength Steel, Metallurgical Transactions, Vol. 5, 1974, pp Berman, D.A., The Effect of Baking and Stress on the Hydrogen Content of Cadmium Plated High Strength Steels, Materials Performance, Vol 24, 1985, pp Boodey, J.B. and Agarwala, V.S., Paper 224: Hydrogen in Metals: Cadmium Plated Steels, in Corrosion 87, NACE, Houston, TX, Thomas, R.L.S., Li, Daoming., Gangloff, R.P., and Scully, R.J., Trap-Governed Hydrogen Diffusivity and Uptake Capacity in Ultrahigh-Strength AerMet 100 Steel, Metallurgical and Materials Transactions A, Vol. 33A, 2002, pp. 1-14

4 K TH (M P a m ) K JIC AerMet 100 T = 25 C C H,diff = 0 wppm dk/dt = 0.2 MPa m/s C H,diff > 0 wppm dk/dt = 2.2 x 10-4 MPa m/s Diffu sible H Concentration (w ppm ) (a) K TH (o r K JIC ) MPa m K JIC AerMet 100 T = 25 C C H,tot = 0 wppm dk/dt = 0.2 MPa m/s C H,tot > 0 wppm dk/dt = 2.2 x 10-4 MPa m/s TotalH Concentration (w p p m ) (b) Figure 1. Threshold stress intensity for hydrogen embrittlement, K TH, at various (a) diffusible H concentrations and (b) total H concentrations. A significant decrease in K TH was observed at all H concentrations dissolved in peak aged AerMet 100.

5 10 µm (a) (b) 10µm 10µm Figure 2. Scanning electron images of the fracture surfaces of AerMet 100 specimens with a diffusible H concentration of (a) 0 wppm and (b) 7.8 wppm. The fracture surface is characterized by ductile-microvoid features in the absence of H (a) and by brittle transgranular features with precharged H (b). Crack growth in both images was from top to bottom.

6 10 8 Load (kn) AerMet 100 H Charged dk/dt = 3.33 x 10-2 MPam 1/2 /s H Charged Baked at 190 o C for 24 hours dk/dt = 4.85 x 10-2 MPam 1/2 /s CMOD (mm) Figure 3. Load versus CMOD data for the charged and charged/baked specimens. Charging was done in saturated Ca(OH) 2 at the charging overpotential of 0.30 V at 60 C for 20 days.