Nickel superalloys - their welding and deposition
TWI Bulletin, March/April 2006
Alloy 718, Waspaloy and single crystal CMSX-10 come in for close scrutiny
Andrea Gregori is currently a principal project leader in Metallurgy and Corrosion at TWI. His work includes management of projects related to weldability, metallurgy and corrosion of steels, stainless steels and nickel alloys and includes research, testing, and failure investigations primarily for the oil and gas and power generation sectors. His experience embraces provision of rapid response to enquiries and consultancy requirements of TWI Industrial Members world-wide, leading multi-disciplinary projects and providing input on the influence of welding and joining technology on the mechanical and corrosion behaviours of materials.
Italian born Dan Bertaso holds qualifications from Universitá degli studi di Bologna where he specialised in oil and gas and mining engineering and also from the Galileo Ferraris in Verona. He gained an MSc in Marine Technology from Cranfield University (UK). His experience includes management of projects involving the development and application of arc welding processes technology. His expertise embraces arc welding and automation for use in a wide range of applications. Before joining TWI in 2001, Dan gained experience in the development of robotic welding procedures for large tubular structures.
A new annual award for the most worthy Industrial Members' report from the Core Research Programme has been won by Andrea Gregori and Dan Bertaso. The award is named after Richard Weck, a former Director General of BWRA and The Welding Institute and it replaces the previous Richard Weck Lecture.
Nickel superalloys are highly susceptible to weld and heat-affected zone (HAZ) cracking during welding and postweld heat treatment (PWHT). However, substantial economic benefit would result from developing successful weld repair techniques and hence there is considerable interest in establishing sound technology for applications in aeroengines and industrial gas turbines.
Composition and ageing condition, together with thermal strain and stress developed during welding and PWHT are the main factors controlling cracking. The cracking mechanisms observed during welding and repair of Ni-superalloys include solidification cracking and HAZ liquation cracking (forming microfissures), the latter being of particular concern because it is difficult to detect using non-destructive examination techniques.
High preheat temperatures often exceeding 900°C have been used prior to this work to control strain and avoid cracking during welding of these alloys. However, this technique presents several practical difficulties and a different approach using low heat input processes and controlling the thermal cycle has been investigated in the work described here. The weldability and repairability by TIG welding and electro-spark deposition (ESD) of polycrystalline Alloy 718 and Waspaloy, and single crystal nickel superalloy CMSX-10 was studied and the effect of PWHT on Waspaloy and Alloy 718 deposits was also examined.
Objectives
- To apply procedures for minimising cracking during welding and repair of Alloy 718 and Waspaloy.
- To investigate the applicability of TIG and ESD processes for the repair of single crystal CMSX-10 superalloy blades.
Experimental approach
Arc welding techniques including pulsed TIG, variable-polarity (VP) square-wave TIG and electro-spark deposition were investigated for welding Alloy 718, Waspaloy and CMSX-10. Variable-polarity square-wave TIG is a type of AC TIG welding in which the waveform is square and asymmetrical compared with standard sinusoidal symmetrical AC TIG. Electrospark deposition is a pulsed micro-welding process used for small-scale repair and coating in which short-duration high-current pulses pass through a consumable electrode in contact with the substrate ( Fig.1). Alloy 718, Waspaloy and Mar-M247 consumables were used and a summary of the parent materials and consumables investigated, including their chemical composition and microstructural characteristics, is reported in Tables 1-2. Forced cooling was applied during welding to control the thermal field and details are provided in TWI Members Report 811/2004.
Table 1 Parent metals and consumables
| Material (form) | Sample size | Heat treatment | Hardness HV0.5 [max-min /mean] | ASTM grain size no. | γ' size [µm] | γ' volume [%] |
| Alloy 718 (plate) | 300x40x2.3mm | Solution heat treated | 256-228/241 | 8 | - | - |
| Waspaloy (disc) | ø500mm, 6mm thick | Solution heat treated and double-aged | 424-414/419 | 4-5 | - | - |
| CMSX-10 (blade) | - | Solution heat treated and primary aged | 430-408/421 | - | 0.41 | 64 |
| Alloy 718 (wire) | ø0.8mm | - | - | - | - | - |
| Waspaloy (wire) | ø1.2mm | - | - | - | - | - |
| Mar-M247 (rod) | ø2.4mm | - | - | - | - | - |
Table 2 Chemical composition of the nickel superalloys
| Material (form) | Element [wt%] |
| C | Si | Mn | P | S | Al |
| Waspaloy (disc) | 0.068 | 0.02 | 0.02 | <0.005 | 0.001 | 1.51 |
| Alloy 718 (plate) | 0.04 | 0.08 | 0.22 | 0.007 | <0.001 | 0.58 |
| CMSX-10 (blade) | 0.010 | <0.001 | <0.01 | - | <0.001 | 5.78 |
| Mar-M247 (rod) | 0.15 | 0.09 | 0.02 | 0.010 | <0.001 | 5.67 |
| Alloy 718 (wire) | 0.053 | 0.09 | - | 0.006 | <0.001 | 0.54 |
| Waspaloy (wire) | 0.032 | 0.01 | 0.02 | <0.005 | <0.001 | 1.30 |
| Waspaloy, UNS N07001* | 0.03-0.10 | 0.75 | 1.0 | 0.030 | 0.030 | 1.20-1.60 |
| Alloy 718, UNS N07718* | 0.08 | 0.35 | 0.35 | 0.015 | 0.015 | 0.20-0.80 |
| Mar-M247** | 0.13-0.17 | 0.15 | 0.20 | 0.015 | 0.010 | 5.30-5.70 |
| CMSX-10*** | 0.04 | 0.05 | 0.04 | - | 0.001 | 5.0-7.0 |
| Material (form) | Element [wt%] |
| B | Co | Cr | Cu | Fe |
| Waspaloy (disc) | 0.0055 | 13.5 | 18.8 | <0.01 | 0.64 |
| Alloy 718 (plate) | 0.002 | 0.15 | 17.9 | - | Balance |
| CMSX-10 (blade) | 0.006 | 3.01 | 1.72 | <0.01 | <0.05 |
| Mar-M247 (rod) | 0.013 | 9.95 | 8.33 | - | 0.24 |
| Alloy 718 (wire) | 0.0020 | - | 17.9 | 0.04 | 19.1 |
| Waspaloy (wire) | 0.0049 | 14.1 | 19.1 | 0.02 | 0.93 |
| Waspaloy, UNS N07001* | 0.003-0.010 | 12.0-15.0 | 18.00-21.00 | 0.50 | 2.0 |
| Alloy 718, UNS N07718* | 0.006 | 1.0 | 17.0-21.0 | 0.30 | Balance |
| Mar-M247** | 0.010-0.020 | 9.00-11.00 | 8.00-8.80 | 0.10 | 0.25 |
| CMSX-10*** | 0.01 | 1.5-9.0 | 1.8-4.0 | - | - |
| - not specified/not analysed. (1) Nb+Ta range from 4.75-5.50%wt. * maximum or range chemical composition as specified by ASTM B637-98. ** maximum or range chemical composition as specified by US Patent 3,720,509. *** maximum or range chemical composition as specified by US Patent 5,366,695. |
| Material (form) | Element [wt%] |
| O | N | Mo | Nb | Ni | Ti |
| Waspaloy (disc) | - | - | 4.24 | 0.04 | Balance | 3.03 |
| Alloy 718 (plate) | <0.001 | 0.008 | 3.12 | 5.11 | 53.6 | 1.04 |
| CMSX-10 (blade) | 0.001 | <0.001 | 0.40 | 0.10 | 68.1 | 0.24 |
| Mar-M247 (rod) | 0.011 | 0.0060 | 0.78 | 0.03 | Balance | 1.14 |
| Alloy 718 (wire) | 0.006 | 0.007 | 2.93 | 5.09 | 53.1 | 1.02 |
| Waspaloy (wire) | 0.0042 | 0.0043 | 4.06 | 0.05 | Balance | 2.99 |
| Waspaloy, UNS N07001* | - | - | 3.50-5.00 | - | Balance | 2.75-3.25 |
| Alloy 718, UNS N07718* | - | - | 2.8-3.3 | (1) | 50.0-55.0 | 0.65-1.15 |
| Mar-M247** | - | - | 0.50-0.80 | 0.10 | Balance | 0.90-1.20 |
| CMSX-10*** | - | - | 0.25-2.0 | 0.5 | Balance | 0.1-1.2 |
| Material (form) | Element [wt%] |
| Re | W | Zr | Hf | Ta |
| Waspaloy (disc) | - | <0.05 | <0.01 | - | <0.05 |
| Alloy 718 (plate) | - | - | - | <0.01 | <0.01 |
| CMSX-10 (blade) | 6.15 | 5.65 | 0.006 | 0.02 | 8.13 |
| Mar-M247 (rod) | - | 9.85 | 0.072 | 1.39 | 3.09 |
| Alloy 718 (wire) | - | - | - | - | - |
| Waspaloy (wire) | 0.01 | 0.21 | 0.062 | - | <0.05 |
| Waspaloy, UNS N07001* | - | - | 0.02-0.12 | - | - |
| Alloy 718, UNS N07718* | - | - | - | - | (1) |
| Mar-M247** | - | 9.50-10.50 | 0.03-0.08 | 1.20-1.60 | 2.80-3.30 |
| CMSX-10*** | 5.0-7.0 | 3.5-7.5 | 0.01 | 0.15 | 7.0-10.0 |
| - not specified/not analysed. (1) Nb+Ta range from 4.75-5.50%wt. * maximum or range chemical composition as specified by ASTM B637-98. ** maximum or range chemical composition as specified by US Patent 3,720,509. *** maximum or range chemical composition as specified by US Patent 5,366,695. |
Welds were examined by fluorescent penetrant inspection to locate any surface breaking cracks. Metallographic examination using light and scanning electron microscopy was performed to identify any fine cracks in the weld metal and heat affected zone and the maximum crack length was measured. Hardness and grain size measurements were also performed. The effect of postweld heat treatment on the hardness of Waspaloy and Alloy 718 deposits on Waspaloy substrate was investigated.
Results and discussion
Crack-free multiple pass deposits on Waspaloy were obtained using pulsed TIG and an arc energy below 0.3kJ/mm ( Fig.2). Thermal strain and stress developed during welding are among the main factors controlling cracking in nickel superalloys and minimising their level is beneficial to subsequent cracking resistance. Investigations of the weldability of nickel superalloys containing high γ'-phase volume fraction ( Fig.3) indicated that crack formation can also be favoured by γ' precipitation during cooling from the γ'-solvus temperature. It appeared that, for the particular welding conditions considered in this study, the application of forced cooling resulted in the reduction and elimination of HAZ microfissures ( Table 3).
Fig.2. Light photomicrographs showing a crack-free six-pass build-up produced on a Waspaloy disc by pulsed TIG using Waspaloy filler with final PWHT at 1150°C for 4h followed by double ageing at 850°C for 4h and 760°C for 16h
Fig.3. Scanning electron photomicrograph showing fine cubic γ' phase with 0.41µm edge size in single crystal alloy CMSX-10
Table 3 Results of the examination of weld beads deposited on Waspaloy
| Filler/description | Forced cooling | Current type | Arc energy [kJ/mm] | Hardness, HV0.5 [max-min/mean] | HAZ grain size ASTM# | Microfissure length [µm]+ |
| HAZ | Weld |
| Waspaloy/single pass | ✓ | DC | 0.3 | - | - | - | 0 |
| Waspaloy/single pass | - | VP | 0.3 | - | - | - | 0 |
| Waspaloy/overlay* | - | VP | 0.3 | 348-282/314 | 323-281/290 | 4-5 | 27 |
| Waspaloy/overlay* | ✓ | VP | 0.3 | 331-219/272 | 311-268/289 | 4-5 | 0 |
| Waspaloy/overlay* | - | DC | 0.3 | 354-245/284 | 299-267/279 | 6 | 0 |
| Waspaloy/overlay* | ✓ | DC | 0.3 | 333-256/292 | 278-258/266 | - | 0 |
| Waspaloy/build-up | - | VP | 0.3 | - | - | - | 17 |
| Waspaloy/build-up | ✓ | VP | 0.3 | 367-331/349 | 298-261/277 (20.5-17.9/29.5) | - | 0 |
| Waspaloy/build-up | ✓ | DC | 0.3 | 381-308/345 | 291-259/270 | - | 0 |
| Waspaloy/overlay* (PWHT) (1) | ✓ | DC | 0.3 | 432-408/420 | 440-413/428 | - | 22 |
| Waspaloy/build-up (PWHT) (1) | ✓ | VP | 0.3 | 440-429/435 | 432-405/416 | - | 0 |
| Waspaloy/build-up (PWHT) (1) | ✓ | DC | 0.3 | 415-413/414 | 446-405/424 | - | 0 |
| Alloy 718/build-up (PWHT) (1) | - | DC | 0.3 | 423-421/422 | 408-363/384 | - | 0 |
| -not applied, measured or analysed. *single layer overlay with 20-30% overlap between passes. +maximum HAZ microfissure length measured by light microscope on a transverse metallographic section at 1000X magnification. (1) PWHT samples: solution annealing at 1015°C±20°C for 4h followed by oil quenching, double-age at 850°C±10°C for 4h and 760°C±10°C for 16h followed by air cooling. Parent material hardness (max-min/mean) after PWHT was 455-423/435 HV0.5 and (43.0-41.9/42.4 HRC). |
The elimination of HAZ cracking was attributed to two factors. First, appropriate forced cooling is able to alleviate thermal strains, distortion and stresses by modifying the thermal field and peak temperatures in the weldment. Moreover, the application of rapid cooling may further reduce cracking by limiting γ' precipitation and associated shrinkage. No difference was observed in this investigation between standard pulsed TIG and VP TIG. Indeed, HAZ microfissures and liquation of grain boundaries were identified in weld deposits obtained by the VP TIG process.
The effect of forced cooling on distortion and stress field was highlighted during welding of Alloy 718 plates where substantial reduction of distortion and buckling was obtained by applying forced cooling ( Fig.4). No HAZ microfissures or weld metal cracking were found and butt welds in Alloy 718, made with and without forced cooling showed no substantial microstructural differences in either the as-welded or PWHT condition.
Fig.4. The effect of forced cooling on deformation of Alloy 718 plates during TIG welding
Third-generation single crystal alloy CMSX-10 is characterised by a rhenium content above 5% and a high concentration of refractory metals (W, Ta) providing high temperature stability and creep-rupture strength, but it is extremely susceptible to cracking during fusion welding. The manually applied ESD deposits on CMSX-10 ( Fig.5) showed no cracking or microfissures in the HAZ. It is important to note that no recrystallisation occurred in the HAZs during ESD deposition. As the single crystal alloys do not contain grain boundary strengtheners, when recrystallisation occurs the newly formed grain boundaries are very weak and susceptible to cracking. Indeed, it was observed that welds produced without preheat by conventional TIG resulted in HAZ and weld metal cracks, and that cracking of the HAZ was associated with recrystallised regions ( Fig.6). Therefore, although improved shielding and preparation of the blade is required to reduce the level of trapped oxides, ESD is of particular interest for localised repair without preheating and coating of small areas on single crystal blades.
Fig.5. Light photomicrograph showing an ESD deposit produced on a CMSX-10 blade airfoil. Manual ESD technique and a Mar-M247 consumable were employed
Fig.6. Light photomicrograph of a CMSX-10 blade repair achieved using conventional pulsed TIG and Mar-M247 consumable, showing the presence of microfissures associted with re-crystallisation in the HAZ
Practical implications
The application of low heat input, ie arc energy less than 0.3kJ/mm, was found to give a low level of very small HAZ microfissures in TIG welds when welding onto aged Waspaloy. Hence, similar or lower heat input levels should be used in practice, to eliminate or minimise cracking during TIG deposition on aged Waspaloy using either a Waspaloy or Alloy 718 consumable.
No significant difference between standard pulsed TIG and variable-polarity square-wave TIG was observed in this study and hence either process may be used for welding onto aged material. Only occasional small HAZ microfissures were found in the present work and further optimisation of the processes is required to see if microfissures can be reliably avoided. Where applicable, forced cooling should be considered to reduce or eliminate the incidence of HAZ microfissuring in multipass deposits.
Forced cooling may also be used to alleviate thermal stresses, and limit distortion and buckling during welding of annealed Alloy 718 plates and potentially may be used for other similar grades in either the annealed or aged condition.
For shroud sealing of single crystal blades, the use of ESD shows promise, as metal can be deposited with no preheat and without HAZ recrystallisation or cracking. However, the deposited metal contained trapped oxide (requiring improved shielding) and this process is slow compared with arc/power beam welding processes. Pulsed micro TIG welding processes have been recently developed with the ability to provide a stable arc at arc energies below 0.3kJ/mm and a higher energy density compared with conventional TIG. The very low heat input to the parent single crystal material when using micro TIG may allow successful repair without preheat providing increased productivity and quality over ESD and ensuring at the same time that no recrystallisation occurs in the HAZ.
Main conclusions
Application of forced cooling during multipass TIG deposition at a heat input of 0.3kJ/mm on aged Waspaloy using matching composition and Alloy 718 consumables resulted in the reduction or elimination of HAZ microfissures compared to standard pulsed TIG and variable-polarity square-wave TIG. The positive effect of forced cooling may be attributed to the reduction of thermal strain, distortion and stresses through modification of the thermal field in the weldment and limitation of γ' precipitation and associated shrinkage. Near-parent matching hardness was achieved in the HAZ and fusion zone after a solution anneal and double age PWHT.
The application of forced cooling during welding of annealed Alloy 718 resulted in a substantial reduction of distortion and buckling and no HAZ microfissures or weld metal cracking were identified. No significant microstructural differences were observed compared to standard TIG welds in the as-welded and PWHT conditions.
CMSX-10 showed a high susceptibility to weld metal and HAZ cracking during conventional TIG welding. HAZ cracking was associated with recrystallisation. No cracking or microfissures were found in the HAZ of ESD deposits on single crystal CMSX-10 alloy. No recrystallisation occurred in the HAZ associated with Mar-M247 deposits produced by ESD.
Recommendations
The application of low heat input, ie arc energy below 0.3kJ/mm, should be used to eliminate cracking during TIG deposition on aged Waspaloy using a Waspaloy or Alloy 718 consumable. Standard TIG or variable-polarity square-wave TIG may be used, although the latter gave small HAZ microfissures. Forced cooling may be used to reduce or eliminate HAZ cracking in multipass deposits.
To alleviate residual stresses and limit distortion and buckling during welding of annealed Alloy 718 plates, forced cooling can be successfully used.