Directional Recrystallization Processing
Funded by NSF
Faculty: Ian Baker, H. J. Frost
Post-doctoral Fellow: A.Y. Badmos
Graduate Students: Jiying Li, Beatrice Iliescu
Undergraduate Students: Stephanie Johns, Matt Dattwyerr, Milene Kennedy, Mark Tomaszewski, Audi Okullo, Ian Gregoria de Souza, Veronica Mendez
Directional recrystallization is the process whereby a sharp hot zone is moved along a specimen and one or a few grains, which are nucleated when the hot zone is applied to one end of the material, grow along the material with the moving hot zone.
Even if several grains are nucleated initially, one grain may grow faster than the others and, by growth selection, a single grain or a few grains may eventually grow into the unrecrystallized material.
The result is a single crystal or columnar grain structure, which can show enhanced creep properties, improved low cycle fatigue resistance and/or impart crack-stopping behavior. This structure can also be produced by directional solidification, but directional recrystallization has several advantages for producing single crystals or columnar grains:
- Processing temperature of 0.3 Tm for primary recrystallization and 0.5 - 0.7 Tm for secondary recrystallization;
- Minimal solute redistribution, allowing single crystals or columnar grain structure to be produced from non-castable alloys;
- Very complex shaped structure can be processed, allowing net shape processing;
- Crystals can be grown from materials which undergo phase transformations, e.g. peritectoid or peritectic, at high temperatures, and which could not be grown by directional solidification;
- In principle, directional recrystallization can easily be scaled up, thus, for example, enabling the large blades that are used in utility gas turbines to be made in greater yields than is currently possible with directional solidification, where distortion and cracking of the core, shell rupture, mold-metal reactions and several different types of defects cause low yields.
The object of this research is to understand how the key microstructure parameters interact with the processing variables to control the microstructure evolution during both primary recrystallization and secondary recrystallization process.
Directional Primary Recrystallization
The important variables that affect the directional primary recrystallized microstructure include:
- Degree of deformation;
- Deformation temperature;
- Applied stress state during deformation;
- Impurities (both solutes and particles) presented in the materials;
- Temperature gradient ahead of hot zone;
- Annealing temperature ( typically between 0.3 ~ 0.5 Tm);
- Rate of hot zone movement;
- Specimen dimensions.
Experimental
Pure copper (>99.9999%) single crystals with (112)[111] orientation were selected for study. The as grown single crystals were cold rolled to an 80% thickness reduction. The experimental set-up for directional annealing is shown in Figure 1. A 4-cm high wire-wound, home-built furnace and cooling plates were attached to a linear track. The water-cooled plates are spring-loaded and firmly in contact with the wide flat surface of the specimen. The specimen is held stationary and the furnace and cooling plates are translated together by a stepper motor at speeds which range from 2 mm/h to 600 mm/h.
The distance between the cooling plates and the top of the furnace is changed in order to change the temperature gradient ahead of the hot zone. The temperature gradients used were 70°C/cm (no cooling plates), 110°C/cm (medium cooling) and 270°C/cm (strong cooling). For the latter condition the cooling plates were immediately adjacent to the top of the furnace. In the present study, hot zone temperatures of 370°C, 420°C and 470°C were used for directional annealing.
Figure 1. Schematic figure of the annealing furnace.
Results
(a) 2mm/h
(b) 10 mm/h
(c) 30 mm/h
(d) 50 mm/h
Figure 2. Microstructure of directionally-annealed sections of rolled copper crystals at 420°C with the different rates of hot zone movement shown, at a temperature gradient of 70°C/cm.
Figure 3. Grain size as a function of hot zone velocity at 370°C, 420°C and 470°C with a temperature gradient of 70°C/cm. When a columnar grain structure is present, the grain size perpendicular and parallel to the columnar growth direction is shown.
Figure 4. Grain size versus hot zone velocity at 420°C for three different temperature gradients (indicated) ahead of the hot zone. When a columnar grain structures is present, the grain size perpendicular and parallel to the columnar growth direction is shown.
Conclusions
Figure 5. Schematic of longitudinal grain size versus hot zone velocity. The effect of increasing temperature is to move the columnar growth region to the right, while the effect of increasing temperature gradient ahead of the hot zone is to move the columnar growth region to the left.
The following conclusions can be drawn from the preceding experiments on the directional annealing of cold-rolled copper single crystals and was schematic shown in Figure 5.
- Directional recrystallization occurs at an optimum hot zone velocity at a given annealing temperature.
- The hot zone velocity for the onset of directional recrystallization and that for its termination both increase with increasing temperature, features that can be explained by the increased rate of nucleation ahead of the hot zone and the increase in grain boundary mobility with increasing temperature.
- Directional recrystallization occurs at the lowest hot zone velocity for the highest temperature gradient, a feature explained by the decreased rate of nucleation ahead of the hot zone.
Directional Secondary Recrystallization
In addition to the starting texture, the variables controlling produce of single crystals or columnar grain structure by directional secondary recrystallization are:
- Thermal gradient in the hot zone;
- Annealing temperature (typically at >0.5Tm);
- Rate of hot zone movement;
- Impurities (both solutes and particles) presented in the materials;
- Temperature gradient ahead of hot zone;
- Grain size;
- Specimen dimensions.
Experimental
Polycrystalline nickel bars with an average grain size of 25 mm were cold rolled to 90% thickness reduction. The sheets were directionally annealed at different speeds in a modified image furnace at 1000°C. The experimental set up is shown in Figure 1. Infrared light from two 3.5 KW halogen lamps is focused by two gold-coated ellipsoidal mirrors into a 3 cm wide and 2 ~ 3 mm high hot zone. The specimens can be translated through the hot zone at speeds from 1 mm/h to 200 mm/h. Ahead of the hot zone, the specimens are cooled by two spring loaded water-cooled plates which contact the wide surfaces of the specimens. The temperature gradient ahead of the hot zone is controlled by changing the distance between the hot zone and cooling plates. Two temperature gradients were used: 1000°C/cm in which the cooling plates were 1 cm below the hot zone and 50°C/cm in which the cooling plates were removed.
Figure 1. Photos of the annealing image furnace.
Results
At a high temperature gradient:
(a) 2 mm/h
(b) 10 mm/h
(c) 100 mm/h
Figure 2. Microstructure of directionally-annealed sections of rolled polycrystalline nickel at 1000°C with the different rates of hot zone movement shown, at a temperature gradient of 1000°C/cm.
At a low temperature gradient:
(a) 5 mm/h
(b) 10 mm/h
(c) 100 mm/h
Figure 3. Microstructures of directionally annealed cold rolled nickel at 1000°C at different hot zone velocities and a temperature gradient of 50°C/cm. The direction of annealing is from fight to left.
Grain size at a different hot zone velocity
Figure 4. Grain size parallel and perpendicular to direction of hot zone movement versus hot zone velocity at 1000°C with a temperature gradient of 1000°C/cm
Microstructure ahead of a hot zone
(a) at 1000°C/cm
(b) high magnification ahead of big grains
(c) at 50°C/cm
(d) orientation of the small grains in Figure (c)
Figure 5. Microstructures of specimens ahead of the hot zone at high temperature gradient and low temperature gradient of 1000°C annealing temperature of 1000°C.
Conclusions
The following conclusions can be drawn from the above directional annealing of cold-rolled polycrystalline nickel at 1000°C with either a high temperature gradient or very low temperature gradient;
- The temperature gradient ahead of the hot zone is a key parameter for directional annealing. Columnar grain structures can only be produced at a high temperature gradient.
- There exists an optimum hot zone velocity for directional recrystallization. Equiaxed grain structures occur outside of this optimum zone.
- Even at the high temperature, nucleation occurs ahead of the hot zone. The nucleated grains have a typical cubic texture for pure nickel.
- At low temperature gradient, nucleation and grain growth occur ahead of the hot zone. grain growth ahead of the hot zone restrains the directional secondary recrystallization.
References
Papers
- "Simulation of Microstructural Evolution During Directional Annealing", A.Y. Badmos, I. Baker, H.J. Frost, Recrystallization and Grain Growth, Proceedings of the 1st Joint International Conference, Eds. - G. Gottstein and D.A. Molodov, Aachen, Germany, Springer-Verlag, 2001, 1041-1046.
- "Directional Recrystallization of Cold-Rolled Copper Single Crystals", J. Li and I. Baker, Acta Materialia, 50 (2002) 805-813.
- "Microstructural Evolution During Directional Annealing", A.Y. Badmos, H.J. Frost and I. Baker, Acta Materialia, 50 (2002) 3347-3359.
- "The Effect of Hot Zone Velocity and Temperature Gradient on the Directional Recrystallization of Polycrystalline Nickel", J. Li, S.L. Johns and I. Baker, Acta Materialia, 50 (2002) 4491-4497.
Presentations
- "Directional Recrystallization Processing", A. Badmos, I. Baker and H.J. Frost, Fall TMS meeting, St Louis, Mo, 8-12th October, 2000.
- "Directional Recrystallization Processing", I. Baker, A. Badmos and H.J. Frost, AFOSR Metallic Materials Contractors Meeting, St Louis, Mo, 13-14th October, 2000.
- "Simulation of Grain Growth during Directional Annealing", A. Badmos, I. Baker, and H.J. Frost, poster at 2001 Fall TMS meeting, Indianapolis, IN, November 4th-8th, 2001.
- "Directional Annealing of Cold-Rolled Copper Single Crystals with and without SiO2 Particles", J. Li, I. Baker and H.J. Frost, 2001 Fall TMS meeting, Indianapolis, IN, November 4th-8th, 2001.
- "Directional Recrystallization Processing", I. Baker, A. Badmos and H.J. Frost, AFOSR Metallic Materials Contractors Meeting, Snowbird, Utah, 21st-22nd August, 2001.
- "Directional Recrystallization Processing", J. Li, I. Baker, A. Badmos and H.J. Frost, Proceedings of the 2001 NSF Design, Service and Manufacturing Grantees and Research Conference, San Juan, Puerto Rico, January 7th-11th, 2002.
- "Simulation of Directional Annealing", I. Baker, J. Li, A. Badmos and H.J. Frost, 2002 Annual TMS meeting, Seattle, WA, February 17th-21st, 2002.
- "Directional Annealing of Worked Alloys", I. Baker, J. Li, S.L. Johns, B. Iliescu and H.J. Frost, 2002 Annual TMS meeting, Seattle, WA, February 17th-21st, 2002.
- "Directional Annealing of Cold-Rolled Copper Single Crystals With and Without SiO2 Particles", J. Li, I. Baker and H.J. Frost, poster at the 2002 Annual TMS meeting, Seattle, WA, February 17th-21st, 2002.
- "Electron Back-Scatter Diffraction Pattern Study of Directionally Recrystallized MA 754 and Cold-Rolled Nickel", B. Iliescu, J. Li and I. Baker, poster at the 2002 Annual TMS meeting, Seattle, WA, February 17th-21st, 2002.