SUPERPLASTICITY & FORMING OF ADVANCED MATERIALS Superplasticity is the capability to deform crystalline solids in tension to unusually large plastic strains, often well in excess of 1000%. This phenomenon results from the ability of the material to resist localized deformation much the same as hot glass does. As high elongations are possible, complex contoured parts can be formed in a single press cycle often eliminating the need for multipart fabrications. This enables the designer to capture several detail parts into a one piece complex, formed structure. Thus materials with superplastic properties can be used to form complex components in shapes that are very near the final dimension. Superplastic forming also enhances design freedom, minimizes the amount of scrap produced, and reduces the need for machining. In addition, it reduces the amount of material used, thereby lowering overall material costs. (i) Superplastic Forming of “commercial purity” Aluminium Present study indicates that using optimum pressure cycle (strain rates) it is possible to superplastically form inexpensive commercial purity aluminum (AA1100) to the same extent as some of the commercially available superplastic aluminum alloys (AA7475) in a die with relatively simple configuration.

Figure 1. Thickness Profiles of the Formed Cones (AA1100 vs AA7475)

1100

7475

(ii) High Strain Rate Superplasticity in Metal-Matrix Composites P/M 6061 Al and composites with different SiCp volume fraction were studied in this work. Tensile tests were conducted at 653K~871K and 0.1s-1. High Strain Rate Superplasticity (HSRS) is possible and the accompanying mechanisms are identified. The SiCp reinforcement has a softening effect on P/M 6061 Al and composites during HSRS deformation. Mechanical analyses show that the decrease in the strength during HSRS deformation is associated with the fine grain size of the 6061 Al matrix. From the microstructural aspect, the decrease in the strength is due to the occurrence of liquid phase. The formation of liquid phase results from the segregation of Mg at grain boundaries and SiCp/Al matrix interfaces. The occurrence of liquid phase and composites reduces the strength of P/M 6061 Al and composites during HSRS deformation, and enhances their ability to achieve large elongation-to-failure. However, the grain size of the Al matrix appears to play a more important role in HSRS than the reinforcement content.

Fig 2 TEM microstructure of as-extruded specimens of 6061 Al and composites. (a).Al6061(unreinforced, 20); (b). MMC1 (10vol%SiCp, 20); (c). MMC2 (18vol%SiCp, 20); (d). MMC3 (10vol%SiCp, 47);

(a)

1 µm 1 µm

(b)

1 µm

1 µm

(c) (d)

1 µm

(iii) 2-Stage Deformation of Magnesium Alloy Recognizing the capability of Mg-3Al-1Zn to dynamically recrystallize, the ‘Two-Stage Deformation Method’ was proposed to further enhance superplasticity. In this method, the first stage (Stage I) of deformation is aimed at refining the coarse microstructure through DRX utilizing the optimum DRX conditions; while the second stage (Stage II) is carried out at the optimum conditions that have been earlier identified for superplastic deformation of coarse-grained Mg-3Al-1Zn, i.e. 1×10-4s-1 at 400°C and 2×10-4s-1 at 450°C. Table 1 lists the results of the elevated temperature tests performed using single and two-stage methods. Table 1. Experimental conditions and elongation-to-failure for samples tested using single and two-stage deformations. All samples were tested at constant strain rates.

Sample No.

Test temperature, °C Strain rate, s-1 Strain*, %

Elongation-to-failure,

±5% 1 250 1×10-4 − 140 2 400 1×10-4 − 250 3 250 400 1×10-4 60 320 4 450 2×10-4 − 265 5 250 450 1×10-4 2×10-4 60 360

* Strain where temperature or/and strain rate change occurred. The time required for raising the temperature to that of Stage II was about 10 to 15 minutes. Although Sample 1 was deformed at optimum DRX conditions and possessed a completely recrystallized microstructure, its elongation-to-failure was limited to 140% due to cavitation failure. Sample 2 was independently deformed at 400°C and 1×10-4s-1 achieving 250% and resulting in coarse-grained microstructure.

Figure 3. Specimens of the two-stage deformation tests. (a) Undeformed sample with initial gauge length of 15mm. (b) Sample 3 failed at 320%, employing Stage I: 250°C, 1×10-4s-1 to 60%; followed by Stage II: 400°C, 1×10-4s-1 to fracture. (c) Sample 5 fractured at 360%, using Stage I: 250°C, 1×10-4s-1 to 60%; followed by Stage II: 450°C, 2×10-4s-1. All samples failed by necking, as indicated by the arrows.

(a) Undeformed sample Gauge=15mm

(b) T=250°C 400°C ε& =1×10-4s-1

fε =320%

(c) T=250°C 450°C ε& =1×10-4 2×10-4s-1 fε =360%

15 mm

69 mm

63 mm

(iv) Grain Boundary Character Distribution (GBCD) on Cavitation Behaviour of Superplastic Aluminium The sliding capability of grain boundaries is inhomogeneous due to the anisotropic GBCD. Cavities are nucleated at sites where high stress concentrations have been induced by GBS and where the local stress exceeds the bonding strength of the particle to the boundary, or the boundary to the boundary. There are interactions between the DFZ formation and the cavity nucleation. The DFZ formation or the cavity nucleation may depend on the local stress condition and the grain boundary bonding strength. The formation of cavity nuclei at the grain boundaries will lower the potential of DFZ formation in their vicinity, and vice versa.

The GBCD, especially the grain boundary bonding strength at test temperature, is the key factor controlling the formation of cavity stringers. The presence of liquid phase on the grain boundaries significantly reduces the bonding strength. The formation of preferential cavity stringer is closely related to the GBCD, GBS and grain boundary debonding. Based on this, a model of cavity stringer formation was proposed.

(a) (b)

(c) Fig. 4 Surface morphologies of the samples after superplastic deformation at 516°C, 10-3s-1. (a) the T-sample strained to fracture, (b) the L-sample after fracture, (c) schematic illustration of the configuration variation of the grains shown in figure (b). The long continuous filament AD and CE showed that the originally contiguous Grains A and D, C and E were far separated, while the originally contiguous grains B and C were connected via a relatively short filament BC. Stressed horizontally.

A B C

D E

(i) Configuration of grains A, B, C, D and E before deformation

(ii) Configuration of grains A, B, C, D and E after deformation

(v) 3-D Modelling of Superplastic Forming Process Finite element simulations for 3D SPF processes presented . The commercial code ABAQUS/Standard was used as computational tool. For verification of the numerical results, the experiment of conical bulging was carried out and the good agreement of thickness distribution between the numerical results and the experimental results was obtained. In general the material flow is constrained by a bead at a constant distance from the die profile in the practical forming process. So the draw in effect should be considered in finite element simulation in order to get more accurate results. The influence of friction depends on the bulging type. In conical bulging, small frictional coefficient can improve the inhomogeneity of the thickness distribution. For the rectangle box bulging, as the friction decreases, the filling ability of the sheet towards to the die corner and the inhomogeneity of thickness distribution can also be improved, but the sheets near the polar top of the die are thinned more seriously. In this case, it is not true that, the smaller the friction, the better the formed part. Lubrication should be performed appropriately. A high m value can improve the inhomogeneity of the thickness distribution, but more forming time will be needed. Because the m value depends strongly on the strain rate of the deformation process, it is recommended to search for an optimum strain rate such that the component can be formed as fast as possible but still under the superplastic conditions.

(a) t=183s

(b) t=659 s

Fig.5 Deformed meshes at different forming time

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 5 10 15 20 25 30 35 40 45 50

Initial distance from center (mm)

Thic

knes

s (m

m)

t=3500st=2880st=2280st=880s

Fig.6 The thickness distribution at different forming time

Acknowledgement: JT ARC 4/95 Grant from MOE & A*STAR.

Related Publications:

1. S.N.Patankar, Tan Ming Jen, “Superplastic Forming of Commercial Purity Aluminium” (Scripta Materialia, Vol.38 (1) (1998) pp 145-148)

2. J.C.Tan, M.J.Tan, “Superplasticity in a rolled Mg-3Al-1Zn alloy by two-stage deformation method” (Scripta Materialia, Vol. 47 (2002) pp101-106)

3. X.Li, M.J.Tan, “High Strain Rate Superplasticity and deformation mechanisms of powder metallurgy 6061 Al/SiCp composites” (Mater. Sci. & Tech., Vol. 18 (5) (2002) pp581-585)

4. C.L.Chen, M.J.Tan, "Effect of grain boundary character distribution (GBCD) on the cavitation behaviour during superplastic deformation of Al7475" (Materials Science and Engineering A Vol. 338 (2002) pp243-252)

5. G.Y.Li, M.J.Tan, K.M.Liew, "Three-dimensional modelling of superplastic forming and analysis of influence of various factors on the forming processes" in (Advances in Concurrent Engineering, Editors: Jardim-Goncalves, Roy & Steiger-Garcao, Publisher : Swets & Zeitlinger B.V., Lisse, The Netherlands, 2002,ISBN: 90 5809 502 9, pages 297-304)

6. S.Thiruvarudchelvan, M.J.Tan, “Two Novel Techniques for Forming Regularly Spaced Deep Recesses on Aluminium Sheet Panels” (Materials Science Forum – accepted for publication)

Tan M.J., S.Thiruvarudchelvan, and Liew K.M.