A Review on Superplastic Forming of Ti-6Al-4V Alloy

JournalMaterials of Science Alloys andand Technology Compounds

AA Review Review on on Superplastic Superplastic Forming Forming of of Ti-6Al-4V Ti-6Al-4V Alloy Alloy --Manuscript Draft--

Manuscript Number: MST16321 Full Title: A Review on Superplastic Forming of Ti-6Al-4V Alloy Article Type: Review Keywords: Superplastic forming; Superplasticity; Ti-6Al-4V; Microstructure; Multi step forming dies Corresponding Author: Sai Pratyush Akula, B.E. Mechanical Engineering Birla Institute of Technology and Science Hyderbad, Telangana INDIA Corresponding Author Secondary Information: Corresponding Author's Institution: Birla Institute of Technology and Science Corresponding Author's Secondary Institution:

First Author: SaiAkhil Pratyush Agnihotri Akula First Author Secondary Information: Order of Authors: SaiAkhil Pratyush Agnihotri Akula AkhilSai Akula Agnihotri Pratyush

AmitAmit Kumar Gupta Order of Authors Secondary Information: Abstract: This paper presents a review on the superplastic forming of Ti-6Al-4V alloy, which has been used to manufacture parts of complex shapes and geometries. This paper outlines the major work carried out on this front in the past three decades. It covers various aspects related to experimental setups, including the manufacture of dies and their modifications to maintain alloy thickness uniformity after forming. A detailed study of the process parameters has also been done to note the most important physical conditions required for successful forming. This is followed by the influence of microstructure, modern applications of superplastic forming of different titanium alloys and is concluded with an insight into the future work and progress in this field. Additional Information: Question Response

Please state the novelty of your 1. Elaborates on the latest advancements in the field of SPF of Ti-6Al-4V submission by providing (at least) four 2. Dedicated to processing of Ti-6Al-4V for aerospace and prosthetic applications aspects of originality or uniqueness in 3. First paper in over three decades to cover the SPF of Ti-6Al-4V alloy your article. Please keep each point to a 4. Consolidated work on theoretical and experimental aspects of SPF of Ti-6Al-4V maximum of 12 words.

Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation Manuscript Click here to download Manuscript A review on superplastic forming of Ti-6Al-4V alloy.docx

1 2 3 4 5 A Review on Superplastic Forming of Ti-6Al-4V Alloy 6 7 8 Sai Pratyush Akula, Akhil Agnihotri, Amit Kumar Gupta 9 10 11 Department of Mechanical Engineering, BITS Pilani, Hyderabad Campus, Telangana, India - 500078 12 13 14 Keywords: Superplastic forming, Superplasticity, Ti-6Al-4V, Microstructure, Multistep forming dies, etc. 15 16 17 ABSTRACT 18 19 20 This paper presents a review on the superplastic forming of Ti-6Al-4V alloy, which has been used to 21 manufacture parts of complex shapes and geometries. This review paper serves as a means to give an 22 23 outline on all the major work that has been carried out on this front in the past three decades. It covers 24 various aspects related to superplastic forming of Ti-6Al-4V alloy like experimental setups which include 25 the manufacture of dies and their modifications to maintain alloy thickness uniformity after forming. A 26 27 detailed study of the process parameters has also been done to note the most important physical 28 conditions required for successful forming. This is followed by the influence of microstructure, modern 29 30 applications of superplastic forming of different titanium alloys and is concluded with an insight into the 31 future work and progress in this field. 32 33 34 1 INTRODUCTION 35 36 37 Metal forming process involves changing the shape of a metal into a desired one without any loss of mass 38 by the means of plastic deformation. It is one of the most important manufacturing techniques due to its 39 40 cost effectiveness, enhanced mechanical properties, flexible operations, high productivity and metal 41 saving. A general forming process is said to include a punch and die mechanism, where the metal takes 42 the shape of the die due to the application of pressure. This mechanical process need not rely on 43 44 temperature to get the desired geometry. Various forming techniques such as forging, rolling (hot and 45 cold) and extrusion etc. are being used today in automotive, appliances, aerospace and other industries. 46 However, a unique kind of forming technique known as superplastic forming is being studied lately, which, 47 48 with the help of elevated temperatures, achieves higher plastic deformation in the metal. 49 50 Superplasticity refers to the property of metals to achieve elongation of over 1000% of without necking, 51 52 usually at elevated temperatures (about 900℃ for Ti-6Al-4V) under a constant, controlled strain rate. The 53 optimum strain rate varies with the superplastic material and is attributed to the viscous behavior shown 54 55 by metals and alloys with refined and stable microstructure at elevated temperatures. 56 57 Superplastic forming (SPF) of sheet metal has been used to produce complex patterns and integrated 58 59 structures that are light-weight and stronger than the assembled components [45]. This process of SPF is 60 often combined with diffusion bonding (DB) to manufacture multisheet structures [40]. This results in 61 62 63 1 64 65 1 2 3 4 reduction in the total number of parts and often gives post-processing benefits such as enhanced product 5 6 performance and cost reduction [7]. This process also has few disadvantages in terms of high working 7 temperature, non-uniform thickness distribution etc. [41]. Titanium has high specific strength and good 8 cold-formability, resulting in applications in the manufacturing of aircraft and automobile parts such as 9 10 compressor blades and turbine discs. However, the technique for improving its formability at elevated 11 temperatures is still a challenge at present [42]. Over the last few decades, a growing interest in the 12 13 superplasticity of Titanium has been seen, as shown in Figure 1 and Figure 2. 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Figure 1: SPF literature focus in the papers 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 Figure 2: Literature trends in SPF since 1980 57 58 59 60 61 62 63 2 64 65 1 2 3 4 Discontinuous Metal Matrix Composites (MMCs) possess high strength but low ductility, and are difficult 5 6 to form with traditional techniques such as rolling, forging or bending [23]. Moreover, due to high 7 hardness, there are only limited number of practically plausible machining techniques. Powder metallurgy 8 does not provide for high dimensional tolerances while casting is only possible in cases where the liquid 9 10 matrix is non-reactive with the reinforcing phase and the environment. In such cases, superplastic forming 11 is desired since it allows for large strains with limited necking and cavitation. 12 13 14 During SPF, differential straining across the sheet takes place, causing large amounts of variation in 15 thickness, thus reducing the sheet’s applicability and increasing post-processing costs [7]. This straining 16 17 occurs in two parts. First, during free forming (when the sheet has not come in contact with the die), 18 which can be controlled if the alloy exhibits high strain rate sensitivities of flow stress. Second, when the 19 alloy comes in contact with the die, and although differential straining is locally controlled, other parts 20 21 continue to experience this phenomenon. Thus, the last part of the alloy to come in contact with the die 22 has experienced the most strain and hence, has the least thickness. This can be controlled by ensuring 23 that the alloy, more or less, comes into contact with the die in a uniform manner and that the 24 25 experimental parameters such as load and forming temperature are specific to the local strain gradients. 26 27 28 Bellows expansion joints find application in areas involving expansions and contractions of pipes due to 29 temperature changes, mechanical vibrations, non-uniform support structure, etc. [24]. Superplastically 30 formed using gas pressure and compressive axial load, Titanium alloy Ti-6Al-4V is an ideal candidate for 31 32 U-type bellows expansion joints as it allows for more than 1000% elongation. The entire process was 33 divided into tube blank fabrication and the actual SPF process. 34 35 36 2 Applications of superplastic forming 37 38 39 40 Titanium alloys are widely used in biomedical applications like cranioplasty since they favour 41 osseointegration with bones, have excellent mechanical features and are biocompatible. This makes the 42 alloy suitable for the making of prostheses. Prostheses are machines which are usually expensive and take 43 44 a lot of time to manufacture. Due to this, superplastic forming has found its application in the making of 45 prostheses using titanium alloys [20]. Ambrogio et al., 2018; has work related to the testing and 46 manufacturing of cranial prosthesis using superplastic forming. The cranial prostheses developed by 47 48 Ambrogio et al., 2018 can be seen in Figure 3 and the die setup can be seen in Figure 4. 49 50 51 52 53 54 55 56 57 58 59 60 Figure 3: Prostheses a) made from incremental forming b) made from superplastic forming. Credits: [20] 61 62 63 3 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Figure 4: a) Die setup for SPF of cranial prostheses b) Formed part. Credits: [20] 19 20 21 22 Sorgente, Palumbo, Piccininni, Guglielmi & Aksenov, 2018; have performed a study on the thickness 23 distribution on the prostheses made with the help of finite element simulations and free inflation tests at 24 25 different pressures. The results of this study can be found in Table 1. 26 27 Table 1: Strain rate vs thickness of the component. Credits: [22] 28 29 Target strain rate, s-1 Minimum thickness Minimum thickness Forming time, s 30 (path), mm (blank), mm 31 -5 32 2 x 10 0.669 0.635 61900 33 4 x 10-5 0.667 0.647 33600 34 8 x 10-5 0.685 0.664 16175 35 1 x 10-4 0.682 0.665 12625 36 -4 37 2 x 10 0.653 0.653 7300 -4 38 4 x 10 0.644 0.643 3775 39 1 x 10-3 0.615 0.614 1425 40 41 42 43 Superplastic forming has a huge application in the area of aerospace products. Aircraft manufacturing 44 giants like Boeing use SPF to make a few components for their aircrafts [4]. Hefti, 2010; in brief explains 45 the methods and problems related to SPF of Boeing 757 aircraft part. According to the author, superplastic 46 47 forming along with diffusion bonding finds its application in the fabrication of the heat shield assembly 48 and a tailcone for a twin-aisle aircraft (Figure 5 and 6). These parts are superplastically formed using fine 49 grain Ti-6Al-4V at a temperature of 750 - 775℃. 50 51 52 53 54 55 56 57 58 59 60 61 62 63 4 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Figure 5: Twin-aisle aircraft heat shield. Credits: [4] 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 Figure 6: Tail cone of a twin-aisle aircraft. Credits: [4] 36 37 38 39 40 The above section has a brief description of the utility of superplastic forming in industries. The methods 41 and the setup for this process will be discussed in the next section. It will contain brief explanations about 42 43 various dies that can be used, how the process is carried out and the influence of various process 44 parameters in manufacturing a component using superplastic forming. 45 46 47 3 Experimental Setup 48 49 3.1 Dies 50

51 3.1.1 Single Step Dies 52 1. Ceramic dies provide superior formability of Ti-6Al-4V due to their adjustable coefficient of 53 thermal expansion [25]. In this regard, a ceramic die was prepared by sintering the green of the 54 55 ZrO2-TiO2 with the polyvinyl alcohol (PVA) in the atmosphere. Yttria semi-stabilized zirconia (3Y- 56 ZrO2) was used as the basic ceramic die material and the addition of TiO2 to ZrO2 allowed for an 57 adjustable coefficient of linear thermal expansion (CTE) and prevented chemical reactions 58 59 between the die and the Ti-6Al-4V alloy. It is imperative to determine the CTE of the die relative 60 to the alloy for dimensional accuracy in blow-forming. To measure the CTE of the ZrO2-TiO2 die, a 61 62 63 5 64 65 1 2 3 4 10 x 50 mm cylinder was manufactured in four steps [26]. Firstly, ZrO2 and TiO2 powders along 5 6 with 25% PVA (Poly Vinyl Alcohol) were mixed in an agitator for 1 hour at 120℃. Then, the green 7 of ZrO2 and TiO2 were pressed into the cylinder mould at 100℃ followed by cooling to room 8 temperature over a period of 3 hours after which de-binding and pre-sintering were performed 9 10 over a range of 25-1000℃ for 24 hours. This was concluded by sintering the ceramic cylinder at 11 1500℃ for 1 hour. The schematic for this process is shown in Figure 7 while Figure 8 shows the 12 13 die setup for SPF forming along with its labelled parts. 14 2. 15 16 17 18 19 20 21 22 23 24 Figure 7: Process schematic of sintering ZrO2 – TiO2 cylinder. Credits: [25] 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 Figure 8: Schematic structure of SPF die – 1 Formed punch; 2 Female die; 3 Outer sleeve; 4 Heating apparatus; 5 45 Power lot; 6 Filling piece. Credits: [26] 46 47 48 49 To find the optimum constituent fraction for the die to match the CTE of Ti-6Al-4V at the 50 51 superplastic temperatures, the volume fraction of TiO2 was varied and corresponding CTEs were 52 measured. Furthermore, the influence of relative density on CTE was also studied and it was 53 observed that higher the relative density, more is the CTE at the superplastic temperature, which 54 55 can be seen below in Figure 9 (a) and 9 (b) respectively. Therefore, by trial and error, the optimal 56 mixed powder with properties of 27% TiO2 volume fraction sintered at 1500℃ for 2 hours with a 57 linear CTE of 8.92×10-6 C-1 was selected as it matched the CTE of Titanium at 900℃. This can be 58 59 seen in Figure 10, which shows the variation of volume fraction and relative density together, 60 61 62 63 6 64 65 1 2 3 4 based on a different linear CTE to match the linear CTE of Ti-6Al-4V, which is 8.92 x 10-6 ℃ at 5 6 930℃. 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 (a) (b) 24 25 Figure 9: Relationship between linear CTE and (a) volume fraction of TiO2 and (b) relative density. Credits: [25], [26] 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 Figure 10: Different volume fraction of TiO2 and corresponding relative density for manufacturing the optimal ZiO2- 47 TiO2 ceramic die. Credits: [25], [26] 48 49 50 51 52 This mixed powder is formed into a die by pressing it into a steel model at 120℃ with a double 53 cling film layer for separation, as shown in Figure 11a. Later, the newly formed die is sintered at 54 1500℃ to increase strength by heating at 8℃/min and holding at maximum temperature for 2 55 56 hours before cooling to room temperature. The final die can be seen in Figure 11b. 57 58 59 60 61 62 63 7 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 (a) (b) 21 Figure 11: (a) Steel model for forming semi-finished ZrO2-TiO2 deep cylinder ceramic die: 1 Pressure head, 2 22 External mould, 3 Mixed power, 4 Internal mould, 5 Heating apparatus, 6 Screw bolts; (b) Finished ZrO2-TiO2 deep 23 cylinder ceramic die. Credits: [25], [26] 24 25 26 3. Tubular blank fabrication was carried out by cutting and bending of the tube [28], welding the tubular 27 blank and checking for defects via radiography followed by sizing and welding cover with gas entrance 28 29 connection. The tubular blank was bended by stress relaxation at 650℃ with a holding time of 30 30 min followed by furnace cooling. Due to welding deformations, sizing of the pipe was performed at 31 32 700℃ with a holding time of 45 min and was cooled similar to the tubular blank. Then, the tubular 33 blank was welded with a square butt and no gap by automatic plasma arc welding (A-PAW). During 34 the process, the blanks were cleaned by emery paper coated with H2O/HNO3/HF solution (76/20/4 35 36 vol.%) and degreased with acetone before welding. Further details regarding welding are given in 37 Table 2 below: 38 39 40 Table 2: Process parameters for different welding positions for manufacturing of the tubular blank. Credits: [28] 41 Welding Position Welding Arc Current(A) Arc Travel speed (m/s) Argon flow (L/min) 42 Method Voltage(V) 43 Covers M-GTAW 35 18 0.1 20 44 Gas entrance M-GTAW 30 20 - 20 45 connection 46 Tubular blank A-PAW 40 20 0.06 25 47 48 49 50 51 52 3.1.2 Multi Step Dies 53 54 55 1. Two-step rectangular die enclosed in a sealed die with uniform pressure applied by the inlet of a 56 pressurized inert gas such as Argon. The to-be-formed diaphragm is rigidly clamped around the 57 58 periphery outside the enclosure, resulting in a significantly thinner border [27]. The gas pressure 59 for blow forming is maintained between 1-2 MPa to give a dome with a near-hemispherical shape 60 61 62 63 8 64 65 1 2 3 4 along with an equilibrium biaxial stress condition at the tip and a plane-strain balanced state at 5 6 the equator [23]. The variation of geometry of the workpiece in various stages of forming is shown 7 in Figure 12. 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Figure 12: (a), (b), (c) and (d) Different stages of pressure blow forming technique in a stepped rectangular die. 33 Credits: [23] 34 35 36 2. Superplastic forming is an effective process for fabricating hollow fan blades, honeycomb 37 structures, bellows expansion joints and large engine components [17]. However, due to slow 38 39 forming rate and high cost of materials, a new hybrid sheet-forming process was used which 40 combined deep drawing and blow forming in one step, resulting in a faster forming process. The 41 major components of the experimental setup include a three-stepped die with 80, 135 and 150% 42 43 cumulative surface expansion ratios, a blank holder with INCONEL disc-springs, a 43mm stroke 44 punch and a sealing system for blow-forming, all of which are shown below in Figure 13: 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 9 64 65 1 2 3 4 5 6 Figure 13: (a) Components of the superplastic-like forming die assembly (b) Die-cavity with non-isothermal heating and showing % surface area expansion at each step. Credits: [17] 7 8 9 3.1.3 Non-traditional dies 10 1. To make an airtight closed chamber for non-traditional die type of superplasticity, an open 11 cylinder of Ti-6Al-4V with a circular base (2.38 mm wall thickness and 16.6 mm length with 2.24 12 13 mm thick disk base) was machined and sealed with another disk-like top of Ti-6Al-4V/TiC 14 composite of 2.24mm thickness [23]. To this, a 6.35mm diameter pipe of Ti-3Al-2.5V was welded 15 to allow gas flow and K-type thermocouple assemblies inside the sealed cylinder. For determining 16 17 the gas-pressure induced radial and circumferential strains, the disks were marked with 18 concentric circles spaced 1 mm apart and parallel lines going through the centre of the disk 19 respectively. A Boron-Nitride (BN) coated 2.5mm thick sleeve to prevent radial strains and allow 20 21 circumferential strains housed the closed cylinder. The entire assembly, as in Figure 14, was 22 placed inside a cylindrical graphite chamber within a stainless-steel vessel. 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 Figure 14: Complete assembly of an airtight closed chamber for SPF forming. Credits: [23] 47 48 49 2. Before forming, the sheets underwent heat treatment up to 1070℃ in an induction heating 50 machine of 25 kW power [7]. Next, they were air-cooled and each sheet was cut according to the 51 52 circular clamping periphery. The forming vessel is made up of a high temperature nickel alloy with 53 a removable top (A). The top houses a stainless-steel die (B) placed inside the lower vessel (C). 54 Heat is supplied through the resistance-heating coils (D) and constant superplastic temperature 55 56 of 920℃ is maintained by a Proportional-Integral-Derivative (PID) controller while Argon gas 57 pressure is regulated by computerized mass-flow valves. The entire assembly setup is given in 58 59 Figure 15 below. 60 61 62 63 10 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Figure 15: Schematic outline of the high temperature forming chamber. (A) – Top pressure vessel (B) Die housing (C) 26 Bottom pressure vessel (D) Heater. Credits: [7] 27 28 29 30 31 3.2 The SPF Process 32 33 3.2.1 With Ti-6Al-4V 34 35 36 1. The structure of setup was divided into pre-forming die, sheet metal and forming die [28] as 37 shown in Figure 16 (a). Two vent holes at the top (1 and 2) and one at the bottom (3) were 38 provided to prevent the sheet sticking to the forming die. The Ti−6Al−4V sheet was heated at a 39 40 rate of 15 ℃/min up to 930 ℃ and held for 60 minutes before pre-forming. Then, the high- 41 pressure nitrogen gas ramped up to 1.4MPa was used to blow the sheet into the pre-forming die 42 43 for 750 seconds. For achieving adaptation to the deformed state, a 30-minute holding time was 44 observed following which more nitrogen was used to blow the sheet into the forming die at 45 1.4MPa in 1500 seconds after which a dwelling time of 1500 seconds was provided. The entire 46 47 variation of pressure vs time for the process is shown in Figure 16 (b). This was achieved by a two- 48 step blowing process: first the gas was blown through vent hole 3 at pre-forming pressure and 49 then it was released through vent hole 1 and 2 while gas at forming pressure was injected through 50 51 vent hole. 52 53 54 55 56 57 58 59 60 61 62 63 11 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Figure 16: (a) Schematic structure of the die for SPF (b) Experimental Pressure vs Time curves. Credits: [28] 19 20 21

22 2. For the actual SPF process shown in Figure 17 (a), (b), (c) and (d), the tubular blank, coated with 23 anti-oxidizing agent, and die were pre-heated to 150-200℃ while being lubricated by graphite 24 powder at their contact surfaces [24]. The entire apparatus was kept in a furnace at 927℃ (the 25 26 approximate superplastic temperature for Ti-6Al-4V) for 30 minutes. Initially, gas at 0.2MPa was 27 injected into the tubular blank for 30 minutes to initiate circumferential plastic deformation, 28 followed by lowering of the punch to close the upper, middle and lower dies tightly. Subsequently, 29 30 gas at 2.5MPa was injected for a period of 10 minutes after which the punch was removed and 31 the setup was kept at room temperature to cool. Following sufficient cooling, the middle die was 32 removed and the two covers were ground from the formed alloy. The bellows were then 33 34 machined to the required dimension by lathe and all the applied coatings were removed by grit 35 blasting. A digital micrometer was used to measure the thickness at various points along the 36 37 finished bellows and it showed a variation of about 9%, which can be seen in Figure 17 (e) below. 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 12 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Figure 17: SPF procedure: (a) original state (b), bulging (c) dies clamping, (d) shape forming, (e) Thickness 21 distribution of as-formed bellows. Credits: [24] 22 23 24 3. The honeycomb structure [29] is produced by selective bonding of four titanium sheets, each with 25 a specific core pattern, and placing them one over the other. Core flat sheets have a thickness of 26 27 0.3mm and face flat sheets have a thickness of 0.8 mm. To ensure that no gas leaks out, all the 28 four sheets are sealed along their periphery and placed into process tooling where they are 29 30 heated to the SPF temperature. Moreover, to reduce post-processing, close tolerances and tight 31 quality control measures should be implemented during the entire process. The four sheets are 32 subjected to forming by applying gas pressure until they come in contact with the die [29]. 33 34 Pressurized gas is injected until diffusion bonding of the sheets resulting in one, monolithic 35 structure is completed. Optimum amount of pressure needs to be applied to ensure superplastic 36 forming at stop-off areas. Its variation with time as shown in Figure 18 (a). 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 (a) (b) 56 57 58 59 60 61 62 63 13 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 (c) 17 Figure 18: (a) Pressure vs time variation for SPF-DB processing (b), Assembly for the processing. 1 – face sheet, 2 – 18 stop off, 3 – core sheet, 4 – stiffening rib, 5 - die (c) Points at which thickness was measured whose values are 19 shown in Table 3. Credits: [29] 20 21 22 Table 3: Thickness distribution of the honeycomb structure. Credits: [29] 23 24 Location 1 2 3 4 5 6 7 8 25 Thickness (mm) 0.76 0.77 0.78 0.76 0.28 0.28 0.28 0.28 26 27 28 29 4. Two categories of sheets: as received and heat treated were subjected to forming at the above 30 conditions for 1000 seconds with strain rates ranging from 1×10−4 to 8×10−4 s−1 [7]. After cooling 31 32 in the atmosphere, thickness was measured for both types of sheets. The lower cup showed a 33 large reduction in thickness and the greatest thinning was found at the innermost corner for both 34 the categories. However, in the heat-treated sheets, there was significantly higher thinning in the 35 36 non-transformed regions than the transformed regions due to the interface effect between the 37 two regions. Comparison with FE simulations is given below in Figures 19 (a) and (b) in which 38 39 length is the peripheral distance along the formed shape direction out from the centre. 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 Figure 19: Comparison of finite element simulation and experimental strains for (a) the as-received blank and (b) the 58 blank with transformed structure. Credits: [7] 59 60 61 62 63 14 64 65 1 2 3 4 5. To carry out SPF, the chamber and vessel initially were maintained at 10 mTorr and evacuated 5 6 again to the initial pressure conditions [23]. To achieve degassing at 10 mTorr pressure, the 7 chamber was heated to 150℃ for 10 hours and ramped to 350, 600 and 800℃ with a holding time 8 of 1 hour at each increment. For the isothermal experiment at 1000℃, 99.9996% pure Argon gas 9 10 at 350 kPa and 120 kPa was introduced in the test chamber and vessel respectively. Titanium foil 11 getters were used to prevent oxidation of the sample and the experiment was conducted till 12 13 failure of the Ti-6Al-4V weld. For the cyclic experiment, Argon gas at the same respective 14 pressures was used but with temperature range of 580℃ and 1020℃ with a holding period of 600 15 s (with ramps of about 60 s) at each increment. These cyclic experiments were conducted till the 16 17 Ti-6Al-4V dome apex developed four cracks of approximately 1 mm length. Constant pressure was 18 maintained via an Argon reservoir and two sets of cyclic experiments were conducted. In the first, 19 both the reinforced and unreinforced disks were allowed to deform with interruptions after every 20 21 10 hours for visual inspection while in the second, unreinforced disk was supported by a steel 22 insert to prevent deformation, allowing only the reinforced disk to deform. A point micrometer 23 was used to calculate the principal strains by measuring the thickness of the formed dome at 24 25 different distances from the apex. 26 27 28 Superplastic forming is not only carried out for Ti-6Al-4V, but for many other titanium alloys as 29 well. These other alloys usually have small proportions of vanadium or molybdenum to impart a 30 few specific desired properties to the alloy and the formed component. The nuances of the 31 32 forming of these other titanium alloys has been explained in the next subsection. 33 34 3.2.2 With Other Titanium Alloys 35 36 37 1. Welded envelopes of VT6S (Ti–6Al–4V), VT14 (Ti–4.2Al–2.7Mo–1.2V) and OT4-1 (Ti–1.8Al–1.45Mo) 38 39 of thicknesses 0.8, 1 and 5 mm respectively were used for superplastic free forming of a spherical 40 shell [16]. For manufacturing the spherical envelope, two circular sheets were taken and a pipeline 41 for Argon entry was allowed for by welding a gas entrance connection to one of the sheets in the 42 43 middle. Their properties are tabulated in Table 3 along with the results in Table 4, in which D0 refers 44 to original sheet diameter while Ds signifies the diameter of the formed sphere. Six sheets were 45 46 welded in pairs across their perimeter based on the relative rolling direction - 0°, 45° and 90°. Three 47 methods, namely argon arc welding, roll seam welding and diffusion bonding were used for the 48 process. To allow for SPF, the fastened envelope was placed on a π-type stand in a furnace. Argon 49 50 was injected into the envelope through the provided pipeline. Gas pressure was controlled 51 according to the predetermined pressure-time cycle for optimum conditions for SPF. During 52 forming, the geometry of the envelope was monitored visually, resulting in an error of 3 mm and 53 54 via a graduated scale on a computer fitted with a camera, resulting in an error of 1 mm. Figure 20 55 shows the final spherical shell formed. 56 57 58 59 60 61 62 63 15 64 65 1 2 3 4 Table 4: Material properties of the Titanium sheets used in manufacturing spherical shells out of welded Titanium 5 envelopes. Credits: [16] 6 7 Chemical Initial Average Optimal conditions for Material constants Procedure 8 Composition thickness grain size d* superplasticity used 9 s0 (mm) (µm) -m 10 Topt sopt xopt K (Mpa s ) m 11 (℃) (MPa) (s-1) 12 Ti-6Al-4V 5 3-5 920 15 2 x 10- 450 0.40 Standard 13 4 14 - 15 Ti-4.2Al- 0.8 1-2 870 5 3 x 10 100 0.52 Standard 3 16 2.7Mo-1.2V - 17 Ti-1.8Al- 1 3-5 890 20 5 x 10 290 0.35 Standard 18 1.45Mo 4 19 20 21 Table 5: Results of experiments on superplastic free forming of spherical shells out of welded titanium envelopes. 22 Credits: [16] 23 24 Sheet Chemical Thickness D0 (mm) Ds (mm) D0/Ds Deviation 25 material composition (mm) (mm) 26 VT6S Ti-6Al-4V 5 200 164 1022 ±2 27 VT6S Ti-6Al-4V 1 108 89 1.21 ±2 28 VT6S Ti-6Al-4V 1 188 158 1.19 ±2 29 VT6S Ti-6Al-4V 1 290 246 1.18 ±3 30 VT6S Ti-4.2Al-2.7Mo- 0.8 188 158 1.19 ±2 31 1.2V 32 OT4-1 Ti-1.8Al-1.45Mo 1 188 160 1.18 ±2 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 Figure 20: Spherical shell formed out of the envelope manufactured by diffusion bonding (D0 = 200 mm, Ds = 164 49 mm). Credits: [16] 50 51 52 53 3.3 Process Parameters 54 55 56 1. In addition, it was found that splitting the die into a pre-forming and forming die as in Figure 21 57 and separately modifying their surfaces resulted in greater uniformity of sheet thickness as 58 compared to the conventional single-piece die blow forming [29]. With this regard, increasing the 59 60 friction of the pre-forming die by mechanical machining and decreasing the friction of the forming 61 62 63 16 64 65 1 2 3 4 die by spraying BN powder were used to obtain more accurate finished products. The friction 5 6 coefficient was monitored by compression ring method and its values for machined surface and 7 BN powder sprayed surface were obtained as 0.55 and 0.15 respectively. 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 (a) (b) 25 Figure 21: (a) Splitting of die into forming and preforming results in uniform thickness distribution (b) Meshed 26 geometry showing the shape of preforming die. Credits: [29] 27 28 29 2. Uniaxial tensile texts were conducted on Ti–22Al–24Nb [30], with traces of Mo, V and Si on 30 Instron 5500R, an electronic universal material testing machine. Since both temperature and strain 31 32 rate affects the superplastic ability, first, optimal temperature was set by conducting tensile tests 33 within 900-980℃ at increments of 30℃ at a strain rate of 0.0001 s-1. Next, optimal strain was selected 34 at 0.0005 s-1 for proceeding study. The superplastic forming tests were conducted in a 1000kN forming 35 36 machine via gas bulge forming until the sheet fitted the mould completely. 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 (a) (b) 53 54 Figure 22: Variation of engineering stress and engineering strain in SPF for various (a) temperatures (b) strain rates. Credits: [30] 55 56

57 58 As can be seen from Figure 22, tensile stress increases to a rapid peak with increase in tensile strain and 59 decays somewhat gradually in the case of 960° and 980℃. With the increase in temperature, 60 61 62 63 17 64 65 1 2 3 4 superplasticity increases in general while the material’s yield strength decreases due to decrease in flow 5 6 stress of the material, allowing for greater deformation. As we increase the temperature beyond 960℃, 7 oxidation effects become dominant and lead to a lower fracture strain at 980°. It is also observed that at 8 960℃, a strain rate of 0.0005s-1 allows the largest deformation and hence, Ti–22Al–24Nb alloy has the 9 10 optimal superplasticity at a temperature of 960 ℃ and strain rate of 0.0005 s−1. 11 12 13 3. As seen above in the various procedures, clamping of the Titanium alloy sheets along their periphery 14 during blow forming results in thinning of the clamped edges [16]. D0 is the original diameter of the circular 15 sheets while Ds is the diameter of the manufactured sphere. It was found out that shells which were 16 17 formed with sheets having relative rolling directions as 45° and 90° had curved welds, i.e. the welds did 18 not lie in the equatorial plane. This resulted in non-uniform plastic strains due to different length of 19 meridians, causing large deviations from the desired spherical shape. It was then concluded that 20 21 symmetrical arrangement of sheets (0° relative rolling direction) was required for desired results. As seen 22 in Table 5, for VT6S of 1 mm thickness, the D0/Ds ratio is 1.21 implying that the sheets ‘possess’ 23 approximately 21% of plasticity resource to account for this reduction in diameter. 24 25 26 During the SPF, the equatorial ring containing the welded joint might lose its circular shape. Folds begin 27 28 to appear in this equatorial zone during the early stages of the process but if the initial sheet thickness is 29 small, these folds may disappear towards the end. If thick sheets or low plasticity welds are used, wrinkles 30 continue to exist in the final formed shell. Thus, elimination of wrinkles and folds is of practical interest 31 32 where high finish of the product is required. 33 34 4. Diffusion bonding combined with superplastic blow forming is the most widely used process for Ti-6Al- 35 36 4V SPF as it leads to the formation of a honeycomb structure which is otherwise not producible [29]. This 37 leads to a significant reduction in cost and weight (around 30-50%) of the formed part and finds 38 application in complex sandwich-like structures with multiple layers of the alloy. 39 40 41 To find the optimal conditions for diffusion bonding, variation of shear strength of the formed joint with 42 43 bonding temperature, pressure and time is studied. From Figure 24(a), it is evident that there is in fact an 44 optimal temperature range (1190-1210 K) for which the shear strength is maximum. This can also be 45 observed in the SEM images taken at 1123 K and 1203 K. It is seen that micro-plastic deformation at the 46 47 material interface is insufficient in the former case while in the latter, due to increased surface area in 48 contact leading to better elemental diffusion, the shear strength increases by almost 1.5 times. 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 18 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Figure 23: SEM images taken at (a) 1123 K and (b) 1203 K showing that micro-plastic deformation is insignificant at 20 lower temperatures. Credits: [29] 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 Figure 24: Variation of shear strength and bonding (a) temperature (b) pressure (c) time during diffusion bonding 57 combined with superplastic blow forming. Credits: [29] 58 59 60 61 62 63 19 64 65 1 2 3 4 Further, as higher pressure leads to increase in contact surface area, we expect an increase in the shear 5 6 strength as in Figure 24(b). This phenomenon can also be seen via the graph plotted over the 5-10 MPa 7 range. Doubling the pressure from 5 MPa to 10 MPa leads to an increase in shear strength of 20 MPa. 8 Hence, effect of pressure on diffusion bonding as compared to the effect of temperature is not as large 9 10 as observed in Figure 25. It should be noted that indefinitely increasing the pressure will lead to large 11 plastic deformations, causing loss of precision in the structure. 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 (a) (b) 29 Figure 25: Microstructure at (a) 7.5 MPa and (b) 10 MPa at constant temperature showing that the effect of pressure 30 is incomparable to the effect of temperature. Credits: [29] 31 32 Holding time has a secondary effect on the shear strength of the joint since it cannot alter the surface 33 area in contact. Therefore, increasing holding time at lower temperatures (1163 K) does not increase the 34 35 shear strength of the joint despite increasing the elemental diffusion of the alloy. However, at higher 36 temperatures (1203 K), due to an increase in surface area in contact, increase in holding time increases 37 the shear strength of the joint. This phenomenon resulted in an optimum time of 30 min at 1203 K and 10 38 39 MPa pressure as observed in Figure 24(c) 40 41 42 5. For both the isothermal and cycling experiments, the unreinforced dome underwent a larger 43 deformation than the reinforced dome [23]. Density measurements were performed by the Archimedes 44 method and it was concluded that there was no significant change in these values. For similar total 45 46 deformation times, it was found that thermal cycling resulted in significantly larger deformations than 47 isothermal holding for both the unreinforced alloy and the Ti-6Al-4V/TiC composite in Figure 26 (a) and 48 (b) respectively. It was also observed that total strain to dome fracture and the deformation rate are more 49 50 in the unreinforced alloy than in the composite in spite of the larger number of cycles in the latter. The 51 three principal strains were plotted as function of original distance from the disk centre for both the set 52 53 of specimens in Figure 27. These strain values differ significantly for each set separately due to gravity, 54 leading to an uneven dome shape. 55 56 57 58 59 60 61 62 63 20 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 (a) (b) 21 Figure 26: Principle strains measured at different locations from the centroid for (a) Ti-6Al-4V and (b) reinforced Ti- 22 6Al-4V/TiC. Credits: [23] 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 Figure 27: Deformed domes. 1 - Ti-6Al-4V cycled (left top) and isothermal (right top); 2 – Composite Ti-6Al-4V/TiC 44 cycled (left middle) and isothermal (right middle); 3 – Samples cycled to fracture: Ti-6Al-4V (left bottom) and Ti-6Al- 45 4V/TiC (right bottom). Credits: [23] 46 47 48 It has been found after intensive study and experimentation that the microstructure of the alloy being 49 50 superplastically formed has a vast impact on the SPF temperature and surface finish of the product. To 51 understand these impacts of microstructure on SPF, the next section has been dedicated to explaining the 52 influence of microstructure on SPF including the commonly used methods to modify the microstructure, 53 54 the mechanism of SPF at the microstructure level, various characterization techniques used and results 55 obtained by multiple research groups. 56 57 58 59 60 61 62 63 21 64 65 1 2 3 4 5 4 Microstructure 6 7 8 Superplastic forming of Ti-6Al-4V is a high temperature forming process for which the temperatures go 9 up to about 1000-1200℃. After a lot of studies and experimentation, it has been found that the 10 microstructure has a huge influence on the superplasticity of the alloy. Hence, further studies have been 11 12 carried out to understand the exact effect of microstructure on the same. Currently, a vast amount of 13 research is focused on modifying the microstructure to lower the superplastic forming temperatures. 14 Studies are also being undertaken on reducing the oxidation of titanium alloy at such high temperatures 15 16 and maintaining a good surface finish, along with uniform thickness of the superplastically formed 17 component. As we further proceed into this section of the paper, we will see more about the various 18 aspects relating to microstructure and superplasticity. 19 20 21 4.1 Methods to modify the microstructure and grain size 22 23 The grain size of the commercially procured Ti-6Al-4V alloy has to be reduced down to be submicron size 24 to achieve relatively lower temperatures for superplastic forming. This modification is done in a different 25 26 way by every researcher. A few common ways to reduce the grain size are hot rolling, equal channel 27 angular pressing, high pressure torsion, multi axial forging and hot pressing. 28 29 30 4.1.1 Hot forging 31 A rod of Ti-6Al-4V alloy was procured. It was isothermally forged in a multistep process in the temperature 32 range of 700-600℃. The forged alloy was then flat rolled at 650℃ down to a thickness of 0.8mm. To retain 33 34 the submicron-sized grains and to produce a sheet with isotropic properties, a low temperature range 35 was chosen. This process led to a grain size of about 0.4µm which is a submicron grain size and showed a 36 -3 -1 37 peak elongation of 1200% at a temperature of 750℃ and a strain rate of 7x10 s [2]. The results obtained 38 in this method can be observed in Table 6. 39 40 Table 6: SPF results of hot forged Ti-6Al-4V sheet. Credits: [2] 41 42 T (℃) Strain Flow stress at 50% of Flow stress at 100% of Elongation (%) Coefficient (m) 43 rate (s-1) strain strain 44 0° 45° 90° 0° 45° 90° 0° 45° 90° 0° 45° 90° 45 7 × 10-4 70 71 71 74 74 74 780 830 810 0.42 0.45 0.44 46 650 7 × 10-3 179.5 180 184 187 186.5 190 720 660 680 0.36 0.32 0.34 3 × 10-2 315 317 309.5 307.5 310 297.5 250 230 270 0.22 0.21 0.24 47 7 × 10-4 44 48 46 54 57 55 900 910 900 0.51 0.51 0.52 48 700 7 × 10-3 110.5 111 118 100 104.5 104 890 900 900 0.47 0.47 0.47 49 3 × 10-2 165 165 165 160 160 159 550 570 600 0.40 0.40 0.42 50 7 × 10-4 27 26 29 37 36 40 1000 970 960 0.65 0.62 0.62 750 7 × 10-3 95 85 80.5 95 95 94 1200 1000 1100 0.51 0.49 0.51 51 3 × 10-2 151.5 146 146.5 152 148 148 500 520 500 0.41 0.43 0.41 52 53 54 4.1.2 High-Pressure Torsion (HPT) 55 Several reports show that High-Pressure Torsion is the optimum technique to produce the smallest grain 56 57 sizes for superplasticity of Ti-6Al-4V. Nanocrystalline Ti-6Al-4V alloys have been processed using this 58 technique [3]. A solution of the alloy is annealed at a temperature of about 1250K and then cooled at 59 various rates to obtain either alpha martensitic or alpha-beta lamellar microstructure. The alloy after 60 61 62 63 22 64 65 1 2 3 4 initial heat treatment is further processed to produce disks of thickness 8mm and diameter 10mm. This is 5 6 done by HPT at room temperature using a rotation speed of 1 rpm through a total of 20 revolutions at an 7 applied pressure of 6 GPa under quasi-constrained conditions. Different kinds of microstructures yield 8 different values of hardness and final microstructures (after the superplastic forming). It was observed 9 10 that the alpha microstructure has smaller grains and higher hardness values than the alpha-beta lamellar 11 microstructure. The initial microstructure also defines the volume fractions of boundaries that act as 12 13 nucleation sites for rapid grain fragmentation and subgrain formation during the further forming steps. 14 The strains observed with different microstructures at different temperatures can be seen in the table 15 below (Table 7): 16 17 18 Table 7: Strains in HPT sheets of Ti-6Al-4V after SF. Credits: [3] 19 20 Initial Microstructure Temperature (K) Strain rate (s-1) δ (%) 21 α + lamellar (α + β) 923 1 × 10-3 575 22 923 1 × 10-3 568 23 α + lamellar (α + β) 998 1 × 10-2 504 24 998 1 × 10-3 676 25 α + 75% lamellar (α + β) 873 1 × 10-3 540 26 873 1 × 10-4 790 27 lamellar (α + β) 973 1 × 10-2 410 28 973 1 × 10-3 690 29 923 1 × 10-2 500 30 Martensite α’ 973 1 × 10-2 610 31 973 1 × 10-3 815 32 33 34 4.1.3 Rapid annealing 35 A commercially available grade of Ti-6Al-4V can be rapidly annealed into beta phase region by using a 36 direct resistance heating technique [7]. Thermocouples are placed at the middle of the specimen to 37 38 monitor the temperature. After a redefined temperature is reached, the power is switched off and the 39 specimen is allowed to cool in atmospheric air. Rapid annealing leads to transformation of the 40 microstructure with acicular alpha grains in beta matrix. This technique does not vary the beta grain size 41 42 by much and they remain about 10-15µm in size, which is finer than the size obtained by usual processing 43 techniques. The dynamic recrystallization undergone during the rapid annealing can be seen in the stress- 44 45 strain curves shown in Figures 27 and 28: 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 23 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Figure 28: Stress-strain curves a) strain rate of 10-3 s-1 and temperature of 1025°C b) strain rate of 10-4 s-1 and 19 temperature of 1070°C. Credits: [7] 20 21 22 23 24 4.2 Influence of microstructure on superplasticity 25 The microstructure of the alloy plays a vital role in the superplasticity of Ti-6Al-4V. Studies have shown 26 27 that the beta phase volume has a huge influence on the superplastic behaviour of the alloy. The beta 28 phase of Ti-6Al-4V is the deformable part of the microstructure grain. Superplastic formation is obtained 29 by the beta phase, inhibiting grain growth and thus promoting the sliding of grain boundaries. Usually, a 30 31 lamellar type of microstructure is preferred for superplastic forming due to the presence of a significant 32 amount of beta phase unlike the case of alpha martensitic. The only drawback of having higher volume of 33 beta phase is that it reduces the hardness and strength of the alloy. Alpha phase is responsible for the 34 35 above-mentioned mechanical properties. 36 37 The superplastic forming process is a high temperature forming process. Hence, there is a possibility of 38 39 oxidation of the surface, which is undesirable [11]. Experimental results have shown that fine grain 40 microstructures have a higher tendency to absorb oxygen than the coarse grain material. This leads to the 41 ℃ 42 formation of multiple oxide layers on the surface, usually above 700 . These layers primarily consist of 43 the oxides of aluminium and titanium and the number of layers also increase with increased exposure to 44 time and temperature. 45 46 47 This increased level of oxygen can be associated with the increased micro hardness and reduced ductility 48 in the alloy [44]. The diffusion starts in the direction parallel through the alpha and beta grains and along 49 50 the grain boundaries. As the concentration of oxygen increases above that, there is a need to convert beta 51 phase to alpha phase, and a surface layer known as the alpha case is formed. This results in a brittle, 52 oxidised material. The oxygen content tends to increase with an increase in alpha titanium grains, as they 53 54 are known to be oxygen stabilisers. To prevent the said problem due to the interaction with oxygen, the 55 processing is done either in an inert atmosphere or coatings are used on the alloy to prevent surface 56 57 interaction with oxygen. Sodium silicate and Delta Glaze 3410 are tested anti-oxidant coatings used for 58 titanium alloys. These prevent the oxidation of the surface and also lead to a better surface finish along 59 with better mechanical properties. The formed part with and without the glaze can be seen in Figure 29. 60 61 62 63 24 64 65 1 2 3 4 After a lot of studies, it has been concluded that the elongation to failure in tensile testing generally 5 6 increases with a decrease in gauge length. 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Figure 29: Surface finish of formed part, with coating (left) and without coating (right). Credits: [11] 22 23 24 25 26 4.3 Mechanisms of superplasticity 27 The material is said to show hardening initially, which is followed by a peak stress and gradual softening 28 [3]. A significant amount of residual stress, dislocations and stacking faults, along with shear 29 30 transformation (as seen in Figure 30) and supersaturated chemical composition is observed at the end of 31 the superplastic forming process. At higher deformations, dislocation slip and phase transformations 32 33 could get activated to accommodate the deformation. One of the major transformations during the 34 process could be that of HCP to FCC. The percentage elongation depends a lot on the grain boundary 35 sliding that can occur in the particular sample taken for superplastic forming [43]. At very low strain rates, 36 37 negligible dynamic coarsening of the microstructure is observed. The beta-phase precipitation after 38 holding at the testing temperature acts as an accommodating mechanism and enhances superplasticity; 39 even a small amount of beta-phase contributes to superplasticity by inhibiting grain growth and 40 41 promoting grain boundary sliding. From this, it can be assumed that the volume fraction of beta phase 42 determines the dominant flow mechanism. The superplastic flow also depends on the crystal lattice as 43 various FCC and HCP lattices have different number of slip systems, as they help inhibit grain growth and 44 45 promote grain boundary sliding (FCC has higher number of slip systems and undergoes intragranular slip 46 to accommodate grain boundary sliding). 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 25 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Figure 30: Microstructure of the tensile test sample observed under an optical microscope a) at gauge b) at grip. Credits: 25 [3] 26 27 28 The microstructural evolution takes place at different rates across the specimen until it becomes almost 29 30 homogenous [5]. The initial processing done for microstructural refinement produces a high density of 31 dislocations and dislocation tangles. Here, the microstructural flow and presence of shear bands are 32 revealed using Selected Area Electron Diffraction (SAED). With increasing shear in the pre-processing 33 34 stages, the SAED patterns become more ring-like, showing grain misorientations and evolution of high 35 angle grain boundaries. During the superplastic forming, strain hardening is observed at higher strain rates 36 and superplastic flow with low stable flow stresses is observed at low strain rates. From the limited set of 37 38 data, it has been observed that the strain hardening constant (m) increases and then decreases with 39 increasing strain rate (Figure 31). It has also been observed that grain boundary sliding in superplasticity 40 41 requires m = 0.5. Common to all the studies, an alpha to beta phase transformation takes place in the 42 alloy during prolonged exposure at the tensile deformation temperature (>750℃). It has also been 43 concluded that high angle grain boundaries are crucial for the flow via extensive grain boundary sliding. 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 26 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Figure 31: Stress-strain curves at varied strain rates. Credits:[5] 23 24 25 26 The superplastic deformation is a result of dislocation gliding, climbing and grain boundary sliding [1]. At 27 the superplastic forming temperature (below 800℃), the mechanism is dominated by the accumulation 28 of dislocations as the thermal activation of grain boundary sliding gets difficult. This increase in dislocation 29 30 density leads to dislocation tangling and subgrain formation. At a relatively low temperature, dynamic 31 recovery can act as an accommodating mechanism. The dislocations at this temperature could be edge 32 33 dislocations, screw dislocations, de-pinning of dislocation or counteraction of dislocation to cancel each 34 other out. For temperatures above 800℃, superplastic forming occurs through grain boundary sliding 35 accompanied by dynamic recrystallization. This is due to the matching of the activation of energy of 36 37 superplastic forming with the grain boundary self-diffusion energy. Most of the dislocations are observed 38 at the alpha-beta phase boundary and not the alpha-alpha phase boundary. Thus, it can be concluded that 39 the superplastic forming of two-phase Ti-6Al-4V vastly depends on the alpha-beta phase ratio. Moreover, 40 41 beta phase is soft and acts as a lubricant during the deformation providing stress relaxation. The 42 aforementioned microstructural changes can be observed in the TEM images shown in Figure 31: 43

44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 27 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 Figure 32: TEM images of tensile test samples (a & b) - 700°C, (c & d) - 750°C, (e & f) - 800°C, (g & h) - 850°C. 54 Credits: [1] 55 56 57 58 After years of research, it has been concluded that very fine grain size and high temperature are the 59 required conditions to achieve superplastic behaviour [19]. However, these two work opposite to each 60 61 62 63 28 64 65 1 2 3 4 other due to the grain growth observed at elevated temperatures. Most of the researchers have 5 6 concluded that grain boundary sliding is mechanism by which superplastic forming can be attained, but it 7 is still unclear if the grain boundary sliding is diffusion-accommodated, dislocation-accommodated or a 8 combination of both. Cavity formation could also be a part of the mechanism for superplasticity. Hence, 9 10 it is fair to assume that there are multiple mechanisms behind superplastic forming. The only certain thing 11 is that two-phase microstructure is critical for superplastic forming. These effects of microstructural 12 13 changes on the stress-strain plot can be seen in Figure 33: 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 Figure 33: Effects of microstructural changes on the stress-strain curve. Credits: [19] 34 35 36 37 4.4 Microstructure study 38 39 The study of microstructure is one of the most crucial aspects in superplastic forming. As we know, there 40 41 are various techniques to observe the microstructure and hence, draw some conclusions which give a 42 greater insight into the microstructural changes during the process. The author refers the readers to the 43 table below to understand the various methods for microstructure study. 44 45 46 47 S.No. Technique used References 48 49 50 1 Optical Microscopy 3; 10; 11; 12; 13; 14; 15; 17; 18; 19; 20; 51 52 2 Scanning Electron Microscopy 1; 10; 12; 16; 18; 53 54 3 Transmission Electron Microscopy 1; 2; 3; 5; 9; 13; 55 56 4 Orientation Imaging Microscopy 1; 8; 19; 57 58 59 5 Vicker’s Microhardness 3; 60 61 62 63 29 64 65 1 2 3 4 5 6 Electron Back Scattered Diffraction 1; 3; 5; 8; 10; 12; 17; 18; 19; 6 7 7 X-Ray Diffraction 2; 3; 5; 17; 8 9 10 8 Thermal Field Emission Scanning 3; 11 Electron Microscope 12 13 9 Selected Area Electron Diffraction 5; 14 15 16 10 Inductively Coupled Plasma 6; 17 Spectroscopy 18 19 11 Knoop Hardness 6; 20 21 12 Ultrasonic C-Scan 18; 22 23 24 25 4.5 Experimental results 26 27 28 The obtained experimental results are given by the table below: 29 30 31 S.No. Grain size Temperature Strain rate Max. Elongation Reference 32 (℃) (s-1) (%) 33 34 35 -3 36 1 2.5µm 800 1x10 862 1 37 38 39 2 0.4µm 750 7x10-3 1200 2 40 41 42 43 3 40nm 700 1x10-3 815 3 44 45 46 47 48 4 1µm 775 4 49 50 51 5 40nm 650 5x10-4 820 5 52 53 54 55 56 -4 57 6 3µm 775 1x10 850 8 58 59 60 61 62 63 30 64 65 1 2 3 4 -3 5 7 0.5µm 800 1x10 700 11 6 7 8 8 649 37 18 9 10 11 12 13 9 10µm 800, 850, 900, 1x10-4 >300 19 14 15 950 16 17 18 19 The work on superplastic forming of Ti-6Al-4V alloy has been carried out by many around the world. A 20 few of the works on SPF/DB have been patented by the creators. A peek into those patents has been 21 provided in the section below. 22 23 24 25 26 27 5 Patents on SPF/DB 28 29 30 A study of SPF/DB found that a combination of superplasticity and diffusion bonding is performed at low 31 pressures [31]. An innovative approach is described, which uses sub-atmospheric pressures between the 32 33 to-be formed sheets instead of externally applied force through hydraulic press. For fine grained Ti-6Al- 34 4V (grain size less than 2 micrometer), an approach to conduct SPF/DB at a reduced temperature and 35 increased strain rate is suggested, thereby decreasing the cost of the process [32]. 36 37 38 The work on SPF and its applications does not end here. This is a field with a lot of potential and scope for 39 40 the future. This is due to its nature, as it can replace the conventional forming processes once the 41 technique has been optimised and perfected. The work on SPF has been going on for about 30 years now, 42 and there have been drastic developments in the methodology in these three decades. After a careful 43 44 study of the research work in this field, a few potential areas of advancements and applications have been 45 brought out in the next section. This will help uplift further research and developments in the said field. 46 47 48 49 6 Scope 50 51 52 For future work, a combination of higher applied stresses and faster temperature cycling in the case of 53 mix-match SPF may allow for an increase in the average strain rate by about an order of 10 [23]. This 54 55 implies that it would be possible to achieve higher deformation rates similar to those achievable in 56 optimized Ti-6Al-4V materials having microstructural superplasticity. Thus, the transformation mix-match 57 SPF could become a well-sought-after forming method for Titanium alloys which are difficult to forge, 58 59 machine or cast. 60 61 62 63 31 64 65 1 2 3 4 The superplasticity of Ti-6Al-4V alloy is improved by additions of traces of hydrogen which decrease the 5 6 dislocation density, promote dislocation motion and facilitate Beta phase flow [13]. 7 8 Rapid Beta heat treatment and the microstructural gradient technique can be implemented in a controlled 9 way to manipulate the local thickness straining [7]. Their feasibility will be improved if the problem at the 10 11 interface zone has been well addressed, as well as better prediction and control of applying the gradients, 12 even though it may be difficult in practice. There is also a remote possibility that if a precise prediction 13 and control of the microstructural gradients can be achieved, highly complex shapes might be even 14 15 formed by free-bulging without a die; i.e. die-less forming. 16 17 18 From an experimental viewpoint, multi-step dies lead to wrinkling, which although removable through 19 “ironing” in the second step, leads to a decrease in the structural strength of the alloy [33]. In addition, 20 adopting a multi-stage die with a gradient in the coefficient of friction results in more uniformity when 21 22 compared to single-step dies [28]. Further research into Finite Element Analysis (FEA)-guided 23 manufacturing of multi-step dies is required so as to optimise the design such that no wrinkles are present 24 in the final die. Experimental conditions given indicate that the optimal conditions for SPF are 900-980℃ 25 26 [30] and 5-10 MPa [29], with strain rates between 0.001/s and 0.0001/s [19]. Moreover, optimal pressure 27 conditions decrease the coefficient of friction between the die and alloy [27]. The optimal pressure and 28 temperature curves are somewhat inverted bell-shaped. Although increasing the SPF pressure optimises 29 30 the process, indefinitely increasing the temperature and pressure leads to loss in structural integrity, 31 causing deformations and wrinkles in the structure. 32 33 34 Forging parameters need to optimised as per the above discussion, which is possible via microstructure 35 refinements and a deeper understanding of grain formation during SPF. Further research in achieving sub- 36 37 micro grained Ti-6Al-4V should be done as shown by [8] that such Ti-6Al-4V showed more deformation, 38 hence, more elongation as compared to traditional Ti-6Al-4V. Furthermore, [34] showed that sub- 39 microcrystalline (SMC) structured Ti-4Al-4V underwent a relatively more uniform cross-sectional strain 40 41 than commercial alloys with microcrystalline (MC) structure. However, SPF temperature for commercial 42 grade Ti-6Al-4V is spread over a wider range (700-1100℃) [35] and occurs at a lower strain rate than that 43 mentioned in [29]. Therefore, further research in obtaining desired physical properties of SPF, as in [2], 44 45 at lower temperatures (700℃) need to be investigated. For instance, as [9] and [1] found, nanometric 46 grains and narrow grain distribution lead to lowering of optimum SPF temperature. Recrystallisation is the 47 main microstructural process which occurs throughout the SPF [17]. Hence, further studies to identify the 48 49 factors which lead to optimal recrystallisation is required. Furthermore, quantitative models to predict 50 thickness distribution based on initial microstructure (along with its evolution through the process) need 51 52 to be built to correlate grain size and phase concentrations for a generalization of the constitutive 53 behavior of the SPF process [37]. Also, [6] extensively showed that fine-grained Ti-6Al-4V showed greater 54 oxygen diffusion rates in the areas of α-case thickness and microhardness depth as compared to the 55 56 commercial counterpart. However, this also means that present manufacturing techniques need to be 57 revised to account for this increased diffusion while accurately studying the effects of this oxygen 58 enrichment on the physical properties of the alloy. 59 60 61 62 63 32 64 65 1 2 3 4 Microstructural reasons of high SPF are given as suppression of porosity and high stability of the grain 5 6 formation regions [38]. Therefore, research into replicating these properties in desired metals could be 7 an interesting avenue for future work, leading to commercial SPF of other Titanium alloys. Other major 8 factors which influence coefficient of friction and strain rate wherein low strain rate resulted in more 9 10 uniform thickness distribution of the formed alloy [39]. Thickness non-uniformity, as shown by [17], leads 11 to stress concentrations causing localized thinning and structural integrity degradation. Hence, a theory 12 13 specifying selective heating areas to avoid such non-uniformity based on the geometry of the forming 14 setup is required. 15 16 17 18 7 Conclusions 19 20 21 This paper presents a review of the advancements in superplastic forming of Titanium and its alloys over 22 23 the last few decades. It provides a detailed description of the various experimental techniques employed 24 to form Titanium into sheet metal and includes various modifications such as multi-step dies and 25 superplastic forming followed by diffusion bonding. A comparative study of process parameters has also 26 27 been done to optimise SPF for better results, followed by a review of the microstructure and its influence 28 on the SPF process. The purpose of this is to elaborate on the different phenomenon that occur at the 29 microstructural level and then to vet the different ways to refine the microstructure. A brief explanation 30 31 of the mechanisms to accommodate high plastic strains along with methods to study the microstructure 32 and a few experimental results have been included in the paper. Finally, further scope for research in SPF 33 of Ti-6Al-4V has been provided. 34 35 36 37 38 References: 39 40 41 1. Zhang, T., Liu, Y., Sanders, D., Liu, B., Zhang, W. and Zhou, C. (2014). Development of fine-grain 42 43 size titanium 6Al–4V alloy sheet material for low temperature superplastic forming. Materials 44 Science and Engineering: A, 608, pp.265-272. 45 2. Salishchev, G., Galeyev, R., Valiakhmetov, O., Safiullin, R., Lutfullin, R., Senkov, O., Froes, F. and 46 47 Kaibyshev, O. (2001). Development of Ti–6Al–4V sheet with low temperature superplastic 48 properties. Journal of Materials Processing Technology, 116(2-3), pp.265-268. 49 3. Shahmir, H., Naghdi, F., Pereira, P., Huang, Y. and Langdon, T. (2018). Factors influencing 50 51 superplasticity in the Ti-6Al-4V alloy processed by high-pressure torsion. Materials Science and 52 Engineering: A, 718, pp.198-206. 53 54 4. Hefti, L. (2010). Fine-grain titanium 6AI-4V for superplastic forming and diffusion bonding of 55 aerospace products. JOM, 62(5), 42-45. doi: 10.1007/s11837-010-0076-9 56 5. Zhang, W., Ding, H., Pereira, P., Huang, Y., & Langdon, T. (2018). Grain refinement and superplastic 57 58 flow in a fully lamellar Ti-6Al-4V alloy processed by high-pressure torsion. Materials Science And 59 Engineering: A, 732, 398-405. doi: 10.1016/j.msea.2018.07.010 60 61 62 63 33 64 65 1 2 3 4 6. Pitt, F., & Ramulu, M. (2004). Influence of Grain Size and Microstructure on Oxidation Rates in 5 6 Titanium Alloy Ti-6Al-4V Under Superplastic Forming Conditions. Journal Of Materials Engineering 7 And Performance, 13(6), 727-734. doi: 10.1361/10599490421394 8 7. Rhaipu, S. (1998). The effect of microstructural gradients on superplastic forming of Ti–6Al–4V. 9 10 Journal Of Materials Processing Technology, 80-81, 90-95. doi: 10.1016/s0924-0136(98)00179-4 11 8. Patankar, S., Escobedo, J., Field, D., Salishchev, G., Galeyev, R., Valiakhmetov, O., & Froes, F. 12 13 (2002). Superior superplastic behavior in fine-grained Ti–6Al–4V sheet. Journal Of Alloys And 14 Compounds, 345(1-2), 221-227. doi: 10.1016/s0925-8388(02)00406-1 15 9. Lutfullin, R., Kruglov, A., Safiullin, R., Mukhametrakhimov, M., & Rudenko, O. (2009). Processing 16 17 properties of nano- and submicro-crystalline Ti–6Al–4V titanium alloy. Materials Science And 18 Engineering: A, 503(1-2), 52-54. doi: 10.1016/j.msea.2008.02.060 19 10. ZOU, C., ZHOU, Q., & WANG, L. (2018). A step deformation method for superplasticity 20 21 improvement of coarse-grained Ti–15V–3Cr−3Sn−3Al. Chinese Journal Of Aeronautics, 31(7), 22 1619-1624. doi: 10.1016/j.cja.2017.11.003 23 11. Patankar, S., & Tan, M. (2001). Influence of Alpha Casing on Superplastic Deformation of Ti-6Al- 24 25 4V. Journal Of Engineering Materials And Technology, 123(1), 144. doi: 10.1115/1.1325007 26 12. Li, X., Chen, N., Wu, H., Chen, J., & Qu, F. (2018). Low-temperature superplastic gas bulging of Ti- 27 28 55 alloy by hydrogen addition. International Journal Of Hydrogen Energy, 43(27), 12455-12459. 29 doi: 10.1016/j.ijhydene.2018.04.138 30 13. Shaoqing, Z., & Linruo, Z. (1995). Effect of hydrogen on the superplasticity and microstructure of 31 32 Ti-6Al-4V alloy. Journal Of Alloys And Compounds, 218(2), 233-236. doi: 10.1016/0925- 33 8388(94)01384-5 34 14. Seshacharyulu, T., Medeiros, S., Frazier, W., & Prasad, Y. (2000). Hot working of commercial Ti– 35 36 6Al–4V with an equiaxed α–β microstructure: materials modeling considerations. Materials 37 Science And Engineering: A, 284(1-2), 184-194. doi: 10.1016/s0921-5093(00)00741-3 38 15. Sastry, S., Lederich, R., Mackay, T., & Kerr, W. (1983). Superplastic Forming Characterization of 39 40 Titanium Alloys. JOM, 35(1), 48-53. doi: 10.1007/bf03338184 41 16. Kruglov, A., Enikeev, F., & Lutfullin, R. (2002). Superplastic forming of a spherical shell out a 42 43 welded envelope. Materials Science And Engineering: A, 323(1-2), 416-426. doi: 10.1016/s0921- 44 5093(01)01376-4 45 17. Liu, J., Tan, M., Aue-u-lan, Y., Guo, M., Castagne, S., & Chua, B. (2013). Superplastic-like forming 46 47 of Ti-6Al-4V alloy. The International Journal Of Advanced Manufacturing Technology, 69(5-8), 48 1097-1104. doi: 10.1007/s00170-013-5101-z 49 18. Bewlay, B., Gigliotti, M., Utyashev, F., & Kaibyshev, O. (2000). Superplastic roll forming of Ti alloys. 50 51 Materials & Design, 21(4), 287-295. doi: 10.1016/s0261-3069(99)00089-8 52 19. Alabort, E., Putman, D., & Reed, R. (2015). Superplasticity in Ti–6Al–4V: Characterisation, 53 modelling and applications. Acta Materialia, 95, 428-442. doi: 10.1016/j.actamat.2015.04.056 54 55 20. Ambrogio, G., Sgambitterra, E., De Napoli, L., Gagliardi, F., Fragomeni, G., & Piccininni, A. et al. 56 (2017). Performances Analysis of Titanium Prostheses Manufactured by Superplastic Forming and 57 Incremental Forming. Procedia Engineering, 183, 168-173. doi: 10.1016/j.proeng.2017.04.057 58 21. Ambrogio, G., Palumbo, G., Sgambitterra, E., Guglielmi, P., Piccininni, A., & De Napoli, L. et al. 59 (2018). Experimental investigation of the mechanical performances of titanium cranial prostheses 60 61 62 63 34 64 65 1 2 3 4 manufactured by super plastic forming and single-point incremental forming. The International 5 Journal Of Advanced Manufacturing Technology, 98(5-8), 1489-1503. doi: 10.1007/s00170-018- 6 7 2338-6 8 22. Sorgente, D., Palumbo, G., Piccininni, A., Guglielmi, P., & Aksenov, S. (2018). Investigation on the 9 thickness distribution of highly customized titanium biomedical implants manufactured by 10 superplastic forming. CIRP Journal Of Manufacturing Science And Technology, 20, 29-35. doi: 11 10.1016/j.cirpj.2017.09.004 12 23. Dunand, D., & Myojin, S. (1997). Biaxial deformation of Ti-6Al-4V and Ti-6Al-4V/TiC composites by 13 14 transformation-mismatch superplasticity. Materials Science and Engineering: A, 230(1-2), 25-32. 15 doi:10.1016/s0921-5093(97)00033-6 16 24. Wang, G., Zhang, K., Wu, D., Wang, J., & Yu, Y. (1970, January 01). Superplastic forming of 17 bellows expansion joints made of titanium alloys. Retrieved from 18 https://www.infona.pl/resource/bwmeta1.element.elsevier-3361ec5d-e861-385c-9108- 19 88ccd0ac0c25 20 21 25. Jiang, S., & Zhang, K. (2009). Superplastic forming of Ti6Al4V alloy using ZrO2-TiO2 ceramic die 22 with adjustable linear thermal expansion coefficient. Transactions of Nonferrous Metals Society of 23 China, 19. doi:10.1016/s1003-6326(10)60081-7 24 26. Jiang, S., & Zhang, K. (2009). Study on controlling thermal expansion coefficient of ZrO2–TiO2 25 ceramic die for superplastic blow-forming high accuracy Ti–6Al–4V component. Materials & Design, 26 30(9), 3904-3907. doi:10.1016/j.matdes.2009.03.023 27 28 27. Balasubramanian, M., Ramanathan, K., & Kumar, V. S. (2013). Mathematical Modeling and Finite 29 Element Analysis of Superplastic Forming of Ti-6Al-4V Alloy in a Stepped Rectangular Die. 30 Procedia Engineering, 64, 1209-1218. doi:10.1016/j.proeng.2013.09.200 31 28. JIANG, S., LU, Z., HE, X., WANG, G. and ZHANG, K. (2012). Superplastic forming Ti–6Al–4V 32 titanium alloy cylinder with near uniform thickness distribution. Transactions of Nonferrous Metals 33 Society of China, 22, pp.s472-s478. 34 35 29. Han, W., Zhang, K. and Wang, G. (2007). Superplastic forming and diffusion bonding for 36 honeycomb structure of Ti–6Al–4V alloy. Journal of Materials Processing Technology, 183(2-3), 37 pp.450-454. 38 30. Wang, C., Zhao, T., Wang, G., Gao, J. and Fang, H. (2015). Superplastic forming and diffusion 39 bonding of Ti–22Al–24Nb alloy. Journal of Materials Processing Technology, 222, pp.122-127. 40 31. Woodward, J. R. (1981). U.S. Patent No. US4429824A. Washington, DC: U.S. Patent and 41 42 Trademark Office. 43 32. Comley, P. N., & Heftl, L. D. (2009). U.S. Patent No. US 7,533,794 B2. Washington, DC: U.S. 44 Patent and Trademark Office. 45 33. Jr., G. L., Friedman, P., & Weinmann, K. (2009). Design and experimental validation of a two-stage 46 superplastic forming die. Journal of Materials Processing Technology, 209(4), 2152-2160. 47 doi:10.1016/j.jmatprotec.2008.05.019 48 49 34. Kaibyshev, O., Safiullin, R., Lutfullin, R., Valiakhmetov, O., Galeyev, R., & Dutta, A. et al. (2006). 50 Advanced superplastic forming and diffusion bonding of titanium alloy. Materials Science And 51 Technology, 22(3), 343-348. doi: 10.1179/174328406x83932 52 35. Prasad, Y., Seshacharyulu, T., Medeiros, S., & Frazier, W. (2001). Influence of oxygen content on 53 the forging response of equiaxed (α+β) preform of Ti–6Al–4V: commercial vs. ELI grade. Journal 54 Of Materials Processing Technology, 108(3), 320-327. doi: 10.1016/s0924-0136(00)00832-3 55 56 36. Salishchev, G., Galeyev, R., Valiakhmetov, O., Safiullin, R., Lutfullin, R., & Senkov, O. et al. (2001). 57 Development of Ti–6Al–4V sheet with low temperature superplastic properties. Journal Of 58 Materials Processing Technology, 116(2-3), 265-268. doi: 10.1016/s0924-0136(01)01037-8 59 60 61 62 63 35 64 65 1 2 3 4 37. Guo, M., Liu, J., Tan, M., & Chua, B. (2014). Microstructure Evolution of Ti-6Al-4V during 5 Superplastic-like Forming. Procedia Engineering, 81, 1090-1095. doi: 6 7 10.1016/j.proeng.2014.10.146 8 38. Imayev, R., Salishchev, G., Shagiev, M., Imayev, V., Gabdullin, N., & Kuznetsov, A. et al. (1999). 9 Low-Temperature Superplasticity of Submicrocrystalline Intermetallics. Materials Science Forum, 10 304-306, 195-200. doi: 10.4028/www.scientific.net/msf.304-306.195 11 39. Chen, Y., Kibble, K., Hall, R., & Huang, X. (2001). Numerical analysis of superplastic blow forming 12 of Ti–6Al–4V alloys. Materials & Design, 22(8), 679-685. doi: 10.1016/s0261-3069(01)00009-7 13 14 40. Grimes, R. (2003). Superplastic forming: evolution from metallurgical curiosity to major 15 manufacturing tool?. Materials Science And Technology, 19(1), 3-10. doi: 16 10.1179/026708303225008725 17 41. Kaibyshev, O., Safiullin, R., Lutfullin, R., Valiakhmetov, O., Galeyev, R., & Dutta, A. et al. (2006). 18 Advanced superplastic forming and diffusion bonding of titanium alloy. Materials Science And 19 Technology, 22(3), 343-348. doi: 10.1179/174328406x83932 20 21 42. Nicholson, R. (1976). Discussion on superplastic-forming operations. Metals Technology, 3(1), 22 167-170. doi: 10.1179/030716976803391944 23 43. Sherby, O., & Wadsworth, J. (1985). Superplasticity and superplastic forming processes. Materials 24 Science And Technology, 1(11), 925-936. doi: 10.1179/mst.1985.1.11.925 25 44. Dawson, A., Blackwell, P., Jones, M., Young, J., & Duggan, M. (1998). Hot rolling and superplastic 26 forming response of net shape processed Ti-6AI-4V produced by centrifugal spray 27 28 deposition. Materials Science And Technology, 14(7), 640-650. doi: 10.1179/mst.1998.14.7.640 29 45. Giuliano, G. (2011). Superplastic forming of advanced metallic materials. Oxford: Woodhead. 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 36 64 65 All images Click here to download Colour figure (online version only) Images.docx

Figure 1: SPF literature focus in the papers

Figure 2: Literature trends in SPF since 1980

Figure 3: Prostheses a) made from incremental forming b) made from superplastic forming.

Figure 4: a) Die setup for SPF of cranial prostheses b) Formed part.

Figure 5: Twin-aisle aircraft heat shield.

Figure 6: Tail cone of a twin-aisle aircraft

Figure 7: Process schematic of sintering ZrO2 – TiO2 cylinder

Figure 8: Schematic structure of SPF die – 1 Formed punch; 2 Female die; 3 Outer sleeve; 4 Heating apparatus; 5 Power lot; 6 Filling piece

(a) (b)

Figure 9: Relationship between linear CTE and (a) volume fraction of TiO2 and (b) relative density

Figure 10: Different volume fraction of TiO2 and corresponding relative density for manufacturing

the optimal ZiO2-TiO2 ceramic die

(a) (b)

Figure 11: (a) Steel model for forming semi-finished ZrO2-TiO2 deep cylinder ceramic die: 1 Pressure head, 2 External mould, 3 Mixed power, 4 Internal mould, 5 Heating apparatus, 6 Screw bolts; (b)

Finished ZrO2-TiO2 deep cylinder ceramic die

Figure 12: (a), (b), (c) and (d) Different stages of pressure blow forming technique in a stepped rectangular die

Figure 13: (a) Components of the superplastic-like forming die assembly (b) Die-cavity with non- isothermal heating and showing % surface area expansion at each step

Figure 14: Complete assembly of an airtight closed chamber for SPF forming

Figure 15: Schematic outline of the high temperature forming chamber. (A) – Top pressure vessel (B) Die housing (C) Bottom pressure vessel (D) Heater

Figure 16: (a) Schematic structure of the die for SPF (b) Experimental Pressure vs Time curves

Figure 17: SPF procedure: (a) original state (b), bulging (c) dies clamping, (d) shape forming, (e) Thickness distribution of as-formed bellows

(a) (b)

(c) Figure 18: (a) Pressure vs time variation for SPF-DB processing (b), Assembly for the processing. 1 – face sheet, 2 – stop off, 3 – core sheet, 4 – stiffening rib, 5 - die (c) Points at which thickness was measured whose values are shown in Table 3

Figure 19: Comparison of finite element simulation and experimental strains for (a) the as-received blank and (b) the blank with transformed structure

Figure 20: Spherical shell formed out of the envelope manufactured by diffusion bonding (D0 = 200

mm, Ds = 164 mm)

(a) (b)

Figure 21: (a) Splitting of die into forming and preforming results in uniform thickness distribution (b) Meshed geometry showing the shape of preforming die

(a) (b)

Figure 22: Variation of engineering stress and engineering strain in SPF for various (a) temperatures (b) strain rates

Figure 23: SEM images taken at (a) 1123 K and (b) 1203 K showing that micro-plastic deformation is insignificant at lower temperatures

Figure 24: Variation of shear strength and bonding (a) temperature (b) pressure (c) time during diffusion bonding combined with superplastic blow forming

(a) (b)

Figure 25: Microstructure at (a) 7.5 MPa and (b) 10 MPa at constant temperature showing that the effect of pressure is incomparable to the effect of temperature

(a) (b)

Figure 26: Principle strains measured at different locations from the centroid for (a) Ti-6Al-4V and (b) reinforced Ti-6Al-4V/TiC

Figure 27: Deformed domes. 1 - Ti-6Al-4V cycled (left top) and isothermal (right top); 2 – Composite Ti-6Al-4V/TiC cycled (left middle) and isothermal (right middle); 3 – Samples cycled to fracture: Ti- 6Al-4V (left bottom) and Ti-6Al-4V/TiC (right bottom)

Figure 28: Stress-strain curves a) strain rate of 10-3 s-1 and temperature of 1025°C b) strain rate of 10- 4 s-1 and temperature of 1070°C

Figure 29: Surface finish of formed part, with coating (left) and without coating (right)

Figure 30: Microstructure of the tensile test sample observed under an optical microscope a) at gauge b) at grip

Figure 31: Stress-strain curves at varied strain rates

Figure 32: TEM images of tensile test samples (a & b) - 700°C, (c & d) - 750°C, (e & f) - 800°C, (g & h) - 850°C

Figure 33: Effects of microstructural changes on the stress-strain curve