Chemical machined thin foils of TiNi shape memory alloy
J.Z. Chen, S.K. Wu
*Institute of Materials Science and Engineering, National Taiwan University, Taipei, Taiwan106 Received 9 November 1998; received in revised form 17 November 1998; accepted 24 November 1998
Abstract
Ti50Ni50 thin foils up to 10 mm thickness are successfully fabricated from 100 mm thickness thin plates using chemical machining.
Ultrasonic agitation is applied to enhance the chemical etching process with mixed solutions of HF/HNO3/H2O. The higher the HF/HNO3
volume ratio in the solution, the higher the etching rate and the smoother the surface will be. Thin foils can also be fabricated with the electropolishing process but with a much slower etching rate. Foil thickness by these processes can only be approximately 10 mm due to the side-etching effect. The martensitic transformation peaks of thin foils shown on the DSC curve are broader for a thinner sample due to the effect of thermal resistance in the DSC sample pans. The enthalpy of transformation also decreases while the foil thickness is reduced due to the effect of the foil surface energy and the energy of plastic deformation. # 1999 Elsevier Science S.A. All rights reserved.
Keywords: TiNi shape memory alloy; Thin foil; Chemical etching; Electropolishing; Thickness effect
1. Introduction
In practical applications, shape memory alloys (SMAs) are usually ®rst mechanically processed into wires or plates and then further fabricated to springs or other shapes [1]. The performance of the shape memory effect pertaining to the SMAs is closely related to the heating/cooling rate of fabricated shapes in the temperature range of Af±Mf. Here,
Mf is the ®nished temperature of the forward martensitic
transformation, and Af is that of the reverse martensitic
transformation. Obviously, the fabricated shapes of SMAs having a larger ratio of surface area/cross-section area can achieve a higher heating/cooling rate and shorten the response time of their shape memory effect. Therefore, from the viewpoint of the response time, how to reduce the cross-section area of fabricated shapes becomes important in the practical applications of SMAs. For example, the thin foils of SMAs can be used as the microactuator because the microactuator of SMAs has a high work output and power density [2±4]. The microactuator device can be fabricated out by the assembly of individual SMA parts [2,5]. However, the thickness of these parts is limited by the conventional mechanical fabrication processes. According to our experi-ence, the conventional cold-rolling/recrystallization process can reduce the thickness of TiNi alloy foil to approximately 30 mm [6]. In this study, thin foils of TiNi alloy, however, can
be thinned to around 10 mm using chemical machining processes. It is well known that the wet-chemical etching solution mixture of hydro¯uoric and nitric acid can be used to electrochemically thin the TiNi foil. However, the details of this thinning process have not yet been reported [5]. In the present paper, the details of the chemical machining of the TiNi alloy is described and the martensitic transformation behavior of acquired thin foils is also discussed.
2. Experimental procedure
The TiNi equiatomic alloy was prepared from the raw materials of titanium (purity 99.7 wt.%) and nickel (purity 99.97 wt.%) using the vacuum arc remelting (VAR) techni-que. The weight loss of the VAR process is less than the order of 0.01%, thus the composition deviation from TiNi can be neglected. The ingot was then hot rolled at 8508C to 1 mm in thickness and wire-cut into pieces of 2 mm 30 mm. The oxide on the surface of samples was removed by chemical etching. Then, the specimen was further thinned to 0.1 mm thickness using the alternate processes of cold rolling and recrystallization annealing. The reduction rate was 30% for each cold rolling process. The inter-annealing (recrystallization annealing) between each cold rolling process was conducted in an evacuated quartz capsule at 7008C for 30 min. The grain size of thin foils after the alternate processes of cold-rolling and
anneal-Materials Chemistry and Physics 58 (1999) 162±165
*Corresponding author.
0254-0584/99/$ ± see front matter # 1999 Elsevier Science S.A. All rights reserved. P I I : S 0 2 5 4 - 0 5 8 4 ( 9 8 ) 0 0 2 7 3 - 9
ing is about 10 mm. The ®nal thickness of thin foils used in this study was chosen between 100 and 10 mm by the process of chemical machining, described as follows.
The chemical thinning process of the TiNi alloy is tested using two methods: the chemical etching with the assistance of ultrasonic agitation, and the electropolishing. There are two chemical solutions, Solutions A and B, for chemical etching. Both of them consist of the solutions of HF, HNO3
and H2O with different volume ratios, in which Solution A is
HF : HNO3: H2O 1 : 4 : 5, and Solution B is 1 : 5 : 20.
For the electropolishing process, the electrolyte is composed of H2SO4 and CH3OH with 1 : 4 in volume ratio. The
applied voltage is 5 V and the operation temperature is 208C.
The shape memory effect of acquired thin foils is mea-sured with the bending test [7] and their martensitic trans-formation temperatures are determined by the Differential Scanning Calorimetry (DSC) with a heating/cooling rate of 108C minÿ1. DSC measurement was conducted using a
Dupont 2000 thermal analyzer equipped with a quantitative scanning system 910 DSC cell for controlling the heating and cooling rates in pure nitrogen gas.
3. Results and discussion
To avoid the effects of concentration polarization, ultra-sonic agitation is applied to enhance the etching process. Fig. 1 shows the etching rate of ultrasonic agitated etching. Solution A has a higher concentration of HF/HNO3 than
solution B and thus the etching rate of solution A, 0.5 mm sÿ1, is higher than that, 0.1 mm sÿ1 of solution B.
Although the lower etching rate is preferred for thickness control, the surface of the thin foils etched by Solution A is much brighter than that etched by Solution B. The bright-ness contrast is so obvious that they can be easily distin-guished visually. The plate thickness can be machined by ultrasonic agitated chemical etching to as thin as 10 mm or
so, which is dif®cult to achieve by a conventional rolling process [6].
Electropolishing can also be applied to chemically machine the TiNi thin plate. The advantage of electro-polishing is that the thinning rate can be controlled by the applied voltage. Fig. 2 shows the thickness reduction rate of electropolishing under the condition of 5 V at 208C. The polishing rate is about 2.5 mm minÿ1 (0.042 mm sÿ1)
which is much slower than that in the case of Fig. 1. The thin foil fabricated in this way is even brighter than that etched by Solution A.
All thin foils fabricated in this study have a shape memory effect in the bending test. The recovery percentages of foils with different thickness are shown in Fig. 3. The foils with thicknesses of 100 and 60 mm are all in excellent shape achieving 100% recovery. However, those with thicknesses of 40, 30 and 20 mm achieve only about 80% recovery. This feature may be due to the fact that, before or during the bending test, the folding induced plastic strain caused by manual operation is more signi®cant in thinner foils.
In order to understand the effect of specimen thickness on the martensitic transformation of the TiNi alloy, thin foils with thicknesses of 100, 50, 20 and 10 mm are tested by DSC
Fig. 1. Etching rates of ultrasonic agitated chemical etching of Solution A and Solution B at room temperatures.
Fig. 2. Thickness reduction rate vs. time in the electropolishing process under the condition of 5 V at 208C.
and the results are shown in Fig. 4. Fig. 4 indicates that the temperatures of transformation peaks, M*and A*, do not
change much for various foil thicknesses. However, the transformation peak broadens and the peak height decreases as the specimen thickness decreases. A similar situation also occurs in the transformation enthalpy, H. Kuninori et al. [8] observed that the area of TiNi alloy thinner than approximately 100 nm in the thin foil of the TEM specimen did not achieve martensitic transformation even when the thin foil was cooled to 98 K. This feature indicates that the surface energy of thin foil can affect the martensitic trans-formation. Therefore, we suggest that the signi®cant drop of the H value of the foil of 10 mm thickness shown in Fig. 4 partly comes from the effect of the surface energy. The broadening of the transformation peak is related to the effect of the sample size (thickness length width) on the traces of the thermal analysis. It is well known that the
heavier sample will enhance the sensitivity of the thermal analysis more greatly [9]. It is dif®cult for the DSC appa-ratus to correctly detect the difference of heat ¯ow when the sample is too small. The suggested minimum weight of the DSC sample used in this study is about 5 mg. In order to obtain the sample weight higher than the minimum value (weight 1020 mg was used in this study), quite a number of thin foils pieces need to be folded into the same DSC pan. The thinner the foil is, the more folding of the foil is needed to put together to get the desired weight. This causes the DSC tracing be different from that of the bulk specimen. The face-to-face contact of folded foils in the DSC pan will induce thermal contact resistance [10]. This thermal contact resistance will delay the heat ¯ow among the folded foils during the DSC test and cause the transformation tempera-ture range become wider. Therefore, the thinner the foil is, the broader the transformation peak will be. These results are explicitly shown in Fig. 4. In addition, as shown in Fig. 3, if the thickness of the thin foil is below 40 mm, the recovery of the SMA is affected by the bending induced plastic deformation. We believe that the folded foils have plastic deformation energy caused by folding and thus decrease the measured H value during the DSC test.
The thickness of TiNi thin foils fabricated by this study is limited by the edge etching attack effect as shown in Fig. 5. The specimen edges are found to be etched more rapidly than thickness reduction when the foil is as thin as 10 mm, even though the thickness reduction rate is as slow as 2.5 mm minÿ1in the electropolishing process. Fig. 5 shows
the zig-zags pro®le edge of the 10 mm thin foil. Therefore,
Fig. 3. Recovery of 100, 60, 40, 30 and 20 mm thickness thin foils. These foils are etched by Solution A with ultrasonic agitation.
Fig. 4. DSC transformation peaks of 100, 50, 20 and 10 mm thickness thin
foils. These foils are etched by Solution A with ultrasonic agitation. Fig. 5. The zig-zags appear on the edge of a 10 mm thickness thin foil. 164 J.Z. Chen, S.K. Wu / Materials Chemistry and Physics 58 (1999) 162±165
the effect of the edge etching attack shown in Fig. 5 limits the further thinning of the TiNi alloy foil by chemical machining.
4. Conclusions
Chemical solution of HF : HNO3: H2O 1 : 4 : 5 in
volume ratio with ultrasonic agitation can be applied for the fabrication of TiNi thin foils. The surface of thin foils fabricated by this solution is very bright and the etching rate can reach 0.5 mm sÿ1. The electropolishing process with
electrolyte of H2SO4: CH3OH 1 : 4 in volume ratio can
also be applied to fabricate TiNi thin foils with an even more brighter surface. When the applied voltage is 5 V at room temperature, the etching rate of electropolishing is only about 0.042 mm sÿ1. The thickness of thin foils fabricated in
this study is limited by the edge etching attack effect and the thinnest thickness obtained is approximately 10 mm.
All TiNi thin foils fabricated in this study exhibit shape memory effect. The proceeding of martensitic transforma-tion is obviously affected by the surface energy and the folded foils if the foil thickness is approaching 10 mm. This feature appears on the decrease of the H value and the broadening of transformation peaks in the DSC test.
Acknowledgements
The authors sincerely acknowledge the ®nancial support of this study by the National Science Council (NSC), Republic of China, through the Grant NSC86-2216-E002-015.
References
[1] Y. Suzuki, in: H. Funakubo (Eds.), Shape Memory Alloys, 1987, p. 176.
[2] K. Kuribayashi, Sensors Actuators 20 (1989) 57.
[3] D. Reynerts, H. Van Brussel, Proc. 1st Int. Conf. Shape Memory and Superelastic Technologies, Monterey, CA, USA, 1994, p. 271. [4] A.D. Johnson, J.D. Busch, Proc. 1st Int. Conf. Shape Memory and
Superelastic Technologies, Monterey, CA, USA, 1994, p. 229. [5] M. Kohl, K.D. Skrobanek, E. Quandt, P. Schlomacher, A. Schler,
D.M. Allen, J. de Physique Colloque C8 (1995) 261.
[6] K.W. Lee, Master Thesis, Department of Mechanical Engineering, National Taiwan University, Taipei, Taiwan, 1996.
[7] H.C. Lin, S.K. Wu, Scripta Metall. Mater. 26 (1992) 59.
[8] T. Kuninori, E. Sukedai, H. Hashimoto, Mater. Trans. JIM 37 (1996) 1404.
[9] W.W. Wendlandt, Thermal Analysis, 3rd ed., Wiley, New York, 1986, p. 265.
[10] J.P. Holman, Heat Transfer, 5th ed., McGraw-Hill, New York, 1981, p. 48.
