Optimization of Nitinol Shape Setting


©2006 ASM International. All rights reserved.
Presented at ASM International M&PMD Conference, Boston, MA.

By: Norman Noble, Inc.


Abstract   • Introduction   • Experimental   • Results & Discussion   • Conclusion
Nitinol is often used in medical devices for its unique mechanical properties. It is also used for the ability to produce complex geometry via shape setting, eliminating costly and difficult machining operations. The process of shape setting requires deformation of a finished or partially finished device. If this deformation is too severe, cracking can result leaving the device unusable. While multiple shape setting operations may be used to minimize the possibility of cracking, limiting the number of these operations is also important. Each shape setting cycle will affect the Austenite to Martensite transformation temperature and mechanical properties, potentially increasing variation in the end product performance or moving the properties outside the allowable limits.   This paper will study whether the level of achievable shape setting deformation can be affected by modifying the post laser cut surface of the pre-shape set component. The study will evaluate the effect of the laser cutting processes, and post laser cutting processes, and their effect on the level of deformation achievable before cracking occurs. Increasing the deformation achievable with a single shape setting cycle may minimize the number of cycles required, decreasing variation of end product performance as well as decreasing manufacturing cost and complexity. This paper will present an analysis of the effects of pre-shape setting surface preparation on the ability to maximize the deformation achievable during shape setting heat treatment.
Nitinol is an important part of the medical device industry. Understanding the fundamental aspects of processing this alloy is important for future innovations. One of the most important properties of Nitinol is the ability of shape setting through heat treating. The process of shape setting permits geometry to be produced that may not be cost effective to achieve with a machining process.

The process of Nitinol shape setting includes controlled deformation of a part during heat treatment. If the level of deformation required results
  in cracking, then multiple shape setting cycles may be required to achieve the final geometry. Each shape setting cycle alters the mechanical and transformation properties of the component, making the prediction of final performance much more difficult. Also, each shape setting cycle will require complex shape setting fixturing that will increase costs. In order to minimize the number of shape setting cycles, it is important to understand the effects of post laser processing on the ability to shape set a Nitinol component. This testing will focus on these effects.
Test samples were produced from 0.35 mm (0.014”) thick Superelastic Nitinol Sheet. The Active Af of the sheet was 7.2 °C.

The samples were 12.7 mm (0.500”) in overall length. The cross section of the test section was 0.35 mm (0.014”) square. Samples were cut in both the longitudinal and transverse to the rolling direction. The sample geometry is depicted in Figure 1 below:

Figure 1: Sample geometry indicating orientation to the rolling direction. Samples are 12.7 mm (0.500”) long.

Each test group was cut using an Nd:Yag laser, and separated into three sub-groups. One of the sub-groups was left as-cut, while the other two were processed with typical post laser cleaning techniques prior to testing. The post laser cutting processes were Post Laser Cleaning (Pickling / Descaling), and Electropolishing. Each group had a minimum of 15 test samples.
  The sample matrix is presented in Table 1 below:

Table 1: Sample Preparation Matrix.
Direction Group 1 Group 2 Group 3
With As Cut Post Laser Cleaned Electropolished
Against As Cut Post Laser Cleaned Electropolished

Testing included deforming the test specimens until failure occurred (see Figure 2). The amount of deformation achievable was measured for each sample. This data was then used to characterize the amount of deformation achievable with each group.

Figure 2: Test Fixture for deforming test samples.
Results & Discussion
Each finishing step changed the surface finish of the test pieces. The post processing operations also changed the amount of deformation that could be achieved prior to failure. The test results are presented in Figure 3:


Figure 3: Scatter plot of data produced during deformation testing. The as-cut parts are on the left, the post laser cleaned are in the middle, and electropolished parts on the right.

The test results clearly show a change in deformation and surface finish with each processing step. With each post processing step, the amount of deformation improved. The most dramatic change can be seen between the raw laser cut parts and the post laser cut cleaned. The maximum deformation was achieved after the parts were electropolished.

It can also be seen that the orientation of the samples with respect to the rolling direction also played a role. This effect was not as noticeable on the raw laser cut parts. However, as the parts were post laser processed, the effects of orientation became more evident, with the electropolished parts showing the most dramatic effects. This result, though not unexpected, is very easy to overlook during process design.

To confirm these relationships, an ANOVA (Analysis of Variance) was performed. The ANOVA indicated that both orientation and post processing had a statistically significant impact on the amount of deformation achievable. The results of the ANOVA are represented graphically as an interaction plot (Figure 4):

Interaction Plot
Figure 4: Interaction plot of the test data.

Metallographic samples were prepared to examine the effects of laser cutting on the microstructure of the samples. The metallographic examination confirms that a thin layer of resolidified material exists on the laser cut surfaces (see Figure 5). This layer has different properties than the base material due to its rapidly solidified material. This change in properties will almost certainly aid in crack formation during tensile loading, and ultimately result in premature failure.

Figure 5: Optical Photomicrograph of recast layer present after laser cutting. This example is approximately 0.0028 mm (0.00011”) thick. 1000X magnification, Krohl’s Reagent.
  Subsequent examination of the products after post laser cutting cleaning and electropolishing revealed that this recast layer had been removed.

SEM fractography was conducted on the fracture faces of each group. The examination revealed that each processing step increased the amount of plastic deformation achievable prior to failure (see Figures 6 and 7). This indicates that the recast layer observed in the metallographic examination
of the unprocessed laser cut samples likely contributed to a stress concentrator that allowed a premature failure to occur. The post processed samples did not show evidence of this recast layer and subsequently showed fracture faces typical of classical tensile overload.

Figure 6: SEM Photomicrograph of fracture face of the laser cut test specimen. The arrows indicate the fracture initiation site in the recast layer still present on the surface. There is microvoid coalescence – indicative of tensile overload; but note the lack of necking – pointing to a stress concentrator. 236X magnification.

Figure 7: SEM Photomicrograph of fracture face of the post laser cut cleaned test specimen. The arrows mark the evidence necking, coupled with microvoid coalescence indicating a classical tensile overload. 236X magnification.

Based upon these observations, several relationships can be drawn:

- Laser cutting leaves a layer of resolidified material on the laser cut edge. This material has different physical properties from the base material, and also likely contains regions of high residual stresses due to its rapid solidification. These two properties act as stress concentrators that promote premature failure during shape setting deformation.

- The post laser cleaning processes contribute to the removal of this layer. The removal of this layer resulted in a drastic improvement in the amount of deformation achievable during shape setting.

- Electropolishing further improves the surface finish, as can be seen in Figure 3. This improvement in surface finish, however, has a negligible impact on the amount of deformation achievable prior to shape setting.
Understanding the fundamental properties of Nitinol is important for optimizing process design. The impact of post laser processing, as well as the orientation of the raw material, should be considered during process design.

In the case of laser cutting, removal of the resolidified material is critical. This can be achieved through post laser cutting cleaning, and subsequent electropolishing. By removing this layer, and considering texture, the amount of deformation achievable during shape setting can be maximized. This, in turn, will minimize the number of shape setting cycles required.
  By minimizing the number of shape setting cycles, the final product consistency can be improved. The ability of achieving final dimensions is maximized, and more importantly, the final properties are more likely to be within specification.

It is important to note that in this study the amount of material removed during these processes was not measured or controlled. It is recommended that any post laser cleaning process be designed to ensure full recast layer removal to repeat the relationships detected in this study.
Reprinted with permission of ASM International®.
All rights reserved. www.asminternational.org