Practical, robust and inexpensive structural health monitoring sensors based on the smart properties of TRIP steels have proven to be feasible. To optimize TRIP high alloy steels for use in structural health monitoring sensor applications.
Synopsis
This work and the experimental work from Section 3.5.9 was the primary source for Journal Paper 1 – Effect of Transformation Temperature on Transformation Rate in TRIP Steels (Bemont, Cornish, & Bright, 2013). This work and that in section 4.4.4 were the primary source for journal paper 3 – Design and testing of subscale aircraft intelligent wing bolts (Vugampore & Bemont, Design and testing of subscale aircraft intelligent wing bolts, 2012 ).
Understanding TRIP Steel
TRIP steel fundamentals
TRIP steel as a smart material
The incubation strain corresponds to the point where the strain-induced transformation is triggered and thus to the lowest strain that can be measured via the martensitic transformation and, for a given cross-sectional area, the lowest measurable strain. This point can be shifted to some extent (it usually corresponds approximately to the yield stress) and the gradient of the curve, synonymous with the martensitic transformation rate, can also be adjusted (by alloy composition and processing) according to the requirements of an application, as will be discussed in chapters 3 and 4.
Some properties of steel
- Steel phases and transformations
- Plastic deformation in crystals
The number of expected slip planes therefore depends on the crystal structure of the material. The width is usually larger when the fcc and hcp allotropes of the material have similar free energies (Talonen, 2007).
TRIP steel metallurgy
- General effects of alloying elements
- Deformation-induced transformation
- The austenite to martensite transformation
- Alloying chemistry for optimising the transformation
- Schaeffler diagrams
- Heat treatment, quenching and processing
Thus, in order for the transformation to proceed in the temperature range between Ms and Md, it is necessary for the remainder of the driving force to be induced mechanically through external loading, as depicted in Figure 2.9 (Porter et al., 2009). Strain-induced martensite forms by a mechanism that requires plastic strain for the parent austenite (Turteltaub & Suiker, 2005), as shown in Figure 2.10 (Stumpf, 2010).
The influence of temperature and strain rate on the transformation
Higher strain rates result in more adiabatic heating of the tensile test specimens, as shown in Figure 2.33 (note that the specimen strained at a rate of 20 in./min likely had a delayed response to the temperature measurement due to this higher strain rate). No hcp structure was found in the Fe-Ni system as in the Fe-Mn system.
Modelling the transformation
- Overview of transformation modelling
- Some earlier models
- Investigation of a general model
Olson and Cohen found that the temperature sensitivity of transformation kinetics can be reduced by reducing the entropy differences of fcc, bcc and hcp through alloy composition (Olson & Cohen, 1975). The temperature dependence of the α parameter can be reduced by reducing the temperature dependence of the folding fault energy by reducing the entropy difference between γ and ε (S є).
Mechanical & physical properties of TRIP steels
TRIP steels also show excellent fatigue resistance properties, again influenced by the transformation of austenite to martensite (Krupp et al., 2001). The fatigue strength and ultimate tensile strength of TRIP steels are compared with other structural steels in Figure 2.43 (Olson et al., 1980). The impressive mechanical properties exhibited by TRIP steels are also partly due to the composite effect of the (often fine) multiphase microstructure, hard martensite in a softer austenite matrix (Tomota et al., 1976) (Bhadeshia & Edmonds, 1979a).
The stress-induced phase transformation is mainly responsible for the impressive ductility and elongation properties (Zackay et al., 1967) (Sherif, 2003). This is due to the austenite grains being split by previously transformed martensitic plates (Perlade et al., 2003). As the transformation to martensite proceeds, the size of newly formed martensitic plates decreases (Perlade et al., 2003).
Martensite formed from plastically deformed austenite is generally stronger than that formed thermally (Zackay et al., 1962).
Observing the transformation
- Observing the transformation microscopically
- Observing the transformation magnetically
- Magnetism and TRIP steel
When carbon was reduced to 0.19 wt%, a significant proportion of ε-martensite was present in the microstructure (Figure 2.56 (b)), and when it was reduced to about 0 wt%, a very significant proportion was found to ε-martensite was present. (Figure 2.56 (c)) (China, 1958). Increasing the proportion of chromium to 15 wt%, with 5 wt% manganese, the microstructure after annealing and water quenching consisted of pools of αʹ martensite in a δ-ferrite matrix (Figure 2.59) (Cina, 1958) . The gradient of the free space curve corresponds to the fundamental physical constant μ0, the magnetic permeability of a vacuum.
When an external magnetic field is applied, the magnetic domains are encouraged to align and a magnetic field much larger than the applied magnetic field can be produced as shown in Figure 2.69 (Gaita, 2004). Near the origin there is a slow rise due to reversible growth of the domain alignment and the magnetic permeability of the core remains low. The magnetic permeability curve shown in Figure 2.71 is derived from the preceding magnetization curve and the relationship μ = B/H (Clarke, 2008).
It can be seen from Figure 2.72 that the magnetization curves of ferromagnetic materials do not reverse on themselves as the external magnetic field increases and then decreases (NDT Resource Center, 2012).
Summary and analysis
It has been found that the first of these phenomena usually occurs when the operating temperature is just above the Ms temperature and the second just below the Md temperature. This implies that the authors have erroneously stated that Ms is lower than it is. Such alloys can thus be advantageous for operation in the range between Ms and Msσ.
It is possible that the nucleation of ferromagnetic α´ from ε martensite may be only minimally sensitive to temperature and strain rate. As discussed in Section 2.5, the level of insensitivity to temperature and strain rate required for the purpose of the current work is considerably greater than that of most previous work. Another option to reduce temperature sensitivity would be to design an alloy to operate in the region where p ≈ 1, where the β curve is flat in Figure 2.39 (b) (Olson & Cohen, 1975).
It may be useful to determine this constant for a specific alloy set in the future.
Material development and characterisation
Introduction
Alloy design overview
Schaeffler diagrams were used to approximate the proportions of different phases in the material and to roughly predict the tendency to form other phases during deformation. As discussed in section 3.4.1, several formulas have been proposed for Ms. and the results of these different formulas vary considerably. Certain elements used in material composition usually have specific effects on the characteristics of the material's transformation.
The elemental composition of the material can also be used to qualitatively predict mechanical and physical property variations, some of which have an indirect influence on the character of the material's martensitic transformation. As discussed in the chapters detailing the proposed applications, it is easy to build in fail-safe mechanisms for situations where the pressure TRIP specimen may reach its ultimate strength. On the other hand, failure of a pull TRIP element will usually lead to complete mechanical failure of the device.
As discussed in Section 2.5, it was known that temperature and strain rate, among other environmental factors, would affect transformation and thus repeatability.
Alloy transformation temperature sensitivity
- Introduction
- Theory
- Summary
Change in the magnitude of the driving or retarding forces at the austenite-martensite interface as a martensite plate grows. In Figure 3.4, the distortion-induced transformation is shown as a progression from 0% to 100% between the upper and lower dotted lines, respectively. When considering strain-induced transformation, a contribution to the strain energy that produces approximately 50% transformation in the alloy in Figure 3.4(a) at standard test temperature (T1) requires a smaller temperature change to achieve zero or full transformation instead than the alloy in Figure 3.4(b) with a lower Mf.
The alloy in Figure 3.4(b) is also expected to transform more gradually over a larger deformation range. However, the maximum extent of transformation should be about the same for the two alloys represented by Figure 3.4(a) and Figure 3.4(b) (similar amounts of retained austenite), and a more gradual transformation is likely to even be beneficial. This is illustrated when comparing Figure 3.4 (b) and (c), where only T0 moves while Ms and Mf remain constant and ΔGMs again remains roughly constant.
It is clear that the temperature sensitivity is significantly reduced for the alloy presented in Fig. 3.4(c), while the rate of transformation with strain is likely to decrease or remain the same, but may also increase depending on the extent of gradient reduction.
Cast alloy design, testing and analysis
- Introduction
- Methodology for the identification of relevant empirical formulae
- Alloy fabrication
- Experimental equipment and method
- Compression testing results
- Metallography .1 Sample preparation .1 Sample preparation
- Etching
- Microstructures
- Assessment of microstructures
- Temperature sensitivity
- Effect of dislocation density
- Effect of high strain rate
- The effect cyclic loading
- The effect of circuitous loading
- Further analysis of compressive testing results
- Conclusions and further work
From Figure 3.7 (derived from Journal Article 1, Appendix 3), it is thus concluded that as the martensite free energy curve shifts downward (orange line), Ms and Md30 increase the temperature by equal amounts and (a contribution of strain energy is required to transform occurs at room temperature) is reduced. We tried to melt the samples in an induction furnace with a vacuum chamber in Figure 3.15. The microstructures of this material are shown in the undeformed and deformed state at different magnifications in Figures 3.37 to 3.40.
At all these magnifications, considerable martensite was revealed in the pre-deformation microstructure (Figure 3.37 (a) to Figure 3.40 (a)). Slip features related to martensitic growth orientation may be present in the microstructure after deformation (middle right, Fig. 3.38 (b)). In the pre-deformation microstructure, it revealed considerable αʹ lath martensite, which is particularly clearly visible in Figure 3.42 (a).
The microstructure after deformation was similar to the pre-deformation microstructure, especially at low magnification (Figure 3.62 (b)). There was some evidence of ε martensite stimulating the growth of αʹ martensite in the microstructure after deformation (Figure 3.65 (b)). It is likely that most of the martensite in the pre-deformation sample (Figure 3.66) was produced during sample preparation.
Hot rolled tensile alloy
- Introduction
- Alloy investigation
- Refining alloy transformation through theory and observation
- Alloy design
- Alloy fabrication .1 Introduction .1 Introduction
- Fabrication of steels TRIP R and TRIP S
The characteristic transformation curves detected in this way were then plotted as shown in Figure 3.119. The empirical and semi-empirical literature equations used to estimate Ms and Md30 temperatures were the same as those used for cast alloys in Section 3.4 and can be found in Table 3.2 and Table 3.3, respectively. For the alloys tested, Table 3.19 shows the alloy compositions, nickel and chromium equivalents, and empirically predicted Ms and Md30 temperatures as specified.
This corresponds to the results in Table 3.19 and to the empirical formulas that were analyzed. The proposed compositions (TRIP S and TRIP R) and their predicted transformation temperatures are listed in Table 3.21, along with the previously tested alloys for comparison. The material is cast in the composition specified in Table 3.19, hot rolled and water quenched in a relatively large amount.
One of the TRIP R ingots had a composition close to that intended, as shown in Table 3.22.