·Home ·Table of Contents ·Materials Characterization and testing

NDE of Fatigue on Metals Thermography, Acoustic Microscopy and Positron Annihilation Method S.U. FASSBENDER, W. KARPEN, A. SOURKOV Fraunhofer Institute Nondestructive Testing, Universität 37, 66123 Saarbrücken, Germany S. SATHISH Center for Materials Diagnostics, University of Dayton, Dayton OH 45469-0121 H. RÖSNER, N. Meyendorf Fraunhofer Institute Nondestructive Testing, Universität 37, 66123 Saarbrücken, Germany Center for Materials Diagnostics, University of Dayton, Dayton OH 45469-0121 Contact

Abstract

Materials exhibit different characteristics at different stages of fatigue. A single experimental technique that can reliably neither predict nor detect the onset of failure in materials is not yet available. This paper presents the results of IR thermography, acoustic microscopy and positron annihilation methods to characterize the fatigue damage accumulation in austenitic steel X6CrNiNb18 and Titanium alloy Ti6Al4V. The results indicate that positron anhiliation is sensitive to very early stages of fatigue damage, while acoustic microscopy can be useful in the analysis of the formation of microcracks and slip bands. Results on IR thermography indicate a possibility that it could be utilized at all stages of fatigue. The limitations and advantages of the three techniques are presented.

Introduction

Fatigue of metals is characterized by an early stage of damage due to changes within the submicrostructure leading to the formation of microcracks. This is followed by an initiation of a crack that grows until the final fracture. Hence, for evaluation of the entire fatigue process from the beginning of the mechanical loading until fracture different methods have to be utilized. At the beginning of the fatigue process sensitivity of the method to submicrostructural changes (atomic scale) is required. As the material degradation progresses microcrack formation and microcrack growth becomes detrimental for the fatigue life of a component, so the method has to be sensitive to cracks [1]. However, in general the fatigue process in a metal that doesn't have any cracks can be assumed to consist of three stages.

Utilization of different methods for fatigue characterization is essential because at different stages of fatigue damage materials show different characteristics. This paper presents application of three different techniques namely positron annihilation, acoustic microscopy and IR thermography to study different stages of the fatigue processes. The positron annihilation method has been applied to evaluate the submicrostructural characteristic defects on the atomic scale (vacancies, dislocations, micro-voids) during very early stages of fatigue. Acoustic microscopy has been used to detect microcracks and to observe the formation of slip bands. Thus, these methods can help to characterize early stages of fatigue damage including the quasi-crack free region (formation of microcracks) and the crack growth. However, both of these methods require special sample preparations and can not be used for continuous monitoring of fatigue. Hence, they have very limited applications in nondestructive testing. Infrared thermography on the other hand is non-contact technique that allows observation of characteristic changes in the heat dissipation during fatigue. It has high potential to be a nondestructive tool for characterization of fatigue. It requires a careful interpretation of the origin of dissipated heat in the material during fatigue.

Preparation of Test Specimens

Two different materials annealed austenitic steel X6CrNiNb18 10 without any texture and an alloy of titanium Ti6-Al4-V were chosen for this study. All the methods have been applied on the some samples. For positron annihilation and acoustic microscopy flat surfaces are necessary. Therefore, dog bone shaped specimens with a rectangular cross-section (5mm*5mm) were prepared. The 4 surfaces of the specimen had to be prepared according to the requirements of the test methods.

For positron annihilation experiments the surface of the fatigued specimen was electrochemically etched to remove surface damage due to polishing. This is essential because positrons have a low penetration depth (approx. 100 µm) into the material and any surface damage due to polishing would influence the measurements.

In order to eliminate contrast due to topography in scanning acoustic microscopy, the specimens were polished with silicon carbide abrasives of reducing grit sizes. A final polish was performed with suspended colloidal silica (0.05 µm) with out any pressure in order to reduce micro-plastic deformation of the surface.

Experimental Methods

Positron annihilation
A positron-emitting source, e.g. ²²NaCl, is required for positron annihilation method. Whenever a positron is generated in the source, a photon is simultaneously emitted with an energy of 1.24 MeV. The positron penetrates into the material and is decelerated to thermal energy (the positron is thermalized). Within the lattice it can be trapped by crystal lattice defects with open volume, i.e. with local negative charge. Finally, the positron annihilates with an electron by emitting two 0.51 MeV photons. The time delay between emission and annihilation of the positron is measured in positron lifetime spectroscopy [2]. Positron lifetime increases in a characteristic manner when the positron is trapped by dislocations, vacancies, or vacancy clusters.

For the evaluation of the data for the positron lifetime spectra two detector channels were set specifically for the start photon 1.28 MeV and the stop photon 0.51 MeV, respectively. The time delay between start and stop photon is converted into the pulse height of an analog signal with a time pulse amplitude converter. A multi-channel analyzer obtains the lifetime spectrum, i. e. the number of events as a function of time delay. The lifetime can be extracted from the slope of the number of events versus time. If several types of defects trapping positrons are present, then several components can be separated from the lifetime spectra.

In the measurements the positron source was placed between the surface of the fatigued specimen and a reference material of annealed high purity iron. Thus, the required sandwich arrangement is realized although the same material for the sandwich configuration was not available.

Acoustic microscopy
An acoustic microscope consists of a high frequency (100 MHz to 2 GHz) ultrasonic transducer mounted on a lens (sapphire) coupled through a liquid to the sample surface (figure 1). The acoustic lens focuses the plane acoustic waves generated by the transducer in presence of a coupling fluid on the surface of the specimen. A surface acoustic wave (generally a leaky Rayleigh wave) is generated at the water-sample interface and propagates on surface of the sample re-radiating energy back into the lens. During the propagation on the surface the surface acoustic wave interacts with the inhomogeneities like cracks, voids and topographic structures. Interference between the incident and reflected surface waves allows the detection of cracks that are an order of magnitude smaller than the spatial resolution of the microscope (approx. 1.5 µm at 1 GHz)[3]. By defocusing the acoustic lens against the sample surface microstructural properties like slip bands can be detected in anisotropic materials. To detect both changes in microstructure and microcracks in every sample acoustic images were acquired at several defocuses.

Fig 1: Principles of the acoustic microscope

IR Thermography
Mechanical loading of a material causes a temperature change, due to energy dissipation caused by hysteretic effects and thermoelasticity. Thermoelasticity is a reversible thermodynamic effect caused by a change of the internal energy during an elastic deformation of the material. The variation of mechanical strain is correlated with a thermoelastic temperature variation. In metals (positive elongation coefficient) an increase of temperature due to compressive stresses and a decrease due to tensile stresses is observed [4]. Heat dissipation leads to an irreversible temperature enhancement [5]. Mechanical damping depends on the changes in microstructure and submicrostructure of the material a measure of the changes in the temperature caused by heat dissipation can be utilized for characterization of fatigue.

For the excitation of the dissipated heat, the specimens were subjected to sinusoidal cyclic loading in a servo-hydraulic fatigue machine. A focal plane array infrared camera was placed in front of the specimen at a distance of 1 m to measure the specimen's surface temperature. The temperature resolution of this short wavelength camera (3-5 µm, Hg-Cd-Te detector) is DT»0.05 K. In order to increase the emissivity, a polymer coating was sprayed on to the specimen surface.

During the cyclic loading process, a variation of temperature DT is measured on the specimen surface. The temperature DT in the adiabatic regime of negligible heat conduction, is the superposition of temperature variations due to thermoelastic effect DT_el and dissipated heat DT_diss. The temperature variation DT measured during the experiment is defined as the difference between the temperature at the center of the specimen and the temperature of an unloaded reference specimen. The use of a reference specimen eliminates uncertainties introduced by room temperature fluctuations.

Fig 2: Experimental set-up used for passive IR thermography

The changes in the temperature due to thermoelasticity are periodic and follow the mechanical loading. Sequences of 10-100 video frames were stored and averaged to calculate the average temperature due to heat dissipation DT_diss and to eliminate the thermoelastic effect DT_el. A software (Visotherm^Ò) developed by Fraunhofer Institute for Nondestructive Testing [6] allowed the performance of automatic, time dependent measurements and storage of frame sequences (figure 2).

Results and Discussion

Positron annihilation
A characteristic increase in the averaged positron lifetime,t_av, was measured on the fatigued specimen of the austenitic steel X6CrNiNb18 10. The positron lifetimes, t_av, increases

significantly until approximately 30% of the lifetime and saturates during the course of further fatigue damage (figure 3). Within the early stages of fatigue the increase of t_av is caused by an incasing number of lattice defects (dislocations, vacancies). The saturation of the averaged positron lifetime corresponds to a 100% trapping of positrons by lattice defects.

Fig 3: Averaged positron lifetime measured on dog-bone specimens fatigued until defined diminished lifetime [7]

Acoustic microscopy
In several fatigued specimens slip bands as well as microcracks were detected. The detection of open microcracks was difficult because the crack opening was very small (comparable to the Rayleigh wavelength l_R). Most of the microcracks could be observed in the acoustic images obtained at focus because the topography of the crack flanks. On the other hand the slip bands can only be detected in the defocussed images because they are characteristics of the microstructure

An analysis of the acoustic images obtained on samples with an elapsed lifetime of 60, 70, 80 and 90%, reveal that the number of microcracks remain nearly constant whereas their average length increases from 22 µm to 27 µm with the number of cycles. In some of the samples cracks with lengths larger than 70 µm were observed. In contrast the number of detectable slip bands decreases. This can be clearly seen in figures 4 and 5 where the distribution of microcracks and slip bands (number per length interval) is displayed. The microcrack density for the samples with 60% and 90% used lifetime were 36/mm² and 40/mm², respectively, whereas the slip band densities were 66/mm² and 10/mm².

Fig 4: Microcrack and slip band distribution for a fatigued specimen with 60 % lifetime.

Fig 5: Microcrack and slip band distribution for a fatigued specimen with 90 % lifetime

An explanation for this unexpected behavior might be the fact that microcracks are generated at the slip bands as can be seen in some of the acoustic images. The topographic changes of the surface due to the tilted crack flanks have the effect that the slip bands are invisible also in the defocussed images because the topography of the surface itself suppresses the imaging of microstructural properties in the defocused acoustic image.

IR thermography
Figure 6 shows the results of the temperature change due to heat dissipation DT_diss in a strain controlled fatigue experiment with strain amplitude e =1% and mean strain 0 at a frequency of 0.05 Hz in a sample of austenitic steel. This temperature shows a steep increase during the first few cycles until equilibrium between heat generation in the sample and heat conduction to the surroundings is reached. Then it increases slightly with fatigue damage. The fatigue experiment was interrupted two times in order to prove that the temperature reaches its former value soon after restarting (figure 6). For this material and loading conditions the temperature change due to fatigue damage is 1.4 K during the whole lifetime.

Fig 6: Temperature DTdiss for an austenitic steel sample during a fatigue experiment

Fig 7: Temperature DTdiss for the titanium alloy Ti6Al4V during a fatigue experiment

The result in figure 7 shows the change of the thermal quantity DT_diss during the stress controlled fatigue loading of the titanium alloy Ti6Al4V on three specimens. These specimens were machined out of the same plate material and fatigued under similar conditions (loading frequency 10 Hz, mean stress 467.5 MPa, stress amplitude 382.5 MPa).

The temperature change DT_diss during the process of fatigue damage is much more significant compared to the previous experiment on the austenitic steel. During the whole lifetime a nearly linear relation between the thermal quantity and the number of cycles has been observed on these specimens. Another interesting feature of the curves is that a short fatigue lifetime is related to a higher increase of DT_diss at the beginning of the experiment and a higher slope of the DT_diss (N) curves. Thus, a qualitative prediction of the lifetime was possible.

The temperature change due to heat dissipation allows a means to characterize the accumulated damage during the entire course of fatigue life. Additionally, the material and the mechanical loading conditions influence the heat dissipation. In general for cyclic loading conditions, the origin of heat dissipation can be attributed to local plastification. Hence, both submicrostructural changes during the early stages of fatigue and the microscopic crack growth might contribute to a temperature change. Thus, the precise origin of the dissipated heat cannot be determined without the knowledge of the material and its fatigue conditions.

Conclusions

NDT methods for the characterization of fatigue damage and loading prehistory are required to increase the service lifetime and to enhance the security standards of technical components. Positron annihilation method is very sensitive to an increase of crystal lattice defects (dislocations, vacancies, microvoids) caused by fatigue damage. The positron lifetime revealed a close correlation with the accumulated fatigue damage of the material in the very early stages of fatigue(30% of the lifetime). Average density and average length of the micro-cracks have been determined by analyzing acoustic microscopic images of fatigued samples. These results show no significant increase in the number of microcracks with increasing number of fatigue cycles, but an unexpected decrease in the number of slip bands has been observed. During the dynamic process of mechanical loading heat dissipation caused by mechanical damping processes has been measured using infrared thermal camera. Changes in the surface temperature of the material has been found to be sensitive to accumulated fatigue damage. IR thermography combined with an effective excitation of the dissipated heat might be used as a NDT method in the future.

Acknowledgments

Effort partially sponsored by the "Gesellschaft für Anlagen- und Reaktorsicherheit (GRS)" under research grant number 150 1096.
Effort partially sponsored by the Defense Advanced Projects Agency (DARPA) Multidisciplinary University Research Initiative (MURI), under Air Force Office of Scientific Research grant number F49620-96-1-0442.
G. Walle, A. Ferry are acknowledged for their help during the course of the experiments.

References

1. S. Suresh, "Fatigue of materials" Cambridge University Press, (1991).
2. Hautojärvi: Positrons in solids, Series: Topics in Current Physics; Vol.12, Springer Verlag, Berlin (1979).
3. A. Briggs, "Acoustic Microscopy" Clarendon Press, Oxford (1992).
4. W. Thomson, "On the Thermo-Elastic and Thermomagnetic Properties of Matter" Quart. J. of Pure & Appl. Math., 1 (1857) pp. 57-77.
5. N. Harwood and Cummings, "Thermoelastic Stress Analysis" Adam Hilger, New York (1991)
6. G. Walle, W. Karpen, U. Netzelmann, H. Rösner, N. Meyendorf, "Zerstörungsfreie Prüfung mittels thermographischer Verfahren" TM Technisches Messen, 66 (1999) pp. 312-321.
7. S. Depetasse, "Studium der Entmischungvorgänge in ausgewählten Aluminiumlegierungen mit der Positronenannihilation&quo; Master Thesis, IZFP Saarbruecken (1999).

© AIPnD , created by NDT.net

|Home|    |Top|