NOISE AND VIBRATION DAMPING OF Fe-Cr-X ALLOYS

D.Pulino-Sagradi, M.Sagradi and J.L.Martin

Débora Pulino-Sagradi (Contacting author) Departamento de Materiais Faculdade de Engenharia Mecânica Universidade Estadual de Campinas CEP 13083-970 , Campinas, S.P., Brazil Tel.: (019) 239-7966 * Fax: (019) 239-3722

Marcello Sagradi Département de Physique École Polytechnique Fédérale de Lausanne Switzerland

Jean-Luc Martin Département de Physique École Polytechnique Fédérale de Lausanne Switzerland

NOISE AND VIBRATION DAMPING OF Fe-Cr-X ALLOYS

ABSTRACT

The aim of the present work is to study the noise and vibration damping capacity of ferromagnetic Fe-16%Cr base alloys (before and after heat treatment) with different Al and Mo contents. The noise damping was evaluated by the level of sound emission after an impact. The vibration damping was studied using a cantilever device. In addition to these tests, the magnetic structure of the materials was also investigated by Kerr effect. It was verified that the materials can decrease noise level in the frequency range of human earring. The vibration damping is influenced by heat treatment and chemical composition of the alloy. The improvement of vibration damping after heat treatment is ascribed to the decrease of internal stresses in materials and changes in magnetic domain structures.

Keywords: high damping alloys, smart materials, magnetomechanical damping, magnetic structure.

INTRODUCTION

Mechanical systems under cyclic loading normally present structural vibrations. These vibrations can cause noise and fatigue problems. The reduction of vibration amplitudes can be partially avoided by using rigid and massive structures or applying springs or other kind of dampers. (Humbeeck, 1985) However, these solutions can not be always used, because they can change the system design or its weight. In such cases, the problems associated with vibrations can be solved by selecting suitable materials. Materials able to absorb the mechanical energy related to vibrations are called damping materials. Rubber is a an example of a damping material, but its use is restrict due to temperature and strength limitations. Another class of damping materials is the metallic alloys. Among these alloys, the ferromagnetic, such as Fe-Cr base alloys, play an important role.(Kawabe and Kuwahara, 1981) Since these alloys have high corrosion resistance and high mechanical strength (Masumoto et al., 1986; Hinai et al., 1988), they are employed to reduce vibrations in turbine blades (Amano et al., 1977), in mechanical and electronic sensitive instruments to maintain their precision (Ritchie et al., 1987), in public transports such as railways (mainly in urban areas, where not only the vibration is a problem, but also the noise is important) (Barbezat et al., 1996), etc. The ferromagnetic materials are composed by magnetic domains. Each domain is separated of its neighbor by a so-called domain wall. These walls can move under external magnetic fields or stresses. When this movement is irreversible, a fraction of energy provided to the material is dissipated as an internal friction. This mechanism (by which the mechanical energy is transformed into heat) is responsible for the intrinsic damping of the material and it is called magneto-mechanical damping.(Cochardt, 1953) Thus, the damping level in ferromagnetic alloys depends on magnetic domains. However, the structure of domains is influenced by the microstructure of the material. The presence of dislocations, second phase particles, alloying elements, etc., create internal stresses. These internal stresses change locally the magnetic properties of the material, influencing the magnetic domains and therefore the damping capacity of the material. (Träuble, 1969) To improve magneto-mechanical damping many authors have been studied the effect of heat treatment and additions of some alloys elements in Fe-Cr base alloys. It was verified that additions of Al (Kawabe and Kuwahara, 1977; Amano et al., 1977; Schneider at al., 1981; Golovin et al., 1990; Golovin et al., 1992; Golovin and Rokhmanov, 1993; Golovin, 1994) and Mo (Suzuki et al., 1977; Schneider et al., 1981; Golovin et al., 1990; Golovin et al., 1992; Golovin and Rokhmanov, 1993; Golovin, 1994) increase damping capacity. However, the exact heat treatment to obtain a maximum damping are not yet defined. The aim of this work is to study the noise and vibration damping of Fe-16%Cr alloys with different Al and Mo contents. For this purpose, the mentioned alloys were heat treated at different temperatures for several times and then tested to evaluate noise and vibration damping capacity and magnetic domain structure.

EXPERIMENTAL

The work was conducted with as cast Fe-16%Cr-(2-8%)Al, Fe-16%Cr-(2 and 4%)Mo, and Fe-16%Cr-2%Mo-1%Cu alloys (wt.%) melted in a vacuum induction melting furnace. All alloys are ferritic. These materials were heat treated in vacuum (( 2x10-6 mbar) at various temperatures (900 to 1200¡C) for different times (1 to 6h) followed by slow cooling (2¡C/min). The mechanical properties of these materials were studied by measuring their Vickers hardness. The noise damping was investigated by sound emission measurements. In these tests, cylindrical bars of 100 mm in length and 10 mm in diameter were let fall down on a non-resonant surface and the noise level emitted at the moment of the impact was recorded by a microphone connected to a signal analyzer. In addition to the Fe-Cr alloys, a bar of an austenitic AISI 304 stainless steel was also tested, since this material does not present magnetomechanical damping. The vibration damping capacity was evaluated using a cantilever device at room temperature. In this procedure, samples with dimensions of 65x7x0.8Êmm were used. For each test the sample was fixed by one end to the sample holder while the other end is kept free. This assembly was then made to oscillate (on mode one of vibration) over a range of frequencies around the resonant frequency with a constant drive-force amplitude. The oscillation of the free end of sample was detected by an optical sensor above the specimen without any physical contact. A plot of vibration amplitude of the sample as a function of frequency provides a curve that allow the damping capacity measurement. The damping was calculated by the inverse of quality factor given by: Q-1Ê=Ê(f/fr where (f is the half-width of a resonance peak of resonant frequency fr. The damping calculated by this way refers to a single drive-force amplitude employed to excite the sample and, thus, to a certain strain amplitude at root beam ((). Applying this procedure for several drive-force amplitudes, a plot of Q-1 versus ( could be obtained. Magnetic domain structures were observed by the magneto-optical Kerr effect (Rave et al., 1993). This effect is based on the interaction between an incident polarized light with the magnetic moment of domains in the material.(Williams et al., 1951) An optical microscopy was suitable arranged for this purpose. The Kerr images were directly obtained by a charged coupled device (CCD) camera connected to a computer. Static and dynamic observations of domain structure were performed. For each observation, samples were previously electrochemically polished to avoid work hardening of their surfaces. Following, an interference layer (Buscher and Reimer, 1993) of ZnSe was evaporated on the polished surface to improve the image contrast. Static observations provide volume-fraction of 90¡ magnetic domains. The volume-fraction of 90¡ magnetic domains was investigated since damping capacity is directly related with them (Xiaodong and Baorong, 1993). To perform this task, images of 40 fields of 2.2x104 mm2 ((0.9 mm2) were acquired in each sample. The volume-fraction was defined as the number of pixels related to 90¡ magnetic domain area divided by the total pixels related to the total surface observed. Application of magnetic field causes movement of domains and thus provide dynamic observations. To estimate the domain mobility, the magnetic field necessary to move the majority of the domains was recorded. For this measurement, 15 fields of 2.2x104 mm2 were tested in each sample.

RESULTS AND DISCUSSION

Noise damping The results of sound emission from some as received materials are presented in figure 1. It was noted that the AISI 304 stainless steel provided two peaks of high intensity, one around 4 kHz and another around 11 kHz. The ratio between the frequencies of the first and the second peak is equal to the ratio between the natural frequencies in mode I and mode II of a free-free excited bar. (Harris and Crede, 1976) Therefore, these peaks correspond to the modal frequencies of the bar tested. Natural frequencies are a function of beam configuration, vibration mode, sample geometry, Young’s modulus and density of the material. However, for the same base alloy the Young’s modulus and density are practically the same and since sample geometry was always the same, these two modal peaks should be presented in the other materials tested. In fact, the Fe-16%Cr-2%Mo, Fe-16%Cr-2%Mo-1%Cu, Fe-16%Cr-2%Al and Fe-16%Cr-8%Al showed these peaks, but the intensity is much lower. On the other hand, in the Fe-16%Cr-(4%Al and 6%Al) alloys the second peak was suppressed. The sharp decrease of noise level of the ferritic alloys (with a suppression of the second peak in some cases) evidences energy loss due to their high damping capacity. Moreover, it is interesting point out that this reduction in the noise level occurs in the frequency range of human earring (less than 16 kHz)(Everest, 1991).

[pic]
Figure 1. Noise level provided by some studied materials after impact.

Vibration damping The curves of resonant peaks at modal frequencies are the first results provided by cantilever method. Examples of these curves are presented in figure 2. It can be seen a non-linear behavior (Broch, 1984), i.e., as the drive-force of excitation increases, the resonant frequency shifts towards lower values. As mentioned before, the natural frequency is a function of beam configuration, vibration mode, sample geometry, Young’s modulus and density of the material. Nevertheless, for a given test, the beam configuration, vibration mode, sample geometry and material density are constant. Therefore, the decrease in resonant frequency can be explained by a change in Young’s modulus of the material with the increase of drive-force. In fact, Masumoto et al. (1979) realized a decrease (up to 10 % of its initial value) of Young’s modulus for Fe-Cr alloys with the increase of vibration amplitude. The increase of drive-force of excitation increases the amplitude of resonant peak. Since increasing drive-force provides higher energy levels to the lame, the amplitude of resonant peaks tend to increase. However, considering the same drive-force, the increase of resonant peak is less pronounced after heat treatment. This reflects the damping enhancement that occurs after heat treatment because the materials after heat treatment absorb much more energy than the other at the same drive-force level.

|[pic] |[pic] |

Figure 2. Resonance peaks provided by a Fe-16%Cr-4%Al alloy in different drive-forces (F/m) of excitation: (a) as received material and (b) heat treated at 1100°C for 6h.

The curves showed in figure 2 allows calculate the intrinsic damping of the material by means the quality factor. Figure 3 shows the curves of damping (as Q-1) versus strain amplitude (() occurring in the fixed end of the sample of as received materials. It can be noted that damping depends strongly on strain. This behavior is well known in ferromagnetic materials and it is related to the difference between the energy provided to the system and the energy absorbed by magnetic domains, as discussed in details elsewhere (Smith and Birchak, 1968; Smith and Birchak, 1969).

|[pic] |[pic] |

Figure 3. Variation of damping capacity (Q-1) versus strain amplitude (() in as received alloys. (a) Materials with different additions of Mo and (b) materials with different additions of Al.

The effect of heat treatment on damping vibration is illustrated in figure 4. It was verified that heat treatment always improve damping capacity of the as received materials. Probably during heat treatment the number of defects, such as dislocations, decreases and, thus, the internal stresses in the material decreases. As a consequence of the internal stress relief, the damping increase. (Smith and Birchak, 1968) This effect is suggested by the initial slope of the Q-1 versus ( curves. It was verified that the initial slope of these curves increase as damping increase. According to Coronel and Beshers (1988), the increase in initial slope of damping curves is caused by the decrease in internal stresses of the material.

|[pic] |[pic] |

Figure 4. Variation in damping capacity of (a) Fe-16%Cr-4%Al and (b) Fe-16%Cr-2%Mo alloys heat treated at different temperatures for 6h.

Besides the decrease in internal stresses, it was also observed that the magnetic domain structure changes after heat treatment. Figure 5 shows an example of one kind of magnetic domain structure with 90° domains found in the heat treated materials. A statistical analysis of such structures indicated that volume-fraction of 90° magnetic domains increases after heat treatment (for example, in a as received Fe-16%Cr-2%Mo alloy the volume-fraction is 0.57%, while after heat treatment at 1100°C for 6h the volume-fraction is 3.6%). Udovenko et al. (1993) also realized that a high-damping state of Fe-Cr alloys is really obtained when the amount of 90¡ domains increases.

[pic]

Figure 5. Magnetic domain structure of a Fe-16%Cr-4%Al alloy heat treated at 1000°C for 6h. Note: black and white fields correspond to 180° domains, while intermediate gray levels correspond to 90° domains.

However, this is not the only change in magnetic domain structure with heat treatment. It was also noted an increase in the mobility of magnetic domain. Figure 6 shows the movement of the magnetic domains under an external magnetic field (similar behavior may be obtained under an external stress). Since internal stresses decrease after heat treatment, probably the number of obstacles to the movement of domains decreases. Moreover, the susceptibility to create 90° domains may be changed after heat treatment. However, the exact reasons for the differences in magnetic domain structures are not yet clear.

| | |
|[pic] |[pic] |

Figure 6. Magnetic domain structures of a Fe-16%Cr-6%Al heat treated at 1100°C for 3h. (a) Without magnetic field and (b) under an applied magnetic field (25 mT).

The effect of chemical composition on damping vibration can be seen in table 1, which presents the maximum damping obtained for each alloy and related Vickers hardness. It was observed that in the alloys with Al addition, as the Al content increases, the damping decreases. Similar results were found in the alloys with Mo, i.e., increasing alloy element makes damping to decrease. Since addition of alloy elements introduce elastic distortions into the crystal lattice (Golovin et al., 1990), internal stresses are created and, thus, the magnetic domain structures change (Luborsky et al., 1983). In fact, as the alloy element content increases, the hardness also increase. Moreover, the alloy elements may change the magnetic moment of the neighbor atoms (Gittsovich et al., 1995) and, hence they may change magnetic domain structures.

Table 1. Maximum damping capacities and related Vickers hardness obtained with different materials.
|contents (wt %) of alloying elements in |heat |Q-1 |hardness |
|Fe-16%Cr |treatment |(10-3) |(HV) |
|2% Al |1000¡C/6h |146 |173 |
|4% Al |1100¡C/1h |105 |178 |
|6% Al |1100¡C/6h |48 |214 |
|8% Al |1000¡C/6h |35 |237 |
|2% Mo |1000¡C/3h |163 |143 |
|4% Mo |1000¡C/6h |143 |188 |
|2% Mo - 1% Cu |1000¡C/1h |91 |171 |

CONCLUSIONS

The study of noise and vibration damping of Fe-16%Cr base alloys with different additions of Al and Mo (before and after heat treatment), allows to obtain the following conclusions:
1. The as received alloys can decrease noise level in the frequency range of human earring.
2. The vibration damping of the investigated alloys can be improved by heat treatment. This fact is ascribed to the decrease of internal stress and changes in magnetic domain structures (increase of volume-fraction of 90° domains and increase of domain mobility)
3. Alloying with Al or Mo over than 2% (in weight) decreases the vibration damping of the studied alloys.
4. The highest vibration damping was obtained in the following materials: Fe-16%Cr-2%Al alloy heat treated at 1000°C for 6h (Q-1 = 146x10-3) and Fe-16%Cr-2%Mo alloy heat treated at 1000°C for 3h (Q-1 = 163x10-3).

ACKNOWLEDGMENTS

The authors thank the Swiss Priority Program on Materials (PPM) and São Paulo State Foundation for Research, Brazil (FAPESP) for financial support of this work.

REFERENCES

Amano, K., Sahashi, M., Tokoro, H. and Nakagawa, M., 1977, “High damping characteristics of Fe-Cr-Al alloys associated with their magnetic properties”, Proceedings, 6th International Conference on Internal Friction and Ultrasonic Attenuation, Tokyo, Japan, pp.763-767.
Barbezat, G., Salito, A. and Karimi, A., 1996, “Lucha al ruido”, Sulzer Technical Review, n.1, pp.42-44.
Broch, J.T., 1984, “Mechanical vibration and shock measurements”, 2.ed., Bruel & Kjaer,. Geostrup, pp.40-71.
Buscher, P. and Reimer, L., 1993 “Imaging and colour coding of magnetic domains by Kerr scanning optical microscopy”, Scanning, vol.15, pp.123-129.
Cochardt, A.W., 1953, “The origin of damping high strength ferromagnetic alloys”, Journal of Applied Mechanics, v.20, pp.196-200.
Coronel, V.F. and Beshers, D.N., 1988, “Magnetomechanical damping in iron”, Journal of Applied Physics, vol.64, pp.2006-2015.
Everest, F.A., 1991, “Fundamentals of sound and vibration”, In: Ballon, G. Handbook for sound engineers: the new audio encyclopedia, 2.ed, SAMS, Carmel, pp.3-21.
Gittsovich, V.N. Semenov, V.G. and Uzdin, V.M., 1995, “Bulk and surface magnetic properties of dilute Fe-Cr alloys”, Journal of Magnetism and Magnetic Materials, vol.146, pp.165-174.
Golovin, I.S., Suvorova, S.O., Sarrak, V.I., Zharkov, R.V., Pepik, S.L. and Demidov, B.E., 1990, “Damping capacity of high chromium complex alloys steels” Izvestiya Akademii Nauk SSSR Metally, pp.154-160.
Golovin I.S., Sarrak, V.I and Suvorova, S.O., 1992, “Influence of carbon and nitrogen on solid solution decay and 475°C embrittlement of high-chromium ferritic steels”, Metallurgical Transactions, vol.23A, pp.2567-2579.
Golovin, I.S. and Rokhmanov, N.Y., 1993, “Question of the mechanism of formation of the damping condition of high-chromium ferritic steels”, Metals Science and Heat Treatment, vol.35, pp.526-533.
Golovin, I.G., 1994, “Mechanism of damping capacity of high-chromium steels and (-Fe and its dependence on some external factors”, Metallurgical and Materials Transactions, vol.25A, pp.111-124.
Harris, C.M. and Crede, C.E., 1976, “Introduction to the handbook”, In: Harris, C.M., Crede, C.E. Shock and vibration handbook, 2.ed., McGraw-Hill, New York, pp.1.1-1.25.
Hinai, M., Sawaya, S. and Masumoto, H., 1988, “Damping capacity, magnetic and mechanical properties and corrosion resistance of Fe-Co-Mo-Cr alloys”, Transactions of the Japan Institute of Metals, vol.29, pp.428-436.
Humbeeck, J.V., 1985 “High damping capacity due to microstructural interfaces”. In: Rath, B.B., Misra, M.S. Role of interfaces on material damping, ASM, Toronto, pp.5-24.
Kawabe, H. and Kuwahara, K., 1977, “Damping- and modulus-measurement of high damping metal “SIA” at low frequencies” Proceedings, 6th International Conference on Internal Friction and Ultrasonic Attenuation, Tokyo, Japan , pp.781-785.
Kawabe, H. and Kuwahara, K., 1981, “A consideration of the strain amplitude dependent damping and modulus in ferromagnetic metals”, Transactions of the Japan Institute of Metals, vol.22, pp.301-308.
Luborsky, F.E., Livingston, J.D. and Chin, G.Y., 1983, “Magnetic properties of metals and alloys”, In: Cahn, R.W. and Haansen, P. Physical Metallurgy, 3.ed., North-Holland Physics Publishing, Amsterdam, pp.1674-1734.
Masumoto, H., Sawaya, S. and Hinai, M., 1979, “On the damping capacity of Fe-Cr alloys”, Transactions of the Japan Institute of Metals, v.20, pp.409-413.
Masumoto, H., Hinai, M. and Sawaya, S., 1986, “Damping capacity and pitting corrosion resistance of Fe-Co-Cr alloys”, Transactions of the Japan Institute of Metals, vol.27, pp.401-407.
Rave, W., Reichel, P., Brendel, H., Leicht, M., McCord, J. and Hubert, A., 1993, “Progress in quantitative magnetic domain observation” IEEE Transactions on Magnetics, vol.29, pp.2551.
Ritchie, I.G., Pan, Z.L., Sprungmann, K.W., SchmidT, H.K. and Dutton, R., 1987, “High damping alloys - the metallurgist’s cure for unwanted vibrations”, Canadian Metallurgical Quarterly, vol.26, pp.239-250.
Schneider, W., Schrey, P., Hausch, G. and Torok, E., 1981, “Damping capacity of Fe-Cr and Fe-Cr based high damping alloys”, Journal de Physique (Paris), n. 10, pp.C5.635-C5.639.
Smith, G.W. and Birchak, J.R., 1968, “Effect of internal stress distribution on magnetomechanical damping”, Journal of Applied Physics, vol.39, pp.2311-2316.
Smith, G.W. and Birchak, J.R., 1969, “Internal stress distribution theory of magnetomechanical hysteresis - An extension to include effects of magnetic field and applied stress”, Journal of Applied Physics, vol.40, pp.5174-5178.
Suzuki, K., Fujita, T. and Hasebe, M., 1977, “Damping capacity and mechanical properties of sintered Fe-Cr-Mo high damping alloys”, Proceedings, 6th International Conference on Internal Friction and Ultrasonic Attenuation, Tokyo, Japan, pp.757-761
Träuble, H., 1969, “The influence of crystal defects on magnetization process in ferromagnetic single crystals”, In: Berkowitz, A.E. and Kneller, E. Magnetism and metallurgy, Academic Press, New York, vol.2, pp.621-687.
Udovenko, V.A., Chudakov, I.V. and Polyakova, N.A., 1993, “The fine crystalline and magnetic structure of high-damping alloys based on the Fe-Cr system”, The Physics of Metals and Metallography, vol.75, pp.247-251.
Williams, H.J., Foster, F.G. and Wood, E.A., 1951, “Observation of magnetic domains by the Kerr effect”, The Physical Review, vol.82, pp.119-120.
Xiaodong, L. and Baorong, W., 1993, “Moving characteristics of magnetic domains in Fe-Cr-Al damping alloys”, Materials Science and Engineering, vol.B18, pp.L1-L3.

-----------------------
b)

a)