谐振式电液疲劳试验机的设计参考外文2

Piezoelectric fatigue testing machines and devicesClaude Bathias*Department of Mechanical Engineering,CNAM/ITMA,2rue Conte,75003Paris,France Received 26November 2004;received in revised form 9September 2005;accepted 29September 2005Available online 6May 2006AbstractDuring the 1990s several methods have been developed around the word in order to test specimens at very high fatigue life.In our laboratory an ultrasonic fatigue testing system was designed and built 15years ago to determine crack growth threshold and S –N curve of metals.Those first results were published in ASTM STP 1231(1994),ASTM STP1411(2002)and in proceedings of Fatigue 2002.This paper sums up the progress of such fatigue testing machines and devices from this date.Ó2006Elsevier Ltd.All rights reserved.Keywords:Gigacycle fatigue;Piezoelectric fatigue machines;High frequency devices1.IntroductionFrom an historical point of view it is said that the first ultrasonic fatigue machine was constructed by Mason in the year of 1950.With the development of computer sci-ences several laboratories have to develop their own machines and design practical test borato-ries of Willertz in the US,Stanzl in Austria,Bathias in France,Ishii in Japan and Puskar in Slovakia,are among leading laboratories in this field.Ultrasonic fatigue test machines in these laboratories are not the same,but some components are the same for all machines.Three most important ones are:a high frequency generator that gener-ates 20kHz sinusoidal electrical signals,a transducer that transforms the electrical signal into mechanical vibration,and a control unit.Early ultrasonic fatigue test machines performed only uni-axial (one-dimensional)and constant amplitude tests,and so the control unit and other parts were not very complicated.In the last two decades,pro-gress has been made to extend the ultrasonic fatigue tech-nique to variable amplitude loading conditions,low or high temperature environments,torsional or multiaxial tests,and so on.Thus,designing a modern ultrasonic fati-gue test machine may involve mechanical,electrical,opti-cal,magnetic and thermal considerations.In France,Bathias used a first ultrasonic fatigue test machine in 1967on the principle of Mason [1].As described in an early review paper [2],the rather restrictive uses of the ultrasonic fatigue test method appeared to be partly due to the lack of commercially available test equipment,forcing the individ-ual investigators to work with improvised facilities not readily amenable to standardized experimental conditions.2.Tension–tension and tension–compression fatigue test machinesAs stated above,the common components of an ultra-sonic fatigue test machine must include the following three components:1.A power generator that transforms 50or 60Hz voltage signal into ultrasonic 20kHz electrical sinusoidal signals.2.A piezoelectric (or magnetostrictive)transducer excited by the power generator,which transforms the electrical signal into longitudinal ultrasonic waves and mechanical vibration of the same frequency.3.An ultrasonic horn that amplifies the vibration coming from the transducer in order to obtainthe/locate/ijfatigueInternational Journalof Fatigue0142-1123/$-see front matter Ó2006Elsevier Ltd.All rights reserved.doi:10.1016/j.ijfatigue.2005.09.020*Tel.:/fax:+33140272322.E-mail address:bathias@cnam.fr .required strain amplitude in the middle section of the specimen.These three parts are special devices required for the production of ultrasonic fatigue load.Other ultrasonic machine components may include recording systems and measuring systems.The function of the system in Fig.1is to make the spec-imen vibrate in ultrasonic resonance at one of its longitudi-nal modes.The displacement amplitude reaches its maximum U0at the end of the specimen,which can be measured by means of a dynamic sensor,while the strain excitation in push–pull cycles(load ratio R=À1)attains the maximum in the middle section of the specimen that produces the required high frequency fatigue stress.The information recorded may include ultrasonic cyclic dis-placement amplitude U0,the evolution of the fatigue crack growth,which enables us to determine the fatigue crack growth rate d a/d N,and stress intensity factor K max,calcu-lated by analytical or numerical methods[1].During ultrasonic fatigue tests,the maximum strain val-ues can be measured directly using miniature strain gauges, suitably positioned on the sample surface.The dynamic displacement amplitude at the specimen extremity,U0,is measured by an opticfibre sensor,which permits measurements of the displacement from1to 199.9l m,with a resolution of0.1l m.The magnification factor of stress can then be calculated according to these measurements.For a virgin specimen(without a crack), the vibratory stress and strain can also be determined at the midsection.This maximum strain value thus deter-mined is then confirmed to be accurate by the use of the above-mentioned strain gauge.In addition,a system of video–camera–television has been used for the control of crack initiation and propagation.This system refines events to1/25th of a second and magnifies specimen surface140–200times.Because specimens of ultrasonic fatigue vibrate in reso-nance,a free end is sufficient for symmetric loading condi-tions(R=À1),which avoids the large and cumbersome arrangements for gripping the specimen as often encoun-tered in conventional fatigue testing(Fig.2).If there is a static load,the situation is different.Superposing a mean stress or displacement upon the symmetric tension–compression cycle can create a complex cyclic loading with R>À1.Therefore,an additional horn will be added at another end of the specimen,as shown in Fig.3.This kind of test machine is particularly useful interest to efficiently study fatigue endurance and fatigue crack growth behaviour of materials subjected to elasto-plastic low cycle fatigue loading superposed on vibration stress cycles with high frequency.(See Fig.3.)A generator with a converter composed of six piezo-ceramics is chosen to provide vibration energy.This ultrasonic generator 900BA is made by Branson Ultrasonic Corporation.It has a maximum power of2kW and provides a sinusoidal signal for the converter that is the source of mechanical vibration.This amplifier maintains automatically the intrinsic frequency of the mechanical system in the range of19.5–20.5kHz.The converter,horn and specimen form a mechanical vibration system with four stress nodes(null stress)and three displacement nodes(null displacement) for an intrinsic frequency of20kHz.Here,the stressand Fig.1.Full resonance system in Bathias’laboratory with schematic view of apparatus.C.Bathias/International Journal of Fatigue28(2006)1438–14451439displacement are considered to be longitudinal.In Fig.4,points B,C (connected points),point A and converter top are stress nodes.The specimen centre is a displacement node,and the stress is the maximum there.The horn must vibrate at a frequency of 20kHz.Depending on the specimen loading,the horn is designed so that the displacement is amplified between B and C usu-ally 3–9times.It means that the geometry between B and C must be determined.The finite element method may be required when the geometrical shape is complex.The mechanical system composed of a converter,a horn and a specimen is linear.All stress and displacement fields are linear.It is necessary to measure the amplitude of one of them.3.High temperature test equipmentThe test equipment consists of a heating device in the middle.Fig.5is a photograph of this system for crack growth testing.In the specimen the temperature is constant over the length of 5–6mm.The television images can be recorded on a videocassette during the tests.The fatigue crack growth rate can easily be determined up to the order of 10À9mm/cycle.For the experiments at elevated temper-atures,a high-frequency inductor is used and the testtem-Fig.2.Vibratory stress and displacement field and computer controlsystem.Fig.3.Ultrasonic fatigue superposed on static loading.1440 C.Bathias /International Journal of Fatigue 28(2006)1438–1445perature can reach to1000°C.The computer system has a thermocouple board that can send the temperature ana-logue signal to an A/D converter,so that the computer is able to control heating of the specimen by using a circuit breaker.Because the Young’s modulus decreases at high temper-atures,the resonance length of ultrasonic fatigue specimen will be shorter than that at ambient temperature.For tem-perature sensitive materials,this change must be taken into account.For example,in a crack growth experiment,we usually start crack initiation at ambient temperature with adequate resonance length.Then,by cutting offboth ends, we can obtain the resonance length at high temperature.An early design of ultrasonic fatigue equipment at high temperature can be found in reference[3].The system described there is capable of studying fatigue crack growth rate and fatigue thresholds at22kHz,at elevated tempera-ture of200–500°C in argon environment in heat chamber, and at20°C using water as a coolant.4.Low temperature test equipmentLet us now discuss the possibility of testing materials at low temperature.A system for ultrasonic fatigue tests at cryogenic temperature has also been developed in our lab-oratory[4].In the laboratory,liquid nitrogen,liquid hydro-gen and liquid helium are used to create a cryogenic temperature atmosphere.However,liquefied gasses are expensive,especially liquid helium.The fatigue tests at very low temperature for titanium alloys used in space rockets need a lot of liquefied gas because the tests would take a very long time,if the conventional fatigue testing were employed.This is another advantage of the ultrasonic fati-gue method that can contribute to the reducing testing time considerably.The machine with a computer control system works at20kHz at cryogenic temperatures(77and20K) for studying fatigue behavior of the titanium alloys used in rocket engines.The device consists of three parts:a cryostat,a mechan-ical vibrator and a controlled power generator.Fig.6 shows the principal aspects of this machine,which is sim-pler than a conventional hydraulic machine.The functions of the converter and the horn are the same as in other ultra-sonic fatigue apparatus:the converter changes an elec-tronic signal into a mechanical vibration;the horn plays the role of displacement amplifier.A Dewar cryostat con-tains liquefied gasses to keep the testing temperature constant.Ultrasonic fatigue tests at cryogenic temperatures for load ratio R>À1are also possible by adding a second horn to the other end of the specimen.Another example of low temperature test system is that of Stanzl’s laboratory[5],where a temperature environ-ment of77K is guaranteed by liquidnitrogen.Fig.5.Specimen installed in the machine including measuring and heatingdevices.Fig.6.Low temperature and high frequency fatigue testing machine.C.Bathias/International Journal of Fatigue28(2006)1438–144514415.Thin sheet test equipmentThe geometry of ultrasonic fatigue specimens is usually with either cylindrical or plane form with reduced section in the central part to form a higher stress to accelerate the test process.Theoretically,as the excitation frequency of the machine coincides with one of the resonance frequencies of the specimen,the specimen will vibrate in that fre-quency.To put this theory into practice,however,is not an easy task.On the one hand,thefinishing of specimens to the design requirement is a delicate work.On the other hand,the horn where the specimen is installed may change the real frequency of vibration of the system,if the connection is not well designed.This is the use espe-cially for load ratio R>À1when another horn is neces-sary or when plane specimens are very thin.It was indicated long ago[3]that the use of a two-horn system and a positive constant mean stress is favoured for avoid-ing transverse vibration.A series of thin sheet tests of ferrous materials[6]have been conducted.The purpose is to determine the fatigue strength(or S–N curve)at109cycles and the threshold of cracking at a small propagation speed of10À12m/cycle, both at a stress ratio of R=0.1.The key to the execution of the test is how tofix the thin specimen to the ends of amplifying horns.We cannot use the same kind of setscrew as that used in the attachment of ordinary specimens(spec-imens with the thickness–length ratio of6–8%).Among the methods for linking two steel components,are riveting, bolt jointing,welding and gluing(Fig.7).6.High pressure piezoelectric fatigue machineIt is well known that it is difficult to carry out a fatigue test under high pressure with a conventional machine.The problem comes from the displacement of an actuator through the wall of an autoclave.The use of a piezoelectric fatigue system eliminates this problem because it is easy to get zero displacement at the location where the horn under a pressure crosses the wall of the autoclave.Thus,a high-pressure piezoelectric fatigue machine which works up to300bars has been built in our labora-tory.The machine is shown in Fig.8.Fig.7.Test system of thinsheet.Fig.8.Autoclave description.1442 C.Bathias/International Journal of Fatigue28(2006)1438–1445With this device,it has been shown that hydrogen under a pressure of100bars has an effect on S–N curve of IN718 and of Ti6Al4V at room temperature.7.Ultrasonic fretting fatigueFretting fatigue is generally promoted by high fre-quency,motions vibratory in low amplitude and com-monly occurs in clamped joints and shrunk-on components[7,8].The surface damage produced by fretting can take the form of fretting wear or fretting fatigue where the materials’fatigue properties can be seriously degraded. Fretting fatigue is a combination of fretting and fatigue processes.It involves a number of factors,including mag-nitude and distribution of contact pressure,the amplitude of relative slip,friction forces,surface conditions,contact materials,cyclic frequency and also environment.Great efforts have been made to quantify fretting fatigue in terms of these factors,but limited success has been achieved. More often,fretting fatigue characteristics are studied in the laboratory experimentally by using a contact pad clamped to a fatigue specimen in order to determine S–N curves with fretting and thereby to establish the fatigue strength reduction factor for a particular material.But these studies,generally performed on the conventional ten-sion–compression fatigue machine with a low frequency, have some drawbacks:(1)The slip amplitude of fretting fatigue is usually cou-pled with the fatigue stress and to change the slip amplitude,pads with different gauge length are needed;(2)The frequency is low and is not appropriate to simu-late the high frequency small elastic vibration cycles of mechanical,acoustical or aerodynamical origin.On the other hand,in some industries,such as the automobiles and the railways,the determination of high cyclic fretting fatigue properties up to108or even109cycles is necessary.The experiment is bound to be time-consuming and uneconomical.In our laboratory,an ultrasonic fretting fatigue test technique at a frequency of20kHz has been developed, in which the fretting slip amplitude can be changed without changing the fretting pads.Fig.9shows a schematic dia-gram of an experimental set-up.It consists of two parts. Thefirst is the ultrasonic fatigue test machine,which has been widely used in the fatigue test for both endurance and crack propagation.Each element in the machine is designed to have a resonant frequency about20kHz and an automatic unit maintains the whole system operating at the resonant frequency.The second is afixture to hold the two cylinder pads pressed against the specimen by two springs.The normal contact force is measured and controlled by the displacement of the springs.Moreover, the use of the springs will eliminate a discernible changes in load should wear occur.The whole experimental system is controlled by a PC.The test system has the functions of regulating test parameters,recording the relative slip amplitude,normal force and stress.By changing the position of pads along the specimen axis,the desired value of relative slip ampli-tude can be obtained.The stress of fretting fatigue test is determined by the position of the pads for the loadratio Fig.9.Schematic experimental system for ultrasonic fretting-fatigue.C.Bathias/International Journal of Fatigue28(2006)1438–14451443R=À1.For R>À1,the total stress is the superposition of static stress applied by the traction machine and the dynamic stress of vibration.According to the position of pads in the specimen axis, the displacement that provokes the fretting can be chosen between0.1l m and tens of microns.By a gauge mounted on the specimen at point x i=0,vibration deformation of the specimen can be measured.It is worth recalling that the reason for Mason’s pioneer work on the ultrasonic fatigue machine was to study fret-ting[9].8.Torsion fatigue testingIn a review article,Stanzl indicated that more recently time a new technique and equipment has been developed [10],which allows one to perform torsion fatigue at ultrasonic frequencies.The mechanical parts of this equipment must be designed so that torsion resonance vibration can be generated(Fig.10).As the shear mod-ulus is smaller than Young’s modulus,all vibrating parts, including the specimens,must be smaller in order to obtain the resonance.Besides this difference,theotherFig.10.Piezoelectric torsion test equipment in Bathias’Laboratory.Fig.11.System for ultrasonic fatigue experiments in three-point bending.1444 C.Bathias/International Journal of Fatigue28(2006)1438–1445experimental details,such as amplitude measurement and control are much the same as those for the axial ultra-sound fatigue loading.The superposition of axial load is possible and has been investigated in measurements on ceramic materials.Superposition of small compressive loads leads to a lifetime twice as long as that of pure cyclic20kHz torsion loads because of the increased fric-tion forces[10,11].9.Three-point bending fatigue testingThe three-point bending ultrasonic fatigue testing sys-tem developed in the Bathias’laboratory is illustrated in Fig.11.This system was developed for testing certain aluminium alloy based metal-matrix composites used in automobile industry or brittle materials such as ceramics or tita-nium–aluminium alloys.The system containing generator, converter,booster,acoustic horn and specimen is con-trolled by a personal computer.A static load or a static dis-placement is introduced with a tensile testing machine through the booster.Thus,the piezoelectric fatigue machine bending is able to work from R=0.1to R=0.5 and beyond,at20kHz.The maximum displacement in the middle of the three-point bending specimen is of the order of30l m.An advantage of this technique is the small size of the specimen,less than1inch!10.ConclusionsDuring the1990s,several devices have been designed using piezoelectric converters working at20or30kHz. Several advantages have been underlined:•This new method is recommended to study the gigacycle fatigue and the threshold regime.•The duration of a test is at least400times shorter than with a conventional machine.It saves money.•The piezoelectric fatigue machines are able to work at negative R ratio and very high R ratio.•This system is able to operate at high temperature,cryo-genic temperature,high gas pressure,fretting-corrosion, torsion and three-point bending. AcknowledgementsThe author is grateful for many comments and discus-sions from Professor PC,Paris.References[1]Wu,T,Bathias C.Contract DRET No.8707101,ITMA/CNAM;1991.[2]Bathias C,de Monicault JM,Baudry G.Automated piezoelectricfatigue machine for severe environments.ASTM STP 2002;1411:3–15.[3]Stanzl S et al.Ultrasonic Fatigue,Fatigue1996;96:1887–98.[4]Ebara R.Corrosion fatigue in practical problems.In:Nisitani H,p,and Exp Fract :Comp.Mech.Publ.;1994.p.347–76.[5]Huo T.Ultrasonic fatigue of Ti and alloys at cryogenic temperature,the`se de doctorat,CNAM;1996.[6]Buchinger L,Stanzl S,Laird C.Dislocation structures produced bylow-amplitude fatigue of copper single crystals at77K.Philos Mag A 1990;62(6):633–51.[7]Wang Q,Bathias C.Re´alisation d’un syste`me de fatigue vibratoire surtoˆles minces,Partie I:Conception et mise au point,Contrat Renault-Rapport96-2,ITMA/CNAM;1996.[8]Lindley TC.Fretting Fatigue in Engineering Alloys.Int J Fatigue1981;19(s1):s39–49.[9]Mason e of high amplitude strains in studying wear andultrasonic fatigue in metals.In:Well JM,Buck Roth OLD,Tien JK, editors.Ultrasonic fatigue,proceedings of thefirst international conference on fatigue and corrosion fatigue up to ultrasonic frequencies.(PA)USA:The Metallurgical Society of AIME;1982.p.87–102.[10]Stanzl S.Fatigue testing at ultrasonic frequencies.J Soc Environ Eng1986;25(1):11–6.[11]Mayer HR,Tschegg EK,Stanzl-Tschegg S.High cycle fatiguemeasurements on ceramic materials in combined loading.In:Fourth international conference on biaxial and multiaxial fatigue.Paris: ESIS;1994.p.357–8.C.Bathias/International Journal of Fatigue28(2006)1438–14451445。