Philosophy

Fatigue crack initiation and growth in a 35CrMo4 steel investigated by infrared thermography

Description
Fatigue crack initiation and growth in a 35CrMo4 steel investigated by infrared thermography
Categories
Published
of 10
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  Fatigue crack initiation and growth in a 35CrMo4 steel investigatedby infrared thermography O. PLEKHOV 2 , T. PALIN-LUC 1 , N. SAINTIER 1 , S. UVAROV 2 and O. NAIMARK 2 1  E.N.S.A.M. CER de Bordeaux, Laboratoire Mat´ eriaux Endommagement Fiabilit´ e et Ing´ enierie des Proc´ ed´ es (LAMEFIP), EA 2727, Esplanade des   Arts et M´ etiers, 33405 Talence Cedex, France, 2  Institute of Continuous Media Mechanics RAS, 1 Koroleva str., 614013 Perm, Russia Received in final form 23 June 2004 ABSTRACT The present work is devoted to the investigation of fatigue crack initiation and growth inmiddle-cycle fatigue ( ∼ 10 5 cycles). Smooth specimens made of 35CrMo4 quenched andtempered steel were loaded in fully reversed plane bending. Temperature field evolutionintimewasrecordedwithaninfraredcamera.Theexperimentalresultsshowthatthelocalheating of metal under fatigue loading is a sensitive and accurate enough manifestationof small fatigue crack initiation. It is shown that the time evolution of the spatial standarddeviation of the temperature field can be used to investigate the damage localization andto monitor both the crack initiation and the current location of the fatigue crack tip. Thisshould help to investigate the behaviour of defect during cyclic loading. Keywords crack growth; crack initiation; infrared thermography; middle-cycle fatigue. INTRODUCTION  The evolution of microstructure in metallic materials un-der cyclic loading has been the object of intensive studiesduring the last century. It has been shown that the fa-tigue of metal is accompanied by the appearance of spe-cific dislocation patterns such as veins or channel struc-tures,persistentslipbands-matrixstructures,labyrinthsorshell structures. 1  This evolution simultaneously involvesa great number of strong nonlinear interactions of defectsat different scales. These phenomena lead to a specificchange of the macroscopic material response. The description of these interactions requires both de-tailed theoretical investigation of the nonlinear laws of defect kinetics and the development of new experimen-tal techniques. Based on the statistical physics approach,Naimark proposed a powerful way  2 to obtain the nonlin-ear kinetic equations for defect density. To progress in thedevelopment of this approach and to identify additionalconstants in the constitutive equations, a detailed inves-tigation of the processes of plastic deformation, damageand failure is strongly needed. Anyphenomenonoflocalization,forexample,plasticde-formation in persistent slip bands or damage localizationsuch as fatigue crack initiation, induces a heterogeneousdistribution of heat sources on the surface of a specimen Correspondence: Thierry Palin-Luc. E-mail: thierry.palin-luc@lamef.bordeaux.ensam.fr that makes it interesting to investigate the infrared radi-ation from the surface. In recent years, progress in thedevelopment of new infrared cameras allows us to use in-frared thermography as a powerful, non-destructive andnon-contacting technique for the investigation of the fa-tigue behaviour of materials.First, observations of temperature evolution by different experimental techniques are presented in Refs [3] and [4].Itwasshownthatinfraredthermographycouldbeconsid-ered as an efficient way to estimate the fatigue propertiesof materials and structures. A technique for a fast estima-tionoftheendurancelimitofmetalswasproposedin1986and called the Risitano’s technique. 5 , 6  This approach wasalso applied by Luong 7 and Blarasin et al  . 8  Then, it hasbeen developed to determine the whole S–N  curve. 9 Differential infrared thermography was also used suc-cessfully to estimate the value of the stress intensity fac-tors. 10 , 11  This technique was adopted for fatigue crack tiplocalization and stress intensity factor estimation in Refs[12] and [13] based on the idea proposed in Ref. [11]. The present work is devoted to the investigation inmiddle-cycle fatigue (around 10 5 cycles) of the tempera-tureheterogeneity,appearingwithfatiguecrackinitiation,at the surface of smooth specimens made of 35CrMo4quenched and tempered steel. The aim of this paper isto find an appropriate technique for detecting in spaceand in time the plastic strain localization, initiation andpropagation of small fatigue cracks by monitoring the c  2005 Blackwell Publishing Ltd. Fatigue Fract Engng Mater Struct  28 , 169–178 169  170 O. PLEKHOV et al  . characteristics of the temperature field on the specimenfree surface. EXPERIMENTAL PROCEDURE  Material and specimen   The material investigated is 35CrMo4 quenched andtempered steel, which has been already studied by  Vivensang. 14 Its chemical composition is given in Table 1. The heat-treatment procedure before specimen machin-ing was as follows: 30 min at 850 ◦ C, then quenched inoil, then 1 h at 550 ◦ C, and then left in air to cool. Afterthis treatment, the mean grain size was around 10 µ m. 14  The microstructure consisted of fine tempered marten-site. The mechanical properties of this steel under qua-sistatic monotonic tension are given in Table 2. The specimens used to carry out the fatigue tests weremachined from round heat-treated bars (external diam-eter 26 mm) and their geometry is illustrated in Fig. 1. The theoretical stress concentration factor, K  t  , of thesesmooth specimens in bending is 1.05. 15 Before experi-ments, all the specimens were polished with emery paperanddiamondpowderuptograde1 µ m.Allthefatiguetests werecarriedoutinfullyreversedsinusoidalplanebending with a resonant electrodynamic fatigue testing machinedesigned at ENSAM-LAMEFIP. The loading frequency  was56Hz.Thetestsautomaticallystoppedwhentheload-ing frequency decreased to more than ∼ 10% of the initialresonant frequency (undamaged specimen). This corre-spondedtoafatiguemacrocracksizeofseveralmillimetresin depth.  Table 1 Chemical composition of the 35CrMo4 steel studied(wt%)C Mn Si S P Ni Cr Mo0.37 0.79 0.30 0.010 0.019 < 0.17 1.00 0.18  Table 2 Mechanical properties of the quenched and tempered 35CrMo4 steel Young modulus, Yield stress, Maximum tensile strength, Endurance limit for Fatigue limit for Elongation after  E  (GPa) Re 0 . 2 (MPa) R m (MPa) 10 7 cycles (MPa) 10 5 cycles (MPa) fracture A% (%)200 950 1068 525 660 11.5 ∗  The size of the corresponding image region on the specimen is dependent on the camera lens. Two lens are now available in our laboratory for the IR camera: the first with a focal distance of 50 mm has a spatial resolution of 0.14 mm for a specimen at 200 mm, the second with amagnification (x1) has a special resolution of 25 µ m, the picture size is around 9 mm × 7 mm. †  Maximum frequency to take IR pictures (the shortest delay between two IR picture is 1/500s). ‡ NETD: Noise Equivalent Temperature Difference. This is the thermal resolution, i.e. the smallest temperature difference which can bemeasured without numerical treatment like the mean value. ∗∗ Number of bits used to digitalize the thermosignal. Infrared (IR) recording conditions During the fatigue tests, an infrared camera (CEDIP JadeIII MWR) was used to record the temperature field evo-lution on the specimen surface. The main technical char-acteristics of this camera are as follows: spectral range 3–5 µ m,maximumpicturesize ∗ 320 × 240pixels,maximumframingrate † 500Hz,NETD ‡ < 25mKat300 ◦ Kanddig-ital conversion ∗∗ 14 bits. To increase the surface emissiv-ity properties, the specimen surface was painted in black (matt paint) after polishing. The lenses of the IR camera were not located exactly perpendicular to the specimensurface. Indeed, to avoid any parasitic reflections comingfrom the specimen surface, the IR camera was located sothat there was a small angle (a few degrees) between thenormalvectortothespecimensurfaceandthecameraaxis.Furthermore, a matt black piece of material was put overand all around the setup (Fig. 2) to reduce artefacts dueto external radiative heat sources. The concave surface of the specimen given in Fig. 1 was observed with the IR camera. Because of the large radius (  R = 40 mm), the the-oreticalstressconcentrationfactorisclosetoone;thusthenormal stress distribution is almost uniform on the areamonitored on this face. The duration of each experiment presented hereafter was between 20 and 25 min. Duringthis short duration, the room temperature changed very slowly and with a low magnitude range ( < 0.5 ◦ C). Forlonger tests, it is necessary to use an unloaded referencespecimen located close to the tested one. All the IR datapresented in this paper were recorded with the synchro-nized mode of the camera. It means that all the picturesof a film were recorded when the bending moment wasmaximum.  Treatment of IR data Damagelocalizationandfatiguecrackinitiationshouldbeaccompanied by a non-uniform temperature field at thespecimensurface.Thispropertywasinvestigatedbycom-puting the spatial standard deviation of the temperature(SDT) field using the following procedure. c  2005 Blackwell Publishing Ltd. Fatigue Fract Engng Mater Struct  28 , 169–178  FATIGUE CRACK INITIATION AND GROWTH 171 Fig. 1 Specimen geometry (arrows indicate plane bending moments; sizes in mm). Fig. 2 Fatigue testing machine and IR camera with the matt black materialremoved to show the setup.  The IR data were treated a posteriori  , i.e. after the fa-tigue test, when macrocrack initiation had been detectedby the resonance frequency drop of the testing machine.Several   j spatial areas were chosen such that each areacontained several points Z  (  x i , y i ) (pixels—from 9 to 100—on the picture) where the temperature was known at each time t  k  . This thermo-signal can be written as thesequence: T  ( t  , Z  ) ={ T  ( t  1 , Z  ) , T  ( t  2 , Z  ) ,..., T  ( t  k  , Z  ) ,..., T  ( t  N , Z  ) } .  The Standard Deviation of Temperature in   j , denotedby SDT   j , was computed through Eq. (1) by employingthe IR experimental data obtained for the N  p points in   j16 (observation window of the IR camera projected onthe specimen surface):SDT 2   j ( t  k  ) = 1  N  p − 1  Z  ∈   j ( T  ( t  k  , Z  ) − ¯ T    j ( t  k  )) 2 , (1) where¯ T    j ( t  k  ) = 1  N  p   N   p i  = 1 T  ( t  k  , Z  (  x i , y i ))isthespatialmean value over   j of the temperature at time t  k  . An example c  2005 Blackwell Publishing Ltd. Fatigue Fract Engng Mater Struct  28 , 169–178  172 O. PLEKHOV et al  . Time (c) SDTStep IIIStep IVStep IIStep I Fig. 3 (a) Temperature pattern. (b) SDTevolution versus time for five different areas.(c) Schema of the four different stages. of five spatial areas (  1 to  5 areas) is shown in Fig. 3 with typical temporal evolution of the standard deviationSDT. These areas are defined by four circles around thehot spot with the following diameters: 0.6, 1, 1.5, 2 and2.5 mm, and one circle far from the hot spot (top of thepicture Fig. 3a). EXPERIMENTAL RESULTS Figure 3 shows the evolution of the SDT as a function of time during the last 123 s of a specimen life (loaded at astress amplitude of 650 MPa; the total life is 83 300 cyclesor ∼ 1500s)forfivedifferentareasizes.Forthefourcurvesrelatedtothefourareasaroundthehotspot(circlesinthemiddle of Fig. 3a) the peak shape is the same. The shapeof the curves tends to a common smooth curve (without any peak) elsewhere (far from the hot spot) as illustratedby the curve in Fig. 3b related to the small area (circle) at the top of Fig. 3a. Four main steps with two knee pointscan be identified as illustrated in Fig. 3c: - Step I : The SDT is stable, the temperature field is homo-geneous, independent of the area of investigation. c  2005 Blackwell Publishing Ltd. Fatigue Fract Engng Mater Struct  28 , 169–178  FATIGUE CRACK INITIATION AND GROWTH 173 Fig. 4 Close up pictures of damage localization (left) and the cross section of the temperature field in the hottest points (right). - Step II : The SDT increases markedly. A heterogeneoustemperature field takes place. The magnitude increase of SDT depends on the area of investigation.- First knee point.- Step III : Simultaneous drop of the SDT in all areas.- Second knee point.- Step IV : Sudden and sharp increase in the SDT. It has to be noticed that the first knee point is observedbefore the sharp change in the resonance frequency of the fatigue testing machine. The knee-point shape of this curve is discussed in detail later. Nevertheless, onecan say that this phenomenon identifies the beginning of the correlation behaviour between the adjacent points inthetemperaturefieldandthethermalbehaviourofthehot  c  2005 Blackwell Publishing Ltd. Fatigue Fract Engng Mater Struct  28 , 169–178
Search
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x