Fatigue Crack Propagation and Charpy Impact Properties CABRILO 4 17.pdf  694 MATERIALE PLASTICE ♦54♦No

  • View

  • Download

Embed Size (px)

Text of Fatigue Crack Propagation and Charpy Impact Properties CABRILO 4 17.pdf  694 MATERIALE PLASTICE...

  • http://www.revmaterialeplastice.ro MATERIALE PLASTICE 54No. 4 2017694

    Fatigue Crack Propagation and Charpy Impact Propertiesin Armor Steel Welds

    ALEKSANDAR CABRILO1*, MIROSLAV CVETINOV21Faculty of Technical Sciences, University of Novi Sad, Trg D. Obradovica 6, 21000 Novi Sad, Serbia,2 Faculty of Sciences, University of Novi Sad, Trg D. Obradovica 4, 21000 Novi Sad, Serbia

    The process of welding armor steel is a complex process because of possible welding faults, appearing inthe weld metal zone in the form of cracks and pores. Austenitic filler material is traditionally used forwelding armor steels. For heavy structural engineering such as armored military vehicles, which are frequentlyunder the effect of dynamic load, it is important to know the dynamic properties of the most sensitive areaof welded joints, the weld metal zone. Due to a significant interest in quantification of material resistance tocrack initiation and propagation, the fatigue crack growth rate was measured in the welded metal zone,while the resistance to crack growth in the weld metal was tested by the amount of austenite transformedinto martensite. Accordingly, the threshold stress concentration factor was 10 MPa m1/2. XRD spectralanalysis revealed direct transformation of - austenite into -martensite.

    Key words: Armor steels, Fatigue crack growth, Austenitic stainless steel and Martensitic transformation.

    Armor steel belongs to the ultra-high tensile strengthand hardness group of steels. The welding of armor steel iscomplicated due to the high percentage of carbon contentin the base metal and the presence of faults in the form ofcracks and pores [1] in the weld metal zone, wherebyfractures may be initiated in the weld metal. Austeniticfiller material is traditionally used for armor steel weldingbecause of hydrogen dilution improved in an austeniticphase [2, 3]. The filler material, in armor steel welded jointshas lower mechanical properties than the base material,i.e. the filler material is the weakest point of the weldedjoint [4]. After the welding process, solidification crackingmay result from high thermal expansion of the austeniticstainless steel [5,6] and invisible defects may be createdin the weld metal zone [7].

    For heavy structural engineering, such as armoredmilitary vehicles frequently being under the effects ofvariable loads [8], mechanical properties of welded jointsand the weld metal zone must be known. Due to variableloads, cracks created in the weld metal may easilypropagate towards the sensitive fusion line, followed bytheir possible rapid growth [9].

    For armored vehicle structures safe and rationaldimensioning, it is necessary to know dynamic effectsextreme values and time periods. Therefore, there is asignificant interest in material resistance related to crackinitiation and propagation, as well as in dynamic forceconditions. A fatigue crack growth rate characteristic inthe linear and threshold region in metal weld is consideredas an important property, since it shows a fault tolerantability of this part of welded joint [10,11].

    Austenitic filler material is unstable and gets transformedinto martensite during fatigue crack propagation due toplastic deformation at the crack tip [12]. During themetastable austenite deformation, two types of martensiticstructures can be formed: - martensite with hexagonalclose packed and -martensite, with body centered cubiccrystal structure. Austenite into martensite transformationis related to the stacking fault energy. If the stacking faultenergy is < 20 J/m2, transformation proceeds according tothe model: . If it is larger, then the direct transformation occurs [13]. It is known that manganese

    and nickel stabilize martensite and prevent martensitictransformation. It should be noted that a upper limit ofstacking fault energy of austenite-martensite phasetransition phase transition varies [14,15]. An amount ofaustenite transformed into martensite is directly related tocrack growth resistance in the weld metal [16].

    The main goal of this study was to investigate the impactenergy by instrumented pendulum and fatigue crackgrowth in the Paris region. Martensitic transformationeffects on fatigue crack growth in the Paris region wereinvestigated by X ray diffraction. Fracture surfaces for theimpact energy and fatigue crack growth tests were alsoinvestigated by Scanning Electron Microscope (SEM).Subsequently, samples in the weld metal region werestudied by tensile strength test, hardness measurements,metallography and chemical analysis.

    Experimental partMaterials and methodsMaterials and welding process

    Gas metal arc welding (GMAW) and AWS ER307 solidwire is used for welding armor steel Protac 500. Weldingdirection is parallel to the rolling direction. Cold rolled plates12 mm thick are cut to the required dimensions (250 x 100mm), while V joint under the angle of 55 is prepared byWater Jet Device figure 1. Robot Kuka and Citronix 400Adevice was used during the welding process testing.Robotic welding is used for human factor effectelimination, in order to allow a fine adjustment ofparameters and results repeatability. Wire diameter is 1.0mm while figure 1 shows V joint dimensions and four -pass welding configuration.

    * email: cabrilo@uns.ac.rs

    Fig. 1. Schematic drawing ofedge preparation andwelding configuration

  • http://www.revmaterialeplastice.roMATERIALE PLASTICE 54No. 4 2017 695

    Base material chemical composition obtained byspectro - chemical analysis is shown in table 1, while thefiller material chemical composition is shown in Table 2.Spectro-chemical analysis was performed after thewelding process.

    As for the fatigue crack growth rate test, it is importantto obtain the welded joint without porosity and cracks.Therefore, radiographic testing was being performed afterthe welding process.

    Mechanical property testsWelded joint tensile strength testing was performed in

    transverse direction of the weld bead. It should be notedthat specimens was cut with Water Jet Device, to eliminatepossibilities of thermal effects to high hardness steel.Tensile strength testing was made on servo - hydraulictesting machine Instron 8033. The loading rate was set as0.125 mm/s until fracture took place. During the tensiletest, extensometer was used to monitor and record thestress-strain curves and strain gauge was employed to verifythe results obtained by the extensometer.

    Metallography testingThe microstructural examination was performed using

    a Leitz-Orthoplan metallographic microscope and ascanning electron microscope JEOL JSM 6460LV at 25 kV.The samples were ground using SiC papers, polished witha diamond paste and finally etched with a mixture HCl andHNO3 reagent to reveal the structure.

    Fatigue crack growth testThree point bending specimen, SEN (B) was used for

    testing [17]. The schematic drawing of specimen forfatigue crack growth test is shown in figure 2. Specimenswere cut by Water Jet Device, to eliminate any possibilityof armor steel thermal treatment. After getting finalmeasures in the grinding process, 5 mm long machinednotch was created on specimens in the direction parallelto welding figure 3, according to the E-647 standard [18].The fatigue pre-crack was inserted before the crack growthrate tests, in accordance with ASTM E647 [18]. The lengthof the fatigue precrack was 4.7 mm. The fatigue pre-crack was realized with a high-frequency CRACTRONICpulsator, at a load ratio R = 0.33, followed by a constantloading frequency of 170 Hz. Fatigue crack growth ratewas tested on high-frequency CRACTRONIC pulsator, themodel with force and frequency control of 145 Hz. Theconstant sinusoidal shape was used, while the testing wasmade under the load ratio R=Kmin/Kmax=0.1.

    During testing procedure, the crack length wasmeasured by RUMUL RMF A-10 measuring foils. In thecourse of experiments, the number of cycles for eachcrack growth of 0.05 mm was automatically recorded. Onthe basis of these records, the diagram of a-N was drawn.

    a-N curves of dependence were used for crack growthrate da/dN determination.

    The initial dynamic load was determined on the basis oftensile characteristics and maintained during the constanttesting period. While the crack was propagating in thenumber of cycles, the value of K was increasing with thecrack length growth.

    For this test, three specimens were used, in same testingconditions and initial loads. The result was average valueof three measurements. A fracture surface was analyzedby Scanning Electron Microscope JEOL JSM 6460LV at 25kV.

    Quantitative phase analysis by X-ray diffractionX-ray diffraction was used to identify a martensitic

    transformation amount formed during the crackpropagation, under the effect of fatigue load. Investigationwas undertaken by X-ray diffraction in BragBrentano :2reflection geometry, at a room temperature. Diffracto-grams were recorded on a Philips X-ray diffractometerhaving a copper tube PW 1830 generator, a PW 1820goniometer fitted with a post-diffracted graphitemonochromator and a scintillation detector attached to aPW 1710 controller (30 kV, 30 mA generator settings, CuKradiation). LaB6 was used as an external standard for peakposition calibration and for instrumental peak broadeningassessment. XRD data were collected over the 2 rangeof 40 to 60, with a step size of 0.05 and an expositiontime of 2 s per step.

    Martensite to austenite ratio was measured on thefracture surface. After the analysis, the 0.05 mm. Thick



    Fig. 2. Schematic drawing of the SEN (B) specimen withdimensions.

    Fig. 3. Specimen orientation with respect to the weld axis forfatigue crack growth test.

  • http://www.revmaterialeplastice.ro MATERIALE PLASTICE