Jtam.dvi JOURNAL OF THEORETICAL AND APPLIED MECHANICS 47, 4, pp. 751-759, Warsaw 2009 INFLUENCE OF THE NOTCH RADIUS ON CHANGES OF THE ∆J PARAMETER UNDER FATIGUE CRACK GROWTH RATE Dariusz Rozumek Opole University of Technology, Faculty of Mechanical Engineering, Opole, Poland e-mail: d.rozumek@po.opole.pl In the paper, the influence of the notch radius on changes of the ∆J parameter under low and high-cycle fatigue is discussed. The tests were carried out on plates made of FeP04-UNI 8092 deep-drawing steel. The specimens were characterised by double symmetric lateral notches with the notch root radius ranging from 0.2mm to 10mm. The MTS 809 servo-hydraulic device was used for tests performed at the Department ofManagement andEngineering inVicenza (PadovaUniversity). All fa- tigue tests were performed under force control, by imposing a constant value of the nominal load ratio, R = 0, and a load amplitude Pa = 6 and7kN (which corresponded to the nominal amplitude of normal stres- ses σa = 100, 117MPa before the crack initiation). The test frequency ranged from 13 and 15Hz. Key words: fatigue crack growth, number of cycles, ∆J parameter, load ratio, notch Notations a – crack length a0 – notch depth da/dN – fatigue crack growth rate n′ – cyclic strain hardening exponent w – specimen width E – Young’s modulus Kt – theoretical stress concentration factor K′ – cyclic strength coefficient N – number of cycles Pa – amplitude of load R – load ratio 752 D. Rozumek Y – correction factor ν – Poisson’s ratio ρ – notch tip radius σmax – maximum stress σnom – nominal stress σu – ultimate tensile stress σy – yield stress ∆J – energy parameter ∆εp – plastic strain range 1. Introduction Fatigue crack growth is directly connected with existing stress concentrators in a material. Cracks usually initiate in the notch (micro notch). In many ca- ses, shape of the notch influences life of the tested material (element). That problem was discussed in a paper by Williams (1952). He showed that, ac- cording to theory of elasticity, the asymptotic stress state near a re-entrant corner is singular and its degree of singularity just depends on the notch ope- ning angle. The stress field intensity is a function of the overall geometry of the component and the far-field loading. Since Williams’s pioneering work, several studies have been undertaken to determine the stress field present at various notch geometry andat singular pointswhere the interface between two elastic solids intersects the traction-free edge. Some recent papers (Lazzarin et al., 1997; Taylor, 1999) suggest that there exists a strong relation between the dimension of this critical zone and El-Haddad et al. (1979) length para- meter l0(∆Kth0). Averaging the maximum principal stress in a finite volume ahead of the notch tip could be a valid idea from the engineering point of view, but presents some drawbacks from the theoretical point of view. In fact, the more l0 increases, the greater the influence of the other principal stresses becomes (values of l0 equal 3.0mm have been determined in some cast irons byTaylor et al.) (1996)). It is the authors’ opinion that an energetic approach is undoubtedly preferable, despite the fact that it necessarily results in some algebraic complications. The approach to fatigue crack growth using the J parameter seems to be themost promising (Tanaka, 1983). The ∆J parame- ter allows one to estimate changes of local and global energies while fatigue crack growth in the tested element (Rozumek et al., 2006; Berto andLazzarin, 2007; Rozumek, 2008). Influence of the notch radius... 753 The aim of this paper is taking into account the influence of the notch radius on changes of the ∆J parameter under fatigue crack growth rate. 2. Materials and test procedure The tests were carried out on plates made of FeP04-UNI 8092 deep-drawing steel,weakened by symmetric lateral notches of varying acuity. In the tests, the MTS 809 servo-hydraulic device was used at the Department of Management and Engineering in Vicenza (Padova University). Table 1 contains chemical compositions of the material, and its some mechanical properties are shown in Table 2. Table 1.Chemical composition (in wt%) of FeP04 steel C Mn Si P S Al Cu Fe 0.05 0.30 0.05 0.032 0.02 0.043 0.07 balance Table 2.Monotonic quasi-static tension properties of FeP04 steel σy [MPa] σu [MPa] E [GPa] ν 210 330 191 0.30 Coefficients of the Ramberg-Osgood equation describing the cyclic strain curve under tension-compression with R = −1 for FeP04 steel are the follo- wing: the cyclic strength coefficient K′ =838MPa, the cyclic strain hardening exponent n′ =0.220. All fatigue tests were performed under force control, by imposing a constant value of the nominal load ratio, R = 0, and a load am- plitude Pa = 6 and 7kN (which corresponded to the nominal amplitude of normal stresses σa = 100, 117MPa before crack initiation). The test frequ- ency ranged from 13 and 15Hz. The specimens had double symmetric lateral notcheswith notch root radii ranging from0.2mmto 10mm(Fig.1). The the- oretical stress concentration factor in the specimen under tension Kt =9.61, 4.30, 3.23 and 1.85 was estimated by making use of the model (Thum et al., 1960). The surfaces of the specimens, w = 50mm wide, had been accurately polished in order to make the cracks originated from the notch tip easily di- stinguishable. The fatigue crack increments weremeasuredwith amicrometer located in a portable microscope withmagnification of 20 times and accuracy 0.01mm. At the same time, the number of loading cycles N was registered. 754 D. Rozumek Fig. 1. Geometry of specimens characterised by: (a) sharp and (b) blunt notches; all dimensions in mm 3. Experimental results and discussion The tests of fatigue crack growth in FeP04 steel subjected to tension were performed in the lowandhighcycles fatigueundera controlled loading.During the tests, a number of cycles to the crack initiation Ni, (i.e. to themoment of occurrence of a visible crack) was determined, and fatigue crack lengths were measured. The test results were shown as graphs of the crack length a versus the number of cycles N and fatigue crack growth rate da/dN versus the ∆J parameter. The cracks initiated (minimal observable crack length about 0.1 to 0.2mm) at the same time on the left and on the right of the slot. From Fig.2 presenting the crack length a versus number of cycles N it appears that after changing the notch root radii ρ from 0.2 to 10mm, the fatigue life increases. It is evident that with the highest radii, the initiation phase, which depends on the stress conditions at the notch tip, prevails. The author proposed to describe the fatigue crack growth range according to the energy approach based on the ∆J parameter, which could be written as (Rozumek, 2004) ∆J1 =πY 2 1 (∆σ2nom E + ∆σnom∆εp √ n′ ) a (3.1) Influence of the notch radius... 755 Fig. 2. Fatigue crack length versus number of cycles for different radii of the notch root If Eq. (3.1) is related to short cracks and includes Kt, it takes the following form ∆J2 =πY 2 1 (∆σ2max E + ∆σmax∆εp √ n′ ) a (3.2) where: ∆σmax = Kt∆σnom is the maximum stress range, ∆σnom – nomi- nal stress range, a is the crack length, Y1 – correction factor dependent on specimen geometry and loading type Y1 =1.12+0.203 (2(a+a0) w ) −1.197 (2(a+a0) w )2 +1.930 (2(a+a0) w )3 a0 –notch depth, ∆εp –plastic strain range corresponding to the stress range. In Eqs. (3.1), (3.2) and (3.3), the exponent n′ concerns the functions in which materials slightly harden or are cyclically stable (Rozumek and Pawli- czek, 2004). A next modification of Eq. (3.1) including the notch influence by introduction of the correction factor Y2, allowing its application for short and long cracks, can be expressed as ∆J3 =πY 2 1Y 2 2 (∆σ2nom E + ∆σnom∆εp √ n′ ) a (3.3) where Y2 = √ 1+e−β with β= 3ρ(w−2a0)a ρ+w−2a0 and ρ – notch tip radius. 756 D. Rozumek Figures 3 and 4 show the fatigue crack growth rate da/dN versus ∆J relations under different notch root radii and load amplitude conditions. Fi- gure 3 presents a comparison of the parameter ∆J with and without the coefficient Kt, and in Fig.4 one can see the influence of different correction factors Y1 and Y2 on variations of the parameter ∆J. Equation (3.2) can be applied only for short fatigue cracks growthbecause local stress concentrations cause high stresses and accompanying strains (see Fig.3), and in the case of long cracks the values calculated according to Eq. (3.2) are overestimated. From Fig.3 it also appears that the smaller is the notch root radius, the gre- ater is the influence of stress concentration. That conclusion can be proved by literature. Lazzarin et al. (1997) proposed the correction factor Y2 including occurrence of the notch for the least (theoretical) notch radius ρ= 0. Intro- ducing the relationships proposed by Lazzarin et al. (1997) into Eq. (3.3), we obtain results in which the values including the notch influence are smaller than the values with no influence of the notch. For the notch radius ρ clo- se to zero, the results described by Eqs. (3.1) and (3.3) coincided, and for ρ= 10mm the greatest differences were obtained. It is inconsistent with the real behaviour of the notch and cracks in the elements (see Rozumek, 2008). Fig. 3. Fatigue crack growth rate da/dN versus the ∆J parameter with and without the factor Kt In order to obtain real results, the notch root radius ρ approaching infinity (ρ → ∞) was assumed and the correction factors Y2 and β were corrected. Then, for the least notch radius ρ, we obtain the greatest differences between the values obtained from Eqs. (3.1) and (3.3) (see Fig.4). Influence of the notch radius... 757 Fig. 4. Fatigue crack growth rate da/dN versus the ∆J parameter with and without the correction factor Y2 for (a) ρ=0.2, 1.25mm and (b) ρ=2.5, 10mm FromFig. 4 it appears that the is the greater the notch radius, the smaller is the difference between the results obtained from Eqs. (3.1) and (3.3). The greatest notch influencewas observed for ρ=0.2mm, up to about a=6mm. Figure 4wasdivided intoFigs.4a and4b.Figure 4a showshow the range of the parameter ∆J varies versus the fatigue crack growth rate for the notch radius ρ = 0.2 and 1.25mm with and without the correction factor Y2. Figure 4b presents the influence of the correction factor Y2 and its lack (Eq. (3.1) and Eq. (3.3)) for the notch radius ρ=2.5 and 10mm. As the notch radius increases from ρ = 0.2 to 10mm, the value of ∆J3 increases in relation to ∆J1. From Fig.4b it appears that for the notch root radius ρ=10mm the results coincide. However, for ρ=0.2 to 2.5mm there is a significant difference between the results including and not including the notch effect. The difference can be seen during crack growth and the notch influence, then it disappears. From Figs.3 and 4 it appears that the notch strongly influences variation of the ∆J parameter both globally and locally. In the elastic-plastic range, the stresses and strains were calculated bymeans of the finite element software FRANC2D. The cyclic stress-strain curve ba- sed on the nonlinear material model was introduced into FRANC2D. In that case, the cyclic stress-strain curve for FeP04 steel described by the Ramberg- Osgood relation was used. It was decided that calculations would be based on incremental elastic-plastic analysis including the kinematic model of material hardening. The calculations were performed for two-dimensional geometrical models of notched specimens. The geometrical model of the specimen was si- 758 D. Rozumek mulated in the CASCA software. The finite element mesh was automatically generated and it containedmore than 3900 six-nodal triangular elements. The crack growth direction was assumed on the basis of observation of the current crack path obtained from the tests. Such simulation of the crack propaga- tion allowed one to obtain the strain and stress maps for each realised crack increment. 4. Conclusions The presented results of the fatigue crack growth in the plane notched spe- cimens made of FeP04 steel and subjected to tension loading allow one to formulate the following conclusions: • Equation (3.2) can be applied only for short fatigue cracks growth be- cause local stress concentrations cause high stresses and accompanying strains. • Equation (3.3) well describes the results including an additional influ- ence of the notch during fatigue crack growth and it can be applied for determination of length and quality of the notch influence. • It has been shown that the notch strongly influences variation of the ∆J parameter, both globally and locally. References 1. BertoF., LazzarinP., 2007,RelationshipsbetweenJ-integral and the strain energy evaluated in a finite volume surrounding the tip of sharp and blunt V - notches, Int. J. of Solids and Structures, 44, 4621-4645 2. El Haddad M.H., Topper T.H., Smith K.N., 1979, Prediction of non- propagating cracks,Engineering Fracture Mechanics, 11, 573-584 3. Lazzarin P., Tovo R., Meneghetti G., 1997, Fatigue crack initiation and propagation phases near notches in metals with low notch sensitivity, Interna- tional Journal of Fatigue, 19, 647-657 4. Rozumek D., 2004, The ∆J-integral range applied for the description of fa- tigue crack growth rate, Proc. of the 12th International Conference on Expe- rimental Mechanics (ICEM12), Bari, Italy, Edit. C. Pappalettere, 275-276 and CD, ps 8 Influence of the notch radius... 759 5. Rozumek D., 2008, Influence of the notch radius on the ∆J-integral range under cyclic tension,Proc. of the Sixth Int. Conference on Low Cyclic Fatigue (LCF6), Berlin, Germany, Edit. P.D. Portella et al., DVM, 505-510 6. Rozumek D.,MachaE., Lazzarin P.,Meneghetti G., 2006, Influence of the notch (tip) radius on fatigue crack growth rate, Journal of Theoretical and Applied Mechanics, 44, 127-137 7. Rozumek D., Pawliczek R., 2004, A description of crack growth and ma- terials fatigue in an energy approach, Studies and Monographs, 165, Opole University of Technology, p.165 [in Polish] 8. TanakaK., 1983,The cyclic J-integral as a criterion for fatigue crack growth, Int. J. Fracture, 22, 91-104 9. Taylor D., 1999, Geometrical effects in fatigue: a unifying theoretical model, Int. Journal of Fatigue, 21, 413-420 10. Taylor D., Hughes M., Allen D., 1996, Notch fatigue behaviour in cast irons explained using a fracture mechanics approach, Int. Journal of Fatigue, 18, 439-445 11. Thum A., Petersen C., Swenson O., 1960, Verformung, Spannung und Kerbwirkung,VDI, Düsseldorf 12. Williams M.L., 1952, Stress singularities resulting from various boundary conditions in angular corners of plates in extension, J. Appl. Mechanics, 19, 526-528 Wpływ promienia karbu na zmiany parametru ∆J przy prędkości wzrostu pęknięć zmęczeniowych Streszczenie W pracy przedstawiono wpływ promienia karbu na zmiany parametru ∆J pod- czas badań wykonywanych w zakresie niskiej i wysokiej liczby cykli zmęczeniowych. Badaniompoddanopróbki płaskiewykonane ze stali FeP04-UNI8092.Próbki charak- teryzowały się dwustronnymi symetrycznymi zewnętrznymi karbami z promieniami dna karbu w zakresie od 0,2 do 10mm. Wyniki badań prezentowane w pracy wy- konywano na maszynie hydraulicznej MTS 809 w Department of Management and EngineeringwVicenzy (UniversytetwPadwie).Wszystkie badanie zmęczeniowebyły wykonywane przy obciążeniu z kontrolowaną siłą przy stałej wartości współczynnika asymetrii cyklu R=0 i amplitudzie obciążenia Pa =6 i 7kN (która korespondowała z nominalną amplitudą naprężenia normalnego σa = 100, 117MPa przed inicjacją pęknięcia). Badania wykonywanow zakresie częstotliwości obciążenia 13 i 15Hz. Manuscript received October 29, 2008; accepted for print February 11, 2009