Resistance of Structural Steel to Crack Formation and Propagation Odpornost gradbenih jekel proti nastanku in širjenju razpoke P.D. Odesskij, Centralnij naučnoisledovatelnij institut strojitelnih konstrukcij, im. V.A: Kučerenko, Moskva N. Kudajbergenov, Kazahskoj Himiko—tehnoiogičeskij institut, Čimket, Kazahstan L. Kosec, FNT, Odsek za metalurgijo in materiale, Ljubljana F. Kržič, FAGG, Jamova 4, Ljubljana Parameters of linear fracture mechanics can be a useful measure for selection of steel with various strength and yield stress (in limits 200 to 1000 MPa). They determine also influence of purity (non-metallic inclusions) and of thermal or thermomechanical treatment. These parameters are effective only if they are measured in the conditions when the plastic deformation at the initial crack growth is limited to minimal value. This happens in corrosion media by measuring Kjc and especially in impact loading by measuring I\'fc. These parameters are closely connected with the microstructure and structure of steel. They are suitable for designing structures resistant to brittle fracture if operational (destruction) conditions of those structures are seized, since they occur at high preceeding plastic deformations. V članku razpravljamo o načinih zboljšanja efektivnih parametrov linearne mehanike loma, ki so merilo kvalitete gradbenih jekel in osnova za izračun konstrukcij odpornih proti krhkemu prelomu. Predmet raziskave je bila skupina maloogljičnih in malolegiranih jekel z napetostjo tečenja 200-1000 MPa. Po kemični sestavi spadajo ta jekla v štiri skupine: maloogljična, mangan silicijeva, manganova mikrolegirana jekla in kompleksno legirana jekla z 0.2-0.6% Mo. Po trdnosti lahko omenjena jekla razdelimo v tri skupine: v skupino z napetostjo tečenja do 290 MPa (normalna trdnost); v skupino z napetostjo tečenja do 390 MPa (povišana trdnost) in v skupino jekel z napetostjo tečenja več kot 390 MPa (visoka trdnost). Lomne karakteristike smo raziskovali pri statičnih in dinamičnih obremenitvah ter v korozijskem mediju. Raziskave smo opravili v skladu z GOST in mednarodnimi standardi, pa tudi po originalni metodiki. Največ smo uporabljali epruvete z ekscentrično obremenitvijo (CTS) (si. 1), dvojno konzolno vpeto klinasto epruveto (si. 2) in cilindrične preizkušance s koncentrično krožno zarezo z utrujenostno razpoko kot koncentratorjem napetosti (si. 3). Pri statičnih obremenitvah smo ugotovili pomembne posebnosti v rasti vrednosti I\1C s trdnostjo jekla takrat, ko je imelo valjano jeklo "racionalno" mikrostrukturo. Ugotovili smo tudi, da raste vrednost I < 45" .20 Figure 2. Contoured double-cantilever double axially notched probe. Slika 2. Dvojno konzolno vpela klinasta epruveta z dvema osnima zarezama. UD 0Dk Figure 3. Cylindrical probe vvith concentric fatigue crack—(type 2 according to GOST 25.506-85). L = length betvveen the clamped parts of probe in llie tensile testing machine. L = 5D; d = (0.6 to 0.7)D: Li > 7D; /0 = 0.5(D - d) > h + 1.5 mm; /o > 3.71 g a; DK = D - 2h = (0.65 io 0.85)D. Slika 3. Cilindrična epruveta s koncentrično utrujenostno razpoko—(lip 2 po GOST 25.506-85). L = razdalja med deloma epruvete, ki se vpneta v trgalni stroj L = 5D\ d = (0.6 do 0.7)D; L\ > 7D: /0 = 0.5(D - d) > h + 1.5 mm; (0 > 3.7 tg a\ DK = D - 2h - (0.65 do 0.85)D. 200 L| a CL -120 "80 T (°C) -40 Figure 5. Fracture toughness of structural steel plate. Dependance of 1\ ic on temperature, plate thickness 20 mm. Probe from Fig. 2. 1) Hardened and tempered molybdenum-alloyed steel: 0.12% C, 0.54% Si. 1.05% Mn, 0.5% Cr, 1.47% Ni, 0.12% V, 0.24% Mo, 0.011% Al, 0.022% N, 0.025% S 2) The same steel, plate thickness 40 mm, Rp = 710 MPa 3) Hardened and tempered manganese-silicon steel (0.1% C, 1.48% Mn, 0.9% Si, 0.031% S, 0.021% P); Rp = 435 MPa 4) Steel above, rolled; Rp = 350 MPa 5) Hot rolled low-carbon steel (0.16% C, 0.24% Si, 0.65% Mn, 0.025% S. 0.025% P); Rp = 265 MPa. Slika 5. Lomna žilavost pločevine iz jekla za gradbene konstrukcije. Odvisnost /\ /c od temperature, debelina pločevine 20 mm. Fpruveta slika 2. 1) Poboljšano legirano jeklo z molibdenom: 0.12% C, 0.54% Si, 1.05% Mn, 0.5% Cr, 1.47% Ni, 0.12% V, 0.24% Mo, 0.011% Al, 0.022% N, 0.025% S 2) Isto jeklo, debelina pločevine 40 mm, Rp = 710 MPa 3) Poboljšano mangan-silicijevo jeklo (0.1% C, 1.48% Mn. 0.9% Si, 0.031% S, 0.021% P); Rp = 435 MPa 4) Jeklo (3) valjano; Rp = 350 MPa 5) Vročevaljano maloogljično jeklo (0.16% C, 0.24% Si. 0.65% Mn, 0.025% S, 0.025% P); 265 MPa. Figure 4. Scheme of crack tip vvith the zone of plastic deformation. Slika 4. Shema razpoke s cono plastične defomiacije. GOST 9454-77 in distilled vvater and in 3% NaCl vvater solution according to RM SEV niethod (Corrosion protec-tion in building engineering. Corrosion cracking of high-strength armature steel. Investigation methods, 1986). Moving rate of the tensile-tester clamping javvs vvas 2 • 10"8 mm/min vvhich vvas sufficient for completing test in one day. Also impact toughness vvith U and V-notched probes (according to GOST 9454-78) vvas measured simul-taneously vvith the uniaxially loaded tensile tests according to GOST 1797-84 vvith (lat probes of the same thickness as investigated plate. 4 Results of Tests Static testing gave a series of relations valid for the crack stability in structural steel (Figs. 5 to 8). It shovvs that in steel of standard purity and rational microstructure the fracture toughness value increases vvith the increased strength and vvith the transition from ferrite-pearlite microstructure POVRŠINA CONA PLASTIČNE DEFORMACIJE CONA PLASTIČNE DEFORMACIJE VRH RAZPOKE VRH RAZPOKE If -196 -122 -80 T (°C) Figure 6. Fracture toughness and the size of plastic deformation zone in normalized steel (0.17% C, 1.56% Mn, 0.4%, Si, 0.11% V. 0.015%. N, 0.008%- S. 0.07% P). Sulphide inclusions are modified by addition of RF., curve (1) vacuum treated steel; R,, = 460 MPa. Slika 6. Lomna žilavost in velikost cone plastične defomiacije v normaliziranem jeklu (0.17% C, 1.56% Mn, 0.4% Si, 0.11% V, 0.015%. N, 0.008% S. 0.07% P). Sullidni vključki so modificirani z dodatkom RZ, (1) jeklo vakuumirano; Rp = 460 MPa. 5 600 650 680 T (°C) Figure 7. Dependance of fracture toughness and the size of plastic deformation zone on tempering temperature for molybdenum-alloyed steel (0.10% C, 0.37% Si, 1.16% Mn, 3.1% Cr, 1.0% Ni. 0.34% Mo. 0.016% S. and 0.04% P), plate thickness 20 mm. probe type from Fig. 3. Slika 7. Odvisnost lomne žilavosti in velikosti cone plastične deformacije od temperature popuščanja za jeklo legirano z molibdenom (0.10% C. 0.37% Si, 1.16% Mn, 3.1% Cr. 1.0% Ni. 0.34% Mo, 0.016% S in 0.04% P) debelina pločevine 20 mm. epruveta si. 3. to microstructures obtained by hardening and tempering (Fig. 5). Plate thickness reduces the A'/e value (Fig. 5) due to the reduced plastic deformation zone (?'r). The investiga-tion results also indicate that fracture toughness of high-strength steel depends on type, amount and distribution of non-metallic inclusions, mainly sulphides (Fig. 6). Pure steel (0.008% S) excells the steel vvith standard amount of sulphur both in the respect of fracture toughness and in size of plastic deformation zone at crack tip ()••/•)• This confirms the influence of inclusions, on vvhich decohesion takes plače (formation of voids), on the conditions of crack initiation *. In heat-treated steels the fracture toughness A'/c is abruptly reduced if tempering temperature is reduced from 650 to 600°C (Fig. 6). The reason is in changed mech-anism vvhich controls the crack initiation. The tvvo-stage process connected to formation of microvoids is substituted by an energy undemanding mechanism of local destruetion vvhich is detailed deseribed elsevvhere4. The reduced tempering temperature reduces the I\ j <• value measured by static loading (Figs. 6 and 7) vvhich is in contradiction vvith the hitherto ideas, especially vvith the changed size of plastic deformation zone rj. This can be explained by applied testing methods vvhich did not allovv a suitably high microstructural sensitivity of parameters deseribing the crack stability in ductile steel vvith rational microstructure. This is confirmed vvith tvvo ad-ditional cases of unsufiicient microsh-uctural sensitivity in estimating A'/c value vvith static tests. Table 2 presents the relation betvveen the fracture toughness of manganese-silicon steel and the chemical contposition and the hardening temperature. Steel santples vvith various chemical com-positions vvere hardened at optimal temperature of 930°C, and at 1050°C. This temperature vvas chosen in order to determine the influence of overheating. In ali the cases the steel samples vvere tempered at 650° C. Investigation results indicated that concentrations of al-loying elements had a small influence on mechanical properties. Overheating of steel highly deteriorates the Charpy-test values, transition the impact toughness value vvhich vvas practically halved. K k; value is not extra highly influenced by overheating; the obtained differencies vvere belovv 5% and they are in the region of measuring errors. The second čase is connected to the seleetion of thermo-mcchanical treatment in manufacturing 50 mm plate vvith yield stress Rp > 450 MPa made of microalloyed man-ganese steel (Table 3). Steel vvas quenched from the rolling temperature and tempered at 650°C. Initial and final rolling temperatures, and the quenching temperature (i.e. interval betvveen ftnished rolling and quenching in vvater) vvere var-ied. Specifications TMT 1, 2, 3, and 4 in the mentioned table represent various regintes of thermomechanical treatment. 1\ jc values vvere measured vvith CTS probes (Fig. 1) having plate thickness. The highest temperature at vvhich the A'/c value vvas correctly measured vvas —40° C. Table gives the dissipation of results of three tests. Table also suggests the seleetion of optimal regime of treatment vvhich enables the yield stress above 490 MPa at simultaneously the highest toughness transition temperature, i.e. TMT-1. Simultaneously it is evident that fracture toughness values (A'/c,-40 and A"/c,-7o) do not enable to judge vvhich thermomechanical treatment is optimal. These measurements only reliably indicate that TMT-4 treatment (hot rolling) is the most unusitable one. The mentioned cases shovv that parameters of linear fracture mechanics are microstructural^ not enough sensitive to enable the regarding seleetion of tested steels. The most probable reason is the high ductility of structural steel; in this čase the ductility of high-strength steel vvhich have oversized plasticity zone at crack tip in static P.D. Odesskij. N. Kudajbergenov, L. Kosec. F. Kržič: Resistance of Structural Steel to Crack Formation and Propagation Table 2. Some parameters of bnttle-fracture resistance of manganese-silicon steels with yield stress Rp > 390 MPa, plate thickness 20 mm Chemical composition of steel % Quenching temperature (°C) Mechanical properties Rv (MPa) KCV~ ,u (J/cm2) (°C) T ' — ' UXJ" JXIC (MPa m1/2) C Mn Si S P 0.09 1.42 0.28 0.030 0.022 930 429 64 -20 165 0.09 1.42 0.28 0.030 0.022 1050 432 29 +20 160 0.09 1.30 0.60 0.022 0.019 930 417 84 180 0.09 1.30 0.60 0.022 0.019 1050 414 41 +20 175 0.09 1.42 1.03 0.030 0.023 930 435 69 -20 175 0.09 1.42 1.03 0.030 0.023 1050 445 41 +20 170 j—Toughness transition temperature determination was based on 50% tough fracture n—Probe in Fig. 2. —70°C was the higliest temperature on which A'/c could be estimated Table 3. Mechanical properties of 50 mm thick plate of microalloyed manganese steel (0.19% C. 1.58% Mn. 0.48% Si, 0.07% V, 0.024%. S) Steel treatment Rv (MPa) XXX (°C) 7^50 <°C) K C V"70 (J/cm2) A-4U ic A'"™ I c (MPa m1/'-') Hot rolled 420 -70 -20 49 55-76 42-55 Hardened and tempered 500 -60 +20 31 61-82 52-67 TMT 1 520 -100 -40 92 67-103 52-64 TMT 2 570 -60 -20 38 52-91 42-64 TMT 3 620 0 32 64-106 39-67 TMT 4 635 -20 +20 28 45-64 27^8 in—Determined by impact toughness criterion .39 J loaded probes though ali the testing conditions described by corresponding standards were fulfilled. Microstructural depcndance of investigated parameters becomes more pronounced in more severe testing conditions vvhich reduce the extent of plastic deformation and the size of plastic deformation zone at the crack tip. This occurs during testing in corrosive media (Fig. 7). In these tests the K\c value increases vvith the increased tempering temperature of steel. The obtained result is the consecjuence of a high density of disordered dislocation loops. Such a structure is essentially less resistant to stress corrosion2, especially if hydrogen embrittlement is developed. In steel vvith such a structure, the KCjc value is essentially lovver than the A'/c one (Fig. 8). Behaviour of hardened and tempered steel during load-ing is characterized by its substructure vvhich is formed in recovery of ferrite and in the beginning of recrystallization. Materials vvith such a structure are not sensitive to stress corrosion; therefore there it is valid: A'/c — ^/c (Fig. 8). Another way to increase the structural sensitivity of A'/c 316 the impact tests vvhere loading rates are increased for six orders of magnitude. In this čase the extent of plastic deformation is reduced to suitable amount, also size of plastic deformation zone at the crack tip is abruptly reduced, and it becomes dependant on steel microstructure. This finding can be confirmed vvith the investigations of steel of big pipelines vvhich measured A"/c values vvere close to the K ic values determined by static testing (Table 4). In static loading the CTS probe vvas applied vvhile for impact tests Charpy test probe vvith fatigue crack vvas used. Steels vvith equal A"/c values exhibit the same sequence £ <2 Figure 8. Dependance of fracture toughness and the size of plastic defonmation zone on tempering temperature for microalloyed manganese steel (0.14% C, 1.64% Mn, 0.52% Si, 0.07% V, 0.007% N, 0.031% S, 0.013% P). KcIC—fracture toughness for tests in corrosive medium. Slika 8. Odvisnost lomne žilavosti in velikosti cone plastične deformacije od temperature popuščanja za mikrolegirano manganovo jeklo. (0.14% C, 1.64% Mn, 0.52% Si, 0.07% V, 0.007% N, 0.031% S. 0.013% P). K'fj—lomna žilavost pri preizkusih v korozijskem mediju. Table 4. Fracture toughness and the size of plastic deformation zone on crack tip Steel treatment Plate thickness Chemical composition Grain size RP /Oc rT Kfc rd t (mm) C S (pm) MPa MPa1/2 mm MPa1/2 mm Microalloyed Mn steel, hardened vvith A1N, 12 0.17 0.012 9 418 80 4.2 13.7 0.15 normal ized Microalloyed Mn steel, hardened vvith VN, 12 0.13 0.011 6 425 80 4.1 22.6 0.14 normalized Mn steel alloyed vvith small • amounts of Mo and Nb, 16 0.13 0.011 4 590 84 3.1 37 0.29 controlled rolled after impact testing. Investigations also showed the advan-tage of controlled rolled plates. Microstructural sensitivity of fracture parameters vvas also essentially increased, espe-cially the Kjc and values vvhich are in a good correla-tion vvith the grain size. 5 Analysis of Results The described testing results shovv that the A'/c value mea-sured at static loading is not sufficiently microstructurally sensitive property (Figs. 6 and 7, Tables 2 and 3). The needed sensitivity can be achieved by increased test sever-ity or vvith more demanding tests applying corrosive media or impact loads (Fig. 7, Table 4). Microstructural sensitiv-ity vvas increased if the extent of plastic deformation vvas reduced. As a rule, at least three of the follovving con-ditions must fulftlled in that čase: temperature belovv 0°C, dynamic tensile load, stress raisers must be present in struc-ture, dimensional factor (great cross section, and the like), and unsuitable steel microstructure. In these cases the parameters of crack stability measured in the conditions of highly limited plastic deformation give good description of fracture conditions. These parameters vvere vvell applied in engineering design of structures. As an example, the crack propagation in the vvall of main pipeline will be described. Crack vvas initiated in the vveld on the line of fusion penetration inside the pipeline, and then it propagated tovvards the extemal surface. The I\'/c values and the critical crack lengths lr in the heat affected zones for various steel are revievved in Table 5. Critical size of defect vvas calculated by expression3: vvhere p is pressure, R pipe radius, and t vvall thickness. The obtained results shovv that crack in manganese-silicon steel becomes unstable at low temperatures, and it starts spontaneously to propagate in the axial direction at relatively small penetration into the pipe vvall. At those temperatures the use of pipes made of the mentioned steel is not allovved. After controlled rolling the critical size of crack becomes greater than the vvall thickness. Thus the stable crack vvhich reaches the pipe surface hinders the sponta-neous propagation of crack. At the given temperature the pipe made of the third steel (cited in Table 5) is safe. Table 5. Fracture characteristics of steel in heat affected zone of pipeline vveld Steel treatment Plate thickness ,.<*<-(30) IC ,(-60) ' C (mm) (MPa m1/2) (mm) Mn-Si steel, normalized 12 24 3 Microalloyed (V) Mn steel, controlled rolled 17 81 24 Mn steel, microalloyed vvith V and Nb, controlled rolled 14 97 32 Fatigue crack vvas normal to the plate surface Determination the fracture toughness at static load at -60°C (A/c > 100 MPanr /2) for manganese-silicon steel indicates the safety of that steel, but unfortunately the prac-tical experiences did not confirm it. Thus the parameters of linear fracture mechanics measured in the conditions of very limited plastic deformation did not exhibit only high structural sensitivity but they are also useful in engineering design of brittle-fracture resistance in the cases vvhen they enough accurately describe the mechanism and conditions for brittle fracture of certain steel. 6 References 1 Werkstoffkunde Eisen und Stahl; Teil 1; Grundlagen der Festigkeit. der Zahigkeit und der Bruchs; Verlag Stahleisen, Diisseldorf, 1983 2 Tylkin M.A.. Bolšakov V.I., Odesskij P.D.: Struktura i svojstva stroiteFnoj stali; Moskva, Metallurgija. 19X3 3 Knott J.: Mikromehanizmy razrušenija i treščinostojkost' konstrukcionnyh splavov; Mehanika razrušenija. (prevod iz angl.), Moskva, Mir. 1979. (101-130) 4 Smith E., et al.: Lokalizacija plastičeskogo tečenija 1 treščinostojkost vysokopročnyh materialov. Mehanika razrušenija, (prevod iz angl.); Moskva. Mir. 1980. (124— 147) 5 Duffy A.R., et al.: Praktičeskie primery rasčeta na sopro-tivlenie hrupkomu razrušeniju truboprovodov pod davle-niem, Razrušenie, 5; Moskva, Masinostroenie. 1977. (146-209)