|ECF® FRACTURE BEHAVIOUR AND DESIGN OF MATERIALS ANO STRUCTURES LOAD CONTROL VS. DISPLACEMENT CONTROL — THBIR INFLUENCE ON THE J-RESISTANCE CURVE ©. Kolednikt ‘Two Jic-multi-specimens are made, where the specimens are loaded under Joad control (LC) and under displacement con- trol (DC), to investigate whether there exits an influence of the loading conditions on the shape of the J-resistance curves. ‘The material tested was a low-strength structural steel. Both eats show the same crack-initiaton toughness, Jy, but for the LC-specimens the slope of the J-Aa-curve is larger. This uean that the erack-groxtth toughness of the L.C-specimens js larger than of the DC-specimens INTRODUCTION “To describe the resistance of materials against initiation and growth of aus Fede cracks in the regime of elastic-plastic fracture mechanics J-resistance sae Umeures) are recorded, Here the J-integral, Jy is plotted as a loading parameter over the crack extension Aa. Weis generally accepted that up to the point of initiation of crack growths J,,the carves are independent ofthe specimen geometry or loading cond ons eee certain specimen dimensions are not t00 small, So J is regarded as a ‘material property. Tr the region of stable erack growth the Jr-curve is influenced by speciinen contr, Within this region there probably exists Tower bound of he sone cesthe carve which fact is useful for design purposes. ic onset of unstable crack growth is known to depend on both specimen geometry and loading conditions i, whether the specimen i loaded under Toad or displacement control ‘ioe te question arises whether the Jjecurvesareinfluenced by the loading oe Ech Sci oni Tur Feibeperphysk, Ostercichiche Akademi der Wissensca ten, A-ST00 Leoben, AustinECF 8 FRACTURE BEHAVIOUR AND DESIGN OF MATERIALS AND STRUCTURES: ions in the region of stable crack growth, too. This is the subject of this A PRECEDING PAPER In ASTM STP 591 S. R. Novak [1] reports of a large effect of the loading, conditions on the shape of stress-intensity-resistance curves (Ka-curves) for a A572 structural steel (yield strength oys = 345 MPa, ultimate tensile strength. ors = 570 MPa). A double clip-gauge method was used to calculate the val- tues of the crack-tip-opening displacement, COD, which were converted into K-values, For the load-controlled tests Compact Specimens were used, the displacement-controlled tests were made on Crack-Line-Wedge-Loading Spec: imens, see ASTM E561 [2]. The specimen width was W = 200 mm, the thick ness B = 38 mm. Fig. 1 shows some results of this work. The Kp-curves of the load-controlled tests ie above the displacement controlled tests. In all tests considerable stable crack growth was observed before “catastrophic” (cleavage?) fracture occurred (designated with a cross in Fig. 1). It is not clear why even the initiation tough= ness is influenced by the loading condition and why the displacement-controlled, tests failed before the load-controlled tests. ‘The results of [1] did not get further attention,may be because the load- and displacement-controlled tests were not performed at the same machine ‘This is the reason why a similar test has been made in the current work. OWN EXPERIMENT ‘The material investigated was a structural steel $t50. ‘The chemical compo: sition in % is 0.34 C, 0.54 Mn, 0,83 Si, 0.019 P and 0.021 S. The tensile test resulted oy = 260 MPa and ors = 520 MPa. Compact Specimens were machined according to ASTM ES13 [3] with a thickness of B= 25 mm and a width of W=50 mm. Two multispecimen JJrc-tests were made on a Schenck servohydraulic machine. The tests of Series 1, (Specimens 1 to 5) were conducted under clip-gauge-displacement control, ‘The loading rate was = 0,002 mm/s. Series 2 (Specimens 6 to 11) was tested under load control. Here the loading rate was decreased successively from 300 N/s down to 5 N/s to get about the same # as for the displacement controlled test. Each specimen was fatigued and immediately tested to exclude 1 possible influence of strain-aging. Specimen 11 was loaded beyond the max: innum load, though the test was performed under load control. ‘The specimen was unloaded partially at the maximum load and reloaded until the F-ine became horizontal. Then it was unloaded again, and so on. For the estimate of J the unloading points were connected to determine the area wnder the F-v-curve.EGE @ FRACTURE BEHAVIOUR AND DESIGN OF MATERIALS AND STRUCTURES TABLE 1; Data of the Tests Performed Under Displacement Control (Spee: T to By and Under Load Displacement Control (Spec. 6 to 11) [Spee [elmo | FIN | Faas] [efrom] | Qotmm] | JTS 1] 2558 [33700 35700 | oor | o.aai [ 8 2 | 2390 | 41740 | 41740 | 0.99 | 0.645 | 115 3. || 2491 | 30220] so220 | 138 | 114 | 160 ‘| 2374 | 41900 | 43600 | 1.60 | 1.90 | 200 5 | 2614 | s9810| ao600 | 1.71 | 242 | 210 Sy 26.59 | a1s40] s1340 | 0.94 | 0.202 [89 7 | 24a | 39400] 39900 | 1.06 | 0.540 | 116 a | are 37610 | 1.12 | 0.509 1 9 | 23.56 s47z0 | 140 | 116 | 168 10 | 23.99 agz00 | 7 | 169 ) 217 n_| 23.51 45120 | 1.96 | 2.93 | 268 In Table 1 the experimental results are summarized, in Fig. 2 the two J ~ ‘Ac-curvesare presented. The two curves show the same crack-initiation tough hes, J, = 66 ki/m, but during ductile tearing the slope of the displacement Controlled test is always lower than of the load-controlled test. So the difference Sed for a given Aa increases with rack extension. Near the maximum load (between points 10 and 4) the relative difference of J is about 14%. DISCUSSION itshoutd be note thatthe et fll the Hutchinson and Pass] coudion aan ee a eh nich uses the we of J fr eck gromth ba Tis bis te tigement length, b= W — @ aan a ee jp Recommendatons [eve the J-vality with eu a) SE we im Ine = (aor BYSE 2 3904/3 @ and Sages = 0.06by 1.5mm. @ ere by is the intial ligament length and the flow stress 07 = Movs + 273). Wthie the fst 1.5 mum of erack extension all three conditions are satisfied thas been worked out by Kolednik [4] that the crack-initiation toughness, “a and the erack-growth toughness are different. J is kind of specific energy 1049ECF 8 FRACTURE BEHAVIOUR AND DESIGN OF MATERIALS AND STRUCTURES to produce the critical COD at the original crack tip. ‘The erack-growth tough- ness can be derived from the Grifith’ energy balance during erack-growth. It is connected to the slope of the Jm-curves. This can be also deduced from John and ‘Turner (5) So we ean conclude that the erack-growth toughness of our material does depend on the loading conditions. Ata given crack extension it is large for the Joad-controlled test than for the displacement-controlled test, For the linear parts of the two Ja-curves (within Aa 1 mm) we have # load contr 106 MJ/m? and { displ.control = 89 Mi/m?. So the difference in the crack- growth toughness of the two tests is 19%. REFERENCES: [i] Novak, S. R., “Resistance to Plane Stress Fracture (R-Curve-Behavior) of A872 Structural Steel", ASTM STP 591, 1975 [2] ASTM B591-86, “Standard Practice for R-Curve-Determination”, Annual Book of ASTM Standards, Vol.0301, 1987. [3] ASTM £591-87, “Standard Test Method for Jig, a Measure of Fracture Toughness” , Annual Book of ASTM Standards, Vol.03.01, 1987. [4] Hutchinson, J. W., and Paris, P. C., in: “Blastic-Plastic-Fracture” ASTM STP 668, 1979. [5] EGF-Recommendations for Determining the Fracture Resistance of Ductile Materials, EGF-P1-87D, 1987. [6] Kolednik, ©., “On the Physical Meaning of the J — Aa Curves”, to be published, {| John, S. J., and Turner, C. E., “Alternative Representations for Scaled R-Curves in a Titanium Alloy”, in:Proccedings of the EEPFM-Symposium Freiburg, Mechanical Engineering Publications, to be published. 1050[ECF 8 FRACTURE BEHAVIOUR AND DESIGN OF MATERIALS ANO STRUCTURES weal 400 Toad controtled, T=22°C displacement controlled, 1=22 load controlled, T=4°C displacement controlled, T= 100 7236 5 6 7 B 9 1 TN aalmm Figure 1: Jpccurves of « structural steel AST2 tested under load control and under displacement control, from (1] st 0 150 +load-controlled test FD + displacement-contolled test Sb 2 25 aoimm Figure 2: Jprcurves of a structural steel St50 tested under lond control and ‘under displacement control 1051