FRACTURE MECHANICAL VALUESOF MODERN PIPELINE-STEELS

S. Felber, Institute for Building Materials, Building Physics, and Fire Protection (E 206),Vienna University of Technology, Karlsplatz 13, A-1040 Vienna, Austria,

Tel ++43 1 58801 43085, Fax ++43 1 58801 43098,email sfelber@mail.zserv.tuwien.ac.at

ABSTRACT

This paper deals with the determination of CTOD- (Crack Tip Opening Displacement-) valuesaccording to BS 7448 using three point bend specimens and CA (Crack-Arrest Fracture-Toughness) values using three-point bend specimens (test proposal of TVFA), compact-crack-arrest specimens (ASTM E 1221), and full-thickness compact-crack-arrest specimens (testproposal of Crosley and Ripling). The problems in determining the crack tip opening displacementand crack-arrest fracture-toughness are discussed in detail, the results are compared. Finally thedifferent resulting fracture mechanical values, as for example CTOD (Crack Tip OpeningDisplacement) values and CA (Crack-Arrest) values etc., of the welds (welding metals and heataffected zones) are compared with each other and with the values of the base material.

The tested materials were the base material, the weld metal, and the heat affected zone of welds,using different welding processes, as Shielded Metal Arc Welding, Gas Metal Arc Welding, orSubmerged Arc Welding, of the pipeline steels X 70 according to API 5L (StE 480.7 TM accordingto DIN 17 172 or L 485MB according to OENORM EN 10 208 – 2) and X80 according to API 5L (L555MB according to OENORM EN 10 208 – 2) and the duplex-steel 1.4462 according to DIN 17440, DIN 17 441, SEW 400, OENORM EN 10 088, part 1 to 3, and TÜV-Austria (1.4462 accordingto OENORM EN 10 027 – 2, X 2 CrNiMoN 22 5 3 according to OENORM EN 10 027 – 1 or S31803according to ASTM UNS).

BENEFITS

Overview about the variations of the fracture mechanical values (CTOD and CA) of the pipeline-steels X70, X80, and the duplex-steel 1.4462 according to different standards or test proposals.Comparison of the values for the base material, the weld metal, and the heat affected zone ofwelds according to different welding processes.

KEY WORDS

Fracture mechanical values, CTOD (Crack Tip Opening Displacement), CA (Crack-Arrest FractureToughness), BS 7448, ASTM E 1221, three-point bend specimens, compact-crack-arrestspecimens, full-thickness compact-crack-arrest specimens, pipeline-steels, X70, X80, duplex-steel, 1.4462, base material, weld metal, heat affected zone, different welding processes.

1 INTRODUCTIONThe resistance of a material against fracture is influenced by its behaviour during the three possiblesteps of the fracture process, which are crack initiation, crack extension, and crack-arrest. If thereis a possibility to avoid the first step of a fracture process, the crack initiation, the other two phaseswill not occur. It is the task of materials testing, especially of fracture mechanics, to find out therelevant parameters.To avoid the initiation of a crack in a pipeline it is of great interest to know the CTOD (Crack TipOpening Displacement) values. This paper deals with the determination of these CTOD valuesaccording to BS 7448 [1] of the pipeline steels X 70, X80, and the duplex-steel 1.4462 (basematerial, weld metal, and heat affected zone).There are two main possibilities to stop a crack in a pipeline: The first one is to apply special crackarrestors, and the second one is to use a material with a high crack-arrest toughness. The firstpossibility is rarely realized, and so it is of great interest to gain knowledge of the fracture toughnessvalues of the pipeline-steels employed in order to limit the crack length. In the past, big specimens,like Robertson plates for example, were used for the determination of the crack-arrest fracture-toughness, KIa, but they are very expensive [2], [3], and so nowadays they are replaced by smallspecimens, like three-point bend specimens [4], [5], [6], compact-crack-arrest specimens [7], andfull-thickness compact-crack-arrest specimens [8].This paper deals with the determination of the CTOD (Crack Tip Opening Displacement) valuesand the CA (Crack-Arrest Fracture-Toughness) values, KIa, of the base material, the weld metal,and the heat affected zone of welds of the pipeline-steels X 70, X80, and the duplex-steel 1.4462.For this purpose tests were performed with each of the three materials, the base material, the weldmetal, and the heat affected zone, for each steel, X70, X80, and 1.4462, employing each of thethree mentioned test specimens. Finally, this paper contains a comment on the differencesbetween the crack-arrest toughness values of the three materials and a comparison of the threeapplied test methods.

2 TESTED MATERIALSThe steels used for the tests were X70 according to API 5L [9] (StE 480.7 TM according to DIN 17172 [10] or L 485MB according to OENORM EN 10 208 – 2 [11]), X80 according to API 5L [9] (L555MB according to OENORM EN 10 208 – 2 [11]), which are nowadays mainly used for onshorepipelines, and 1.4462 according to DIN 17 440 [12], DIN 17 441 [13], SEW 400 [14], OENORM EN10 088, part 1 to 3 [15] to [17] , and TÜV-Austria [18] (1.4462 according to OENORM EN 10 027 – 2[19], X 2 CrNiMoN 22 5 3 according to OENORM EN 10 027 – 1 [20] or S31803 according to ASTMUNS). Figure 1 shows the chemical composition (mass-%) and Figure 2 the mechanical propertiesaccording to the before mentioned standards for the pipeline-steels X70, X80, and the duplex-steel1.4462.

Figure 1. Chemical composition of the pipeline-steels X70, X80, and the duplex-steel 1.4462(L ... longitudinal welded pipe, S ... spiral welded pipe, B ... company B, N ... company N,

W ... rolled, Pl ... plate, P ... Pipe)

rolled (W)casted (Pl)casted (P)

mas

s-%

mas

s-%

Cr,

Ni,

Mo

in th

e st

eel 1

.446

2

Figure 2. Mechanical properties of the pipeline-steels X70, X80, and the duplex-steel 1.4462(L ... longitudinal welded pipe, S ... spiral welded pipe, B ... company B, N ... company N,

W ... rolled, Pl ... plate, P ... Pipe)

The microscopic structures of the tested steels are shown in the Figures 3 and 4. For furtherinformations see [21].

Pipeline-SteelsX70 (L) X70 (S) X80 (B) X80 (N)Microscopic picture (etched with 2% HNO3, 1st line (100:1), 2nd line (200:1))

Figure 3. Structure of the tested pipeline-steels X70 and X80 (L ... longitudinal welded pipe,S ... spiral welded pipe, B ... company B, N ... company N) [21]

Duplex-SteelsRolled (W) Casted (Pl) Casted (P)

Microscopic picture (etched with LBI, 1st line (100:1), 2nd line (200:1))

Figure 4. Structure of the tested duplex-steels 1.4462 (W ... rolled, Pl ... plate, P ... pipe) [21]

3 MANUFACTURING OF THE SPECIMENSThe CTOD (Crack Tip Opening Displacement) specimens according to BS 7448 [1] were taken outof the base material, different welds, and their heat affected zone of all three types of steels. Thedimensions can be seen in Figure 5.

Figure 5. Geometry and dimensions of the CTOD-specimens

For the determination of crack-arrest fracture toughness values nine specimens were taken fromthe base material, the weld metal, and the heat affected zone of each of the three different types ofspecimens: - Three-Point-Bend (3PB)-Specimens according to a test proposal of the TVFA [4], [5], [6], - Compact-Crack-Arrest (CCA)-Specimens according to ASTM E 1221 - 88 [7], and - Full-Thickness Compact-Crack-Arrest (FTCCA)-Specimens according to a test proposal of

Crosley and Ripling [8].For the three-point-bend (3PB) test method a side-grooved three-point-bend-specimen with specialdimensions, for the compact-crack-arrest (CCA) test method a side-grooved CT-specimen with aspecial thickness, and for the full-thickness compact-crack-arrest (FTCCA) test method a CT-specimen, which has a thickness equal to the wall thickness of the tested plate, were used. The

side-grooves provided for a straight crack front. The dimensions of the specimens, according to theinstructions of the three test procedures, can be seen in Figure 6.

Figure 6. Geometry and dimensions of the CA-specimens

All the specimens were manufactured in the following way. First, the specimens were takenoversized from the corresponding plates. All of them were worked out of the center-position of thethickness of the plate. The crack plane was always vertical to the surface of the plate, and itincluded the rolling-direction in case of the base material. In the weld metal specimens the crackplane was situated in the middle of the weld and parallel to the direction of the axis of the weld. Thespecimens used for testing the heat affected zone had to be etched in order to make the geometryof the weld clearly visible. Subsequently the position of the notch was fixed in a way that let thecrack plane run almost completely through the heat affected zone and also parallel to the directionof the weld axis.Afterwards the CA-specimens were prepared for applying the brittle starter welds. A chevron notchwithout a brittle crack starter weld is provided by Crosley and Ripling (1990) for the FTCCA-specimens, but this did not yield usable crack-initiation results. Therefore to all CA-specimen typescrack starter welds were applied in a single pass technique by the Boehler Welding TechnologyCompany with the electrodes FOX DUR 500 (with a diameter of 5 mm for the 3PB-specimens, andof 2 mm for the CCA- and FTCCA-specimens). The electrode Wearshield MM, recommended in anadditional paper to ASTM E 1221, was not expedient for this purpose. Research, made into thissubject, resulted in a weld metal with many cracks, leading to crack initiation at the tip of one of thesmall cracks in the weld metal of the starter weld. Using these brittle electrodes non uniform testconditions can therefore be provided for all the tests [22], [23].Afterwards a crack starter notch with the dimensions shown in Figure 5 and 6 was machined to allthe specimens by means of spark erosion. Then the specimens were machined to their finaldimensions, including the fatigue cracks in the CTOD-specimens and the side grooves in theCTOD-, 3PB-, and the CCA-specimens with a depth of 1.25 mm and a notch root radius of 0.2 mmby spark erosion too. The notches were made by spark erosion because milling of the notches ofthe CA-specimens would have overloaded the milling tool and, in addition, spark erosion providesfor smoother surfaces. Then the specimens were machined to their final dimensions. The sidegrooves were machined to obtain a straight crack front in the specimens.

4 DESCRIPTION OF THE TESTSAll the tests were carried out on the electromechanical 250 kN multi-purpose Schenck Trebeltesting machine RME 250 with integrated load measurement. A Cryoson cooling chamber, intowhich liquid nitrogen was injected, was used to attain the test temperature. To control the testtemperature, a thermocouple was welded to the surface of the CTOD-, 3PB-, and CCA-specimensat a location near to one of the side grooves, about 10 mm ahead of the starter notch and to theFTCCA-specimens near the expected crack extension plane about 5 mm ahead of the starternotch. The loading arrangement for the CTOD-specimens was according to BS 7448 and for the3PB-specimens similar to that one suggested in ASTM E 399-87. The specimens were placed onrolls and loaded by a loading ram. The velocity of the loading ram was 2 mm/min. To test the CCA-and FTCCA-specimens they were placed on a support block containing a hole. Then a wedge wasforced by a loading ram into a split-pin, deposited in the 12 mm-hole of the specimens until a crackstarted. The thickness of the support block and the existence of the support block hole allowed thewedge to push on 50 mm before it reached the base plate of the testing machine under the supportblock. The velocity of the loading ram was 20 mm/min. The lubricant Ha Tec was applied to thesurfaces of the support block, the wedge, the split-pin, and the specimen itself to reduce friction.The temperature dependency of the viscosity of the lubricant leads to different steepnesses in theindividual load-displacement records. For further details see [2] to [6], [21], [24] to [27].The crack mouth opening displacement was measured at the specimen surface between twointegrated knife edges at the CTOD- and 3PB-, and between two screwed-on knife edges at theCCA- and FTCCA-specimens. The clip gauge used had been specially designed for this purposeby the TVFA at the University of Technology in Vienna. During the test the load applied to the wedgeversus the displacement was directly recorded by an XY-plotter. The crack length at the tip of thefatigue precrack, a0, and the arrested crack length, aa, are defined as the arithmetic average of thereal crack lengths measured at three positions (at the center line of the fracture surface andhalfway between that center line and the bottoms of the side grooves or the ungrooved lateralsurfaces). To get reliable measurements, the crack fronts had to be marked by means of heat-tinting after the tests. For that purpose the specimens were heated to a temperature of 300 °C andheld at this temperature for a period of approximately 50 to 60 minutes. Finally the specimens werebroken completely in two at a temperature of about –196 °C.

5 TEST RESULTSIn Figs. 7, 8, and 9 photographs of some typical fracture surfaces of the tested specimens arereproduced. The bright zones show the fracture surfaces which resulted from the final brittlebreaking of the specimens at a temperature of about –196 °C.

Figure 9. Fracture surface of a CCA-specimen of the base material of X70 (temperature –80 °C)

In Fig. 9 showing a CCA-specimen taken from the base material, an extreme ligamentation can beseen, which resulted from the thermomechanical treatment of the material.Normaly in case of side grooves the crack front remained nearly straight during the test. Testing ofnon side grooved specimens, however, resulted in a bent crack front ("fingernail-shaped").

Figure 7. CTOD-values over the temperature for the base material and the weld metal of 1.4462with corresponding fracture surfaces (W ... rolled, SP ... welding specimen) [21]

Figure 8. CTOD-values over the temperature for different weld metals of 1.4462 with correspondingfracture surfaces (W ... rolled, Pl ... plate, P ... pipe, SP ... welding specimen) [21]

The CTOD and the KIa-values were calculated using the equations according to the standards orthe test proposals mentioned before. KIa could only be determined if a fast running crack (cleavagecrack) occured and was arrested. Furthermore, attention was paid to the criteria of validity.Figure 10 shows the results of the CTOD-tests on the pipeline-steel X70 (base material (BM), weldmetal (WM), and heat affected zone (HAZ)). On the base material tests were carried out at –30 °Cand 0 °C, on the weld metal and the heat affected zone at 0 °C. At a test temperature of 0 °C theaverage CTOD-value was 0,25 mm, the one of the weld metal was 0,19 mm and the one of theheat affected zone was 0,33 mm. Additionally the figure shows CTOD-values which are calculatedout of crack-arrest values and marked with a star in the figure. It can be clearly seen, that the weldmetal gives lower values than the heat affected zone and the base material.

Figure 10. CTOD-values over the temperature for longitudinal welded pipes out of X70 and differentshielded metal arc welds (BM ... base material, WM ... weld metal,

HAZ ... heat affected zone, SP ... welding specimen)

Figure 11 shows the CTOD-values over the temperature for different welding processes on thepipeline-steels X70 and X80 [21]. The values of the submerge arc welds (12) are a little bit higherthan the shielded metal arc welds (111). The big variations in the values of the base materials resultof the different melts tested.

Figure 11. CTOD-values over the temperature for different welds on pipeline-steels(BM ... base material, WM ... weld metal, HAZ ... heat affected zone,

111 ... shielded metal arc welding, 12 ... submerge arc welding, 135 ... gas metal arc welding)

The CTOD-tests on the rolled base material of the duplex-steel 1.4462, see Figure 7, wereperformed in a temperature range of –180 to –20 °C. The resulting CTOD-values were in between0,23 and 0,36 mm. For the weld metal tests were performed between –100 and –20 °C resultinginto CTOD-values between 0,08 and 0,3 mm [21]. The comparison of the shielded metal arc weld(SMAW, 111, DUPSP2) in figure 8 to the submerge arc weld (SAW, 12, DUPSP3) gave CTOD-

values between 0,35 and 0,43 mm for SMAW with basic electrodes and CTOD-values between0,21 and 0,33 mm for SAW. All tests were performed at –20 °C. The tests carried out between –80and –20 °C resulted into CTOD-values in between 0,18 and 0,53 mm.Figure 12 shows some crack-arrest test results for the base material, the weld metal, and the heataffected zone for X70.

Figure 12. Crack-arrest fracture-toughness over the temperature for X70 (base material, weldmetal, and heat affected zone, 3PB ... Three-Point-Bend, CCA ... Compact-Crack-Arrest,

FTCCA ... Full-Thickness Compact-Crack-Arrest)

No crack-arrest occured in the weld metal (WM) of the 3PB-specimens and therefore no values areplotted. Below a temperature of +20 °C the specimens out of the weld metal broke in two bycleavage, and above this temperature stable crack extension occurred without crack jump,whereas at the temperature of +20 °C both cases happened. The values of the weld metal are thelowest one and show a steep increase. At test temperatures higher than the cited upper limits for allthree materials and all three specimen types only stable crack extension occured, and therefore itwas not possible to determine a stress intensity factor for crack-arrest. Because only 9 specimensper material and specimen type were intended to test, the lowest test temperature was determinedby the upper limit and the temperature steps between the test temperatures. Temperatures belowthese values are not relevant for this steel, because they are far below its range of application, andtherefore no further tests were made at lower temperatures.

6 DISCUSSION OF THE RESULTSIn this work the fracture mechanics CTOD and crack-arrest toughness value, KIa, for the basematerial, the weld metal, and the heat affected zone of the pipeline-steels X 70, X80, and theduplex-steel 1.4462 were determined for a wide temperature range using:- CTOD-specimens according to the British standard BS 7448,- 3PB-specimens according to a test proposal by the TVFA,- CCA-specimens according to the American standard ASTM E 1221, and- FTCCA-specimens according to a test proposal by Ripling and Crosley.The results of all the tests show very clearly that there are great differences in the CTOD-valuesdepending on the used melts of the steels and the applied welding processes.In case of the crack-arrest tests it could be found that the temperature ranges, between which intests a fast-running crack occurs and consequently gets arrested, are depending on the testmethod used. On the basis of these test results it can be deduced that the test proposal of theTVFA at the University of Technology in Vienna is most easily applicable. The tests can beperformed easily and quickly. A disadvantage which could be mentioned is that in these tests only asmall temperature range was observed within which crack-arrest occured.The handling of the CCA- and FTCCA-specimens in the tests was more difficult, but thetemperature range for the determination of the KIa-values was larger and extended to lowertemperatures. Regarding the results, those of the 3PB-specimens were almost always on the safeside or at least within the scatterband of those of the two other specimen types.

REFERENCES[1] BS 7448, Part 1, 1991: Fracture mechanics toughness tests, Part 1: Method for determination of KIc,

critical CTOD and critical J values of metallic materials.[2] Felber, S.: Vergleich verschiedener Probenformen bezüglich der Ermittlung der Rißauffangzähigkeit.

Doctoral thesis, University of Technogy, Vienna, May 1994.[3] Felber, S., Schneeweiß, G., Varga, T.: Crack-Arrest in a Pipeline Steel. Second International Confe-

rence on Pipeline Technology, Ostende, Belgium, September 11 - 14, 1995, pp. 491-502.[4] Schneeweiß, G., Varga, T.: Zur praktischen Bedeutung und experimentellen Ermittlung der Rißauf-

fangzähigkeit KIa. VGB-Conference, Essen, Germany, 19.-21.9.1990, G4, S. 1-22.

[5] Schneeweiß, G., Varga, T.: Erfahrungen mit Dreipunkt-Biegeproben zur Ermittlung der Rißauffang-zähigkeit KIa an niedriglegiertem Stahl. Swiss Mat. 1 (1989) No. 6, S. 29-37.

[6] Kanada, A.H., Varga, T., Schneeweiß, G.: Proposed Test Method for Determining Plane-Strain Crack-Arrest Fracture Toughness, KIa, of Ferritic Steels, Using Three-Point Bend [(SE)B] Specimens. Draft

Proposed Test Method.[7] ASTM E 1221 - 88: Standard Test Method for Determining Plane-Strain Crack-Arrest Fracture

Toughness, KIa, of Ferritic Steels.

[8] Material Research Laboratory Inc.: Test Method for Determining Full-Thickness Crack Arrest FractureToughness, Ka, of Metallic Materials. Draft 2 - May 1990.

[9] API Spec 5L: Specification for Line Pipe. Ed. April 1995.[10] DIN 17 172: Stahlrohre für Fernleitungen für brennbare Flüssigkeiten und Gase, Technische

Lieferbedingungen. Ausg. Mai 1978.[11] OENORM EN 10 208 - 2: Stahlrohre für Rohrleitungen für brennbare Medien, Technische Lieferbe-

dingungen, Rohre der Anforderungsklasse B (enthält auch Berichtigung AC:1996). Ausg. 1. April 1997.[12] DIN 17 440: Nichtrostende Stähle, Technische Lieferbedingungen für Blech, Warmband und gewalzte

Stäbe für Druckbehälter, gezogenen Draht und Schmiedestücke. Ausg. September 1996.[13] DIN 17 441: Nichtrostende Stähle, Technische Lieferbedingungen für kaltgewalzte Bänder und

Spaltbänder sowie daraus geschnittene Bleche für Druckbehälter. Ausg. Februar 1997.[14] SEW 400: Nichtrostende Walz- und Schmiedestähle. Ausg. Februar 1991.[15] ÖNORM EN 10 088 - 1: Nichtrostende Stähle, Verzeichnis der nichtrostenden Stähle. Ausg. 1. Oktober

1995.[16] ÖNORM EN 10 088 - 2: Nichtrostende Stähle, Technische Lieferbedingungen für Blech und Band für

allgemeine Verwendung. Ausg. 1. Oktober 1995.[17] ÖNORM EN 10 088 - 3: Nichtrostende Stähle, Technische Lieferbedingungen für Halbzeug, Stäbe,

Walzdraht und Profile für allgemeine Verwendung. Ausg. 1. Oktober 1995.[18] TÜV Werkstoffblatt 225: Ferritisch-austenitischer Stahl X 2 CrNiMoN 22 5. Ausg. Oktober 1994.[19] ÖNORM EN 10 027 - 2: Bezeichnungssysteme für Stähle, Nummernsystem. Ausg. 1. Dezember 1992.[20] ÖNORM EN 10 027 - 1: Bezeichnungssysteme für Stähle, Kurznamen, Hauptsymbole. Ausg. 1.

Dezember 1992.[21] Felber, S.: Pipelinebau (Untersuchungen an ferritisch-perlitischen Stählen, bainitischen Stählen und

Duplex-Stählen für Fernleitungen). Habilitation, TU Wien, Dezember 2000.[22] Mäser, N.: Ermittlung der Kenngrößen zur Beschreibung des Rißauffangverhaltens des Behälterstahles

ALDUR 58 D an Dreipunktbiegeproben in Abhängigkeit der Versprödungsschweißung. Master´s thesis,University of Technology, Vienna, 1995.

[23] Lipp, M., Mäser, N., Felber, S., Varga T.: Vergleich verschiedener Versprödungsschweißungen undverschiedener Kerbradien bei Rißauffanguntersuchungen. Schriftenreihe der Technischen UniversitätWien, Schweißtechnik und verwandte Verfahren 2000, Internationale Konferenz über Schweißtechnik,Werkstoffe und Werkstoffprüfung, Bruchmechanik und Qualitätsmanagement, Welding Technology andRelated Fields 2000, International Conference on Welding Technology, Materials and Materials Testing,Fracture Mechanics and Quality Management, Vienna, Austria, 25. - 27. September 2000, S. 251 - 262.

[24] Loibnegger, F.: Zur Rißeinleitung und ihrer Bedeutung in Stahl. Doctoral thesis, University of Techno-logy, Vienna, October 1990.

[25] Felber, S., Varga, T., Schneeweiss, G.: Crack-Arrest Measurements in a Pressure Vessel Steel. 13thInternational Conference on Structural Mechanics in Reactor Technology, Porto Alegre, Brasil, August13 - 18, 1995.

[26] Felber, S., Alber, W., Oberndorfer, M., Schneeweiß, G., Varga, T.: Crack-Arrest Measurements inPipeline Steels. Eurojoin 2, Florence, Italy, May 16 - 18, 1994, pp. 707-717.

[27] Felber, S., Oberndorfer, M., Alber, W.-R., Kanada, A., Schneeweiß, G., Varga, T.: Crack-ArrestMeasurements in Pipeline Steels. Workshop on User´s Experience in Crack-Arrest Testing, ASTMConference in Atlanta, USA, May 17, 1993.