Residual stresses in steel specimens induced by laser cladding and their effect on fatigue strength

Available online at www.sciencedirect.com Physics Procedia 39 (212 ) 354 361 LANE 212 Residual stresses in steel specimens induced by laser cladding and their effect on fatigue strength Henry Köhler a,, Knut Partes a, Joana Rebelo Kornmeier b, Frank Vollertsen a a BIAS Bremer Instiut für angewandte Strahltechnik, Klagenfurter Str. 2, 28359 Bremen, Germany b Forschungsneutronenquelle Heinz Maier-Leibnitz ZWE, FRM II, Lichtenbergstr. 1, 85747 Garching, Germany Abstract Residual stresses resulting from circumferential laser cladding of fatigue test specimen of austenitic steel X5CrNi18-1 and heat treatable steel 42CrMo4 with Stellite 21 are evaluated by neutron diffraction. Below the interface of cladding and base material the former shows compressive residual stresses, the latter shows tensile residual stresses. Locations of crack initiation during fatigue testing correlate to the findings. Austenitic steel specimens crack at the surface whereas heat treatable steel components and specimens crack close to the interface. Fatigue strength of both material systems drops due to laser cladding whereby the drop of heat treatable steel is significant. It can be concluded that residual stress distribution needs to be optimized in order to conserve maximal fatigue strength. 212 211 Published by Elsevier B.V. Ltd. Selection and/or review under responsibility of Bayerisches Laserzentrum GmbH Open access under CC BY-NC-ND license. Keywords: Laser cladding; residual stresses; neutron diffraction; austenitic steel; heat treatable steel 1. Introduction Laser cladding is applied widely for reconditioning and surface tailoring i.e. in terms of tribological and corrosive behavior. Examples can be found in the generation of functional surfaces of forming tools [1], in the field of reconditioning and surface layer tailoring of large shafts [2] and marine engine components [3]. Process parameter windows are often limited by the susceptibility of cracking caused by residual stresses. To prevent cracking over a wider range of process parameters for example additional inductive heating can be applied to reduce the magnitude of residual stresses [4]. To maintain certain temperatures or isotherms during processing closed-loop control systems i.e. on basis of pyrometers [5] or * Corresponding author. Tel.: +49-421-218-589 ; fax: +49-421-218-5863. E-mail address: koehler@bias.de. 1875-3892 212 Published by Elsevier B.V. Selection and/or review under responsibility of Bayerisches Laserzentrum GmbH Open access under CC BY-NC-ND license. doi:1.116/j.phpro.212.1.48

Henry Köhler et al. / Physics Procedia 39 ( 212 ) 354 361 355 2D sensors [6] have been developed. However, it is not yet approved to apply laser cladding for highly dynamically loaded parts as it is not predictable how fatigue properties are altered. Preliminary results proved that fatigue strength of round [7] and flat specimen as well as of complete crankshaft segments [8] seriously decreased after laser cladding with the cobalt-based alloy Stellite 21. From these studies it could be evaluated that locations of crack initiation in low-alloy steel 42CrMo4 were close to the interface of cladding and base material. Results of preliminary fatigue tests of the specific specimen types and respective material combinations are summarized in table 1. The fatigue strength of round specimen under alternating bending load (R = -1) dropped by 69.9% and under repeated load (R = ) by 6.6%. The fatigue strength of this material combination under alternating tensile-compressive load dropped by 59.3%. Crank pins of 42CrMo4 were tested according to a CIMAC design standard (International Council on Combustion Engines) in which a defined geometry related load was to be borne by the component. Instead of bearing the defined maximum number of load cycles (5 million) the components broke at a mean number of 1.3 million cycles. On the other hand, high-alloy steel X5CrNi18-1 cracked at the surface. The fatigue strength dropped by 2.6% at alternating bending load and by 9.3% at repeated load. Table 1. Fatigue strength of low and high alloyed steel specimen and components determined in preliminary investigations, *results kindly provided by IWT Stiftung Institut für Werkstofftechnik, Bremen; **results kindly provided by SLV Halle Specimen type, geometry and material combination Fatigue strength in for R = Fatigue strength in for R = -1 Reference Uncladded base material X5CrNi18-1 (round, diam.1 mm) 42CrMo4 (round, diam. 1 mm) 42CrMo4 (flat, height 1 mm) 42CrMo4 (crankshaft crank pin, diameter 152 mm) 324* 464* - - 424* 572* 332** - (no defect within 5 million load cycles)** [8] [8] [7] [7] Cladded with Stellite 21 X5CrNi18-1 (round, diam.12 mm) 294 (-9.3%)* 413 (-2.6%)* [8] 42CrMo4 (round, diam.12 mm) 183 (-6.6%)* 172 (-69.9%)* [8] 42CrMo4 (flat, height 1 mm) - 135 (-59.3%)** [7] 42CrMo4 (crankshaft crank pin, - - (defect after [7] diameter 152 mm) 1.3 million load cycles)** The heat affected zone (HAZ) of the cladded steel 42CrMo4 showed martensitic hardening. Due to martensite formation and the accompanied expansion of the unit cell compressive residual stresses can be expected in the HAZ. Contrary to the findings these should increase fatigue strength. The investigations presented in this article were performed to evaluate the role of residual stresses as reason for the drop of fatigue strength, especially in steel 42CrMo4. Therefore residual stresses were to be analyzed. Commonly applied methods to determine residual stresses by x-ray diffraction i.e. by layer wise etching as executed in connection with laser hardening in [9] bear the risk of influencing the stress situation due to material removal. Neutron diffraction has proved to be highly suitable to determine macroscopic residual stresses in steels as it is for example shown in [1].

356 Henry Köhler et al. / Physics Procedia 39 ( 212 ) 354 361 2. Experimental Basis of the presented investigations are the round specimen mentioned in table 1. These consisted of steel 42CrMo4 tempered to 35 HV.3 and steel X5CrNi18-1. The uncladded base material as well as the cladded specimen geometry is shown in figure 1. A batch of specimen was laser cladded with Stellite 21 circumferentially. A Trumpf HL 46D Nd:YAG laser was used as heat source. Cladding material was delivered coly into the process zone by the Precitec processing head YC5. The cladding process was closed-loop controlled on basis of pyrometric measurements in the melt pool center with an Impac Infrared IGAR 12 LO pyrometer, similarly as reported in [5]. A track offset of.8 mm was applied. A set-temperature of 165 C, a feed velocity of 1 m/min of the heat source along the surface and a powder feed rate of 15 g/min were applied. The closed-loop control led to a mean output power of 9 W at the center position of the fatigue test specimens. The laser power resulting by the process control and the respective temperature signal is given in figure 2. Cladded specimens were ground after laser processing. Fig. 1. Side view on uncladded base material with d 1 (black) and cladded specimen with d 2 (red) with positions of residual stress measurements (A: all specimens, B: additional positions cladded specimens were measured) a) 2 25 b) 2 Temperature C 16 14 12 2 4 6 s Time W 15 5 Output power of heat source Temperature C 16 14 12 2 4 6 s Time 25 W 15 5 Output power of heat source Fig. 2. Temperature and power signal during laser cladding of a) X5CrNi18-1 and b) 42CrMo4 bending specimen with Stellite 21 Cross sections of the laser cladded specimen are shown in figure 3. Very homogeneous cladding results were achieved in terms of clad height, dilution and connection of cladding to base material. The cladding and the interface of cladding and base material were free of defects. The following material combinations were investigated by neutron diffraction: X5CrNi18-1 base material, X5CrNi18-1 base material laser cladded with Stellite 21 and ground, 42CrMo4 base material and 42CrMo4 base material laser cladded with Stellite 21 and ground.

Henry Köhler et al. / Physics Procedia 39 ( 212 ) 354 361 357 a) b) Fig. 3. Cross section of a) X5CrNi18-1 and b) 42CrMo4, each laser cladded with Stellite 21 [8] The residual stress analyses were determined by neutron diffraction at FRM II, Technische Universitat München (TUM), on the STRESS-SPEC instrument. This diffractometer is dedicated to texture and residual stress analysis [13]. A PSD area detector of 3 x 3 cm² collected the diffracted radiation. A bent silicon Si (4) monochromator was set at a take-off angle which provides a wavelength around 1.6734 A. The measurements were performed around a scattering angle of S austenite and S for the ferrite (heat treatable steel). A gauge volume of 1x1x1 mm³ was defined by the primary and secondary slits for all the principal directions, and directions. Residual stresses in the considered directions are calculated by Eq. 1: E (1 ) E i i j, j 1,2, 3 (1 )(1 2 ) (1 2 )(1 ) j and j i (1) where i, i are the principal stress and strains, and E and the Young s modulus and Poisson s ratio for the direction perpendicular to the {} diffraction planes. For the austenitic steel X5CrNi18-1 the diffraction peak of the plane {311} (Fe- line) and for the heat treatable steel 42CrMo4 the diffraction peak of the plane {211} (Fe- line) was analyzed. According to [11] Young s moduli and Poisson ratio of these planes are E 311 = 175 GPa and 311 =.31 for austenite and E 211 = 22 GPa and 211 =.28 for ferrite. Stress-free references samples were cut out of the center position of each specimen type. Lattice displacements and resulting residual stresses were evaluated by determination of the reflection peak shift relative to the respective reference. For the measurements of residual stresses in every specimen line scans in center position along the radius (, from center to surface) were performed as sketched in figure 1. The measurements were made in, and direction. The gauge volume position closest to the workpiece surface was partly placed outside the base material. To quantitatively determine residual stress in these locations additional x-ray diffraction measurements were performed at IWT Stiftung Institut für Werkstofftechnik, Bremen.

358 Henry Köhler et al. / Physics Procedia 39 ( 212 ) 354 361 3. Results and Discussion The evaluation of residual stress measurements along line A in figure 1 is shown in figure 4. The outer most position measured by neutron diffraction is marked by a gray box. Residual stresses determined by x-ray diffraction are given in detail in the upper right of figure 4a and figure 4b. Comparing the resulting stresses determined by x-ray and neutron diffraction it can be evaluated that the measurement at the outer position is more accurate for 42CrMo4 than for X5CrNi18-1. Taking into account that the neutron diffraction measurement gives a mean value for a gauge volume of 1 mm³ both results can be interpreted as qualitatively correct. The uncladded X5CrNi18-1 specimen shows higher compressive residual stress at the surface than in the core (figure 4a). The X5CrNi18-1 laser cladded with Stellite 21 shows compressive residual stress at the interface cladding to base material. Towards the core residual stresses shift from compressive to tensile (figure 4c). The uncladded 42CrMo4 specimen shows higher compressive residual stresses at the surface than in the core (figure 4b). In the cladded and ground steel 42CrMo4 high tensile residual stresses could be measured near the interface of base material and cladding (figure 4d). a) 5 b) 5-2 4,9 mm 5, 2 - - X5CrNi18-1 - - 1 2 3 4 mm 6 42CrMo4 1 2 3 4 mm 6 c) 5 d) 5 2 2 - - X5CrNi18-1 42CrMo4 - - 1 2 3 4 mm 6 1 2 3 4 mm 6 Stellite 21-4,9 mm 5, Stellite 21 Fig. 4. Residual stresses in center position (line A in figure 1), a) and b) in uncladded base materials; c) and d) in cladded specimens; gray boxes show indicative results; inserts in the upper right of a) and b) give residual stresses evaluated by x- ray diffraction at the surfaces To test if the induced residual stresses are homogeneous along the longitudinal axis of the cladded specimen, measurements have been executed 5 mm offset from the specimens center position as sketched in figure 1 along line B. The results are shown in figure 5. The high alloyed steel X5CrNi18-1 shows compressive residual stress close to the base material surface in the same magnitude as determined in center position. In the low alloyed steel 42CrMo4 tensile residual stress can be evaluated close to the

Henry Köhler et al. / Physics Procedia 39 ( 212 ) 354 361 359 surface which switches to compressive residual stress 1 mm below the surface. Both offset measurements in low as well as in high alloyed steel are qualitatively identical to the ones determined in the respective center positions. a) 5 b) 5 2 2 - - X5CrNi18-1 42CrMo4 - - 1 2 3 4 mm 6 1 2 3 4 mm 6 Stellite 21 Stellite 21 Fig. 5. Residual stresses in cladded base materials at 5 mm offset from center position (line B in figure 1) The results correlate with fatigue tests reported earlier [7, 8]: Uncladded steel 42CrMo4 and uncladded and cladded X5CrNi18-1 cracked at the surfaces as shown in figure 6a for cladded X5CrNi18-1. On the other hand laser cladded 42CrMo4 specimens cracked below the specimen surface close to the interface between cladding and base material as shown in figure 6b. The measurements indicate that the applied laser cladding process induced a residual stress situation which influences the fatigue properties negatively. Especially fatigue strength of low alloyed steel is strongly reduced by laser cladding with the process parameters applied in the presented studies. Fig. 6. Fracture surface of a) X5CrNi18-1 and b) 42CrMo4, each laser cladded with Stellite 21 [8] Despite it can clearly be identified that the induced residual stress fields do not benefit fatigue strength, the measurements indicate that the stress fields are uniform along the longitudinal specimen axis. This is the case for high as well as low alloyed steel. The closed-loop process control was applied in order to maintain a homogeneous cladding quality along the weld trajectory in the first place. On basis of the results it can furthermore be assumed that the uniform peak temperature also causes a uniform residual stress field.

36 Henry Köhler et al. / Physics Procedia 39 ( 212 ) 354 361 Conservation of momentum demands that the integral residual stress along one cross section results in zero. The metallographic cross section of the cladded steel 42CrMo4 shows that the complete cross section 5 mm around the specimen center was transformation hardened. This could also be proofed by hardness indentations presented in [8]. Tensile residual stresses in the base materials close to the surface can either be caused by the interaction with the inner part of the base material or by interaction with the cladding material. In the austenitic steel X5CrNi18-1 low compressive residual stresses are found in the HAZ and low tensile residual stresses are found in the core of the specimen. In the first place these findings disagree with [12] where low tensile residual stresses in the HAZ and higher compressive residual stresses in the depth were evaluated in flat specimen of a similar material combination. It can be anticipated that the difference between literature and the present investigations is based on the different specimen shape and circumferential processing trajectory. 4. Summary By the application of a closed-loop control a constant peak temperature could be set during the laser process. This resulted in highly homogeneous cladding quality. Furthermore it is assumed that the constant peak temperature induced uniform residual stress fields, both in low alloyed as well as in high alloyed steel. High tensile residual stresses could be evaluated in the HAZ of the steel 42CrMo4 despite compressive residual stresses were expected due to martensite formation. Results indicate that this is due to the hardening of the complete cross section of the specimen. Especially fatigue strength of the cladded heat treatable steel 42CrMo4 proved to decrease. The results correlate well with findings from fatigue tests: Locations of crack initiation coincide with the areas of high tensile residual stresses. Acknowledgements The authors thank the Deutsche Forschungsgemeinschaft (DFG, German Research Community) for funding the project VO_53/31 (Laserbeschichten und Simulation der Temperatur- und Eigenspannungssituation) within the collaborative project DFG PAK 299 (Vorhersage der Dauerfestigkeit Laserbeschichteter Komponenten). Furthermore we thank our project partners BIME Bremer Institut für Strukturmechanik und Produktionsanlagen and IWT Stiftung Institut für Werkstofftechnik for the successful work within our common collaborative project. References [1] Schmidt, M.; Kolleck, R.; Grimm, A.; Veit, R.; Bartkowiak, K.: Direct laser deposition of Cu alloy on forming tool surfaces Process window and mechanical properties; CIRP Annals - Manufacturing Technology 59 (21), 211 214. [2] Walz, C.; Nägeler, S.: Laserbeschichten mit Festkörperlasern im industriellen Umfeld; 6. Laseranwenderforum (LAF'8), eds.: F. Vollertsen, T. Seefeld. BIAS-Verlag; Bremen, Bd. 36, (28), 131-138. [3] Wagner, F.; Welzenbach, M.; Buschhoff, J.; Gall, S.: Forschung wird Wirtschaft oder Laser-Pulver-Auftragsschweißen im Schiffsreparaturbereich; 6. Laseranwenderforum (LAF'8), eds.: F. Vollertsen, T. Seefeld. BIAS-Verlag; Bremen, Bd. 36, (28), 139-146. [4] Brückner, F.; Lepski, D. & Beyer, E.: Modeling the influence of process parameters and additional heat sources on

Henry Köhler et al. / Physics Procedia 39 ( 212 ) 354 361 361 residual stresses in laser cladding; Journal of thermal spray technology (27), 16, 355-373. [5] Doubenskaia, M.; Bertrand, Ph.; Smurov, I.: Pyrometry in laser surface treatment; Surface & Coatings Technology 21 (26), 1955 1961. [6] Hofman, J. T.: Development of an observation and control system for industrial laser cladding; Universiteit Twente, University of Twente (29). [7] Koehler, H.; Partes, K.; Seefeld, T.; Vollertsen, F.: Influence of laser reconditioning on fatigue properties of crankshafts; Lasers in Manufacturing (LIM 211), eds.: M. Schmidt, M. F. Zaeh, T. Graf. Elsevier B.V. Amsterdam (211), 512-518. [8] Koehler, H.; Schumacher, J.; Schuischel, K.; Partes, K.; Bomas, H.; Jablonski, F.; Vollertsen, F.; Kienzler, R.: An approach to calculate fatigue properties of laser cladded components; Production Engineering Research and Development, Vol. 6, Issue 2, Springer Berlin/Heidelberg (212), 137-148. [9] Habedank, G.; Woitschig, J.; Seefeld, T.; Jüptner, W.; Vollertsen, F.; Baierl, R.; Bomas, H.; Mayr, P.; Schröder, R.; Jablonski, F. & Kienzler, R.: Endurance limit of pulsed laser hardened component-like specimens Experiment and simulation; Materials Science and Engineering (28), A488, 358 371. [1] Withers, P. J.: Mapping residual and internal stress in materials by neutron diffraction; C. R. Physique 8 (27), 86-82. [11] Eigenmann, B.; Macherauch, E.: Röntgenographische Untersuchung von Spannungszuständen in Werkstoffen Teil 3; Mat.-wiss. u. Werkstofftech. 27 (1996), 426-437. [12] Pilloz, M.; Pelletier, J. M. & Vannes, A. B.: Residual stresses induced by laser coatings: phenomenological analysis and predictions; Journal of Material Science (1992) 27, 124-1244. [13] Hofmann, M.; Schneider, R.; Seidl, G. A.; Rebelo Kornmeier, J.; Wimpory, R. C.; Garbe, U.; Brokmeier, H. -G.: The new materials science diffractometer STRESS-SPEC at FRM-II; 8th International Conference on Neutron Scattering; Sydney, Australia (25); Physica B-Condensed Matter Volume 385-86, 135-137.

ID Title Pages 187365 Residual Stresses in Steel Specimens Induced by Laser Cladding and their Effect on Fatigue Strength 8 http://fulltext.study/journal/1772 http://fulltext.study