Showing papers by "Otmar Vöhringer published in 2005"

Influence of heating rate, cooling rate and numbers of pulses on the microstructure of AISI 4140 after short-time-hardening

Abstract: In recent years, significant progress has been made in application of lasers to surface hardening processes. While there are many investigations on the transformation behaviour during cooling based on initially homogeneous austenite states, there is a lack of data concerning austenite formation at very high heating rates and the impact of the initial austenitic condition on the final microstructure. Moreover the influence of cyclic austenite-martensite-transformation on the microstructure has been studied rarely. Therefore, in this paper the effect of heating rates up to 10000 K/s, cooling rates up to 3000 K/s and the impact of cyclic austenite-martensite-transformations on the microstructure of AISI 4140 (German grade 42CrMo4) are presented.

Similarity Rules for the Shot Peening Process Based on Finite Element Simulations

Abstract: Multiple shot impacts are analyzed using a three-dimensional Finite Element model with a user defined temperatureand strain-rate-dependent material formulation. Based on this analysis, a method was developed for the prediction of changes in the material state in areas close to the surface by applying similarity mechanics. This approach provides a tool for the quick estimation of interesting work piece properties after peening especially when a vast field of process parameter variations has to be analyzed. The predictions of the applied similarity mechanics were compared with an explicit Finite Element analysis. SUBJECT INDEX Finite Element analysis, similarity mechanics, residual stresses INTRODUCTION Finite Element (FE)-simulation of shot peening processes may be used to predict surface states after shot peening without costly experiments and provide clear physical results already in the design process. Especially when variations of process parameters are of high interest the claim for fast results can be met by applying similarity mechanics on this process simulation. Based on the 3D shot peening FEsimulation of (Schwarzer, 2003) and the similarity rules applied on a 2D single shot simulation in (Kotschenreuther, 2003) a promising tool for the quick estimation of interesting work piece properties after peening for a vast field of process parameters will be presented. METHODS Model Geometry and Material Definition The FE-simulations were performed using ABAQUSlexplicit by applying the procedure described in (Schwarzer, 2003). The work piece is defined as a cuboid with a quadratic base. The dimensions of this plate can be set to any designated value. The plate is meshed with hexahedral elements with reduced integration and an decreasing element size from the edges of the cuboid to the impact centre. Shock waves occurring after an impact shall not be reflected. Therefore the plate is framed by so-called "infinite elements". The shot is formulated as a rigid body half sphere with the mass and moment of inertia of a full spherical shot based on measurements of the cast steel shot. Size, MODELLING AND SIMULATION kinematics, impact location and order are parameterized and allow a vast field of a single or combined variation of the typical shot peening parameters. The behaviour of the work piece material, quenched and tempered steel AlSl 4140 (German grade: 42CrMo4), was properly described using a constitutive law considering the influence of temperature and strain-rate on the flow stress on the basis of thermally activated dislocation slip (Schulze, 2000). It was implemented into the finite element code using a user subroutine VUMAT (Kotschenreuther, 2003), (Schwarzer, 2003) together with the work hardening behaviour using a "generalized Voce" constitutive equation according to (Tome, 1984). The material behaviour was calculated adiabatically since 3D-elements with thermo-mechanical coupling were not available in the FE code. The lack of heat conduction can be tolerated for the realistic impact duration of few microseconds. The complete formulation consists of 9 material parameters that were determined by a numerical fit of data obtained from tensile tests at different temperatures and strain-rates to the material law similarly to (Schulze 2000). By comparing the shape of simulated and experimentally produced shot impacts, the material law could be validated. To describe the contact between shot and work piece, isotropic Coulomb friction with a coefficient of friction ,U = 0.4 is used. The cast steel shot is modelled as a rigid body with a density of 7.85 g/cm3. To achieve a realistic modelling of a shot peening process an arrangement of the spheres was chosen that provides a closest packed dimple pattern on the surface with a certain impact order of the shots (Fig. 1). The grey marked inner area which can be approximated with a circle was used for the calculation of residual stress profiles. As shown in (Schwarzer, 2003) the fact that the shots impact one after the other instead Fig. 1: Arrangement and order of of impacting simultaneously has great impacts as well as area evaluated to influence on the developing residual calculate the residual stress profile stresses due to inhomogeneous plastic and roughness deformations and therefore changes in the local stress states. Similarity Mechanics According to the Buckingham theorem (Buckingham, 1914), a dimensionless output value a of a physical problem depends only on p = n q dimensionless numbers D,, ..., D, where n is the number of influence parameters and q is the rank of the dimension matrix (Kotschenreuther, 2003). In this study the influence parameters are: shot diameter d, shot velocity v, impact angle a, distance between dimple centres I determining coverage, initiate temperature 6, friction coefficient p, Young's modulus El Poisson's ratio v, density of plate and shot p,, and ps and the above mentioned 9 remaining material parameters, hence altogether 19 influence parameters. Further parameters characterizing the impact arrangement and sequence are not considered here and were always kept constant. In this case the rank of the dimension matrix equals the number of the relevant fundamental units: length, mass, time and temperature. That means that to fully 96 ESP9 : SHOT PEENING describe the scaling problem of the impacts it is necessary to consider p = 19 4 = 15 dimensionless numbers. The knowledge of the functional dependencies of a dimensionless output a on the 15 numbers a = a (Dl , ..., Dl5) would then completely describe the scaling problem. As previously described in (Kotschenreuther, 2003) the choice of the dimensionless numbers is done in a way that whenever possible, each varying influence parameter only affects a single number. As can be seen below the shot velocity for example only affects the so called Cauchy number, giving the ratio of the impact velocity to the internal sound velocity of the plate. Three process parameters and hence their according similarity numbers were varied: where S i is the size number representing the shot diameter. Ca is the Cauchy number describing the shot velocity. a is the angle between the normal to the work piece surface and the impact velocity vector. The remaining 12 dimensionless numbers were kept constant and describe the coverage, the initial process temperature, the friction and the material properties. Dimensionless Output Values The results of the simulation were evaluated using the characteristics of the residual stress profile in a dimensionless manner. In order to get normalized values the residual stress components perpendicular and parallel to the projection of the velocity RS onto the surface at the surface and in the maximum OF , 0: ; and a,,, were related to the Young's Modulus E. The depths of maximum and zero residual stresses z , , , and zo as well as the roughness Rt of the dimples were related to the shot diameter d. Estimation of Output Values The determination of the full functional dependence of an output variable a on the 15 similarity numbers would exceed a maintainable computational effort. Therefore only the 3 process parameters were taken into account and as further simplification, a product ansatz, stating 1 a (Si, Ca, Q) = a, (Si) . a, , (Ca) . a, (a) a0 (2) was made, where a0 represents the output value for standard material and process parameters and a; (0 are functional descriptions of the influence of the dimensionless input parameter i on the output parameter when all other input parameters were kept constant at standard values Sio = 1.10 10' (acc. to d = 0.564 mm), Cao = 6.77. 10 -3 (acc. to v = 35 m/s) and GO = 0'. With tolerable computational costs of hardly more than a dozen of calculation variations for the deduction of the one dimensional functional relationships it is possible to determine interesting output values for arbitrary input values within minutes by using the product ansatz whereas a single FE-analysis and its evaluation take about 24 hours in this example. The impressive time advantage of the product MODELLING AND SIMULATION 97 ansatz is then given when the number of different parameter variations that are of possible interest exceeds the number of necessary FE-analyses for the determination of the functional dependence of the output values from one input value. RESULTS The two charts in Fig. 2 show the dependence of the dimensionless output values on the Si-number with a variation range of 0.5 3.5 Sio while all other 14 input values remained constant. Results of the computations with standard values are highlighted with larger symbols. With increasing Si-number the normalized maximum and surface residual stresses decrease slightly although the impact energy, proportional to the shot mass, increases with the third power of the diameter. This development of the residual stresses can be explained considering Hertzian stress theory with higher and sharper stress distribution caused by decreasing radii of the contact partners. Despite an impact angle of a0 = 0" the residual stress components in two perpendicular directions spanning the surface of the work piece differ systematically of about 20 %. This may be an artefact caused by the asymmetric impact order in the simulation. A total constancy of the dimensionless ratio z / d over Si was determined for the depth of the maximum stresses, the zero-crossing depth of the stresses and the output value for the roughness depth Rt / d meaning that the bare depth and roughness values increase linearly with ascending shot diameter. Fig. 2: Dimensionless output values vs. Si for constant Ca = Cao and a = a. v [rnls] 0 10 20 30 40 50 60 70 80 6 . 0 ~ 1 0 ~ ~

Determination of Residual Stress Gradients in Al2O3 Ceramics after Thermal and Mechanical Loading

Abstract: Al2O3 ceramic plates of diameter 166 mm and thickness 5 mm were subjected to frictional loading regimes. The resultant residual stress effect in the material was analyzed using the sin2y and AP methods based on a modified measurement condition for the pole angle y and the azimuth angle F on a 4 cycle goniometer. Furthermore, the frictional loading of the Al2O3 plates lead to clear differences in the residual stress state near the surface. An underlying correlation was observed between the residual stress state of the Al2O3 material and its residual strength with this effect being ascribed to the magnitude of frictional load applied.

Auswirkung zyklischer Temperaturänderungen beim Laserstrahlhärten auf den Randschichtzustand von vergütetem 42CrMo4

Abstract: Kurzfassung Bei der Laserstrahlhartung von Stahlen lassen sich bei geeigneter Temperatur-Zeit-Fuhrung durch Gefuge, Eigenspannungs- und Verfestigungszustand charakterisierte Randschichtzustande erzeugen, die verbesserte mechanische und tribologische Eigenschaften aufweisen. Das Spektrum der bei derartigen Behandlungen erzielbaren Zustande ist, insbesondere bei zyklischen Temperaturanderungen, bislang nicht untersucht. In der vorliegenden Arbeit werden daher systematische Untersuchungen zum Laserstrahlharten mit zyklischen Temperaturanderungen am niedriglegierten Vergutungsstahl 42CrMo4 vorgestellt. Die unter Variation von Erwarmgeschwindigkeit, Abkuhlgeschwindigkeit, Zyklenzahl und Minimaltemperatur entstehenden Gefuge werden anhand der sich einstellenden Hartetiefenprofile analysiert und bewertet.

Interaction between residual stresses and surface defects in alumina ceramics under tribological loading

Abstract: In the presented research work Al2O3 ceramic plates were subjected to frictional and thermal loadings with the main purpose to investigate typical failure mechanisms of ceramic friction components under such loading conditions. This was done by calculating the local temperatures and stresses using the finite element method and subsequently verifying the model developed using experimental results. By conducting residual stress analyses, 4-point bending tests, ceramographic analyses and fractographic observations on the ceramic components, a detailed interpretation of the crack initiation and propagation was possible. Eigenspannungen und Schadigungen in Al2O3 – Keramik nach trockener Friktionsbeanspruchung Keramikscheiben aus Al2O3 wurden an einem Laserprufstand thermisch und an einem Friktionsprufstand thermisch-mechanisch belastet, um gezielt Schadigungen in das Material einzubringen. Anschliesend wurde der Schadigungszustand durch Analysen des Eigenspannungszustandes, 4-Punkt Biegeversuche sowie keramographische und fraktographische Untersuchungen charakterisiert. Parallel wurde eine Finite-Elemente-Simulation durchgefuhrt, um die auftretenden Temperaturen und Spannungen zu berechnen und eine Vorhersage des Schadigungsablaufes zu ermoglichen.

The effects of shot peening and deep rolling on the surface layer and the mechanical properties of sinterer iron

Abstract: The surface states and the mechanical behaviour of sintered iron (ASC 100.29) with different initial densities were studied after shot peening and deep rolling. This included the investigation of topography, depth profiles of microhardness, density and residual stresses and the determination of the behaviour at tensile, impact and alternating bending fatigue loading. The influence of the parameters used during the mechanical surface treatments on these properties will be discussed. SUBJECT INDEX Sintered Iron, Densification, Alternating Bending Fatigue INTRODUCTION The increase of the fatigue properties of powder-metallurgical components is limited in the case of conventional sintering treatments. Therefore several enhanced but very expensive sintering strategies were developed, as it is well accepted that the fatigue strength of sintered materials mainly depends on alloying and on density [1,2]. In [3] and [4] it is shown that the size and the number of pores strongly lower the material's fatigue strength, since the true load bearing cross-sectional area is reduced. However the local stress concentrations induced by the notch effects of pores close to the surface are more important for fatigue behaviour as they promote the detrimental initiation of cracks [5]. Thus, for materials states with relatively low density, mechanical surface treatments may be an economically fine alternative due to densification effects in the surface layer, which usually will be the highestly loaded area of a component. Therefore not only the work hardening and residual stress state after mechanical surface treatments like shot peening or deep rolling but also local densification effects are of major interest. The separation of the effects of these changes in the surface state imposed by those densification treatments on the behaviour at fatigue loading has not yet been investigated systematically. In this paper results are shown for the peened and deep rolled flat samples made from unalloyed sintered iron (ASC 100.29) which have two different initial densities. The characteristics of near surface regions as well as the fatigue strength will be compared and evaluated for both mechanically surface treated variants. MATERIAL AND SPECIMEN GEOMETRY AND EXPERIMENTAL PROCEDURE The investigations were performed using unalloyed sintered iron specimens (Hoganas ASC 100.29) and compact Armco-iron for comparison. Two sets of flat hour-glass specimens (90x1 1x5 mm) were pressed to nominal densities of 6.9 g/cm3 and 7.2 g/cm3, resp.. They were sintered for 30 minutes at 1150 "C in an atmosphere of HZ (10%)-NH3 (10%). The compact Armco iron (7.87 g/cm3) was subjected to a grain coarsening annealing at 600" C in N2 atmosphere in a fluidized AI2O3 air bed for two hours and cooled down in air in order to adjust the grain size to that of the porous 242 ICSP9 : SHOT PEENING material of 35 to 40 pm. The chemical composition in weight-% of the sintered and the Armco-iron is presented in Tab. 1. Tab. 1: Chemical composition (wt.-%) of the materials states investigated The shot peening (SP) treatments were performed using an air blast machine at room temperature. Cast iron shot S170 56 HRC was used at peening pressures of 1.6, 4 and 8 bar with a media flow rate of 1.6 kglmin and a resultant plain coverage of the sample surface. The Almen intensities were measured to be 0.33, 0.50 and 0.66 mmA, resp.. The deep rolling (DR) treatments were carried out using a so-called ball-point tool with a hydrostatic spherical carbide element (0 6.35 mm), which allows deep rolling of free formed surfaces of non-rotational symmetric specimens at a constant pressure. Plain coverage was attained by meandering the surface with a 0.04 mm spacing between each track in such a manner that maximum compressive residual stresses are generated perpendicular to the specimen's axis. Hence the pressures were varied between 55 and 380 bar and feed rates were adjusted to 1000 mmlmin so that the hydrostatic ball conveyed loads ranging from 137 to 950 N. The porosity of shot peened specimens was measured using computer aided image analysis of optical micrographs. The S-N-curves were determined by performing amplitude controlled alternating bending tests at a frequency of 25 Hz using 25 to 30 specimens. The results were analysed applying the arcsin JP method [ 6 ] Residual stresses were measured using CrKa-radiation at the (21 1)-interference lines of the ferritic phase at 13 angles between -60" and +60° according to the sin2q-method [7], using a Young's modulus of 210000 ~ l m m ' and a Poisson's ratio of v = 0.28. Subsequent X-ray measurements after electrolytic removal of thin surface layers enabled the determination of the depth distribution of the residual stresses. The mechanical elastic properties of porous materials are governed by the shape and concentration of the pores [8]. Therefore the residual stress values were corrected by a method described by 181, applying the results of the pore depth profiles. To compensate the effects imposed by the material removal, the results finally were corrected according to the method of [9]. Cr 0.018 0.026 0.018 [g/cm3] 6.9 7.2 , 7.87 RESULTS Shot peening and deep rolling of the sintered iron for both nominal densities 6.9 and 7.2 glcm3, resulted in significant differences in the surface roughness, Rz. Shot peening of sintered iron results in a linear increase of the roughness Rz with increasing pressure varying from 1.6 to 8 bar. The Rz values in this pressure range, measured by a confocal white light microscope, are more than doubled from 19 and 14 pm in the inital state to 47 and 42 pm, resp.. On the other hand deep rolling at pressures of 55 and 100 bar yields roughness values of 0.60 and 0.55 pm, resp., which are smaller by more than one order of magnitude compared to the untreated state. At a pressure of 180 bar the onset of slight pitting at the specimens surface can be observed, which increases Rz to 1.70 pm. At a maximum pressure of 380 bar the measured roughness of 13.70 pm almost equals that observed at the initial state. It can be inferred that shot peening increases the Rz value while deep rolling decreases it markedly. C 0.015 0.049 0.010 Ni

Modelling the stress state of a thermal barrier coating system at high temperatures

Abstract: A model is proposed to assess the time- and temperature-dependent creep and oxidation behavior of typical thermal barrier coating (TBC) systems consisting of EB-PVD manufactured partially stabilized ZrO2 and a plasma-sprayed MCrAlY bond coat (BC) on a typical Ni-base sheet alloy. The creep behavior of the BC was investigated by compression tests at the NiCoCrAlY alloy PWA 1370. With the objective of describing the formation of the so-called thermally grown oxide, an oxidation kinetic for the BC alloy is presented, which predicts the oxide thickness as a function of time and temperature. Using a finite element model of the complete TBC system, the influence of time- and temperature-dependent oxide growth and deformation behavior of the components on the stress development was studied. The obtained results show a development of compressive stresses above the peaks and of tensile stresses in the valleys of the rough TBC/BC interface. With this, the delamination stress between top coat and BC, which ...