5.5 Materialmodellierung
Filtern
Dokumenttyp
- Vortrag (32)
- Zeitschriftenartikel (29)
- Beitrag zu einem Tagungsband (9)
- Posterpräsentation (3)
- Buchkapitel (2)
- Forschungsdatensatz (2)
- Dissertation (1)
- Forschungsbericht (1)
Sprache
- Englisch (67)
- Deutsch (11)
- Französisch (1)
Schlagworte
- Creep (9)
- Nickel-base superalloys (8)
- Fatigue (7)
- LCF (7)
- Thermomechanical fatigue (7)
- Additive manufacturing (5)
- Anisotropy (5)
- Crack growth (5)
- Crystal plasticity (5)
- HIP (5)
- Data Fusion (4)
- Microstrucrue Design (4)
- Scale-bridging (4)
- Virtual experiments (4)
- Crack (3)
- Crack propagation (3)
- Crack tip opening displacement (3)
- EBSD (3)
- Ferritic-martensitic steels (3)
- Phase-field (3)
- Residual stress (3)
- Single-crystals (3)
- Superalloy (3)
- Additive Fertigung (2)
- CALPHAD (2)
- Concrete (2)
- Continnum damage model (2)
- Cyclic softening (2)
- Damage (2)
- Density-based Thermodynamics (2)
- Density-based model (2)
- Densty-based Thermodynamics (2)
- Diffusion (2)
- Dislocation (2)
- Dislocations (2)
- Elasticity (2)
- Environmentally assisted cracking (2)
- Finite element analysis (2)
- Fracture mechanics (2)
- Kriechen (2)
- Low cycle fatigue (2)
- Mass transport (2)
- Micromechanical model (2)
- Microstructure (2)
- Notches (2)
- Ontology (2)
- P92 (2)
- Parametric modeling (2)
- Phase-field simulation (2)
- Simulation (2)
- Simulation of concrete (2)
- Single-Crystal (2)
- Superalloys (2)
- Superlegierung (2)
- Symmetric dwell periods (2)
- Thermodynamics (2)
- Under cyclic loading (2)
- Viscoplasticity (2)
- 9-12% Cr ferritic-martensitic steels (1)
- 9-12%Cr steel (1)
- 9–12%Cr steel (1)
- AGIL (1)
- AM (1)
- Abgasturbolader Heißteile (1)
- Accelerated integration scheme (1)
- Accelerated temporal integration (1)
- Additive Manufacturing (1)
- Ageing (1)
- Aging (1)
- Al alloys (1)
- Al-Cu alloy (1)
- Alloy design (1)
- Alloy microstructure (1)
- Austenitic cast iron (1)
- Austenitischer Stahl (1)
- Automated finite element analysis (1)
- Automatische Finite Element Simulation (1)
- CMSX4 (1)
- Characterization (1)
- Chemo-mechanical coupling (1)
- Contact mechanics (1)
- Craze-crack mechanism (1)
- Creep-fatigue (1)
- Crystal Defects (1)
- Crystal Plasticity Modelling (1)
- Cyclic loading (1)
- Cyclic oxidation (1)
- Data Management (1)
- Data Structures (1)
- Data structure (1)
- Datenfusion (1)
- Defects engineering (1)
- Defects phase diagram (1)
- Defects thermodynamics (1)
- Deformation mechanisms (1)
- Differential scanning calorimetry (DSC) (1)
- Digitalisierung (1)
- Digitalization (1)
- Digitization (1)
- Eigenspannung (1)
- Elastic constants (1)
- Elastic energy (1)
- Electron microscopy (1)
- Entwicklung (1)
- Environmental Stress Cracking (1)
- Environmental stress cracking (ESC) (1)
- FNCT (1)
- Fatigue crack growth (1)
- Ferritic–martensitic steel (1)
- Finite-Elemente-Methode (1)
- Fourier series (1)
- Full notch creep test (FNCT) (1)
- Full-Notch Creep Test (1)
- Full-notch creep test (1)
- Fullerite (1)
- Grade S960QL steel (1)
- Gradient-enhanced fatigue model (1)
- Grain Boundary Phase Diagram (1)
- Grain Boundary Segregation (1)
- Grain Boundary Spinodal (1)
- Grain boundaries (1)
- Grain boundary (1)
- Grain boundary phase diagram (1)
- Grain boundary thermodynamics (1)
- Hertzian cracks (1)
- High-Entropy Alloys (1)
- High-entropy alloys (1)
- Hochtemperaturermüdung (1)
- Hot isostatic pressing (HIP) (1)
- Incremental lifetime models (1)
- Interfacial Spinodal (1)
- Interfacial anisotropy (1)
- Isostatic hot pressing (HIP) (1)
- Knowledge Representation (1)
- L-PBF 316L (1)
- Laser beam melting (LBM) (1)
- Laser powder bed fusion (1)
- Laser scanning microscopy (LSM) (1)
- Long-term aging (1)
- Low strain (1)
- Machine learning (1)
- Material digital (1)
- Material modeling (1)
- Mechanical anisotropy (1)
- Mechanical behavior (1)
- Mechanical testing (1)
- Mechanische Eigenschaften (1)
- Mechanistic Modelling (1)
- Mehrskalenmodell (1)
- Microstructural evolution (1)
- Microstructure characterisation (1)
- Microstructure evolution (1)
- Microstructure modification (1)
- Mikrostruktur (1)
- Misfitting precipitate (1)
- Model (1)
- Modeling (1)
- Modell (1)
- Nanocrystalline alloys (1)
- Nanoindentation (1)
- Ni-Resist (1)
- Nickel alloys (1)
- Nickel-base alloy (1)
- Nickel-base superalloy (1)
- Non-destructive testing (1)
- Numerische Simulation (1)
- Ontologie (1)
- Ordering (1)
- Ostwald ripening (1)
- P92 steels (1)
- PE-HD (1)
- Phase Diagram (1)
- Phase field model (1)
- Phase stability (1)
- Plasticity (1)
- Polyethylene (1)
- Pore (1)
- Pores (1)
- Porosity (1)
- Power plant (1)
- Precipitate shape (1)
- Precipitation (1)
- Rafting (1)
- Relaxation fatigue (1)
- Rissausbreitung (1)
- Risse (1)
- Rissform (1)
- Rissverlauf (1)
- S-phase (1)
- S355 steel sheet (1)
- Selective laser melting (SLM) (1)
- Semantic Web Technologies (1)
- Semantisches Web (1)
- Single crystal superalloys (1)
- Single-crystal (1)
- Slow crack growth (SCG) (1)
- Spinodal Decomposition (1)
- Spinodal decomposition (1)
- Strain energy (1)
- TEM (1)
- TMF (1)
- Tempered Martensite Ferritic Steels (1)
- Tempered martensite ferritic steels (1)
- Tensile test (1)
- Tensile testing (1)
- Thermo-Mechanical Fatigue (1)
- Thermo-mechanical fatigue (1)
- Thermographie (1)
- Thermomechanische Ermüdung (1)
- Transmission electron microscopy (TEM) (1)
- Vacancies (1)
- Viskoplastisch (1)
- Wissensrepräsentation (1)
- X-ray computed tomography (CT) (1)
- Zugversuch (1)
- γ″ phase (1)
- θ′ phase (1)
- θ′-(Al2Cu) precipitate phase (1)
Organisationseinheit der BAM
- 5.2 Metallische Hochtemperaturwerkstoffe (79) (entfernen)
Eingeladener Vortrag
- nein (32)
Modelling of environmentally assisted material degradation in the crack phase-field framework
(2019)
The simulation of crack propagation was conducted with a diffusive crack model in a variational framework. Moreover, the physically sound introduction of mass transport and coupling mechanisms due to environmentally assisted effects could be realised in this framework. The objective consists of the application of the phase-field Approach towards the simulation of environmentally assisted material degradation with the advantage of a non-required predefined crack path and a mesh-independent non-local formulation that facilitates the damage evolution with respect to material softening. The sharp crack is regularised by the introduction of a phase-field order parameter leading to a diffusive crack formulation.
Besides the equations originating from the linear momentum balance an additional evolution equation for the crack phase-field is introduced. Furthermore, mass transport is simulated by a Diffusion equation. The description delivered by the variational phase-field framework is able to simulate crack propagation according to published numerical test cases.
Additionally, the calculation of stress intensity factors is possible as well as crack resistance curves that describe stable crack propagation.
The current competitive situation on electricity markets forces power plants into cyclic operation regimes with frequent load shifts and starts/shutdowns. In the present work, the cyclic mechanical behavior of ferritic-martensitic 9-12 % Cr steels under isothermal and thermomechanical loading was investigated for the example of grade P92 material. A continuous softening was observed under all loading conditions. The introduction of hold periods to the applied cycles reduced material lifetime, with most prominent effects at technologically relevant small strain levels. The microstructural characterization reveals a coarsening of the original “martensitic” lath-type microstructure to a structure with polygonal subgrains and reduced dislocation density. The microstructural data forms the input for a physically-based modelling approach.
9-12% Cr ferritic-martensitic stainless steels are widely used as high temperature construction materials in power plants due to their excellent creep and oxidation resistance. The growing share of renewable energy sources in power generation forces many of these plants into more flexible operation with frequent load shifts or shutdowns. These cyclic operation profiles constitute a major lifetime issue.
The present contribution reports on current findings obtained in a multidisciplinary project which combines cyclic mechanical and cyclic oxidation testing with detailed microstructural analyses. Mechanical analyses are carried out on P92 and P91 steel grades to give an overview of softening phenomena and lifetimes obtained in isothermal cyclic loading (low cycle fatigue, LCF), non-isothermal cyclic loading (thermo-mechanical fatigue, TMF), and service-like combinations of creep and fatigue periods (creep-fatigue interaction).
Oxidation testing focuses on the grades P92 and VM12 with the intention of clarifying the impact of frequent passes through intermediate temperature levels on the kinetics of steam-side oxidation and the characteristics of the evolving oxide scales. An attempt is made to evaluate their composition, strength, integrity and adhesion after up to 250 temperature cycles. Flat coupons as well as curved tube sections are tested to assess the mutual influence of geometry on oxide scale integrity.
Complementary microstructural investigations by scanning and transmission electron microscopy plus EBSD are used for phase identification and substrate/oxide interface characterisation. The evolutions of grain size and dislocation density under different test conditions are quantified.
9-12% Cr ferritic-martensitic stainless steels are widely used as high temperature construction materials in power plants due to their excellent creep and oxidation resistance. The growing share of renewable energy sources in power generation forces many of these plants into more flexible operation with frequent load shifts or shutdowns. These cyclic operation profiles constitute a major lifetime issue. The present contribution reports on current findings obtained in a multidisciplinary project which combines cyclic mechanical and cyclic oxidation testing of different 9-12% Cr grades with detailed microstructural analyses.
Mechanical analyses are carried out on P92 and P91 steel grades to give an overview of softening phenomena and lifetimes obtained in isothermal cyclic loading (low cycle fatigue, LCF), non-isothermal cyclic loading (thermo-mechanical fatigue, TMF), and service-like combinations of creep and fatigue periods.
Complementary microstructural investigations by scanning and transmission electron microscopy plus EBSD are used for phase identification, substrate/oxide interface characterization and quantification of the microstructure evolution under cyclic conditions.
The employment of renewable energy resources, which are naturally intermittent, for electricity generation has altered the working conditions of conventional power plants from continuous (baseload) to cyclic or flexible operation. For a long time, 9-12% Cr ferritic-martensitic stainless steels have been widely used in power plants due to their favorable characteristics such as high creep strength at high temperatures and oxidation and corrosion resistance. The components of power plants are subjected to long term cyclic loadings including fatigue and creep-fatigue at high temperatures. As ferritic-martensitic steels are known to exhibit cyclic softening when subjected to such loading scenarios, it is crucial to study the material response in such conditions. Since it is impossible to test the material behavior exactly as the operation conditions of power plants, due to technical difficulties and cost issues, it is necessary to develop physically-based material models that can predict the material behavior in more realistic situations.
In recent years, many material models have been proposed to describe the behavior of 9-12% Cr ferritic-martensitic stainless steels, which follow phenomenological or physically-based approaches. Phenomenological models provide a stress-strain relation based on empirical observations although they usually lack physical background. To alleviate this drawback and to allow for more flexibility and wider ranges of strain-rate and temperature, physically-based models are suggested. In this approach, microstructural evolution, dislocation movement and/or kinetics of plastic deformation processes are included. The physically-based models allow for a better extrapolation from the experiment results to other operation conditions and their material constants can be interpreted physically. Compared to the former approach, the material behavior can be described more accurately and flexibly and the number of material constants is less in general. In the presented work, the existing micromechanical models developed for P92 steel are compared and extended to allow for new dislocation-based strengthening/cyclic mechanisms. Their performance is assessed in the light of mechanical test data from creep-fatigue and thermo-mechanical fatigue experiments and detailed characterizations of the microstructure evolution in the fatigued material.
Es konnte anhand einer numerischen Voruntersuchung gezeigt werden, dass anhand der kombinierten Auswertung der im Versuch verwendeten Sensorik eine Einteilung der unter Ermüdung in Bohrlochproben auftretenden Rissformen in verschiedene Hauptkategorien (Eckriss, Oberflächenriss, Durchgangsriss) möglich ist.
Es wird eine neu entwickelte Methode zur Thermographiebasierten Rissmessung vorgestellt. Darüber hinaus wird eine numerische Vorarbeit präsentiert, die zeigt, dass anhand der gemeisamen Auswertung der Versuchsdaten aus unterschiedlicher Sensorik die Möglichkeit besteht, die unter Ermüdungsbelastung in Bohrlochproben auftretenden Risse in Geometriekategorien zu unterteilen.
The methods of computational damage mechanics are well-established for the description of degradation of materials under monotone loading. An extension to structural damage induced by cyclic loading is however significantly limited. This is due to enormous computational costs required to resolve each load cycle by conventional temporal incremental integration schemes while a typical fatigue loading history comprises between thousands and millions of cycles. Despite the permanent increase of computational resources and algorithmic performance, a successful approach is rather based on the development of novel multiscale in time integration schemes.
A Fourier transformation-based temporal integration (FTTI) is represented, which takes advantage of temporal scale separation incorporated into the cycle jump method. The response fields are approximated by a Fourier series whose coefficients undergo the evolution on a long-time scale. This is correlated with the evolution of the history variables, including damage, by means of the adaptive cycle jump method of various orders. The necessary extrapolation rates are obtained from the underlying solution of a short-time scale problem, which results from the oscillatory boundary condition and fulfills the global equilibrium of the Fourier coefficients. In this way, a remarkable speedup is achieved because the number of cycles to be fully integrated dramatically decreases.
The key idea behind the FTTI method is that the global in space equilibrium problem is linear since it is decoupled from the evolution equations. The latter are solved in the quadrature points under response fields prescribed throughout the whole load cycle. Consequently, integration of a single load cycle is much more efficient than the conventional single scale integration where the global equilibrium iteration and the local iteration of the evolution equations are coupled. This results in an additional speedup of the FTTI method.
The performance of the FTTI technique is demonstrated for two different constitutive behaviors: a viscoplastic model with a damage variable governed by the local equivalent viscoplastic strain; a quasi-brittle response where the damage variable is driven by a non-local equivalent strain. The latter is implicitly introduced as proposed by Peerlings. Both, the explicit and implicit extrapolation schemes are validated. The FTTI solutions agree very well with the reference cycle-by -cycle solutions, while significantly reducing the computational costs. The adaptive determination of the jump length can properly recognize the particular responses throughout the fatigue loading history (stationary fatigue, acceleration of fatigue damage when approaching failure) as well as stress redistribution phenomena.
A constitutive model for the mechanical behaviour of single crystal Ni-base superalloys under high temperature conditions has been developed in the framework of a Cooretec project in cooperation with Siemens AG, MTU Aero Engines AG and University Bayreuth. In addition to the conventional material properties e.g. elastic constants, the model requires the parameters of the initial microstructure as an input. Thus, the γ’-precipitate size and the channel width of the γ-matrix were obtained from SEM micrographs. The model uses the slip system theory and describes the movement, multiplication and annihilation of dislocations in the channels. Furthermore, the cutting of precipitates is another mechanism contributing to the plastic flow. The evolution of the morphology due to rafting and its effects on the deformation have been implemented according to. The kinematic hardening is introduced as a stress tensor to realistically represent the strain hardening of arbitrary oriented single crystals.
The mechanical behaviour of single crystal specimens has been experimentally investigated in tension tests at different strain rates and in creep tests under various loads. The constitutive model has been calibrated based on the experimental data for temperatures of 950°C and 850°C and the [001] and [111] crystallographic orientations.
Finally, a micromechanical model was created to simulate the creep response of additive manufactured polycrystalline structures. An EBSD image is taken to obtain the grain geometry and their respective orientation. The grain boundaries are discretised using cohesive elements, whereas the single crystal model was applied to each grain in the representative volume. The polycrystal model is generated using Dream3D, NetGen and other software previously developed at the BAM.
Virtual-lab-based determination of a macroscopic yield function for additively manufactured parts
(2018)
This work presents a method for the yield function determination of additively manufactured parts of S316L steel. A crystal plasticity model is calibrated with test results and used afterwards to perform so-called virtual experiments, that account for the specific process-related microstructure including crystallographic and morphological textures. These simulations are undertaken on a representative volume element (RVE), that is generated from EBSD/CT-Scans on in-house additively manufactured specimen, considering grain structure and crystal orientations. The results of the virtual experiments are used to determine an anisotropic Barlat yield function, that can be used in a macroscopical continuum-sense afterwards. This scale-bridging approach enables the calculation of large-scale parts, that would be numerically too expensive to be simulated by a crystal plasticity model.
Virtual-lab-based determination of a macroscopic yield function for additively manufactured parts
(2018)
This work presents a method for the yield function determination of additively manufactured parts of S316L steel. A crystal plasticity model is calibrated with test results and used afterwards to perform so-called virtual experiments, that account for the specific process-related microstructure including crystallographic and morphological textures. These simulations are undertaken on a representative volume element (RVE), that is generated from EBSD/CT-Scans on in-house additively manufactured specimen, considering grain structure and crystal orientations. The results of the virtual experiments are used to determine an anisotropic Barlat yield function, that can be used in a macroscopical continuum-sense afterwards. This scale-bridging approach enables the calculation of large-scale parts, that would be numerically too expensive to be simulated by a crystal plasticity model.
The sound numerical prediction of welding-induced thermal stresses, residual stresses, and distortions strongly depends on the accurate description of a welded material’s thermomechanical deformation behaviour. In this work, we provide experimental data on the viscoplastic deformation behaviour of a grade-s960ql steel up to a temperature of 1000 ◦C. In addition, a multi-phase viscoplastic material model is proposed, which accounts for the experimentally observed isothermal deformation behaviour of grade-s960ql steel base and austenitised material, as well as for athermal contributions that originate from solid-state phase transformations. The multi-phase viscoplastic and a classic rateindependent isotropic hardening material model were applied in the numerical simulations of both-ends-fixed bar Satoh tests and a single-pass gas metal arc weld. The influence of material modelling choices on the agreement between numerical simulation and experimental results is discussed, and recommendations for further work are given.
With the introduction of a mass transport mechanism the entire problem is subjected to a time frame that dictates the time-dependent action of soluted species on mechanical properties. A numerical framework within the phase-field approach is presented with an embrittlement-based coupling mechanism. The underlying functionals are expressed in terms of the displacement, mass concentration and crack phase-field. Within the phase-field approach the modelling of sharp crack discontinuities is replaced by a diffusive crack model facilitating crack initiation and complex crack topologies without the requirement of a predefined crack path. The isotropic hardening of the elasto-plastic deformation model and the local fracture criterion are affected by the species concentration. This allows for embrittlement and leads to an accelerated crack propagation. An extended mass transport equation for hydrogen embrittlement, accounting for mechanical stresses and deformations, is implemented. For stabilisation purposes a staggered scheme is applied to solve the system of partial differential equations by a multi-field finite-element method. A thermodynamically consistent coupling relation that accommodates the required mechanisms is presented.
The creep behavior of single-crystals of the nickel-base superalloy CMSX-4 was investigated at 1288°C, which is the temperature of the hot isostatic pressing (HIP) treatment applied to this superalloy in the industry. It was found that at this super-solvus temperature, where no Gamma’-strengthening occurs, the superalloy is very soft and rapidly deforms under stresses between 4 and 16 MPa. The creep resistance was found to be very anisotropic, e.g. the creep rate of [001] crystals was about 11 times higher than that of a [111] crystal. The specimens of different orientations also showed a very different necking behavior. The reduction of the cross-section area psi of [001] crystals reached nearly 100%, while for a [111] crystal psi=62%. The EBSD analysis of deformed specimens showed that despite such a large local strain the [001] crystals didn’t not recrystallize, while a less deformed [111] crystal totally recrystallized within the necking zone. From the shape of deformed specimens and TEM investigations it was concluded that the main strain contribution resulted from <011> {111} octahedral slip.
A numerical pre-study has shown that cracks in a flat sample featuring a drilled hole can be classified into one of three crack shape classes based on the combined evaluation of various types of test data.
The creep behavior of single crystals of the nickel-based superalloy CMSX-4 was investigated at 1288 °C, which is the temperature of the hot isostatic pressing treatment applied to this superalloy in the industry. It was found that at this super-solvus temperature, where no gammaPrime-strengthening occurs, the superalloy is very soft and rapidly deforms under stresses between 4 and 16 MPa. The creep resistance was found to be very anisotropic, e.g., the creep rate of [001] crystals was about 11 times higher than that of a [111] crystal. The specimens of different orientations also showed a very different necking behavior. The reduction of the cross-sectional area psi of [001] crystals reached nearly 100 pct, while for a [111] crystal psi = 62 pct. The EBSD analysis of deformed specimens showed that despite such a large local strain the [001] crystals did not recrystallize, while a less deformed [111] crystal totally recrystallized within the necking zone. The recrystallization degree was found to be correlated with deformation behavior as well as with dwell time at high temperature. From the analysis of the obtained results (creep anisotropy, stress dependence of the creep rate, traces of shear deformation, and TEM observations), it was concluded that the main strain contribution resulted from <01-1>{111} octahedral slip.
Les aubes de turbines à gaz utilisées en particulier pour les turboréacteurs de l’aéronautique sont élaborées par fonderie en superalliage monocristallin à base de nickel. Le procédé de fonderie, ainsi que les traitements thermique d’homogénéisation réalisés à très haute température, induisent la présence de pores au sein des pièces qui affectent les propriétés mécaniques et la durée de vie des aubes. Afin de réduire cette porosité les motoristes effectuent un traitement de compression isostatique à chaud (CIC) au cours duquel la porosité diminue par fermeture des pores. Afin de mieux comprendre les mécanismes impliqués au cours du traitement de CIC, nous avons lancé un programme de recherche dans le cadre du projet ERA-Net MICROPORE. La modélisation par champ de phase des mécanismes en jeu est présentée au cours de ce colloque. Nous présentons dans cette affiche un des volets de la caractérisation expérimentale du projet.
Des échantillons de superalliage CMSX4 sont observés après traitement de mise en solution et CIC sous 103 MPa à 1288°C pour différentes durées. Les pores présents sont caractérisés par microscopie électronique à balayage (MEB) afin de suivre l’évolution du taux de porosité au cours du traitement. Une caractérisation plus détaillée de pores partiellement refermés est menée par MEB et grâce à la diffraction des électrons rétrodiffusés (EBSD). Une vision tridimensionnelle de ces défauts est obtenue par des coupes métallographiques effectuées par découpe ionique (FIB).
Le projet ERA – Net MICROPORE est financé en Allemagne par la DFG (projects EP 136/1-1 and FE933/2-1) et en France par l’ANR (projects ANR15-MERA-000-03 and ANR15-MERA-0003-04).
A key limitation of the most constitutive models that reproduce a Degradation of quasi-brittle materials is that they generally do not address issues related to fatigue. One reason is the huge computational costs to resolve each load cycle on the structural level. The goal of this paper is the development of a temporal Integration scheme, which significantly increases the computational efficiency of the finite element method in comparison to conventional temporal integrations.
The essential constituent of the fatigue model is an implicit gradient-enhanced formulation of the damage rate. The evolution of the field variables is computed as amultiscale Fourier series in time.On a microchronological scale attributed to single cycles, the initial boundary value problem is approximated by linear BVPs with respect to the Fourier coefficients. Using the adaptive cycle jump concept, the obtained damage rates are transferred to a coarsermacrochronological scale associated with the duration of material deterioration. The performance of the developedmethod is hence improved due to an efficient numerical treatment of the microchronological problem in combination with the cycle jump technique on the macrochronological scale. Validation examples demonstrate the convergence of the obtained solutions to the reference simulations while significantly reducing the computational costs.
An improved diffusion model is proposed for pore annihilation during HIP of single-crystal nickel-base superalloys. The model assumes the pore dissolution by emission of vacancies and their sink to the low angle boundaries. Calculation, considering distribution of the pore sizes, predicts the kinetics of pore annihilation similar to the experimental one.