Strain–life curves of steels and methods for determining the curve parameters. Part 2. Methods based on the use of artificial neural networks

The authors look into the possibility of using artificial neural networks for predicting the deformation characteristics of steels (the parameters of the Basquin–Manson–Coffin strain–life curve equation) based on static strength and plasticity characteristics, by constructing four independent neural networks with different configurations of input and output data. The prediction of parameters of the Basquin–Manson–Coffin equation and the fatigue life calculations by means of artificial neural networks are demonstrated to provide a better accuracy in comparison to the available conventional methods.     

Evaluation of three test methods for determining the alkali–silica reactivity of glass aggregate

Consumption of natural raw materials and pollution have become significant problems due to technological developments and continual increase in demand. Accordingly, great efforts are being made in order to recover wastes including glass. One of the possible applications is utilizing waste glass in concrete; however, alkali–silica reaction (ASR) is of major concern. In this study, tests were conducted by applying three different procedures: ASTM C1293, RILEM AAR-2, and microbar test methods. In microbar testing, glass aggregate was used as coarse aggregate, whereas the other two methods dealt with investigating the reactivity of the finer fraction of the waste glass. The effects of chemical composition, particle size and amount of glass in the mixture were studied. According to the results, flint glass expanded to a greater extent than amber and green glass. Expansions, within the specified time periods dictated by the methods, remained low; however, extended durations resulted in very high length change values of the flint glass-including mixtures, particularly in the AAR-2 and microbar tests.     

An iterative procedure for determining effective stress–strain curves of sheet metals

An iterative correction procedure using 3D finite element analysis (FEA) was carried out to determine more accurately the effective true stress–true strain curves of aluminum, copper, steel, and titanium sheet metals with various gage section geometries up to very large strains just prior to the final tearing fracture. Based on the local surface strain mapping measurements within the diffuse and localized necking region of a rectangular cross-section tension coupon in uniaxial tension using digital image correlation (DIC), both average axial true strain and the average axial stress without correction of the triaxiality of the stress state within the neck have been obtained experimentally. The measured stress–strain curve was then used as an initial guess of the effective true stress–strain curve in the finite element analysis. The input effective true stress–strain curve was corrected iteratively after each analysis session until the difference between the experimentally measured and FE-computed average axial true stress–true strain curves inside a neck becomes acceptably small. As each test coupon was analyzed by a full-scale finite element model and no specific analytical model of strain-hardening was assumed, the method used in this study is shown to be rather general and can be applied to sheet metals with various strain hardening behaviors and tension coupon geometries.     

Fatigue life estimation under multiaxial random loading by means of the equivalent Lemaitre stress and multiaxial S–N curve methods

In this paper, the frequency domain formula of equivalent Lemaitre stress taking into account the hydrostatic stress effect is first introduced, and the corresponding method for estimating fatigue life under multiaxial random loading is developed based on multiaxial S–N curve. The proposed method is systematically validated with the random bending-torsion fatigue tests and numerical simulations. It has been shown that the hydrostatic stress has a significant influence on multiaxial fatigue life; the results predicted by the proposed method agree well with the experiment, and are more accurate than those obtained for the equivalent von Mises stress method.     

Development of methods and instruments of the Russian service for determination of the Earth’s orbital parameters

The state and future outlook of research to determine the Earths orbital parameters, i.e., Universal Time and the coordinates of the pole, carried out by the State Time and Frequency Service and in calculations of the Earths orbital parameters are described.Translated from Izmeritelnaya Tekhnika, No. 1, pp. 24–27, January, 2005.     

Development of structure and effect of processing parameters on Strength–structure relationships for two ferritic stainless steels

Abstract

The development of substructure as material passes through the roll gap has been examined for a ferritie stainless steel and it is shown that the final structure evolves close to the roll gap exit. Temperature variations as the material was rolled were monitored and the subsequent effect on substructure investigated. The variation of substructure with processing for two ferritic steels is discussed and the effect of this variation on room temperature properties is presented. It is shown that the strength–structure relationship is highly dependent upon retained martensite which degrades the tensile strength.

MST/380     

Modeling the response of physical and mechanical properties of Cr–Mo prealloyed sintered steels to key manufacturing parameters

In this study an experimental investigation using response surface methodology has been undertaken in order to model and evaluate the physical and mechanical properties of Cr–Mo prealloyed sintered steels with respect to the variation of powder metallurgy process parameters such as compacting pressure, sintering temperature and Cr content of the prealloyed steel powder. Mathematical models were developed at 95% confidence level to predict the physical properties such as sintered density and electrical resistivity and mechanical properties such as transverse rupture strength, apparent (=macro-)hardness, and impact energy. Analysis of variance was used to validate the adequacy of the developed models. The obtained mathematical models are useful not only for predicting the physical and mechanical properties with higher accuracy but also for selecting optimum manufacturing parameters to achieve the desired properties.     

Mechanistic approach for prediction of creep deformation,damage and rupture life of different Cr–Mo ferritic steels

Abstract

A mechanistic approach based on finite element analysis of continuum damage as proposed by Kachanov has been used to assess and compare creep deformation, damage and rupture behaviour of 2·25Cr–1Mo, 9Cr–1Mo and modified 9Cr–1Mo ferritic steels. Creep tests were carried out on the steels at 873 K over a stress range of 90–230 MPa. Modified 9Cr–1Mo steel was found to have highest creep deformation and rupture strength whereas 2·25Cr–1Mo steel showed the lowest among the three ferritic steels. Creep damage in the steels has been manifested as the microstructural degradation. 2·25Cr–1Mo steel was more prone to creep damage than 9Cr–steels. Finite element estimation of creep deformation and rupture lives were found to be in good agreement with the experimental results.     

Modeling the brittle–ductile transition in ferritic steels. Part II: analysis of scatter in fracture toughness

A dislocation simulation model has been proposed to predict the brittle–ductile transition in ferritic steels in Part I. Here we extend the model to address the problem of inherent scatter in fracture toughness measurements. We carried out a series of Monte Carlo simulations using distributions of microcracks situated on the plane of a main macrocrack. Detailed statistical analysis of the simulation results showed the following: (a) fracture is initiated at one of the microcracks whose size is at the tail of the size distribution function, and (b) the inherent scatter arises from the distribution in the size of the critical microcrack that initiates the fracture and not from the variation of the location of the critical microcrack. Utilizing the weakest-link theory, Weibull analysis shows good agreement with the Weibull modulus values obtained from fracture toughness measurements.     

Comparison of ASME K IR curve and dynamic master curve in determining ductile–brittle transition temperature of A48P2 steel

Abstract

The ductile–brittle transition temperature (DBTT) of grade A48P2 steel is characterised based on the American Society of Mechanical Engineers (ASME) fracture toughness K _IR curve and dynamic master curve approaches. The indexing parameter for the K _IR curve, reference temperature RT_NDT, is determined from drop weight and Charpy tests to be ?45°C. The dynamic master curve is constructed following ASTM standard E1921 guidelines; however, instead of precracked tests, the dynamic fracture toughness K _Jd is determined from Charpy V notch tests using a modified Schindler's procedure. A Weibull plot is constructed using the K _Jd data, and it is found that the points comply reasonably with the forced fit line of slope 4. The reference temperature for constructing the dynamic master curve, termed T^dy,Sch₀, thus determined is ?56°C. The ASME K _IR curve is shown to be conservative compared with the dynamic master curve constructed using T^dy,Sch₀.     

Study of validity of the single-diode model for solar cells by I–V curves parameters extraction using a simple numerical method

The I–V measurement of solar cells is one of the most employed electrical characterization techniques in the photovoltaics research field due to the valuable information one can obtain from such a curve. Parameters like Voc, Isc and the maximum power can be easily observed at a glance. Furthermore, additional information can be extracted by a more exhaustive analysis involving the equivalent electrical circuit of a solar cell which is based on ideal electrical components with a well-defined behavior. This equivalent circuit is typically assumed to be formed by a current source in parallel with a single diode and a shunt resistor, connected to a series resistor. However, this model does not take into account the separate contributions of the different carrier transport mechanisms in solar cells; for example, carrier diffusion and recombination-generation in the depletion region. Therefore, the single diode model may not be suitable in many practical cases. In this work, a simple numerical method was developed to extract the parameters for both single diode and double diode models from experimental I–V curves of solar cells. The developed numerical algorithm was applied to extract the parameters for a published benchmark solar cell which has been used for testing this kind of algorithms. The extracted parameters using our simple method are comparable with other more sophisticated and computer power demanding algorithms. Thereafter, we applied the developed algorithm to extract the single-diode parameters from simulated “experimental” I–V curves, where two carrier transport mechanisms are present, trying to understand under what conditions the single diode approximation would provide meaningful parameters for such experimental curves. It is shown that the extracted parameters can vary strongly, particularly for the dark saturation current and ideality factor, without much variation of the root mean square error between the experimental data and the model, causing these values to be unreliable and its physical interpretation misleading. We show that the same algorithm can be applied to a double diode (two exponentials) model providing physically meaningful parameters without much computing power requirements. In summary, there is no further justification for using a single diode model to interpret the experimental I–V curves of real solar cells.     

Some metallurgical aspects of Dynamic Strain Aging effect on the Low Cycle Fatigue behavior of C–Mn steels

Carbon–Manganese steels and associated welds are commonly used, and sometimes to sustain loads in the Low Cycle Fatigue domain. Nevertheless, the metallurgy of these C–Mn steels is rather complex, due to the interaction of solute atoms (carbon and nitrogen) with dislocations during deformation which leads to metallurgical instabilities: Lüders strain, Static Strain Aging (SSA) and Dynamic Strain Aging (DSA). The DSA phenomenon is an interaction during the test between solute atoms and dislocations which are submitted to an supplementary anchorage if the temperature is sufficient to allow the diffusion of solute atoms leading to a discontinuous plastic deformation localized in bands associated with serrations on the stress–strain curve. In C–Mn, the temperature domain where the phenomenon is present is from 150 °C to 300 °C. If these metallurgical instabilities induce an increase in hardness, unfortunately they produce a decrease of ductility detrimental to components safety. The results of the DSA effect on LCF behavior in C–Mn and Low Alloyed steels reported in the literature are very confused and contradictories. In this study, two C–Mn steels with a different sensitivity to DSA are investigated in the Low Cycle fatigue domain. As reported from some authors, the fatigue life seems enhance or reduce in the temperature domain where the DSA is maximum, but the decrease of the strain rate always decreases the number of cycles to failure.     

Calibration of fracture mechanics parameters and J–R curve determination in polyethylene side-grooved arc-shaped specimens

The J–R curve is determined in arc-shaped samples directly obtained from medium density polyethylene gas pipes by the modified normalization method. The elastic and plastic factors (η_el and η_pl) used in the J-integral expression are geometry sensitive and their values for side-grooved arc-shaped samples in a three-point bending configuration are experimentally determined. The values of the elastic compliance and crack depth-to-width ratios are found to be related by the expression recommended in the ASTM E813-87 standard for the standard bend specimens. The η_pl factor is determined from the exponent of the power law that best fits the experimental data of separation parameter versus ligament area and is also close to the theoretical value for three-point bending geometry. The J–R curve is also predicted by applying the procedure proposed in the EPRI/GE handbook and compared with the experimental multiple-specimen results.     

Probabilistic fatigue S–N curves including the super-long life regime of a railway axle steel

A concurrent probability method is proposed for estimating probabilistic fatigue S–N curves including the super-long life regime. This work is based on experimental investigation of LZ50 railway axle steel. Test results reveal that fatigue cracks are initiated from the weakest surface phase and the damage preferentially follows a competition mechanism between the surface quality and subsurface inclusions. The curves are estimated by the test data in the mid-long life regime and the fatigue limits, which were connected together in concurrent probability levels. The accuracy of the curves was verified by the test data in the super-long life regime.     

Alloying effects on the microstructure and phase stability of Fe–Cr–Mn steels

Austenitic Fe–Cr–Mn stainless steels interstitially alloyed with nitrogen have received considerable interest lately, due to their many property improvements over conventional Fe–Cr–Ni alloys. The addition of nitrogen to Fe–Cr–Mn stabilizes the fcc structure and increases the carbon solubility. The benefits of increased interstitial nitrogen and carbon content include: enhanced strength, hardness, and wear resistance. This study examines the effect of carbon, silicon, molybdenum, and nickel additions on the phase stability and tensile behavior of nitrogen-containing Fe–Cr–Mn alloys. Nitrogen and carbon concentrations exceeding 2.0 wt.% were added to the base Fe–18Cr–18Mn composition without the formation of nitride or carbide precipitates. Minor additions of molybdenum, silicon, and nickel did not affect nitrogen interstitial solubility, but did reduce carbon solubility resulting in the formation of M₂₃C₆ (M=Cr, Fe, Mo) carbides. Increasing the interstitial content increases the lattice distortion strain, which is directly correlated with an increase in yield stress.     

Evolution of microstructure and strength during the ultra-fast tempering of Fe–Mn–C martensitic steels

Ultra-fast tempering cycles, combining a rapid heating rate (R _H = 300 °C/s) from room temperature to a peak temperature within the range 400–700 °C and subsequent rapid cooling, were performed on a Fe–Mn–C martensitic steel. The influence of the peak temperature reached during the cycle was determined on both the tensile properties of the steel and on its microstructure. The mechanisms controlling the microstructural evolution occurring during rapid tempering were studied by combining both TEM observations and 3D reconstructions by FIB/SEM. A theoretical analysis coupled with the acquired experimental data was then proposed to explain the evolution of mechanical properties. The results obtained support the assumption that carbide precipitation during fast tempering plays a key role in the evolution of mechanical properties.     

Mechanical properties of wood flake–polyethylene composites. Part I: effects of processing methods and matrix melt flow behaviour

The structure–property relationship of wood flake–high-density polyethylene (HDPE) composites was studied in relation to the matrix agent melt flow behaviour and processing technique. The flake distribution and flake wetting were optimised to obtain acceptable mechanical properties in these composites using two processing techniques, namely twin-screw compounding and mechanical blending. The microstructure of the composites revealed that the twin-screw compounded composites based on medium melt flow index (MMFI) HDPE always achieved better flake wetting and distribution, and therefore had higher mechanical properties, than those mechanically blended composites or twin-screw compounded composites with low MFI (LMFI) HDPE. For 50:50 wt% composites the overall flake wetting, depending on processing technique and matrix flow behaviour, is ranked as compounded MMFI>compounded LMFI>blended MMFI>blended LMFI. However, the uniformity of flake distribution of the composites follows a somewhat different pattern, i.e. compounded MMFI>blended MMFI>compounded LMFI>blended LMFI. Evidence shows that the medium MFI HDPE penetrates into lumens of wood fibres in wood flakes. This phenomenon combined with flake wetting and flake distribution had a profound effect on the mechanical properties, in particular the impact strength.