The rolling process is widely used in the metal forming industry and has been so for many years. However, the process has attracted renewed interest as it recently has been adapted to very small scales where conventional plasticity theory cannot accurately predict the material response. It is well-established that gradient effects play a role at the micron scale, and the objective of this study is to demonstrate how strain gradient hardening affects the rolling process. Specifically, the paper addresses how the applied roll torque, roll forces, and the contact conditions are modified by strain gradient plasticity. Metals are known to be stronger when large strain gradients appear over a few microns; hence, the forces involved in the rolling process are expected to increase relatively at these smaller scales. In the present numerical analysis, a steady-state modeling technique that enables convergence without dealing with the transient response period is employed. This allows for a comprehensive parameter study. Coulomb friction, including a stick–slip condition, is used as a first approximation. It is found that length scale effects increase both the forces applied to the roll, the roll torque, and thus the power input to the process. The contact traction is also affected, particularly for sheet thicknesses on the order of 10 μ m and below. The influences of the length parameter and the friction coefficient are emphasized, and the results are presented for multiple sheet reductions and roll sizes.

Alumina ceramics, due to their excellent properties of high hardness, corrosion resistance, and low thermal expansion coefficient, are important industrial materials with a wide range of applications, but these materials also present difficulty in machining with low material removal rates and high tool wear. This study is concerned with laser-assisted machining (LAM) of high weight percentage of alumina ceramics to improve the machinability by a single point cutting tool while reducing the cutting forces. A multiscale model is developed for simulating the machining of alumina ceramics. In the polycrystalline form, the properties of alumina ceramics are strongly related to the glass interface existing in their microstructure, particularly at high temperatures. The interface is characterized by a cohesive zone model (CZM) with the traction–separation law while the alumina grains are modeled as continuum elements with isotropic properties separated by the interface. Bulk deformation and brittle failure are considered for the alumina grains. Molecular dynamics (MD) simulations are carried out to obtain the atomistic structures and parameterize traction–separation laws for the interfaces of different compositions of alumina ceramics at high temperatures. The generated parameterized traction–separation laws are then incorporated into a finite element model in Abaqus to simulate the intergranular cracks. For validation purposes, simulated results of the multiscale approach are compared with the experimental measurements of the cutting forces. The model is successful in predicting cutting forces with respect to the different weight percentage and composition of alumina ceramics at high temperatures in LAM processes.

An analytical expression is developed for the relationship between far-field acoustic emission (AE) signal and the propagation of a finite plane crack due to a nonuniform residual stress in a weldment. The AE sensor is situated at the boundary parallel to the crack surface. The quasi-static crack opening displacement (COD) rate is used as the source function. The AE initial response on a traction-free surface is obtained as a function of instantaneous crack length and speed. A closed-form expression is obtained for the normal surface displacement at the epicenter when the crack propagates at uniform speed for a finite duration between initiation and arrest. Results from an earlier analysis of the (variable) speed of crack propagation in a weldment are used to develop a more exact numerical prediction of the surface motion. The effects of initial crack length and bluntness on AE surface motion are investigated in both time and frequency domains.

Numerical results of plane-strain strip drawing are presented emphasizing the interfacial friction conditions between workmaterial and drawing die. The results are based on a large-strain elasto-plastic finite element code with a special boundary condition capability to treat interface friction. Two different modeling approaches are developed. In the interface velocity model, finite element solutions are based on the velocity boundary conditions along the die-work interface measured through transparent sapphire dies. In the interface traction model, a technique called “General Linear Boundary Condition,” is developed to incorporate many different existing friction models. Predicted values of drawing force, die separating force, global friction, and residual stresses from both models are favorably compared with the corresponding experimental measurements.

A systematic laboratory investigation, of the effect of sanding on wheel/rail wear and adhesion, is reported. The tests were conducted on the IIT-GMEMD Wheel Rail Simulation Facility, using Hertzian simulation of the conditions for a typical North American locomotive wheel. Experiments at adhesion coefficients of 0.24, 0.35 and 0.5, with and without sand application on a clean dry rail, show that ratios of terminal and average wear rates, for the two categories of tests, range approximately from 10 to 100. Archard wear coefficients for the two categories are in the range 10 × 10 −5 to 50 × 10 −5 for tests without sand and in the range 130 × 10 −5 to 520 × 10 −5 for tests with sand. The ratios of wear coefficients, for tests with and without sand, range from 5 to 30. Adhesion creepage characteristics of a locomotive wheel simulation under different contaminants, such as water, sand, diesel fuel, lubricating oil, sand after oil and oil after sand, are also reported. It was observed that the liquid contaminants decreased adhesion levels and increased creepage. Sand application reduces the high creepage due to liquid contaminants, simultaneously increasing the attainable adhesion levels. Contrary to earlier findings, it was observed that the degree of sanding does influence the adhesion creepage characteristics, low sanding rates resulting in higher adhesion levels and lower creepage as compared to medium/high sanding rates. The wear and adhesion studies indicate that the beneficial effects of sanding in improving adhesion are more than offset by the increased wear rates. This increase is one to two orders of magnitude, compared to no sanding condition, depending on levels of traction. It is concluded that application of sand should be avoided as far as possible and high adhesion should be achieved by other means in order to minimize wear and the resulting track and equipment costs.

An experimental study of the real contact stresses for U.S. locomotives and rails including the effects of plasticity and wear has been performed under laboratory Hertzian simulation using the IIT-GMEMD wheel-rail simulation facility. Experiments were performed under both traction and braking conditions to account for differences observed earlier in the two modes. Wheel/rail tests were conducted using adhesion coefficients of 0.02, 0.15, and 0.25. Average contact stresses for various stages of wear were determined by measuring the contact areas. A synthesis of all the data generated showed that for operation of purely tractive wheels of a typical U.S. locomotive on a rail, the stabilized average contact stress ranges from approximately 100 to 25 ksi as the continuous operating adhesion coefficient ranges from 0 to 0.25. In order to determine the contact stresses for locomotives under field conditions, measurements of contact stresses were made on three different locomotives with wheels of different degrees of wear. Contact stresses for locomotives were found to be higher than stabilized contact stresses established by laboratory simulation tests. The locomotive wheel contact stresses were found to be closer to freight car wheel stabilized contact stresses established in an earlier study than for the laboratory locomotive simulation. It is suggested that this is due to the fact that 20 to 50 times as many cars operate on the same rails as do locomotives. On the basis of these experiments it is recommended that for U.S. locomotive wheels an average stabilized contact stress of approximately 65 ksi rather than the current 138 ksi would be quite stable. Profile and dynamic stability should be achieved simultaneously in such an approach. Currently available 2-D theories have been used to compare the experimental data showing poor agreement and reasons for discrepancy.

A series solution is obtained for thick-walled cylinders subjected to a temperature distribution which varies both radially and axially. The solution is based on three-dimensional linear theory of thermoelasticity, with appropriate approximations by neglecting small terms and using St. Venant’s principle. The internal and external curved surfaces are assumed traction-free. In each series of the solution the first term is identical to the thin shell theory and the subsequent correcting terms are expressed in increasing powers of the thickness-to-radius ratio. An illustrative example problem is solved by using the present solution, and the numerical results are compared with those based on the thin shell theory.

Plasma torch treatments of track rails to improve locomotive wheel traction were investigated as a potential substitute for rail sanding. Rolling friction tests were conducted under simulated rail-wheel load conditions at about 200,000 psi contact stress. Although traction was improved significantly in laboratory tests, total power required to significantly improve the coefficient of friction for a diesel electric operation at low speeds appears excessive. Power required at 80 mph is prohibitive. Moreover, safety, maintenance, and mechanical aspects require a considerable amount of further development before such an application can be considered possible on American railroads.

State-of-the-art methods in regular use for the calculation of the maximum soil-limited net traction of tracked vehicles ashore are briefed. A few simple caveats in applying these to similar vehicles crawling on the sea floor are suggested. Results of limited preliminary tests with a small single-tracked machine in submerged natural silty sand, with supporting in situ soil strength measurements, are presented and commented in context.

Wheel-rail adhesion has been the subject of vital interest to locomotive builders and the railroad. This area continued to receive considerable attention during the past decade due to the significant increase in horsepower available for traction of internally powered locomotives. For background information, this publication will briefly review the factors that affect adhesion and some of the published data on available adhesion between locomotive wheels and rail as a function of speed. Instrumentation, testing procedure, and results of adhesion tests on an SD-45 locmotive are discussed, and developed adhesion versus speed curves for different operating conditions are presented.

Conventional kinematic concepts of rolling can be seriously in error if applied to heavily loaded contacts, particularly if traction is large. Two relevant theorems of rolling are stated, which arose from consideration of an extreme case pertaining to pneumatic tires. For a comprehensive analysis of the latter, a mechanical analog is proposed.