Why define atoms in real space?

The physics of a system is determined by a variation of the action integral, i.e., by a variation of the space–time volume integral of the Lagrange function. If one demands that the properties of an atom in a molecule be derived from physics, the atom must generate its own space–time volume, requiring that its boundaries be defined in real space. The variations in the action are related to the actions of generators of infinitesimal unitary transformations. In the general case, the action integral is altered by generators acting in both the spacelike and timelike surface bounding the space–time volume, whereas for a total isolated system, the physics is totally determined by their action in just the spacelike surfaces at the two time endpoints. It is shown and illustrated for a one-dimensional system that the definition of an atom corresponds to the possibility of choosing a subsystem in such a way that the contributions to the change in action resulting from the evolution in time of its spatial boundaries vanishes identically. The properties of these subsystems and of the total system of which they are a part are, therefore, determined by one and the same action principle. This choice of subsystem corresponds to the possibility of augmenting the Lagrange function by the divergence of the gradient of the electron density a step that, while leaving the equations of motion unchanged, modifies the generating operators in the required manner. © 1994 John Wiley & Sons, Inc.     

Theory of relaxed density matrices: Application to second-order response properties

The evaluation of second derivatives of the electronic energy for nonvariational wave functions using an energy functional is discussed. It is shown that, in certain cases, the formation of the first-order relaxed density matrix leads to an efficient algorithm for the calculation of second-order response properties. Detailed formulas are given for second-order Møller–Plesset perturbation theory. © 1994 John Wiley & Sons, Inc.     

Density functionals for exchange and correlation energies: Exact conditions and comparison of approximations

A set of exact conditions is compiled for the purpose of developing and testing approximations for the exchange-correlation energy as a functional of the electron density. Special emphasis is placed upon recently developed density-scaling relationships. Commonly used generalized gradient approximations are compared against several of these conditions. A direct tabular comparison of these functionals (not of calculated properties) with one another is also made. © 1994 John Wiley & Sons, Inc.     

Mechanical properties of polyester polymer-concrete

The properties of a polymer-concrete composed of polyester matrix and locally available rock aggregate are investigated. The formula of the concrete is found by an experimental-calculation approach in such a way as to attain a closer packing of the aggregate particles on the one hand, and to ensure the needed processing characteristics (placeability) of the mix on the other. It is shown experimentally that the material obtained has a rather high compression strength. Under prolonged compression loads, the polymer-concrete exhibits a noticeable creep behavior with a linear relation between the creep strains and stresses. After the action of half the ultimate load over 3000 h, the total strains exceed the instantaneous ones by 2.0 to 2.2 times. The accumulation of irreversible strains is also observed; however, their contribution to the total strain is small. It is found that the stress-strain relation can be represented by the equation of linear hereditary creep theory.Institute of Polymer Mechanics, University of Latvia, Riga, LV-1006, Latvia. Translated from Mekhanika Kompozitnykh Materialov, Vol. 35, No. 2, pp. 147–162, March–April, 1999.