On the iterative solution of the algebraic equations in fully implicit Runge-Kutta methods

This paper deals with the iterative solution of stage equations which arise when some fully implicit Runge-Kutta methods, in particular those based on Gauss, Radau and Lobatto points, are applied to stiff ordinary differential equations. The error behaviour in the iterates generated by Newton-type and, particularly, by single-Newton schemes which are proposed for the solution of stage equations is studied. We consider stiff systems y'(t) = f(t,y(t)) which are dissipative with respect to a scalar product and satisfy a condition on the relative variation of the Jacobian of f(t,y) with respect to y, similar to the condition considered by van Dorsselaer and Spijker in [7] and [17]. We prove new convergence results for the single-Newton iteration and derive estimates of the iteration error that are independent of the stiffness. Finally, some numerical experiments which confirm the theoretical results are presented. This revised version was published online in June 2006 with corrections to the Cover Date.     

Can a finite element method perform arbitrarily badly?

In this paper we construct elliptic boundary value problems whose standard finite element approximations converge arbitrarily slowly in the energy norm, and show that adaptive procedures cannot improve this slow convergence. We also show that the -norm and the nodal point errors converge arbitrarily slowly. With the -norm two cases need to be distinguished, and the usual duality principle does not characterize the error completely. The constructed elliptic problems are one dimensional.

     

Monte Carlo simulations of electron dynamics in N2/CO2 mixtures

Motivated by experimental investigations of electrical discharges in N₂/CO₂/H₂O, Monte Carlo (MC) electron dynamics simulations in atmospheric N₂/CO₂ mixtures were performed. The goal was to obtain electron energy distribution functions (EEDFs), mean free path, drift velocity, collision frequency and mean energy of electrons, rate coefficients of electron-impact reactions, ionisation and attachment coefficients, as functions of the reduced electric field strength (E/N) and of the concentration of individual gas components. The results obtained by MC simulations were fitted with polynomials of up to the 3rd order with reasonable accuracy for E/N above 80 Td. The studied parameters below 80 Td were strongly non-linear as functions of E/N. This is mostly due to the influence of elastic collisions of electrons with CO₂ molecules prevailing in CO₂-dominant mixtures for E/N < 40 Td, and vibrational excitation collisions of N₂ species prevailing in N₂-dominant mixtures for E/N from 40 to 80 Td. The effect of these electron-impact processes was specific for each of the studied parameters.     

Quenching of the Haldane gap in LiVSi<Subscript>2</Subscript>O<Subscript>6</Subscript> and related compounds

We report results of susceptibility χ and ⁷Li NMR measurements on LiVSi₂O₆. The temperature dependence of the magnetic susceptibility χ(T) exhibits a broad maximum, typical for low-dimensional magnetic systems. Quantitatively it is in agreement with the expectation for an S=1 spin chain, represented by the structural arrangement of V ions. The NMR results indicate antiferromagnetic ordering below T_N=24 K. The intra- and interchain coupling J and J_p for LiVSi₂O₆, and also for its sister compounds LiVGe₂O₆, NaVSi₂O₆ and NaVGe₂O₆, are obtained via a modified random phase approximation which takes into account results of quantum Monte Carlo calculations. While J_p is almost constant across the series, J varies by a factor of 5, decreasing with increasing lattice constant along the chain direction. The comparison between experimental and theoretical susceptibility data suggests the presence of an easy-axis magnetic anisotropy, which explains the formation of an energy gap in the magnetic excitation spectrum below T_N, indicated by the variation of the NMR spin-lattice relaxation rate at T≪T_N.     

Effective Hamiltonian approach to doubly degenerate electronic states

Effective vibronic Hamiltonian models are built for E ⊗e Jahn-Teller systems and analytical solutions are obtained through Lie algebraic methods. Although approximate, we show that these models allow in particular to recover the possible ground state crossover when quadratic couplings are present. The equivalence of E ⊗e and G' ⊗e vibronic systems in cubic symmetry is precisely established through a particular realization of the electronic operators for an orbital quadruplet. We show how this equivalence is broken by a rovibronic interaction which, for a G' ⊗e system, still gives an exactly solvable model.     

Superfluidity in many fermion systems: Exact renormalisation group treatment

The application of the exact renormalisation group to symmetric as well as asymmetric many-fermion systems with a short-range attractive force is studied. Assuming an ansatz for the effective action with effective bosons, describing pairing effects, a set of approximate flow equations for the effective coupling including boson and fermionic fluctuations has been derived. The phase transition to a phase with broken symmetry is found at a critical value of the running scale. The mean-field results are recovered if boson-loop effects are omitted. The calculations with two different forms of the regulator are shown to lead to similar results. We find that, being quite small in the case of the symmetric many-fermion system the corrections to mean-field approximation become more important with increasing mass asymmetry.     

Electromagnetic proton-neutron knockout off <Superscript>16</Superscript>O : New achievements in theory

Results for the cross-sections of the exclusive ¹⁶O(e, e'pn)¹⁴N and ¹⁶O(γ, pn)¹⁴N knockout reactions are presented and discussed in different kinematics. In comparison with earlier work, a complete treatment of the center-of-mass (CM) effects in the nuclear one-body current is considered in connection with the problem of the lack of orthogonality between initial bound and final scattering states. The effects due to CM and orthogonalization are investigated in combination with different treatments of correlations in the two-nucleon overlap function and for different parametrizations of the two-body currents. The CM effects lead in super-parallel kinematics to a dramatic increase of the ¹⁶O(e, e'pn) cross-section to the 1₂ ⁺ excited state (3.95MeV) of ¹⁴N . In all the situations considered the results are very sensitive to the treatment of correlations. A crucial role is played by tensor correlations, but also the contribution of long-range correlations is important.     

Deposition of Ni<Subscript> 13</Subscript> and Cu<Subscript> 13</Subscript> clusters on Ni(111) and Cu(111) surfaces

The soft deposition of Ni₁₃ and Cu₁₃ clusters on Ni(111) and Cu(111) surfaces is studied by means of constant-energy molecular-dynamics simulations. The atomic interactions are described by the Embedded Atom Method. It is shown that the shape of the nickel clusters deposited on Cu(111) surfaces remains rather intact, while the copper clusters impacting on Ni(111) surfaces collapse forming double and triple layered products. Furthermore, it is found that for an impact energy of 0.5 eV/atom the structures of all investigated clusters show the lowest similarity to the original structures, except for the case of nickel clusters deposited on a Cu(111) surface. Finally, it is demonstrated that when cluster and substrate are of different materials, it is possible to control whether the deposition results in largely intact clusters on the substrate or in a spreading of the clusters. This separation into hard and soft clusters can be related to the relative cohesive energy of the crystalline materials.