The effects of turbulence in stationary gas tungsten arc (GTA) welds in AISI 304 stainless steel were examined using a finite element thermofluids model. The model includes buoyancy, Lorentz and Marangoni driven fluid flow, a large deformation model of the free surface, and a k–∊ turbulence model. To facilitate implementation of the wall function boundary conditions for the k–∊ turbulence model, a dynamic numerical grid remapping technique was used to clearly separate elements in the liquid from those in the solid. The influences of sulphur content of AISI 304 stainless steels on the turbulent viscosity, fluid flow, and weld pool dimensions were simulated. Good correlation between experimentally observed and predicted weld pool shapes and dimensions was obtained. Also, the effect of sulphur concentration on AISI 304 weld pool dimensions was correctly predicted. The simulations indicate that the flow in such stationary GTA weld pools in AISI 304 stainless steel is not laminar and that quantitatively accurate predictions of weld pool fluid flow and weld shapes and dimensions will only be possible if the effects of turbulence in GTA weld pools are modelled correctly.

A comprehensive thermofluids model of stationary gas tungsten arc welding (GTAW) has been developed and used to examine the effects of thermofluids phenomena on the predicted temperatures and flow field in the weld pool, as well as their impact on the resultant weld pool dimensions in 304 stainless steel and 6061 aluminium plates. A dynamic numerical grid remapping technique was used within the finite element based model to model the geometry of the solid/liquid interface and free surface of the weld pool. Initial work showed that correlation between experimental and predicted weld pool dimensions was only possible provided that the effects of turbulence were modelled using a well posed k - ∊ turbulence model. However, the two-equation k - ∊ turbulence model introduces additional complexity and requires considerable additional computational effort. To overcome these shortcomings, a simpler vorticity based turbulence model has been developed in which the turbulent viscosity and thermal conductivity are based upon the magnitude of the vorticity in the flow field as predicted through solution of the continuity, momentum and energy equations. Excellent correlation was obtained between the weld pool dimensions predicted by the vorticity based turbulence model, the predictions from the k - ∊ turbulence model and the experimental data from welds made in 304 stainless steel with three different sulphur impurity concentrations and in 6061 aluminium. The advantage of the vorticity based turbulence model is that it is significantly less computationally intensive than the standard k - ∊ turbulence model and has, therefore, the potential of providing tractable and practical computations of fully three-dimensional welding processes.

No abstract

Bismuth triborate, BiB 3 O 6 (BiBO), a promising new material for effective frequency conversion, was synthesized by top-seeded growth from the nearly stoichiometric melt at high temperature. The crystal space group is non-centrosymmetric monoclinic C2 with Z = 2. The lattice constants were a = 7.116.2/, b = 4.993.2/ and c = 6.508.3/Å with b = 105.62.3/ • . The atomic Wyckoff positions are Bi in 2a, B(1) in 2b and B(2) and O in 4c. Raman spectra show that the structural rigidity of the BiBO crystals can be mainly ascribed to B-O bond stretching and bending. The B(1)O 4 units weaken the conjugated p-orbit system. The BiBO crystal has a short ultraviolet absorption edge. Therefore, the two principal structural units, B(2)O 3 and B(1)O 4 , and the bonding are of considerable interest.

The cluster formations in solid-liquid interface boundary layers of KDP crystals have been studied by Raman spectroscopy and ab initio cluster calculations with the density-functional theory. The calculations are made on H 2 PO 4 , H 4 P 2 O 8 , and H 4 P 4 O 16 clusters, which model the cluster structure in different growth layers. Excellent agreement has been achieved between vibrational frequencies predicted by theory and those observed in experiments. The present results also show that the desolvation process of polymer clusters takes place within the characteristic boundary layers. The growth unit of crystal growth is identified as H 4 P 2 O 8 dimer cluster.