DETECTION OF DNA-PSORALEN PHOTOADDUCTS in situ

Abstract— An immunological method, with the use of specific immune serum, has been developed for detection of 8-methoxypsoralen (8-MOP) photoadducts to DNA, formed in situ in cell nuclei, after combined treatment with 8MOP and UV-A irradiation (Zarçbska et al. , 1978). Lymphocytes fixed on slides or in suspension, and cryostat sections of different mammalian tissues, served as antigenic substrate, after treatment with 8-MOP and UV-A in vitro. Specific fluorescence in these substrates was detected in the nuclei after treatment with 30 ˜ 140 kJ/m² UV-A in the presence of 0.1-0.3 μg/cm² 8-MOP. PHA-stimulated-lymphocytes appeared to be the most sensitive substrate.
However, hairless mice treated with high doses of UV-A in vivo , 70 ˜ 360 kJ/m² did not reveal a specific fluorescence of epidermal nuclei, unless a high local concentration of 8-MOP was attained.
The apparent discrepancy in the level of photoadduct detection between the in vitro and in vivo treated specimens was explained by the low number of DNA-8-MOP-photoadducts formed in vivo under these experimental conditions. The relevance of these findings to the role of DNA-8-MOP-photoadducts formed during PUVA photochemotherapy is discussed.     

Shape factors for nozzle-corner cracks

Epoxy pressure vessels with fatigue cracks in the nozzle corner were burst.¹ The remains were cut into beams forK _Ic measurements. With nominal stress from burst pressures, crack sizes from measurements and stress intensities fromK _Ic, the shape factors were determined from the fundamental equation of fracture mechanics.     

A comparison of boundary element and finite element methods for modeling axisymmetric polymeric drop deformation

A modified boundary element method (BEM) and the DEVSS‐G finite element method (FEM) are applied to model the deformation of a polymeric drop suspended in another fluid subjected to start‐up uniaxial extensional flow. The effects of viscoelasticity, via the Oldroyd‐B differential model, are considered for the drop phase using both FEM and BEM and for both the drop and matrix phases using FEM. Where possible, results are compared with the linear deformation theory. Consistent predictions are obtained among the BEM, FEM, and linear theory for purely Newtonian systems and between FEM and linear theory for fully viscoelastic systems. FEM and BEM predictions for viscoelastic drops in a Newtonian matrix agree very well at short times but differ at longer times, with worst agreement occurring as critical flow strength is approached. This suggests that the dominant computational advantages held by the BEM over the FEM for this and similar problems may diminish or even disappear when the issue of accuracy is appropriately considered. Fully viscoelastic problems, which are only feasible using the FEM formulation, shed new insight on the role of viscoelasticity of the matrix fluid in drop deformation. Copyright © 2001 John Wiley & Sons, Ltd.     

An assessment of a parallel,finite element method for three‐dimensional,moving‐boundary flows driven by capillarity for simulation of viscous sintering

A parallel, finite element method is presented for the computation of three‐dimensional, free‐surface flows where surface tension effects are significant. The method employs an unstructured tetrahedral mesh, a front‐tracking arbitrary Lagrangian–Eulerian formulation, and fully implicit time integration. Interior mesh motion is accomplished via pseudo‐solid mesh deformation. Surface tension effects are incorporated directly into the momentum equation boundary conditions using surface identities that circumvent the need to compute second derivatives of the surface shape, resulting in a robust representation of capillary phenomena. Sample results are shown for the viscous sintering of glassy ceramic particles. The most serious performance issue is error arising from mesh distortion when boundary motion is significant. This effect can be severe enough to stop the calculations; some simple strategies for improving performance are tested. Copyright © 2001 John Wiley & Sons, Ltd.     

Characterisation of Interfaces Between Liquid Tin and Alumina in the Presence of Titanium Alloy Additions

Neutron reflection spectroscopy has been used to characterise the composition of interfaces between liquid Sn-Ti alloys and Al₂O₃. The reflectivity profiles for both 1% and 3% Ti alloys are consistent with the presence of a layer approximately 2 nm thick at the liquid/solid interface, showing extensive segregation of Ti. The composition of this layer is not pure Ti but shows a greater Ti content than any known Ti oxide composition. In all interfaces exposed to liquid metals at elevated temperature (including pure Sn) there is a layer of thickness 20–100 nm on the alumina surface with slightly lower neutron scattering length density than pure alumina. This is interpreted as evidence for surface roughening of the Al₂O₃ surface during exposure to the liquid metal.     

ON QUOTIENTS OF BRAID GROUPS

ABSTRACT.

Let G be the group ?[t, t ^?1] x ?. By studying the action of the braid group B_n on the set Gⁿ , we obtain representations of B_n into a wreath product of the symmetric group and the general linear group over ?[t, t ^?1]. This in particular recovers the Burau representation of the braid group. Furthermore, some quotients of the braid group are obtained by using the representations found.     

On setting a pressure datum when computing incompressible flows

The conventional approach to set the pressure level in a finite element discretization of an enclosed, steady, incompressible flow is to discard a continuity residual and set the associated pressure basis function coefficient to a desired value. Two issues surrounding this setting of a pressure datum are explored. First, it is shown that setting a boundary traction at a single node, in lieu of a Dirichlet velocity condition, is a preferred alternative for use with pressure‐stabilized finite element methods. Second, it is shown that setting a pressure datum can slow or even stop the convergence of a GMRES‐based iterative solver; though by some appearances a solution may appear to be converged, significant local errors in the velocity may exist. Under such circumstances it is preferable to solve the consistent singular system of equations, rather than setting a pressure datum. It is shown that GMRES converges in such cases, implicitly setting a pressure level that is determined from the initial guess. © 1999 John Wiley & Sons, Ltd.