MAGNETOMETRY USING SQUID

SUMMARY OF COMMUNICATIONS stone with dropstone on the roche moutonnee from Salto,to cross-bedded and “convoluted” sandstone beds in theriver channel, toward NW. They are interpreted as proxi-mal and distal glacial facies respectively, associated withan advance and retreat of the glacier.Those rocks are overlain by an extensive andthick shale (pellitic) section that may representthe deposit of a marine transgressive post-glacialepisode. Dropstones in the shale denote presenceof icebergs and permanence of glacial inﬂuence dur-ing sedimentation. — ( December 14, 2001 ). ∗ Supported by FAPESP 00/6039-6.**E-mail: jbviviani@aol.com MULTIDISCIPLINARY THEMES Alcides N. Sial and Valderez P. Ferreira(Organizers) EFFICIENT ALGORITHMS FOR PACKING BOXESINTO CONTAINER Sostenes L. S. Lins Departamento de Matematica, CCEN, Universidade Federalde Pernambuco, Recife, PE, Brazil. The problem of packing boxes into bigger boxes(containers) is a practical problem which, by its economi-cal signiﬁcance, deserves a serious and inspired scientiﬁcapproach. From a strictly mathematical point of view, itis an incredibly challenging problem which badly needsgood heuristics to solve it. We have been involved in try-ing to ﬁnd such heuristics since 1997. These heuristicsuse new concepts in graph theory (the tets), in data struc-ture (the phormas), and used some classical algorithms astopological sorting and the coding of combinatorial ob-jects to approach real world packing problems. A distin-guished feature of our approach is the visual treatment ofthe spatial packing: we produce a sequence of homoge-neous increments (a loading plan) in the packings whichmakes it easy to visualize and to actually produce the so-lutions found.We have produced a set of effective heuristics fordealing with real world box packing problems: we permitvarious types of boxes, various types of containers, de-mand requirements on the boxes and the information onwhichboxtypescanchangeitsverticaldirection. Theim-plementation of our heuristics produces very good pack-ings as compared with non-scientiﬁc ones: typically weput 7% to 12% more boxes. The full paper related to thesematters is scheduled to be published this year (2002) inthe European Journal of Operations Research.A computer package named ExpedPlex is under de-velopment, where the pertinent algorithms are being efﬁ-ciently implemented. — (


SUMMARY OF COMMUNICATIONS
stone with dropstone on the roche moutonnée from Salto, to cross-bedded and "convoluted" sandstone beds in the river channel, toward NW. They are interpreted as proximal and distal glacial facies respectively, associated with an advance and retreat of the glacier.
Those rocks are overlain by an extensive and thick shale (pellitic) section that may represent the deposit of a marine transgressive post-glacial episode.
Dropstones in the shale denote presence of icebergs and permanence of glacial influence during sedimentation.
- ( December 14, 2001  The problem of packing boxes into bigger boxes (containers) is a practical problem which, by its economical significance, deserves a serious and inspired scientific approach. From a strictly mathematical point of view, it is an incredibly challenging problem which badly needs good heuristics to solve it. We have been involved in trying to find such heuristics since 1997. These heuristics use new concepts in graph theory (the tets), in data structure (the phormas), and used some classical algorithms as topological sorting and the coding of combinatorial objects to approach real world packing problems. A distinguished feature of our approach is the visual treatment of the spatial packing: we produce a sequence of homogeneous increments (a loading plan) in the packings which makes it easy to visualize and to actually produce the solutions found.
We have produced a set of effective heuristics for dealing with real world box packing problems: we permit various types of boxes, various types of containers, demand requirements on the boxes and the information on which box types can change its vertical direction. The implementation of our heuristics produces very good packings as compared with non-scientific ones: typically we put 7% to 12% more boxes. The full paper related to these matters is scheduled to be published this year (2002)  Superconducting Quantum Interference Device (SQUID) is by far the most sensitive device ever built. It is a magnetic flux detector that can be used to measure magnetic field as low as few femtoTesla. Because of its high sensitivity, SQUID are being employed in detecting magnetic field generated by neural brain activity, noninvasive detection of metallic pieces inside the human body, nondestructive evaluation of corrosion and flaws, and in characterizing magnetic properties of materials, just to list few applications. Perhaps, the main limitation to replace most of the available magnetic sensors with SQUID's is the need of using liquid helium to operate them. However, with the recent progress made in the high temperature superconductor research field, there are SQUID being built that operate with liquid nitrogen and this makes them even more attractive.
Nowadays, there is a couple of companies around the world which sales SQUID magnetometers. However, they are too dedicated with almost no flexibility, the price is still too high and it is quite difficult to get maintenance when they fail. Another drawback is that these systems are true black boxes, not allowing graduate students and technicians to be trained in this very important technology. In my talk, I will describe a homemade magnetometer that uses a SQUID as the sensing element to investigate magnetic properties of materials. Our SQUID system operates from room temperature down to 1.5 K in a superconducting solenoid that generates magnetic field as high as 8 T.
The SQUID system is particularly important when only a small amount of the magnetic material is available or to investigate system that presents weak magnetism. Some results of the magnetic studies made in magnetic polymers, thin-films and manganites will also be presented.