Abstract: A method of making a structure having an optimized sterical shape of a material 4 comprises the steps of forming a pre-form 1 of the structure of the material 4, setting geometric boundary conditions for the material 4 being in a state in which it is able to rearrange under the influence of internal tensions in the volume of the pre-form 1 so that the material rearranges itself into the structure in a self-organized way, and transferring the rearranged material 4 into a state in which the rearrangement of the material 4 is conserved.

Abstract: The invention relates to a process for magnetic resonance examinations, in particular nuclear spin tomography, whereby a static magnetic field is produced in an examination zone and at least one gradient magnetic field is superposed, whereby unwanted alternating electromagnetic forces are produced, whereby by means of magnetostriction, forces opposing the electromagnetic forces are produced by the same gradient magnetic field and these forces act longitudinally and/or radially and/or around the circumference of the examination zone. Furthermore, the invention relates to a device for magnetic resonance examinations, in particular nuclear spin tomography, with a first generating unit for a static magnetic field, a second generating unit (20) for at least one gradient magnetic field and a support unit (10) for the second generating unit (20), whereby a magnetostrictive material system (11, 12) is provided on the support unit (10).

Abstract: A method for identification of dynamic characteristic quantities in a linear system having an electronic model, the electronic model having a complex signal flow with a complex residual a.sub.i =a'.sub.i +ja".sub.i, and a complex inherent value b.sub.i =b'.sub.i +jb".sub.i of the `i`th degree freedom describing model parameters a'.sub.i,a".sub.i,b.sub.i ', and b".sub.i for each degree of the system. The various signal flows are arranged parallel to one another for the various degrees of freedom, and are excited by the same signal `f` as the linear system. A difference e(n) determined at the time `n` between a model response `y` and a similar system response `d`, the system response having been time delayed by one unit of time, is used for adapting the model parameters for the system for the point in time (n+1), using variable step sized coefficients .alpha.'.sub.i (n),.alpha.".sub.i (n),.beta.'.sub.i (n),&.beta.".sub.i (n) for determining the accuracy of the adaptation to the system.