MAGNETOHYDRODYNAMICS Vol. 38 (2002), No. 1-2, pp. 121–127 MECHANICALLY FORCED AND THERMALLY DRIVEN FLOWS IN LIQUID SODIUM W.L. Shew, D.R. Sisan, D.P. Lathrop University of Maryland, College Park, MD, USA 20742 We describe experiments designed to better understand the effects of turbulent flow and rotation on magnetic field self-generation. Three different types of liquid sodium experiments are detailed: mechanically forced experiments with an asymmetric geom- etry, a rotating convection experiment, and a planned larger experiment that is both mechanically forced and rotating. Introduction. The dynamo process of magnetic field self-generation is be- lieved to be present in many astrophysical objects: the sun, most of the major planets, and large scale objects such as galaxies. This belief is supported by a large body of research, described in excellent reviews by Glatzmaier & Roberts [9] and Busse & Carrigan [2]. Most of this research is numerical or analytical, which requires approximations to make the problem tractable. While valuable, this re- search leaves questions unanswered. Experiments, which can naturally model the turbulent conditions of astrophysical bodies, may address the unanswered ques- tions and directly test the reliability of numerical and analytical approximations. In our experiments, some issues we are trying to better understand are: the dy- namics of the Earth’s field, the mechanism that determines the saturated field strength, and the way in which magnetic field generation by turbulent flow is affected by rotational and Lorentz forces. With these goals in mind, we are attempting to produce a self-generating laboratory dynamo. Previous results from a mechanically forced experiment in- dicate a trend towards self-generation [13]. Motivated by this progress, we are attempting several other related experiments, described in this paper. We are also motivated by the successes of Gailitis et al. at Riga [14, 15] and M¨ uller & Stieglitz at Karlsruhe [10], who have successfully produced laboratory dynamos. In Sec. 1 of this paper, we describe a geometric variation of the mechanically- forced apparatus designed (but failing) to improve magnetic field stretching. In Sec. 2, we present a rotating convection apparatus, now nearing its first run. Finally, in Sec. 3, we detail the design for a larger mechanically forced, rotating sodium experiment in a three meter diameter sphere. 1. Herzenberg-type flow. The above mentioned trend towards dynamo action was achieved in an apparatus [13] that consisted of a 30.5 cm diameter spherical shell with four thin pole-to-pole baffles on the inside surface that extended 1.5 cm into the flow. The liquid sodium was forced with propellers mounted on coaxial shafts that entered the sphere at the two poles. The shafts were driven with two variable speed 7.5 kW motors. The primary diagnostic was magnetic pulse decay measurements. A pair of external coils supplied a magnetic pulse for one second and then quickly turned off (∼ 1 ms). The subsequent decaying magnetic field was measured with hall probes just outside the sphere. The decay rate is taken as the largest (slowest decaying) finite time Lyapunov exponent for the system [16]. 121