Está en la página 1de 4
7R.C. Nicklin and B. Rafert, Am. J. Phys. $0, 108 (1982). °R.C.Camp, T. A.Smay, and C.J. Triska, Mieroprocesor Systems Engi- neering (Matrix, Portland, OR 1979). L. Scanlon, 6502 Softeare Design (Howard Sams, Indianapolis, IN, 1580) "P. Horowitz and W. Hil, The Art of Electronics (Cambridge, Cam bridge, CB2 IRP, 1980). Emphasis ison the 8085 procestor. “A. Lesea and R. Zaks, Microprocessor Interfacing Techniques (Sybex, Berkeley, CA, 1985), 3rd ed The floating plank R. Delbourgo "J, Carr, Designing Microprocessor Based Instrumentation (Reston, Res ton, VA, 1982). RK. Newcomb, BYTE §,202 (Feb, 1980), "8, Caria, BYTE 6, 66 (May 1981). "North American Philips Corporation, Cheshire Di dustrial Park, Cheshire, CT 06410. "8. Licson, The Handbook of Microcomputer Interfacing (Tab, Blue Ridge Summit, PA, 1983). "Available from E & L Instruments Ine, Detby, CT. on, Cheshire ln Physics Department, University of Tasmania, Hobart, Australia 7005 (Received 17 July 1986; accepted for publication 20 December 1986) The stable floating configuration of a long plank of rectangular cross section depends on the relative density of the plank to the fluid and on the ratio of the sides. The complete solution of this metacentric problem is given. LINTRODUCTION Courses in continuum mechanics include a component of hydrostatics asa preliminary to the more difficult task of understanding fiuid mechanics. Although hydrostatics is considered to be an easy topic, there is one part that cer- tainly is not, namely the determination of equilibrium con- figuration of floating bodies. This is discussed in some de- tail in Sommerfeld’s classic text.' From our point of view the relevant points are that (i) when a body floats in eq librium, the weight and buoyancy forces are equal and op- posite and pass through the same point, either the center of gravity Gor the center of buoyancy C; (ii) this equilibrium ‘may be stable or not as the body undergoes atlt through an angle 6 and the center of buoyancy moves toa new position Co, such that the magnitudes of weight and buoyancy forces remain equal; (iii) forsmall @ the metacenter M lies at the center of curvature of the buoyancy arc CC; (iv) if the equilibrium is to be stable, the couple on the body must actin such a way that it should restore equilibrium and this ‘corresponds to a metacenter which lies above the center of gravity. Thus the problem is one of finding the metacentric point M of the buoyancy forces as the body undergoes a virtual rotation and this depends critically on the geometry of the system. In fact, the design of keels of boats, for in- stance, is very much an exercise of discovering the shape ‘which maximizes the metacentric height for large rotations of the boat. This is no simple matter and requires as much flair as it does science. A seemingly trivial example of stability concerns a two- dimensional rectangular body floating in a fluid; in particu- lar this can be modeled by a long plank having a rectangu- lar cross section of sides a, b, where we may disregard the long side if it lies nearly horizontally. The analysis of this example is surprisingly complicated and I often set this as a challenging problem to final year undergraduates. ‘The 799 Am. J. Phys. 85 (9), September 1987 book on hydrostatics by Lamb? and the text by Ramsey? treat parts of this problem and mention that Huygens de- voted much effort into studying the various possibilities, with a final solution having to await Korteweg. However, the ideas are rather scattered and I thought it would be worthwhile to provide the complete analysis in this Journal in order to expose the richness of this type of problem. I make no claim to originality; indeed theres little one needs to know, beyond first-year university mathematics and physics, in order to follow the arguments. tis clear that the stable configuration can only depend con the aspect ratio r = 6 /a and on the density of the plank relative to that of the fluid s (< 1). Without loss of genera- lity, wemay choose r< 1, by swapping the two sides ifneces- sary. At the same time we can takess<} due to the following duality argument.”* Let G be the location of the center of mass of the entire body, C the center of mass of the immersed part (ic, the center of buoyancy), and C’ the center of mass of the ex- posed part. See Fig. 1(a). Obviously, G lies on the line between Cand C’. If V denotes the immersed volume and V’ the exposed volume, Archimedes’ principle tells us that SHV) =VorsV! = (1 = 5). a Thus an interchange s-+(1— 5) corresponds to V=+V’. Hence to each equilibrium configuration there is a dual one where s~1 — 5 and the body is rotated through 180°: see Fig. 1(b). In fact one can also prove that duality preserves stability.’ For this reason I shall restrict the discussion to <1, (It is easy enough to make the appropriate substitu- tions to cover the case where J { (b) The dual fat corresponding to s5-+1—5.(¢) Relative positon ofthe metacenter M, center of mass 6, and center of buoyancy C,for stable tation. C moves to Cand O to Oy as the body stilted through (dashed shape). Mis the center of curvature of the buoyancy are CC, fo small 0. (A) the longer side is immersed asymmetrically as in Fig. 2(4), (5) only one corner is immersed ‘“quasisymmetrical- ly” asin Fig. 2(e), (6) only one corner is immersed “asym- metrically” as in Fig. 2(f). Im the last two cases, the defini- tion of “quasisymmetric” refers to a central root of a quartic that corresponds to a symmetrically floating square plank when r= 1; it retains that significance even for r< 1 I shall analyze cases (1) and (3) in Sec. II; cases (2) and (4) are simply derived from these by inverting r. Cases (5) and (6) are more difficult and are left to Sec. I11. Evidently, there is a critical value of the specific gravity s for which ceases (2) and (4) merge into case (6) and this will be mentioned at the relevant place in the analysis. Section IV contains the conclusions, comprising Fig. 3, in which all the various possibilities are neatly graphed. A glance at ‘hat diagram will reveal which of the configurations the plank will settle down to. In order to determine the metacenter M, as the body undergoes a rotation 6 (with the same immersed volume), it proves more convenient to virtually rotate the fluid level rather than the body. For small ,M islocated at the center of curvature of the buoyancy arc and has to lie vertically above G if the equilibrium is to be stable—see Fig. 1(c) Since, for small virtual 8, MC is a constant and GC< MC, GC has to be a minimum as a function of 8, to guarantee stability. This is the rule which gives al the results; the rest is algebra and some calculus. j} —4 @ o Fig. 2.Cases(a)(Q corre- (1)-(6) stated inthe nteo- duction @ « 800 —Am.J. Phys, Vol. $5, No. 9, September 1987 spond 10. configurations fay Ei ite -f 4: Fig. 3. The various regions of stability marked on an plot. Cotigura- tions: (1) —shading, (2) | shading, (3) / shading, (4) \ shading, (3) otted shading, and (6) solid shading. 1. CONFIGURATIONS WITH TWO CORNERS IMMERSED Examine Fig. 4 which corresponds to a general inclined configuration (it includes the symmetric choice @ = 0.as a special case). We will require GC to be vertical for the 8 in question. Using the Cartesian axes as drawn, and remem- bering 6 = ra, the center of gravity is at G= (17) and the center of buoyancy (the center of mass of the im- ‘mersed trapezium) is Teadily calculated to lie at C= las + 7 tan 8 /12s, r+ tan 6/65) Fig. 4 Hydrostatic configuration with two corners immersed. Here GCIs normal 1 the surface, which is drawn at inclination 8 for convenience. R.Delbourgo 200 Hence we can find the coordinate vector CG. Since CGis to bbe vertical one needs Yoo = — Xeq tan 8. We thereby ob- tain the equilibrium condition, tIs(s— 1) + 7/6 + 717/12] =0, @ where t= tan 0. At the same time we have 4(CG)7/a? = (s— 14 717/128)? + (Ft /6s)7,, (3) which has to be a minimum for stability. The vanishing of the first derivative of Eq. (3) just coincides with Eq. (2) and the requirement of positive second derivative gives s(5—1) + 7/6 + 717/430. (4) ‘There are two equilibrium solutions of Eq, (2), namely the symmetrical float t=, stable [according to (4)) if P>6s(1—s) and asymmetrical float 1? = 12s(1 —3)/ P—2, stable if P<6s(1—s). Thus the curve P= 65(1—s) demarcates the boundary between a sym- metric and an asymmetric equilibrium stance. This is not quite the full story. Because both corners are immersed by assumption, there is an upper limit to 6, given by tan @ = 2s/r (at which stage the trapezuim in Fig. 4 re- duces to a triangle with one vertex at the origin of coordi- nates). This critical asymmetric case corresponds to P= 6s— 8s", ‘The conclusion of this section is that case 1: symmetric, stable occurs when 6s(1 —s)

J, whereupon 1, go complex, thereby excluding the asymmetric solu- tions. On the other hand if = ¢, or f_, reality necessitates q4, Putting all this together, one can assert that for a square plank the symmetric state is the stable one for 4 I by simply redefining the sidesa, 6, and when 5>4, by applying duality (rotation of the body through 180" about the surface). tis perhaps surprising that such a simple excercise as this turns out tobe so complicated in practice. The example only serves to underline how subtle metacentric problems ‘can be in general.! I have assumed throughout this article that the plank is long and that the other dimension can be neglected. With a shorter block this is clearly a false pre- mise. Ramsey° provides a brief discussion of the floating cube but his treatment is far from exhaustive and does not apply to the general rectangular block anyway. The inter- ested reader is invited to generalize my presentation to that situation. "A. Sommerfeld, Mechanics of Deformable Bodies, Vol. IT (Academic, London, 1967) °H. Lamb, Staries (Cambridge U. P, London, 1921), SASS. Ramsey, Hydrostaties (Cambridge U.P, London, 1936) Ehrenfest theorem and the classical trajectory of quantum motion J. Nag Department of Physics, Jadavpur University, Caleutta 700032, India V. J. Menon and S.N. Mukherjee Department of Physics, Banaras Hindu University, Varanasi 221005, India (Received 14 May 1986; accepted for publication 29 October 1986) Ehrenfest theorem asserts that the quantum mechanical motion of a particle when considered in the expectation value sense should agree with classical mechanics in the correspondence limit. An explicit verification of this result is presented in the orie-dimensional case for motion in an infinite potential well (large quantum number limit) and a brief mention is made of the case of a smooth potential (+0 limit) INTRODUCTION The question of the behavior of quantum systems in the classical limit has been considered in detail in the litera- ture.'* The Ehrenfest theorem® deals with this issue in a very general way by demonstrating that d (X)/dt and d (P)/dt obey Hamilton-like equations where the expecta~ tion values (X ), (P) of the position and momentum opera~ tors are taken with respect to a suitable wave packet state Ig(#)), say, in the Schrédinger picture. However, the ex- plicit calculation of the (X’) value and actual verification that it does reduce to the Newtonian trajectory in the corre- spondence limit (i.e. for large quantum numbers and small Planck's constant) is a much harder problem, This issue seems to have been tackled only for a few, albeit ifluminat- ing, cases such as the free particle, the harmonic oscillator, and the nonlinear Schrédinger equation.’ 802 _Am.J. Phys. $5 6), September 1987 The aim of the present article is to demonstrate such a correspondence between the quantum and classical mo- tions in two more examples using different techniques in ‘each case. Section II considers a particle in an infinite square well potential whose bound quantum mechanical ‘energy eigenfunctions are well known,‘ and shows how the series for (X') goes over to the Fourier decomposition of the underlying classical trajectory in the limit of large quan- ‘tum numbers. In Sec. III we discuss a general smooth short-range interaction V(x) whose continuum eigensolu- tions possess a WKB representation® in the #0 limit and ‘make the passing observation that the probability density |§(a0) P develops a peak at x = x,, where x, is the stan dard Hamilton’s solution forthe trajectory: result already derived by Brown? for a multiparticle system using the Hamilton-Jacobi theory. Finally, Sec. IV discusses some features of our method along with possible difficulties if © 1987 American Association of Physics Teachers 802

También podría gustarte