• Abstract and Figures
• Public Full-text

Viscous control of minimum uncertainty state T. Koide∗ Instituto de F´ısica, Universidade Federal do Rio de Janeiro, C.P. 68528, 21941-972, Rio de Janeiro, RJ, Brazil A minimum uncertainty state for position and momentum is obtained in quantum viscous hydro- dynamics which is defined through the Navier-Stokes-Korteweg (NSK) equation. This state is the generalization of the coherent state and its uncertainty is given by a function of the coefficient of viscosity. The uncertainty can be smaller than the standard minimum value in quantum mechanics, ~=2, when the coefficient of viscosity is smaller than a critical value which is similar in magnitude to the Kovtun-Son-Starinets (KSS) bound. I. INTRODUCTION second coefficient of viscosity, respectively. The traceless symmetric stress tensor is defined by The uncertainty relation is an important feature in ij 1 j i 1 quantum physics and its comprehension requires un- E = @iv + @jv − (r · v) δij : ceasing improvement [1{5]. A similar relation was re- 2 D cently proposed in classical viscous hydrodynamics [6, 7]. Normally, hydrodynamics is described using the mass dis- There, the minimum uncertainty for position and mo- tribution. For the sake of comparison with quantum me- mentum of fluid elements is given by a function of the chanics, however, we use the distribution of constituent coefficient of viscosity. This suggests that the uncertainty particles of the quantum viscous fluid ρ, which is normal- of a quantum state can be controlled by viscosity. ized by the number of constituent particles N. Then the Viscosity is one of dissipative effects but there is no es- mass distribution is given by Mρ with M being the mass tablished quantum theory of dissipation. The dissipative of constituent particles. The last term on the first line is effect is introduced, for example, by the coarse-graining the gradient of the quantum potential. Ignoring the con- of environment variables using the influence functional tributions on the second line and choosing κ = ~2=(4M2), method or the projection operator method. The derived Eq. (1) agrees with Madelung's Euler equation. It is in- dissipative equation however can describe an unphysical teresting to note that Eq. (1) is a special case of the state with negative probability [8]. See also Refs. [9{ Navier-Stokes-Korteweg (NSK) equation which was pro- 16] for other approaches of quantum dissipation. At the posed to describe liquid-vapor fluids near phase transi- same time, the behavior of quantum many-body systems tions. Then the term associated with the quantum po- is approximately given by viscous fluids in heavy-ion col- tential represents the capillary action [20, 21]. lision physics [17]. The viscous effect is indeed considered The uncertainty relation for the quantum viscous fluid to be indispensable because it is believed that the coef- is derived by applying the method developed in Refs. ficient of viscosity cannot be smaller than the Kovtun- [6, 7] to Eq. (1). However, the existence of the corre- Son-Starinets (KSS) bound [18]. sponding minimum uncertainty state has not yet been As was discussed by Madelung, the Schr¨odingerequa- known. The purpose of this paper is to derive the min- tion can be expressed in the form of the Euler equation imum uncertainty state. The uncertainty of the derived with the so-called quantum potential [19]. Thus it is nat- state is a function of the coefficient of viscosity and can ural to consider that quantum viscous hydrodynamics is be smaller than the standard minimum value in quantum described by the Navier-Stokes-Fourier (NSF) equation mechanics ~=2 for a sufficiently weak viscosity. with the quantum potential term. In the D-dimensional This paper is organized as follows. In Sec. II, quantum Cartesian coordinate system, this equation is given by viscous hydrodynamics is formulated in the framework of the stochastic variational method [6, 7, 22{28]. In Sec. arXiv:2104.11777v1 [quant-ph] 23 Apr 2021 2p i 1 r ρ III, the uncertainty relation is derived by applying the (@t + v · r)v = − @iV + 2κ∂i p M ρ method in Refs. [6, 7]. The minimum uncertainty state D in quantum viscous hydrodynamics is derived in Sec. IV. 1 n η o 1 X − @ P − µ + (r · v) + @ ηEij ; Section V is devoted to concluding remarks. Mρ i D Mρ j j=1 (1) II. STOCHASTIC VARIATIONAL METHOD where v, V , P , η and µ are the velocity field, the external potential, the pressure, the coefficient of viscosity and the To define the uncertainty relation in viscous systems, we formulate Eq. (1) in the stochastic variational method (SVM) [6, 7, 22{28]. As is well known, the behavior of a fluid can be described by the ensemble of fluid elements. ∗Electronic address: [email protected],[email protected] We thus consider the variation of the trajectory of a fluid 2 element in SVM. A fluid element is an abstract volume Two Fokker-Planck equations for ρ are obtained using element and constituent particles inside of it are assumed the two SDE's independently but these are equivalent. to be thermally equilibrated. As is shown later, however, To satisfy this condition, u±(x; t) should satisfy the con- the derivation of the uncertainty relation is irrelevant to sistency condition, the assumption of the local thermal equilibrium. For the sake of simplicity, we thus identify a fluid element with a u+(x; t) = u−(x; t) + 2νr ln ρ(x; t) : (5) constituent particle in the following discussion. See Ref. [7] for more details on the uncertainty relation for fluid See Ref. [7] for details. The consistency condition corre- elements. sponds to the commutation relation in quantum mechan- In SVM, the viscous and quantum potential terms are ics as discussed later. It is also noteworthy that a similar induced through the fluctuations of constituent particles. condition plays an important role in the derivation of the Then the trajectory of a constituent particle is supposed modified NSF equation called bivelocity hydrodynamics to be given by the forward stochastic differential equation [6, 7, 30{32]. (SDE), We apply SVM to the classical Lagrangian, p M dr2 " dbr(t) = u+(br(t); t)dt + 2νdWc(t) (dt > 0) : L (r; dr=dt) = − V − ; (6) cla 2 dt ρ The second term on the right-hand side represents the noise of Brownian motion. We used (b) to denote where " is an internal energy density given by a func- stochastic variables and dAb(t) = Ab(t + dt) − Ab(t) for tion of the particle distribution and the entropy density. an arbitrary Ab(t). The standard Wiener process is de- Applying the classical variation, this Lagrangian gives scribed by Wc(t) which satisfies the Euler equation [7]. As mentioned before, the vis- cous and quantum potential terms are induced through i j 0 E[dWc(t)] = 0 ; E[dWc (t)dWc (t )] = jdtj δt t0 δij ; (2) the fluctuating trajectory in SVM and hence quantum viscous hydrodynamics (1) is obtained by applying SVM where E[ ] denotes the ensemble average for the Wiener to this Lagrangian. To find the corresponding stochastic process. Note that u (r(t); t) is stochastic because of + b Lagrangian, we have to replace d=dt with D and D r(t), but u (x; t) is a smooth function. The field u (x; t) + − b + + in Eq. (6). Due to this ambiguity in the replacement, is associated with the velocity of constituent particles. we introduce two real parameters α and α . Then the SVM determines its form by applying the variational A B stochastic Lagrangian is defined by principle. The noise intensity ν controls the stochasticity of the trajectory. Lsto(r; D+r; D−r) The standard definition of velocity is not applicable b b b M D r " because the left and right-hand limits of the inclination = (D r; D r)M +b − V − ; (7) +b −b D r do not agree in the stochastic trajectory. To distinguish 2 −b ρ this difference, we consider the backward time evolution with of the trajectory described by the backward SDE, p 1 + α 1 + α 1 − αB M = 2 A 2 B 4 2 : dbr(t) = u−(br(t); t)dt + 2νdWc(t) (dt < 0) ; 1 αB 1 1 4 − 2 2 − αA 2 + αB where Wc(t) satisfies the same correlation properties as In the vanishing limit of ν,D± coincide with d=dt and Eq. (2) using jdtj = −dt. The field u−(x; t) is associated with the velocity backward in time. Because of this am- then the stochastic Lagrangian (7) is reduced to the cor- biguity of velocity, Nelson introduced two different time responding classical one (6) independently of αA and αB. derivatives [29]: one is the mean forward derivative D The parameters αA and αB are absorbed into the defini- + tions of κ and η as shown later in Eq. (11). and the other the mean backward derivative D−, which are defined by In the classical variation, a trajectory is entirely deter- mined for a given velocity. This is however not the case br(t + dt) − br(t) with SVM due to the noise terms in the two SDE's. Thus D±r(t) = lim E r(t) = u±(r(t); t) : b dt!0± dt b b only the averaged behavior of the stochastic Lagrangian (3) is optimized by variation. The action is then Here the expectation value is the conditional average for Z tf fixing br(t) and we used that br(t) is Markovian. When I [r] = dt E[L (r; D r; D r)] ; (8) these are operated to a function of r(t), we find sto b sto b +b −b b ti 2 D±f(br(t); t) = @t + u±(x; t) · r ± νr f(x; t) ;(4) with an initial time (ti) and a final time (tf ).

Load more

  7

Copy Link