The Transition Problem between Time-Independent Motions of a Body in a Viscous Liquid
Giovanni P. Galdi, Toshiaki Hishida · 2025
Read The Transition Problem between Time-Independent Motions of a Body in a Viscous Liquid by Giovanni P. Galdi, Toshiaki Hishida (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-00571.txt.
Introduction
Giovanni Galdi and Toshiaki Hishida study a body switching gaits in a thick liquid in a 2025 maths paper. The body first moves by steady translation through the fluid. At time zero it smoothly shifts to a new rigid motion. The shift completes by time one in their setup. They ask if the flow links the two steady states. For small enough speeds, the answer is yes. A unique solution of the Navier-Stokes problem joins both states. The result settles the transition problem in this slow regime.
Catalog record
| CREATOR | Giovanni P. Galdi, Toshiaki Hishida |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 61 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-00571
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
The Transition Problem between Time-Independent Motions of a Body in a Viscous Liquid Giovanni P. Galdi, Toshiaki Hishida Abstract A body ℬ \mathscr{B} moves in an unbounded Navier-Stokes liquid by time-independent translatory motion. Suppose that at time t = 0 t=0 , ℬ \mathscr{B} smoothly changes its motion to an arbitrary rigid motion, reached at time t = 1 t=1 . We then show that the associated Navier-Stokes problem has a unique solution connecting the two steady-states generated by the motion of ℬ \mathscr{B} , provided all the involved velocities of ℬ \mathscr{B} are sufficiently small. Mathematics Subject Classification. 35Q30, 76D05. Keywords. Navier–Stokes flow past a body, transition problem, steady flow, attainability, asymptotic behavior in time, Oseen evolution operator. 1 Introduction In 1965, Robert Finn posed the question of providing a rigorous mathematical proof that a (finite) body \mathscr{B} , completely immersed in a viscous Navier-Stokes liquid that fills the three-dimensional whole space, \Omega , outside \mathscr{B} , can be accelerated from rest to a state of motion characterized by a constant translational velocity \gamma [ 5 ] . This is the so-called “starting problem.” Basically, it consists in showing the convergence of nonstationary motions of the liquid, generated by the acceleration of \mathscr{B} , to the time independent (steady-state) motion corresponding to \gamma . The main difficulty in answering this question is due to the fact that the classical ( L^{2} ) energy estimate –fundamental in establishing global-in-time results– is not available in such a case, because \Omega is an exterior domain. It suggests, instead, that one should resort to a suitable L^{q} approach, with q\geq 3 . An extended theory of this kind was successfully obtained by Shibata about thirty years later [ 25 ] (see also [ 21 ] ), which, short after, led to the complete solution of Finn’s problem, at least for a “small” \gamma [ 9 ] . 1 1 1 It is highly unlikely that the problem can be solved for “large” \gamma , unless one adds, for example, appropriate control forces on the body. Over time, the results in [ 9 ] have been improved and generalized in several respects; see, e.g., [ 10 , 18 , 24 , 28 , 29 , 30 ] . Of particular relevance to our investigation is the question addressed by Sazonov in [ 24 ] . Specifically, he is interested in the more general situation where the initial state of \mathscr{B} is not necessarily rest, but a generic translatory motion with velocity \gamma_{0} (say). Sazonov refers to this as “transition problem.” In [ 24 , Theorems 3.1 and 3.3] , he is able to solve this problem, among others, provided that \gamma_{0} and \gamma are parallel, |\gamma-\gamma_{0}| is “small” and the steady-state corresponding to \gamma is stable, in the sense of spectral theory. However, the crucial point that makes such a result rather unsatisfactory is that it is proved under the further assumption \Omega\equiv\mathbb{R}^{3} , that is, in the absence of the body. Objective of this paper is to give a positive answer to the transition problem, in a somewhat general formulation. More specifically, we assume that the initial state is generated by \mathscr{B} that translates with constant velocity \gamma_{1} , while, in the final state, \mathscr{B} is allowed to perform an arbitrary rigid motion with velocity {\sf V} , characterized by (constant) translation, \gamma_{2} , and rotation, \omega_{0} . No assumptions are made about the direction of \gamma_{i} , i=1,2 , and \omega_{0} , but only about their magnitude which is assumed to be less than a given constant. Let v_{i} , i=1,2 be the (velocity fields of the) steady-state solutions corresponding to \gamma_{1} and {\sf V} , respectively. We then show that, in a suitable function class, there exists a unique solution whose velocity field coincides with v_{1} at time t=0 and, as t\to\infty , converges to v_{2} , with a distinct order of decay. The starting point of our analysis is to rewrite the original set of equations as a single integral equation in an appropriate function space (weak- L^{3} space); see ( 3.11 )–( 3.12 ). This is done employing the evolution operator T=T(t,s) constructed in [ 14 ] . It is then easy to recognize that, formally, the transition problem is equivalent to finding sufficient conditions that guarantee that the solutions to this integral equation converge to 0 as t\to\infty . To make this argument rigorous, we begin to give a precise definition of global (in time) solution (see Definition 3.1 ). Successively, in Theorem 3.1 , we prove that, if the data \gamma_{i} , i=1,2, and \omega_{0} are below a certain constant involving the initial acceleration of \mathscr{B} , then solutions exist, are unique, depend continuously upon the data and decay to 0, as t\to\infty in different norms, with corresponding (algebraic) decay rate; see ( 3.25 ). The proof of this result is obtained by the contraction mapping theorem. For its success, we need, basically, three fundamental estimates: the first one, regarding the “forcing term” g , the second one involving linear terms with coefficients depending on v_{2} , and the third one related to the nonlinear term. The evaluation of the first two terms can be deduced from the known results on the large time behavior of the evolution operator T(t,s) proved in [ 10 , 15 , 16 ] and recalled in Propositions 4.1 and 4.2 , provided g and v_{2} meet appropriate global summability properties along with corresponding estimates with regard to the data; see Lemmas 5.1 and 5.2 . Such properties are indeed shown in Lemma 4.1 under the given assumption, namely, that \mathscr{B} goes from an initial translatory motion to an arbitrary rigid motion. It is worth emphasizing that similar estimates are not known if, in the initial state, the angular velocity of \mathscr{B} is not zero, in which case the transition problem remains open; see Remark 4.1 . We wish also to remark that the summability properties of g and their dependence on the data may as well affect the asymptotic behavior in time of the solution; see Remark 3.1 . The estimate of the term in the integral equation involving the nonlinearity (see ( 5.2 )) is performed in Lemma 5.1 , also with the help of the results reported in Propositions 4.1 and 4.2 . Our objective here is twofold. On the one hand, to show quadratic bounds, in suitable norms, that allow the use of a contraction-mapping argument. On the other hand, to prove, again in suitable norms, algebraic decay estimates in time. Finally, with the help of Lemma 5.3 we provide the uniqueness property. The paper is organized as follows. In Section 2 we furnish the precise mathematical formulation of our transition problem. In the following Section 3 , after introducing some standard notation, we show that the problem can be equivalently rewritten as an integral equation. We thus give the definition of a solution to this equation and state our main results in Theorem 3.1 . We also furnish a few remarks about this theorem and some of its consequences. Then, In Section 4 , we collect several preparatory results, mostly, from [ 10 , 15 , 16