Abstract
We consider a flow of non-Newtonian incompressible heat conducting fluids with dissipative heating. Such system can be obtained by scaling the classical Navier–Stokes–Fourier problem. As one possible singular limit may be obtained the so-called Oberbeck–Boussinesq system. However, this model is not suitable for studying the systems with high temperature gradient. These systems are described in much better way by completing the Oberbeck–Boussinesq system by an additional dissipative heating. The satisfactory existence result for such system was however not available. In this paper we show the large-data and the long-time existence of dissipative and suitable weak solution. This is the starting point for further analysis of the stability properties of such problems.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction
The Rayleigh–Bénard problem of thermal convection is one of the most canonical examples of instability in the fluid flow and it has attracted lot of attention not only in the modelling and physical understanding of such phenomenon but also in the mathematical community (e.g. [9, 14, 16]). Indeed, this became a source of difficult mathematical problems (e.g. [17, 20]) and has driven the study of the so-called singular limits during last decades [6]. Such singular limits may have many forms depending on the scaling, and the most classical model (the asymptotic limit) is the so-called Oberbeck–Boussinesq system, that is used exactly for Rayleigh–Bénard convection. This is the case when the fluid is heated below with the bottom temperature \(\theta ^{b}\) and with prescribed lower temperature \(\theta ^{t}\) on the top of two parallel plates and when the fluid can be understood as mechanically incompressible and all changes in the density are just because of variations in the temperature. It appears that such model is the very accurate approximation of real fluid in case that the temperature gradient and consequently \((\theta ^b - \theta ^t\)) is not large. On the other hand, for large temperature gradient, the Oberbeck–Boussinesq may significantly fail in giving the correct predictions and therefore there are many attempts how to generalize the approximative model. One of such attempt is to add a dissipative heating into the system. This approach was successfully used in [21, 28], where the authors formally derived a generalization of the Oberbeck–Boussinesq system in the following form
Here, \(\text {Gr}\) is the Grashof number, \(\text {Pr}\) denotes the Prandtl number, and the new additional scaling parameter is the dissipative number \(\text {Di}\). Further, \(\textbf{v}\) means the velocity field, the body force is just the gravity, i.e. it has direction in the d-th component and is constant \(\textbf{f}:=(-1)\textbf{e}_d\) and the auxiliary functions \(\Theta\) scales the difference between the temperature on the top and on the bottom of plates
Here \(\theta\) is the real physical temperature. Note here that assuming \(\text {Di}=0\), we arrive to the classical Oberbeck–Boussinesq system without dissipative heating. This system with \(\text {Di}=0\) is rigorously derived and treated in [6] as the singular limit of compressible Navier–Stokes–Fourier system with certain nonlocal boundary conditions. The existence analysis for such problem is then established in [4]. It should be mentioned here that all considered models in [4, 6] are Newtonian, i.e. the Cauchy stress \(\mathbf{S}\) is linear with respect velocity gradient.
1.1 Beyond Oberbeck–Boussinesq system
Here, in this paper, we want to deal with \(\text {Di}>0\) and with possibly nonlinear \(\mathbf{S}\), and as a starting point for further analysis, we want to establish the existence of a solution. To simplify the notation and also computation (but without any essential impact on the analysis), we denote a new unknown \(\vartheta :=\theta +\Theta\), consider general body force \(\textbf{f}\) and scale the equations such that \(\text {Gr}=\text {Pr}=\text {Di}=1\) and we obtain the final system
which is supposed to be satisfied in \(Q:=(0,T) \times \Omega \subset (0, +\infty )\times {{\mathbb {R}}}^d\) with \(\Omega\) being a Lipschitz domain. Note here, that (1a) is the incompressibility constraint, (1b) is the balance of linear momentum and (1c) is the balance of internal energy. Here \(\textbf{v}: Q\rightarrow {{\mathbb {R}}}^d\) denotes the velocity field, \(\textbf{D}:= (\nabla \textbf{v}+ (\nabla \textbf{v})^{T})/2\) is the symmetric part of the velocity gradient \(\nabla \textbf{v}\), \(\pi :Q\rightarrow {{\mathbb {R}}}\) is the pressure, \(\vartheta :Q\rightarrow {{\mathbb {R}}}\) is the temperature; \(\textbf{S} : Q\rightarrow {{\mathbb {R}}}^{d\times d}_\textrm{sym}\) denotes the viscous part of the Cauchy stress tensor and \(\textbf{q}:Q \rightarrow {{\mathbb {R}}}^d\) is the heat flux.
1.2 Cauchy stress tensor and heat flux
The heat flux \(\textbf{q}\) is represented by the Fourier law
with the heat conductivity \(\kappa : {{\mathbb {R}}}\rightarrow (0, +\infty )\) being a continuous function of the temperature satisfying, for all \(\vartheta \in (0, +\infty )\) and for some \(0<\underline{\kappa }\le \overline{\kappa } <+\infty ,\)
We also assume that the Cauchy stress is given as \(\textbf{S} =\textbf{S} ^*(\vartheta , \textbf{D}\textbf{v})\), where \(\textbf{S} ^*: (0,\infty )\times {{\mathbb {R}}}^{d\times d}_\textrm{sym} \rightarrow {{\mathbb {R}}}^{d\times d}_\textrm{sym}\) is a continuous mapping fulfilling for some \(p>2d/(d+2)\), some \(0< \underline{\nu }, \overline{\nu }<+\infty\) and for all \(\vartheta \in {{\mathbb {R}}}_+\), \(\textbf{D}_1,\textbf{D}_2 \in {{\mathbb {R}}}^{d\times d}_\textrm{sym}\)
The prototypic relation \(\textbf{S} \sim \nu (\vartheta ) |\textbf{D}\textbf{v}|^{p-2}\textbf{D}\textbf{v}\) falls into the class (4).
1.3 Boundary and initial conditions
The system (1a)–(1c) is completed by the following initial conditions
We prescribe the following Navier’s slip (\(\alpha \ge 0\)) boundary condition for the velocity
and the Neumann type boundary condition for the temperature
Here, \(\textbf{n}\) is the unit outward normal and \(\textbf{v}_{\tau }\) stands for the projection of the velocity field to the tangent plane, i.e. \(\textbf{v}_{\tau }:= \textbf{v}-(\textbf{v}\cdot \textbf{n})\textbf{n}\). The first condition in (6) expresses the fact that the solid boundary is impermeable, the second condition in (6) is Navier’s slip boundary condition, and the condition in (7) states that there is no heat flux across the boundary. In fact, this is not the correct boundary condition and one should consider here either the Dirichlet boundary condition or inspired by [6] some version of nonlocal boundary condition. However, since we want to present the first large-data result for the problems with dissipative heating, we assume the simplest boundary conditions. However, we are sure that the similar result can be obtained for more realistic boundary conditions and it will be a part of the forthcoming paper about the stability, where the Dirichlet boundary condition plays the crucial role.
2 Definition of solution and main theorem
We assume that the initial data \(\textbf{v}_0, \vartheta _0\) and the external body force \(\textbf{f}\) satisfy
and that
We look for \((\textbf{v},\vartheta ,\pi ):[0,T]\times \Omega \rightarrow {{\mathbb {R}}}^d\times {{\mathbb {R}}}^{+}\times {{\mathbb {R}}}\) solving the following set of equations in \((0,T)\times \Omega\)
completed by a weak formulation of the internal energy inequality
and satisfying the boundary conditions
and the initial conditions
The inequality (11) can be also replaced by (this is usually used in the setting of compressible fluids, where the a priori estimates are not sufficient to define (11) in sense of distributions)
where the entropy \(\eta\) is defined as \(\eta :=\log \vartheta\).
Below we give a precise formulation of the notion of weak solution but before that we introduce some notation that will be needed in what follows.
2.1 Basic definitions and function spaces
Let \(\Omega \subset {{\mathbb {R}}}^d\) be a bounded domain with Lipschitz boundary \(\partial \Omega\), i.e. \(\Omega \in \mathcal {C}^{0,1}\). We say \(\Omega \in \mathcal {C}^{1,1}\) if the mappings that locally describe the boundary \(\partial \Omega\) belong to \(\mathcal {C}^{1,1}\).
We consider the standard Lebesgue, Sobolev and Bochner spaces endowed with the classical norms. For our purposes, we introduce for arbitrary \(q\in [1,\infty )\) the subspaces of vector-valued Sobolev functions given by
and
Similarly, we consider the classical Sobolev space \(W^{1,q}(\Omega )\) and use the standard abbreviation for its dual space \(W^{-1,q'}:=(W^{1,q}(\Omega ))^*\). Also, in what follows whenever there is \(v\in X^*\), \(u\in X\), the symbol \(\langle v,u\rangle\) means the duality paring in X. In case there was possible ambiguity, we would write \(\langle v,u\rangle _X\). Notice that all above mentioned space are Banach spaces that are in addition separable provided that \(p,q <\infty\). In addition, they are reflexive whenever \(p,q\in (1,\infty )\). Further, we introduce few inequalities used and needed in the text. Since we deal only with the symmetric gradient, we need some form of the Korn inequality. Since we want to deal with general boundary conditions, we use the following form (see [12, Lemma 1.11] or [5, Theorem 11])
which is valid for all \(p\in (1,\infty )\) provided that \(\Omega\) is Lipschitz. Further, we also frequently use in the paper the following interpolation inequality
2.2 Definition of weak and suitable weak solutions and main theorem
Here, we introduce the notion of weak and suitable weak solution to (9)–(12). We consider only the dimension \(d=3\).
Definition 1
(Weak solution) Let \(\Omega \subset {{\mathbb {R}}}^3\) be a bounded domain of class \(C^{1,1}\) and let (0, T) with \(T>0\) be the time interval. Let \(\textbf{v}_0,\vartheta _0\) and \(\textbf{f}\) be given functions satisfying (8) and let p be given in the interval \((6/5, +\infty )\). We say that a triplet \((\textbf{v},\vartheta ,\pi )\) is a weak solution to the problem (9)–(12) if
fulfills the following weak formulations: The linear momentum equation (9) is satisfied in the following sense
for any \(\varvec{\varphi }\in \mathcal {C}_0^\infty ([0, T); W_{\textbf{n}}^{1, q}\cap L^{\infty }(\Omega ))\cap L^2((0,T)\times \partial \Omega )\) with \(q= \max \{p, \frac{5p}{5p-6}\}\); The global energy inequality holds in the following sense
for any nonnegative \(\varphi \in \mathcal {C}_0^\infty ([0,T))\);
The entropy inequality (13) is satisfied as
for any nonnegative \(\varphi \in \mathcal {C}_0^\infty ([0, T); W^{1,\infty }({\Omega }))\). The initial conditions are attained in the following sense
The above definition fulfills the basic assumption on the consistency, i.e. if we have a weak solution that is in addition smooth then it is also the classical solution, we refer here e.g. to the classical book [18], where such approach is justified. On the other hand, if we want to study further properties of the solution, for example the stability, we usually require more refined notion of the solution, namely the suitable weak solution. However, it also requires more assumption on the growth parameter p.
Definition 2
[Suitable weak solution] Let \(\Omega \subset {{\mathbb {R}}}^3\) be a bounded domain of class \(C^{1,1}\) and let (0, T) with \(T>0\) be the time interval. Let \(\textbf{v}_0,\vartheta _0\) and \(\textbf{f}\) be given functions satisfying (8) and let \(p\in (9/5, +\infty )\) be given. We say that a triplet \((\textbf{v},\vartheta ,\pi )\) is a suitable weak solution to the problem (9)–(12) if Definition 1 is satisfied with (23) replaced by
which is valid for any \(\varphi \in \mathcal {C}_0^\infty ([0,T);W^{1,\infty }({\Omega }))\). Moreover, we require that (11) is satisfied in the following sense
for any nonnegative \(\varphi \in \mathcal {C}_0^\infty ([0,T);W^{1,\infty }({\Omega }))\).
Next, we formulate the main theorem of this paper.
Theorem 1
Let \(\Omega \subset {{\mathbb {R}}}^3\) be a bounded domain with \(\mathcal {C}^{1,1}\) boundary. Assume that \(\textbf{S} ^*\) and \(\kappa\) satisfy (3) and (4) with \(p>6/5\). Then for any data \(\textbf{v}_0,\vartheta _0, \textbf{f}\) fulfilling (8), there exists a weak solution to (9)–(12) in the sense of Definition 1. Moreover, if \(p>8/5\) then (23) holds with the equality sign. In addition, if \(p>9/5\) then there exists a suitable weak solution in sense of Definition 2. Furthermore, if \(p\ge 11/5\), then (24) and (27) holds with equality sign and the following is true
Note that in above Theorem 1 we assume a stronger assumption on the boundary, namely \(\Omega \in \mathcal {C}^{1,1}\). The reason is the necessity of having a priori estimates of the pressure \(\pi\) that appears in (23) and cannot be omitted by using divergence-free functions as test functions as it is usual in studying Navier–Stokes equations without the temperature. At this, we would like to discuss the main novelty of the paper. It seems that the only relevant existence result are due to [25, 26], where the authors treated the same system but with \(\textbf{S} ^*\) being linear with respect to the velocity gradient, i.e. the case \(p=2\), so the result of this paper is much more general. Second, in [25], the authors treated only the steady case and in [26], the authors were not able to show the validity of (26), i.e. they did not show the existence of a suitable weak solution, which is the main weak point in their result. We also refer to [22, 23], which is an extension of [25] to more general boundary conditions, but deal only with the steady case. In our setting, we are able to prove the existence of a suitable weak solution. In addition, for \(p\ge 11/5\), we obtain the internal energy equality. Furthermore, in spirit of results [1,2,3], we see that the existence result obtained in this paper is the starting point for studying the stability analysis for the underlying problem. We would like to remind that there are naturally appearing numbers like 6/5, 8/5, 9/5, 11/5, which are dictated by the nature of the problem, and these borderline are usual in the theory for non-Newtonian models of heat conducting incompressible fluids. Adding the dissipative heating to the system does not bring any change in these borderlines. The only change, but rather essential, is the way how the uniform estimates are obtained, which makes the result highly nontrivial extension of works [1, 2] (see also further references therein). In addition, (28) as well as energy equalities valid for \(p\ge 11/5\) seem to be the essential assumption to obtain the stability result for arbitrary weak solution, see [2, 3] or [1] where the same system but without dissipative heating and in dimension two is treated.
The proof is split into several steps. In Sect. 3, we introduce an approximation, where the nonlinear convective terms are truncated by an auxiliary cut-off function. The existence of solutions for the k-approximation, is for the sake of completeness and clarity included in Appendix Appendix A. Then, in Sect. 4.1, we derive estimates that are uniform with respect to k-parameter. Finally, letting \(k\rightarrow +\infty\) in Sect. 4.2, we complete the proof of Theorem 1.
3 Definition of approximating systems and their solutions
We start this part with definition of auxiliary cut-off functions. For any arbitrary natural number \(k\ge 1\), we define
and a function \(g_k:\mathbb {R}^{+}\rightarrow [0,1]\) such that it is continuous and satisfies
Finally, we introduce an auxiliary function \(\eta\), which is used in the proof of attainment of the initial data. Let \(T > 0\) be given, and let \(0 < \varepsilon \ll 1\) and \(t \in (0, T-\varepsilon )\) be arbitrary. Consider \(\eta \in \mathcal {C}^{0,1}([0, T])\) as a piece-wise linear function of three parameters, such that
We define an approximative problem \(\mathcal {P}^{k}\) (for simplicity we write \((\textbf{v},\pi ,\vartheta )\) instead of \(\textbf{v}^{k},\pi ^{k},\vartheta ^{k}\)) such that we truncate the convective term (in order to be able to use the Minty method), we truncate the source term (in order to have proper estimates at the beginning) and we also modify the boundary conditions (to avoid problem with low integrability). More precisely, we consider the problem:
in \((0,T)\times \Omega\) complemented with the boundary conditions
and the following initial conditions
For this problem, we have the following existence result, which is formulated in any dimension d. Note that the result is dimension-independent due to the presence of truncation functions.
Lemma 1
Let \(\Omega \subset {{\mathbb {R}}}^d\) be a bounded domain with \(\mathcal {C}^{1,1}\) boundary. Assume that \(\textbf{S} ^*\) and \(\kappa\) satisfy (3) and (4) with \(p>2d/(d+2)\). Then for any \(k\in \mathbb {N}\) and any data \(\textbf{v}_0,\vartheta _0, \textbf{f}\) fulfilling (8), there exists a triplet \((\textbf{v},\vartheta ,\pi )=(\textbf{v}^k,\vartheta ^k,\pi ^k)\) satisfying
and
attaining the initial conditions (36) in the following sense
satisfying equation (33) in the following sense: for any \(\varvec{\varphi }\in W^{1,p}_{\textbf{n}}\) and for a.a. \(t\in (0, T)\) there holds
and satisfying (34) in the following sense: for any \(f\in \mathcal {C}^2({{\mathbb {R}}})\) satisfying \(f''\in \mathcal {C}_0({{\mathbb {R}}})\), for any \(\varphi \in W^{1,2}(\Omega )\cap L^\infty (\Omega )\) and for a.a. \(t\in (0, T)\) there holds
Proof
The complete proof is presented at the Appendix Appendix A for the most important case \(d=3\). For other dimensions the proof is however almost identical. \(\square\)
We would like to emphasize here, that the above existence result is in fact very strong. Although the equation (33) is satisfied in the classical weak sense (45), the equation (33) is satisfied in the renormalized weak sense as (46). This enables us to deduce the proper uniform estimates rigorously.
4 Limit in the approximating system
In the previous section we established the existence of a weak solution to the k-approximating system (32)–(34). Key k-uniform estimates and the limits as \(k\rightarrow +\infty\) are derived in this section. We focus only on dimension \(d=3\), the proof for \(d=2\) is in fact even easier.
4.1 Uniform estimates
For \((\textbf{v},\vartheta ,\pi )=(\textbf{v}^k,\vartheta ^k,\pi ^k)\) we derive estimates that are uniform with respect to k-parameter (the relevant quantities are then bounded by a generic constant C, where \(C(\Vert \textbf{f}\Vert _{\infty }, \Vert \textbf{v}_0\Vert _2, \Vert \vartheta _0\Vert _1, \Vert \log \vartheta _0\Vert _1)\)).
We set \(\varvec{\varphi }=\textbf{v}\) in (45) and in (46) we set \(\varphi =1\) and \(f(\vartheta )=\vartheta \). Summing both identities and using the fact that \({\text {div}}\textbf{v}= 0\) to eliminate convective termsFootnote 1 we deduce
Integration with respect the time variable then leads to
Next, for any \(\varepsilon >0\) we set \(f(\vartheta ):= \log (\vartheta + \varepsilon )\) and \(\varphi := 1\) in (46), to deduce
Integrating this identity over the time interval (0, t), it follows
and taking the supremum over \(t\in (0, T)\), we have
Then employing (47), the fact that \(\log \vartheta _0\in L^1(\Omega )\) and taking the limit as \(\varepsilon \rightarrow 0\) we get the following k-independent estimate
In order to improve the uniform bound for the temperature, we fix arbitrary \(\sigma \in (0, 1)\) and set \(f(\vartheta )=\vartheta ^{\sigma }\) and \(\varphi =1\) in (46) to obtain
Integrating it over time and using the fact that \(\sigma \in (0,1)\) and the uniform estimate (47), we deduce
To bound the term on the right hand side, we recall the interpolation inequality (here we consider \(d=3\))
and using also the Hölder inequality, the a priori bound (47) and the assumption (3), we get the estimate
Collecting (49) and (50) we deduce the following estimates that are uniform with respect to \(k\in {\mathbb {N}}\)
Consequently, we deduced that for arbitrary \(\sigma \in (0,1)\) there holds
Next, we derive the final estimates for \(\vartheta \). Using the interpolation inequality
on the function \(z:=\vartheta ^{\sigma }\), with \(\sigma :=\frac{3q}{5}\) for \(q<5/3\), we obtain from (52) that
Similarly, combining (51) and (54) and using the Hölder inequality, we get
The last bound can be viewed by setting
Finally, we use that \(\vartheta ^{\frac{\sigma }{2}}\) is uniformly bounded in \(L^2(0, T; L^6(\Omega ))\) and consequently \(\vartheta ^\sigma \) is uniformly bounded in \(L^1(0, T; L^3(\Omega ))\), combine it with the interpolation inequality (valid for some \(\sigma \in (0,1)\) and \(\lambda \in (0,1)\))
and use the uniform estimates (47) and (52), we see that
Having the estimate (56), we can now proceed with further bounds on the velocity field. Setting \(\varvec{\varphi }:=\textbf{v}\) in (45) and integrating in time, it yields that
As consequence it follows from (4b) that
The interpolation inequality
together with the Korn inequality (14) and the uniform estimates (57), (58) ensure that
We finish this part by introducing the estimates on the pressure. It follows from (45) that for all \(\varphi \in W^{2,p'}(\Omega )\) satisfying \(\nabla \varphi \cdot \textbf{n} =0\) on \(\partial \Omega \), and for almost all time \(t\in (0,T)\) there holds (see also (A.59))
In addition, recall that
Then, we use the fact that \(\Omega \in \mathcal {C}^{1,1}\) and the theory for Laplace equation (see also [10,11,12,13] for details) and obtain that
Applying the \(z'\)-power and integrating the result over (0, T) we finally get
where the last bound follows from the estimates (47), (51), (58) and (59).
Finally, we recall the estimates on the time derivative. By using the very classical procedure, we can deduce from (45) with the help of above uniform estimates (47), (51), (58) and (59) that
where \(z'\) is defined in (60). Similarly, considering (46) with \(f(s):=\log s\), we can use the above estimates (47), (51), (58) and (59) to observe that for any \(\omega >5\), we have (see e.g. [13] for similar estimate)
4.2 Limit as \(k\rightarrow +\infty\)
Let us consider \(\varvec{\varphi }\in C_0^\infty ([0, T); W_{\textbf{n}}^{1, q}\cap L^{\infty }(\Omega ))\cap L^2((0,T)\times \partial \Omega )\) with \(q= \max \{p, \frac{5p}{5p-6}\}\) in (45), integrate it over the time interval (0, T), then after the integration by parts in the time derivative term we get
where we abbreviate \(\textbf{S} ^k=\textbf{S} ^*(\vartheta ^k, \textbf{D}\textbf{v}^k).\) Next, we consider \(\varphi \in \mathcal {C}_0^\infty ([0,T);\mathcal {C}^{\infty }({\Omega }))\) in (46), integrate it over the time interval (0, T), and after the integration by parts with respect to the time variable, we get
Now, we make two special choices of f, namely we consider \(f(s)=s\) and then \(f(s)=\log s\) (such choices can be rigorously justified taking the mollification with compactly supported functions). With the first choice, we deduce
With the second choice, we also use the abbreviation \(\eta ^k=\log \vartheta ^k\) and we see that
Finally, to get the energy identity, we consider \(\varphi \in \mathcal {C}^\infty ((0,T)\times \Omega )\) fulfilling \(\varphi (T)=0\) and use \(\textbf{v}^k \varphi\) and \(\varphi\) as test functions in (63) and (65) respectively. Doing this and then taking the sum of the outcome we deduce that (using also the fact that \({\text {div}}\textbf{v}^k=0\) and integration by partsFootnote 2)
where \(\mathcal {G}_k\) is such that \(\mathcal {G}_k'(s)=g_k(s)\). In particular, for any \(\varphi \in \mathcal {C}_0^\infty ([0,T))\), i.e. \(\varphi\) independent of spatial variable, there holds
We want to discuss the limit in formulations (63) and (65)–(68). By virtue of the uniform estimates (57), (58) we can extract a subsequence \((\textbf{v}^k,\vartheta ^k, \pi ^k, \textbf{S} ^k, \eta ^k)\) such that the following convergence results hold
Moreover, employing the Aubin–Lions compactness Lemma we deduce that
Consequently, we have that \(\eta =\log \vartheta\) and \(\overline{(\vartheta )^{\sigma }}=(\vartheta )^{\sigma }\). The above strong and weak convergence results are sufficient to pass to the limit in most term.
However, it is not sufficient for identification of \(\textbf{S} =\textbf{S} ^*(\vartheta , \textbf{D}\textbf{v})\). This identification follows from the procedure developed in [15], see also [11]. Indeed, there is shown that there exists a nondecreasing sequence of measurable sets \(\{Q_n\}_{n=1}^{\infty }\) fulfilling \(\lim _{n\rightarrow \infty }|Q{\setminus } Q_n|=0\) such that for every \(n\in \mathbb {N}\) we have
Further, using the growth assumption (4b), the convergence result (80), the fact that \(\textbf{D}\textbf{v}\in L^{p'}(Q; {{\mathbb {R}}}^{3\times 3})\) and the Lebesgue dominated convergence theorem, we obtain
Thus, using the monotonicity assumption (4a), the weak convergence result (70) and (82), we observe that for any \(n\in \mathbb {N}\)
Hence, we have that
Thus, it is a simple consequence of (82) and (70) that for every \(n\in \mathbb {N}\)
Finally, using (70), (73) and (84), we have for arbitrary \(\textbf{B}\in L^p(Q; {{\mathbb {R}}}^{3\times 3})\) that
Hence, using the Minty method, i.e. setting \(\textbf{B}:=\textbf{D}\textbf{v}\pm \varepsilon \textbf{C}\) in the above inequality, dividing by \(\varepsilon >0\) and then letting \(\varepsilon \rightarrow 0_+\), we have
Since \(\textbf{C}\) is arbitrary and \(|Q\setminus Q_n|\rightarrow 0\) as \(n\rightarrow \infty\), we observe from the above inequality that \(\textbf{S} =\textbf{S} ^*(\vartheta , \textbf{D}\textbf{v})\).
Having identified the nonlinearity \(\mathbf{S}\), we may now focus on the limiting procedure in desired equations. Indeed, it is now easy to use (69)–(80) and to let \(k\rightarrow \infty\) in (63) to deduce (22). Here, it is essential that \(\textbf{v}^k\) converges strongly in \(L^{2+\varepsilon }(Q)\), which is due to the assumption \(p>6/5\). Next, we let \(k\rightarrow \infty\) in (68). Using (78) and (80), we can pass to the limit in the first term on the left hand side. For the second term on the left hand side we use (79) and the Fatou lemma to obtain the inequality (23). In order to obtain the equality sign in (23), we can use the result in [12], where it is shown thatFootnote 3
provided that \(p>8/5\). Consequently, we obtain (23) with the equality sign.
Next, we focus on the energy and the entropy (in)equalities (24) and (27). To do so, we first show how to pass to the limit with possibly inequality signs in highest order terms. Let us consider arbitrary nonnegative \(\varphi \in L^{\infty }(Q)\). Then using (4a), (4b) and the convergence result (84), we deduce that for any \(n\in \mathbb {N}\)
Consequently, letting \(n\rightarrow \infty\) in the above inequality, we deduce
Similarly, using in addition (80) and the nonnegativity of \(\vartheta ^k\), we deduce
Very similarly, we can use the weak lower semicontinuity, the weak convergence (77), the pointwise convergence of \(\vartheta ^k\) (80) and the boundedness of \(\kappa (\cdot )\), see (3), there holds
These convergence results are sufficient to take the limit in (66) and to obtain (24). Then we can proceed to the limit in the equation (65) to deduce (27), provided we show
To show the above convergence result, we recall the strong convergence results (78) and (80). Thus, it is sufficient to show that \(\vartheta ^k \textbf{v}^k\) is uniformly bounded in \(L^{1+\varepsilon }(Q)\) for some \(\varepsilon >0\). Using the Hölder inequality and the classical interpolation, we have (here \(\sigma \in (0,1)\) will be specified later, but typically it will be almost equal to one)
provided that (we are using the uniform bounds coming from (70) and (75))
Note that if
then we can always find \(\sigma \in (0,1)\) such that the inequality (90) is satisfied. As a consequence, we deduce (89). Thus, we may proceed to the limit also in (65) to obtain (27).
Next, we want to let \(k\rightarrow \infty\) in (67) to get (26). We already discuss almost all terms except the one \(|\textbf{v}^k|^2\textbf{v}\). Thus, for identification of the limit also in this term, we need that \(\textbf{v}^k\) is compact at least in \(L^3(Q)\). Comparing it with the strong convergence result (78), we see that the condition \(p>9/5\) is exactly the one guaranteeing the compactness of the velocity in \(L^3(Q)\).
Finally, in case that \(p\ge 11/5\), one may follow the standard monotone operator theory and to conclude that (84) holds true even in \(L^1(Q)\) (and not only in a subset \(Q_n\subset Q\)). Thus, we have the weak convergence on the whole set Q and therefore we are able to identify the limits in terms containing \(\textbf{S} ^k:\textbf{D}\textbf{v}^k\) with equality signs and consequently, we may consider equalities in (24) and (27). The relation (28) is proved in the same way as (43), see the computation following (A.57).
We also omit the proof of attaining of the initial conditions (25) and refer the reader e.g. to [13]. The proofs of both main theorems are thus complete.
Data availability
This declaration is not applicable.
Notes
Compare it with very similar computations in Appendix.
To evaluate the convective term we proceed as follows
$$\begin{aligned} \begin{aligned}&\int _{\Omega } g_k(|\textbf{v}^k|^2)\, (\textbf{v}^k\otimes \textbf{v}^k): \nabla (\textbf{v}^k \varphi ) \ \,\textrm{d} {x}\\&=\int _{\Omega } \varphi g_k(|\textbf{v}^k|^2)\, (\textbf{v}^k\otimes \textbf{v}^k): \nabla \textbf{v}^k +g_k(|\textbf{v}^k|^2)\, (\textbf{v}^k\otimes \textbf{v}^k): (\textbf{v}^k \otimes \nabla \varphi ) \ \,\textrm{d} {x}\\&\quad =\frac{1}{2}\int _{\Omega } \varphi g_k(|\textbf{v}^k|^2)\, \textbf{v}^k \cdot \nabla |\textbf{v}^k|^2 +2g_k(|\textbf{v}^k|^2)\, |\textbf{v}^k|^2 \textbf{v}^k \cdot \nabla \varphi \ \,\textrm{d} {x} \\&=\frac{1}{2}\int _{\Omega } \varphi \textbf{v}^k \cdot \nabla \mathcal {G}_k(|\textbf{v}^k|^2) +2g_k(|\textbf{v}^k|^2)\, |\textbf{v}^k|^2 \textbf{v}^k \cdot \nabla \varphi \ \,\textrm{d} {x}\\&\quad =\int _{\Omega } \left( g_k(|\textbf{v}^k|^2)\, |\textbf{v}^k|^2 -\frac{\mathcal {G}_k(|\textbf{v}^k|^2)}{2}\right) \textbf{v}^k \cdot \nabla \varphi \ \,\textrm{d} {x} \end{aligned} \end{aligned}$$It follows from the following interpolation (assuming \(p\le 2\), for \(p>2\) it is obvious)
$$\begin{aligned} \int _0^T \int _{\partial \Omega } |\textbf{v}|^2 \le C\int _0^T \Vert \textbf{v}\Vert ^{\frac{8p-12}{5p-6}}_2 \Vert \textbf{v}\Vert ^{\frac{2p}{5p-6}}_{\frac{5p-6}{2p},2}\,\textrm{d} t \le C\int _0^T \Vert \textbf{v}\Vert ^{\frac{8p-12}{5p-6}}_2 \Vert \textbf{v}\Vert ^{\frac{2p}{5p-6}}_{1,p}\,\textrm{d} t . \end{aligned}$$Since \(\frac{2p}{5p-6}<p\) for \(p>8/5\), we see that the desired compactness follows from the a priori estimates.
The computation goes as follows:
$$\begin{aligned} \begin{aligned}&\int _{\Omega }(\textbf{v}^{n,m} \otimes \textbf{v}^{n,m} \, g_{k}(|\textbf{v}^{n,m}|^2)):\nabla \textbf{v}^{n,m}\,\textrm{d} {x}\\&=\frac{1}{2} \int _{\Omega }g_{k}(|\textbf{v}^{n,m}|^2)\textbf{v}^{n,m} \cdot \nabla |\textbf{v}^{n,m}|^2\,\textrm{d} {x}\\&=\frac{1}{2} \int _{\Omega }\textbf{v}^{n,m} \cdot \nabla G_{k}(|\textbf{v}^{n,m}|^2)\,\textrm{d} {x}=-\frac{1}{2} \int _{\Omega } G_{k}(|\textbf{v}^{n,m}|^2) {\text {div}}\textbf{v}^{n,m} \,\textrm{d} {x}=0, \end{aligned} \end{aligned}$$and \(G_k\) denotes the primitive function to \(g_k\).
It is a consequence of the following computation
$$\begin{aligned} \begin{aligned} \int _{\Omega }\mathcal {T}_k(\vartheta ^{n,m})\textbf{v}^{n,m}\cdot \nabla \vartheta ^{n,m}\,\textrm{d} {x}=\int _{\Omega }\textbf{v}^{n,m}\cdot \nabla \mathfrak {T}_k(\vartheta ^{n,m})\,\textrm{d} {x}=-\int _{\Omega }\mathfrak {T}_k(\vartheta ^{n,m}) {\text {div}}\textbf{v}^{n,m}\,\textrm{d} {x}=0, \end{aligned} \end{aligned}$$where \(\mathfrak {T}_k\) denotes a primitive function to \(\mathcal {T}_k\).
References
Abbatiello A, Bulíček M, Kaplický P (2019) On the existence of classical solution to the steady flows of generalized Newtonian fluid with concentration dependent power-law index. J Math Fluid Mech 21(1):15–22
Abbatiello A, Bulíček M, Kaplický P (2022) On solutions for a generalized Navier–Stokes–Fourier system fulfilling the entropy equality. Philos Trans R Soc A 380(2236):20210351
Abbatiello A, Bulíček M, Kaplický P (2024) On the exponential decay in time of solutions to a generalized Navier–Stokes–Fourier system. J Differ Equ 379:762–793
Abbatiello A, Feireisl E (2023) The Oberbeck–Boussinesq system with non-local boundary conditions. Quart Appl Math 81(2):297–306
Bauer S, Pauly D (2016) On Korn’s first inequality for mixed tangential and normal boundary conditions on bounded Lipschitz domains in R^N. Ann Univ Ferrara Sez VII Sci Mat 62(2):173–188
Bella P, Feireisl E, Oschmann F (2023) Rigorous derivation of the Oberbeck–Boussinesq approximation revealing unexpected term. Commun Math Phys 403(3):1245–1273
Boccardo L, Dall’Aglio A, Gallouët T, Orsina L (1997) Nonlinear parabolic equations with measure data. J Funct Anal 147(1):237–258
Boccardo L, Gallouet T (1996) Summability of the solutions of nonlinear elliptic equations with right-hand side measures. J Convex Anal 3(2):361–365
Bodenschatz E, Pesch W, Ahlers G (2000) Recent developments in Rayleigh–Bénard convection. Annu Rev Fluid Mech 32:709–778
Bulíček M, Feireisl E, Málek J (2009) A Navier–Stokes–Fourier system for incompressible fluids with temperature dependent material coefficients. Nonlinear Anal Real World Appl 10(2):992–1015
Bulíček M, Gwiazda P, Málek J, Świerczewska Gwiazda A (2012) On unsteady flows of implicitly constituted incompressible fluids. SIAM J Math Anal 44(4):2756–2801
Bulíček M, Málek J, Rajagopal KR (2007) Navier’s slip and evolutionary Navier–Stokes-like systems with pressure and shear-rate dependent viscosity. Indiana Univ Math J 56(1):51–85
Bulíček M, Málek J, Rajagopal KR (2009) Mathematical analysis of unsteady flows of fluids with pressure, shear-rate, and temperature dependent material moduli that slip at solid boundaries. SIAM J Math Anal 41(2):665–707
Chandrasekhar S (1961) Hydrodynamic and hydromagnetic stability. International Series of Monographs on Physics. Clarendon Press, Oxford
Diening L, Růžička M, Wolf J (2010) Existence of weak solutions for unsteady motions of generalized Newtonian fluids. Ann Sc Norm Super Pisa Cl Sci 9(1):1–46
Ecke RE, Shishkina O (2023) Turbulent rotating Rayleigh–Bénard convection. Annu Rev Fluid Mech 55:603–638
Foias C, Manley O, Temam R (1987) Attractors for the Bénard problem: existence and physical bounds on their fractal dimension. Nonlinear Anal 11(8):939–967
Feireisl E (2004) Dynamics of viscous compressible fluids. Oxford lecture series in mathematics and its applications, vol 26. Oxford University Press, Oxford
Grisvard P (1985) Elliptic problems in nonsmooth domains, Monographs and studies in mathematics, vol 24. Pitman (Advanced Publishing Program), Boston, MA
Li J, Titi ES (2016) Global well-posedness of the 2D Boussinesq equations with vertical dissipation. Arch Ration Mech Anal 220(3):983–1001
Kagei Y, Růžička M, Thäter G (2000) Natural convection with dissipative heating. Commun Math Phys 214(2):287–313
Kim T (2022) Steady Boussinesq system with mixed boundary conditions including friction conditions. Appl Math 67(5):593–613
Kim T, Cao D (2020) Mixed boundary value problems of the system for steady flow of heat-conducting incompressible viscous fluids with dissipative heating. Methods Appl Anal 27(2):87–124
Málek J, Nečas J, Rokyta M, Růžička M (1996) Weak and measure-valued solutions to evolutionary PDEs, vol 13. Applied mathematics and mathematical computation. Chapman & Hall, London
Naumann J, Pokorný M, Wolf J (2012) On the existence of weak solutions to the equations of steady flow of heat-conducting fluids with dissipative heating. Nonlinear Anal Real World Appl 13(4):1600–1620
Naumann J, Wolf J (2010) Existence of weak solutions to the equations of natural convection with dissipative heating. Advances in mathematical fluid mechanics. Springer, Berlin, pp 367–384
Prignet A (1997) Existence and uniqueness of “entropy’’ solutions of parabolic problems with \(L^1\) data. Nonlinear Anal 28(12):1943–1954
Rajagopal KR, Růžička M, Srinivasa AR (1996) On the Oberbeck–Boussinesq approximation. Math Models Methods Appl Sci 6(8):1157–1167
Walter W (1970) Differential and integral inequalities, Ergebnisse der Mathematik und ihrer Grenzgebiete, Band 55. Springer, New York (translated from the German by Lisa Rosenblatt and Lawrence Shampine)
Funding
Open access funding provided by Università degli Studi della Campania Luigi Vanvitelli within the CRUI-CARE Agreement. M. Bulíček and D. Lear acknowledge the support of the project No. 20-11027X financed by Czech Science Foundation (GAČR). D. Lear was partially supported by the RYC2021-030970-I research Grant, and the AEI Grants PID2020-114703GB-I00 and PID2022-141187NB-I00.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Ethical approval
The authors declare that they do not have competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
The research activity of A. Abbatiello is performed under the auspices of GNFM - INdAM. M. Bulíček and D. Lear acknowledge the support of the Project No. 20-11027X financed by Czech Science Foundation (GAČR). M. Bulíček is a member of the Nečas Center for Mathematical Modelling. D. Lear was partially supported by the RYC2021-030970-I research Grant, and the AEI Grants PID2020-114703GB-I00 and PID2022-141187NB-I00.
Solvability of the k-approximation—Lemma 1
Solvability of the k-approximation—Lemma 1
This appendix is devoted to the proof of Lemma 1, i.e. we assume that \(k>0\) is given and fixed. We provide the complete rigorous proof that is however very much inspired by [13], which can be used also as a tool for interested reader. We refer also to [2, 3], where the existence, the stability and the convergence to the equilibria are studied in details.
We proceed as follows. First, we introduce the two-level Galerkin approximation: one for the velocity field and the second one for the temperature. Then, we pass to the limit in the temperature equation and derive all necessary a priori estimates. Finally, we pass to the limit also in the equation for the velocity and complete the proof. Note that because of the presence of the cut-off functions, the proof is relatively standard and we use just the monotone operator theory together with the relatively classical approach for parabolic equations with \(L^1\) data.
For the sake of simplicity, we set \(d=3\) from the very beginning (since it is physically most relevant case) and to shorten the proof we consider only the case \(\alpha \equiv 0\) here. However, the general dimension d and also the case when \(\alpha >0\) is treated similarly. We also recall that in what follows the constant C denotes some universal constant depending only on the data of the problem, i.e. on \(\textbf{v}_0\), \(\theta _0\) and \(\textbf{f}\). If there is any dependence on different quantities, it will be clearly denoted.
1.1 Galerkin approximations
We start by considering an orthogonal basis \(\{\textbf{w}_i\}_{i=1}^{\infty }\) of the space \(W_{\textbf{n},{\text {div}}}^{3,2} \hookrightarrow W^{1,\infty }(\Omega )^d\) that is orthonormal in \(L_{\textbf{n},{\text {div}}}^2\), see [24, Appendix A.4] how to construct such basis. Similarly, let \(\{w_j\}_{j=1}^{\infty }\) be a basis of \(W^{1,2}(\Omega )\) which is again orthonormal in the space \(L^2(\Omega )\).
Moreover, defining the subspaces
and the associated projections
we are in position to consider the usual Galerkin approximations.
Now, we construct Galerkin approximations \(\{\textbf{v}^{n,m}, \vartheta ^{n,m}\}_{n,m=1}^{\infty }\) of the form
where \(\textbf{c}^{n,m}(t):=(c_1^{n,m}(t), \ldots , c_n^{n,m}(t))\) and \(\textbf{d}^{n,m}(t):=(d_1^{n,m}(t), \ldots , d_m^{n,m}(t))\) solve the following system of ordinary differential equations
for all \(i\in \{1,\ldots ,n\}\) and
for all \(j\in \{1,\ldots ,m\}\). Here, we have used the following abbreviations
In addition, we assume that \(\textbf{v}^{n,m}\) and \(\vartheta ^{n,m}\) satisfy the following initial conditions
where \(\textbf{v}_0^{n,m}\) is independent of m–parameter and \(\vartheta _0^{n,m}\) has the following meaning. First, we use the convention that
Then, we compute the standard regularization of an integrable function \(\eta _0\) with the kernel \(r_{1/n}\) having the support in a ball of radii 1/n. It means, we define \(\eta _0^{n}:=r_{1/n}*\eta _0\). Then, since \(\vartheta _0\) and \(\ln \vartheta _0\) are assumed to belong to \(L^1(\Omega )\), we have
Finally, we apply the projection onto the linear hull of \(\{w_j\}_{j=1}^{m}\) to get \(P^m(\vartheta _0^{n}).\) Note that as an immediate consequence of the properties of the projectors we have
In addition, since \(\log \vartheta _0\in L^{1}(\Omega )\) by hypothesis, we also have that
We focus on solvability of (A.1)–(A.2). Defining the auxiliary vector-valued functions
Then, the system (A.1)–(A.2) can be rewritten as
where \(\mathcal {F}\) is a Carathéodory function. Thus, using the classical Carathéodory theory, see [29, Chapter 1], we deduce the existence of solution to (A.1)–(A.2) at least for a short time interval. The uniform estimates derived in the next subsection enable us to extend the solution onto the whole time interval (0, T). Then, we set \(m\rightarrow \infty\) and then \(n\rightarrow \infty\). Note that some of the estimates are independent of the order of approximation and are frequently used later (after using weak lower semicontinuity of norm in a reflexive space).
1.2 Estimates independent of m
In this part, we start by assuming that \(n\in \mathbb {N}\) is arbitrary, but fixed, and we let \(m\rightarrow \infty\).
1.2.1 Estimates independent of m for the velocity
Multiplying the i-th equation in (A.1) by \(c_i^{n,m},\) then taking the sum over \(i=1,\ldots , n\) we get
Integrating the result over time interval (0, t), we obtain
where we have used the identity
which followsFootnote 4 from the fact \({\text {div}}\textbf{v}^{n,m}=0\) and integration by parts. Next, we apply (4b)\(_1\) to the second term on the left-hand side of (A.7). For the right hand side, we use the assumption that \(\textbf{f}\) is bounded and apply the \(L^2-L^2\) Hölder’s inequality and the definition of \(\mathcal {T}_k\), see (29) and the fact that \(\Vert \textbf{v}^{n,m}_0\Vert _2 \le \Vert \textbf{v}_0\Vert _2\) to obtain for all \(t\in (0,T)\)
First, the use of the Gronwall inequality directly leads to the \(L^{\infty }-L^2\) estimate for \(\textbf{v}^{n,m}\). Using this information and the Korn inequality (14) we deduce from (A.9) the following m-independent estimate
1.2.2 Estimates independent of m for the temperature
Next, multiplying the j-th equation in (A.2) by \(d_j^{n,m}\), taking the sum over \(j=1,\ldots , k\) we get
and integrating the result over time we arrive at
where we have used the abbreviation (A.3) and the fact that
which followsFootnote 5 from the fact that \({\text {div}}(\textbf{v}^{n,m})=0\) and integration by parts. Proceeding as before, we are going to study each term in (A.11). For the left hand side, we use the lower bound (3) on the second term. For the right hand side, we use the fact that \(\{\textbf{w}_i\}_{i=1}^{\infty }\in W_{\textbf{n},{\text {div}}}^{3,2}\hookrightarrow W^{1,\infty }(\Omega )^d\). Consequently, since the velocity field is for almost all time \(t\in (0,T)\) taken from the finite dimensional space \(\textbf{W}^n\), can deduce from (4b) that
where the second inequality follows from (A.10). Combining both (recall that \(\Vert \vartheta _0^{n,m}\Vert _2 \xrightarrow []{m\rightarrow \infty } \Vert \vartheta _0^{n}\Vert _2\)) we obtain
and applying the Gronwall lemma and recalling (A.10), we have
Finally, using the three-dimensional version of the interpolation inequality (15), it follows from (A.14) and from the embedding \(W^{3,2}\hookrightarrow W^{1,\infty }\) that
1.2.3 Estimates independent of m for time derivatives
In order to deduce the compactness of the velocity and the temperature, we also estimate the norms of their time derivative. More specifically, our goal in this section is to prove that
We start with the velocity field. Since \(\{\textbf{w}_i\}_{i=1}^{\infty }\) is orthonormal in \(L^2(\Omega )^d\), we have that
Therefore, multiplying the i-equation in (A.1) by \(\dot{c}_i^{n,m}(t),\) summing over \(i=1,\ldots ,n\) and integrating it over time, we obtain (after using the estimate (A.14)) that
Now, we are going to study each term separately. Using the fact that \(\{\textbf{w}_i\}_{i=1}^{\infty }\) forms a basis of \(W_{\textbf{n},{\text {div}}}^{3,2}\), the fact that \(\textbf{W}^n\) is finite dimensional, the Hölder inequality and the \(\delta\)-Young inequality and the a priori bound (A.14), we have
Proceeding similarly and using (4b)\(_2\) we get
Combining it all, taking \(\delta\) small enough and applying (A.17) and (A.14), we obtain
which gives the first part of (A.16).
Next, we focus on the estimates for \(\partial _t \vartheta ^{n,m}\). Using the orthogonality of the basis of \(W^m\) and the regularity of \(\partial _t \vartheta ^{n,m}\), we have that for all \(t\in (0,T)\)
Now, using (A.2) we get (we omit writing (t, x))
and proceeding as before, we obtain
Since, the properties of the basis \(W^m\) gives that \(\Vert P^m(\varphi )\Vert _{1,2}\le \Vert \varphi \Vert _{1,2}\), we can use (A.20) in (A.19), apply the second power and then integrate over \(t\in (0,T)\) to get
and consequently, using (A.14), we deduce that
1.3 Limit \(m\rightarrow \infty\) for fixed \(n\in \mathbb {N}\)
In this part, we let \(m\rightarrow \infty\) but keep \(n\in \mathbb {N}\) fixed. Our goal is to identify limits in the equations (A.1)–(A.2) as well as in the constitutive relations (A.3).
1.3.1 Weak and strong limits based on a priori estimates
Having a priori uniform estimates (A.14)–(A.16), we can let \(m\rightarrow \infty\) and find subsequences \(\{c^{n,m}, \vartheta ^{n,m}\}_{m=1}^{\infty }\), that we do not relabel, such that
Moreover, using the Aubin–Lions compactness lemma on the sequence \(\{\vartheta ^{n,m}\}_{m=1}^{\infty }\), i.e. using (A.22d)–(A.22f), and the compact embedding \(W^{1,2}(0,T)\hookrightarrow \hookrightarrow \mathcal {C}(0,T)\) on the sequence \(\{\textbf{c}^{n,m}\}_{m=1}^{\infty }\), i.e. using (A.22b), we have for a subsequence that we do not relabel
Moreover, it is a simple consequence of our choice of basis and (A.23b) that
and consequently
In addition, using (4b), (A.23a) and (A.25) we can use the Lebesgue dominated convergence theorem to get after denoting \(\textbf{S} ^{n,m}:=\textbf{S} ^{*}(\vartheta ^{n,m},\textbf{D}(\textbf{v}^{n,m}))\)
Similarly, recalling that \(\textbf{q}(f):=-\kappa (f)\nabla f\), combining (A.22d) and (A.23a) we get for \(\textbf{q}^{n,m}:=\textbf{q}(\vartheta ^{n,m})\) that
1.3.2 Limit in the equations for \(\textbf{v}^{n,m}\) and \(\vartheta ^{n,m}\)
The convergence results established in (A.22)–(A.27) are sufficient to let \(m\rightarrow \infty\) in (A.1) and (A.2). Indeed, we take an arbitrary \(\varphi \in \mathcal {C}^{\infty }_0(0,T)\) and multiply the i-th equation in (A.1) and the j-th equation in (A.2) by \(\varphi\) and then integrate over time \(t \in (0, T)\). Then, using the convergence results (A.22)–(A.27) it is easy to pass to the limit in all terms to get the following systems
for all \(i=1,\ldots ,n\) and
for all \(j=1,\ldots ,\infty .\) Because \(\varphi \in \mathcal {C}^{\infty }_0(0, T)\) can be chosen arbitrarily we can conclude that
for all \(i=1,\ldots ,n\) and all times \(t\in (0,T).\) From the same reason and from the fact that \(\{w_j\}_{j=1}^{\infty }\) forms a basis of \(W^{1,2}(\Omega )\) we conclude that
is valid for all \(\psi \in W^{1,2}(\Omega )\) and for all \(t\in (0,T).\)
1.3.3 Attainment of the initial condition for \((\textbf{v}^n,\vartheta ^n)\)
We start with the initial condition for the velocity. Since \(\textbf{v}_0^{n,m}:=\textbf{P}^{n}(\textbf{v}_0)\) we have that \(\textbf{v}_0^{n,m}\equiv \textbf{v}_0^n\) is independent of m–parameter. Equivalently, we have \(\textbf{c}^{n,m}_0=\textbf{c}^n_0,\) for all \(m\in {\mathbb {N}}\). Due to (A.23b) we have \(\textbf{c}^{n,m}(t) \rightarrow \textbf{c}^n(t)\) strongly in \(\mathcal {C}([0,T])\) and consequently we get \(\textbf{c}^n(0)=\textbf{c}^n_0.\) Now, from the definition of \(\textbf{v}^n(t,x)\) and \(\textbf{v}^n_0(x)\) it is clear that
for all \(x\in \Omega\). It remains to show that \(\vartheta ^n(0,x)=\vartheta ^n_0(x).\) First, note that a priori estimates (following from (A.22d) and (A.22e)) together with the standard parabolic embedding imply that
Thus, it makes a good sense to define an initial condition. To prove our goal we integrate the equation (A.2) over time (0, t) to get
Here, we have used the fact that \(\vartheta ^{n,m}_0=\vartheta ^{n,m}(0)\). Now, using the previous convergence results (A.22)–(A.27) and the convergence of the initial condition (A.5b), we can let \(m\rightarrow \infty\) and to obtain for almost all \(t\in (0,T)\) that
Since \(\vartheta ^n\in \mathcal {C}(0,T;L^2(\Omega ))\), the above identity can be extended for all \(t\in (0,T)\). Consequently, letting \(t\rightarrow 0_+\) in the above identity, we observe that
But as \(\vartheta ^n\in \mathcal {C}(0,T;L^2(\Omega ))\) and weak limit as time tends to zero is \(\vartheta ^n_0\), we have that
1.4 Estimates independent of n
In this part, we derive estimates that are n-independent and that help us to pass to the limit \(n\rightarrow \infty\). Some of the estimates are also independent of k-approximation but if there is dependence on k, it will be clearly denoted.
1.4.1 Nonnegativity for \(\vartheta ^n\)
First, we show that the temperature \(\vartheta ^n\) is nonnegative, i.e.
We consider \(\psi (t,x):=\chi _{[0,s]}(t)\min \{0,\vartheta ^n(t,x)\}\le 0\) as a test function in (A.29). Integrating it over time \(t\in (0,T)\) we obtain
Next, we show that the right hand side is non-positive. Indeed, using the definition of \(\textbf{q}^n\), see (A.27), and of \(\psi\) we have that
almost everywhere in \((0,T)\times \Omega\). Similarly, using (4) we can compute
and consequently, since \(\psi\) is non-positive, we have that
almost everywhere in \((0,T)\times \Omega\). Further, since \(\vartheta ^n_*=\max \{0,\vartheta ^n\}\), it directly follows from the definition of \(\psi\) that
Finally, to estimate the convective term, we introduce the primitive function
and we get after integration by parts (compare with (A.12)), that
where we used the fact that \({\text {div}}\textbf{v}^n=0\) in \(\Omega\) and \(\textbf{v}^{n} \cdot \textbf{n}=0\) on \(\partial \Omega\). Hence, we arrive at
Finally, the fact that \(\psi (x,0)= 0\) a.a. \(x\in \Omega\), implies that \(\Vert \psi (s)\Vert _2=0\) for all \(s\in (0,T).\) Then one can easily obtain the desired conclusion (A.30). Consequently, we can now replace \(\vartheta _{*}^{n}\) by \(\vartheta ^n\) everywhere.
1.4.2 Renormalization of the temperature equation
For our result, the very special choice of the test functions in the temperature equation plays an essential role. Therefore, we state already at this point the renormalized version of the temperature equation, which then helps us to get the a priori estimates and even more, will be important for proving the entropy (in)equality. For future use, it is convenient to record a renormalized version of the approximate equation (A.29). Using (A.22d) we have that \(\vartheta ^n \in L^2(0,T;W^{1,2}(\Omega ))\) and moreover we know \(\theta ^n\ge 0\). Therefore, if we consider \(f\in \mathcal {C}^2([0,\infty ))\) fulfilling \(\Vert f'\Vert _{W^{1,\infty }(0,\infty )}<\infty\), we deduce that for arbitrary \(\phi \in W^{1,2}(\Omega )\cap L^{\infty }(\Omega )\) the function \(\psi := f'(\vartheta ^n) \phi \in W^{1,2}(\Omega )\) for almost all \(t\in (0,T)\). Consequently, such \(\varphi\) can be used in (A.29) and we deduce
which is valid for all \(\phi \in W^{1,2}(\Omega )\cap L^{\infty }(\Omega )\), any \(f\in \mathcal {C}(0,\infty )\) fulfilling \(f'\in W^{1,\infty }(0,\infty )\) and for almost all \(t\in (0,T).\) Please notice here that for the first term, we used the regularization approach as follows (the computation is done for almost all t and \(\vartheta ^n_{\varepsilon }\) denotes the classical regularization with respect to the t-variable)
1.4.3 Energy and entropy estimates independent of n and k
First, we derive the estimates based on the kinetic and internal energy. It is noticeable that they are independent of n and also of k. We set \(\psi \equiv 1\in W^{1,2}(\Omega )\) in (A.29) and deduce the identity
which is valid for almost all \(t\in (0,T)\). Then, we multiply the i-the equation in (A.28) by \(c^n_i\), sum over \(i=1,\ldots ,n\) and we obtain
for a.a. \(t\in (0,T).\) Here, we have used the symmetry of \(\textbf{S} ^n\), the fact that \({\text {div}}\textbf{v}^n=0\) and the integration by parts (see the similar computation for the convective term in (A.8)). Summing the above identities and using the nonnegativity of the temperature \(\vartheta ^n\), we get after integration with respect to time the energy equality
where the second inequality follows from the assumptions on \(\vartheta _0\) and \(\textbf{v}_0\) and from the properties of the projection \(\textbf{P}^{n}\). Therefore, we have
and
Next, we show the uniform bound on the entropy. We fix \(0<\epsilon \ll 1\), and consider \(f(s)=\ln (\epsilon +s)\) and \(\phi =1\) in (A.31). Note that due to the nonnegativity of the temperature \(\vartheta ^n\), such choice of f is admissible. Using the following inequalities
and moving all terms with positive sign to one side, we observe that for almost all \(t\in (0,T)\) we have
Thus, using the fact that \(\textbf{f}\) is bounded, see (8a), we deduce after using the Hölder inequality and after integration over time that
Consequently, using the assumptions on \(\vartheta _0^n\), letting \(\epsilon \rightarrow 0_+\), using the Fatou lemma, the simple algebraic inequality \(\max \{0,\ln (\epsilon +\vartheta ^n(t))\}\le C(1+\vartheta ^n(t))\) and the uniform bounds (A.34) and (A.35), we deduce
where C is a uniform constant depending only on data.
1.4.4 Gradient estimates independent of n possibly depending on k
Here, we derive the estimates for the velocity and the temperature gradients that are independent of n but may depend on the truncation k. We start with the velocity filed. Integrating in time the equation (A.33) and using the estimate (A.35), we obtain
Thus, the assumption (4b) and the Korn inequality (14) and a priori estimate (A.35) imply that
Next, we focus on the estimate for the temperature. The starting point is to show that for all \(\sigma \in (0,1)\) there holds
To show (A.38), we define the auxiliary function
where \(0<\sigma <1\). Then we use (A.31) with this f and also set \(\phi :=1\). Doing so, we deduce
Here, we used the definition of \(\textbf{q}^n\) in (A.27). Noticing again that
almost everywhere in \((0,T)\times \Omega\), we can integrate the above identity over time, use the Hölder inequality and the already obtained bounds (A.34) and (A.35) to deduce (A.38).
Next, we focus on estimates for \(\vartheta ^n\) following from (A.38). First, due to nonnegativity of \(\vartheta ^n\), it follows form (A.38) that for all \(\sigma \in (0,1)\)
Moreover, using also the uniform bound (A.34), we obtain (recall that \(\sigma \in (0,1)\))
Consequently, combining the above inequalities and also (A.34), we have that
Thus, using the interpolation inequality (15) (we use its three dimensional variant only here), we deduce the classical estimate for the temperature of the form
which is valid for all \(\sigma \in (0,1)\). Therefore, it directly follows from the above inequality that
To derive an estimate on \(\nabla \vartheta ^n\), we consider \(1\le r < 5/4.\) Combining the Hölder inequality and the previous estimate (A.39), we obtain
provided that we can choose \(\sigma \in (0,1)\) such that
Since \(r\in [1,\frac{5}{4})\), we can always find \(\sigma \in (0,1)\) so that the above inequality holds and therefore
1.4.5 Estimates for time derivatives independent of n
In order to deduce the compactness of the velocity and the temperature, we also need to get a bound on the norms of their time derivatives. More specifically, our goal in this section will be to prove the uniform bounds in the following spaces
We start with the velocity field. Assume that \(\varvec{\varphi }\in W_{\textbf{n},{\text {div}}}^{3,2}\) is arbitrary and fulfills \(\Vert \varvec{\varphi }\Vert _{W_{\textbf{n},{\text {div}}}^{3,2}}\le 1\). We also recall the orthogonality of the basis \(\{\textbf{w}_i\}_{i=1}^{\infty }\) as well as the continuity of the projection \(\textbf{P}^n\) in the space \(W_{\textbf{n},{\text {div}}}^{3,2}\). Then we have due to the regularity of \(\partial _t \textbf{v}^n\) that
Now, using (A.28) we get
and proceeding as before and using the growth assumption (4b), we obtain that for all \(t\in (0,T)\)
Consequently, we have
and raising this inequality to the power \(p'\) and integrating the result over (0, T), using also the already obtained bound (A.37), we deduce
Finally, we focus on the estimate for time derivative of \(\vartheta ^n\). Recall that
is valid for all \(\psi \in W^{1,2}(\Omega )\) and for a.a. \(t\in (0,T).\) Let us consider \(\psi \in W^{1,z}(\Omega )\) with \(z>5\) fulfilling \(\Vert \psi \Vert _{1,z} \le 1\). Then using the fact that \(W^{1,z}(\Omega )\hookrightarrow L^{\infty }(\Omega )\), we get by using the Hölder inequality that for almost all \(t\in (0,T)\) the following inequality holds
Hence, using (4b), and the fact that \(z>2\), we deduce that for almost all \(t\in (0,T)\) there holds
Since \(z>5\) we have that \(z'<\frac{5}{4}\) and therefore integration of the above inequality over (0, T) and applying the uniform bounds (A.35), (A.37) and (A.40) leads to
1.5 Identification of limits as \(n\rightarrow \infty\)
In this final part we let \(n\rightarrow \infty\) and complete the proof of Lemma 1.
1.5.1 Convergence results based on a priori estimates
Having a priori uniform estimates (A.35), (A.37), (A.40), (A.39) and (A.42), using the definition of \(\textbf{q}^{n}\) and the assumption (3), and using the reflexivity of underlying spaces (or separability of their pre-dual spaces), we can let \(n\rightarrow \infty\) and find subsequences \(\{\textbf{v}^{n}, \vartheta ^{n}\}_{n=1}^{\infty }\) (that are not relabeled) such that
Moreover, using the generalized version of the Aubin–Lions compactness lemma, the estimate (A.44) and the convergence results (A.45b) and (A.45e), we obtain
Then, going back to the uniform bounds (A.36) and (A.38), we also have for a proper subsequence
In addition, using the Fatou lemma, the convergence result (A.46a) and the estimate (A.36), we have
1.5.2 Limit for the velocity equation
Having the convergence results (A.45) and (A.46), we can easily pass to the limit \(n\rightarrow \infty\) in (A.28) and to conclude for all \(\textbf{w}\in W_{\textbf{n},{\text {div}}}^{3,2}\) and a.a. \(t\in (0,T)\) that
In addition, as the space \(W_{\textbf{n},{\text {div}}}^{3,2}(\Omega )\) is dense in \(W_{\textbf{n},{\text {div}}}^{1,p}\) for all \(p\ge 1\) we have that (A.49) holds true for all \(\textbf{w}\in W_{\textbf{n},{\text {div}}}^{1,p}\) and almost all \(t\in (0,T)\) and also that
Consequently, the standard parabolic interpolation gives that
To complete the part of the proof of Lemma 1, which is related to the velocity field, we need to identify \(\mathbf{S}\) and also the initial condition for \(\textbf{v}(0)\).
1.5.3 Attainment of the initial condition for \(\textbf{v}\)
The arguments concerning the attainment of the initial conditions \(\textbf{v}_0\) are standard. For sake of completeness, we included all details below. Since \(\textbf{v}\in \mathcal {C}(0,T;L^2(\Omega )^3),\) we get that
In what follows, we show that
and these convergence results together (due to the uniqueness of weak convergence) identify the limit (44)\(_1\), that we want to prove.
Let \(0<\varepsilon \ll 1\) and \(t\in (0,T-\varepsilon )\) be arbitrary. We consider the auxiliary function \(\eta\) defined in (31), multiply (A.28) by this \(\eta\), and integrate the result over (0, T) to obtain for every \(i=1,\ldots ,n\)
Next, we integrate by parts in the first term, use that \(\eta (T) = 0\), and the equality \(\textbf{v}^n(0)=\textbf{P}^{n}(\textbf{v}_0)\) to get
This identity is ready for the use of the convergence results (A.45)–(A.46) as well as the convergence of the projection \(\textbf{P}^{n}(\textbf{v}_0)\) to obtain for any \(i\in \mathbb {N}\) that
Using the definition of \(\eta\), i.e. using that \(\eta (\tau ) = 1\) for \(\tau \in [0, t)\) and \(\eta (\tau ) = 0\) for \(\tau \in (t + \varepsilon , T]\), and \(\eta '(\tau )=-\frac{1}{\varepsilon }\) for \(\tau \in (t, t + \varepsilon )\), we obtain
Next, since \(\textbf{v}\in \mathcal {C}(0,T;L_{\textbf{n},{\text {div}}}^2)\), we can easily let \(\varepsilon \rightarrow 0_+\) to get
for all \(t\in (0,T)\). Therefore, we see that
This holds for every \(i \in \mathbb {N}\), and since \(\{\textbf{w}_i\}_{i=1}^{\infty }\) is a basis of the space \(W_{\textbf{n},{\text {div}}}^{3,2}\), this is nothing more than
Finally, since \(W_{\textbf{n},{\text {div}}}^{3,2}\) is dense in \(L_{\textbf{n},{\text {div}}}^2\), the weak convergence result that we expected is also true, and it identifies the strong limit of the initial condition in \(L_{\textbf{n},{\text {div}}}^2\) as required in (44)\(_1\).
1.5.4 Identification of \(\mathbf{S}\)
Next, we show that
First, we notice that we can extend the solution to the interval \((0,T+1)\), e.g., by extending \(\textbf{f}\) by zero outside of (0, T). Multiplying the i-th equation in (A.28) by \(c_i^n\), summing the result over \(i=1,\ldots , n\) and integrate over \((0,T+s)\), where \(s\in (0,1)\), we deduce the following identity (the convective term vanishes)
Using the fact that \(\textbf{S} ^n:\textbf{D}(\textbf{v}^n)\ge 0\), we can integrate the above identity over \(s\in (0,\varepsilon )\) to get
Then, we may directly use the convergence results (A.45) and (A.46), together with the Fatou lemma to get
Finally, letting \(\varepsilon \rightarrow 0_+\), we get (using that \(\textbf{v}\in \mathcal {C}([0,T]; L^2(\Omega ))\))
Then, setting \(\textbf{w}:=\textbf{v}\) in (A.49), integrating over (0, T) and using the fact that \(\textbf{v}(0)=\textbf{v}_0\), we get the identity
Comparing (A.51) and (A.52), we directly derive
Consequently, using this inequality, the monotonicity and the growth assumption (4a)–(4b), the strong convergence result (A.46a), the weak convergence results (A.45b), (A.45f) and the Lebesgue dominated convergence theorem, we deduce that for all \(\overline{\textbf{D}}\in L^p(Q,\mathbb {R}^{d\times d})\) there holds
The choice \(\overline{\textbf{D}}:=\textbf{D}(\textbf{v})\pm \lambda \widetilde{\textbf{D}}\) with \(\lambda >0\) then leads (after division by \(\lambda\) and letting \(\lambda \rightarrow 0_+\)) to
for all \(\overline{\textbf{D}}\in L^p(Q,\mathbb {R}^{d\times d})\). This directly implies (A.50).
In addition, we show that
Indeed, we set \(\overline{\textbf{D}}\equiv \textbf{D}(\textbf{v})\) in (A.54) and we get
However, thanks to (4a), this implies
Since
which follows from (A.45b), (A.46a) and (4b), we see that (A.56) directly implies (A.55).
1.5.5 Limit in the temperature gradient and the renormalized temperature equation
First, we let \(n\rightarrow \infty\) in (A.29). Using the weak convergence results (A.45b)–(A.45e), the strong convergence results (A.46), the essential convergence stated in (A.55), the convergence of the initial condition (A.4a) and using also the integration by parts with respect to time variable, we deduce that for all \(\psi \in \mathcal {C}_0^1(-\infty , T; \mathcal {C}^1(\overline{\Omega }))\) there holds
In addition, having the weak compactness stated in (A.56), we can follow step by step the procedure developed in [27], see also [1], to pass to the limit in (A.31) and to obtain (46). Finally, to obtain (43), we set \(f'(s):=(s-m)_+/s\) and \(\phi \equiv 1\) in (A.31) to get after integration over (0, T) that
Using the weak lower semicontinuity and also all the above established convergence results, we may let \(n\rightarrow \infty\) to get
Thus, we see that taking \(\limsup\) as \(m\rightarrow \infty\), one conclude (43).
1.5.6 Attainment of the initial condition for \(\vartheta\)
This is a very classical part and we refer the reader e.g. to [13] or [7, 8].
1.5.7 On the pressure
In this final subsection, we sketch the proof the existence of the pressure \(\pi \in L^{p'}(Q)\). We refer for details to [12] or [13]. By the Helmholtz decomposition we observe that
Having \((\textbf{v},\vartheta )\) we introduce \(\pi\) as the solution of the following problem
where \((-\Delta _N)\) denotes the Laplace operator together with homogeneous Neumann boundary conditions. We consider
The \(L^p\)-regularity theory for the Neumann problem (A.59) implies, see [19, Proposition 2.5.2.3], that
Note that the weak formulation of (A.59) is the following identity
for all \(\varphi \in W^{2,p}(\Omega )\) with \(\nabla \varphi \cdot \textbf{n}=0\) on \(\partial \Omega\) and a.a. \(t\in (0,T).\) Having such pressure in hands, it is then easy to show that (A.49) holds as
for all \(\textbf{w}\in W_{\textbf{n}}^{1,p}(\Omega )\) and a.a. \(t\in (0,T).\)
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Abbatiello, A., Bulíček, M. & Lear, D. On the existence of solutions to generalized Navier–Stokes–Fourier system with dissipative heating. Meccanica (2024). https://doi.org/10.1007/s11012-024-01791-5
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s11012-024-01791-5
Keywords
- Navier–Stokes–Fourier equations
- Dissipative heating
- Oberbeck–Boussinesq system
- Non-Newtonian fluids
- Large-data existence