Dedicated with appreciation to Tai-Ping Liu on his 60 th birthday

Size: px
Start display at page:

Download "Dedicated with appreciation to Tai-Ping Liu on his 60 th birthday"

Transcription

1 ENTROPY STABLE APPROXIMATIONS OF NAVIER-STOKES EQUATIONS WITH NO ARTIFICIAL NUMERICAL VISCOSITY EITAN TADMOR AND WEIGANG ZHONG Abstract. We construct a new family of entropy stable difference schemes which retain the precise entropy decay of the Navier-Stokes equations, d/dt ρsd = λ +µq /θ + κθ /θ d. To this end we employ the entropy conservative differences of [Tadmor4] to discretize Euler convective flues, and centered differences to discretize the dissipative flues of viscosity and heat conduction. The resulting difference schemes contain no artificial numerical viscosity in the sense that their entropy dissipation is dictated solely by viscous and heat flues. Numerical eperiments provide a remarkable evidence for the different roles of viscosity and heat conduction in forming sharp monotone profiles in the immediate neighborhoods of shocks and contacts. Dedicated with appreciation to Tai-Ping Liu on his 6 th birthday Contents. Introduction. Entropy dissipation 4.. The entropy variables 4.. Eamples of entropy pairs for Navier-Stokes equations 6 3. Entropy stable schemes for Navier-Stokes equations Entropy conservative schemes Entropy stable semi-discrete schemes for Navier-Stokes equations 3.3. Time discretization 3 4. Numerical eperiments 4 References 5. Introduction We consider the full Navier-Stokes NS equations for compressible viscous flows in one-space dimension, ρ m + m qm + p =λ +µ t E qe + p q + κ q /.. θ Here, ρ = ρ, t is the density of the flow, m = m, t is the momentum and E = E, t stands for the total energy per unit volume. The three equations epress, respectively, conservation laws of mass, momentum and total energy for the flow, driven by convective flues on the left together with viscous Date: December, Mathematics Subject Classification. 76N,65M. Key words and phrases. Navier-Stokes equations, viscosity, heat conduction, entropy decay, entropy variables, difference schemes, entropy conservative schemes, entropy stability.

2 Eitan Tadmor and Weigang Zhong and heat flues on the right. These flues involve the velocity q := m/ρ, the pressure p = p, t which is determined by an ideal polytropic equation of state, p =γ e, e := E m ρ,. and the absolute temperature, θ = θ, t >, such that C v ρθ = e. The constant γ > is the specific heat ratio and e = e, t is the internal energy. On the RHS of. we have the viscous and heat flues, depending on the constant Lamé coefficients of the viscosity λ, µ > and the constant conductivity κ>. Finally, C v > is the specific heat at constant volume; for simplicity, we set C v = while rescaling κ κ/c v. If the heat flu is ecluded from the full NS equations, i.e. κ =, we obtain the viscous NS equations ρ m + m qm + p =λ +µ t E qe + p q..3 q / On the other hand, if we turn off the viscosity, i.e. λ = µ =, then the NS equations amount to t ρ m E + m qm + p qe + p = κ θ..4 If the heat flu and viscosity are both taken away, the system. is reduced to the compressible Euler equations, ρ m + m qm + p =..5 t E qe + p The additional viscous and heat flu terms on the RHS of the various NS equations.,.3 or.4, are dissipative terms in the sense that they are responsible for the dissipation of the total entropy. To this end, we now discuss the entropy balance associated with the above equations. We begin with the specific entropy S := ln pρ γ. A straightforward manipulation on.,. yields the transport equation, S t + qs = κ ρ ln θ +λ +µ q e + κ ρ θ..6 θ Multiplied by ρ,.6 becomes ρs t + ms = κln θ +λ +µ q θ + κ θ..7 θ On the other hand, pre-multiplying the continuity equation, ρ t + m =,bys and adding it to.7, ρs + ms + κln θ = λ +µ q t θ κ θ..8 θ Spatial integration of.8 then yields d ρs d = λ +µ dt q θ d κ θ d..9 θ Since the epression on the right is negative, we conclude that the total entropy, ρs d, is decreasing in time, thus recovering the second law of thermodynamics, e.g., [GrootMazur984]. In fact, equation.9 specifies the precise entropy decay rate, which is dictated by the viscous and heat flues through their dependence on the nonnegative κ, λ, and µ.

3 Entropy stable approimations for the Navier-Stokes equations 3 The purpose of this paper is to present a systematic study of difference schemes which respect the above entropy dissipation statements. The typical approach by practitioners in the field of Computational Fluid Dynamics is to address the general issue of entropy stability by adding enough artificial numerical viscosity often an ecessive amount of it, in order to mask various discretizations errors and enforce the decay of the total entropy ρs d. Our aim here is to construct more faithful approimations of the NS equations, with a discrete analogue for the precise entropy decay statement in.9. Our prototype result is the following. Theorem.. Consider the NS equations., t u + fu =ɛ du. Here, u = [ ρ, m, E ] is the vector of conservative variables, fu is the corresponding 3-vector of flues fu = [ m, qm + p, qe + p ], and ɛdu stands for the combined viscous and heat flues, ɛdu =λ +µ [,q,q / ] + κ [,,θ ] where ɛ signals the vanishing amplitudes of viscosity and heat conductivity. We approimate these NS equations by a semi-discrete scheme of the form d dt u νt+ f f = ɛ du ν+ du ν du ν du ν..a ν+ ν ν ν ν+ ν Let f = f u ν+ ν, u ν+ is the numerical flu given the by the eplicit formula, f ν+ =γ 3 j= m j+ m j [ l j, v ν+ v ν lj, v = vu := E e S + γ +, q ]. θ,.b θ Here, {l j = l j } 3 ν+ j= are three linearly independent directions in v-space at our disposal consult eamples 3., 3. and 3.3 below; {r j = r j } 3 ν+ j= is the corresponding orthogonal system and {mj = m j } 4 ν+ j= are the intermediate values of the momentum specified along the corresponding path, vj+ = v j + l j, v ν+ v ν r j, starting with v = v ν and ending with v 4 = v ν+. Then, the resulting scheme.a,.b is entropy stable and the following discrete entropy balance holds d dt ρ ν S ν ν = ν λ +µ qν+ /θ ν ν+ ν+ ν+ κ θν+ /θ ν ν+ ν+ ν+.. We briefly describe the content of the paper, leading to the above result. The entropy balance. is a precise discrete analogue of.9. The scheme.a,.b contains no artificial numerical viscosity in the sense that entropy dissipation is driven solely by the viscous and heat flues. In the particular case that viscosity and the heat conduction are absent, κ = λ = µ =, then the entropy balance.8 is reduced to the formal entropy equality of the Euler equations.5, ρs + t ms =,. We let ẑ ν+ and z ν+ denote the arithmetic and geometric means, ẑ ν+ = z ν + z ν+ / and zν+ = z νz ν+.

4 4 Eitan Tadmor and Weigang Zhong which in turn, implies the entropy conservation ρs,t d = ρs, d. Similarly, setting ɛd = we omit the dissipative terms in NS equations, and the difference scheme.a becomes entropy conservative, ρν S ν t ν = ρ ν S ν ν. Entropy conservative schemes are studied in section 3, following [Tadmor4]. The key ingredient here is the construction of their entropy conservative flues, such as f in.b. These flues employ ν+ the so called entropy variables, v = vu, which are discussed in section. The main results are then summarized in theorems 3. and 3.. Finally, in section 4 we present a series of numerical simulations with the new schemes. The entropy conservative approimations of Euler equations are purely dispersive and as such, their solutions eperience dispersive oscillations, interesting for their own sake, consult [La986, HFMM986, Tadmor986, LLV993, LevermoreLiu996, LeFlochRhode] and the references therein. Turning to the NS equations, our simulations provide a remarkable evidence for the different roles that viscosity and heat conduction have in removing the dispersive oscillations, to yield sharp monotone profiles of well-resolved shock and contact layers. No limiters were added, but instead, the viscous and heat conduction terms in NS equations are found to serve as accurate edge detectors Remark.. The viscous NS equations dissipate a general family of entropies, ρhs, where hs is an arbitrary increasing function. Indeed, arguing along the above lines we multiply the continuity equation by hs and adding it to.7 h S to find ρhs + t mhs+κln θ h S = θ..3 θ = κh θ SS θ h S λ +µ q θ + κ In the case that the heat conduction is absent, the first term on the RHS of.3 vanishes, and we are left with d q ρhs d = λ +µ dt θ h Sd..4 Thus, the viscous NS equations imply the dissipation of a family of entropies, ρhs d for all h S > ; consult [Harten983]. Each one of these entropies carries its own entropy conservative flu f. The eplicit construction of such flues is outlined in theorem 3. below. Combining ν+ these entropy conservative flues together with centered differencing of the additional viscous terms, λ +µ[,q,q /], yield a generalization of theorem. which recovers the precise entropy balance.4. We note in passing that when heat conduction is present, however, the negativity of the first term on the right of.3 requires h S =, so that we are left with one canonical entropy, hs S discussed in theorem.; consult [HFMM986]. Remark.. It is straightforward to generalize the recipe for faithful entropy stable approimations of multidimensional NS equations. The etension is carried out dimension by dimension and as indicated in the one-dimensional setup of theorem., one has the freedom of choosing different paths in phase space.. Entropy dissipation.. The entropy variables. We turn our attention to general systems of hyperbolic conservation laws t u + fu =,, t R [,,.

5 Entropy stable approimations for the Navier-Stokes equations 5 governing the N-vector of conserved variables, u =[u,...,u N ]. The entropy equality t Uu+ F u =.. is an additional conservation law for an entropy function, Uu, which is balanced by an entropy flu F u. The Euler equations.5 are viewed as a prototype eample for such systems, with the three conservative variables u = [ ρ, m, E ] [ ] balanced by the flu f = m, qm + p, qe + p and endowed with entropy pairs U, F = ρhs, mhs. We now briefly recall the circle of ideas linking the dissipation of the total entropy, Uu,td, and the realization of u as a vanishing viscosity limit, in analogy to the vanishing NS limits and their relation to entropic solutions of the Euler equations. We refer to e.g., [Liu99] and [Dafermos], for a more comprehensive discussion. Let Uu,Fu be a given entropy pair associated with.. Note that Uu satisfies the entropy equality. if and only if it is linked to an entropy flu function F u through the compatibility relation U u f u = F u..3 Indeed, multiplying. by Uu on the left, one recovers the equivalence between. and.3 for all u s solving.. Of course, these formal manipulations are valid only under the smooth regime. To justify these steps in the presence of shock discontinuities, the conservation law. is realized by appropriate vanishing viscosity limits. To this end, we define the entropy variables vu :=U u u. We make the additional assumption that the entropy Uu isconve, so that the mapping u v is a one-to-one. Following, [Godunov96, Mock98], we claim that the change of variables, u = uv, puts the system. into the equivalent symmetric form, t uv+ fuv =..4 The system.4 is symmetric in the sense that the Jacobian matrices of flues are, u v v = u v v and fv v = f v v..5 Indeed, a straightforward computation utilizing the compatibility relation.3, shows that uv and fv are, respectively, the gradients of the corresponding potential functions, φ and ψ, uv =φ v v, φv := v, uv Uuv,.6 fv =ψ v v, ψv := v, fv F uv..7 Hence the Jacobian matrices Hv :=u v v and Bv :=f v v in.5 are symmetric, being Hessians of the potentials φv and ψv. Moreover, the conveity of U implies that H is positive definite, H =U uu >. Physically relevant solutions of. are postulated as limits of the vanishing viscosity solutions, t uɛ + fuɛ =ɛ duɛ,.8 where du is any admissible dissipative flu, and ɛ stands for vanishing amplitudes such as the viscosity coefficients λ, µ, the heat conductivity κ, etc. Here, the admissibility of the dissipative flu requires the Jacobian d u to be H-symmetric positive-definite, that is, d H =d H, d H := d u H..9 For brevity of notations we often write fv for fuv whenever the different dependence of fu and fv is made clear by the distinction between the conservative variables u and entropy variables v.

6 6 Eitan Tadmor and Weigang Zhong If we epress the dissipation flu in terms of the entropy variables, dv = duv, then admissibility requires that the v-jacobian of this flu will be positive symmetric, d H = d v v =d v v. Thus, in this case.8 reads t uvɛ + fvɛ =ɛ dvɛ, and all v-dependent flues the temporal, spatial and the dissipation flu have symmetric Jacobains. We now integrate.8 against v = Uu, employ the compatibility relation U u f = F and use differentiation by parts on the admissible dissipation on the right to find t Uuɛ + F ɛ u ɛ ɛ v ɛ, du ɛ = ɛ v ɛ, d v v ɛ.. Letting ɛ, we obtain the entropy inequality, [La973] t Uu+ F u,. which is the generalization of the entropy decay statements for NS equations.8 and.3... Eamples of entropy pairs for Navier-Stokes equations. The NS equations admit the family of conve entropy pairs Uu = ρhs, Fu = mhs, h.. Here S =lnpρ γ is the specific entropy and the conveity of the corresponding Uu s as functions of u =ρ, m, E holds iff h S γh S >, [Harten983]. We consider two prototype eamples. Eample.. The simplest choice of hs is the specific entropy S itself, hs =S =ln pρ γ..3 Straightforward computation gives us the following entropy pair, entropy variables, and potentials. Entropy pair Uu = ρs and F u = ms; Entropy variable consult [Harten983] with the inverse mapping uv = p v 3 v = w γ v v 3 vu = v 3 v v v 3 E/e S + γ + q/θ /θ, where w =.4 γ γ S e v 3 γ γ, S = γ v + v ; v 3 Potential pair φ =γ ρ and ψ =γ m. In this case, the general statement of entropy balance in. with the entropy pair U, F = ρs, ms amounts to the one we have in.9, d dt ρs d = ɛ v ɛ, d vv ɛ d = λ +µ q θ d κ θ d..5 θ Eample.. A particularly convenient form of entropy variables is associated the entropy function consult [Harten983, Tadmor4], Uu = ρhs with hs = γ + γ e S γ+,.6 where we have the following.

7 Entropy stable approimations for the Navier-Stokes equations 7 Entropy pair Uu = +γ γ ρp +γ and F u = +γ γ qρp Entropy variable vu = u Uu = ρp γ +γ with the inverse mapping uv = ρp γ γ+ v 3 v v = E m ρ +γ ; [ ] γ γ v v 3 v γ v 3 v v Potential pair φ, ψ = ρp γ+,mpρ γ γ+ In case that the heat conduction is absent κ =, we apply the general statement of entropy balance. with the entropy pair, U, F = +γ γ ρp +γ,qρp +γ, obtaining ρp +γ d = ɛ v, ɛ d v v ɛ d d dt λ +µ γ h Sq d = λ +µ + γθ +γ ; e S/+γ q d..7 θ Remark.. As noted in [Harten983], the flu fv is a homogeneous function of degree η =: + γ/ γ, fαv =α η fv, α R. Homogeneity implies that f v vv = ηfv which in turn, enables us to rewrite the spatial flu in. in a skew-adjoint form, fu = f v v +f v v /η + ; consult [Tadmor984a]. 3. Entropy stable schemes for Navier-Stokes equations 3.. Entropy conservative schemes. We turn our attention to consistent approimations of., based on semi-discrete conservative schemes of the form d dt u νt = f ν+ f ν ν. 3. Here, u ν t denotes the discrete solution along the grid line ν,t, ν := ν+ ν is the possibly variable mesh spacing and f ν+ is the Lipschitz-continuous numerical flu which occupies a stencil of p + -gridvalues, = fu ν p+,, u ν+p. f ν+ The scheme is said to be consistent with the system. if f satisfies fu, u,, u =fu, u R N. By making the changes of variables u ν = uv ν, we obtain the equivalent form of 3. d dt uv νt = f ν+ f ν ν. 3. The essential difference lies with the numerical flu, f ν+, which is now epressed in terms of the entropy variables, f ν+ = f v ν p+,, v ν+p :=f u v ν p+,, u v ν+p, consistent with the differential flu, fv, v,, v =fv fuv. 3.3 The semi-discrete schemes 3. and 3. are completely identical. It proved useful, however, to work with the entropy variables rather than the usual conservative ones, since system. is symmetrized with respect to these entropy variables. The entropy variables-based formula 3. has the advantage that it provides a natural ordering of symmetric matrices, which in turn enables us to compare the

8 8 Eitan Tadmor and Weigang Zhong numerical viscosities of different schemes, consult [Tadmor984b, Tadmor987] for details. In particular, we will be able to utilize the entropy conservative discretization of [Tadmor4] for the Euler convective part of the equations, and thus recover the precise entropy balance dictated by viscous and heat flues in NS equations. Let U, F be a given entropy pair. We proceed with the construction of an entropy conservative scheme, in the sense of satisfying a discrete entropy equality analogous to., d dt U u νt + ν F ν+ F ν =. 3.4 Here, F ν+ = F u ν p+,, u ν+p is a consistent numerical entropy flu, such that F u, u,, u = F u, u R N. The numerical flu of such entropy conservative schemes will play an essential role in the construction of entropy stable schemes, by adding a judicious amount of physical viscosity. The key step in the construction of entropy conservative schemes is the choice of an arbitrary piecewise-constant path in phase space, connecting two neighboring gridvalues v ν and v ν+ through the intermediate states {v j } N ν+ j= at the spatial cell [ ν, ν+ ]. Let {r j } N ν+ j= be an arbitrary set of N linearly independent N-vectors, and let {l j } N ν+ j= be the corresponding orthogonal set. We introduce the intermediate gridvalues {v j } N ν+ j=, which define a piecewise constant path in phase space v ν+ v j+ ν+ v N+ ν+ = v ν = v j ν+ + = v ν+ l j, v ν+ ν+ r j, j =,,,N, ν+ v ν+ := v ν+ v ν. 3.5 Theorem 3. Tadmor4, Theorem 6.. Consider the system of conservation laws.. Given the entropy pair U, F, then the conservative scheme d dt u νt = f f ν+ ν ν 3.6 with a numerical flu f ν+ N ψ v j+ ψ v j f ν+ ν+ = ν+ j= l j, v ν+ ν+ l j ν+ 3.7 is an entropy-conservative approimation, consistent with.,.. Here, v are the entropy variables, v = U u u and ψv is the entropy potential.7 ψv = v, fuv F uv. The proof is based on the requirement of entropy conservation in [Tadmor987, Theorem 5.], v ν+, f = ψ ν+ ν+, ψ ν+ := ψv ν+ ψv ν. 3.8 The numerical flu 3.7 satisfies this entropy conservation requirement, for N ψ v j+ ψ v j v ν+, f ν+ ν+ = l j, v ν+ j= l j ν+ ν+, v ν+ ν+ = N j= ψ v j+ ψ v j ν+ ν+ = ψ v N+ ψ v ν+ ν+ = ψ ν+.

9 Entropy stable approimations for the Navier-Stokes equations 9 In addition, f ν+ path, v j+ ξ :=v j + v j+ /+ξ l j, v ν+ ν+ ν+ ν+ ν+ r j ν+ we have is a consistent flu satisfying 3.3. Indeed, if we let v j+ ξ denote intermediate ν+ ψ v j+ ψ v j = ν+ ν+ ξ= d dξ ψ v j+ ξ ν+ connecting v j ν+ dξ and v j+, then by.7, ν+ = f v j+ ξ dξ, r j l j, v ξ= ν+ ν+ ν+ ν+. Inserted into 3.7, we can rewrite the entropy-conservative flu 3.7 in the equivalent form N f = f v j+ ν+ ξ dξ, r j l j, 3.9 ν+ ν+ ν+ j= ξ= and the consistency relation 3.3 now follows, f v, v = N fv, r j l j = fv. ν+ ν+ j= We emphasize that the recipe for construction entropy-conservative flues in 3.7 allows an arbitrary choice of a path in phase space. We demonstrate this recipe with three eamples. Eample 3.. Set {r j } along the standard Cartesian coordinates, r j = e ν+ j, j =,,...,N. In this case we have [ v j vν+ = ν+,..., v ν+ j, v ν j,...,, v ν N] j =, 3,...,N, and the entropy conservative flu 3.7 is given by the particularly simple eplicit formula = ψ v ν+ ψ vν ψ v j+ vν+ ν+ ψ v j ψ v,..., ν+ ν+ ψ v N v ν vν+ j ν+,..., v ν vν+ N. 3. v ν f ν+ We carried out numerical eperiments with these flues for the approimate solution of Euler and NS equations. The formulation is particularly simple though the computed intermediate values might lie outside the physical space ρ, p >. Eample 3.. A more physically relevant choice than the Cartesian path is offered by a Riemann path which consists of {u j } N ν+ j=, stationed along an approimate set of right eigenvectors, { rj }, ν+ of the Jacobian f u u ν+. Set v j = vu j, j=,,...,n, and let l j s be the orthogonal system ν+ ν+ to {v j+ v j } N j=. This will be our choice of a path for computing entropy stable approimations of NS equations in section 3. below. The resulting flu, miing conservative and entropy variables, admits the somewhat simpler form N ψ u j+ ψ u j f ν+ ν+ = l j where ψu = U ν+ j= l j ν+ u u fu F u. 3., v ν+ ν+ Eample 3.3. If all r j s are chosen to approach the same direction of v ν+, then by 3.9 the flu 3.9 collapses to the entropy-conservative flu f ν+ = ξ= f v ν+ ξ dξ, v ν+ j ξ := v ν + v ν+ +ξ v ν+. 3. N

10 Eitan Tadmor and Weigang Zhong The resulting flu 3. was introduced in [Tadmor986] and was the forerunner for the family of entropy conservative flues outlined in theorem 3.. It has the drawback, however, that its evaluation requires a nonlinear integration in phase space. Thus, with the loss of linear independence, we lose here the eplicit evaluation of the entropy conservative flu offered in 3.7 and demonstrated in the previous two eamples. 3.. Entropy stable semi-discrete schemes for Navier-Stokes equations The compressible Euler equations. Let U, F be an admissible entropy pair associated with the Euler equations.5, let v = vu denote the corresponding entropy variables outlined in eamples. and. above. To conserve the total entropy Uu,td, we appeal to the semi-discrete scheme 3.6 with the entropy-conservative numerical flu 3.7, f ν+ = 3 j= ψ v j+ ν+ ψ l j, v ν+ ν+ v j ν+ l j. ν+ To compute f, we distinguish between two cases. If v ν+ ν = v ν+, we employ the equivalent form of the numerical flu in 3.9, f ν+ = 3 j= ξ= f v j+ ξ dξ, r j l j, ν+ ν+ ν+ which implies that all the intermediate gridvalues coincide, v ν = v = v = v 3 = v 4 = v ν+ ν+ ν+ ν+ ν+ and the entropy-conservative flu amounts to f = f ν+ ν = f ν+. Otherwise, if v ν v ν+, we choose to work along the path which is dictated by an approimate Riemann solver. Specifically, we use the eigensystem of the Roe matri, [A] ν+ = Au ν, u ν+, [Roe98], γ 3 [A] ν+ := q 3 γ q ν+ ν+ γ. 3.3a γ q3 ν+ q ν+ H ν+ H ν+ γ q ν+ γ q ν+ Here q and H are the average values of the velocity q and total enthalpy H =E +p/ρ at Roe-average state, q ν+ = q ν ρν + q ν+ ρν+ ρν +, Hν+ = H ν ρν + H ν+ ρν+ ρ ν+ ρν b ρ ν+ The r j s are the right eigenvectors { r j r j } 3 ν+ j= of the Roe matri 3.3a given by omitting the subscript ν+ of all averaged variables r = q c, r = q, r 3 = q + c, 3.3c H q c q / H + q c with the corresponding left eigenvector set { l j l j ν+ } 3 j= given by + δ/4γ l = + δ/ c, l = δ /γ δ/4γ δ/ c, l 3 = δ/ c γ / c γ / c γ / c. 3.3d

11 Entropy stable approimations for the Navier-Stokes equations Here δ := γ q/ c, and c is the average sound speed given by c =γ H q. We are now able to form the intermediate path in u space as in 3.5 u j+ = u j + lj, u r j, j =,, Since the mapping between u and v is one-to-one, then these intermediate gridvalues in u space, {u j } 4 j=, correspond to intermediate gridvalues {vj } 4 j= in v space. We let {rj } 3 j= be the right vectors connecting these v-values, r j := v j+ v j, and let {l j } 3 j= be the corresponding left orthogonal set. We summarize the algorithm of computing the entropy-conservative flu f in the ν+ following. Algorithm 3.. If u ν = u ν+ then f ν+ = fv ν ; else Set u ν+ := u ν and compute recursively the intermediate states, u j+ ν+ = u j + lj ν+ ν+, u ν+ r j, j =,, ν+ Here, { l j ν+ } and { rj } are the left and right eigensystems of the Roe matri in 3.3c, 3.3d. ν+ Set r j = vu j+ vu j and compute {l j } 3 ν+ ν+ ν+ j= as the corresponding orthogonal system, l j, r k = δ ν+ ν+ jk, r j := v j+ v j 3.6 ν+ ν+ ν+ 3 ψ v j+ ψ v j Compute the entropy-conservative numerical flu, f ν+ ν+ = l j ν+ j= l j ν+, v ν+ ν+ Remark 3.. Observe that if {u j+ u j } 3 j= are linearly independent then, since u v is symmetric positive definite, the corresponding set of directions in v-phase space, {v j+ v j } 3 j=, are also linearly independent, at least when u ν+ is in a small neighborhood of u ν. It guarantees the eistence of the orthogonal set {l j } 3 j=. But what happens when ˆl j, u = for certain j s? for eample, if u ν is connected to u ν+ through a k-shock, then the Roe matri [A] ν+ retains the perfect resolution of such a shock by enforcing ˆl j, u =, j k and we can omit the contribution of these sub-paths to the conservative flu f. The general approach is to construct a precise mirror image of the Roe-path in v-phase space in terms of the right and left orthogonal systems, r j := [H] ˆr j, l j := [H] ν+ ν+ ˆlj, j =,,..., 3.7 where [H] ν+ denotes an averaged symmetrizer such that u ν+ =[H] ν+ v ν+ and there are many such averages. Then, {r j } 3 j= forms the path in v-phase space, vj+ = v j + l j, vr j, which retains the desired Roe property of perfect resolution of shocks. Indeed, if u is a k-shock with speed s then it satisfies omitting subscripts f = s u = s[h] v. But the Roe matri in 3.3a is constructed so that [Roe98], f = [A] u = [A][H] v, and we conclude v = r k. Thus, l j, v =, j k. The corresponding entropy conservative numerical flu reads f = ψ v j + ξ j r j ψ v j ν+ ν+ ν+ l j, ξ j = ν+ ξ ν+ j {j ξj } l j, v ν+ ν+

12 Eitan Tadmor and Weigang Zhong 3... The Navier-Stokes equations. We turn to the construction of entropy-stable schemes for the full NS equations.. To this end, we rewrite the equation as a system of conservation laws t u + fu =ɛ du, u = ρ m m, fu = qm + p, 3.8a E qe + p with additional diffusive terms ɛdu :=λ +µ q q / + κ θ. 3.8b For the convection part on the LHS, we use the same entropy-conservative differencing used for the Euler equations. For the dissipative terms on the RHS, we employ standard centered differences. We arrive at our main result. Theorem 3.. Let U, F be a given entropy pair of the NS equations 3.8a,3.8b, which respect the entropy inequality.. Consider the semi-discrete approimation d dt u νt+ f f = ɛ d ν+ d ν d ν d ν. 3.9a ν+ ν ν ν ν+ ν Here f is an entropy conservative numerical flu 3.7, ν+ 3 ψ v j+ ψ v j f ν+ ν+ = l j, 3.9b ν+ j= l j ν+, v ν+ ν+ which is outlined in algorithm 3. above. {i} The resulting scheme 3.9a,3.9b is entropy-dissipative in the sense that d U ν t ν = ɛ v dt ν+, d ν+ v ν ν ν+ v ν+. 3. ν+ This entropy balance is a discrete analogue of the entropy balance statements.5 and.7. {ii} In the specific case of the canonical entropy pair U, F = ρs, ms, the entropy decay 3. amounts to. d U ν t ν = dt ν qν+ /θ ν+ ν+ ν+ κ ν θν+ /θ ν+ ν+ ν+ λ +µ. 3. ν Proof. We multiply 3.9a by [U u ] ν = v ν, then sum up all spatial cells to get the balance of the total entropy, d U ν t ν + v ν, f f = ɛ v dt ν+ ν ν, d ν+ d ν d ν d ν. 3. ν ν ν ν+ ν Since we chose f as the entropy conservative flu, a straightforward manipulation on the entropy ν+ conservation requirement 3.8 yields the conservative difference, v ν, f f = F ν+ ν ν+ F ν, 3.3

13 Entropy stable approimations for the Navier-Stokes equations 3 where F ν+ = v ν + v ν+, f ν+ ψv ν +ψv ν+. On the other hand, summation by parts on the RHS of 3. yields ɛ ν v ν, d ν+ d ν ν+ d ν d ν ν nuhf = ν = ν = ν ɛ ν+ ɛ ν+ ɛ ν+ By 3.3 and 3.4b, the semi-discrete entropy balance amounts to Here d dt U ν t ν = ν ν ɛ ν+ v ν+ v ν+ v ν, d ν+ d ν v ν+ v ν+, d ν+, d ν+ v v ν+ ν+, d ν+ v v ν+ ν+ 3.4a. 3.4b. 3.5 d ν+ v ν+ = ξ= d v v ν+ ξ dξ, where v ν+ ξ is given by 3.. By the admissibility of the dissipative NS flues d v and the RHS of 3.5 is indeed non-positive. Thus, the semi-discrete scheme 3.9a guarantees the total entropy dissipation. In the specific case of the entropy pair U, F = ρs, ms, the entropy variable are found in.4, and we eplicitly compute the inner products in 3.4a as omitting all subscripts, ν ɛ v, d = ν = λ +µ ν = λ +µ ν { q λ +µ q λ +µ θ q + θ ν+ θ ν q + θ ν+ θ ν q κ θ κ ν κ ν θ } θ θ θ θ. θ ν+ θ ν The discrete entropy balance. now follows. We emphasize the main point made here, namely, we introduce no ecessive entropy dissipation due to spurious, artificial numerical viscosity: by 3., the semi-discrete scheme contains the precise amount of numerical viscosity to enforce the correct entropy dissipation dictated by the NS equations Time discretization. To complete the computation of a semi-discrete scheme, it needs to be augmented with a proper time discretization. To enable a large time-stability region and maintain simplicity, the three-stage third-order Runge-Kutta RK3 method will be used, consult [GST], u ν t n + t =u ν t n + t 9 K ν, +3K ν, +4K ν,3, 3.6

14 4 Eitan Tadmor and Weigang Zhong where K ν, = {f ν+ u n ν, u n ν+ f ν u n ν, u n } ν ν K ν, = { f u n ν+ ν ν + t K ν,, u n ν+ + t K ν+, f u n ν ν + t K ν,, u n ν + t } K ν, K ν,3 = { f u n ν+ ν + 3 t ν 4 K ν,, u n ν+ + 3 t 4 K ν+, f u n ν ν + 3 t 4 K ν,, u n ν + 3 t } 4 K ν, The resulting fully-discrete schemes has a spatial stencil involving seven-point gridvalues, with two ghost boundary values on the left boundary and two ghost boundary on the right required to close the system. For simplicity, these ghost values are etrapolated from the given Dirichlet boundary values. We note in passing that though the fully eplicit RK3 time discretization need not conserve the entropy, it introduces a negligible amount of entropy dissipation; for a general framework of entropy-conservative fully discrete schemes consult [LMR]. 4. Numerical eperiments We consider ideal polytropic gas equations as an approimation of air with γ =.4, C v = 76, κ =.3, λ+µ =.8 5 We simulate the Sod s shocktube problem, [Sod978], where the Euler and NS equations are solved over the interval [, ] subject to Riemann initial conditions {.,.,.5 <.5 ρ, m, E t= =.5,.,.5.5 <<. In the following figures, we display the numerical solutions for the fully discrete scheme 3.6 with the numerical flu 3.7, or in its equivalent yet simpler form 3.. Uniform space and time grid sizes, and t, are used. Both viscous and inviscid cases are eplored. We use different spatial resolutions for the same problem, and adjust time step according to the CFL condition. Different choices of entropy function are also tested in the numerical eperiments. We group our results into four sets.. Euler equations. The first four sets of figures are devoted to the Euler equations with zero viscous and heat flues.5. With the choice of the entropy pair +γ Uu, Fu = γ ρp +γ, +γ γ qρp +γ, 4. Figure 4. depicts the density, velocity, and pressure fields at t =.5 and t =.; here we use =. and t =.5. Comparing these to the corresponding results of the canonical entropy pair Uu, Fu = ρ lnpρ γ, m lnpρ γ, 4. in figure 4., we see that the different choices of entropies do not affect the behavior of the numerical solutions. Figures 4.d and 4.d demonstrate the conservation of the total entropies: the negligible amount of entropy decay 4 is introduced by the RK3 time discretization. t Net, we make the same comparison for the refined the spatial mesh, taking =.5, =.. Figure 4.3 presents the computed solutions of density, velocity, and pressure fields at t =.5 and.

15 Entropy stable approimations for the Navier-Stokes equations 5 t =. with the entropy pair 4. while figure 4.4 depicts the solutions with the canonical entropy pair 4.. The total entropy is shown in figures 4.3d and 4.4d. The above results demonstrate the purely dispersive character of the entropy conservative schemes. Dispersive oscillations on the mesh scale are observed in shocks and contact regions, due to the absence of any dissipation mechanism, consult [La986, LevermoreLiu996]. The numerical solutions do not blow up. Actually, as we refine the mesh, these dispersive oscillations approach a modulated wave envelope. The study of these modulated waves in the conservative Euler equations would be an etremely challenging task. A similar entropy conservative Lagrangian formulation of Euler equations of [TrulioTrigger96] motivated the discussion in [La986]. Navier-Stokes equations with heat flu. We solve the Navier-Stokes equations.4. The results are summarized in the net three sets of figures We follow the same pattern of plotting density, velocity, pressure and total entropy. As before, the choice of entropy pairs 4. in figures 4.5 and 4.6 are very similar. The presence of heat flu causes the oscillations to be dramatically reduced around the contact discontinuity. Furthermore, oscillations are significantly damped around the shock; when the mesh is well-refined, figure 4.7 shows that heat conduction causes these oscillations to be well localized in the immediate neighborhood of the shocks. If the mesh is underresolved, a small portion of dispersive oscillations persist in the neighborhood of shocks. 3. Navier-stokes equations with viscosity and no heat flu. We solve the viscous NS equations.3. The results are summarized in figures Since the results are essentially independent of the choice of entropy, we chose to quote here only the results for the canonical pair 4.. The viscosity in NS equations is doing a better job than heat flu in damping oscillations around the shock discontinuity. The plots of total entropy, reveal a greater entropy decay than the NS equations with heat conduction. On the other hand, we still observe an oscillatory behavior around the contact discontinuity, even with the refined mesh in figure Full Navier-stokes equations with viscous and heat flues. In figures we record the results for the full NS equations.. As before, the difference due to different entropy functions is undetectable and we chose to record here only the canonical entropy. As epected, these numerical solutions are the smoothest ones found in our numerical eperiments. especially in very fine meshes, depicted in figure 4.. Small oscillations remain with underresolved meshes. Not only the oscillations around the shocks are damped out by viscosity, but the oscillations around the contact discontinuity are significantly reduced due to the heat flu. Compared with the results of NS equations with heat conduction.4 in figures , oscillations in the neighborhood of the shock are better damped here thanks to the viscosity terms. The remaining sharp spike at the tip of shock discontinuity is due to the relatively small viscosity coefficient of air. Acknowledgment. Weigang Zhong wants to thank Ning Jiang, Dongming Wei, and Bin Cheng for their very intriguing discussions. Research was supported in part by NSF DMS #DMS4-774 and ONR #N4-9-J-76. References [Dafermos] C. Dafermos, Hyperbolic Conservation Laws in Continuum Mechanics, Springer,. [Godunov96] S. K. Godunov, An interesting class of quasilinear systems, Dokl. Acad. Nauk. SSSR, 39, 96, [GST] S. Gottlieb, C.-W. Shu, and E. Tadmor, Strong stability-preserving high-order time discretization methods, SIAM Review, 43,,, 89. [GrootMazur984] S. R. de Groot and P. Mazur, Non-equilibrium Thermodynamics, Dover, 984. [Harten983] A. Harten, On the symmetric form of systems of conservation laws with entropy, J. Comput. Phys , 5-64.

16 6 Eitan Tadmor and Weigang Zhong [HFMM986] T. J. R. Hughes, with L. P. Franca and M. Mallet, A new finite element formulation for computational fluid dynamics: I. Symmetric forms of the compressible Euler and Navier-stokes equations and the second law of thermodynamics, Computer Methods in Applied Mechanics and Engineering, , 3 34; with M Mallet and A Mizukami, A new finite element formulation for computational fluid dynamics: II. Beyond SUPG Computer Methods in Applied Mechanics and Engineering, , ; with M. Mallet, A new finite element formulation for computational fluid dynamics: III. The generalized streamline operator for multidimensional advective-diffusive systems, Computer Methods in Applied Mechanics and Engineering, , 35 38; with M. Mallet, A new finite element formulation for computational fluid dynamics: IV. A discontinuity- capturing operator for multidimensional advective-diffusive systems, Computer Methods in Applied Mechanics and Engineering, , [La973] P. D. La, Hyperbolic Systems of Conservation Laws and the Mathematical Theory of Shock Waves, CBMS- NSF, SIAM, 973. [La986] P. D. La, On dispersive difference schemes, Physica D, 8, 986, [LLV993] P. D. La, C. D. Levermore and S. Venakides, The generation and propagation of oscillations in dispersive IVPs and their limiting behavior, in Important Developments in Soliton Theory 98-99, T. Fokas and V.E. Zakharov eds, Springer-Verlag, 993. [LeFlochRhode] P. G. LeFloch and C. Rohde, High-order schemes, entropy inequalities, and nonclassical shocks, SIAM J. Numer. Anal., 37, no. 6, 3 6. [LMR] P. G. LeFloch, J. M. Mercier, and C. Rohde, Full discrete, entropy conservative schemes of arbitrary order, SIAM J. Numer. Anal., 4, no. 5, [LevermoreLiu996] C. D. Levermore and J. -G. Liu Oscillations arising in numerical eperiments, Physica D, 99, 996, 9. [Liu99] T. -P. Liu, Viscosity criterion for hyperbolic conservation laws, in Viscous Profiles and Numerical Methods for Shock Waves, SIAM. M. Shearer ed. 99, pp [Mock98] M. S. Mock, Systems of conservation of mied type, J. Diff. Eqns, 37, 98, [Roe98] P. L. Roe, Approimate Riemann solvers, parameter vectors and difference schemes, J. Comp. Phys., 43 98, [Sod978] G. Sod, A survey of several difference methods for hyperbolic systems of nonlinear conservation laws, JCP, 7 978, 3. [Tadmor984a] E. Tadmor, Skew-selfadjoint form for systems of conservation laws, J. Math. Anal. Appl., 3, 984, [Tadmor984b] E. Tadmor, Numerical viscosity and the entropy condition for conservative difference schemes, Math. Comp., 43, 984, [Tadmor986] E. Tadmor, Entropy conservative finite element schemes in Numerical Methods for Compressible Flows - Finite Difference Element and Volume Techniques, Proc. winter annual meeting of the ASME AMD-Vol. 78 T. E. Tezduyar and T.J.R. Hughes, eds. 986, [Tadmor987] E. Tadmor, The numerical viscosity of entropy stable schemes for systems of conservation laws, I, Math. Comp., 49, 987, 9 3. [Tadmor4] E. Tadmor, Entropy stability theory for difference approimations of nonlinear conservation laws and related time-dependent problems, Acta Numerica, 4, [TrulioTrigger96] J. G. Trulio and K. R. Trigger, Numerical solution of one dimensional hydrodynamical shock problem, UCRL Report 65, 96.

17 Entropy stable approimations for the Navier-Stokes equations 7.9 D Euler,density,γ=.4,C v =76,κ=,entropy=+γ/ γ*ρ p /+γ, t=.5e 5, =.,T ma =. 5 t=..6.4 D Euler,velocity,γ=.4,C v =76,κ=,entropy=+γ/ γ*ρ p /+γ, t=.5e 5, =.,T ma =. 5 t=...7 ρ u a Density b Velocity.9 D Euler,pressure,γ=.4,C v =76,κ=,entropy=+γ/ γ*ρ p /+γ, t=.5e 5, =.,T ma =. 5 t= D Euler,entropy conservative scheme w/ 3rd R K mtd,entrop y vs time,entropy = +γ/ γ*ρ p /+γ p.5 U total entropy t c Pressure d Total entropy v.s. time Figure 4.. Euler equations with spatial gridpoints, Uu = +γ γ pρ +γ, t =.5 5, = 3

18 8 Eitan Tadmor and Weigang Zhong.9 D Euler,density,γ=.4,C v =76,κ=,entropy= ρ lnpρ γ, t=.5e 5, =.,T ma =. 5 t=..6.4 D Euler,velocity,γ=.4,C v =76,κ=,entropy= ρ lnpρ γ, t=.5e 5, =.,T ma =. 5 t=...7 ρ u a Density b Velocity.9 D Euler,pressure,γ=.4,C v =76,κ=,entropy= ρ lnpρ γ, t=.5e 5, =.,T ma =. 5 t= D Euler, entropy conservative scheme w/ 3rd R K mtd, entr opy vs time,entropy = ρ lnpρ γ p.5.3. U total entropy t c Pressure d Total entropy v.s. time Figure 4.. Euler equations with spatial gridpoints, Uu = ρ ln pρ γ and same t and as figure 4.

19 Entropy stable approimations for the Navier-Stokes equations 9.9 D Euler,density,γ=.4,C v =76,κ=,entropy=+γ/ γ*ρ p /+γ, t=.5e 5, =.5,T ma =. 5 t=..6.4 D Euler,velocity,γ=.4,C v =76,κ=,entropy=+γ/ γ*ρ p /+γ, t=.5e 5, =.5,T ma =. 5 t=...7 ρ u a Density b Velocity.9 D Euler,pressure,γ=.4,C v =76,κ=,entropy=+γ/ γ*ρ p /+γ, t=.5e 5, =.5,T ma =. 5 t= D Euler,entropy conservative scheme w/ 3rd R K mtd,entrop y vs time,entropy = +γ/ γ*ρ p /+γ p U total entropy t c Pressure d Total entropy v.s. time Figure 4.3. Euler equations with 4 spatial gridpoints, Uu = +γ γ pρ +γ, t =.5 5, =.5 4

20 Eitan Tadmor and Weigang Zhong.9 D Euler,density,γ=.4,C v =76,κ=,entropy= ρ lnpρ γ, t=.5e 5, =.5,T ma =. 5 t=..6.4 D Euler,velocity,γ=.4,C v =76,κ=,entropy= ρ lnpρ γ, t=.5e 5, =.5,T ma =. 5 t=...7 ρ u a Density b Velocity.9 D Euler,pressure,γ=.4,C v =76,κ=,entropy= ρ lnpρ γ, t=.5e 5, =.5,T ma =. 5 t= D Euler, entropy conservative scheme w/ 3rd R K mtd, entr opy vs time,entropy = ρ lnpρ γ.38.7 p.5 U total entropy t c Pressure d Total entropy v.s. time Figure 4.4. Euler equations with 4 spatial gridpoints, Uu = ρ ln pρ γ and same t and as Figure 4.3

21 Entropy stable approimations for the Navier-Stokes equations D Euler,density,γ=.4,C v =76,κ=.3,entropy=+γ/ γ*ρ p /+γ, t=.5e 5, =.,T ma =. 5.9 t=..4. D Euler,velocity,γ=.4,C v =76,κ=.3,entropy=+γ/ γ*ρ p /+γ, t=.5e 5, =.,T ma =. 5 t=..7 ρ u a Density b Velocity.9 D Euler,pressure,γ=.4,C v =76,κ=.3,entropy=+γ/ γ*ρ p /+γ, t=.5e 5, =.,T ma =. 5 t= D Euler,entropy conservative scheme w/ 3rd R K mtd,entrop y vs time,entropy = +γ/ γ*ρ p /+γ p.5 U total entropy t c Pressure d Total entropy v.s. time Figure 4.5. Navier-Stokes equations with heat conduction and no viscous term. spatial gridpoints, Uu = +γ γ pρ +γ, t =.5 5, = 3

22 Eitan Tadmor and Weigang Zhong.9 D Euler,density,γ=.4,C v =76,κ=.3,entropy= ρ lnpρ γ, t=.5e 5, =.,T ma =. 5 t=..4. D Euler,velocity,γ=.4,C v =76,κ=.3,entropy= ρ lnpρ γ, t=.5e 5, =.,T ma =. 5 t=..7 ρ u a Density b Velocity.9 D Euler,pressure,γ=.4,C v =76,κ=.3,entropy= ρ lnpρ γ, t=.5e 5, =.,T ma =. 5 t= D Euler, entropy conservative scheme w/ 3rd R K mtd, entr opy vs time,entropy = ρ lnpρ γ p.5.3 U total entropy t c Pressure d Total entropy v.s. time Figure 4.6. Navier-Stokes equations with heat conduction and no viscous terms. spatial gridpoints, Uu = ρ ln pρ γ and same t and as Figure 4.5

23 Entropy stable approimations for the Navier-Stokes equations 3.9 D Euler,density,γ=.4,C v =76,κ=.3,entropy= ρ lnpρ γ, t=.5e 5, =.5,T ma =. 5 t=..4. D Euler,velocity,γ=.4,C v =76,κ=.3,entropy= ρ lnpρ γ, t=.5e 5, =.5,T ma =. 5 t=..7 ρ u a Density b Velocity.9 D Euler,pressure,γ=.4,C v =76,κ=.3,entropy= ρ lnpρ γ, t=.5e 5, =.5,T ma =. 5 t= D Euler, entropy conservative scheme w/ 3rd R K mtd, entr opy vs time,entropy = ρ lnpρ γ p.5 U total entropy t c Pressure d Total entropy v.s. time Figure 4.7. Navier-Stokes equations with heat conduction and no viscous terms. 4 spatial grids, Uu = ρ ln pρ γ, t =.5 5, =.5 4

24 4 Eitan Tadmor and Weigang Zhong.9 D N S,density,λ+µ=.8e 5,γ=.4,C v =76,κ=,entropy= ρ lnpρ γ, t=.5e 5, =. 5 t=..4. D N S,velocity,λ+µ=.8e 5,γ=.4,C v =76,κ=,entropy= ρ lnpρ γ, t=.5e 5, =. 5 t=..7 ρ u a Density b Velocity.9 D N S,pressure,λ+µ=.8e 5,γ=.4,C v =76,κ=,entropy= ρ lnpρ γ, t=.5e 5, =. 5 t= D N S, entropy conservative scheme w/ 3rd R K mtd,entropy vs time,entropy = ρ lnpρ γ p.5.3 U total entropy t c Pressure d Total entropy v.s. time Figure 4.8. Navier-Stokes equations with viscous terms and no heat conduction. spatial gridpoints, Uu = ρ ln pρ γ, t =.5 5, = 3

25 Entropy stable approimations for the Navier-Stokes equations 5.9 D N S,density,λ+µ=.8e 5,γ=.4,C v =76,κ=,entropy= ρ lnpρ γ, t=.5e 5, =.5 5 t=.. D N S,velocity,λ+µ=.8e 5,γ=.4,C v =76,κ=,entropy= ρ lnpρ γ, t=.5e 5, =.5 5 t=..7 ρ u a Density b Velocity.9 D N S,pressure,λ+µ=.8e 5,γ=.4,C v =76,κ=,entropy= ρ lnpρ γ, t=.5e 5, =.5 5 t= D N S, entropy conservative scheme w/ 3rd R K mtd,entropy vs time,entropy = ρ lnpρ γ p.5 U total entropy t c Pressure d Total entropy v.s. time Figure 4.9. Navier-Stokes equations with viscous terms and no heat conduction. 4 spatial grids, Uu = ρ ln pρ γ, t =.5 5, =.5 4

26 6 Eitan Tadmor and Weigang Zhong.9 D N S,density,λ+µ=.8e 5,γ=.4,C v =76,κ=.3,entropy= ρ lnpρ γ, t=.5e 5, =. 5 t=..4. D N S,velocity,λ+µ=.8e 5,γ=.4,C v =76,κ=.3,entropy= ρ lnpρ γ, t=.5e 5, =. 5 t=..7 ρ u a Density b Velocity.9 D N S,pressure,λ+µ=.8e 5,γ=.4,C v =76,κ=.3,entropy= ρ lnpρ γ, t=.5e 5, =. 5 t= D N S, entropy conservative scheme w/ 3rd R K mtd,entropy vs time,entropy = ρ lnpρ γ p.5 U total entropy t c Pressure d Total entropy v.s. time Figure 4.. Navier-Stokes equations with viscosity and heat conduction. spatial gridpoints, Uu = ρ ln pρ γ, t =.5 5, = 3

27 Entropy stable approimations for the Navier-Stokes equations 7.9 D N S,density,λ+µ=.8e 5,γ=.4,C v =76,κ=.3,entropy= ρ lnpρ γ, t=.5e 5, =.5 5 t=.. D N S,velocity,λ+µ=.8e 5,γ=.4,C v =76,κ=.3,entropy= ρ lnpρ γ, t=.5e 5, =.5 5 t=..7 ρ u a Density b Velocity.9 D N S,pressure,λ+µ=.8e 5,γ=.4,C v =76,κ=.3,entropy= ρ lnpρ γ, t=.5e 5, =.5 5 t= D N S, entropy conservative scheme w/ 3rd R K mtd,entropy vs time,entropy = ρ lnpρ γ p U total entropy t c Pressure d Total entropy v.s. time Figure 4.. Navier-Stokes equations with viscosity and heat conduction. 4 spatial gridpoints, Uu = ρ ln pρ γ, t =.5 5, =.5 4

28 8 Eitan Tadmor and Weigang Zhong Department of Mathematics, Center for Scientific Computation and Mathematical Modeling CSCAMM and Institute for Physical Science and Technology IPST, University of Maryland, MD 74. address: Department of Mathematics, Center for Scientific Computation and Mathematical Modelling CSCAMM, University of Maryland, College Park, MD 74 address:

ENTROPY STABLE APPROXIMATIONS OF NAVIER STOKES EQUATIONS WITH NO ARTIFICIAL NUMERICAL VISCOSITY

ENTROPY STABLE APPROXIMATIONS OF NAVIER STOKES EQUATIONS WITH NO ARTIFICIAL NUMERICAL VISCOSITY Journal of Hyperbolic Differential Equations Vol. 3, No. 3 (006 59 559 c World Scientific Publishing Company ENTROPY STABLE APPROXIMATIONS OF NAVIER STOKES EQUATIONS WITH NO ARTIFICIAL NUMERICAL VISCOSITY

More information

Computation of entropic measure-valued solutions for Euler equations

Computation of entropic measure-valued solutions for Euler equations Computation of entropic measure-valued solutions for Euler equations Eitan Tadmor Center for Scientific Computation and Mathematical Modeling (CSCAMM) Department of Mathematics, Institute for Physical

More information

Novel entropy stable schemes for 1D and 2D fluid equations

Novel entropy stable schemes for 1D and 2D fluid equations Novel entropy stable schemes for D an D flui equations Eitan Tamor an Weigang Zhong Department of Mathematics, Center for Scientific Computation an Mathematical Moeling (CSCAMM, Institute of Physical Science

More information

On the entropy stability of Roe-type finite volume methods

On the entropy stability of Roe-type finite volume methods Proceedings of Symposia in Applied Mathematics On the entropy stability of Roe-type finite volume methods Mária Lukáčová - Medvid ová and Eitan Tadmor Abstract. We study the entropy stability of a class

More information

A NUMERICAL STUDY FOR THE PERFORMANCE OF THE RUNGE-KUTTA FINITE DIFFERENCE METHOD BASED ON DIFFERENT NUMERICAL HAMILTONIANS

A NUMERICAL STUDY FOR THE PERFORMANCE OF THE RUNGE-KUTTA FINITE DIFFERENCE METHOD BASED ON DIFFERENT NUMERICAL HAMILTONIANS A NUMERICAL STUDY FOR THE PERFORMANCE OF THE RUNGE-KUTTA FINITE DIFFERENCE METHOD BASED ON DIFFERENT NUMERICAL HAMILTONIANS HASEENA AHMED AND HAILIANG LIU Abstract. High resolution finite difference methods

More information

ON THE GLOBAL REGULARITY OF SUB-CRITICAL EULER-POISSON EQUATIONS WITH PRESSURE

ON THE GLOBAL REGULARITY OF SUB-CRITICAL EULER-POISSON EQUATIONS WITH PRESSURE ON THE GLOBAL REGULARITY OF SUB-CRITICAL EULER-POISSON EQUATIONS WITH PRESSURE EITAN TADMOR AND DONGMING WEI Abstract We prove that the one-dimensional Euler-Poisson system driven by the Poisson forcing

More information

Variational formulation of entropy solutions for nonlinear conservation laws

Variational formulation of entropy solutions for nonlinear conservation laws Variational formulation of entropy solutions for nonlinear conservation laws Eitan Tadmor 1 Center for Scientific Computation and Mathematical Modeling CSCAMM) Department of Mathematics, Institute for

More information

A New Fourth-Order Non-Oscillatory Central Scheme For Hyperbolic Conservation Laws

A New Fourth-Order Non-Oscillatory Central Scheme For Hyperbolic Conservation Laws A New Fourth-Order Non-Oscillatory Central Scheme For Hyperbolic Conservation Laws A. A. I. Peer a,, A. Gopaul a, M. Z. Dauhoo a, M. Bhuruth a, a Department of Mathematics, University of Mauritius, Reduit,

More information

A numerical study of SSP time integration methods for hyperbolic conservation laws

A numerical study of SSP time integration methods for hyperbolic conservation laws MATHEMATICAL COMMUNICATIONS 613 Math. Commun., Vol. 15, No., pp. 613-633 (010) A numerical study of SSP time integration methods for hyperbolic conservation laws Nelida Črnjarić Žic1,, Bojan Crnković 1

More information

Energy-Preserving and Stable Approximations for the Two-Dimensional Shallow Water Equations

Energy-Preserving and Stable Approximations for the Two-Dimensional Shallow Water Equations Energy-Preserving and Stable Approximations for the Two-Dimensional Shallow Water Equations Eitan Tadmor* and Weigang Zhong Department of Mathematics, Center for Scientific Computation and Mathematical

More information

Non-linear Wave Propagation and Non-Equilibrium Thermodynamics - Part 3

Non-linear Wave Propagation and Non-Equilibrium Thermodynamics - Part 3 Non-linear Wave Propagation and Non-Equilibrium Thermodynamics - Part 3 Tommaso Ruggeri Department of Mathematics and Research Center of Applied Mathematics University of Bologna January 21, 2017 ommaso

More information

A high order adaptive finite element method for solving nonlinear hyperbolic conservation laws

A high order adaptive finite element method for solving nonlinear hyperbolic conservation laws A high order adaptive finite element method for solving nonlinear hyperbolic conservation laws Zhengfu Xu, Jinchao Xu and Chi-Wang Shu 0th April 010 Abstract In this note, we apply the h-adaptive streamline

More information

Gas Dynamics Equations: Computation

Gas Dynamics Equations: Computation Title: Name: Affil./Addr.: Gas Dynamics Equations: Computation Gui-Qiang G. Chen Mathematical Institute, University of Oxford 24 29 St Giles, Oxford, OX1 3LB, United Kingdom Homepage: http://people.maths.ox.ac.uk/chengq/

More information

Antony Jameson. AIAA 18 th Computational Fluid Dynamics Conference

Antony Jameson. AIAA 18 th Computational Fluid Dynamics Conference Energy Estimates for Nonlinear Conservation Laws with Applications to Solutions of the Burgers Equation and One-Dimensional Viscous Flow in a Shock Tube by Central Difference Schemes Antony Jameson AIAA

More information

arxiv: v1 [math.na] 30 Jan 2018

arxiv: v1 [math.na] 30 Jan 2018 ENTROPY CONSERVATIVE SCHEMES AND THE RECEDING FLOW PROBLEM AYOUB GOUASMI, SCOTT MURMAN, AND KARTHIK DURAISAMY arxiv:1801.1013v1 [math.na] 30 Jan 018 Abstract. This work delves into the family of entropy

More information

Piecewise Smooth Solutions to the Burgers-Hilbert Equation

Piecewise Smooth Solutions to the Burgers-Hilbert Equation Piecewise Smooth Solutions to the Burgers-Hilbert Equation Alberto Bressan and Tianyou Zhang Department of Mathematics, Penn State University, University Park, Pa 68, USA e-mails: bressan@mathpsuedu, zhang

More information

HFVS: An Arbitrary High Order Flux Vector Splitting Method

HFVS: An Arbitrary High Order Flux Vector Splitting Method HFVS: An Arbitrary High Order Flu Vector Splitting Method Yibing Chen, Song Jiang and Na Liu Institute of Applied Physics and Computational Mathematics, P.O. Bo 8009, Beijing 00088, P.R. China E-mail:

More information

arxiv: v1 [cs.na] 22 Sep 2012

arxiv: v1 [cs.na] 22 Sep 2012 Kinetic energy preserving and entropy stable finite volume schemes for compressible Euler and Navier- Stokes equations Praveen Chandrashekar TIFR Center for Applicable Mathematics, Bangalore 5665, India

More information

Hyperbolic Systems of Conservation Laws. in One Space Dimension. II - Solutions to the Cauchy problem. Alberto Bressan

Hyperbolic Systems of Conservation Laws. in One Space Dimension. II - Solutions to the Cauchy problem. Alberto Bressan Hyperbolic Systems of Conservation Laws in One Space Dimension II - Solutions to the Cauchy problem Alberto Bressan Department of Mathematics, Penn State University http://www.math.psu.edu/bressan/ 1 Global

More information

The inviscid limit to a contact discontinuity for the compressible Navier-Stokes-Fourier system using the relative entropy method

The inviscid limit to a contact discontinuity for the compressible Navier-Stokes-Fourier system using the relative entropy method The inviscid limit to a contact discontinuity for the compressible Navier-Stokes-Fourier system using the relative entropy method Alexis Vasseur, and Yi Wang Department of Mathematics, University of Texas

More information

Finite volumes for complex applications In this paper, we study finite-volume methods for balance laws. In particular, we focus on Godunov-type centra

Finite volumes for complex applications In this paper, we study finite-volume methods for balance laws. In particular, we focus on Godunov-type centra Semi-discrete central schemes for balance laws. Application to the Broadwell model. Alexander Kurganov * *Department of Mathematics, Tulane University, 683 St. Charles Ave., New Orleans, LA 708, USA kurganov@math.tulane.edu

More information

Entropy stable schemes for degenerate convection-diffusion equations

Entropy stable schemes for degenerate convection-diffusion equations Entropy stable schemes for degenerate convection-diffusion equations Silvia Jerez 1 Carlos Parés 2 ModCompShock, Paris 6-8 Decmber 216 1 CIMAT, Guanajuato, Mexico. 2 University of Malaga, Málaga, Spain.

More information

FUNDAMENTALS OF LAX-WENDROFF TYPE APPROACH TO HYPERBOLIC PROBLEMS WITH DISCONTINUITIES

FUNDAMENTALS OF LAX-WENDROFF TYPE APPROACH TO HYPERBOLIC PROBLEMS WITH DISCONTINUITIES 6th European Conference on Computational Mechanics (ECCM 6) 7th European Conference on Computational Fluid Dynamics (ECFD 7) 1115 June 018, Glasgow, UK FUNDAMENTALS OF LAX-WENDROFF TYPE APPROACH TO HYPERBOLIC

More information

c 2012 Society for Industrial and Applied Mathematics

c 2012 Society for Industrial and Applied Mathematics SIAM J. NUMER. ANAL. Vol. 5, No., pp. 544 573 c Society for Industrial and Applied Mathematics ARBITRARILY HIGH-ORDER ACCURATE ENTROPY STABLE ESSENTIALLY NONOSCILLATORY SCHEMES FOR SYSTEMS OF CONSERVATION

More information

A Central Compact-Reconstruction WENO Method for Hyperbolic Conservation Laws

A Central Compact-Reconstruction WENO Method for Hyperbolic Conservation Laws A Central Compact-Reconstruction WENO Method for Hyperbolic Conservation Laws Kilian Cooley 1 Prof. James Baeder 2 1 Department of Mathematics, University of Maryland - College Park 2 Department of Aerospace

More information

A Fourth-Order Central Runge-Kutta Scheme for Hyperbolic Conservation Laws

A Fourth-Order Central Runge-Kutta Scheme for Hyperbolic Conservation Laws A Fourth-Order Central Runge-Kutta Scheme for Hyperbolic Conservation Laws Mehdi Dehghan, Rooholah Jazlanian Department of Applied Mathematics, Faculty of Mathematics and Computer Science, Amirkabir University

More information

Entropy-stable discontinuous Galerkin nite element method with streamline diusion and shock-capturing

Entropy-stable discontinuous Galerkin nite element method with streamline diusion and shock-capturing Entropy-stable discontinuous Galerkin nite element method with streamline diusion and shock-capturing Siddhartha Mishra Seminar for Applied Mathematics ETH Zurich Goal Find a numerical scheme for conservation

More information

Semi-Discrete Central-Upwind Schemes with Reduced Dissipation for Hamilton-Jacobi Equations

Semi-Discrete Central-Upwind Schemes with Reduced Dissipation for Hamilton-Jacobi Equations Semi-Discrete Central-Upwind Schemes with Reduced Dissipation for Hamilton-Jacobi Equations Steve Bryson, Aleander Kurganov, Doron Levy and Guergana Petrova Abstract We introduce a new family of Godunov-type

More information

Solution of Two-Dimensional Riemann Problems for Gas Dynamics without Riemann Problem Solvers

Solution of Two-Dimensional Riemann Problems for Gas Dynamics without Riemann Problem Solvers Solution of Two-Dimensional Riemann Problems for Gas Dynamics without Riemann Problem Solvers Alexander Kurganov, 1, * Eitan Tadmor 2 1 Department of Mathematics, University of Michigan, Ann Arbor, Michigan

More information

The Riemann problem and a high-resolution Godunov method for a model of compressible two-phase flow

The Riemann problem and a high-resolution Godunov method for a model of compressible two-phase flow The Riemann problem and a high-resolution Godunov method for a model of compressible two-phase flow D.W. Schwendeman, C.W. Wahle 2, and A.K. Kapila Department of Mathematical Sciences, Rensselaer Polytechnic

More information

Foundations of Computational Mathematics: Hong Kong 2008

Foundations of Computational Mathematics: Hong Kong 2008 London Mathematical Society Lecture Note Series: 363 Foundations of Computational Mathematics: Hong Kong 008 Edited by Felipe Cucker City University of Hong Kong Allan Pinkus Technion Mike Todd Cornell

More information

Fourier analysis for discontinuous Galerkin and related methods. Abstract

Fourier analysis for discontinuous Galerkin and related methods. Abstract Fourier analysis for discontinuous Galerkin and related methods Mengping Zhang and Chi-Wang Shu Abstract In this paper we review a series of recent work on using a Fourier analysis technique to study the

More information

Hierarchical Reconstruction with up to Second Degree Remainder for Solving Nonlinear Conservation Laws

Hierarchical Reconstruction with up to Second Degree Remainder for Solving Nonlinear Conservation Laws Hierarchical Reconstruction with up to Second Degree Remainder for Solving Nonlinear Conservation Laws Dedicated to Todd F. Dupont on the occasion of his 65th birthday Yingjie Liu, Chi-Wang Shu and Zhiliang

More information

The RAMSES code and related techniques I. Hydro solvers

The RAMSES code and related techniques I. Hydro solvers The RAMSES code and related techniques I. Hydro solvers Outline - The Euler equations - Systems of conservation laws - The Riemann problem - The Godunov Method - Riemann solvers - 2D Godunov schemes -

More information

AMath 574 February 11, 2011

AMath 574 February 11, 2011 AMath 574 February 11, 2011 Today: Entropy conditions and functions Lax-Wendroff theorem Wednesday February 23: Nonlinear systems Reading: Chapter 13 R.J. LeVeque, University of Washington AMath 574, February

More information

Author(s) Huang, Feimin; Matsumura, Akitaka; Citation Osaka Journal of Mathematics. 41(1)

Author(s) Huang, Feimin; Matsumura, Akitaka; Citation Osaka Journal of Mathematics. 41(1) Title On the stability of contact Navier-Stokes equations with discont free b Authors Huang, Feimin; Matsumura, Akitaka; Citation Osaka Journal of Mathematics. 4 Issue 4-3 Date Text Version publisher URL

More information

Hyperbolic Systems of Conservation Laws. I - Basic Concepts

Hyperbolic Systems of Conservation Laws. I - Basic Concepts Hyperbolic Systems of Conservation Laws I - Basic Concepts Alberto Bressan Mathematics Department, Penn State University Alberto Bressan (Penn State) Hyperbolic Systems of Conservation Laws 1 / 27 The

More information

Commun Nonlinear Sci Numer Simulat

Commun Nonlinear Sci Numer Simulat Commun Nonlinear Sci Numer Simulat 7 (0) 3499 3507 Contents lists available at SciVerse ScienceDirect Commun Nonlinear Sci Numer Simulat journal homepage: www.elsevier.com/locate/cnsns Application of the

More information

Hyperbolic Problems: Theory, Numerics, Applications

Hyperbolic Problems: Theory, Numerics, Applications AIMS on Applied Mathematics Vol.8 Hyperbolic Problems: Theory, Numerics, Applications Fabio Ancona Alberto Bressan Pierangelo Marcati Andrea Marson Editors American Institute of Mathematical Sciences AIMS

More information

Équation de Burgers avec particule ponctuelle

Équation de Burgers avec particule ponctuelle Équation de Burgers avec particule ponctuelle Nicolas Seguin Laboratoire J.-L. Lions, UPMC Paris 6, France 7 juin 2010 En collaboration avec B. Andreianov, F. Lagoutière et T. Takahashi Nicolas Seguin

More information

1 Introduction. J.-L. GUERMOND and L. QUARTAPELLE 1 On incremental projection methods

1 Introduction. J.-L. GUERMOND and L. QUARTAPELLE 1 On incremental projection methods J.-L. GUERMOND and L. QUARTAPELLE 1 On incremental projection methods 1 Introduction Achieving high order time-accuracy in the approximation of the incompressible Navier Stokes equations by means of fractional-step

More information

Projection Dynamics in Godunov-Type Schemes

Projection Dynamics in Godunov-Type Schemes JOURNAL OF COMPUTATIONAL PHYSICS 142, 412 427 (1998) ARTICLE NO. CP985923 Projection Dynamics in Godunov-Type Schemes Kun Xu and Jishan Hu Department of Mathematics, Hong Kong University of Science and

More information

AA214B: NUMERICAL METHODS FOR COMPRESSIBLE FLOWS

AA214B: NUMERICAL METHODS FOR COMPRESSIBLE FLOWS AA214B: NUMERICAL METHODS FOR COMPRESSIBLE FLOWS 1 / 43 AA214B: NUMERICAL METHODS FOR COMPRESSIBLE FLOWS Treatment of Boundary Conditions These slides are partially based on the recommended textbook: Culbert

More information

Direct Numerical Simulations of a Two-dimensional Viscous Flow in a Shocktube using a Kinetic Energy Preserving Scheme

Direct Numerical Simulations of a Two-dimensional Viscous Flow in a Shocktube using a Kinetic Energy Preserving Scheme 19th AIAA Computational Fluid Dynamics - 5 June 009, San Antonio, Texas AIAA 009-3797 19th AIAA Computational Fluid Dynamics Conference, - 5 June 009, San Antonio, Texas Direct Numerical Simulations of

More information

Entropic Schemes for Conservation Laws

Entropic Schemes for Conservation Laws CONSTRUCTVE FUNCTON THEORY, Varna 2002 (B. Bojanov, Ed.), DARBA, Sofia, 2002, pp. 1-6. Entropic Schemes for Conservation Laws Bojan Popov A new class of Godunov-type numerical methods (called here entropic)

More information

Relaxation methods and finite element schemes for the equations of visco-elastodynamics. Chiara Simeoni

Relaxation methods and finite element schemes for the equations of visco-elastodynamics. Chiara Simeoni Relaxation methods and finite element schemes for the equations of visco-elastodynamics Chiara Simeoni Department of Information Engineering, Computer Science and Mathematics University of L Aquila (Italy)

More information

Solving the Euler Equations!

Solving the Euler Equations! http://www.nd.edu/~gtryggva/cfd-course/! Solving the Euler Equations! Grétar Tryggvason! Spring 0! The Euler equations for D flow:! where! Define! Ideal Gas:! ρ ρu ρu + ρu + p = 0 t x ( / ) ρe ρu E + p

More information

Numerical methods for the Navier- Stokes equations

Numerical methods for the Navier- Stokes equations Numerical methods for the Navier- Stokes equations Hans Petter Langtangen 1,2 1 Center for Biomedical Computing, Simula Research Laboratory 2 Department of Informatics, University of Oslo Dec 6, 2012 Note:

More information

CHAPTER 7 SEVERAL FORMS OF THE EQUATIONS OF MOTION

CHAPTER 7 SEVERAL FORMS OF THE EQUATIONS OF MOTION CHAPTER 7 SEVERAL FORMS OF THE EQUATIONS OF MOTION 7.1 THE NAVIER-STOKES EQUATIONS Under the assumption of a Newtonian stress-rate-of-strain constitutive equation and a linear, thermally conductive medium,

More information

Finite Volume Schemes: an introduction

Finite Volume Schemes: an introduction Finite Volume Schemes: an introduction First lecture Annamaria Mazzia Dipartimento di Metodi e Modelli Matematici per le Scienze Applicate Università di Padova mazzia@dmsa.unipd.it Scuola di dottorato

More information

Application of the Kurganov Levy semi-discrete numerical scheme to hyperbolic problems with nonlinear source terms

Application of the Kurganov Levy semi-discrete numerical scheme to hyperbolic problems with nonlinear source terms Future Generation Computer Systems () 65 7 Application of the Kurganov Levy semi-discrete numerical scheme to hyperbolic problems with nonlinear source terms R. Naidoo a,b, S. Baboolal b, a Department

More information

Effects of a Saturating Dissipation in Burgers-Type Equations

Effects of a Saturating Dissipation in Burgers-Type Equations Effects of a Saturating Dissipation in Burgers-Type Equations A. KURGANOV AND P. ROSENAU Tel-Aviv University Abstract We propose and study a new variant of the Burgers equation with dissipation flues that

More information

Chapter 5: The Hydrodynamical Riemann Problem

Chapter 5: The Hydrodynamical Riemann Problem Chapter 5: The Hydrodynamical Riemann Problem 5.) Introduction 5..) General Introduction to the Riemann Problem We have seen in Chapter 4 that even Burgers equation, the simplest non-linear scalar conservation

More information

A Finite Volume Code for 1D Gas Dynamics

A Finite Volume Code for 1D Gas Dynamics A Finite Volume Code for 1D Gas Dynamics Michael Lavell Department of Applied Mathematics and Statistics 1 Introduction A finite volume code is constructed to solve conservative systems, such as Euler

More information

Numerical Solutions for Hyperbolic Systems of Conservation Laws: from Godunov Method to Adaptive Mesh Refinement

Numerical Solutions for Hyperbolic Systems of Conservation Laws: from Godunov Method to Adaptive Mesh Refinement Numerical Solutions for Hyperbolic Systems of Conservation Laws: from Godunov Method to Adaptive Mesh Refinement Romain Teyssier CEA Saclay Romain Teyssier 1 Outline - Euler equations, MHD, waves, hyperbolic

More information

A recovery-assisted DG code for the compressible Navier-Stokes equations

A recovery-assisted DG code for the compressible Navier-Stokes equations A recovery-assisted DG code for the compressible Navier-Stokes equations January 6 th, 217 5 th International Workshop on High-Order CFD Methods Kissimmee, Florida Philip E. Johnson & Eric Johnsen Scientific

More information

Divergence Formulation of Source Term

Divergence Formulation of Source Term Preprint accepted for publication in Journal of Computational Physics, 2012 http://dx.doi.org/10.1016/j.jcp.2012.05.032 Divergence Formulation of Source Term Hiroaki Nishikawa National Institute of Aerospace,

More information

Affordable, entropy-consistent, Euler flux functions

Affordable, entropy-consistent, Euler flux functions Affordable, entropy-consistent, Euler flux functions (with application to the carbuncle phenomenon) Phil Roe Aerospace Engineering University 0f Michigan Ann Arbor Presented at HYP 2006 1 Entropy pairs

More information

Riemann Solvers and Numerical Methods for Fluid Dynamics

Riemann Solvers and Numerical Methods for Fluid Dynamics Eleuterio R Toro Riemann Solvers and Numerical Methods for Fluid Dynamics A Practical Introduction With 223 Figures Springer Table of Contents Preface V 1. The Equations of Fluid Dynamics 1 1.1 The Euler

More information

Weak-Strong Uniqueness of the Navier-Stokes-Smoluchowski System

Weak-Strong Uniqueness of the Navier-Stokes-Smoluchowski System Weak-Strong Uniqueness of the Navier-Stokes-Smoluchowski System Joshua Ballew University of Maryland College Park Applied PDE RIT March 4, 2013 Outline Description of the Model Relative Entropy Weakly

More information

NONCLASSICAL SHOCK WAVES OF CONSERVATION LAWS: FLUX FUNCTION HAVING TWO INFLECTION POINTS

NONCLASSICAL SHOCK WAVES OF CONSERVATION LAWS: FLUX FUNCTION HAVING TWO INFLECTION POINTS Electronic Journal of Differential Equations, Vol. 2006(2006), No. 149, pp. 1 18. ISSN: 1072-6691. URL: http://ejde.math.txstate.edu or http://ejde.math.unt.edu ftp ejde.math.txstate.edu (login: ftp) NONCLASSICAL

More information

E = where γ > 1 is a constant spesific to the gas. For air, γ 1.4. Solving for p, we get. 2 ρv2 + (γ 1)E t

E = where γ > 1 is a constant spesific to the gas. For air, γ 1.4. Solving for p, we get. 2 ρv2 + (γ 1)E t . The Euler equations The Euler equations are often used as a simplification of the Navier-Stokes equations as a model of the flow of a gas. In one space dimension these represent the conservation of mass,

More information

A General Technique for Eliminating Spurious Oscillations in Conservative Schemes for Multiphase and Multispecies Euler Equations

A General Technique for Eliminating Spurious Oscillations in Conservative Schemes for Multiphase and Multispecies Euler Equations A General Technique for Eliminating Spurious Oscillations in Conservative Schemes for Multiphase and Multispecies Euler Equations Ronald P. Fedkiw Xu-Dong Liu Stanley Osher September, 2 Abstract Standard

More information

Chapter 3 Burgers Equation

Chapter 3 Burgers Equation Chapter 3 Burgers Equation One of the major challenges in the field of comple systems is a thorough understanding of the phenomenon of turbulence. Direct numerical simulations (DNS) have substantially

More information

Non-linear Scalar Equations

Non-linear Scalar Equations Non-linear Scalar Equations Professor Dr. E F Toro Laboratory of Applied Mathematics University of Trento, Italy eleuterio.toro@unitn.it http://www.ing.unitn.it/toro August 24, 2014 1 / 44 Overview Here

More information

H. L. Atkins* NASA Langley Research Center. Hampton, VA either limiters or added dissipation when applied to

H. L. Atkins* NASA Langley Research Center. Hampton, VA either limiters or added dissipation when applied to Local Analysis of Shock Capturing Using Discontinuous Galerkin Methodology H. L. Atkins* NASA Langley Research Center Hampton, A 68- Abstract The compact form of the discontinuous Galerkin method allows

More information

Entropy stable schemes for compressible flows on unstructured meshes

Entropy stable schemes for compressible flows on unstructured meshes Entropy stable schemes for compressible flows on unstructured meshes Tata Institute of Fundamental Research Center for Applicable Mathematics Bangalore deep@math.tifrbng.res.in http://math.tifrbng.res.in/

More information

AA214B: NUMERICAL METHODS FOR COMPRESSIBLE FLOWS

AA214B: NUMERICAL METHODS FOR COMPRESSIBLE FLOWS AA214B: NUMERICAL METHODS FOR COMPRESSIBLE FLOWS 1 / 29 AA214B: NUMERICAL METHODS FOR COMPRESSIBLE FLOWS Hierarchy of Mathematical Models 1 / 29 AA214B: NUMERICAL METHODS FOR COMPRESSIBLE FLOWS 2 / 29

More information

Chapter 1. Introduction

Chapter 1. Introduction Chapter 1 Introduction Many astrophysical scenarios are modeled using the field equations of fluid dynamics. Fluids are generally challenging systems to describe analytically, as they form a nonlinear

More information

Applying Asymptotic Approximations to the Full Two-Fluid Plasma System to Study Reduced Fluid Models

Applying Asymptotic Approximations to the Full Two-Fluid Plasma System to Study Reduced Fluid Models 0-0 Applying Asymptotic Approximations to the Full Two-Fluid Plasma System to Study Reduced Fluid Models B. Srinivasan, U. Shumlak Aerospace and Energetics Research Program, University of Washington, Seattle,

More information

Modelling and numerical methods for the diffusion of impurities in a gas

Modelling and numerical methods for the diffusion of impurities in a gas INERNAIONAL JOURNAL FOR NUMERICAL MEHODS IN FLUIDS Int. J. Numer. Meth. Fluids 6; : 6 [Version: /9/8 v.] Modelling and numerical methods for the diffusion of impurities in a gas E. Ferrari, L. Pareschi

More information

Dedicated to the 70th birthday of Professor Lin Qun

Dedicated to the 70th birthday of Professor Lin Qun Journal of Computational Mathematics, Vol.4, No.3, 6, 39 5. ANTI-DIFFUSIVE FINITE DIFFERENCE WENO METHODS FOR SHALLOW WATER WITH TRANSPORT OF POLLUTANT ) Zhengfu Xu (Department of Mathematics, Pennsylvania

More information

YINGJIE LIU, CHI-WANG SHU, EITAN TADMOR, AND MENGPING ZHANG

YINGJIE LIU, CHI-WANG SHU, EITAN TADMOR, AND MENGPING ZHANG CENTRAL DISCONTINUOUS GALERKIN METHODS ON OVERLAPPING CELLS WITH A NON-OSCILLATORY HIERARCHICAL RECONSTRUCTION YINGJIE LIU, CHI-WANG SHU, EITAN TADMOR, AND MENGPING ZHANG Abstract. The central scheme of

More information

Numerische Mathematik

Numerische Mathematik Numer. Math. ) :545 563 DOI.7/s--443-7 Numerische Mathematik A minimum entropy principle of high order schemes for gas dynamics equations iangxiong Zhang Chi-Wang Shu Received: 7 July / Revised: 5 September

More information

Entropy and Relative Entropy

Entropy and Relative Entropy Entropy and Relative Entropy Joshua Ballew University of Maryland October 24, 2012 Outline Hyperbolic PDEs Entropy/Entropy Flux Pairs Relative Entropy Weak-Strong Uniqueness Weak-Strong Uniqueness for

More information

Finite volume approximation of the relativistic Burgers equation on a Schwarzschild (anti-)de Sitter spacetime

Finite volume approximation of the relativistic Burgers equation on a Schwarzschild (anti-)de Sitter spacetime Turkish Journal of Mathematics http:// journals. tubitak. gov. tr/ math/ Research Article Turk J Math 2017 41: 1027 1041 c TÜBİTAK doi:10.906/mat-1602-8 Finite volume approximation of the relativistic

More information

An Improved Non-linear Weights for Seventh-Order WENO Scheme

An Improved Non-linear Weights for Seventh-Order WENO Scheme An Improved Non-linear Weights for Seventh-Order WENO Scheme arxiv:6.06755v [math.na] Nov 06 Samala Rathan, G Naga Raju Department of Mathematics, Visvesvaraya National Institute of Technology, Nagpur,

More information

Hierarchical Reconstruction with up to Second Degree Remainder for Solving Nonlinear Conservation Laws

Hierarchical Reconstruction with up to Second Degree Remainder for Solving Nonlinear Conservation Laws Hierarchical Reconstruction with up to Second Degree Remainder for Solving Nonlinear Conservation Laws Dedicated to Todd F. Dupont on the occasion of his 65th birthday Yingjie Liu, Chi-Wang Shu and Zhiliang

More information

High Order Accurate Runge Kutta Nodal Discontinuous Galerkin Method for Numerical Solution of Linear Convection Equation

High Order Accurate Runge Kutta Nodal Discontinuous Galerkin Method for Numerical Solution of Linear Convection Equation High Order Accurate Runge Kutta Nodal Discontinuous Galerkin Method for Numerical Solution of Linear Convection Equation Faheem Ahmed, Fareed Ahmed, Yongheng Guo, Yong Yang Abstract This paper deals with

More information

Multi-Domain Hybrid Spectral-WENO Methods for Hyperbolic Conservation Laws

Multi-Domain Hybrid Spectral-WENO Methods for Hyperbolic Conservation Laws Multi-Domain Hybrid Spectral-WENO Methods for Hyperbolic Conservation Laws Bruno Costa a, Wai Sun Don b, a Departamento de Matemática Aplicada, IM-UFRJ, Caia Postal 6853, Rio de Janeiro, RJ, C.E.P. 945-97,

More information

On the Dependence of Euler Equations on Physical Parameters

On the Dependence of Euler Equations on Physical Parameters On the Dependence of Euler Equations on Physical Parameters Cleopatra Christoforou Department of Mathematics, University of Houston Joint Work with: Gui-Qiang Chen, Northwestern University Yongqian Zhang,

More information

Weighted Essentially Non-Oscillatory limiters for Runge-Kutta Discontinuous Galerkin Methods

Weighted Essentially Non-Oscillatory limiters for Runge-Kutta Discontinuous Galerkin Methods Weighted Essentially Non-Oscillatory limiters for Runge-Kutta Discontinuous Galerkin Methods Jianxian Qiu School of Mathematical Science Xiamen University jxqiu@xmu.edu.cn http://ccam.xmu.edu.cn/teacher/jxqiu

More information

Notes: Outline. Shock formation. Notes: Notes: Shocks in traffic flow

Notes: Outline. Shock formation. Notes: Notes: Shocks in traffic flow Outline Scalar nonlinear conservation laws Traffic flow Shocks and rarefaction waves Burgers equation Rankine-Hugoniot conditions Importance of conservation form Weak solutions Reading: Chapter, 2 R.J.

More information

Sung-Ik Sohn and Jun Yong Shin

Sung-Ik Sohn and Jun Yong Shin Commun. Korean Math. Soc. 17 (2002), No. 1, pp. 103 120 A SECOND ORDER UPWIND METHOD FOR LINEAR HYPERBOLIC SYSTEMS Sung-Ik Sohn and Jun Yong Shin Abstract. A second order upwind method for linear hyperbolic

More information

AN EXACT RESCALING VELOCITY METHOD FOR SOME KINETIC FLOCKING MODELS

AN EXACT RESCALING VELOCITY METHOD FOR SOME KINETIC FLOCKING MODELS AN EXACT RESCALING VELOCITY METHOD FOR SOME KINETIC FLOCKING MODELS THOMAS REY AND CHANGHUI TAN Abstract. In this work, we discuss kinetic descriptions of flocking models, of the so-called Cucker- Smale

More information

On discontinuity capturing methods for convection diffusion equations

On discontinuity capturing methods for convection diffusion equations On discontinuity capturing methods for convection diffusion equations Volker John 1 and Petr Knobloch 2 1 Universität des Saarlandes, Fachbereich 6.1 Mathematik, Postfach 15 11 50, 66041 Saarbrücken, Germany,

More information

3. FORMS OF GOVERNING EQUATIONS IN CFD

3. FORMS OF GOVERNING EQUATIONS IN CFD 3. FORMS OF GOVERNING EQUATIONS IN CFD 3.1. Governing and model equations in CFD Fluid flows are governed by the Navier-Stokes equations (N-S), which simpler, inviscid, form is the Euler equations. For

More information

Applications of the compensated compactness method on hyperbolic conservation systems

Applications of the compensated compactness method on hyperbolic conservation systems Applications of the compensated compactness method on hyperbolic conservation systems Yunguang Lu Department of Mathematics National University of Colombia e-mail:ylu@unal.edu.co ALAMMI 2009 In this talk,

More information

ALGEBRAIC FLUX CORRECTION FOR FINITE ELEMENT DISCRETIZATIONS OF COUPLED SYSTEMS

ALGEBRAIC FLUX CORRECTION FOR FINITE ELEMENT DISCRETIZATIONS OF COUPLED SYSTEMS Int. Conf. on Computational Methods for Coupled Problems in Science and Engineering COUPLED PROBLEMS 2007 M. Papadrakakis, E. Oñate and B. Schrefler (Eds) c CIMNE, Barcelona, 2007 ALGEBRAIC FLUX CORRECTION

More information

Received 6 August 2005; Accepted (in revised version) 22 September 2005

Received 6 August 2005; Accepted (in revised version) 22 September 2005 COMMUNICATIONS IN COMPUTATIONAL PHYSICS Vol., No., pp. -34 Commun. Comput. Phys. February 6 A New Approach of High OrderWell-Balanced Finite Volume WENO Schemes and Discontinuous Galerkin Methods for a

More information

n v molecules will pass per unit time through the area from left to

n v molecules will pass per unit time through the area from left to 3 iscosity and Heat Conduction in Gas Dynamics Equations of One-Dimensional Gas Flow The dissipative processes - viscosity (internal friction) and heat conduction - are connected with existence of molecular

More information

Strong Stability Preserving Properties of Runge Kutta Time Discretization Methods for Linear Constant Coefficient Operators

Strong Stability Preserving Properties of Runge Kutta Time Discretization Methods for Linear Constant Coefficient Operators Journal of Scientific Computing, Vol. 8, No., February 3 ( 3) Strong Stability Preserving Properties of Runge Kutta Time Discretization Methods for Linear Constant Coefficient Operators Sigal Gottlieb

More information

THE ENTROlm~ DISSIPATION BY NUMERICAL VISCOSITY IN NONLINeAR CONSERVATIVE DIFFERENCE SC~I~ES Eitan Tadmor* School of Mathematical Sciences, Tel-Aviv University and Institute for Computer Applications in

More information

A Fifth Order Flux Implicit WENO Method

A Fifth Order Flux Implicit WENO Method A Fifth Order Flux Implicit WENO Method Sigal Gottlieb and Julia S. Mullen and Steven J. Ruuth April 3, 25 Keywords: implicit, weighted essentially non-oscillatory, time-discretizations. Abstract The weighted

More information

ACCURATE NUMERICAL SCHEMES FOR APPROXIMATING INITIAL-BOUNDARY VALUE PROBLEMS FOR SYSTEMS OF CONSERVATION LAWS.

ACCURATE NUMERICAL SCHEMES FOR APPROXIMATING INITIAL-BOUNDARY VALUE PROBLEMS FOR SYSTEMS OF CONSERVATION LAWS. ACCURATE NUMERICAL SCHEMES FOR APPROXIMATING INITIAL-BOUNDARY VALUE PROBLEMS FOR SYSTEMS OF CONSERVATION LAWS. SIDDHARTHA MISHRA AND LAURA V. SPINOLO Abstract. Solutions of initial-boundary value problems

More information

ARTICLE IN PRESS Mathematical and Computer Modelling ( )

ARTICLE IN PRESS Mathematical and Computer Modelling ( ) Mathematical and Computer Modelling Contents lists available at ScienceDirect Mathematical and Computer Modelling ournal homepage: wwwelseviercom/locate/mcm Total variation diminishing nonstandard finite

More information

PDE Solvers for Fluid Flow

PDE Solvers for Fluid Flow PDE Solvers for Fluid Flow issues and algorithms for the Streaming Supercomputer Eran Guendelman February 5, 2002 Topics Equations for incompressible fluid flow 3 model PDEs: Hyperbolic, Elliptic, Parabolic

More information

APPROXIMATE PERIODIC SOLUTIONS for the RAPIDLY ROTATING SHALLOW-WATER and RELATED EQUATIONS

APPROXIMATE PERIODIC SOLUTIONS for the RAPIDLY ROTATING SHALLOW-WATER and RELATED EQUATIONS 1 APPROXIMATE PERIODIC SOLUTIONS for the RAPIDLY ROTATING SHALLOW-WATER and RELATED EQUATIONS BIN CHENG Department of Mathematics University of Michigan Ann Arbor, MI 48109 E-mail:bincheng@umich.edu EITAN

More information

Improvement of convergence to steady state solutions of Euler equations with. the WENO schemes. Abstract

Improvement of convergence to steady state solutions of Euler equations with. the WENO schemes. Abstract Improvement of convergence to steady state solutions of Euler equations with the WENO schemes Shuhai Zhang, Shufen Jiang and Chi-Wang Shu 3 Abstract The convergence to steady state solutions of the Euler

More information

POSITIVITY PROPERTY OF SECOND-ORDER FLUX-SPLITTING SCHEMES FOR THE COMPRESSIBLE EULER EQUATIONS. Cheng Wang. Jian-Guo Liu

POSITIVITY PROPERTY OF SECOND-ORDER FLUX-SPLITTING SCHEMES FOR THE COMPRESSIBLE EULER EQUATIONS. Cheng Wang. Jian-Guo Liu DISCRETE AND CONTINUOUS Website: http://aimsciences.org DYNAMICAL SYSTEMS SERIES B Volume 3 Number May003 pp. 0 8 POSITIVITY PROPERTY OF SECOND-ORDER FLUX-SPLITTING SCHEMES FOR THE COMPRESSIBLE EULER EQUATIONS

More information