Processing math: 100%

2024年1月23日 星期二

Solution to Partial Differential Equations: Methods and Applications (Robert McOwen) Section 2.3

Solution to Partial Differential Equations: Methods and Applications (Robert McOwen) Section 2.3

    1. For complex-valued functions u and v on Ω, let u,v=Ωuˉvdx, where ˉv denotes the complex conjugate of v. If the coefficients aα(x) in (32a) are complex-valued functions, define the adjoint L so that Lu,v=u,Lv for all uCm(Ω), vCm0(Ω).
    2. For (complex) vector-valued functions u(x)=(u1(x),,uN(x)) on Ω, let u,v=Ω(u1ˉv1++uNˉvN)dx. If the coefficients aα(x) in (32a) are N×N matrix-valued functions, then (32a) defines a system of mth-order operators. Define the adjoint L so that Lu,v=u,Lv for all uCm(Ω,RN), vCm0(Ω,RN).
  1. Solution
    1. From (32a), we find that

      Lu,v=Ω(Lu)(ˉv)dx=Ω(|α|maα(x)Dαu)ˉvdx=|α|mΩ(Dαu)(aα(x)ˉv)dx=|α|m(1)|α|Ω(u)(Dα(aα(x)ˉv))dx=Ω(u)[|α|m(1)|α|Dα(aα(x)ˉv)]dx:=u,Lv,

      where the adjoint operator L is given by

      Lv:=¯|α|m(1)|α|Dα(aα(x)ˉv)=|α|m(1)|α|Dα(¯aα(x)v)for vCm0(Ω).

    2. Suppose aα(x)=[aijα(x)]1i,jN is a N×N matrix-valued function. Then we have

      Lu,v=Ω[(Nj=1|α|ma1jα(x)Dαuj)ˉv1++(Nj=1|α|maNjα(x)Dαuj)ˉvN]dx=Ω(Nj=1|α|ma1jα(x)Dαuj)ˉv1dx++Ω(Nj=1|α|maNjα(x)Dαuj)ˉvNdx=Nj=1|α|m[Ω(Dαuj)(a1jα(x)ˉv1)dx++Ω(Dαuj)(aNjα(x)ˉvN)dx]=Nj=1|α|m(1)|α|[Ω(uj)Dα(a1jα(x)ˉv1)dx++Ω(uj)Dα(aNjα(x)ˉvN)dx]=Nj=1Ω(uj)|α|m(1)|α|Dα(a1jα(x)ˉv1++aNjα(x)ˉvN)dx=u,Lv,

      where the adjoint operator L is given by

      Lv=¯|α|m(1)|α|Dα(aTα(x)ˉv)=|α|m(1)|α|Dα(¯aTα(x)v)for vCm0(Ω,RN).

      Here AT denotes the tranpose of the matrix A.

  2. Let Lu=uμη as in Example 1. Derive the transmission condition (44) along the characteristic μ=0.
  3. SolutionSuppose u is a weak solution of Lu=0, i.e., Ωuμηvdμdη=Ω(Lu)vdμdη=0 for vC20(Ω). Recall that the operator L is self-adjoint, i.e., L=L. From (40), we have

    0=Ωu(Lv)dμdη=Ω+u+(Lv)dμdη+Ωu(Lv)dμdη=Ω+u+vμηdμdη+Ωuvμηdμdη=0u+vμηdηdμ+0uvμηdηdμ=0u+ηvμdηdμ0uηvμdηdμ=0u+ηvμdμdη0uηvμdμdη=[u+ηv|μ=μ=00u+μηvdμ]dη[uηv|μ=0μ=0uμηvdμ]dη=u+η(0,η)v(0,η)dηuη(0,η)v(0,η)dη=[u+η(0,η)uη(0,η)]v(0,η)dη.

    Thus, the transmission condition is

    [u+η(0,η)uη(0,η)]v(0,η)dη=0for vC20(Ω).


  4. Consider the first-order equation ut+cux=0.
    1. If fC(R), show that u(x,t)=f(xct) is a weak solution.
    2. Can you find any discontinuous weak solutions?
    3. Is there a transmission condition for a weak solution with jump discontinuity along the characteristic x=ct?
  5. Solution
    1. To show u(x,t)=f(xct) is a weak solution, it suffices to verify that R2f(xct)(Lv)(x,t)dxdt=0 for vC10(R2), where L=tcx is the adjoint operator L=t+cx. Let s=xct, t=t and ˜v(s,t)=v(s+ct,t) for s,tR2. Clearly, ˜vC10(R2) and satisfies ˜vt(s,t)=vt(s+ct,t)+cvx(s+ct,t). Thus, the improper double integral becomes

      R2f(xct)(Lv)(x,t)dxdt=f(xct)(vt(x,t)+cvx(x,t))dxdt=f(s)˜vt(s,t)dsdt=f(s)˜vt(s,t)dtds=f(s)0ds=0for vC10(R2).

      Here the Jacobian is 1. Therefore, u(x,t)=f(xct) is a weak solution for ut+cux=0.
    2. When f is discontinuous but Lebesgue integrable, u(x,t)=f(xct) is still a weak solution because the computation in part (a) also holds true.
    3. Suppose u is a weak solution with jump discontinuity along the characteristic x=ct. Let ˜u(s,t)=u(s+ct,t). Then ˜u is a weak solution of ˜L˜u:=˜ut(s,t)=0 with jump discontinuity along the line s=0, where ˜L=t. Clearly, the adjoint operator ˜L of ˜L is t. Let R2±={(s,t)R2:±s>0,tR} and ˜u±=˜u|R2±. Then for any test function vC10(R2), we have

      0=R2˜u(s,t)˜Lv(s,t)dsdt=R2˜u(s,t)(tv(s,t))dsdt=0˜u+(s,t)vt(s,t)dtds0˜u(s,t)vt(s,t)dtds=0[˜u+(s,t)v(s,t)|t=t=+˜u+t(s,t)v(s,t)dt]ds+0[˜u(s,t)v(s,t)|t=t=+˜ut(s,t)v(s,t)dt]ds=0˜u+t(s,t)v(s,t)dtds+0˜ut(s,t)v(s,t)dtds=0.

      Thus, there is no transmission condition for a weak solution with jump discontinuity along the characteristic x=ct.

  6. If fL1loc(Ω), define Ff,vΩf(x)v(x)dx. Show that Ff is a distribution in Ω.
  7. SolutionLet a,bR and v,wD(Ω). It is easy to verify that

    Ff(av+bw):=Ff,av+bw=Ωf(x)[av(x)+bw(x)]dx=aΩf(x)v(x)dx+bΩf(x)w(x)dx=aFf,v+bFf,w=aFf(v)+bFf(w),

    which shows Ff is a linear mapping on D(Ω). Moreover, let {vj}j=1 be a sequence of functions in D(Ω) with supp(vj)K for some compact set KΩ (independent of j) and Dαvj0 uniformly in K. Then for any given ε>0, there exists a NN such that |vj(x)|<ε(K|f(x)|dx)1 for j>N and xK. For j>N, we have

    |Ff(vj)0|=|Ff,vj|=|Ωf(x)vj(x)dx|=|Kf(x)vj(x)dx|K|f(x)||vj(x)|dxK|f(x)|ε(K|f(x)|dx)1dx=ε.

    This implies limjFf(vj)=0. Therefore, Ff is continuous on D(Ω) and Ff is a distribution in Ω.

    1. If ξ1,,ξk are points in Ω and a1,,ak are real numbers, show that Fa1δξ1++akδξk is a distribution in Ω.
    2. For Ω=Rn, find an infinite sequence {ξk}k=1 for which fa1δξ1+ is a distribution. (This gives an example of a distribution that cannot be realized as a distributional derivative of an integrable function.)
  8. Solution
    1. Let a,bR and v,wD(Ω). It is easy to verify that

      F(av+bw)=F,av+bw=ki=1aiδξi,av+bw=Ωki=1aiδξi(x)[av(x)+bw(x)]dx=ki=1aiΩδξi(x)[av(x)+bw(x)]dx=ki=1ai[av(ξi)+bw(ξi)]=aki=1aiv(ξi)+bki=1aiw(ξi)=aki=1aiΩδξi(x)v(x)dx+bki=1aiΩδξi(x)w(x)dx=aΩki=1aiδξi(x)[v(x)]dx+bΩki=1aiδξi(x)[w(x)]dx=aki=1aiδξi,v+baki=1aiδξi,w=aF,v+bF,w=aF(v)+bF(w),

      which shows F is a linear mapping on D(Ω). Moreover, let {vj}j=1 be a sequence of functions in D(Ω) with supp(vj)K for some compact set KΩ (independent of j) and Dαvj0 uniformly in K. Since K{ξ1,,ξk} is finite, we may write it as {ξj1,,ξjr} for rN. Set A=max1ir|aji|. Then for any ε>0, there exists NN such that |vj(x)|ε/(rA) for j>N and xK. For j>N, we have

      |F(vj)0|=|F,vj|=|ΩF(x)vj(x)dx|=|Ωki=1aiδξi(x)vj(x)dx|ri=1|aji||Kδξji(x)vj(x)dx|=ri=1|aji||vj(ξji)|ri=1AεrA=ε.

      This implies limjF(vj)=0. Therefore, F is continuous on D(Ω) and F is a distribution in Ω.
    2. Choose ξj=je1:=(j,0,,0) for jN, and then we prove F=i=1aiδξi is a distribution in Rn. Let a,bR and v,wD(Rn). Then there exists a compact subset KRn such that supp(v)K and supp(w)K. Since K is bounded, K{ξj}j=1 is finite and K{ξj}j=1={ξj1,,ξjr} for rN. Then one may check that

      F(av+bw)=F,av+bw=j=1ajδξj,av+bw=Rnj=1ajδξj(x)[av(x)+bw(x)]dx=Kri=1ajiδξji(x)[av(x)+bw(x)]dx=ri=1ajiKδξji(x)[av(x)+bw(x)]dx=ri=1aji[av(ξji)+bw(ξji)]=ari=1ajiv(ξji)+bri=1ajiw(ξji)=ari=1ajiKδξji(x)v(x)dx+bri=1ajiKδξji(x)w(x)dx=aKri=1ajiδξji(x)v(x)dx+bKri=1ajiδξji(x)w(x)dx=aRnj=1ajδξj(x)v(x)dx+bRnj=1ajδξj(x)w(x)dx=aj=1ajδξj,v+bj=1ajδξj,w=aF,v+bF,w=aF(v)+bF(w),

      which shows F is a linear mapping on D(Rn). Moreover, let {vk}k=1 be a sequence of functions in D(Rn) with supp(vk)K for some compact set KRn (independent of j) and Dαvk0 uniformly in K. Since K is bounded, K{ξj}j=1 is finite and K{ξj}j=1={ξji}ri=1 for rN. Set A=max1ir|aji|. Then for any ε>0, there exists NN such that |vk(x)|<ε/(rA) for k>N and xK. For j>N, we have

      |F(vk)0|=|F,vk|=|RnF(x)vk(x)dx|=|Rnj=1ajδξj(x)vk(x)dx|=|Kri=1ajiδξji(x)vk(x)dx|=|ri=1ajivk(ξji)|ri=1|aji||vk(ξji)|ri=1AεrA=ε.

      This implies limkF(vk)=0. Therefore, F is continuous on D(Rn) and F is a distribution in Rn.

    1. Prove the Lemma on p. 67 when fC(Rn) has a compact support.
    2. Find an example to show that the result fails to hold if gL1loc(Rn) is replaced by gD(Rn).
  9. Solution
    1. Suppose that K=supp(f) is a compact subset of Rn. Let xRn and yˉB1(x) arbitrarily and define h(z)=f(xz)f(yz) for zRn. Clearly, h is continuous function compact support, and

      supp(h)(xK)(ˉB1(x)K):=˜K,

      where

      xK={xzRn:zK} and ˉB1(x)K={yzRn:yˉB1(x),zK}.

      On the other hand, f is uniformly continuous in Rn. Then for any ε>0, there exists δ=δ(ε)>0 such that |f(xz)f(yz)|<ε(˜K|g(z)|dz)1 whenever |xy|<δ and zRn. Thus, we find that

      |(fg)(x)(fg)(y)|=|Rn[f(xz)f(yz)]g(z)dz|=|˜K[f(xz)f(yz)]g(z)dz|˜K|f(xz)f(yz)||g(z)|dz<ε.

      This shows limyx(fg)(y)=(fg)(x), which means fgC(Rn).
    2. For n=1, let g=δD(R) and f(x)=|x| for xR. Then one may check that

      (fg)(x)=Rf(xy)g(y)dy=R|xy|δ(y)dy=Rddy|xy|δ(y)dy=Rsgn(xy)δ(y)dy=sgn(x),

      which is not continuous on R.

  10. Let Γ be a hypersurface (such as a sphere) in Rn, let a(z) be a continuous function of zΓ, and let dz denote the surface measure on Γ.
    1. Show that F,v=Γa(z)v(z)dz is a distribution in Rn. (It is natural to denote this distribution by F=aδΓ.)
    2. Suppose we can choose a unit normal ν along Γ. Formulate a definition for the "conormal distribution" aνδΓ in Rn.
  11. Solution
    1. Note that Γ is a hypersurface in Rn so Γ is a closed set in Rn. For c1,c2R and v1,v2D(Rn), we have

      F,c1v1+c2v2=Γa(z)[c1v1(z)+c2v2(z)]dz=c1Γa(z)v1(z)dz+c2Γa(z)v2(z)dz=c1F,v1+c2F,v2,

      which shows F is a linear mapping on D(Rn). Here we have used the fact that v1,v2C(Γ)D(Rn). Moreoever, let {vj}j=1 be a sequence of functions in D(Rn) with supp(vj)K for some compact set KRn (independent of j) and Dαvj0 uniformly in K. Since a is continuous on the compact set ΓK, the maximum of |a| over ΓK is finite and can be denoted by A:=maxzΓKa(z). For any ε>0, there exists NN such that

      |vj(x)|ε[AΓKdz]1 for j>N and xK.

      For j>N, we have

      |F,vj0|=|Γa(z)vj(z)dz|=|ΓKa(z)vj(z)dz|ΓK|a(z)||vj(z)|dzΓKAε[AΓKdz]1dz=ε.

      This implies limjF(vj)=0. Therefore, F is continuous on D(Rn) and F is a distribution in Rn.
    2. Suppose aC1(Γ). Then we define the conformal distribution aνδΓ by

      aνδΓ,v=Γ(a(z)v(z))ν(z)dzfor vD(Rn).

      For c1,c2R and v1,v2D(Rn), we have

      aνδΓ,c1v1+c2v2=Γ(a(z)[c1v1(z)+c2v2(z)])ν(z)dz=c1Γ(a(z)v1(z))ν(z)dzc2Γ(a(z)v2(z))ν(z)dz=c1aνδΓ,v1+c2aνδΓ,v2,

      which shows aνδΓ, is a linear mapping on D(Rn). Here we have used the fact that v1,v2C(Γ)D(Rn). Moreover, let {vj}j=1 be a sequence of functions in D(Rn) with supp(vj)K for some compact set KRn (independent of j) and Dαvj0 uniformly in K. Since a is continuously differentiable on the compact set ΓK, aC1(ΓK):=maxzΓK|a(z)|+maxzΓK|a(z)| is finite. For any ε>0, there exists NN such that

      vj(x)ε2[aC1(ΓK)ΓKdz]1,|vj(x)|ε2[aC1(ΓK)ΓKdz]1for j>N and xK.

      For j>N, we have

      |aνδΓ,vj|=|Γ(a(z)vj(z))ν(z)dz|=|Γ[vj(z)a(z)+a(z)vj(z)]ν(z)dz|=|ΓK[vj(z)a(z)+a(z)vj(z)]ν(z)dz|ΓK(|vj(z)||a(z)|+|a(z)||vj(z)|)dzΓKaC1(ΓK)(|vj(z)|+|vj(z)|)dzaC1(ΓK)ΓKε[aC1(ΓK)ΓKdz]1dz=ε.

      This implies limjaνδΓ,vj=0. Therefore, aνδΓ, is continuous on D(Rn) and F is a distribution in Rn.

  12. Let

    fn(x)={n2for 1n<x<1n;0for |x|1n.

    Show that fn(x)δ(x) as distributions on R.
  13. SolutionFor vC0(R), we find that

    Rfn(x)v(x)dx=n21/n1/nv(x)dx.

    Since v is continuous on R, for any ε>0, there exists δ>0 such that |v(x)v(0)|<ε whenever |x|<δ. Then for n>1/δ, we have

    |Rfn(x)v(x)dxv(0)|=n2|1/n1/n[v(x)v(0)]dx|n21/n1/n|v(x)v(0)|dx<n21/n1/nεdx=ε.

    This shows limnRfn(x)v(x)dx=v(0)=Rδ(x)v(x)dx, i.e., fn(x)δ(x) as distributions on R.

  14. If fn(x) and f(x) are integrable functions such that for any compact set KΩ we have K|fn(x)f(x)|dx0 as n, then fnf as distributions.
  15. SolutionLet vC0(R) and K=supp(v). Then we observe that

    |fn,vf,v|=|fnf,v|=|Ω(fn(x)f(x))v(x)dx|=|K(fn(x)f(x))v(x)dx|K|fn(x)f(x)||v(x)|dxvC0(K)K|fn(x)f(x)|dx0as n.

    This means limnfn,v=f,v for vD(Ω), i.e., fnf as distributions on Ω.

  16. Let aR, a0.
    1. Find a fundamental solution for L=d/dxa on R (i.e., solve dF/dxaF=δ).
    2. Show that a fundamental solution for L=d2/dx2a2=(d/dx+a)(d/dxa) on R is given by

      F(x)={a1sinhaxif x>0;0if x<0.

  17. Solution
    1. Multiplying equation by eax, it gives

      ddx(eaxF)=eaxdFdxaeaxF=eaxδ(x)=δ(x),

      which implies

      eaxF(x)=H(x)={1if x>0;0if x<0.

      Hence F(x)=eaxH(x)={eaxif x>0;0if x<0.

      Note that the adjoint operator L of L is d/dxa. It is easy to verify that

      LF,v=F,Lv=RF(x)(dvdxav)dx=0eaxdvdxdxa0eaxv(x)dx=eaxv(x)|0+a0eaxv(x)dxa0eaxv(x)dx=v(0)=δ,v,

      which shows LF=δ.
    2. Note that L is self-adjoint, i.e., L=L. Then for vD(R), it is easy to verify that

      LF,v=F,Lv=RF(x)(d2vdx2a2v(x))dx=0a1sinh(ax)(d2vdx2a2v(x))dx=a1sinh(ax)dvdx|00cosh(ax)dvdxdx0asinh(ax)v(x)dx=cosh(ax)v(x)|0+0asinh(ax)v(x)dx0asinh(ax)v(x)dx=v(0)=δ,v,

      which implies LF=δ.

  18. Regularity of u defined by (60):
    1. If f(x)L1(R) has a compact support, show that (60) defines a continuous weak solution of u=f(x).
    2. If, in addition to (a), f(x) is a bounded function on R, show uC1(R), and u(x)=12(xf(y)dyxf(y)dy).
    3. If, in addition to (a), f(x) is continuous, show that u is in fact a classical solution: uC2(R) with u=f.
  19. Solution
    1. Since fL1(R) has a compact support, fL1loc(R). On the other hand, g(x)=12|x|C(R). Thus, by Lemma at Page 67, u(x)=(gf)(x)=12R|xy|f(y)dy is continuous on R.
    2. From (60), we have

      u(x)=12x(xy)f(y)dy+12x(yx)f(y)dy.

      A direct computation gives

      u(x+h)u(x)=12x+h(x+hy)f(y)dy12x(xy)f(y)dy+12x+h(yxh)f(y)dy12x(yx)f(y)dy=h2x+hf(y)dy+12x+hx(xy)f(y)dyh2x+hf(y)dy12x+hx(yx)f(y)dy.

      Since f is bounded, there exists M>0 such that |f(y)|M for yR. Then it is easy to observe that

      |x+hx±(xy)f(y)dy||x+hx(yx)Mdy|=Mh22,

      which implies limh01hx+hx±(xy)f(y)dy=0 by the squeeze theorem. Therefore, by the definition, we find

      u(x)=limh0u(x+h)u(x)h=12x+hf(y)dy12x+hf(y)dy+12hx+hx(xy)f(y)dy12hx+hx(yx)f(y)dy=12xf(y)dy12xf(y)dy.

      Here we have used the fact that functions xf(y)dy and xf(y)dy are continuous in x because fL1(R) is bounded with compact support. This also shows uC(R), and hence uC1(R).
    3. When continuous function fL1(R) has a compact support, f is bounded on R. Thus, by (b), uC1(R). Moreover, by the fundamental theorem of calculus, functions xf(y)dy and xf(y)dy are differentiable in x, and then u is also differentiable. By the fundamental theorem of calculus, we arrive at u(x)=12f(x)12(f(x))=f(x) for xR. Since fC(R), uC(R) and then uC2(R).

    1. Use the fundamental solution (59) to solve the initial u=f(x) for x>0 with u(0)=u0 and u(0)=u0, where fC([0,)) and f=O(|x|2ϵ) as |x|.
    2. Do the same for the boundary value problem u=f(x) for 0<x< with u(0)=0=u().
  20. Solution
    1. Extend ˜f by ˜f(x)=f(x) if x>0; 0 if x0. Then we consider the equation ˜u(x)=˜f(x) for xR. Plugging (59) into (56), we obtain the particular solution

      up(x)=(Ff)(x)=RF(xy)˜f(y)dy=12R|xy|˜f(y)dy=120|xy|f(y)dyfor xR.

      To fulfill the initial condition, we need to find a function uh satisfying

      uh(x)=0,uh(0)=u0up(0)=u0120yf(y)dy,uh(0)=u0up(0)=u0+120f(y)dy.

      Here we have used the result of Exercise 11(b). It is easy to obtain

      uh(x)=uh(0)+uh(0)x=u0120yf(y)dy+(u0+120f(y)dy)xfor xR.

      Thus, we find

      ˜u(x)=uh(x)+up(x)=u0120yf(y)dy+(u0+120f(y)dy)x+120|xy|f(y)dyfor xR.

      In particular, the solution is u(x)=˜u|x0(x) for x0.

      It is remark that the solution u can be represented as

      u(x)=u(0)+x0u(t)dt=u0+x0[u(0)+t0u(s)ds]dt=u0+u0x+x0t0f(s)dsdt=u0+u0x+x0xsf(s)dtds=u0+u0x+x0(xs)f(s)dsfor x0..

    2. Extend ˜f by ˜f(x)=f(x) if 0<x<; 0 if x0 or x. Then we consider the equation ˜u(x)=˜f(x) for xR. Plugging (59) into (56), we obtain the particular solution

      up(x)=(Ff)(x)=RF(xy)˜f(y)=12R|xy|˜f(y)dy=120|xy|f(y)dyfor xR.

      To fulfill the boundary condition, we need to find a function uh satisfying

      uh(x)=0,uh(0)=up(0)=120yf(y)dy,uh()=up()=120(y)f(y)dy.

      It is easy to obtain

      uh(x)=(120(y)f(y)dy)x(120yf(y)dy)(x)=x0(2y)f(y)dy120yf(y)dyfor xR.

      Thus, we find

      ˜u(x)=uh(x)+up(x)=x0(2y)f(y)dy120yf(y)dy+120|xy|f(y)dyfor xR.

      In particular, the solution is u(x)=˜u|0x(x) for 0x.

      Note that u()=0=u(0), which implies 0u(t)dt=0, and hence

      0=0[u(0)+t0u(s)ds]dt=u(0)+0t0f(s)dsdt.

      Thus, u(0)=10t0f(s)dsdt. Therefore, the solution can be represented as

      u(x)=u(0)+x0u(t)dt=x0[u(0)+t0u(s)ds]dt=(10t0f(s)dsdt)x+x0t0f(s)dsdt=x0(s)f(s)ds+x0(xs)f(s)dsfor 0x.


    1. Using δ(μ,η)=δ(μ)δ(η), show that each of the following functions is a fundamental solution of L=2/μη:

      F1(μ,η)=H(μ)H(η),F2(μ,η)=H(μ)H(η),F3(μ,η)=H(μ)H(η),F4(μ,η)=H(μ)H(η).

    2. Use part (a) with the change of variables μ=x+ct, η=xct to obtain four distinct fundamental solutions for the one-dimensional wave operator, each having support in one of the wedges determined by the lines x=±ct.
  21. Solution
    1. We firstly note that L is self-adjoint, i.e., L=L. For vD(R2), it is easy to verify that

      LF1,v=F1,Lv=R2H(μ)H(η)vμη(μ,η)dμdη=00vμη(μ,η)dμdη=0vη(μ,η)|μ=μ=0dη=0vη(0,η)dη=v(0,η)|η=0=v(0,0)=δ,v,LF2,v=F2,Lv=R2H(μ)H(η)vμη(μ,η)dμdη=00vμη(μ,η)dμdη=0vη(μ,η)|μ=μ=0dη=0vη(0,η)dη=v(0,η)|η=0η==v(0,0)=δ,v,LF3,v=F3,Lv=R2H(μ)H(η)vμη(μ,η)dμdη=00vμη(μ,η)dμdη=0vη(μ,η)|μ=0μ=dη=0vη(0,η)dη=v(0,η)|η=η=0=v(0,0)=δ,vLF4,v=F4,Lv=R2H(μ)H(η)vμη(μ,η)dμdη==00vμη(μ,η)dμdη=0vη(μ,η)|μ=0μ=dη=0vη(0,η)dη=v(0,η)|η=0η==v(0,0)=δ,v.

      This shows Fi is a fundamental solution of LFi=δ for i=1,2,3,4.
    2. By (a) with the change of variables μ=x+ct and η=xct, the fundamental solutions for 4c2˜uμη(μ,η)=utt(x,t)c2uxx(x,t)=δ are

      U1(x,t)=(2c)1H(x+ct)H(xct),U2(x,t)=(2c)1H(x+ct)H(x+ct),U3(x,t)=(2c)1H(xct)H(xct),U4(x,t)=(2c)1H(xct)H(x+ct).


  22. If K(x,t) is a distribution in Rn depending on the parameter t, then we say K depends continuously on t if, for every vC0(Rn), f(t)K(,t),v=K(x,t)v(x)dx defines a continuous function of t. Use this idea to define K is continuously differentiable in t.
  23. SolutionK is continuously differentiable in t if for every vC0(Rn), f(t)=RnK(x,t)v(x)dx defines a continuously differentiable function of t.

  24. According to Section 1.2, a weak solution of uy+(G(u))x=0 in a rectangle Ω=(a,b)×(c,d) satisfies

    ddyx2x1u(x,y)dx+G(u(x,y))|x=x2x=x1=0

    for every [x1,x2](a,b) and c<y<d. Show that this implies

    Ωuvy+G(u)vxdxdy=0for every vC10(Ω)

    (i.e., the notion of weak solution of Section 1.2 is consistent with that of Section 2.3).
  25. SolutionLet xi=a+(ba)i/m for i=0,1,,m and yj=c+(dc)j/n for j=1,,n. Set Ωij=[xi1,xi]×[yj1,yj] for 1in and 1jm. Then {Ωij}1in,1jm forms a partition of Ω. Put Δx=(ba)/m and Δy=(dc)/n. Then using the summation by parts, we have

    mi=1nj=1[u(xi,yj)v(xi,yj)v(xi,yj1)Δy+G(u(xi,yj))v(xi,yj)v(xi1,yj)Δx]ΔyΔx=mi=1[u(xi,yn)v(xi,yn)u(xi,y1)v(xi,y0)Δynj=2v(xi,yj1)u(xi,yj)u(xi,yj1)Δy]ΔyΔx+nj=1[G(u(xm,yj))v(xm,yj)G(u(x1,yj))v(x0,yj)Δxmi=2v(xi1,yj)G(u(xi,yj))G(u(xi1,yj))Δx]ΔyΔx=mi=1nj=2[v(xi,yj1)u(xi,yj)u(xi,yj1)Δy]ΔyΔxmi=2nj=1[v(xi1,yj)G(u(xi,yj))G(u(xi1,yj))Δx]ΔyΔx=m1i=1n1j=1{v(xi,yj)[u(xi,yj+1)u(xi,yj)Δy+G(u(xi+1,yj))G(u(xi,yj))Δx]}ΔyΔx

    Here we have used the fact that vC0(Ω) so v(x0,yj)=v(xm,vj)=v(xi,y0)=v(xi,yn)=0 for 1im and 1jn. Note that

    G(u(xi+1,yj))G(u(xi,yj))=xi+1xiuy(x,yj)dxfor 1im1, 1jn1.

    Then the double sum becomes

    m1i=1n1j=1{v(xi,yj)[u(xi,yj+1)u(xi,yj)Δy1Δxxi+1xiuy(x,yj)dx]}ΔyΔxbadcv(x,y)[uy(x,y)uy(x,y)]dxdy=0as n,m.

    This shows Ωuvy+G(u)vxdxdy=0 for vC10(Ω).

    Warning: This problem is quite difficult so I cannot promise this solution is correct. Please provide some references or theorem that I may use to make this solution better.


  26. The mth-order operator (32a) is elliptic at x if its principal symbol (32c) has no nonzero characteristics [i.e., σL(x,ξ)0 for all ξRn{0}]. In this case, show that m must be an even integer.
  27. SolutionSuppose by contradiction that m is an odd integer. Then σL(x,ξ)=|α|=maα(x)ξα is a polynomial of odd degree. Since aα(x) is not all zero, we may assume ae1(x)=a(m,0,,0)(x)0 by rearranging the indices. Thus, we have

    σL(x,ξ)=ame1(x)ξm1+|α|=m,αme1aα(x)ξα.

    Since m is odd and aα(x) are real numbers, by the fundamental theorem of algebra, there exists ξ1=ξ1(ξ2,,ξm,x) for any (ξ2,,ξm)0. This means the principal symbol has nonzero characteristics, which leads a contradiction. Hence m must be an even integer.

沒有留言:

張貼留言