Partial Differential Equations Study Notes

Definition and Notation of Partial Differential Equations

  • Definition: A Partial Differential Equation (PDE) is a differential equation involving partial derivatives of one or more dependent variables with respect to one or more independent variables.

  • Standard Notations:

    • p=∂z∂xp = \frac{\partial z}{\partial x}

    • q=∂z∂yq = \frac{\partial z}{\partial y}

    • r=∂2z∂x2r = \frac{\partial^2 z}{\partial x^2}

    • s=∂2z∂x ∂ys = \frac{\partial^2 z}{\partial x \, \partial y}

    • t=∂2z∂y2t = \frac{\partial^2 z}{\partial y^2}

Formation of Partial Differential Equations

  • By Elimination of Arbitrary Constants: Differentiating an equation f(x,y,z,a,b)=0f(x, y, z, a, b) = 0 partially with respect to xx and yy and eliminating the constants aa and bb yields a first-order PDE of the form f(x,y,z,p,q)=0f(x, y, z, p, q) = 0

  • By Elimination of Arbitrary Functions: Eliminating arbitrary functions ff from an expression f(u,v)=0f(u, v) = 0 (where uu and vv are functions of x,y,zx, y, z) yields a first-order linear PDE Pp+Qq=RP p + Q q = R

Types of Integrals / Solutions of PDEs

  • Complete Integral: A solution containing as many arbitrary constants as independent variables (ϕ(x,y,z,a,b)=0\phi(x, y, z, a, b) = 0)

  • Particular Integral: A solution obtained by assigning specific numerical values to arbitrary constants in the complete integral

  • Singular Integral: Represents the envelope of the family of surfaces given by the complete integral, obtained by eliminating aa and bb from ϕ=0\phi = 0, ∂ϕ∂a=0\frac{\partial \phi}{\partial a} = 0, and ∂ϕ∂b=0\frac{\partial \phi}{\partial b} = 0

  • General Integral: A solution of the form f(u,v)=0f(u, v) = 0 containing arbitrary functions

Lagrange's Solution of First-Order Linear PDEs

  • Form: Pp+Qq=RP p + Q q = R, where P,Q,RP, Q, R are functions of x,y,zx, y, z

  • Lagrange's Auxiliary Equations:   dxP=dyQ=dzR\frac{dx}{P} = \frac{dy}{Q} = \frac{dz}{R}

  • Method: Solve the auxiliary equations to find two independent integrals u(x,y,z)=c1u(x, y, z) = c_1 and v(x,y,z)=c2v(x, y, z) = c_2. The general solution is expressed as f(u,v)=0f(u, v) = 0

Standard Forms for Non-Linear First-Order PDEs

  • Standard Form I: f(p,q)=0f(p, q) = 0

    • Complete Integral: z=ax+by+cz = ax + by + c, where b=ϕ(a)b = \phi(a) from f(a,b)=0f(a, b) = 0

  • Standard Form II: f(p,q,z)=0f(p, q, z) = 0

    • Method: Substitute X=x+ayX = x + ay, which converts the PDE into an ordinary differential equation (ODE) f(dzdX,adzdX,z)=0f\left(\frac{dz}{dX}, a\frac{dz}{dX}, z\right) = 0

  • Standard Form III: f(p,x)=g(q,y)f(p, x) = g(q, y)

    • Method: Equate both sides to an arbitrary constant aa (f(p,x)=af(p, x) = a and g(q,y)=ag(q, y) = a), solve for pp and qq, and integrate dz=p dx+q dydz = p\,dx + q\,dy

  • Standard Form IV (Clairaut's Form): z=px+qy+f(p,q)z = px + qy + f(p, q)

    • Complete Integral: z=ax+by+f(a,b)z = ax + by + f(a, b)

Charpit's Method for General First-Order Non-Linear PDEs

  • Charpit's Auxiliary Equations:   dp∂f∂x+p∂f∂z=dq∂f∂y+q∂f∂z=dz−p∂f∂p−q∂f∂q=dx−∂f∂p=dy−∂f∂q\frac{dp}{\frac{\partial f}{\partial x} + p \frac{\partial f}{\partial z}} = \frac{dq}{\frac{\partial f}{\partial y} + q \frac{\partial f}{\partial z}} = \frac{dz}{-p \frac{\partial f}{\partial p} - q \frac{\partial f}{\partial q}} = \frac{dx}{-\frac{\partial f}{\partial p}} = \frac{dy}{-\frac{\partial f}{\partial q}}

  • Method: Solve for pp and qq using the auxiliary equations, then integrate dz=p dx+q dydz = p\,dx + q\,dy

Homogeneous Linear PDEs with Constant Coefficients

  • General Form: F(D,D′)z=f(x,y)F(D, D')z = f(x, y), where D=∂∂xD = \frac{\partial}{\partial x} and D′=∂∂yD' = \frac{\partial}{\partial y}

  • Auxiliary Equation (A.E.): Formed by substituting D=mD = m and D′=1D' = 1 into F(D,D′)=0F(D, D') = 0

  • Complementary Function (C.F.):

    • Distinct real roots α1,α2,…,αn\alpha_1, \alpha_2, \dots, \alpha_n: C.F.=f1(y+α1x)+f2(y+α2x)+⋯+fn(y+αnx)C.F. = f_1(y + \alpha_1 x) + f_2(y + \alpha_2 x) + \dots + f_n(y + \alpha_n x)

    • Repeated root α\alpha (twice): C.F.=f1(y+αx)+xf2(y+αx)C.F. = f_1(y + \alpha x) + x f_2(y + \alpha x)

Classification of Second-Order Linear PDEs

  • General Equation: Rr+Ss+Tt+f(x,y,z,p,q)=0R r + S s + T t + f(x, y, z, p, q) = 0

  • Classification Criterion (based on discriminant S2−4RTS^2 - 4RT):

    • Hyperbolic: S2−4RT>0S^2 - 4RT > 0

    • Parabolic: S2−4RT=0S^2 - 4RT = 0

    • Elliptic: S2−4RT<0S^2 - 4RT < 0

Classical Second-Order Partial Differential Equations

  • Wave Equation (1D): ∂2u∂t2=c2∂2u∂x2\frac{\partial^2 u}{\partial t^2} = c^2 \frac{\partial^2 u}{\partial x^2}

    • D'Alembert's Solution: u(x,t)=12[ϕ(x−ct)+ϕ(x+ct)]+12c∫x−ctx+ctψ(s) dsu(x, t) = \frac{1}{2}[\phi(x - ct) + \phi(x + ct)] + \frac{1}{2c} \int_{x - ct}^{x + ct} \psi(s)\,ds

  • Heat Equation (1D): ∂u∂t=c2∂2u∂x2\frac{\partial u}{\partial t} = c^2 \frac{\partial^2 u}{\partial x^2}

    • Solved using the method of separation of variables yielding Fourier sine/cosine series solutions.

  • Laplace Equation (2D): ∇2u=∂2u∂x2+∂2u∂y2=0\nabla^2 u = \frac{\partial^2 u}{\partial x^2} + \frac{\partial^2 u}{\partial y^2} = 0

    • Functions satisfying Laplace's equation are called harmonic functions.

Sturm-Liouville Problems and Green's Function

  • Sturm-Liouville Equation: ddx(p(x)dydx)+q(x)y=λy\frac{d}{dx}\left(p(x) \frac{dy}{dx}\right) + q(x) y = \lambda y

  • Properties: Eigenvalues λn\lambda_n are real, ordered, and countable; corresponding eigenfunctions yn(x)y_n(x) are orthogonal.

  • Green's Function: Used to construct integral solutions for non-homogeneous boundary value problems Ly=f(x)L y = f(x) in the form y(x)=∫abG(x,t)f(t) dty(x) = \int_a^b G(x, t) f(t)\,dt


Definition and Notation of Partial Differential Equations

  • Definition: A Partial Differential Equation (PDE) is a differential equation involving partial derivatives of one or more dependent variables with respect to two or more independent variables.

  • Standard Notations (Monge's Notation):

    • p=∂z∂xp = \frac{\partial z}{\partial x}

    • q=∂z∂yq = \frac{\partial z}{\partial y}

    • r=∂2z∂x2r = \frac{\partial^2 z}{\partial x^2}

    • s=∂2z∂x ∂ys = \frac{\partial^2 z}{\partial x \,\partial y}

    • t=∂2z∂y2t = \frac{\partial^2 z}{\partial y^2}

Formation of Partial Differential Equations

  • By Elimination of Arbitrary Constants:

    • Differentiating an equation f(x,y,z,a,b)=0f(x, y, z, a, b) = 0 partially with respect to xx and yy gives equations involving pp and qq.

    • Eliminating constants aa and bb yields a first-order PDE f(x,y,z,p,q)=0f(x, y, z, p, q) = 0.

    • If the number of arbitrary constants equals the number of independent variables, a first-order PDE is obtained.

  • By Elimination of Arbitrary Functions:

    • Eliminating arbitrary functions ff from an expression f(u,v)=0f(u, v) = 0 (where u=u(x,y,z)u = u(x, y, z) and v=v(x,y,z)v = v(x, y, z)) yields a first-order linear PDE of Lagrange's form Pp+Qq=RP p + Q q = R.

Types of Integrals / Solutions of PDEs

  • Complete Integral: A solution ϕ(x,y,z,a,b)=0\phi(x, y, z, a, b) = 0 containing as many arbitrary constants (a,ba, b) as independent variables (x,yx, y).

  • Particular Integral: A solution obtained by assigning specific numerical values to the arbitrary constants in the complete integral.

  • Singular Integral: Represents the envelope of the family of surfaces given by the complete integral. It is obtained by eliminating aa and bb from:

    • ϕ(x,y,z,a,b)=0\phi(x, y, z, a, b) = 0

    • ∂ϕ∂a=0\frac{\partial \phi}{\partial a} = 0

    • ∂ϕ∂b=0\frac{\partial \phi}{\partial b} = 0

  • General Integral: Obtained by setting b=ϕ(a)b = \phi(a) in the complete integral to get ϕ(x,y,z,a,ϕ(a))=0\phi(x, y, z, a, \phi(a)) = 0, differentiating with respect to aa, and eliminating aa. It corresponds to the form f(u,v)=0f(u, v) = 0.

Lagrange's Solution of First-Order Linear PDEs

  • Form: Pp+Qq=RP p + Q q = R, where P,Q,RP, Q, R are functions of x,y,zx, y, z.

  • Lagrange's Auxiliary Equations:   dxP=dyQ=dzR\frac{dx}{P} = \frac{dy}{Q} = \frac{dz}{R}

  • Method of Solution:

    1. Grouping Method: Pair any two ratios to integrate directly and find u(x,y,z)=c1u(x, y, z) = c_1.

    2. Method of Multipliers: Find multipliers (l,m,n)(l, m, n) such that lP+mQ+nR=0l P + m Q + n R = 0, implying l dx+m dy+n dz=0l\,dx + m\,dy + n\,dz = 0, which integrates to v(x,y,z)=c2v(x, y, z) = c_2.

  • General Solution: Expressed as f(u,v)=0f(u, v) = 0 or u=ϕ(v)u = \phi(v).

Standard Forms for Non-Linear First-Order PDEs

  • Standard Form I: f(p,q)=0f(p, q) = 0

    • Complete Integral: z=ax+by+cz = a x + b y + c, where bb is determined from f(a,b)=0f(a, b) = 0.

  • Standard Form II: f(p,q,z)=0f(p, q, z) = 0

    • Method: Substitute X=x+ayX = x + a y, transforming the PDE into an ordinary differential equation (ODE):     f(dzdX,adzdX,z)=0f\left(\frac{dz}{dX}, a \frac{dz}{dX}, z\right) = 0

  • Standard Form III: f(p,x)=g(q,y)f(p, x) = g(q, y)

    • Method: Equate both sides to an arbitrary constant aa (f(p,x)=a⇒p=h1(x,a)f(p, x) = a \Rightarrow p = h_1(x, a) and g(q,y)=a⇒q=h2(y,a)g(q, y) = a \Rightarrow q = h_2(y, a)), then integrate dz=p dx+q dydz = p\,dx + q\,dy to get:     z=∫h1(x,a) dx+∫h2(y,a) dy+bz = \int h_1(x, a)\,dx + \int h_2(y, a)\,dy + b

  • Standard Form IV (Clairaut's Form): z=px+qy+f(p,q)z = p x + q y + f(p, q)

    • Complete Integral: z=ax+by+f(a,b)z = a x + b y + f(a, b).

Charpit's Method for General First-Order Non-Linear PDEs

  • Applicable to general first-order equations f(x,y,z,p,q)=0f(x, y, z, p, q) = 0.

  • Charpit's Auxiliary Equations:   dp∂f∂x+p∂f∂z=dq∂f∂y+q∂f∂z=dz−p∂f∂p−q∂f∂q=dx−∂f∂p=dy−∂f∂q\frac{dp}{\frac{\partial f}{\partial x} + p \frac{\partial f}{\partial z}} = \frac{dq}{\frac{\partial f}{\partial y} + q \frac{\partial f}{\partial z}} = \frac{dz}{-p \frac{\partial f}{\partial p} - q \frac{\partial f}{\partial q}} = \frac{dx}{-\frac{\partial f}{\partial p}} = \frac{dy}{-\frac{\partial f}{\partial q}}

  • Method: Solve for pp and qq using auxiliary equations and the original PDE, then integrate dz=p dx+q dydz = p\,dx + q\,dy

Homogeneous and Non-Homogeneous Linear PDEs with Constant Coefficients

  • General Form: F(D,D′)z=f(x,y)F(D, D') z = f(x, y), where D=∂∂xD = \frac{\partial}{\partial x} and D′=∂∂yD' = \frac{\partial}{\partial y}.

  • Auxiliary Equation (A.E.): Formed by substituting D=mD = m and D′=1D' = 1 into F(D,D′)=0F(D, D') = 0

  • Complementary Function (C.F.):

    • Distinct real roots α1,α2,…,αn\alpha_1, \alpha_2, \dots, \alpha_n:     C.F.=f1(y+α1x)+f2(y+α2x)+⋯+fn(y+αnx)C.F. = f_1(y + \alpha_1 x) + f_2(y + \alpha_2 x) + \dots + f_n(y + \alpha_n x)

    • Repeated root α\alpha (mm times):     C.F.=f1(y+αx)+xf2(y+αx)+⋯+xm−1fm(y+αx)C.F. = f_1(y + \alpha x) + x f_2(y + \alpha x) + \dots + x^{m-1} f_m(y + \alpha x)

  • Particular Integral (P.I.):

    • For f(x,y)=exp⁡(ax+by)f(x, y) = \exp(a x + b y): P.I.=1F(a,b)exp⁡(ax+by)P.I. = \frac{1}{F(a, b)} \exp(a x + b y) provided F(a,b)≠0F(a, b) \neq 0.

    • For f(x,y)=sin⁡(ax+by)f(x, y) = \sin(a x + b y) or cos⁡(ax+by)\cos(a x + b y): Substitute D2→−a2D^2 \rightarrow -a^2, DD′→−abD D' \rightarrow -a b, and D′2→−b2{D'}^2 \rightarrow -b^2.

Cauchy's Problem and Existence / Uniqueness

  • Initial Value Problem: Consists of finding a solution surface passing through a given initial space curve x=x0(t)x = x_0(t), y=y0(t)y = y_0(t), z=z0(t)z = z_0(t).

  • Uniqueness Condition: For Pp+Qq=RP p + Q q = R, a unique solution exists if:   P(x0,y0,z0)dy0dt−Q(x0,y0,z0)dx0dt≠0P(x_0, y_0, z_0) \frac{dy_0}{dt} - Q(x_0, y_0, z_0) \frac{dx_0}{dt} \neq 0

  • No Solution / Infinite Solutions: If Pdy0dt−Qdx0dt=0P \frac{dy_0}{dt} - Q \frac{dx_0}{dt} = 0, then:

    • Infinite Solutions if Rdx0dt=Pdz0dtR \frac{dx_0}{dt} = P \frac{dz_0}{dt}.

    • No Solution if Rdx0dt≠Pdz0dtR \frac{dx_0}{dt} \neq P \frac{dz_0}{dt}.

Classification and Canonical Forms of Second-Order Linear PDEs

  • General Form: Rr+Ss+Tt+f(x,y,z,p,q)=0R r + S s + T t + f(x, y, z, p, q) = 0

  • Classification Criterion (discriminant Δ=S2−4RT\Delta = S^2 - 4 R T):

    • Hyperbolic: S2−4RT>0S^2 - 4 R T > 0 (Two real characteristic curves). Canonical form: ∂2z∂u ∂v=ϕ(u,v,z,zu,zv)\frac{\partial^2 z}{\partial u \,\partial v} = \phi(u, v, z, z_u, z_v).

    • Parabolic: S2−4RT=0S^2 - 4 R T = 0 (One real characteristic curve). Canonical form: ∂2z∂v2=ϕ(u,v,z,zu,zv)\frac{\partial^2 z}{\partial v^2} = \phi(u, v, z, z_u, z_v).

    • Elliptic: S2−4RT<0S^2 - 4 R T < 0 (Complex conjugate characteristic curves). Canonical form: ∂2z∂ξ2+∂2z∂η2=ϕ(ξ,η,z,zξ,zη)\frac{\partial^2 z}{\partial \xi^2} + \frac{\partial^2 z}{\partial \eta^2} = \phi(\xi, \eta, z, z_\xi, z_\eta).

Classical Second-Order Partial Differential Equations

  • Wave Equation (1D): ∂2u∂t2=c2∂2u∂x2\frac{\partial^2 u}{\partial t^2} = c^2 \frac{\partial^2 u}{\partial x^2}

    • D'Alembert's Solution for infinite domain (−∞<x<∞-\infty < x < \infty):     u(x,t)=12[ϕ(x−ct)+ϕ(x+ct)]+12c∫x−ctx+ctψ(s) dsu(x, t) = \frac{1}{2}\left[\phi(x - c t) + \phi(x + c t)\right] + \frac{1}{2 c} \int_{x - c t}^{x + c t} \psi(s)\,ds

    • Fourier Series Solution for finite domain (0≤x≤l0 \le x \le l) with u(0,t)=u(l,t)=0u(0, t) = u(l, t) = 0:     u(x,t)=∑n=1∞sin⁡(nπxl)[ancos⁡(nπctl)+bnsin⁡(nπctl)]u(x, t) = \sum_{n=1}^{\infty} \sin\left(\frac{n \pi x}{l}\right) \left[a_n \cos\left(\frac{n \pi c t}{l}\right) + b_n \sin\left(\frac{n \pi c t}{l}\right)\right]

  • Heat Equation (1D): ∂u∂t=c2∂2u∂x2\frac{\partial u}{\partial t} = c^2 \frac{\partial^2 u}{\partial x^2}

    • Solved via separation of variables: u(x,t)=∑n=1∞cnsin⁡(nπxl)exp⁡(−c2n2π2tl2)u(x, t) = \sum_{n=1}^{\infty} c_n \sin\left(\frac{n \pi x}{l}\right) \exp\left(-\frac{c^2 n^2 \pi^2 t}{l^2}\right).

  • Laplace Equation (2D): ∇2u=∂2u∂x2+∂2u∂y2=0\nabla^2 u = \frac{\partial^2 u}{\partial x^2} + \frac{\partial^2 u}{\partial y^2} = 0

    • Functions satisfying Laplace's equation are called harmonic functions.

    • In polar coordinates: ∂2u∂r2+1r∂u∂r+1r2∂2u∂θ2=0\frac{\partial^2 u}{\partial r^2} + \frac{1}{r} \frac{\partial u}{\partial r} + \frac{1}{r^2} \frac{\partial^2 u}{\partial \theta^2} = 0.

Sturm-Liouville Problems and Green's Function

  • Sturm-Liouville Equation: ddx(p(x)dydx)+q(x)y+λr(x)y=0\frac{d}{dx}\left(p(x) \frac{dy}{dx}\right) + q(x) y + \lambda r(x) y = 0

  • Key Properties:

    • Eigenvalues λn\lambda_n are real, countable, and ordered (λ1<λ2<…\lambda_1 < \lambda_2 < \dots).

    • Eigenfunctions yn(x)y_n(x) corresponding to distinct eigenvalues are orthogonal with weight function r(x)r(x).

  • Green's Function: Used for non-homogeneous BVP L[y]=f(x)L[y] = f(x):   y(x)=∫abG(x,t)f(t) dty(x) = \int_{a}^{b} G(x, t) f(t)\,dt   where G(x,t)G(x, t) is continuous at x=tx = t, satisfies boundary conditions, and possesses a jump discontinuity in derivative at x=tx = t:   $$\left.\frac{\partial G}{\