The Central Idea
A differential equation by itself often determines a family of solutions.
For example,
\[ y'=2x \]has the general solution
\[ y=x^2+C. \]The differential equation determines the shape of the family, but the constant \(C\) remains unknown.
Additional conditions select a particular member of the family.
Initial Conditions
An initial condition specifies information about a solution at one value of the independent variable.
For a first-order equation, a typical condition is
\[ \boxed{y(x_0)=y_0}. \]This tells us that the solution passes through the point
\[ (x_0,y_0). \]Initial-Value Problems
A differential equation together with initial data is called an initial-value problem, or IVP.
A first-order IVP often has the form
\[ \boxed{ y'=f(x,y), \qquad y(x_0)=y_0 }. \]The differential equation controls the slope, while the initial condition selects the starting point.
Example: First-Order Initial Condition
Solve
\[ y'=4x, \qquad y(1)=7. \]Integrate:
\[ y=2x^2+C. \]Apply the initial condition:
\[ 7=2(1)^2+C. \]Therefore,
\[ C=5. \]The particular solution is
\[ \boxed{y=2x^2+5}. \]Second-Order Initial Conditions
A second-order equation usually requires two independent conditions.
A common initial-value problem is
\[ \boxed{ y''=f(x,y,y'), \qquad y(x_0)=y_0, \qquad y'(x_0)=v_0 }. \]The first condition specifies the state, while the second specifies its initial rate of change.
Mechanical Interpretation
If \(x(t)\) represents position, then
\[ x(0)=x_0 \]gives the initial position.
The condition
\[ x'(0)=v_0 \]gives the initial velocity.
These two pieces of information naturally match a second-order equation involving acceleration.
Example: Second-Order Initial-Value Problem
Solve
\[ y''-3y'+2y=0, \]with
\[ y(0)=3, \qquad y'(0)=4. \]The characteristic equation is
\[ r^2-3r+2=0. \]Therefore,
\[ r=1,\qquad r=2. \]The general solution is
\[ y=C_1e^x+C_2e^{2x}. \]Apply Both Initial Conditions
Differentiate:
\[ y'=C_1e^x+2C_2e^{2x}. \]At \(x=0\),
\[ C_1+C_2=3 \]and
\[ C_1+2C_2=4. \]Subtracting gives
\[ C_2=1. \]Therefore,
\[ C_1=2. \]Hence
\[ \boxed{y=2e^x+e^{2x}}. \]Why the Number of Conditions Matters
A first-order general solution usually contains one arbitrary constant.
A second-order general solution usually contains two.
More generally, an \(n\)th-order differential equation commonly requires \(n\) independent conditions to determine a particular solution.
The conditions must be independent and compatible.
Boundary Conditions
A boundary condition constrains the solution at one or more specified locations, often at the ends of an interval.
For example,
\[ \boxed{ y(a)=A, \qquad y(b)=B }. \]The conditions occur at two different points.
Boundary-Value Problems
A differential equation together with boundary conditions is called a boundary-value problem, or BVP.
A typical second-order BVP is
\[ \boxed{ y''+p(x)y'+q(x)y=g(x), \qquad y(a)=A, \qquad y(b)=B }. \]Boundary conditions are especially common in problems involving spatial domains.
Initial Conditions versus Boundary Conditions
Initial conditions typically provide all required data at one point.
Boundary conditions distribute the data across one or more locations.
For example:
\[ y(0)=2, \qquad y'(0)=3 \]is initial data.
By contrast,
\[ y(0)=2, \qquad y(5)=7 \]is boundary data.
Example: Simple Boundary-Value Problem
Solve
\[ y''=0 \]with
\[ y(0)=2, \qquad y(4)=10. \]Integrate:
\[ y'=C_1. \]Integrate again:
\[ y=C_1x+C_2. \]Complete the Boundary Example
From
\[ y(0)=2, \]we obtain
\[ C_2=2. \]From
\[ y(4)=10, \]we get
\[ 4C_1+2=10. \]Therefore,
\[ C_1=2. \]Thus,
\[ \boxed{y=2x+2}. \]Dirichlet Boundary Conditions
A boundary condition that specifies the value of the function is called a Dirichlet condition.
Examples include
\[ y(0)=0 \]and
\[ y(L)=5. \]In heat-transfer problems, these might represent fixed endpoint temperatures.
Neumann Boundary Conditions
A boundary condition that specifies a derivative is called a Neumann condition.
Examples include
\[ y'(0)=0 \]or
\[ y'(L)=q. \]Such conditions can represent flux, slope, force, or an insulated boundary, depending on the application.
Robin or Mixed Boundary Conditions
A boundary condition may combine the function and its derivative:
\[ \boxed{ ay+b y'=c }. \]These are often called Robin or mixed boundary conditions.
They occur when a boundary response depends on both state and flux.
Example: Neumann Conditions
Suppose
\[ y''=6x \]with
\[ y'(0)=1 \]and
\[ y(0)=4. \]Integrate:
\[ y'=3x^2+C_1. \]Apply
\[ y'(0)=1, \]giving
\[ C_1=1. \]Complete the Neumann Example
Integrate again:
\[ y=x^3+x+C_2. \]Apply
\[ y(0)=4. \]Therefore,
\[ C_2=4. \]Thus,
\[ \boxed{y=x^3+x+4}. \]Existence and Uniqueness for Initial-Value Problems
For a first-order problem
\[ y'=f(x,y), \qquad y(x_0)=y_0, \]suitable continuity conditions on \(f\) can guarantee that a local solution exists.
Additional regularity in the \(y\)-direction can guarantee that the solution is locally unique.
A Useful Introductory Criterion
If
\[ f(x,y) \]and
\[ \frac{\partial f}{\partial y} \]are continuous near
\[ (x_0,y_0), \]then the initial-value problem
\[ y'=f(x,y), \qquad y(x_0)=y_0 \]has a unique local solution.
Why Uniqueness Matters
If uniqueness holds, two distinct solution curves cannot pass through the same initial point and then evolve differently.
The initial state determines one local trajectory.
This gives deterministic meaning to many differential-equation models.
Boundary-Value Problems Behave Differently
A boundary-value problem may have:
- exactly one solution;
- no solution;
- infinitely many solutions.
Boundary constraints can interact with the differential equation in ways that do not arise in ordinary well-posed initial-value problems.
Example: A Boundary Problem with No Solution
Consider
\[ y''=0 \]with
\[ y'(0)=0 \]and
\[ y'(1)=1. \]Since
\[ y''=0, \]the derivative \(y'\) must be constant.
It cannot equal both \(0\) and \(1\).
Therefore,
\[ \boxed{\text{no solution exists}}. \]Example: Infinitely Many Boundary Solutions
Consider
\[ y''+\pi^2y=0 \]with
\[ y(0)=0, \qquad y(1)=0. \]The general solution is
\[ y=C_1\cos(\pi x)+C_2\sin(\pi x). \]From
\[ y(0)=0, \]we obtain
\[ C_1=0. \]At \(x=1\),
\[ y(1)=C_2\sin\pi=0 \]for every \(C_2\).
Therefore,
\[ \boxed{ y=C_2\sin(\pi x) } \]gives infinitely many solutions.
Eigenvalue Boundary Problems
Some boundary-value problems contain a parameter whose allowed values are determined by the boundary conditions.
A classic example is
\[ y''+\lambda y=0, \]with
\[ y(0)=0, \qquad y(L)=0. \]Nonzero solutions exist only for special values of \(\lambda\).
Deriving the Eigenvalues
For \(\lambda>0\), write
\[ \lambda=\mu^2. \]Then
\[ y=C_1\cos(\mu x)+C_2\sin(\mu x). \]The condition
\[ y(0)=0 \]gives
\[ C_1=0. \]The condition
\[ y(L)=0 \]requires
\[ C_2\sin(\mu L)=0. \]Allowed Boundary Eigenvalues
For a nonzero solution,
\[ C_2\neq0. \]Therefore,
\[ \sin(\mu L)=0. \]Hence
\[ \mu L=n\pi, \qquad n=1,2,3,\ldots \]so
\[ \boxed{ \lambda_n = \left(\frac{n\pi}{L}\right)^2 }. \]The corresponding solution shapes are
\[ \boxed{ y_n(x) = C\sin\left(\frac{n\pi x}{L}\right) }. \]Why Eigenvalue Problems Matter
Boundary eigenvalue problems appear in:
- vibrating strings;
- heat conduction;
- wave equations;
- quantum mechanics;
- structural modes;
- separation of variables for partial differential equations.
The boundary conditions select the allowable modes.
Boundary Conditions in Heat Problems
Suppose \(T(x)\) represents steady temperature along a rod.
Conditions such as
\[ T(0)=T_0, \qquad T(L)=T_L \]specify endpoint temperatures.
A condition such as
\[ T'(0)=0 \]may represent zero heat flux at an insulated end.
Boundary Conditions in Mechanics
For a beam or string, boundary conditions can encode how endpoints are supported or constrained.
Examples include:
\[ y(0)=0 \]for a fixed displacement,
\[ y'(0)=0 \]for a zero slope, or more advanced combinations involving moments and shear.
Boundary Conditions in Diffusion and Transport
In diffusion problems, boundary conditions may specify:
- concentration at a boundary;
- flux through a boundary;
- exchange with an environment;
- symmetry conditions.
The condition is part of the physical model, not merely an algebraic afterthought.
Conditions and Model Interpretation
The same differential equation can represent different systems when paired with different conditions.
For example,
\[ y''+y=0 \]describes a family of oscillations.
Initial conditions determine amplitude and phase.
Boundary conditions may instead select particular spatial modes.
Over-Specified Problems
A problem can include too many independent conditions.
For a first-order equation with one arbitrary constant, two independent conditions may conflict.
For example,
\[ y'=1 \]has
\[ y=x+C. \]If we require
\[ y(0)=0 \]and
\[ y(1)=5, \]the first condition gives
\[ y=x. \]But then
\[ y(1)=1, \]not \(5\).
The conditions are incompatible.
Under-Specified Problems
A problem can also provide too little information.
Suppose
\[ y''=0 \]with only
\[ y(0)=2. \]The general solution is
\[ y=C_1x+C_2. \]The condition gives
\[ C_2=2, \]but \(C_1\) remains arbitrary.
Therefore, infinitely many solutions remain.
Independent Conditions
Having the correct number of conditions is not enough by itself.
The conditions must provide independent information.
Two conditions that reduce to the same equation for the constants do not uniquely determine two unknown constants.
The Shooting Method: Basic Idea
A boundary-value problem can sometimes be approached by converting it into a sequence of initial-value problems.
Suppose we know
\[ y(a)=A \]and
\[ y(b)=B \]but do not know
\[ y'(a). \]We can guess an initial slope, solve the resulting IVP, and see how close the solution comes to the required value at \(b\).
Shooting as a Numerical Strategy
If the computed endpoint value is too high or too low, adjust the guessed initial slope.
Repeating this process converts the boundary problem into a root-finding problem for the missing initial slope.
This is the basic intuition behind the shooting method.
Verification of Initial-Value Solutions
A valid IVP solution must satisfy:
- the differential equation;
- every initial condition;
- the required interval or domain conditions.
Checking only the differential equation is incomplete.
Verification of Boundary-Value Solutions
A valid BVP solution must satisfy:
- the differential equation throughout the interval;
- every boundary condition;
- any regularity or domain requirements.
Endpoint checks are essential.
Common Error: Confusing Initial and Boundary Conditions
Conditions such as
\[ y(0)=2, \qquad y'(0)=3 \]are initial conditions because they occur at the same point.
Conditions such as
\[ y(0)=2, \qquad y(5)=7 \]are boundary conditions because they constrain different locations.
Common Error: Using the Wrong Number of Conditions
A second-order equation usually needs two independent conditions.
One condition generally leaves one constant undetermined.
Three independent conditions may over-specify the problem unless they happen to be compatible.
Common Error: Applying a Condition Before Differentiating Correctly
If a condition involves
\[ y', \]first differentiate the complete general solution carefully.
Then substitute the specified point.
Do not apply a derivative condition directly to the undifferentiated expression.
Common Error: Ignoring Compatibility
Conditions must agree with the differential equation.
A set of conditions can be algebraically stated but mathematically impossible.
If the resulting equations for the arbitrary constants are inconsistent, the boundary or initial data are incompatible.
Common Error: Assuming Every Boundary Problem Has a Unique Solution
Boundary-value problems can have zero, one, or many solutions.
Uniqueness theorems for initial-value problems do not automatically transfer to boundary-value problems.
Common Error: Ignoring Physical Meaning
A boundary condition is part of the model.
For example:
- fixed temperature and zero heat flux are different physical conditions;
- fixed displacement and fixed slope are different mechanical conditions;
- known concentration and known mass flux describe different transport constraints.
The mathematics should reflect the actual system.
A Reliable Initial-Value Workflow
- Identify the order of the differential equation.
- Determine how many independent constants appear in the general solution.
- Write the general solution.
- Differentiate as needed for derivative conditions.
- Substitute the initial point into every condition.
- Solve the resulting algebraic system for the constants.
- Write the particular solution.
- Verify the differential equation.
- Verify every initial condition.
- Check the interval of validity.
A Reliable Boundary-Value Workflow
- Identify the differential equation and its order.
- Identify the type and location of each boundary condition.
- Find the general solution when possible.
- Apply all boundary conditions.
- Solve for the arbitrary constants.
- Check whether the algebraic system is consistent.
- Determine whether the solution is unique, absent, or nonunique.
- Verify the differential equation and all boundary conditions.
- Interpret the conditions in the model context.
Example: Full Initial-Value Workflow
Solve
\[ y''+4y=0 \]with
\[ y(0)=3, \qquad y'(0)=-2. \]The characteristic equation is
\[ r^2+4=0. \]Thus,
\[ r=\pm2i. \]Therefore,
\[ y=C_1\cos2x+C_2\sin2x. \]Complete the Full Initial-Value Workflow
Differentiate:
\[ y'=-2C_1\sin2x+2C_2\cos2x. \]From
\[ y(0)=3, \]we get
\[ C_1=3. \]From
\[ y'(0)=-2, \]we get
\[ 2C_2=-2. \]Therefore,
\[ C_2=-1. \]Hence
\[ \boxed{ y=3\cos2x-\sin2x }. \]Example: Full Boundary-Value Workflow
Solve
\[ y''+y=0 \]with
\[ y(0)=1, \qquad y\left(\frac{\pi}{2}\right)=0. \]The general solution is
\[ y=C_1\cos x+C_2\sin x. \]From
\[ y(0)=1, \]we obtain
\[ C_1=1. \]At
\[ x=\frac{\pi}{2}, \]we require
\[ 0=C_2. \]Therefore,
\[ \boxed{y=\cos x}. \]Practice 1: First-Order Initial Value
Solve
\[ y'=6x, \qquad y(0)=4. \]Practice 2: Second-Order Initial Value
Solve
\[ y''=0, \qquad y(0)=1, \qquad y'(0)=3. \]Practice 3: Boundary Values
Solve
\[ y''=0, \qquad y(0)=2, \qquad y(5)=12. \]Practice 4: Classify the Conditions
Are
\[ y(0)=2, \qquad y'(0)=5 \]initial conditions or boundary conditions?
Practice 5: Classify the Conditions
Are
\[ y(0)=1, \qquad y(3)=4 \]initial conditions or boundary conditions?
Practice 6: Dirichlet or Neumann?
Classify
\[ y'(L)=0. \]Practice 7: Compatibility
Determine whether
\[ y'=2 \]can satisfy both
\[ y(0)=1 \]and
\[ y(1)=10. \]Practice 8: Number of Conditions
How many independent conditions are typically needed to determine a unique particular solution of a third-order differential equation?
Practice 9: Boundary Eigenvalues
For
\[ y''+\lambda y=0, \qquad y(0)=0, \qquad y(\pi)=0, \]find the positive eigenvalues.
Practice 10: Verify Conditions
Determine whether
\[ y=2\cos x+3\sin x \]satisfies
\[ y''+y=0, \qquad y(0)=2, \qquad y'(0)=3. \]Practice Answers
Practice 1
Integrate:
\[ y=3x^2+C. \]Apply
\[ y(0)=4. \]Therefore,
\[ C=4. \]Thus,
\[ \boxed{y=3x^2+4}. \]Practice 2
Since
\[ y''=0, \]integrate:
\[ y'=C_1. \]From
\[ y'(0)=3, \]we get
\[ C_1=3. \]Integrate again:
\[ y=3x+C_2. \]Apply
\[ y(0)=1, \]giving
\[ C_2=1. \]Therefore,
\[ \boxed{y=3x+1}. \]Practice 3
The general solution is
\[ y=C_1x+C_2. \]From
\[ y(0)=2, \]we get
\[ C_2=2. \]From
\[ y(5)=12, \]we obtain
\[ 5C_1+2=12. \]Therefore,
\[ C_1=2. \]Thus,
\[ \boxed{y=2x+2}. \]Practice 4
Both conditions are specified at the same point \(x=0\).
Therefore, they are
\[ \boxed{\text{initial conditions}}. \]Practice 5
The function is constrained at two different locations.
Therefore, they are
\[ \boxed{\text{boundary conditions}}. \]Practice 6
Because the derivative is specified, the condition is
\[ \boxed{\text{Neumann}}. \]Practice 7
Solve
\[ y'=2. \]Then
\[ y=2x+C. \]From
\[ y(0)=1, \]we get
\[ C=1. \]Therefore,
\[ y=2x+1. \]But
\[ y(1)=3, \]not \(10\).
Therefore,
\[ \boxed{\text{the conditions are incompatible}}. \]Practice 8
A third-order equation typically requires
\[ \boxed{3} \]independent conditions.
Practice 9
Since
\[ L=\pi, \]the eigenvalues are
\[ \lambda_n = \left( \frac{n\pi}{\pi} \right)^2. \]Therefore,
\[ \boxed{ \lambda_n=n^2, \qquad n=1,2,3,\ldots }. \]Practice 10
Differentiate:
\[ y'=-2\sin x+3\cos x \]and
\[ y''=-2\cos x-3\sin x. \]Therefore,
\[ y''+y=0. \]Also,
\[ y(0)=2 \]and
\[ y'(0)=3. \]Thus,
\[ \boxed{\text{all conditions are satisfied}}. \]What to Remember
Differential equations usually produce families of solutions, while conditions select the solution relevant to a particular problem.
- initial conditions specify data at one point;
- boundary conditions constrain one or more locations;
- an \(n\)th-order equation typically requires \(n\) independent conditions;
- Dirichlet conditions specify function values;
- Neumann conditions specify derivative values;
- Robin conditions combine function and derivative information;
- initial-value problems often have strong local uniqueness results;
- boundary-value problems may have zero, one, or many solutions;
- boundary conditions can select special eigenvalues and modes;
- conditions must be compatible with both the equation and the model.
The central structural idea is
\[ \boxed{ \text{differential equation} + \text{conditions} \longrightarrow \text{particular solution} }. \]Where We Go Next
Initial and boundary conditions select particular trajectories, but we also need ways to understand how entire families of trajectories behave geometrically.
The next lesson, Phase-Plane Representations, develops state-space plots, vector fields, nullclines, equilibria, trajectories, stability, nodes, saddles, spirals, and qualitative analysis of two-dimensional systems.