MAE 376: Applied Mathematics for MAE - Review Test 2

0.0(0)
Studied by 0 people
call kaiCall Kai
learnLearn
examPractice Test
spaced repetitionSpaced Repetition
heart puzzleMatch
flashcardsFlashcards
GameKnowt Play
Card Sorting

1/45

flashcard set

Earn XP

Description and Tags

Vocabulary practice flashcards covering error definitions, vector and matrix norms, condition numbers, eigenproblems, numerical differentiation, Runge-Kutta IVP solvers, numerical stability, and finite difference BVPs.

46 Terms

1
New cards

True Percent Relative Error (εt\varepsilon_t)

The exact relative error calculated when the true analytical solution is known: εt=true value−approximationtrue value×100%\varepsilon_t = \frac{\text{true value} - \text{approximation}}{\text{true value}} \times 100\%

2
New cards

Approximate Percent Relative Error (εa\varepsilon_a)

An error estimate used in iterative numerical methods when the true value is unknown: εa=present approximation−previous approximationpresent approximation×100%\varepsilon_a = \frac{\text{present approximation} - \text{previous approximation}}{\text{present approximation}} \times 100\%

3
New cards

Round-off Error

Numerical error arising from the computer representation of numbers with a finite number of significant digits, which decreases as the step size increases.

4
New cards

Truncation Error

Numerical error resulting from approximating or truncating exact mathematical operations or continuous quantities in numerical methods, which generally increases as the step size increases.

5
New cards
<p>Minimum Error Point (Sweet Spot)</p>

Minimum Error Point (Sweet Spot)

The optimal step size where total numerical error reaches its minimum, formed by the balance between increasing truncation error and decreasing round-off error.

6
New cards

Norm

A mathematical operator that takes an object (such as a vector or matrix) and maps it to a single non-negative real number representing its magnitude, length, or size.

7
New cards

Euclidean Norm (Vector 2-Norm)

The geometric length of an nn-dimensional vector uu, defined via the dot product as ∥u∥2=∑i=1nui2\left\|u\right\|_2 = \sqrt{\sum_{i=1}^n u_i^2}

8
New cards

Vector p-Norm (∥u∥p\left\|u\right\|_p)

The generalized norm of an nn-dimensional vector uu, defined by the formula ∥u∥p=(∑i=1n∣ui∣p)1/p\left\|u\right\|_p = \left(\sum_{i=1}^n |u_i|^p\right)^{1/p}

9
New cards

Vector 1-Norm (∥u∥1\left\|u\right\|_1)

A specific vector norm computed as the sum of the absolute values of all vector components: ∥u∥1=∑i=1n∣ui∣\left\|u\right\|_1 = \sum_{i=1}^n |u_i|

10
New cards

Vector Infinity Norm (∥u∥∞\left\|u\right\|_\infty)

A vector norm equal to the maximum absolute value among all elements of the vector: ∥u∥∞=max⁡1≤i≤n∣ui∣\left\|u\right\|_\infty = \max_{1 \le i \le n} |u_i|

11
New cards

Frobenius Norm (∥A∥F\left\|A\right\|_F)

A matrix norm quantifying the overall size of an n×nn \times n matrix AA, defined as the square root of the sum of squares of all entries: ∥A∥F=∑i=1n∑j=1naij2\left\|A\right\|_F = \sqrt{\sum_{i=1}^n \sum_{j=1}^n a_{ij}^2}

12
New cards

Column-Sum Norm (Matrix 1-Norm)

A matrix norm determined by taking the maximum of the sums of absolute values across each column: ∥A∥1=max⁡j∑i=1n∣aij∣\left\|A\right\|_1 = \max_j \sum_{i=1}^n |a_{ij}|

13
New cards

Row-Sum Norm (Matrix Infinity Norm)

A matrix norm determined by taking the maximum of the sums of absolute values across each row: ∥A∥∞=max⁡i∑j=1n∣aij∣\left\|A\right\|_\infty = \max_i \sum_{j=1}^n |a_{ij}|

14
New cards

Condition Number (κ(A)\kappa(A))

A diagnostic measure for linear systems Ax⃗=b⃗A\vec{x} = \vec{b}, defined as κ(A)=∥A∥∥A−1∥\kappa(A) = \left\|A\right\| \left\|A^{-1}\right\|, which bounds how relative errors in b⃗\vec{b} propagate to relative errors in x⃗\vec{x}.

15
New cards

Well-Conditioned Matrix

A matrix with a condition number κ(A)\kappa(A) that is small (close to 11), ensuring that the relative error in x⃗\vec{x} is not much larger than the relative error in b⃗\vec{b}.

16
New cards

Ill-Conditioned Matrix

A matrix characterized by a large condition number κ(A)\kappa(A), which indicates that small perturbations in b⃗\vec{b} can lead to much larger relative errors in the solution x⃗\vec{x}.

17
New cards

Eigenvalue and Eigenvector

A scalar λ\lambda and a corresponding nonzero vector x⃗\vec{x} that satisfy the matrix equation Ax⃗=λx⃗A\vec{x} = \lambda\vec{x}, or equivalently (A−λI)x⃗=0⃗(A - \lambda I)\vec{x} = \vec{0}.

18
New cards

Power Method

An iterative numerical method used to approximate the dominant (largest in magnitude) eigenvalue and its associated eigenvector by repeated matrix-vector multiplication and normalization using the infinity norm.

19
New cards

Inverse Power Method

A numerical algorithm adapted from the power method to compute the smallest eigenvalue and its corresponding eigenvector of a matrix.

20
New cards

Taylor Series Expansion

An infinite series approximating a function f(x)f(x) around an expansion point xix_i, which is mathematically exact for an nnth-order polynomial when using n+1n+1 terms and produces a truncation error of order O(Δxn+1)O(\Delta x^{n+1}).

21
New cards
<p>Forward Finite Difference (First Derivative)</p>

Forward Finite Difference (First Derivative)

A first-order accurate (O(h)O(h)) numerical differentiation stencil estimating f′(xi)f'(x_i) using values at xix_i and the forward point xi+1x_{i+1}.

22
New cards
<p>Backward Finite Difference (First Derivative)</p>

Backward Finite Difference (First Derivative)

A first-order accurate (O(h)O(h)) numerical differentiation stencil estimating f′(xi)f'(x_i) using values at xix_i and the preceding point xi−1x_{i-1}.

23
New cards
<p>Centered Finite Difference (First Derivative)</p>

Centered Finite Difference (First Derivative)

A second-order accurate (O(h2)O(h^2)) numerical differentiation stencil estimating f′(xi)f'(x_i) using the symmetric span between xi+1x_{i+1} and xi−1x_{i-1}, which quarters its truncation error when the step size hh is halved.

24
New cards

Centered Finite Difference (Second Derivative)

A second-order accurate numerical approximation for the second derivative defined as d2Tdx2=Ti+1−2Ti+Ti−1Δx2\frac{d^2 T}{d x^2} = \frac{T_{i+1} - 2T_i + T_{i-1}}{\Delta x^2}

25
New cards
<p>Forward Euler Method (RK1)</p>

Forward Euler Method (RK1)

A first-order accurate (O(Δt)O(\Delta t)) explicit single-step method for IVPs that computes the solution at ti+1t_{i+1} by extrapolating linearly along the tangent slope evaluated at (ti,yi)(t_i, y_i).

26
New cards
<p>Mid-point Method (RK2)</p>

Mid-point Method (RK2)

A second-order accurate (O(Δt2)O(\Delta t^2)) Runge-Kutta scheme comprising 11 predictor stage (computing the slope at the half-interval ti+1/2t_{i+1/2}) and 11 corrector stage (advancing from yiy_i to yi+1y_{i+1} using the midpoint slope).

27
New cards
<p>Fourth-Order Runge-Kutta Method (RK4)</p>

Fourth-Order Runge-Kutta Method (RK4)

A fourth-order accurate (O(Δt4)O(\Delta t^4)) numerical method for initial value problems that computes 44 slope evaluations per time step (consisting of 33 predictors and 11 corrector) combined in a weighted average.

28
New cards

Numerical Accuracy vs. Numerical Stability

Accuracy refers to how closely a numerical solution approximates the exact solution at each step, while stability describes whether numerical errors decay or amplify over successive computational steps.

29
New cards

Amplification Factor (AA)

A scalar multiplier relating the error or solution amplitude between consecutive steps (yi+1=Ayiy_{i+1} = A y_i); stability requires ∣A∣<1|A| < 1 to ensure errors diminish over time.

30
New cards

Conditional Stability (Explicit/Forward Euler)

A stability property where explicit methods require the step size to be strictly bounded, specifically Δt<2∣λ∣\Delta t < \frac{2}{|\lambda|} for dydt=λy\frac{dy}{dt} = \lambda y, to keep ∣A∣=∣1+Δtλ∣<1|A| = |1 + \Delta t \lambda| < 1.

31
New cards

Unconditional Stability (Implicit/Backward Euler)

A stability property where a method satisfies the error decay criterion (∣A∣<1|A| < 1) for all positive step sizes Δt\Delta t, as seen in Backward Euler for dydt=λy\frac{dy}{dt} = \lambda y with λ<0\lambda < 0.

32
New cards

Ghost Point

A fictitious node (such as T0T_0) introduced outside a physical boundary to allow centered finite difference approximation of a derivative boundary condition while maintaining second-order accuracy.

33
New cards

Neumann Boundary Condition

A boundary condition in a differential equation problem that prescribes the derivative of the dependent variable at the boundary (e.g., T′(0)=0T'(0) = 0).

34
New cards

Dirichlet Boundary Condition

A boundary condition in a differential equation problem that fixes the exact value of the dependent variable at the boundary (e.g., T(10)=400 KT(10) = 400\,K ).

35
New cards

Step size error

(O(h)→O(12h)O(h)\to O\left(\frac12h\right) )

(O(h2)→O(14h2)O(h^2)\to O\left(\frac14h^2\right) )

36
New cards

Order of accuracy is not the same as size of error

A (second-order) central difference has error O(Δx²), which you can write out as O((L/(N−1))²). The “trick” is that the error depends on the step size, not on N alone. N and L both feed into it.

How someone could be tricked

  1. Confusing “order” with “N.” A question might say “the method is O(Δx²)” and ask for the order of accuracy in terms of N. The answer is not “order N.” In terms of grid points it’s O(1/N²) for fixed L, but the order of accuracy itself is still 2. The order is the exponent on Δx, nothing else.

  2. Forgetting the domain length. Two people use the same N = 101 points, one on L = 1 and one on L = 100. Same N, same “second-order” method, but Δx is 0.01 versus 1. The second error is roughly 10,000 times larger (since 100² = 10⁴). Same method, same N, wildly different accuracy. Someone who thinks “more points = more accurate” without checking L gets burned.

  3. The N vs. N − 1 trap. N points give N − 1 intervals, so Δx = L/(N−1), not L/N. It’s an easy off-by-one mistake, and it matters when N is small.

  4. Thinking a higher-order method is automatically better. An O(Δx²) scheme with a large Δx can easily be worse than an O(Δx) scheme with a tiny Δx. Order tells you how fast the error shrinks as you refine, not how small it is at a given grid.

  5. Doubling N doesn’t exactly halve Δx. Going from N to 2N gives Δx = L/(2N−1), not L/(2N−2). To cleanly halve Δx you need to go from N to 2N − 1 points. This matters if you’re doing a convergence study and expecting the error to drop by exactly 4×.


37
New cards

Accuracy

how closely a numerical solution approximates the exact solution at each step

38
New cards

Stability

describes whether numerical errors decay or amplify over successive computational steps

39
New cards

Stable

errors diminish over time

40
New cards

Unstable

errors grow

41
New cards

Central difference, second derivative

(f(xi+1)−2f(xi)+f(xi−1))(Δx)2≈d2fdx2\frac{(f(x_{i+1})-2f(x_{i})+f(x_{i-1}))}{(\Delta x)^2}\thickapprox\frac{d^2f}{dx^2}

O(Δx2)O(\Delta x^2)^{} (second order accuracy)

42
New cards

Central difference, first derivative

(f(xi+1)−f(xi−1))(2Δx)≈dfdx\frac{(f(x_{i+1})-f\left(x_{i-1})\right)}{(2\Delta x)}\thickapprox\frac{df}{dx}

O(Δx2)O(\Delta x^2)^{} (second order accuracy)

43
New cards

Forward difference, first derivative

(f(xi+1)−f(xi))Δx≈dfdx\frac{(f(x_{i+1})-f(x_{i}))}{\Delta x}\thickapprox\frac{df}{dx}

O(Δx)O(\Delta x^{})^{} (first order accuracy)

44
New cards

Backward difference, first derivative

(f(xi)−f(xi−1))Δx≈dfdx\frac{(f(x_{i})-f(x_{i-1}))}{\Delta x}\thickapprox\frac{df}{dx}

O(Δx)O(\Delta x^{})^{} (first order accuracy)

45
New cards

Halving Error (RK1 euler, RK2 midpoint, RK4)

For error of order p, halving Δx reduces the error by a factor of about 2ᵖ

RK1: error ∝Δt\varpropto\Delta t^{} → ½ t → (1/2) error (drops by factor of 2)

RK2: error ∝Δt2\varpropto\Delta t^2 → ½ t → (1/2)² = ¼ error (drops by factor of 4)

RK4: error ∝Δt4\varpropto\Delta t^4 → ½ t → (1/2)^4 = 1/16 error (drops by factor of 16)

46
New cards

Ghost Point Example (Neumann BC)

  1. Use central finite difference to approximate

  2. Rearrange equation for unknown node

\frac{dy_{ghost}}{\differentialD x}=\frac{y_{i+1}-y_{i-1}}{2\Delta x}=BC\to y_{i+1}=y_{i-1}+2BC\Delta x

  1. substitute into discretized equation