Home / Courses / CSIR NET Physics / Practice / Fourier and Laplace transforms

Fourier and Laplace transforms

Mathematical methods · 70 questions and templates for CSIR NET Physics, in order from warm-up to top rank. Check each answer, then read the full solution.

Loading questions… (they need JavaScript; nothing you type is sent anywhere)

Written questions with solutions

The same questions as a list, so you can print them or read without the practice box. Templates (fresh numbers) appear only in the practice box above.

  1. MCQWarm-up

    The Laplace transform of eat (for s>a) is
    1. 1s−a
    2. 1s+a
    3. as2+a2
    4. ss−a
    Answer and solution

    Answer: A

    ∫0∞eate−stdt=∫0∞e−(s−a)tdt=1s−a, which converges only when s>a.

    Why people get this wrong: The sign: e−at gives 1/(s+a).

  2. NATWarm-up

    If F(s) is the Laplace transform of t2, the value of F(2) is ______.
    Answer and solution

    Answer: 0.25 (±0.001)

    ℒ{tn}=n!sn+1, so ℒ{t2}=2s3 and F(2)=28=0.25.

    Why people get this wrong: Forgetting the factorial gives 1/8.

  3. MCQWarm-up

    The Laplace transform of sin⁡ωt is
    1. ss2+ω2
    2. ωs2+ω2
    3. ωs2−ω2
    4. 1s2+ω2
    Answer and solution

    Answer: B

    sin⁡ωt=Imeiωt and ℒ{eiωt}=1s−iω=s+iωs2+ω2. Imaginary part: ωs2+ω2. (The real part, ss2+ω2, is for cos⁡ωt.)

    Why people get this wrong: Mixing up sine and cosine: sine has ω on top, so F(s)→0 like ω/s2, matching sin⁡≈ωt near t=0.

  4. MCQWarm-up

    With f~(k)=∫−∞∞f(x)e−ikxdx, the Fourier transform of δ(x−a) is
    1. 1
    2. eika
    3. e−ika
    4. δ(k−a)
    Answer and solution

    Answer: C

    The delta function picks out the integrand at x=a: ∫δ(x−a)e−ikxdx=e−ika. A spike in x contains every k with equal strength; only the phase depends on where the spike is.

    Why people get this wrong: The sign of the exponent follows the convention; with e+ikx in the definition you would get eika.

  5. MCQWarm-up

    The Fourier transform of a Gaussian e−x2/(2σ2) is
    1. a Lorentzian
    2. a sinc function
    3. a decaying exponential
    4. a Gaussian
    Answer and solution

    Answer: D

    Completing the square in the exponent gives f~(k)=2πσe−σ2k2/2 (with f~=∫fe−ikxdx): again a Gaussian, of width 1/σ.

    Why people get this wrong: The Lorentzian is the transform of e−a|x|; the sinc is the transform of a rectangle.

  6. NATWarm-up

    The Laplace transform of te−2t, evaluated at s=1, is ______ (to three decimal places).
    Answer and solution

    Answer: 0.111 (±0.002)

    Shift rule: ℒ{e−2tf(t)}=F(s+2). With F=ℒ{t}=1/s2: 1(s+2)2. At s=1: 19≈0.111.

    Why people get this wrong: The exponential shifts s to s+2 (not s−2) for a decaying factor.

  7. NATStandard

    The rectangular pulse f(x)=1 for |x|<2 and 0 otherwise is transformed. Use f~(k)=∫−∞∞f(x)e−ikxdx. The smallest positive k at which f~(k)=0 is ______ (to two decimal places).
    Answer and solution

    Answer: 1.57 (±0.01)

    f~(k)=∫−22e−ikxdx=2sin⁡2kk. It first vanishes where 2k=π: k=π/2≈1.57. A wider pulse has its first zero nearer the origin: narrower spectrum.

    Why people get this wrong: Using the full width 4 instead of the half-width 2 gives π/4.

  8. MCQStandard

    With f~(k)=∫f(x)e−ikxdx, the transform of the shifted function f(x−a) is
    1. eikaf~(k)
    2. e−ikaf~(k)
    3. f~(k−a)
    4. f~(k)
    Answer and solution

    Answer: B

    Substitute u=x−a: ∫f(u)e−ik(u+a)du=e−ikaf~(k). Shifting in x only changes the phase; |f~| is unchanged.

    Why people get this wrong: f~(k−a) is the transform of eiaxf(x): a shift in k, not in x.

  9. MSQStandard

    With F=ℒ{f}, G=ℒ{g} and u the unit step, which rules are correct?
    1. ℒ{tf(t)}=−F′(s)
    2. ℒ{f(t)g(t)}=F(s)G(s)
    3. ℒ{eatf(t)}=F(s−a)
    4. ℒ{f(t−a)u(t−a)}=e−asF(s) for a>0
    Answer and solution

    Answer: A, C, D

    Differentiating ∫fe−stdt with respect to s brings down −t. Multiplying by eat replaces s by s−a. Delaying by a multiplies by e−as. A product of transforms corresponds to the convolution ∫0tf(τ)g(t−τ)dτ, not to the product fg.

    Why people get this wrong: Products in one domain are convolutions in the other.

  10. NATStandard

    If ℒ{f}=1s(s+1), the value of f(2) is ______ (to three decimal places).
    Answer and solution

    Answer: 0.865 (±0.003)

    Partial fractions: 1s(s+1)=1s−1s+1, so f=1−e−t and f(2)=1−e−2≈0.865. (It is the step response of a first-order system.)

    Why people get this wrong: Getting the sign of the second fraction wrong gives 1+e−t, which does not start at 0.

  11. MCQStandard

    With f~(k)=∫f(x)e−ikxdx, the transform of e−a|x| (a>0) is
    1. 1a+ik
    2. aa2+k2
    3. 2aa2+k2
    4. 2a2+k2
    Answer and solution

    Answer: C

    Split at 0: ∫0∞e−(a+ik)xdx+∫−∞0e(a−ik)xdx=1a+ik+1a−ik=2aa2+k2, a Lorentzian. Check: f~(0)=∫e−a|x|dx=2/a ✓.

    Why people get this wrong: 1a+ik is only the x>0 half (the transform of e−axu(x)).

  12. NATStandard

    The Laplace transform of cosh⁡2t, evaluated at s=3, is ______ (to two decimal places).
    Answer and solution

    Answer: 0.6 (±0.01)

    cosh⁡2t=12(e2t+e−2t), so ℒ=12(1s−2+1s+2)=ss2−4. At s=3: 35=0.6.

    Why people get this wrong: Using s2+4 (the cosine formula) gives 3/13.

  13. MCQStandard

    For y′+2y=0 with y(0)=3, the Laplace transform Y(s) is
    1. 3s−2
    2. 3ss+2
    3. 1s+2
    4. 3s+2
    Answer and solution

    Answer: D

    ℒ{y′}=sY−y(0), so sY−3+2Y=0 and Y=3s+2, i.e. y=3e−2t.

    Why people get this wrong: Dropping the initial value y(0) from ℒ{y′} loses the 3.

  14. MCQStandard

    With f~(k)=∫f(x)e−ikxdx and f→0 at ±∞, the transform of f′(x) is
    1. ikf~(k)
    2. −ikf~(k)
    3. f~′(k)
    4. kf~(k)
    Answer and solution

    Answer: A

    Integrate by parts: ∫f′e−ikxdx=[fe−ikx]+ik∫fe−ikxdx=ikf~(k). Derivatives become multiplication by ik, which is why the Fourier transform turns linear ODEs with constant coefficients into algebra (and why pˆ=−iℏd/dx is ℏk in momentum space).

    Why people get this wrong: The sign follows the convention: with e+ikx it would be −ik.

  15. MSQStandard

    With f~(k)=∫f(x)e−ikxdx, which statements are true?
    1. For every complex f, |f~(k)| is an even function of k
    2. If f is real and even, f~ is real and even
    3. If f is real and odd, f~ is purely imaginary and odd
    4. If f is real, f~(−k)=f~(k)∗
    Answer and solution

    Answer: B, C, D

    For real f, conjugating ∫fe−ikxdx gives ∫feikxdx=f~(−k). Even f: only cos⁡kx survives, real and even. Odd f: only −isin⁡kx survives, imaginary and odd. But a complex f such as e−x2eik0x has a spectrum centred at k0, not symmetric.

    Why people get this wrong: The symmetry |f~(−k)|=|f~(k)| needs f real.

  16. NATStandard

    If ℒ{f}=2s+3s(s+1)(s+3), then limt→∞⁡f(t) is ______.
    Answer and solution

    Answer: 1

    The final value theorem limt→∞⁡f=lims→0⁡sF(s) applies because the other poles (−1, −3) are in the left half-plane. sF=2s+3(s+1)(s+3)→33=1.

    Why people get this wrong: The theorem fails if sF has poles on the imaginary axis or to the right (for example for sin⁡t).

  17. NATExam level

    For f(x)=e−|x|, use f~(k)=∫−∞∞f(x)e−ikxdx. The value of ∫−∞∞|f~(k)|2dk is ______ (to two decimal places).
    Answer and solution

    Answer: 6.28 (±0.01)

    Parseval (Plancherel) in this convention: ∫|f~|2dk=2π∫|f|2dx. Here ∫e−2|x|dx=1, so the answer is 2π≈6.28. (Directly: ∫4(1+k2)2dk=4⋅π2=2π ✓.)

    Why people get this wrong: Forgetting the 2π that comes with the non-symmetric convention.

  18. NATExam level

    Let r(x)=1 for |x|<1 and 0 otherwise. The convolution (r∗r)(x)=∫r(y)r(x−y)dy at x=0.5 is ______ (to two decimal places).
    Answer and solution

    Answer: 1.5 (±0.01)

    The integrand is 1 where both |y|<1 and |x−y|<1, i.e. on the overlap of (−1,1) and (x−1,x+1), of length 2−|x| for |x|<2. At x=0.5: 1.5. The convolution of two boxes is a triangle; its transform is the square of the box's transform, (2sin⁡k/k)2.

    Why people get this wrong: Convolution is the overlap area as one box slides past the other, not the product of the boxes.

  19. MCQExam level

    A Gaussian f(x)∝e−x2/(2σ2) has transform f~(k)∝e−k2/(2σk2). Then
    1. σk=σ
    2. σk=1/σ
    3. σk=σ2
    4. σk=2π/σ
    Answer and solution

    Answer: B

    Completing the square: ∫e−x2/(2σ2)e−ikxdx=2πσe−σ2k2/2, so σk=1/σ and σσk=1. The probability densities |f|2 and |f~|2 have widths σ/2 and 1/(2σ), whose product 12 gives ΔxΔp=ℏ/2.

    Why people get this wrong: Squeezing a function in x spreads its transform in k; they never shrink together.

  20. NATExam level

    Solve y′′+3y′+2y=0 with y(0)=1, y′(0)=0 by Laplace transform. The value of y(1) is ______ (to three decimal places).
    Answer and solution

    Answer: 0.6 (±0.003)

    ℒ{y′′}=s2Y−sy(0)−y′(0)=s2Y−s and ℒ{y′}=sY−1. So (s2+3s+2)Y=s+3 and Y=s+3(s+1)(s+2)=2s+1−1s+2. Hence y=2e−t−e−2t and y(1)=0.7358−0.1353≈0.600. Check: y(0)=1, y′(0)=−2+2=0 ✓.

    Why people get this wrong: The 3y′ term also carries −3y(0); missing it gives Y=s/((s+1)(s+2)).

  21. MSQExam level

    Which of these functions have a Laplace transform (for all large enough s)?
    1. et2
    2. t−1/2
    3. e5tsin⁡t
    4. 1/t
    Answer and solution

    Answer: B, C

    We need ∫0∞|f|e−stdt to converge for large s. et2 outgrows every est. t−1/2 is integrable at 0: ℒ=Γ(12)/s1/2=π/s. e5tsin⁡t is of exponential order: 1(s−5)2+1 for s>5. 1/t diverges logarithmically at t=0.

    Why people get this wrong: Two ways to fail: growing too fast at infinity, or blowing up too strongly at t=0.

  22. MCQExam level

    With f(x)=12π∫−∞∞f~(k)eikxdk, the function whose transform is f~(k)=11+k2 is
    1. 12e−|x|
    2. e−|x|
    3. πe−|x|
    4. e−x2
    Answer and solution

    Answer: A

    From the pair e−a|x|↔2aa2+k2 with a=1: e−|x|↔21+k2, so 11+k2 comes from 12e−|x|. (By residues: for x>0 close up round k=i: 12π⋅2πi⋅e−x2i=12e−x.)

    Why people get this wrong: Dropping the factor 2 in the pair, or the 1/2π in the inverse.

  23. NATExam level

    The Laplace transform of tsin⁡t, evaluated at s=1, is ______ (to two decimal places).
    Answer and solution

    Answer: 0.5 (±0.01)

    ℒ{tf}=−F′(s) with F=1s2+1: −F′=2s(s2+1)2. At s=1: 24=0.5.

    Why people get this wrong: Forgetting the minus sign in −F′ gives −0.5; a transform of a function that is positive near 0 should be positive for large s.

  24. MCQExam level

    If ℒ{f}=F(s), then ℒ{∫0tf(τ)dτ} is
    1. sF(s)
    2. F(s)−f(0)
    3. F(s)/s
    4. −F′(s)
    Answer and solution

    Answer: C

    Let g=∫0tf, so g′=f and g(0)=0. Then sG−0=F, giving G=F/s. (Equivalently, g=f∗1 and ℒ{1}=1/s.)

    Why people get this wrong: Integration divides by s; differentiation multiplies.

  25. NATExam level

    If ℒ{f}=1s2(s+1), the value of f(1) is ______ (to three decimal places).
    Answer and solution

    Answer: 0.368 (±0.003)

    Partial fractions: 1s2(s+1)=1s2−1s+1s+1, so f=t−1+e−t and f(1)=e−1≈0.368. Check: f(0)=0 and f′(0)=0, as the 1/s3 fall-off of F demands.

    Why people get this wrong: Missing the −1/s term; the coefficients must add to give no 1/s behaviour overall at large s.

  26. MCQExam level

    With f~(k)=∫f(x)e−ikxdx and a>0, the transform of f(x)=sin⁡axπx is
    1. 2sin⁡kak
    2. e−a|k|
    3. a Gaussian of width a
    4. 1 for |k|<a and 0 for |k|>a
    Answer and solution

    Answer: D

    Duality: the box 1 for |k|<a has inverse transform 12π∫−aaeikxdk=sin⁡axπx. So the sinc function transforms into the box. (This is the ideal low-pass filter.)

    Why people get this wrong: 2sin⁡kak is the transform of the box itself, going the other way.

  27. MSQExam level

    With f~(k)=∫f(x)e−ikxdx, which statements are true?
    1. ∫−∞∞eikxdk=2πδ(x)
    2. The transform of f(x)=1 is 2πδ(k)
    3. The transform of a periodic function is a smooth function of k
    4. The transform of cos⁡k0x is π[δ(k−k0)+δ(k+k0)]
    Answer and solution

    Answer: A, B, D

    Transform-then-invert of δ(x): δ(x)=12π∫eikxdk. The constant 1 is the inverse of 2πδ(k). cos⁡k0x=12(eik0x+e−ik0x) gives two spikes of weight π. A periodic function contains only the harmonic wavenumbers, so its transform is a comb of delta functions, not a smooth function.

    Why people get this wrong: Missing the 2π: it appears wherever the convention puts it, so track it.

  28. NATExam level

    The Laplace transform of (t−2)u(t−2) (u the unit step), evaluated at s=1, is ______ (to three decimal places).
    Answer and solution

    Answer: 0.135 (±0.002)

    Second shift theorem: ℒ{f(t−a)u(t−a)}=e−asF(s) with f=t, F=1/s2. So e−2ss2; at s=1: e−2≈0.135.

    Why people get this wrong: (t−2)u(t−2) is a ramp starting at t=2, not tu(t−2); the latter would add 2e−2s/s.

  29. NATExam level

    Let y(t)=∫0tsin⁡τsin⁡(t−τ)dτ. Using Laplace transforms, the value of y(π) is ______ (to three decimal places).
    Answer and solution

    Answer: 1.571 (±0.002)

    y=sin⁡∗sin⁡, so Y(s)=1(s2+1)2. The inverse is y(t)=12(sin⁡t−tcos⁡t) (check: it vanishes with its first derivative at 0, and y′′+y=sin⁡t, the equation whose transform gives Y). At t=π: 12(0−π⋅(−1))=π/2≈1.571.

    Why people get this wrong: Multiplying the functions instead of convolving, or taking the inverse of 1/(s2+1)2 as tsin⁡t (that one is 2s/(s2+1)2).

  30. MCQExam level

    With f~(k)=∫−∞∞f(x)e−ikxdx, the transform of xe−x2/2 is
    1. ik2πe−k2/2
    2. k2πe−k2/2
    3. −ik2πe−k2/2
    4. −k22πe−k2/2
    Answer and solution

    Answer: C

    The transform of g=e−x2/2 is 2πe−k2/2. With this sign convention, dg~dk=∫(−ix)ge−ikxdx, so the transform of xg is idg~/dk=i2π(−k)e−k2/2=−ik2πe−k2/2. Check: xe−x2/2 is real and odd, so its transform must be purely imaginary and odd, which only the first and third options are; the sign fixes the third.

    Why people get this wrong: The sign depends on the convention e−ikx versus e+ikx; with e+ikx the answer would be +ik2πe−k2/2.

← All topics · Timed mock test →