← Back to Math 055

7.3 Operational Properties I: Class Notes

7.3a

Recall

ℒ{f(t)}=∫0∞e−stf(t)dt=F(s)

Theorem: First Translation Theorem

If ℒ{f(t)}=F(s) and a is any real number, then ℒ{eatf(t)}=F(s−a).

(This is also called a shifting theorem.)

Proof

By Definition 7.1.1,

ℒ{eatf(t)}=∫0∞e−steatf(t)dt=∫0∞e−(s−a)tf(t)dt=F(s−a)

−(s−a) is playing the role of −s.

Notation: ℒ{eatf(t)}=ℒ{f(t)}s→s−a

We also get an inverse rule: ℒ−1{F(s−a)}=eatf(t)

Example

Find either F(s) or f(t), as indicated.

ℒ{t10e−7t}

Solution

t10 is what we’re transforming, and e−7t shifts s to s+7. Since ℒ{t10}=10!s11,

Answer: F(s)=10!(s+7)11

ℒ{e−2tcos⁡4t}

Solution

Transform cos⁡4t, then shift s to s+2:

Answer: F(s)=s+2(s+2)2+16

ℒ−1{1s2+2s+5}

Solution

Complete the square: s2+2s+5=(s+1)2+4. The s+1 is a shift, and 4=k2 gives k=2 and sine (no s in the numerator). Multiply and divide by 2:

ℒ−1{1s2+2s+5}=ℒ−1{12·1·2(s+1)2+4}

Answer: f(t)=12e−tsin⁡2t

Tip: Recall from partial fractions: an irreducible quadratic factor gets a linear numerator, as in Ax+Bx2+4.

ℒ−1{(s+1)2(s+2)4}

Solution

Expand the numerator. The denominator is a repeated linear factor, so the partial fractions are

s2+2s+1(s+2)4=As+2+B(s+2)2+C(s+2)3+D(s+2)4

s2+2s+1=A(s+2)3+B(s+2)2+C(s+2)+D

Compare coefficients and substitute s=−2:

s3:A=0s=−2:1=Ds2:1=Bconst:1=8A+4B+2C+D=4+2C+1⇒C=−2

Each term is a power of s+2, a shift: use part (b) with n=1, n=2, and n=3. The middle numerator is already 2!=2, and the last needs 3!=6, so multiply and divide by 6:

ℒ−1{1(s+2)2−2(s+2)3+16·1·3!(s+2)4}=te−2t−t2e−2t+16t3e−2t=e−2t(t−t2+16t3)

Answer: f(t)=e−2t(t−t2+16t3)

7.3b

Example

Use the Laplace transform to solve the initial-value problem.

y′−y=1+tet,y(0)=0

Solution

Recall

ℒ{1}=1s,ℒ{t}=1s2ℒ{y′(t)}=sY(s)−y(0),ℒ{y(t)}=Y(s)

In tet, f(t)=t and et is a shift, so ℒ{tet}=1(s−1)2. With y(0)=0:

sY(s)−Y(s)=ℒ{1}+ℒ{tet}=1s+1(s−1)2Y(s)=1s(s−1)+1(s−1)3

Partial fractions: 1s(s−1)=As+Bs−1, so 1=A(s−1)+Bs, giving A=−1, B=1. The last term needs 2!=2 in the numerator (n=2):

Y(s)=−1s+1s−1+12·1·2(s−1)3

Answer: y(t)=−1+et+12t2et

2y″+20y′+51y=0,y(0)=2,y′(0)=0

Solution

Transform each term (colors match the equation); y′(0)=0 and y(0)=2:

2[s2Y(s)−sy(0)−y′(0)]+20[sY(s)−y(0)]+51Y(s)=02s2Y(s)−4s+20sY(s)−40+51Y(s)=0(2s2+20s+51)Y(s)=4s+40Y(s)=4s+402s2+20s+51

Multiply the numerator and denominator by 12, then complete the square:

s2+10s+512=s2+10s+25+12=(s+5)2+12

Y(s)=2s+20s2+10s+25+12=2s+20(s+5)2+12

The shift s+5 has to show up in the numerator too: 2s+20=2(s+5)+10. For sine and cosine, k2=12, so k=12; the sine term needs 12 in its numerator, and 10=102·12:

Y(s)=2·s+5(s+5)2+12+102·1/2(s+5)2+12

Answer: y(t)=e−5t(2cos⁡12t+102sin⁡12t)

Example

Solve the boundary-value problem.

y″+8y′+20y=0,y(0)=0,y′(π)=0

Solution

The value y′(0) is not given, so call it c1=y′(0). The terms with y(0)=0 drop out:

s2Y(s)−sy(0)−y′(0)+8sY(s)−8y(0)+20Y(s)=0(s2+8s+20)Y(s)−c1=0Y(s)=c1s2+8s+16+4=12·2c1(s+4)2+4

Here k=2 (sine) with the shift s+4. Writing c=c1/2 for the constant,

y(t)=ce−4tsin⁡2t

Now use y′(π)=0:

y′(t)=−4ce−4tsin⁡2t+2ce−4tcos⁡2ty′(π)=2ce−4π=0⇒c=0

(At t=π, sin⁡2π=0 and cos⁡2π=1.)

Answer: y=0

7.3c

The First Translation Theorem showed how to translate on the s-axis. A translation on the t-axis is essentially an adjustment for moving the lower limit in the definition of a Laplace transform. This is accomplished via the Heaviside function or unit step function.

Definition

The unit step function is

𝒰(t−a)={0,0≤t<a1,t≥a

Unit step function U(t minus a): 0 for t less than a, then 1 for t at least a

Graph of the unit step function 𝒰(t−a). It is 0 (“off”) for 0≤t<a, with an open circle at (a,0); it is 1 (“on”) for t≥a, with a filled dot at (a,1).

To make it first on and then off:

1−𝒰(t−a)={1−0,0≤t<a1−1,t≥a={1,0≤t<a0,t≥a

Consider

sin⁡t𝒰(t−π/2)={0,0≤t<π/2sin⁡t,t≥π/2

Graph of sin t times U(t minus pi/2): zero until pi/2, then the sine curve from height 1 on

Graph of sin⁡t𝒰(t−π/2): 0 for 0≤t<π/2 (open circle at (π/2,0)), then the sine curve from (π/2,1) onward, crossing 0 at t=π and reaching −1 at t=3π/2. The sine wave is cut off, not shifted.

Shifting the sine yields

sin⁡(t−π/2)𝒰(t−π/2)={0,0≤t<π/2sin⁡(t−π/2),t≥π/2

Graph of sin(t minus pi/2) U(t minus pi/2): zero, then the sine wave shifted right by pi/2

Graph of sin⁡(t−π/2)𝒰(t−π/2): 0 for 0≤t<π/2, then a full sine wave that starts at (π/2,0), rises to 1 at t=π, and returns to 0 at t=3π/2. It is the graph of sin⁡t moved π/2 units to the right.

Theorem: Second Translation Theorem

If F(s)=ℒ{f(t)} and a>0, then ℒ{f(t−a)𝒰(t−a)}=e−asF(s).

Watch the signs: compare with ℒ{eat}=1s−a. Here the shift t−a comes with e−as.

Proof

ℒ{f(t−a)𝒰(t−a)}=∫0∞e−stf(t−a)𝒰(t−a)dt=∫0ae−stf(t−a)·0dt+∫a∞e−stf(t−a)dt

The first integral is 0. In the second, let v=t−a; then dv=dt and t=v+a. The limits change: t=a becomes v=0, and t=∞ stays v=∞.

=∫0∞e−s(v+a)f(v)dv=e−as∫0∞e−svf(v)dv=e−asF(s)

This is the shift: the lower limit moved from a to 0.

If we let f(t)=1, then we have ℒ{𝒰(t−a)}=e−ass.

In inverse form, ℒ−1{e−ass}=𝒰(t−a).

Example

Find either F(s) or f(t), as indicated.

ℒ{(3t+1)𝒰(t−1)}

Solution

The theorem needs the form ℒ{f(t−a)𝒰(t−a)}, but 3t+1 is not written in terms of t−1: these don’t match. Call g(t)=3t+1 and use the alternative form below.

Alternative form of the second translation theorem

ℒ{g(t)𝒰(t−a)}=e−asℒ{g(t+a)}

Why it works. Start from the original theorem:

ℒ{f(t−a)𝒰(t−a)}=e−asℒ{f(t)}

Let g(t)=f(t−a); then f(t)=g(t+a). So

ℒ{g(t)𝒰(t−a)}=ℒ{f(t−a)𝒰(t−a)}=e−asℒ{f(t)}=e−asℒ{g(t+a)}

Back to the example, with g(t)=3t+1 and a=1: replace t by t+1.

ℒ{(3t+1)𝒰(t−1)}=e−sℒ{3(t+1)+1}=e−sℒ{3t+4}=e−s(3s2+4s)

Answer: F(s)=e−s(3s2+4s)

7.3d

Inverse form of the Second Translation Theorem

If f(t)=ℒ−1{F(s)}, then ℒ−1{e−asF(s)}=f(t−a)𝒰(t−a).

You have to make the t−a happen: replace t by t−a everywhere in f(t).

Example

Find either F(s) or f(t), as indicated.

ℒ−1{(1+e−2s)2s+2}

Solution

Expand the square and split into three inverse transforms. Each uses f(t)=ℒ−1{1s+2}=e−2t, shifted by a=2 or a=4:

ℒ−1{(1+e−2s)2s+2}=ℒ−1{1+2e−2s+e−4ss+2}=ℒ−1{1s+2}+ℒ−1{2e−2ss+2}+ℒ−1{e−4ss+2}=e−2t+2e−2(t−2)𝒰(t−2)+e−2(t−4)𝒰(t−4)

Answer: f(t)=e−2t+2e−2(t−2)𝒰(t−2)+e−2(t−4)𝒰(t−4)

ℒ−1{e−2ss2(s−1)}

Solution

Set the e−2s aside for now and find partial fractions for the rest:

1s2(s−1)=As+Bs2+Cs−1

Tip: Or use As+Bs2+Cs−1; it gives the same result.

1=As(s−1)+B(s−1)+Cs2

s2:A+C=0s:−A+B=0const:−B=1

So B=−1, then A=B=−1, and C=−A=1. Bring back e−2s and use the inverse form with a=2:

ℒ−1{e−2s(−1s−1s2+1s−1)}=𝒰(t−2)[−1−(t−2)+et−2]

Answer: f(t)=𝒰(t−2)[−1−(t−2)+et−2]

Example

Use the Laplace transform to solve the given initial-value problem.

y′+y=f(t),   y(0)=0, where

f(t)={1,0≤t<1−1,t≥1

Solution

Graph of f(t): 1 for t from 0 to 1, then negative 1 for t at least 1

Graph of f(t): a horizontal segment at height 1 for 0≤t<1 (open circle at (1,1)), then a horizontal ray at height −1 for t≥1 (filled dot at (1,−1)).

First express f(t) using the unit step function. Write it as a sum of two pieces:

f(t)={1,0≤t<10,t≥1+{0,0≤t<1−1,t≥1

f(t)=1−𝒰(t−1)−𝒰(t−1)=1−2𝒰(t−1)

Check: f(t)={1−0,0≤t<11−2,t≥1 ✓

Now solve y′+y=1−2𝒰(t−1) with y(0)=0:

sY(s)−y(0)+Y(s)=1s−2e−ssY(s)=1s(s+1)−2e−ss(s+1)

Partial fractions: 1s(s+1)=As+Bs+1, so 1=A(s+1)+Bs, giving A=1, B=−1.

Y(s)=1s−1s+1−2e−s(1s−1s+1)

For the last term, use the inverse form with a=1: replace t by t−1 in 1−e−t.

Answer: y(t)=1−e−t−2𝒰(t−1)[1−e−(t−1)]