Differential-Equations · Unit 3 · Video 5 · Interactive Practice

Why the Starting Temperature Stops Mattering: Steady State and Transient in dT/dt+kT=kTedT/dt + kT = kT_e

IKey Formulas

FormulaNameWhat it gives
dTdt+kT=kTe(t),k>0\dfrac{dT}{dt} + kT = kT_e(t), \quad k > 0Conduction equation, standard linear formConstant coefficient; the right-hand side is kTekT_e, never TeT_e
u=ekdt=ektu = e^{\int k\,dt} = e^{kt}Integrating factor(ektT)=kTe(t)ekt\left(e^{kt}T\right)' = kT_e(t)\,e^{kt}
T=ekt0tkTe(t1)ekt1dt1+T0ektT = e^{-kt}\displaystyle\int_0^t kT_e(t_1)\,e^{kt_1}\,dt_1 + T_0\,e^{-kt}Definite-integral solution of T(0)=T0T(0) = T_0Every quantity is data of the problem
T0ekt0(t)T_0\,e^{-kt} \to 0 \quad (t \to \infty)The transientThe only place T0T_0 appears; the first term is the steady state

Key Insight: The initial condition enters the solution in exactly one place, the transient. A positive conductivity kills it, so every chamber ends on the same steady state — and steady means the transient is gone, not that the curve is flat.

IIFrom Integrating Factor to the Two Terms

Five moves carry dT/dt+kT=kTedT/dt + kT = kT_e to a solution split into a part that stays and a part that dies.

Step 1 — Standard linear form
dTdt=k(TeT)dTdt+kT=kTe(t)\frac{dT}{dt} = k\left(T_e - T\right) \quad\Longrightarrow\quad \frac{dT}{dt} + kT = kT_e(t)
Units force the kk on the right: dT/dtdT/dt and kTkT are both degrees per unit time, while TeT_e alone is degrees and cannot match.

IIIFour Chambers, One Curve

Chambers starting at 00, 1010, 3030 and 40°C40\degree\text{C} in a 20°C20\degree\text{C} bath: how long until they agree?

The one-percent mark sits at t=ln100/k4.6/kt = \ln 100 / k \approx 4.6/k: a better-insulated wall (smaller kk) makes the chambers remember their start proportionally longer, but never forever.

IVSteady Does Not Mean Constant

Change what the bath does: the steady state changes with it, while the same transient dies away.

VQuiz Questions

Problem 1 · The Integrating Factor

Given: the constant-coefficient equation dTdt+kT=kTe(t)\dfrac{dT}{dt} + kT = kT_e(t), with kk constant — find the integrating factor uu.

✅ Correct! With p=kp = k the exponent is kdt=kt\int k\,dt = kt, so u=ektu = e^{kt} and multiplying through gives (ektT)=kTe(t)ekt\left(e^{kt}T\right)' = kT_e(t)\,e^{kt}.
❌ Check the sign of pp. Standard form is y+py=qy' + p\,y = q, so here p=+kp = +k. Reading the coefficient off dT/dt=kT+kTedT/dt = -kT + kT_e instead is exactly the sign trap the standard form exists to prevent.
❌ Only the coefficient of TT goes in the exponent. Te(t)T_e(t) lives on the right-hand side; it is q/kq/k, not part of pp.
❌ Not quite. The integrating factor is u=epdtu = e^{\int p\,dt}, where pp is the coefficient of TT once the equation is in standard form.
Show solution

The equation is already in standard linear form T+pT=qT' + p\,T = q with

p=k(a constant),q=kTe(t)p = k \quad\text{(a constant)}, \qquad q = kT_e(t)

The integrating factor is u=epdtu = e^{\int p\,dt}, and no arbitrary constant is needed because only one uu is wanted:

u=ekdt=ektu = e^{\int k\,dt} = e^{kt}

Multiplying both sides by uu collapses the left-hand side into a single derivative:

ektdTdt+kektT=kTe(t)ekt(ektT)=kTe(t)ekte^{kt}\frac{dT}{dt} + k\,e^{kt}T = kT_e(t)\,e^{kt} \quad\Longrightarrow\quad \left(e^{kt}T\right)' = kT_e(t)\,e^{kt}

Check by the product rule: (ektT)=ektT+kektT\left(e^{kt}T\right)' = e^{kt}T' + k\,e^{kt}T, which is the left-hand side.

Problem 2 · The Right-Hand Side

Given: Newton's law of cooling, dTdt=k(TeT)\dfrac{dT}{dt} = k\left(T_e - T\right) with k>0k > 0write it in standard linear form.

✅ Correct! Expanding k(TeT)=kTekTk(T_e - T) = kT_e - kT and moving kT-kT across gives kTkT on the left and kTekT_e on the right — the kk appears on both sides, once with the unknown and once with the external value.
❌ The kk must appear on the right as well. Units settle it: dT/dtdT/dt and kTkT are degrees per unit time, so kk carries 1/time1/\text{time}; TeT_e by itself is degrees and cannot balance the equation.
❌ Both signs are flipped. That equation is dT/dt=k(TTe)dT/dt = k(T - T_e): a warmer bath would now cool the chamber, and the T0T_0 term would grow like e+kte^{+kt} instead of dying away.
❌ The coefficient of TT is kk, not 11. You cannot drop the conductivity on the left while keeping it on the right.
❌ Not quite. Multiply out the right-hand side first: k(TeT)=kTekTk\left(T_e - T\right) = kT_e - kT.
Show solution

Expand the right-hand side and collect the TT terms on the left:

dTdt=kTekTdTdt+kT=kTe\frac{dT}{dt} = kT_e - kT \quad\Longrightarrow\quad \frac{dT}{dt} + kT = kT_e

The units check. Both sides of a differential equation must carry the same units:

  • dT/dtdT/dt: degrees per unit time
  • kk: 1/time1/\text{time}, so kTkT is degrees per unit time ✅
  • TeT_e: degrees only ❌ — it cannot sit alone on the right

The same shape holds for the diffusion twin, dCdt+k1C=k1Ce\dfrac{dC}{dt} + k_1C = k_1C_e.

Problem 3 · A Chamber in a Fixed Bath

Given: a bath held at Te=20°CT_e = 20\degree\text{C}, conductivity k=0.5k = 0.5 per hour, and T(0)=60°CT(0) = 60\degree\text{C}find the temperature and evaluate it after two hours.

What is T(t)T(t)?

What is TT after 22 hours?

✅ Correct! Steady state 20(1e0.5t)20\left(1 - e^{-0.5t}\right) plus transient 60e0.5t60e^{-0.5t} collects to 20+40e0.5t20 + 40e^{-0.5t}, and at t=2t = 2 the exponent is 1-1: 20+40(0.3679)=34.7°C20 + 40(0.3679) = 34.7\degree\text{C}.
❌ Build it as steady state plus transient. The steady-state term is 20(1e0.5t)20\left(1 - e^{-0.5t}\right) and the transient is T0e0.5t=60e0.5tT_0e^{-0.5t} = 60e^{-0.5t}; then collect the exponentials.
❌ That is the bath value plus the transient. The steady-state term is 20(1e0.5t)20\left(1 - e^{-0.5t}\right), not 2020 — its own 20e0.5t-20e^{-0.5t} must be collected with the transient, leaving 20+40e0.5t20 + 40e^{-0.5t}. Check at t=0t = 0: your formula gives 80°C80\degree\text{C}, not 6060.
❌ Check the exponent. It is kt=0.5(2)=1-kt = -0.5(2) = -1, so the factor is e1=0.3679e^{-1} = 0.3679, not e2e^{-2}.
❌ The 6060 and the 2020 have swapped roles. The bath value 2020 is the level the solution approaches; 40=T0Te40 = T_0 - T_e is the initial gap that decays.
Show solution

Step 1 — the two terms. For a constant bath the steady-state integral can be done:

ekt0tkTeekt1dt1=Teekt[ekt1]0t=Te(1ekt)=20(1e0.5t)e^{-kt}\int_0^t kT_e\,e^{kt_1}\,dt_1 = T_e\,e^{-kt}\left[e^{kt_1}\right]_0^t = T_e\left(1 - e^{-kt}\right) = 20\left(1 - e^{-0.5t}\right)

The transient is T0ekt=60e0.5tT_0e^{-kt} = 60e^{-0.5t}.

Step 2 — add and collect.

T=2020e0.5t+60e0.5t=20+40e0.5tT = 20 - 20e^{-0.5t} + 60e^{-0.5t} = 20 + 40e^{-0.5t}

Check the start: T(0)=20+40=60°CT(0) = 20 + 40 = 60\degree\text{C} ✅, and T20°CT \to 20\degree\text{C} as tt \to \infty because k>0k > 0.

Step 3 — evaluate at t=2t = 2. The exponent is 0.5(2)=1-0.5(2) = -1:

T(2)=20+40e1=20+40(0.36788)=34.7°CT(2) = 20 + 40e^{-1} = 20 + 40(0.36788) = 34.7\degree\text{C}

Note the regrouping: 20+40e0.5t20 + 40e^{-0.5t} is not the steady-state/transient split. The steady state is 2020e0.5t20 - 20e^{-0.5t} and the transient is 60e0.5t60e^{-0.5t}; only the transient carries T0T_0.

Problem 4 · A Bath That Never Settles

Given: k=1k = 1 per hour and a swinging bath Te(t)=20+8sintT_e(t) = 20 + 8\sin t, so that T=20+4(sintcost)16et+T0etT = 20 + 4\left(\sin t - \cos t\right) - 16e^{-t} + T_0e^{-t}.

Which part of the solution carries T0T_0?

What happens as tt \to \infty?

✅ Correct! T0T_0 multiplies ete^{-t} and nothing else, and k>0k > 0 kills it. What survives is 20+42sin ⁣(tπ4)20 + 4\sqrt{2}\sin\!\left(t - \tfrac{\pi}{4}\right) — steady, because the transient is gone, but far from constant.
❌ Look at where T0T_0 appears. The steady-state term ekt0tkTe(t1)ekt1dt1e^{-kt}\int_0^t kT_e(t_1)e^{kt_1}dt_1 is built from the bath alone; the initial condition enters only through C=T0C = T_0, multiplying ekte^{-kt}.
❌ Drop only what decays. Both 16et-16e^{-t} and T0etT_0e^{-t} vanish, leaving 20+4sint4cost20 + 4\sin t - 4\cos t; now combine that into a single sine.
❌ Steady state does not mean constant. Only the exponentials die. The oscillation the bath drives is part of the steady-state term and lasts forever.
❌ The chamber cannot follow the whole swing. Conduction through the wall takes time: 4sint4cost4\sin t - 4\cos t has amplitude 42+42=425.66\sqrt{4^2 + 4^2} = 4\sqrt{2} \approx 5.66, not 88, and it peaks after the bath does.
Show solution

Where T0T_0 lives. The definite-integral solution is

T=ekt0tkTe(t1)ekt1dt1steady state — the bath’s doing+T0ekttransient — the start’s doingT = \underbrace{e^{-kt}\int_0^t kT_e(t_1)\,e^{kt_1}\,dt_1}_{\text{steady state — the bath's doing}} + \underbrace{T_0\,e^{-kt}}_{\text{transient — the start's doing}}

The first term is the solution that starts at 0°C0\degree\text{C}; the second is the only place the initial condition appears. Here k=1>0k = 1 > 0, so T0et0T_0e^{-t} \to 0.

The long run. Dropping every ete^{-t} term leaves

T20+4sint4costT \longrightarrow 20 + 4\sin t - 4\cos t

Write 4sint4cost=Rsin(tφ)4\sin t - 4\cos t = R\sin(t - \varphi) with R=42+42=425.657R = \sqrt{4^2 + 4^2} = 4\sqrt{2} \approx 5.657 and tanφ=4/4=1\tan\varphi = 4/4 = 1, so φ=π/4\varphi = \pi/4:

T20+42sin ⁣(tπ4)T \longrightarrow 20 + 4\sqrt{2}\,\sin\!\left(t - \frac{\pi}{4}\right)

The chamber oscillates about the bath's mean with a smaller amplitude and a phase lag — a steady state that is never constant.

Solved: 0 / 4