The Heat Equation — Question 5

PDF ↗

Question 5

A homogeneous rod with insulated ends is observed to have temperature u(x,t)=Ta+Be−λtcos⁡(πx/L),0≤x≤L,B,λ,L>0.u(x,t)=T_a+B e^{-\lambda t}\cos(\pi x/L),\qquad 0\leq x\leq L,\quad B,\lambda,L>0. The unknown constant conductivity and volumetric heat capacity satisfy Cut=kuxxC u_t=k u_{xx} and k,C>0k,C>0. Assume B,L,λB,L,\lambda are known.

Tasks

  1. Verify the boundary conditions and infer the diffusivity α=k/C\alpha=k/C from the observed field.

  2. Prove that temperature observations of this field alone do not identify kk and CC separately. Describe the full family of positive parameter pairs.

  3. An independent measurement gives rightward heat flux per area qm>0q_m>0 at x=L/2,t=0x=L/2,t=0. Recover kk and CC uniquely and check their units.

  4. Explain why the midpoint is useful for this flux measurement but useless for measuring the temperature decay. Identify which temperature sensor locations can recover λ\lambda from two distinct measurement times.

Original worksheet page 1: question and worked solution for 9-1-005
Show solutionHide solution

Question 5 – Solution

Strategy. The temperature equation determines a ratio of material parameters; a flux measurement provides an independent scale.

Step 1: Verify the field and recover diffusivity. Differentiation gives ut=−λBe−λtcos⁡(πx/L),uxx=−π2L2Be−λtcos⁡(πx/L).u_t=-\lambda B e^{-\lambda t}\cos(\pi x/L),\qquad u_{xx}=-\frac{\pi^2}{L^2}B e^{-\lambda t}\cos(\pi x/L). Also ux=−(πB/L)e−λtsin⁡(πx/L)u_x=-(\pi B/L)e^{-\lambda t}\sin(\pi x/L), so ux=0u_x=0 at both ends. Matching the equation at points where the cosine is nonzero gives α=kC=λL2π2.\boxed{\alpha=\frac{k}{C}=\frac{\lambda L^2}{\pi^2}.} The same equality then verifies the equation everywhere, including the node.

Step 2: Identify the nonuniqueness. For every choice C*>0C_*>0, the pair C=C*,k=λL2π2C*\boxed{C=C_*,\qquad k=\frac{\lambda L^2}{\pi^2}C_*} produces exactly the observed field. Multiplying both parameters by the same positive factor leaves the equation and insulated boundary conditions unchanged. More exact observations of this same temperature field cannot resolve that scaling ambiguity.

Step 3: Use the independent flux data. At the midpoint initially, q=−kux=kBπ/Lq=-ku_x=kB\pi/L. Therefore k=qmLBπ,C=kα=qmπBλL.\boxed{k=\frac{q_mL}{B\pi},\qquad C=\frac{k}{\alpha}=\frac{q_m\pi}{B\lambda L}.} With qmq_m in W/m2\mathrm{W/m^2}, BB in kelvins and λ\lambda in s−1\mathrm{s^{-1}}, these have units W/(mK)\mathrm{W/(mK)} and J/(m3K)\mathrm{J/(m^3K)}, respectively. Positivity of the measured quantities ensures physically positive parameters.

Step 4: Choose informative sensor locations. At x=L/2x=L/2, u=Tau=T_a for all time: both utu_t and uxxu_{xx} vanish, so a ratio of these derivatives would be 0/00/0. But |ux||u_x| is largest there, giving a strong flux signal. At any fixed x≠L/2x\ne L/2 in [0,L][0,L], the excess temperature is nonzero and λ=1t2−t1log⁡|u(x,t1)−Tau(x,t2)−Ta|,t2>t1.\boxed{\lambda=\frac 1{t_2-t_1} \log\left|\frac{u(x,t_1)-T_a}{u(x,t_2)-T_a}\right|,\quad t_2>t_1.} The absolute values allow sensors on either side of the node; the ratio eliminates the unknown local amplitude.

Original worksheet page 2: question and worked solution for 9-1-005

Original worksheet layout. Use Enlarge or open the PDF for a closer view.