Please help me formulate (and solve) a differential equation for a practical DIY problem concerning air sterilization (for covid)

89 Views Asked by At

Here I present the DIY air sterilisation project which this is about.

I want to calculate the overall sterilization rate of the device under some simplifying assumptions. (UVC is the hard ultra violet (UV) light killing the germs.)

schema drawing enter image description here I have

  • the effective UVC power $E$ of my germicidal lamp. The lamp is a Hg-"neon" tube along the axis of the cylindrical device.
  • the length $L$ of the lamp
  • the radius $R_l$ of the lamp
  • the inner radius $R_c$ of the cylinder
  • a radius-invariant constant airflow through the cylinder (we are way into turbulent flow) with the speed $v$, so that I can calculate a mean residence time of the virus traveling through the killzone for
  • a time $t = L/v$
  • the UVC irradiation intensity $I$ at a given radius $r$ in the cylinder $I(r) = \frac{E}{2\pi \, L \, r}$
  • a irradiation dose $d(r) = I(r)*t$
  • a virus-specific susceptibilty rate $k$ to calculate the specific killrate $P(r)=1-e^{-k \, d(r)}$
  • a infinitisimal thin hollow cylinder (thickness $dr$) at radius $r$ around the axis of the cylinder with the volume $V(r) = 2 \pi L \, r \, dr$ , which is proportional to the amount of germs exposed to the specific irradiation dose $d(r)$.

Now I look for the overall kill rate for the volume between $R_l$ and $R_c$.

Would it be correct to calculate $\frac{1}{R_c - R_l}\cdot\int_{R_l}^{R_c} P(r) \cdot V(r)dr = \frac{1}{R_c - R_l}\cdot\int_{R_l}^{R_c}2 \pi L \, r \, (1-e^{-k \frac{E}{2\pi\,v \, r}}) dr$ ?

Is that the differential equation I need? What is a solution for that?

Here are some matlab lines which are unfortunatly not correct as the result is independent of most of the variables.

syms R_l R_c L r E k v V(r);
u = symunit;

assume(0<R_l <R_c);

t = L*u.m/(v*u.m/u.s);
I = E*u.W/(2*pi*L*u.m*r*u.m);
d = I * t;
P = 1-exp(-k*(u.cm^2/u.mJ)*d);
diff(V*u.m^3) == 2*pi* L*u.m * r*u.m *diff(r*u.m);
killrate = V*P


latex(killrate);

mean_killrate = int( killrate ,r, R_l, R_c)/(R_c -R_l);

mean_killrate_small = eval( subs(mean_killrate, {R_l R_c L E k v}, {0.016/2 0.125/2 0.113 0.95 0.41 3}))

mean_killrate_medium = eval( subs(mean_killrate, {R_l R_c L E k v}, {0.016/2 0.125/2 0.205 4 0.41 3}))

mean_killrate_large = eval( subs(mean_killrate, {R_l R_c L E k v}, {0.040/2 0.125/2 0.51 27 0.41 3}))

edit: updated the matlab code to use diff() for differential terms and measurement units (to find inconcistencies in the equations).