Question 1

How do I find the thermal resistance of a finned heatsink?

Accepted Answer

Run the fin-array chain: the fin parameter m = sqrt(2h/(k*t)), the corrected length L_c = L + t/2, the fin efficiency eta_f = tanh(m*L_c)/(m*L_c), the wetted areas A_f, A_b, A_t, the overall surface efficiency eta_o = 1 - (N*A_f/A_t)*(1-eta_f), and finally R_array = 1/(h*eta_o*A_t). Add the optional base-plate conduction R_base = t_base/(k_base*A_base) to get the achievable sink-to-ambient resistance theta_SA. For h=40 W/m2K, k=180, N=11 fins, t=1mm, L=30mm, W=100mm, A_base=0.005 m2, theta_SA is about 0.40 C/W.

Question 2

What is fin efficiency and overall surface efficiency?

Accepted Answer

Fin efficiency eta_f = tanh(m*L_c)/(m*L_c) is the ratio of the heat a real fin dissipates to the heat an ideal isothermal fin at the base temperature would dissipate; it falls below 1 because the fin tip is cooler than the root. Overall surface efficiency eta_o = 1 - (N*A_f/A_t)*(1-eta_f) blends the finned area and the exposed base into a single factor for the whole array, so R_array = 1/(h*eta_o*A_t). Many one-line calculators omit eta_o and therefore over-state the heat the heatsink can shed.

Question 3

How do I check a heatsink against a required theta_SA?

Accepted Answer

Get the required sink-to-ambient resistance from the power-loss heatsink calculator (theta_SA,req = (Tj_max - Ta)/Pd - theta_JC - theta_CS), paste it into the Required theta_SA field here, and compare: the design passes when the achievable theta_SA is less than or equal to the required theta_SA. A small clip heatsink with h=5, N=3, L=10mm gives about 18.2 C/W which FAILS a required 11.46 C/W, so you need more fins, a higher h, or forced air.

Question 4

When is the fin-efficiency equation valid?

Accepted Answer

The 1-D fin equation assumes the Biot number Bi = h*t/(2k) is small (roughly below 0.2-0.25), so heat conducts along the fin much faster than it convects off the sides. The corrected length L_c = L + t/2 folds the convective tip into an equivalent adiabatic-tip fin, so you must not also add a separate tip-convection term. Very long fins with m*L_c above about 5 carry almost no heat near the tip; the tool warns on both conditions.

Question 5

Can I estimate the convection coefficient h?

Accepted Answer

Yes, but treat it as preliminary. The tool offers a natural-convection estimator for a vertical parallel-plate channel (Elenbaas / Bar-Cohen-Rohsenow composite Nusselt correlation) and a forced-convection estimator for the channel between fins (laminar Nu about 7.54, h = Nu*k_air/D_h). Both are shown behind an estimate-only banner; the defensible path is a user-supplied measured or vendor h. Confirm the estimate before trusting a PASS/FAIL.

Heatsink Fin θSAThermal Resistance Calculator

Heatsink Fin θSA
Thermal Resistance Calculator