Opening the book…
The calculation of instantaneous heat flux \(q\left(\theta \right)\) across the engine cycle relies on a quasi-steady assumption. The gas-side convective heat transfer coefficient \(h_c\) is determined by engine geometry (bore \(B\), length ratios), gas properties (\(k,\mu ,c_p,\rho\)), flow field parameters (\(v\)), chemical energy release rate \({\dot{q}}_{ch}\), engine speed \(N\), and crank angle \(\theta\):
\[f\left(h_c,B,z,y_1,\dots ,y_m,v,u_1,\dots ,u_n,k,\mu ,c_p,\rho ,{\dot{q}}_{ch},N,\theta \right)=0 \nonumber \]
Applying dimensional analysis yields four fewer nondimensional groups, incorporating the Nusselt (\(\mathrm{Nu}\)), Reynolds (\(\mathrm{Re}\)), and Prandtl (\(\mathrm{Pr}\)) numbers:
\[\mathrm{\Phi }\left(\dfrac{h_cB}{k},\dfrac{\rho S_pB}{\mu },\dfrac{c_p\mu }{k},\dfrac{B}{L},\dfrac{\dot{q}_{ch}}{\rho c_pNT},r_c,R,y_1,\dots ,y_m,\theta ,u_1,\dots ,u_n\right)=0 \nonumber \]
where mean piston speed \(S_p=2LN\) is chosen as the characteristic velocity \(v\). Correlations generally adopt the classical turbulent boundary-layer form, Nusselt number:
\[\mathrm{Nu}=a{\mathrm{Re}}^m{\mathrm{Pr}}^n \nonumber \]
Taylor and Toong defined an effective mean gas temperature \(T_{g,a}\) such that the cycle-integrated heat flow sums to zero:
\[\int{A}h_c\left(T-T_{g,a}\right) d\theta =0 \nonumber \]
The overall gas-side heat transfer correlates against charge flow rate Reynolds number:
\[\mathrm{Nu}=\dfrac{\dot{Q}B}{\pi B^2/4\left(T_{g,a}-T_c\right)k_g}=\dfrac{4\dot{Q}}{\pi Bk_g\left(T_{g,a}-T_c\right)} \nonumber \]
\[\mathrm{Re}=\dfrac{\dot{m}B}{\mu_g\left(\pi B^2/4\right)}=\dfrac{4\dot{m}}{\pi {\mu }_gB} \nonumber \]
Annand observed that diesel engine correlations yield approximately \(25\%\) higher Nusselt numbers than SI engines at equivalent Reynolds numbers, primarily due to soot radiation.
Annand developed the following correlation and evaluates the cylinder-average charge temperature \({\overline{T}}_g=\dfrac{pVM}{m\overline{R}}\):
\[\left(\dfrac{h_cB}{k}\right)=a{\left(\dfrac{\rho S_pB}{\mu }\right)}^b \nonumber \]
Parameter range: \(b=0.7\), and \(0.35\le a\le 0.8\), scaling proportionally with the intensity of in-cylinder charge motion.
Woschni assumes \(\mathrm{Nu}=0.035{\mathrm{Re}}^m\) with \(m=0.8\), Woschni correlation with exponent can be summarized as:
\[h_c =3.26 B{\left[\mathrm{m}\right]}^{-0.2} p{\left[\mathrm{kPa}\right]}^{0.8} T{\left[\mathrm{K}\right]}^{-0.55} w{\left[\mathrm{m/s}\right]}^{0.8} \nonumber \]
The effective local gas velocity \(w\) accounts for a four stroke, water-cooled, four valve direct injection CI engine without swirl was expressed as:
\[w=\left[C_1{\overline{S}}_p+C_2\dfrac{V_dT_r}{p_rV_r}\left(p-p_m\right)\right] \nonumber \]
The effective local gas velocity \(w\) accounts for piston motion and combustion-induced density change:
| Operating Period | \(C_1\) (No Swirl) | \({ C}_1\) (With Swirl) | \(C_2\) |
|---|---|---|---|
| Gas Exchange | \(6.18\) | \(6.18+0.417\dfrac{v_s}{{\overline{S}}_p}\) | \(0\) |
| Compression | \(2.28\) | \(2.28+0.308\dfrac{v_s}{{\overline{S}}_p}\) | \(0\) |
| Combustion & Expansion | \(2.28\) | \(2.28+0.308\dfrac{v_s}{{\overline{S}}_p}\) | \( 3.24\times {10}^{-3}~\mathrm{m/(s}\cdot \mathrm{K)}\) |
(Note: Swirl speed \(v_s=\dfrac{B{\omega }_p}{2}\), where \({\omega }_p\) is the paddle-wheel rotation speed. \(p_m\) denotes the motored cylinder pressure).
Boundary-Layer Flat-Plate Models
Applied to swirling flows in Direct-Injection (DI) diesel engines at radius \(r\):
\[\dot{q}\left(r\right)=0.023\dfrac{k}{r}{\left(\dfrac{\omega r^2}{\nu }\right)}^{0.8}\left[{\overline{T}}_g\left(r\right)-T_w\left(r\right)\right] \nonumber \]
Two-Zone Models (SI Engines) separate the cylinder into unburned (\(u\)) and burned (\(b\)) zones:
\[{\dot{Q}}_u=A_{u,w}h_{c,u}\left(T_u-T_w\right),\qquad {\dot{Q}}_b=A_{b,w}h_{c,b}\left(T_b-T_w\right) \nonumber \]
Thermal Boundary-Layer Models divide the burned gas region into an adiabatic core (\(T_{ac}\)) and a thermal boundary layer of thickness \(δ\):
\[\dot{q}=\dfrac{k_e\left(T_{ac}-T_w\right)}{δ} \nonumber \]
Boundary layer growth: \(δ\propto t^{0.5}\) (laminar regime); \(δ\propto t^{0.8}\) (turbulent regime).
Intake Manifold modeled via steady, turbulent internal pipe flow correlations. Liquid fuel evaporation introduces two-phase heat transfer complexity. For exhaust system & valves, exhaust heat transfer exhibits the highest heat flux rates in the operating cycle due to high-velocity blowdown gases. For valve open period (\(L_v/D_v\le 0.2\) - Jet Regimes):
\[\mathrm{Nu}=K{\mathrm{Re}}^n_j \nonumber \]
where
\[\mathrm{Re}_j=\dfrac{v_jD_v}{ν} \nonumber \]
Standard pipe flow based on port velocity \(v_p\) and diameter \(D_p\).
For valve closed period, the correlation:
\[\mathrm{Nu}=0.022 {\overline{\mathrm{Re}}}^{0.8}_D \nonumber \]
For downstream exhaust pipe:
\[\mathrm{Nu}=0.0483 {\mathrm{Re}}^{0.783} \nonumber \]
(Note: Pulsating flow increases heat transfer rates to roughly twice those predicted by steady-flow turbulent pipe correlations).
This page titled 7.2: Convective heat transfer was last modified on Wed, 18 Nov 2026 07:18:33 GMT and is shared under a CC BY 4.0 license and was authored, remixed, and/or curated by .
The book could not be loaded just now.