Heat Transfer Interview Questions

The three modes are conduction (heat transfer through a solid medium without bulk motion, e.g., heat flowing through a metal rod), convection (heat transfer between a surface and a moving fluid, e.g., air cooling a hot engine), and radiation (heat transfer through electromagnetic waves without any medium, e.g., solar energy reaching Earth). In most engineering applications, multiple modes occur simultaneously.

Fourier's law states that the rate of heat conduction through a material is proportional to the negative temperature gradient and the cross-sectional area: Q = -kA(dT/dx). Here, k is thermal conductivity (W/m-K), A is area, and dT/dx is the temperature gradient. The negative sign indicates heat flows from high to low temperature. This law forms the foundation for all conduction analysis in thermal engineering.

Thermal conductivity (k) is a material property that quantifies its ability to conduct heat, measured in W/m-K. Metals have high thermal conductivity (copper ~400 W/m-K, aluminum ~237 W/m-K) due to free electrons. Insulators like fiberglass (~0.04 W/m-K) and air (~0.026 W/m-K) have low values. For most solids, k decreases with temperature, while for gases it increases. This property is crucial in selecting materials for heat sinks, insulation, and heat exchangers.

Newton's law of cooling states that the rate of convective heat transfer is proportional to the temperature difference between a surface and the surrounding fluid: Q = hA(Ts - T∞). Here, h is the convective heat transfer coefficient (W/m²-K), A is surface area, Ts is surface temperature, and T∞ is fluid temperature. This law is fundamental in designing cooling systems, HVAC equipment, electronic cooling, and any application involving fluid-surface heat exchange.

Natural (free) convection occurs due to buoyancy forces arising from density differences caused by temperature gradients; fluid motion is driven by gravity (e.g., warm air rising from a radiator). Forced convection involves external means like fans, pumps, or wind to move the fluid (e.g., car radiator with a fan). Forced convection typically has higher heat transfer coefficients (h = 25-250 W/m²-K for air) compared to natural convection (h = 5-25 W/m²-K for air) due to higher fluid velocities.

The Stefan-Boltzmann law states that the radiant heat emitted by a blackbody is proportional to the fourth power of its absolute temperature: Q = σAT⁴, where σ = 5.67×10⁻⁸ W/m²-K⁴. For real surfaces, emissivity (ε) is introduced: Q = εσAT⁴. Emissivity ranges from 0 to 1, representing how effectively a surface emits radiation compared to a blackbody. Polished metals have low emissivity (~0.05), while oxidized surfaces and paints have high values (~0.9).

Thermal resistance (R_th) quantifies the opposition to heat flow, analogous to electrical resistance. For conduction: R_th = L/(kA), for convection: R_th = 1/(hA), and for radiation: R_th = 1/(h_rad×A). Just as voltage drives current through electrical resistance (V = IR), temperature difference drives heat flow through thermal resistance (ΔT = Q×R_th). Series resistances add directly, while parallel resistances combine reciprocally, enabling analysis of complex thermal circuits.

Major types include: Shell-and-tube (most common in refineries, power plants due to high pressure/temperature capability), Plate heat exchangers (compact, high efficiency, used in HVAC, dairy industry), Double-pipe (simple, small capacity applications), Finned-tube (air-cooled condensers, car radiators), and Regenerators (gas turbines, steel plants). Selection depends on fluid types, pressure/temperature requirements, space constraints, and maintenance accessibility. Shell-and-tube can handle fouling better, while plate exchangers offer higher effectiveness in a smaller footprint.

Fins are extended surfaces used to increase the heat transfer rate by expanding the surface area available for convection. They are particularly effective when one fluid has a much lower heat transfer coefficient than the other (e.g., air vs. liquid). Common applications include air-cooled engines, electronic heat sinks, radiators, and air conditioning condensers. The effectiveness of fins depends on their geometry, material thermal conductivity, and the convective environment. High-conductivity materials like aluminum and copper are preferred for fin construction.

Critical radius of insulation is the outer radius at which adding more insulation actually increases heat transfer instead of decreasing it. For a cylinder, r_cr = k/h, where k is insulation thermal conductivity and h is external convective coefficient. Below critical radius, adding insulation increases surface area for convection more than it adds thermal resistance, thus increasing heat loss. This concept is crucial for small-diameter pipes and electrical wires. For typical insulation (k~0.05 W/m-K) in air (h~10 W/m²-K), r_cr ≈ 5mm.

Key dimensionless numbers include: Reynolds number (Re = ρVL/μ) - ratio of inertial to viscous forces, determines flow regime; Nusselt number (Nu = hL/k) - ratio of convective to conductive heat transfer; Prandtl number (Pr = μCp/k) - ratio of momentum to thermal diffusivity, a fluid property; Grashof number (Gr = gβΔTL³/ν²) - ratio of buoyancy to viscous forces in natural convection. These numbers enable correlation of experimental data and scaling of heat transfer results across different conditions.

The overall heat transfer coefficient (U) combines all thermal resistances in a heat transfer system into a single value. For a plane wall with convection on both sides: 1/UA = 1/(h₁A₁) + L/(kA) + 1/(h₂A₂). For heat exchangers, U relates to heat transfer rate as Q = UA×ΔT_mean. Typical values: water-to-water (1000-2500 W/m²-K), water-to-air (30-60 W/m²-K), steam condensers (1500-4000 W/m²-K). U decreases over time due to fouling, requiring fouling factors in design calculations.

A blackbody is an ideal surface that absorbs all incident radiation regardless of wavelength and direction, with emissivity ε = 1. It emits the maximum possible radiation at any given temperature. A graybody is a more realistic model where emissivity is constant across all wavelengths but less than 1 (0 < ε < 1). Real surfaces are often approximated as graybodies for engineering calculations. Examples: lampblack (ε ≈ 0.95) behaves nearly as a blackbody, while polished aluminum (ε ≈ 0.05) is far from it.

In steady-state heat transfer, temperature at any point in the system remains constant over time (∂T/∂t = 0), meaning heat entering equals heat leaving. Example: a well-insulated pipe carrying hot water at constant flow. In transient (unsteady) heat transfer, temperatures change with time as the system moves toward equilibrium. Example: quenching a hot metal part, heating food in a microwave. Transient analysis requires solving the heat diffusion equation with time-dependent terms and is characterized by the Biot and Fourier numbers.

Fouling is the accumulation of unwanted deposits on heat exchanger surfaces, including scaling (mineral deposits), biological growth, corrosion products, and particulate matter. Fouling increases thermal resistance, reduces heat transfer efficiency, increases pressure drop, and can lead to flow blockages. Fouling factors (R_f, typically 0.0001-0.002 m²-K/W) are added to design calculations to account for performance degradation. Management strategies include proper water treatment, regular cleaning, material selection, and designing for cleanability. Fouling can reduce efficiency by 10-30% if unmanaged.

The Log Mean Temperature Difference (LMTD) method calculates heat transfer using Q = UA×LMTD, where LMTD = (ΔT₁ - ΔT₂)/ln(ΔT₁/ΔT₂). For counterflow: ΔT₁ = Th,in - Tc,out and ΔT₂ = Th,out - Tc,in. For parallel flow: ΔT₁ = Th,in - Tc,in and ΔT₂ = Th,out - Tc,out. A correction factor F is applied for multi-pass and cross-flow configurations. LMTD method is best when inlet and outlet temperatures are known. For complex geometries or when only inlet temperatures are known, the NTU-effectiveness method is preferred.

The NTU-effectiveness method uses Number of Transfer Units (NTU = UA/C_min) and heat capacity ratio (C_r = C_min/C_max) to determine effectiveness (ε = Q_actual/Q_max). Effectiveness relations depend on heat exchanger configuration. For counterflow: ε = [1-exp(-NTU(1-C_r))]/[1-C_r×exp(-NTU(1-C_r))]. Advantages: works when only inlet temperatures are known (rating problem), handles any exchanger configuration through effectiveness-NTU charts, directly gives performance metrics. LMTD requires iterative solutions when outlets are unknown, making NTU method more practical for rating existing exchangers.

Fin efficiency (η_f) is the ratio of actual heat transfer to the heat transfer if the entire fin were at base temperature. For a straight rectangular fin: η_f = tanh(mL)/(mL), where m = √(hP/kA_c), h is convective coefficient, P is perimeter, k is conductivity, and A_c is cross-sectional area. Factors affecting efficiency: high thermal conductivity increases efficiency, longer fins reduce efficiency, higher convective coefficient reduces efficiency, thicker fins are more efficient. Typical efficiencies range from 70-95%. Design optimization balances fin length, material, and geometry for maximum heat dissipation per unit weight.

For a composite wall, thermal resistances are combined like electrical circuits. Series layers (convection + conduction + convection): R_total = 1/(h₁A) + Σ(L_i/k_iA) + 1/(h₂A). For parallel paths (like studs in insulation), calculate equivalent resistance using 1/R_eq = 1/R₁ + 1/R₂. Heat transfer rate: Q = ΔT_overall/R_total. Interface temperatures are found using Q×R to that point. This approach simplifies complex geometries, accounts for contact resistances at interfaces (typically 0.5-5×10⁻⁴ m²-K/W), and enables quick parametric studies for insulation optimization in building walls, industrial furnaces, and electronic enclosures.

The Biot number (Bi = hL_c/k) compares external convective resistance to internal conductive resistance, where L_c is characteristic length (volume/surface area). When Bi < 0.1, internal resistance is negligible, and the lumped capacitance method applies (uniform temperature throughout the body). When Bi > 0.1, temperature gradients within the body are significant, requiring spatial analysis using Heisler charts or numerical methods. For a sphere being quenched (h=1000 W/m²-K, k=200 W/m-K, D=50mm): Bi = 1000×0.0083/200 = 0.04, so lumped analysis is valid. This criterion is crucial for hardening, annealing, and thermal processing design.

View factor (F₁₂) is the fraction of radiation leaving surface 1 that directly reaches surface 2, depending only on geometry. Key rules: Summation rule (ΣF₁ⱼ = 1 for an enclosure), Reciprocity (A₁F₁₂ = A₂F₂₁), Self-view factor for concave surfaces (F₁₁ > 0). For parallel plates of equal size with separation H and dimension L: F₁₂ approaches 1 as H→0. Heat exchange between two surfaces: Q₁₂ = A₁F₁₂σ(T₁⁴-T₂⁴) for blackbodies. View factors are tabulated or calculated using integration. Understanding view factors is essential for furnace design, spacecraft thermal control, and industrial heating applications.

The velocity boundary layer is the region where fluid velocity increases from zero at the wall to 99% of free-stream velocity. The thermal boundary layer is where temperature changes from wall temperature to 99% of free-stream temperature. Their relative thickness depends on Prandtl number: for Pr > 1 (oils), thermal BL is thinner; for Pr < 1 (liquid metals), thermal BL is thicker; for Pr ≈ 1 (gases), they are similar. This relationship affects heat transfer: thinner thermal BL means steeper temperature gradient and higher heat transfer. For turbulent flow, both boundary layers have similar structures with a viscous sublayer, buffer zone, and turbulent core.

Key considerations include: Tube-side vs shell-side fluid placement (corrosive, high-pressure, fouling fluids typically tube-side), TEMA standards selection (R for refinery, C for commercial, B for chemical), baffle design (segmental, disc-and-doughnut) affecting shell-side flow pattern and pressure drop, tube arrangement (triangular for better heat transfer, square for easier cleaning), number of shell and tube passes for desired approach temperature, material selection for corrosion resistance and thermal expansion, and proper venting/drainage. Pressure drop constraints often limit velocities (tube: 1-2.5 m/s for liquids, shell: 0.3-1 m/s) balancing heat transfer enhancement.

Radiation shields are low-emissivity surfaces placed between hot and cold surfaces to reduce radiative heat exchange. Each shield creates additional resistance. For N shields between parallel plates, heat transfer reduces to: Q_with_shields = Q_without_shields/(N+1) for shields with same emissivity as surfaces. For different emissivities: each shield adds resistance of (1/ε₁ + 1/ε₂ - 1). Example: polished aluminum shields (ε=0.05) are highly effective in cryogenic applications, vacuum flasks, and spacecraft insulation (MLI uses multiple reflective layers). A single shield can reduce radiation by 50% or more depending on emissivities.

A heat pipe is a passive two-phase heat transfer device consisting of a sealed container with working fluid and wick structure. At the evaporator end, heat vaporizes the fluid; vapor travels to the cooler condenser end where it liquefies, releasing latent heat. The wick (sintered, grooved, or mesh) returns liquid to the evaporator via capillary action. Heat pipes offer extremely high effective thermal conductivity (10,000-100,000 W/m-K), are passive with no moving parts, and work against gravity with proper wick design. Applications include laptop/smartphone cooling, satellite thermal control, permafrost stabilization, and CPU heat sinks.

Selection depends on: Flow geometry (internal/external, flat plate/cylinder/sphere), Flow regime (laminar/turbulent based on Re), Convection type (forced/natural/mixed based on Gr/Re² ratio), and Boundary conditions (constant wall temperature or constant heat flux). For external flow over flat plate: use Blasius (laminar, Re<5×10⁵) or Dittus-Boelter type. For internal pipe flow: use Sieder-Tate or Gnielinski correlation. For natural convection: use Churchill-Chu correlation. Always check validity range (Re, Pr, L/D limits). For non-standard geometries, CFD or experimental correlations are needed.

Pool boiling has distinct regimes as wall superheat increases: Natural convection (ΔT<5°C, no bubbles), Nucleate boiling (ΔT=5-30°C, bubbles form at nucleation sites, very high heat transfer up to 1 MW/m²), Transition boiling (unstable, vapor film forms and collapses), and Film boiling (stable vapor layer insulates surface, much lower heat transfer). The Critical Heat Flux (CHF) occurs at peak nucleate boiling; exceeding it causes rapid temperature rise (burnout). Design must maintain nucleate boiling regime. Condensation modes: filmwise (lower h, ~5000-25000 W/m²-K) vs dropwise (higher h but difficult to maintain).

Key challenges include: Maintaining optimal temperature range (20-40°C) for performance and longevity, Managing temperature uniformity (ΔT<5°C) across cells to prevent uneven aging, Handling high heat loads during fast charging (2-3x normal), Preventing thermal runaway propagation between cells, Cold weather pre-conditioning requiring heating, Balancing cooling system weight/cost against performance, and Integration with cabin HVAC. Solutions include liquid cooling (glycol loops, cold plates), air cooling (limited to lower power density), phase change materials, and refrigerant direct cooling. Tesla uses serpentine ribbon coolant channels; Chevrolet Bolt uses aluminum cold plates.

Advantages: Higher heat transfer coefficients (3-5x) due to turbulent flow in corrugated channels, Compact design (smaller footprint for same duty), Easy capacity modification by adding/removing plates, Lower fouling tendency due to high turbulence, Complete counterflow possible (high effectiveness), Easy inspection and cleaning. Limitations: Lower pressure capability (typically <25 bar for gasketed type), Temperature limited by gasket material (<180°C typically), Not suitable for highly viscous fluids, Gasket failures can cause leaks or fluid mixing, Higher initial cost per unit area. Brazed and welded plate exchangers extend pressure/temperature limits but lose flexibility.

Optimization involves: Fin spacing (too close increases boundary layer interference, too wide wastes space; optimal spacing depends on flow velocity and fin height), Fin thickness (thicker improves efficiency but adds weight and cost), Fin height (taller fins have lower efficiency; optimal length where marginal heat transfer equals added material), Material selection (aluminum offers best cost/performance, copper for high-power applications), Surface enhancement (anodizing, micro-fins increase area), and Flow arrangement (bypass effects in ducted vs unducted). For natural convection, optimal spacing ≈ 7mm for typical heat sinks. CFD and thermal testing validate designs against junction temperature limits.

Thermal contact resistance occurs at interfaces between solid surfaces due to surface roughness creating air gaps (only 1-2% of nominal area actually contacts). Typical values: 0.5-5×10⁻⁴ m²-K/W for machined metal surfaces. Minimization strategies: Increase contact pressure (reduces air gap volume), Apply thermal interface materials (TIMs) - thermal greases (1-5 W/m-K), thermal pads (3-10 W/m-K), phase change materials, liquid metals (20-70 W/m-K), Improve surface finish (reduce roughness), Use softer materials that conform to surfaces. In electronics, TIM selection is critical; a 0.1mm bondline with 5 W/m-K TIM adds 0.02 K/W resistance to each component.

Fouling factors (R_f) add thermal resistance to the clean overall coefficient: 1/U_dirty = 1/U_clean + R_f,i + R_f,o. TEMA standards provide typical values: cooling tower water (0.00035 m²-K/W), seawater (0.00009), clean steam (0.00009), fuel oil (0.0009), river water (0.00035-0.0005). Design must balance over-sizing (cost, weight) against maintenance frequency. Fouling rate depends on fluid velocity (higher velocity reduces deposition), wall temperature (scaling increases with temperature), fluid chemistry, and material compatibility. Excess area of 10-25% is common. Modern approaches use fouling monitors and predictive maintenance rather than fixed factors.

For Bi < 0.1, energy balance gives: ρVc(dT/dt) = -hA(T-T∞). Solving: (T-T∞)/(Ti-T∞) = exp(-t/τ), where τ = ρVc/(hA) is the time constant. Physical interpretation: τ is the time to reach 63.2% of temperature change; 5τ gives 99% approach to equilibrium. For faster response (smaller τ): increase h (forced convection), increase A/V ratio (smaller objects, fins), decrease ρc (lower thermal mass). Example: quenching steel ball (D=20mm) in oil (h=1000 W/m²-K): τ = 7800×(0.01/3)×500/(1000) = 13 seconds. This analysis applies to thermocouple response, small electronics, and heat treatment processes.

Maximum effectiveness depends on: Flow arrangement (counterflow can theoretically reach ε=1 with infinite area; parallel flow limited to ε_max=0.5 when C_r=1), Capacity rate ratio (C_r=0 gives ε=1-exp(-NTU) for all configurations), Physical constraints (pressure drop increases with NTU, fouling, cost), and Temperature pinch (minimum approach temperature for practical operation, typically 5-10°C). For condensers/evaporators with phase change (C_r→0), high effectiveness is easily achievable. Economic optimum typically gives ε=0.7-0.85 balancing capital cost against operating energy savings. Above NTU≈3, diminishing returns make additional area uneconomical.

Mixed convection occurs when both buoyancy and forced flow significantly affect heat transfer. The criterion uses Richardson number: Ri = Gr/Re² (or Gr/Re^2.5 for turbulent flow). When Ri << 1, forced convection dominates; Ri >> 1, natural convection dominates; Ri ≈ 1, mixed convection must be analyzed. For assisting flow (buoyancy aids forced flow): Nu^n ≈ Nu_forced^n + Nu_natural^n (n≈3 for laminar). For opposing flow: effects partially cancel. Mixed convection is common in vertical tubes at low flow rates, electronic equipment with fans, and building ventilation. Accurate analysis often requires CFD due to complex interactions.

Comprehensive strategy includes: Cell-level protection (ceramic-coated separators, PTC thermistors, CID vents), Module design with fire-resistant barriers between cell groups (mica sheets, intumescent materials expanding at ~150°C), Active cooling maintaining cells at 25-35°C with redundant pumps and multiple cooling circuits, Early detection via distributed temperature sensing (fiber optic or NTC thermistors every 2-3 cells) with ML-based anomaly detection, Controlled venting through pressure relief directing gases to safe exhaust pathways away from passengers, Passive propagation delay using phase change materials (paraffin with 200 J/g latent heat) providing 3-5 minute buffer, and BMS integration shutting down affected modules while maintaining others. Testing per UNECE R100 validates 5-minute no-propagation requirement.

Conjugate heat transfer (CHT) simultaneously solves heat conduction in solids and convection in adjacent fluids with coupled boundary conditions at interfaces. It becomes necessary when: temperature distribution in solid significantly affects fluid behavior (thermal developing flows), solid thermal resistance is comparable to convective resistance (Bi≈1), detailed local heat flux distribution is needed (electronic cooling), or when simplified boundary conditions are inadequate (turbine blade cooling). CHT requires matching temperature and heat flux at interfaces: T_solid = T_fluid and k_s(∂T/∂n)_s = k_f(∂T/∂n)_f. Computational cost is higher than decoupled analysis but essential for accurate hot-spot prediction in gas turbines, electronic packages, and compact heat exchangers.

Multi-scale approach: Chip level (μm-mm) - detailed power maps from electrical simulation, spreading resistance in die/TIM/IHS, compact thermal models (Delphi-style CTMs) with 3-10 nodes representing key junctions; Package level (mm-cm) - detailed 3D FEA including leadframe/substrate/solder, validated against thermal test chips (JEDEC standards), boundary condition sensitivity analysis; Board level (cm) - PCB as orthotropic material (in-plane 30 W/m-K, through-plane 0.3 W/m-K), component interactions via radiation and board conduction; System level (m) - coarse CFD with components as volumetric heat sources using CTMs, fan curves and vent impedances. Each level provides boundary conditions for detailed models and validates against measurements. Key challenge: information transfer between scales maintaining accuracy while managing computational cost.

Participating media (gases like CO₂, H₂O, and particles) absorb, emit, and scatter radiation. Analysis requires solving the Radiative Transfer Equation (RTE): dI_λ/ds = κ_λI_bλ - κ_λI_λ + σ_s/(4π)∫I_λ(s')Φ(s',s)dΩ. Methods include: P1 approximation (fast but inaccurate for optically thin/thick extremes), Discrete Ordinates Method (DOM/S_N) solving along discrete directions, Monte Carlo (accurate but computationally expensive), and Weighted Sum of Gray Gases (WSGG) for combustion gases reducing spectral complexity. Key parameters: absorption coefficient (κ) from HITRAN database or WSGG coefficients, optical thickness (τ=κL). For furnaces, gas emissivity charts (Hottel charts) provide engineering estimates. Combined with CFD, participating radiation adds 30-50% to computational cost but is essential for furnaces, combustion chambers, and high-temperature processes above 1000K.

Design approach: PCM selection - erythritol (Tm=118°C, latent heat 340 kJ/kg) or HDPE (120-140°C) for this temperature range, considering congruent melting, cycling stability, and cost; Heat transfer enhancement - due to low PCM conductivity (~0.5 W/m-K), use finned tubes (10-20% fin volume), metal foam inserts (aluminum 10-15 PPI), or encapsulated PCM spheres (1-5cm diameter) for high surface area; System sizing - for 100 kWh storage: m_PCM = Q/(L + c_pΔT) ≈ 1000 kg considering sensible storage, container volume 1.5-2× PCM volume for expansion; Thermal resistance network analysis ensuring adequate charge/discharge rates; Material compatibility (corrosion, thermal stress); and Safety features (expansion volume, containment). Validate with 1D Stefan problem analysis or CFD with enthalpy-porosity method for phase tracking.

Inverse heat transfer problems determine unknown boundary conditions, heat sources, or material properties from measured temperature data - opposite of the direct problem. Examples: determining surface heat flux from internal temperature measurements, identifying heat source location from thermal images. Challenges: ill-posed nature (small measurement errors cause large solution oscillations), non-uniqueness, instability. Solution methods: Regularization techniques (Tikhonov adding smoothness constraints, truncated SVD), Iterative methods (conjugate gradient minimizing objective function J=Σ(T_measured-T_computed)² + regularization term), Bayesian approaches quantifying uncertainty, Neural networks for real-time applications. Applications include: estimating heat transfer coefficients in quenching, determining weld pool thermal history, identifying defects via thermal NDT, and estimating metabolic heat generation in medical imaging. Proper sensor placement and uncertainty quantification are critical for reliable solutions.

Design constraints: q'' = 500 W/cm², junction limit T_j < 125°C, coolant inlet 25°C. Microchannel geometry: hydraulic diameter 200-500μm balances heat transfer and pressure drop; aspect ratio 4-8 for high effectiveness. Using Churchill correlation for developing laminar flow (Re~500): with D_h=300μm, w_ch=100μm, H=800μm, 50 channels across 1cm give Nu≈8-10, h≈50,000 W/m²-K. Thermal resistance: R_total = R_conv + R_cond + R_TIM + R_spreading. At m_dot=0.01 kg/s water: ΔT_fluid=12°C, ΔT_conv=5°C, ΔT_cond(Cu base)=3°C, requiring T_base<50°C for junction compliance. Pressure drop: ΔP≈50 kPa acceptable for micropumps. Manifold design prevents flow maldistribution. Manufacturing: CNC or laser machining in copper, diffusion bonding for cover. Validate with infrared thermography on powered module.

RANS models: k-ε (robust, overpredicts mixing, uses wall functions with y+>30; standard for industrial flows), k-ω SST (better adverse pressure gradients, boundary layers, requires y+<5), Reynolds Stress Models (anisotropic effects, reattachment, but computationally expensive and less stable). For heat transfer: turbulent Prandtl number Pr_t≈0.85-0.9 assumed constant, limiting accuracy in low-Re and separated regions. LES resolves large eddies, models small scales; better for separated flows but 10-100× cost of RANS; requires fine mesh (y+<1) and time-dependent solution. DNS resolves all scales but impractical for engineering Re. For heat transfer accuracy: RANS adequate for attached flows (±10%), LES needed for separation, jets, complex geometries (±5%). Enhanced wall treatment (adaptive blending) improves near-wall heat transfer prediction. Validation against experimental Nu correlations essential.

TEC performance depends on operating current: Q_c = αIT_c - 0.5I²R - K(T_h-T_c), where α is Seebeck coefficient, R is electrical resistance, K is thermal conductance. Maximum cooling (Q_c,max) occurs at I_Qmax = αT_c/R but gives COP≈0.3-0.5. Maximum COP occurs at lower current I_COP = αΔT/[R(√(1+ZT_m)-1)] giving COP≈1-2 but 30-50% lower Q_c. Optimization approach: for fixed heat load, select TEC with Q_c,max > 1.5× required load allowing operation at higher COP point; multi-stage TECs for ΔT>50°C (each stage adds ~20°C capability); proper heat sink design critical (hot side Rth < cold side); parallel TECs at reduced current vs single TEC at high current tradeoff; consider Peltier modules with ZT>1 materials (Bi₂Te₃ alloys). System COP must include TEC COP and heat sink fan power for fair comparison with vapor compression.

Heisler charts provide solutions for 1D transient conduction in plane walls, cylinders, and spheres with convective boundary conditions. They plot dimensionless temperature θ = (T-T∞)/(Ti-T∞) versus Fourier number Fo = αt/L² for various Biot numbers. Usage: calculate Bi = hL/k and Fo, read centerline θ₀ from first chart, use second chart for temperature at other positions θ/θ₀ vs x/L. For multidimensional bodies (short cylinder, rectangular block): apply product solution θ_total = θ_plane × θ_cylinder. Limitations: constant thermal properties assumption (problematic for large ΔT), uniform initial temperature required, constant boundary conditions, applicable only after initial transient (Fo>0.2), chart reading errors (±5%), and single-term series approximation. For complex geometries, variable properties, or time-varying boundaries, numerical methods (FDM, FEM) with smaller time steps provide better accuracy.

Challenges: flow regime transitions (bubbly, slug, annular, mist) dramatically change heat transfer mechanisms; non-equilibrium effects (vapor superheat, subcooled boiling); interfacial phenomena (bubble dynamics, entrainment); and strong coupling between hydrodynamics and heat transfer. Approaches: Homogeneous model (simplest, assumes phases at same velocity, uses mixture properties); Separated flow models (Lockhart-Martinelli for pressure drop, Chen correlation for boiling combining nucleate and convective contributions); Drift-flux models (relative velocity via drift velocity); Two-fluid models (separate conservation equations for each phase, closure via interfacial transfer terms). For heat exchangers with phase change: zone analysis dividing into subcooled, two-phase, superheated regions. CFD approaches: VOF/Level-set for resolved interfaces, Eulerian-Eulerian for dispersed flows. Correlations remain industry standard due to modeling complexity; Shah, Kandlikar correlations widely validated for in-tube flow boiling.

Target: PUE = Total facility power / IT load power < 1.3 means cooling + overhead < 300kW. Strategy: Economizer-first approach using outside air or water when T_ambient < 20°C (70%+ hours in temperate climates); Hot/cold aisle containment achieving supply T=20-25°C (ASHRAE A1); Raised floor or overhead supply with variable speed CRAC units responding to rack-level temperature sensors; Rear-door heat exchangers for high-density racks (>20kW) using chilled water; Free cooling chillers with water-side economizer (plate HX) bypassing compressor when T_wb < 15°C; Variable primary chilled water flow with ΔT=8-12°C; EC fans throughout (60%+ efficiency at part load); Warm water cooling (40-45°C supply) enabling direct-to-chip cooling and waste heat recovery potential. Controls: CFD-informed setpoints, predictive algorithms for weather-based staging. Monitoring: real-time PUE calculation, leak detection, redundancy management. Achievable PUE: 1.2-1.25 with this approach.

Challenges: extreme environments (sun-facing surfaces >150°C, shadowed <-150°C), no convection in vacuum (radiation only), orbital variations (eclipse, solar angle changes), long mission life with degradation, and tight mass/power budgets. Passive strategies: Multi-layer insulation (MLI) with 20+ aluminized layers reducing heat loss by 99%; Surface coatings controlling α/ε ratio (white paint for radiators: low α, high ε; gold for insulation: low both); Heat pipes redistributing power; Thermal mass for eclipse protection. Active strategies: Heaters with thermostatic control preventing cold-soak; Louvers modulating radiator area; Fluid loops for high-power payloads; Cryocoolers for IR sensors requiring <100K. Thermal model: node-network analysis with orbital heating terms (Q_solar, Q_albedo, Q_IR) varying with β-angle. Thermal balance test in vacuum chamber validates flight hardware. Design margins: +10°C qualification, +5°C acceptance from operating limits.

Pinch analysis (Linnhoff method) minimizes external utility requirements by maximizing heat recovery between process streams. Steps: Extract stream data (supply/target temperatures, heat capacity rates); Construct composite curves (hot streams releasing heat, cold streams requiring heat) on T-H diagram; Identify pinch point (minimum approach temperature ΔT_min, typically 10-20°C) where curves closest; Calculate minimum hot utility (above pinch) and cold utility (below pinch) from composite curve end-points; Design network following pinch rules: no heat transfer across pinch, no external cooling above pinch, no external heating below pinch. Grand composite curve identifies utility levels and potential for heat engines/pumps. For complex networks: mathematical optimization (MINLP) finds cost-optimal configurations considering capital/operating cost tradeoffs, forbidden matches, and controllability. Industrial applications achieve 20-40% energy savings through systematic retrofit or grassroots design.

Laser Powder Bed Fusion (LPBF) thermal modeling challenges: extreme heating/cooling rates (10⁵-10⁷ K/s), small melt pool (~100μm), multiple scales (μm powder to cm part), phase changes (melting, solidification, solid-state), and process-microstructure-property linkages. Modeling approaches: Analytical models (Rosenthal solution for moving point source provides initial estimates); Part-scale FEM (coarse mesh, layer-by-layer activation, homogenized properties, 'inherent strain' for distortion); Meso-scale (resolved scan tracks, temperature-dependent properties, latent heat via enthalpy method); Micro-scale (powder-scale discrete element + CFD for melt pool dynamics, Marangoni convection, keyholing). Key phenomena: surface heat loss via convection (h~10-20 W/m²-K in inert gas) and radiation, thermal cycling causing residual stress, re-melting of previous layers affecting microstructure. Calibration against thermocouple data, melt pool dimensions from high-speed imaging, and residual stress measurements. Used for process parameter optimization, support structure design, and distortion compensation.