GATE Mechanical: Heat Transfer, previous-year questions solved
Real Heat Transfer questions from the last four GATE Mechanical papers: conduction with heat generation and variable conductivity, fins and lumped cooling, free and forced convection, radiation networks, and heat exchangers. Each one has a step-by-step solution that matches the official answer key, and a note on where the idea shows up in real engineering work: electronics cooling, EV batteries, furnaces, power-plant condensers and thermal simulation.
Questions from official GATE ME papers (IIT Kharagpur 2022, IISc Bengaluru 2024, IIT Roorkee 2025, IIT Guwahati 2026); solutions by Skill-Lync. · Updated 16 Nov 2026 · 24 questions
Conductivity rises across a slab as k = A + Bx. What shape is the temperature profile?
- (A)
- (B)
- (C)
- (D)
Answer: (B) Official key: B
- No heat sources and steady state, so the heat flux is the same at every : (heat flows towards , so is negative and positive).
- . The conductivity grows with , so the slope must fall with : steep near , flatter near .
- Integrating: a curve that rises steeply and bends over (concave down) with no hump: plot (B).
(A) is the constant- answer, (C) is what you'd get if fell with , and (D) can't happen: a dip inside the slab would need a heat sink.
Insulation, refractory bricks and composite laminates all have conductivity that varies with position or temperature. Thermal engineers read the slope of a measured temperature profile the same way: where it is steep, the material conducts poorly.
A slab with k = CT generates 1280 kW/m³ and both faces are at 600 K. How hot is the centre?
Answer: 1000 K Official key: 1000 to 1000
- Governing equation: with . Use the Kirchhoff substitution: .
- So . By symmetry at ; integrating twice from the centre to a face:
- .
- .
Treating as constant at its face value (1200 W/m·K) gives , which is wrong: grows towards the hot centre, so the real peak is lower.
Conductivity of real materials changes a lot with temperature: plain carbon steel conducts only about half as well at 800 °C as at room temperature, and fuel pellets in a nuclear reactor are analysed with exactly this Kirchhoff transformation. FEA and CFD solvers handle it by letting k be a function of temperature, which makes the problem non-linear.
A 20 mm slab generates 100 MW/m³ with faces at 150 °C and 110 °C. Where is the hottest point?
Answer: 6 mm Official key: 6 to 6
- With uniform generation the profile is a parabola: .
- Right face: gives , so and .
- Maximum where : from the left face.
The peak is 159 °C. It sits nearer the hotter face: heat from the generation flows out both sides, but more of it goes to the cooler right face.
This is the model for any part that heats itself: a busbar or winding carrying current, a battery cell, a fuel plate. Engineers use the peak location and value to place temperature sensors and to check that the hottest point stays under the material or insulation limit.
A rod with heat generation q̇ = cos(2πx/L) is cooled at one end. What happens to its mid-point temperature if h doubles?
- (A)
- (B)
- (C)
- (D)
Answer: (C) Official key: C
- Energy balance on the whole rod: the only exit for heat is the end , so heat leaving = total generation .
- So , which means whatever the value of .
- The interior profile is fixed by , and : does not appear, so at the mid-point stays .
A numerical solve of the same rod (4,000 cells) gives the same mid-point temperature for and , equal to . The middle half of the rod has negative (a heat sink), which is why the mid-point sits below .
Always do the overall energy balance before solving the differential equation. In practice it is the first check on any thermal simulation: if the heat generated does not equal the heat leaving through the boundaries, the model or its boundary conditions are wrong.
4.5 kW/m² flows through a 0.1 m slab from an 80 °C face. How much entropy is generated?
Answer: 1.18 W/m²·K Official key: 1.12 to 1.24
- Fourier's law gives the cold-face temperature: , so . (.)
- At steady state the slab's entropy does not change, so entropy generated = entropy out − entropy in: .
- .
Temperatures must be in kelvin here: using °C gives a meaningless answer.
Heat flowing through any temperature difference destroys some ability to do work. Heat-exchanger designers trade this off: a bigger exchanger (smaller ΔT) wastes less exergy but costs more steel. Second-law (exergy) analysis of power and process plants starts from this formula.
A slab surface loses heat by convection and radiation. What is the temperature of its other face?
Answer: 319 K Official key: 315 to 323
- Heat leaving surface 2 by convection: .
- By radiation to 0 K surroundings: .
- At steady state the same 929.6 W/m2 is conducted through the slab: .
Leaving out radiation gives 314 K, just outside the key: radiation carries a quarter of the heat even at 300 K.
Electronics enclosures, solar panels and building walls all lose heat by convection and radiation in parallel. In CFD this is a combined convection + radiation boundary condition; at low wind speeds radiation is often a third or more of the total, so leaving it out makes parts look hotter than they are.
A very long square fin is replaced by a round fin of the same size. What is the ratio of heat transfer rates?
- (A)
- (B)
- (C)
- (D)
Answer: (A) Official key: A
- A very long (infinite) fin transfers , where is the perimeter and the cross-section area.
- Square of side : , , so . Circle of diameter : , , so .
- Ratio (square/circle) .
is the trap: it forgets the square root.
Pin-fin heat sinks for LEDs and power electronics come with both shapes. A long square pin moves about 27% more heat than a round pin of the same width, while round pins have lower pressure drop and work with airflow from any direction.
A 5 mm radius sphere at 400 °C is quenched in water at 20 °C. How long until its centre reaches 50 °C?
Answer: 25.39 s Official key: 24.50 to 26.00
- Check the lumped model first. Characteristic length , so . The sphere is nearly uniform inside, so the centre temperature is the sphere temperature.
- Time constant .
- , so .
Quench cooling rates decide the hardness of heat-treated steel parts, and the lumped model is how heat-treatment shops estimate them for small parts. Thermocouple response time is the same calculation, which is why fast probes use tiny beads: scales with the bead radius.
In the natural convection momentum equation, what does the term gβ(T − T∞) represent?
- (A)Ratio of inertial force to viscous force.
- (B)Ratio of buoyancy force to viscous force.
- (C)Viscous force per unit mass.
- (D)Buoyancy force per unit mass.
Answer: (D) Official key: D
Every term in this equation has units of acceleration (m/s2), that is, force per unit mass. The left side is inertia, is the viscous force, and comes from the density difference acting under gravity: the buoyancy force per unit mass.
(A) and (B) are ratios, which are dimensionless groups (Reynolds and Grashof numbers), not terms of a dimensional equation.
This Boussinesq buoyancy term is what drives natural convection in electronics enclosures, transformer oil, room air and solar water heaters. CFD codes switch it on as a 'gravity + Boussinesq' setting; forget it and a fan-less enclosure shows no airflow at all.
Fluid rises near one vertical plate and falls near the other. How do T₁, T₂ and T₃ compare?
- (A) and
- (B) and
- (C) and
- (D) and
Answer: (A) Official key: A
The only thing moving the fluid is buoyancy. Near plate 1 the fluid moves up, so it is lighter than the bulk: it has been heated, and . Near plate 2 it moves down, so it is heavier: it has been cooled, and .
Together: , option (A).
This loop of rising and falling fluid is how double-glazed windows lose heat across the gas gap, and how passive cooling works in electronics racks and solar chimneys. Window makers size the gap to keep this circulation weak.
A ball of ice at −7 °C sits in still air at 30 °C. Which way does the air flow?
- (A)Upward
- (B)Downward
- (C)NO motion
- (D)Sideward
Answer: (B) Official key: B
Air touching the ice is cooled from 30 °C towards −7 °C. Cooler air is denser, so it is heavier than the surrounding air and sinks. Warm air moves in from the sides and top to replace it. The overall flow is downward, a cold plume falling from the bottom of the sphere.
It is the mirror image of the rising plume above a hot body.
You can feel this falling cold air below an open freezer or an AC indoor unit. Cold-room and display-fridge designers place evaporators high for this reason, so the cold air falls through the products.
Which non-dimensional term is an estimate of the Nusselt number?
- (A)Ratio of internal thermal resistance of a solid to the boundary layer thermal resistance
- (B)Ratio of the rate at which internal energy is advected to the rate of conduction heat transfer
- (C)Non-dimensional temperature gradient
- (D)Non-dimensional velocity gradient multiplied by Prandtl number
Answer: (C) Official key: C
At the wall, all the heat passes by conduction through the still fluid layer: . Making and dimensionless (, ) gives : the non-dimensional temperature gradient at the surface.
(A) is the Biot number (it uses the solid's conductivity), (B) is the Péclet number ().
CFD codes compute exactly this way: from the temperature gradient in the first cell next to the wall. That's why thermal simulations need a fine mesh at the wall; a coarse first cell gives a wrong gradient and a wrong Nusselt number.
The free-stream velocity over a flat plate doubles in laminar flow. By what factor does the Nusselt number change?
- (A)
- (B)2
- (C)1.26
- (D)1
Answer: (A) Official key: A
- Laminar flat plate (Blasius / Pohlhausen): (the average Nu has 0.664, same exponents).
- and Pr doesn't change, so .
- .
uses the Prandtl exponent by mistake. In turbulent flow () the ratio would be about 1.74.
Doubling fan speed on a heat sink in laminar flow buys only about 41% more heat transfer while fan power rises roughly eightfold. That's why electronics cooling designers usually add fin area before they add airflow.
How long must a tube with 2000 W/m² heat flux be to heat 0.01 kg/s of water from 293 K to 353 K?
Use the specific heat of water as 4181 J kg−1 K−1.
Answer: 6.7 m Official key: 6.5 to 6.8
- Heat picked up by the water: .
- With a uniform flux that heat enters through the wall area : .
- .
No convection coefficient is needed: with a fixed flux, the energy balance alone sets the length. would only be needed to find the wall temperature.
Solar water-heater tubes, electric heaters and boiler tubes are sized from this energy balance first. Then the wall temperature is checked against the limit (boiling, scaling or the material rating) using a Nusselt-number correlation, as in the next question.
A 40 m pipe carries 1 kg/s under 15 kW/m² heat flux. What is the wall temperature at the exit?
Answer: 321 °C Official key: 317 to 324
- Heat added: .
- Exit bulk temperature: .
- Dittus-Boelter: , so .
- Wall minus bulk: , so .
With constant heat flux and fully developed flow, the wall-to-fluid difference is the same all along the pipe, so the wall is hottest at the exit.
Engineers check boiler tubes, nuclear fuel channels and electric heaters for exactly this: the exit wall temperature is the highest and decides whether the fluid boils, cokes or degrades. Dittus-Boelter is the quick hand check; CFD or the Gnielinski correlation is used when ±25% isn't good enough.
A cast-iron plate absorbs 600 W/m² of sunlight and has a low-temperature absorptivity of 0.21. How hot does it get?
Under the aforementioned conditions, the radiation equilibrium temperature of the plate is __________ °C (round off to the nearest integer).
Answer: 218 °C Official key: 210 to 225
- At low (infrared) wavelengths Kirchhoff's law gives emissivity = absorptivity = 0.21. That value applies to what the plate emits and to the infrared it absorbs from the 25 °C surroundings.
- Energy balance: .
- , so .
Ignoring the radiation coming back from the surroundings gives 201 °C, outside the key range.
This is the idea behind selective coatings on solar water heaters: absorb almost all the sunlight but emit little infrared, so the absorber runs hot. Black chrome and sputtered cermet coatings do this, with solar absorptivity above 0.9 and infrared emissivity near 0.1.
A surface's emissivity peaks at 6000 Å. At what temperature is its total emissivity highest?
Answer: 4830 K Official key: 4825 to 4835
- Total emissivity is the spectral emissivity averaged over the blackbody spectrum at . It is largest when most of the blackbody energy falls where is high, that is, when the blackbody peak sits on the emissivity peak.
- Emissivity peak: .
- .
The same idea guides the choice of pyrometers and infrared cameras: they are built for the wavelength band where the target emits most strongly. For the sun (about 5800 K) that peak is near 0.5 µm, which is why solar absorber coatings are tuned for the visible band.
A 6 m hemispherical furnace dome at 1200 K radiates to a black base at 300 K. What is the heat transfer rate?
Use Stefan-Boltzmann constant = W/(m2 K4)
Answer: 2727 kW Official key: 2720 to 2740
- Areas: base ; dome ().
- The flat base sees only the dome: .
- Two-surface enclosure, base black (no surface resistance): .
- .
Furnace and kiln designers use this resistance network to size burners and to predict how fast the load (here the base) heats. Above about 800 °C radiation carries most of the heat in a furnace, so emissivity of the refractory lining matters a lot.
In a black cylindrical furnace, how much heat does the 800 K bottom radiate to the 500 K curved wall?
All surfaces of the furnace can be assumed as black.
The Stefan-Boltzmann constant, W m−2 K−4.
Answer: 309.1 kW Official key: 308 to 310
- Summation rule for the flat bottom (it can't see itself): .
- Bottom area .
- Black surfaces exchange directly: .
The top at 1500 K doesn't enter this answer: with black surfaces each pair exchanges heat independently of the others.
View factors and the summation rule are the core of radiation modelling in furnaces, spacecraft and building daylighting. CFD radiation models (surface-to-surface, discrete ordinates) compute thousands of these factors, and engineers check the results against a hand calculation like this one.
A hot plate radiates to a room and to an insulated plate at right angles. What is the total radiation resistance?
Answer: 2.54 m⁻² Official key: 2.50 to 2.60
- Surfaces: 1 hot (), 2 insulated (reradiating: it sends back everything it receives), 3 the large room (acts black). , so . Equal areas give and .
- Surface resistance of 1: .
- From to the room there are two parallel paths: direct, ; and through surface 2, . In parallel: .
- Total: .
The 1000 K and 300 K temperatures are not needed for the resistance, only for the heat loss itself ().
Insulated refractory walls in furnaces act as reradiating surfaces: they don't absorb net heat but bounce it back towards the load. Furnace designers add them deliberately to make heating more even.
Blood is cooled from 37 °C to 25 °C by water warming from 4 °C to 18 °C. What is the exchanger's effectiveness?
Use the following fluid properties:
| Fluid | Density (kg/m3) | Specific heat (J/kg-K) |
|---|---|---|
| Blood | 1050 | 3740 |
| Water | 1000 | 4200 |
Answer: 0.42 Official key: 0.40 to 0.44
- Blood: , . .
- Water: . This is the smaller capacity rate, so .
- .
The water flow rate isn't given and isn't needed: the water's temperature rise over the maximum possible difference gives directly, because water is the fluid. Using the blood's () is the trap.
Heart-lung machines really do include a heat exchanger like this to cool and later rewarm the patient's blood. Its limits are unusual: perfusion guidelines cap the water-to-blood temperature difference (about 10 °C), because large gradients can release dissolved gas from the blood as bubbles.
150 kg/s of steam condenses at 200 °C, heating a fluid entering at 100 °C. With ε = 0.9, what is the fluid flow rate?
Answer: 1000 kg/s Official key: 1000 to 1000
- A condensing fluid stays at 200 °C, so its capacity rate is effectively infinite and the tube fluid is .
- , so and the tube fluid warms by 90 K.
- Heat released: , so .
Feedwater heaters in power plants work like this: steam bled from the turbine condenses on the shell side and preheats the boiler feedwater in the tubes. With one fluid changing phase, holds for any flow arrangement, which simplifies their design.
A condenser with 20,000 tubes at 1 kg/s each condenses 430 kg/s of steam at 50 °C. What is its effectiveness?
Answer: 0.625 Official key: 0.590 to 0.633
- Heat released by the steam: .
- Cooling water: , so it warms by , leaving at 42.50 °C.
- The steam is at constant temperature, so the water is : .
Roughly 1,000 MW goes into the cooling water here, typical for a plant of about 500 MW electrical output. Condenser performance sets the turbine back-pressure: a fouled condenser raises it and cuts the turbine's output.
A 1 m double-pipe exchanger: inner fluid 20 to 30 °C, outer fluid 60 to 40 °C. What is the heat transfer rate?
Answer: 50.5 W Official key: 49.5 to 51.5
- From the figure the flows are opposite (counterflow): fluid 2 (inside, ) goes 20 → 30 °C, fluid 1 (annulus, ) goes 60 → 40 °C.
- End temperature differences: and , so .
- Resistances per metre: inside ; wall ; outside .
- .
The copper wall () adds less than 0.01% of the resistance; the two convection films control everything.
This is why heat-exchanger designers attack the weaker film first: fins on the gas side of a car radiator or an AC coil, or turbulators inside tubes. Making the wall thinner or more conductive does almost nothing once it is metal.