GATE Mechanical: Calculus and Vector Calculus (Engineering Mathematics), previous-year questions solved, part 2
Part 2 of 2 of the Calculus and Vector Calculus hub: 10 more official GATE ME questions on Fourier series and vector calculus: gradient, divergence, curl and flux, each solved step by step and checked against the official answer key. Part 1 covers limits, maxima and minima, Taylor series and definite and double integrals.
Questions from official GATE ME papers (IIT Kharagpur 2022, IIT Kanpur 2023, IISc Bengaluru 2024, IIT Roorkee 2025, IIT Guwahati 2026); solutions by Skill-Lync. · Updated 16 Nov 2026 · 10 of 19 questions on this page
Fourier series of x³ on −1 ≤ x < 1: sine terms, cosine terms or both?
- (A)only sine terms
- (B)only cosine terms
- (C)both sine and cosine terms
- (D)only sine terms and a non-zero constant
Answer: (A) Official key: A
- is odd: . On the symmetric interval , its product with an even function is odd and integrates to zero, so every , including the constant , is zero.
- The product with is even, so the survive: , with .
- Only sine terms: option (A).
Trap (D): a non-zero constant would mean a non-zero average over one period, but an odd function's average over is zero. Numerically, to all came out below and to matched the formula above.
Fourier series are how vibration engineers read signals. An FFT of an accelerometer trace splits engine, gearbox or bearing vibration into harmonics, and symmetry shows up directly: a waveform with odd symmetry has only sine content, one with half-wave symmetry only odd harmonics. Order analysis in NVH testing and harmonic analysis in ANSYS both rest on this decomposition.
Square wave between 4 and 0, 1 s at each level: what is the constant Fourier term?
- (A)1
- (B)2
- (C)3
- (D)4
Answer: (B) Official key: B
- The constant term of a Fourier series is the average value of the function over one period.
- Period s: for 1 s and for 1 s. Average : option (B).
Trap (D): with the series written as , the coefficient is , and the constant term is . Typing itself gives 4. Averaging the wave numerically over a period gives 2.
The constant (DC) term is the mean of a periodic signal: the average torque of an engine over a cycle, the mean voltage of a PWM motor drive (duty cycle times supply voltage, here 50% of 4), the mean stress in a fatigue load cycle. Fatigue checks use exactly this split: mean stress (the DC term) plus alternating stress (the harmonics).
Component of a = 5i + 7j + 2k orthogonal to b = 3i − j + 6k, in their plane: what is its magnitude?
Answer: 8.32 Official key: 7.90 to 8.70
- Split into a part along and a part perpendicular to it. Both parts lie in the plane of and .
- , , .
- Parallel part: .
- Perpendicular part, by Pythagoras: , which rounds to 8.32.
Same answer another way: , since the cross product's magnitude is . Trap: the parallel part 2.95 is the projection, the opposite of what is asked.
Splitting a vector into components along and across a direction is everyday mechanics: normal and tangential force on an inclined contact (friction uses the normal part), axial and shear load on a bolt, and the normal and shear stress on a weld throat or a fibre direction in a composite. FEA post-processors do this projection when you ask for stress in a local coordinate system.
Flux of ∇φ, φ = (x² + y² + z²)/2, out of a unit sphere
- (A)
- (B)
- (C)
- (D)0
Answer: (A) Official key: A
- , the position vector.
- Direct route: on the unit sphere (radius 1), so everywhere. The integral is the surface area, .
- Divergence theorem route: , times the volume gives : option (A).
Trap (C): is the sphere's volume, what you get if you forget the divergence is 3, not 1. A direct numerical surface integral over the sphere gives 12.566460 ().
The divergence theorem is the foundation of the finite volume method used by ANSYS Fluent, OpenFOAM and STAR-CCM+: the change of mass, momentum or energy inside each cell is computed from the fluxes through its faces. It is also Gauss's law for heat: the total heat leaving a closed surface equals the heat generated inside it.
Closed surface integral of z dx dy over a sphere of radius R
- (A)0
- (B)
- (C)
- (D)
Answer: (C) Official key: C
- In surface-integral notation, means the flux of the field : is the projection of the surface element.
- Divergence theorem: , so the integral equals the enclosed volume, : option (C).
Trap (A): the integrand is odd, which tempts you to answer 0. But is odd too (negative on the lower half, where is negative), so is positive everywhere and nothing cancels. A numerical surface integral of for gives 4.18899 against .
Flux integrals like this give a body's volume, buoyancy and pressure force from its surface alone: integrating pressure times the normal over a hull's wetted surface gives Archimedes' buoyancy. CAD systems compute volume, mass and centroid of a solid model the same way, by integrating over its boundary surfaces.
Flux of F = 3x i + 5y j + 6z k out of a unit cube centred at (1, 2, 3)
- (A)14
- (B)27
- (C)28
- (D)31
Answer: (A) Official key: A
- Divergence theorem: , a constant.
- Flux : option (A).
The centroid is a distraction: with a constant divergence, only the volume matters. Face by face it is just as quick: the two faces are 1 apart, so their net flux is ; likewise and for and . Trap (D): plugging the centroid into gives , and adding the components gives 31. A face-by-face numerical integral gives 14 with the cube at and 14 at the origin.
This is a finite volume cell in miniature. A CFD solver checks that, for every cell, the net flux through its faces balances what is created or stored inside; a constant divergence means a uniform source, like uniform heat generation in a battery cell or a resistive conductor. Mass imbalance across cells is one of the residuals you watch converge in Fluent or OpenFOAM.
Flux of B = x i + y j − 2z k through the curved surface of a cone (h = 2, r = 3)
If denotes the unit outward normal to the curved surface of the cone, the value of the integral equals _________ . (Answer in integer)
Answer: 0 Official key: -0.001 to 0.001
- Close the surface with the flat base and use the divergence theorem: , so the net flux out of the whole closed cone (curved surface plus base) is zero.
- Base: it lies in with outward normal , so everywhere on it. The base carries no flux.
- So the curved surface carries none either: .
It is zero by cancellation, not because nothing flows. The horizontal part pushes out through the curved surface (), and the vertical part pulls exactly back in. A direct numerical integral over the curved surface gives and for the two parts, which add to less than : zero to rounding.
A divergence-free field like this one is what an incompressible flow looks like: whatever enters a region through one face must leave through another. CFD solvers for water, oil and low-speed air enforce exactly this on every cell (the continuity equation), and a non-zero net flux through a closed control volume is the mass-imbalance residual you watch converge.
What is the divergence of the curl of a vector field?
- (A)the magnitude of this vector field
- (B)the argument of this vector field
- (C)the magnitude of the curl of this vector field
- (D)zero
Answer: (D) Official key: D
- Write in components and take its divergence:
- The six mixed second derivatives cancel in pairs ( against , and so on) for any field with continuous second derivatives. The result is identically zero: option (D).
Companion identity worth memorising: , the curl of a gradient is zero. We checked the first one numerically on the field at three points: the curl itself is far from zero, its divergence is zero to rounding error.
Because , a magnetic field written as the curl of a vector potential automatically has no divergence, which is how electromagnetic FEA (motor and transformer design in tools like Ansys Maxwell) guarantees physical fields. In fluids, the vorticity (the curl of velocity) is divergence-free for the same reason, a fact used in vortex methods and in checking CFD results.
Directional derivative of f = x² + xy² at (1, 0) along (i + √3 j)/2
Answer: 1.0 Official key: 0.9 to 1.1
- Gradient: . At : .
- The direction already has length , so no normalising is needed.
- , so the answer is 1.0.
Trap: is the largest rate of change, along the gradient itself, not along the given direction (60° away from it, and ). A finite-difference step along gives 1.000000.
Directional derivatives tell you how fast temperature, stress or pressure changes along a particular line: the temperature gradient through a weld along the plate, the stress gradient at a notch used in fatigue (notch sensitivity), or a sensitivity in design optimisation, where the solver needs the change in mass or stress along a proposed design change.
Is the gradient ∇φ perpendicular or parallel to a surface of constant φ?
- (A)always perpendicular to the surface of constant
- (B)always parallel to the surface of constant
- (C)the minimum rate of change of scalar
- (D)always zero
Answer: (A) Official key: A
- Move a small step within a surface of constant . Then doesn't change: .
- So is perpendicular to every direction lying in the surface: it is normal to the surface of constant , option (A).
- (C) is backwards: is the maximum rate of change, reached along . (D) holds only at stationary points.
We checked it on the ellipsoid : at 200 random points, the gradient's dot product with both surface tangent vectors was zero to of their lengths.
Heat flows along , perpendicular to isotherms, which is why heat-flux arrows in a thermal FEA plot cross the temperature contours at right angles. CAD uses the same fact to get the normal of an implicit surface, and contact algorithms in crash and forming simulation use surface normals from gradients to push penetrating nodes back out.