Chemegate
Home/Topics/Applied/Engineering Mathematics
Topic Guide

Engineering Mathematics
for GATE CH

Linear algebra, calculus, differential equations, probability, and numerical methods for the GATE Chemical Engineering paper.

Practice

Solved PYQs →

Overview

Engineering Mathematics is the one GATE CH section that isn't chemical engineering at all, it's the shared mathematics syllabus (linear algebra, calculus, differential equations, probability, numerical methods) common across every GATE engineering discipline, scoped down to what a chemical engineer actually needs. Combined with General Aptitude, it makes up a fixed, syllabus-guaranteed chunk of the paper every single year, which makes it one of the most reliable places to bank marks with focused, calculation-heavy practice rather than conceptual chemical engineering judgment.

The syllabus covers linear algebra (matrices, eigenvalues/eigenvectors, systems of linear equations), calculus (limits, partial derivatives, gradient/directional derivatives, multiple integrals, vector calculus), differential equations (first-order and linear higher-order ODEs, plus common named equations like Euler-Cauchy), complex numbers, probability and statistics (distributions, conditional probability, expectation), and numerical methods (root-finding via Newton-Raphson, numerical integration via the trapezoidal rule, and basic curve-fitting via least squares).

Unlike the core chemical engineering sections, near-zero physical intuition is needed here, success is almost entirely a function of mechanical fluency with standard techniques and formulas, practiced under time pressure. Because this section repeats a fairly narrow set of question archetypes year over year (eigenvalues of a small matrix, Newton-Raphson to a fixed number of iterations, a standard probability distribution calculation), it rewards drilling past-year-style problems more than any other section on the exam.

Real GATE CH PYQ Frequency (2024–2026)

2024202520263-Yr Avg
10999.3

Computed directly from 28 real questions tagged to this topic across our GATE CH 2024–2026 archive, averaging 9.3 questions/year. Browse the underlying 28 questions for this topic in the PYQ archive, or see the full topic weightage comparison across all topics.

GATE Weightage

Engineering Mathematics typically contributes 8–10 questions (about 13–15 marks) to the GATE CH paper, one of the single largest scoring blocks, on par with or exceeding any individual core chemical engineering subject, alongside a further fixed 15 marks of General Aptitude.

Sub-areaApprox. Marks
Linear algebra (matrices, eigenvalues, systems)~3–4
Calculus (multivariable, vector calculus)~3–4
Differential equations & complex numbers~3–4
Probability, statistics & numerical methods~3–4

Sub-area split is a directional estimate (our archive doesn't tag marks at this granularity), for the real, computed topic-level total, see "Real GATE CH PYQ Frequency" above.

Key Subtopics

1

Linear algebra

Matrix operations, determinants, rank, systems of linear equations (unique/infinite/no solution via rank comparison), and eigenvalues/eigenvectors of small matrices.

2

Differential calculus & limits

Limits, continuity, partial derivatives, and the gradient/directional derivative for scalar fields of several variables, the calculus of functions like temperature or concentration fields.

3

Integral calculus & vector calculus

Multiple integrals, line integrals, and vector calculus identities (divergence, curl, Green's/Stokes' theorem) frequently framed around a vector field.

4

Complex numbers

Algebra of complex numbers, polar/exponential form, and roots of complex numbers, also foundational for the Laplace-transform-heavy Process Control section.

5

Ordinary differential equations

First-order linear/separable ODEs, higher-order linear ODEs with constant coefficients, and named forms like the Euler-Cauchy equation.

6

Probability & statistics

Standard probability distributions (binomial, Poisson, normal), conditional probability and Bayes' theorem, and basic descriptive statistics (mean, variance, expectation).

7

Numerical methods

Root-finding via Newton-Raphson (and its convergence behavior), numerical integration via the trapezoidal rule, and simple least-squares curve fitting.

Essential Formulas

Full formula reference →

det⁡(A−λI)=0\det(A - \lambda I) = 0

Characteristic equation, solve for eigenvalues λ of matrix A

(A−λI)v=0(A-\lambda I)v = 0

Eigenvector equation, solve for the eigenvector v corresponding to eigenvalue λ

∇f=(∂f∂x,∂f∂y,∂f∂z)\nabla f = \left(\dfrac{\partial f}{\partial x}, \dfrac{\partial f}{\partial y}, \dfrac{\partial f}{\partial z}\right)

Gradient of a scalar field, points in the direction of steepest increase

Du^f=∇f⋅u^D_{\hat u}f = \nabla f \cdot \hat u

Directional derivative of f in the direction of unit vector û

z=reiθ=r(cos⁡θ+isin⁡θ)z = re^{i\theta} = r(\cos\theta + i\sin\theta)

Polar/exponential form of a complex number (Euler's formula)

y′′+ay′+by=0  ⇒  y=C1em1x+C2em2xy'' + a y' + by = 0 \;\Rightarrow\; y = C_1e^{m_1x}+C_2e^{m_2x}

General solution of a second-order linear homogeneous ODE with constant coefficients (distinct real roots m1, m2 of the auxiliary equation)

x2y′′+axy′+by=0x^2 y^{\prime\prime} + ax y^{\prime} + by = 0

Euler-Cauchy equation, solved via the substitution y = x^m, giving an algebraic equation in m

xn+1=xn−f(xn)f′(xn)x_{n+1} = x_n - \dfrac{f(x_n)}{f^{\prime}(x_n)}

Newton-Raphson iteration for root-finding, converges quadratically near a simple root

∫abf(x) dx≈h2[f(x0)+2∑i=1n−1f(xi)+f(xn)]\int_a^b f(x)\,dx \approx \dfrac{h}{2}\left[f(x_0)+2\sum_{i=1}^{n-1}f(x_i)+f(x_n)\right]

Trapezoidal rule for numerical integration, step size h = (b−a)/n

P(A∣B)=P(B∣A)P(A)P(B)P(A|B) = \dfrac{P(B|A)P(A)}{P(B)}

Bayes' theorem for conditional probability

Visual Reference

A Keuffel & Esser log-log duplex decitrig slide rule
Before calculators, this was numerical methods — a slide rule computes to slide-rule accuracy exactly what the Newton-Raphson and trapezoidal-rule questions here compute to several decimal places. Daderot, CC0, via Wikimedia Commons.

Derivations & Physical Insight

Why Newton-Raphson Converges Quadratically Near a Simple Root

Newton-Raphson approximates a function f(x) near the current guess xn by its tangent line, a first-order (linear) Taylor expansion, and finds where that tangent line crosses zero as the next guess. Writing the true root as x* = xn + εn (where εn is the current error) and Taylor-expanding f around xn to second order reveals how the error shrinks from one iteration to the next.

Substituting the Taylor expansion of f(x*) = 0 into the Newton-Raphson update formula and simplifying shows that the new error εn+1 is proportional to the square of the old error εn (scaled by f''(x*)/2f'(x*)), not proportional to εn itself. That squared relationship is exactly what "quadratic convergence" means: once the guess is reasonably close to a simple root, the number of correct digits roughly doubles with every iteration, which is why Newton-Raphson converges dramatically faster than a simple bisection method (which only halves the error interval each time), provided f'(x) isn't close to zero near the root and the initial guess is close enough to avoid divergence.

εn+1≈f′′(x∗)2f′(x∗) εn2\varepsilon_{n+1} \approx \dfrac{f^{\prime\prime}(x^{*})}{2f^{\prime}(x^{*})}\,\varepsilon_n^2

Why Eigenvectors Are the Directions a Matrix Leaves Unrotated

A matrix A acting on a general vector both rotates and stretches/shrinks it. An eigenvector v is defined as a special direction that A does not rotate at all, applying A to v just scales it by a factor λ (the eigenvalue), so Av = λv exactly, with no change in direction (only length, and possibly sign if λ is negative).

This is solved by rewriting Av = λv as (A − λI)v = 0, a homogeneous linear system that only has a nonzero solution v when the coefficient matrix (A − λI) is singular, i.e., when its determinant is zero. That determinant condition, det(A − λI) = 0, is the characteristic equation, and its roots are exactly the eigenvalues for which a nontrivial "unrotated direction" exists. In chemical engineering, this recurs directly in reactor stability analysis and process control (where the eigenvalues of a system matrix determine whether small perturbations grow or decay over time) and in solving coupled linear ODEs (like multiple interacting CSTRs), where eigenvectors give the natural, decoupled modes of the system.

Worked Practice Problems

Original practice problems in the GATE CH style, not copied from any question bank. Work them before reading the solution.

1-mark · NAT

Eigenvalues of a 2×2 Matrix

Problem: Find the eigenvalues of the matrix A = [[4, 1], [2, 3]].

Given: A = [[4, 1], [2, 3]].

det⁡(A−λI)=(4−λ)(3−λ)−(1)(2)=0\det(A-\lambda I) = (4-\lambda)(3-\lambda) - (1)(2) = 0
λ2−7λ+12−2=λ2−7λ+10=0\lambda^2 - 7\lambda + 12 - 2 = \lambda^2 - 7\lambda + 10 = 0
(λ−5)(λ−2)=0(\lambda-5)(\lambda-2) = 0

Answer: λ = 5 and λ = 2

2-mark · NAT

Newton-Raphson, One Iteration Toward a Root

Problem: Use one iteration of Newton-Raphson, starting from x0 = 2, to find an improved estimate of the root of f(x) = x³ − 2x − 5.

Given: f(x) = x³ − 2x − 5, f'(x) = 3x² − 2, x0 = 2.

f(x0)=23−2(2)−5=8−4−5=−1f(x_0) = 2^3 - 2(2) - 5 = 8 - 4 - 5 = -1
f′(x0)=3(2)2−2=12−2=10f'(x_0) = 3(2)^2 - 2 = 12-2 = 10
x1=x0−f(x0)f′(x0)=2−−110x_1 = x_0 - \dfrac{f(x_0)}{f^{\prime}(x_0)} = 2 - \dfrac{-1}{10}

Answer: x1 = 2.1 (the true root is ≈2.0946, so one iteration already lands within 0.006, illustrating quadratic convergence).

Topic-wise PYQ Frequency

High

Linear algebra (eigenvalues, systems of equations)

GATE 2024 Q26/Q30 (linear algebra, eigenvalues) and GATE 2025 Q45 and GATE 2026 Q23 (eigenvalues) confirm this is tested every year, often multiple questions.

High

Differential equations & Euler-Cauchy

GATE 2024 Q52 and GATE 2025 Q44 (differential equations) and GATE 2026 Q62 (Euler-Cauchy equation) show consistent testing.

High

Probability & statistics

GATE 2024 Q12 (probability distributions), GATE 2025 Q14/Q37 (probability), and GATE 2026 Q34 (conditional probability) confirm this is a reliable, recurring numerical.

Medium

Numerical methods (Newton-Raphson, trapezoidal rule)

GATE 2024 Q50 and GATE 2025 Q46 (Newton-Raphson) and GATE 2026 Q35 (trapezoidal rule) and Q63 (least squares) show regular testing.

Medium

Vector calculus & complex numbers

GATE 2024 Q53 (directional derivatives), GATE 2025 Q13/Q36 (complex numbers), and GATE 2026 Q21/Q37/Q40/Q41 (gradient, vector fields, complex numbers, polar coordinates) confirm frequent testing.

Recommended Study Order

  1. 1

    1. Linear algebra

    The highest-yield subtopic and a prerequisite for eigenvalue-based reasoning that appears in Process Control and Reaction Engineering as well.

  2. 2

    2. Differential equations & complex numbers

    Complex numbers are needed for solving linear ODEs with complex characteristic roots, and both feed directly into the Laplace-transform work in Process Control.

  3. 3

    3. Calculus (multivariable & vector calculus)

    Builds on standard single-variable calculus, focus extra practice on gradient, directional derivative, and line/surface integral mechanics.

  4. 4

    4. Probability & statistics

    A self-contained, high-frequency subtopic, practice standard distributions and conditional probability until fast.

  5. 5

    5. Numerical methods

    Quick to learn and mechanical, Newton-Raphson and the trapezoidal rule are two of the highest marks-per-minute-of-study formulas on the whole exam.

Common Pitfalls

✗ Sign errors when expanding a 3×3 determinant for the characteristic equation.

✓ Use cofactor expansion consistently and double-check signs (+,−,+ pattern along a row/column), this is the single most common source of wrong eigenvalues under time pressure.

✗ Forgetting that Newton-Raphson can diverge or converge to the wrong root if the initial guess is poor or f'(x) is near zero.

✓ Sanity-check the first iteration's direction against a rough sketch or estimate of the function before trusting the numerical result.

✗ Mixing up permutations and combinations, or conditional probability direction (P(A|B) vs. P(B|A)), under exam time pressure.

✓ Write out Bayes' theorem explicitly rather than reasoning informally whenever a problem gives a conditional probability in one direction and asks for the reverse.

✗ Using the wrong number of trapezoidal-rule intervals or forgetting the factor-of-2 weighting on interior points.

✓ Double-check the trapezoidal rule's coefficient pattern (endpoints weight 1, all interior points weight 2) before computing, especially with an odd number of intervals.

Put It Into Practice

Work the eigenvalue and Newton-Raphson problems above by hand without a calculator shortcut, then drill the same formula-recognition speed across the Engineering Mathematics questions in the full GATE previous-year test set.

Take a Practice Test →Browse Solved PYQs →

Keep Exploring

Related

Process Control

Transfer functions, first/second order dynamics, PID control, stability, and tuning for GATE CH.

Open →

Plant Design & Economics

Cost estimation, depreciation, payback period, NPV, and rate of return for GATE CH.

Open →

GATE CH Formula Sheet

Every core GATE Chemical Engineering formula in one place, grouped by topic with variable definitions.

Open →

GATE Practice Tests by Year

Full-length, timed GATE CH practice tests (2024–2026) with instant, step-by-step-explained results.

Open →