GATE Civil: Numerical Methods (Newton-Raphson, trapezoidal rule, Euler steps, interpolation), previous-year questions solved

Every numerical methods question from the ten official GATE Civil papers of 2022-2026: 11 questions (7 NAT, 4 MCQ) on root finding (Newton-Raphson, fixed-point iteration, locating a root), the trapezoidal rule, Euler and backward-difference steps for ODEs, finite-difference formulas and Newton interpolation, each solved step by step exactly as the question prescribes and checked by a script against the official key. Engineering mathematics is Section 1 of the GATE CE syllabus and stays the same in GATE 2027.

Questions from the ten official GATE CE papers, sessions CE1 and CE2 of GATE 2022 (IIT Kharagpur), 2023 (IIT Kanpur), 2024 (IISc Bengaluru), 2025 (IIT Roorkee) and 2026 (IIT Guwahati); solutions by Skill-Lync. · Updated 17 Nov 2026 · 11 questions

GATE 2022 · CE1 · Q.29NAT · 1 markNumerical methods: roots, numerical integration, ODE steps and interpolation

What is x5 of the iteration x(n+1) = (x(n) + P/x(n))/2 for P = 10 starting from x1 = 1?

Consider the following recursive iteration scheme for different values of variable with the initial guess :

For , is obtained to be 1.414, rounded-off to three decimal places. For , is obtained to be 1.732, rounded-off to three decimal places.

If , the numerical value of is ________. (round off to three decimal places)

Answer: 3.162 Official key: 3.100 to 3.200

  1. The scheme is Newton-Raphson for : simplifies to the given formula, so it converges to . The two worked cases, 1.414 and 1.732, are and .
  2. With and : .
  3. .
  4. .
  5. , so 3.162, already equal to to three decimals.

Trap: stopping at after two updates instead of four. Note also the start: from a guess of 1 the first update overshoots to 5.5, and the convergence becomes quadratic (the number of correct digits roughly doubles each step) only once the iterate is close to the root.

Where this shows up in industry

Newton-Raphson iteration is how software solves equations with no closed form: the Colebrook equation for the pipe friction factor in water-supply design, the normal depth of a canal from Manning's equation, and the equilibrium iterations inside each load step of a nonlinear pushover analysis in ETABS or SAP2000.

GATE 2022 · CE1 · Q.48NAT · 2 marksNumerical methods: roots, numerical integration, ODE steps and interpolation

Using Euler's method with step 0.2, what is y at x = 1.4 for dy/dx = 4(x + 2) - y with y(1) = 3?

Consider the differential equation

For the initial condition at , the value of at obtained using Euler's method with a step-size of 0.2 is ________. (round off to one decimal place)

Answer: 6.4 Official key: 6.3 to 6.5

  1. Euler's method: with and .
  2. Step 1 from : , so .
  3. Step 2 from : , so 6.4.
  4. For comparison, the exact solution of this linear equation is , which gives . Euler overestimates here because the true slope falls along each step while Euler keeps the slope from the start of the step.

Trap: stopping after one step (4.8) or quoting the exact value (6.2), which is outside the accepted range; the question asks for the Euler estimate.

Where this shows up in industry

Euler's method and its refinements march time-dependent problems forward step by step: flood routing through a reservoir with the storage equation, consolidation settlement with time, and the time-history response of a building to an earthquake record are all solved this way, with smaller steps where accuracy matters.

GATE 2023 · CE1 · Q.36MCQ · 2 marksNumerical methods: roots, numerical integration, ODE steps and interpolation

Does the trapezoidal rule overestimate or underestimate the area under a smooth hump-shaped curve?

A function , that is smooth and convex-shaped between interval is shown in the figure. This function is observed at odd number of regularly spaced points. If the area under the function is computed numerically, then ________.

Smooth hump-shaped curve f(x) between x_l and x_u, observed at 11 regularly spaced points; the approximated function joins those points with straight chords that sit just below the curvef(x)(0,0)xlxuxActual functionApproximated function
  1. (A)the numerical value of the area obtained using the trapezoidal rule will be less than the actual
  2. (B)the numerical value of the area obtained using the trapezoidal rule will be more than the actual
  3. (C)the numerical value of the area obtained using the trapezoidal rule will be exactly equal to the actual
  4. (D)with the given details, the numerical value of area cannot be obtained using trapezoidal rule

Answer: (A) the numerical value of the area obtained using the trapezoidal rule will be less than the actual Official key: A

  1. The trapezoidal rule joins neighbouring observed points with straight chords and adds up the trapezia under them.
  2. The curve in the figure is a hump: it bends downwards, . Every chord of such a curve lies below the arc, so each strip loses a sliver of area. The error term agrees: error , meaning actual minus computed is positive.
  3. Hence (A): the trapezoidal value is less than the actual area. Example: from 0.45 to 2.75 with 11 points gives 1.8167 against the exact 1.8247, 0.44% low.
  4. (B) holds for a curve bending upwards; (C) only for a straight line; (D) is false because the trapezoidal rule works with any number of points. The odd count matters for Simpson's 1/3 rule, not for the trapezoidal rule.

Trap: the word convex. In the calculus sense a convex function has and the trapezoidal rule would overestimate; the figure shows the opposite shape, so read the curve, not the adjective.

Where this shows up in industry

Trapezoidal and Simpson's rules are how site engineers compute earthwork volumes from cross-sections at regular chainages on a highway or canal, and how discharge is computed from velocity readings across a river section. Knowing the bias direction tells you whether a quantity estimate is likely to be short.

GATE 2023 · CE1 · Q.52NAT · 2 marksNumerical methods: roots, numerical integration, ODE steps and interpolation

One backward-difference step of du/dt + 2tu^2 = 1 from u = 1.75 at t = 3.14 s: what is u_n - u_(n-1)?

The differential equation,

is solved by employing a backward difference scheme within the finite difference framework. The value of at the time-step, for some , is 1.75. The corresponding time () is 3.14 s. Each time step is 0.01 s long. Then, the value of is ________ (round off to three decimal places).

Answer: -0.151 Official key: -0.152 to -0.149

  1. The backward (implicit) scheme evaluates the equation at the new time level s:
  2. .
  3. Substitute: , which rearranges to .
  4. Positive root: .
  5. -0.151. (Using s in the term gives -0.1506, also inside the key.)

Trap: the explicit forward step uses the old value, , outside the key. The backward scheme is implicit: the unknown sits inside the nonlinear term, so a quadratic has to be solved.

Where this shows up in industry

Implicit time stepping is what keeps finite element consolidation, transient seepage and flood routing models stable with practical time steps. PLAXIS consolidation analyses and HEC-RAS unsteady flow use implicit schemes, solving nonlinear equations at every step just as this question does by hand.

GATE 2024 · CE1 · Q.11MCQ · 1 markNumerical methods: roots, numerical integration, ODE steps and interpolation

In which range does the smallest positive root of x^5 - 5x^4 - 10x^3 + 50x^2 + 9x - 45 = 0 lie?

The smallest positive root of the equation

lies in the range

  1. (A)
  2. (B)
  3. (C)
  4. (D)

Answer: (A) Official key: A

  1. Group the terms in pairs: .
  2. So , and the quartic factors as .
  3. The five roots are . Check: .
  4. The smallest positive root is , which lies in : (A).

Trap: (B) holds the next root, , and the largest root 5 lies in none of the ranges. If you cannot spot the factors, a sign check brackets the answer: and , so a root lies between 0 and 2, and bisection would close in on it.

Where this shows up in industry

Polynomial roots appear whenever a civil engineer solves a characteristic equation, for example the cubic in flow depth from the specific energy equation of an open channel, or the frequency equation of a multi-storey frame. Bracketing a root by a sign change is the first step of the bisection and Newton routines inside design spreadsheets.

GATE 2024 · CE1 · Q.32NAT · 1 markNumerical methods: roots, numerical integration, ODE steps and interpolation

Estimate f(1.5) by second-order Newton interpolation from f(1) = 0, f(2) = 0.3010 and f(3) = 0.4771

Consider the data of given in the table.

012
123
00.30100.4771

The value of estimated using second-order Newton’s interpolation formula is ____________ (rounded off to 2 decimal places).

Answer: 0.17 Official key: 0.16 to 0.18

  1. Equal spacing , so use Newton’s forward formula with .
  2. Differences: , , .
  3. , so 0.17.

Trap: stopping at the first-order term gives 0.1505, which rounds to 0.15, outside the key. The data are , and the true value shows the remaining error of a quadratic fitted to a curved function.

Where this shows up in industry

Engineers interpolate in tables every day: IS code tables for permissible stresses or design coefficients, stage-discharge and area-capacity tables of reservoirs, and IDF rainfall data. A second-order fit through three points is more accurate than straight-line interpolation where the tabulated curve bends.

GATE 2024 · CE2 · Q.13MCQ · 1 markNumerical methods: roots, numerical integration, ODE steps and interpolation

Which is the fourth-order central divided difference formula for the second derivative with step h?

The second derivative of a function is computed using the fourth-order Central Divided Difference method with a step length .

The CORRECT expression for the second derivative is

  1. (A)
  2. (B)
  3. (C)
  4. (D)

Answer: (A) Official key: A

  1. Any difference formula for a derivative must give zero for a constant function, so its coefficients must add up to zero. A: . B and C add up to 2 and 2, so they fail at once.
  2. The formula must also be symmetric about (a central difference for an even derivative), which B and C break.
  3. D is A with every sign flipped: it gives . For (where ), A gives while D gives .
  4. Taylor expansion of A matches exactly for to , so its error is of order : (A).

Trap: memorising the numbers 1, 16, 30 but not the signs. Check the coefficient sum and test on ; on at with , A is within of the true value.

Where this shows up in industry

Higher-order finite differences are used in finite-difference models of beam and plate deflection, in consolidation and groundwater flow codes, and when engineers differentiate measured deflection or strain data along a girder during load testing to estimate curvature and bending moment.

GATE 2025 · CE1 · Q.49NAT · 2 marksNumerical methods: roots, numerical integration, ODE steps and interpolation

Euler method with h = 0.5 for dy/dx = y + 2x - x^2, y(0) = 1: value of y at x = 1

Consider the differential equation given below. Using the Euler method with the step size () of 0.5, the value of at is equal to __________ (rounded off to 1 decimal place).

Answer: 2.6 Official key: 2.4 to 2.8

  1. Euler step: with and . Two steps reach .
  2. Step 1, from : ; at .
  3. Step 2, from : ; .
  4. So 2.6.
  5. For comparison, the exact solution is (check: ), giving . Euler's large step underestimates because the slope keeps rising within each step.

Trap: answering with the exact value ; the question asks for the Euler estimate. Taking one step of instead of two steps of 0.5 gives .

Where this shows up in industry

Explicit time stepping of this kind drives flood routing, reservoir level simulation and groundwater models used by water resources departments; the step size controls the error, exactly as the gap between 2.6 and the exact 3.72 shows, so modellers check results by halving the step.

GATE 2026 · CE1 · Q.48NAT · 2 marksNumerical methods: roots, numerical integration, ODE steps and interpolation

Starting from x = 0.5, what is the Newton-Raphson second approximation to the root of e^-x - x?

Starting with the first approximation as , the second approximation for the root of the following function by the Newton-Raphson method is ______ (rounded off to two decimal places).

Answer: 0.57 Official key: 0.55 to 0.58

  1. and .
  2. At : , so and .
  3. Newton step: .
  4. Rounded: 0.57. One more step gives , already the root 0.567143; Newton converges quadratically.

Trap: the question calls 0.5 the first approximation, so the second approximation is the first Newton step. A sign slip () gives 0.43, moving away from the root.

Where this shows up in industry

Newton-Raphson is the workhorse for implicit equations in civil design: normal depth from Manning's equation for a trapezoidal canal, friction factor from the Colebrook equation for a water main, and the equilibrium iterations inside nonlinear and pushover analysis in ETABS or SAP2000.

GATE 2026 · CE1 · Q.49NAT · 2 marksNumerical methods: roots, numerical integration, ODE steps and interpolation

For data (-2, 28), (1, 4), (2, 16), what is the value of the second-degree interpolating polynomial at x = 0?

Values of for different values of are tabulated below.

12
28416

If a second-degree interpolating polynomial is used to represent , the value of is ______ (rounded off to the nearest integer).

Answer: 2 Official key: 2 to 2

  1. Lagrange basis values at for the nodes :
  2. ; ; .
  3. .
  4. Check: passes through all three points (, , ) and its constant term is 2.

Trap: straight-line interpolation between and gives , ignoring the curvature that the third point supplies. Solving the three equations , , directly also works: subtracting the first from the third gives , then and , and is simply .

Where this shows up in industry

Interpolating polynomials are how engineers read between tabulated values, for example between rows of design tables in IS codes, between gauged stages when building a stage-discharge rating curve, or between survey levels when setting out a vertical curve.

GATE 2026 · CE2 · Q.12MCQ · 1 markNumerical methods: roots, numerical integration, ODE steps and interpolation

For a fifth-degree polynomial with positive coefficients, do forward and backward differences under or overestimate?

A fifth-degree polynomial in is defined for . All coefficients of the polynomial are positive. The first derivative of the polynomial is obtained numerically at a point by using the first-order forward as well as the first-order backward difference methods. Identical step lengths are used for both the methods.

Following statements are made.

(I) Forward difference method underestimates the true derivative.
(II) Backward difference method overestimates the true derivative.

Which one of the following options is CORRECT?

  1. (A)Both statements (I) and (II) are FALSE.
  2. (B)Both statements (I) and (II) are TRUE.
  3. (C)Statement (I) is TRUE and statement (II) is FALSE.
  4. (D)Statement (I) is FALSE and statement (II) is TRUE.

Answer: (A) Both statements (I) and (II) are FALSE. Official key: A

  1. With all coefficients positive and : , so is convex (curving upward).
  2. Forward difference by Taylor: . It overestimates, so (I) is FALSE.
  3. Backward difference: . It underestimates, so (II) is FALSE.
  4. Example: at , : true slope 15, forward 17.156, backward 13.144.

Trap: the statements reverse the sign of the error. For a convex curve the chord ahead of the point is steeper than the tangent and the chord behind is flatter, so the answer is (A) Both statements (I) and (II) are FALSE.

Where this shows up in industry

Engineers take first differences of measured data all the time: settlement rate from settlement-time records under an embankment, gradient from a longitudinal level survey, curvature from inclinometer readings along a pile or diaphragm wall; knowing the sign of the truncation error tells you which way the estimate is biased.

From GATE prep to a structural design career

Talk to a counsellor.

Talk to a career counsellor

Free 20-minute call. We'll map the right course to where you want to work.

By submitting you agree to our Terms and Privacy Policy.

Talk to a career counsellor

Talk to a career counsellor

Free 20-minute call. We'll map the right course to where you want to work.

By submitting you agree to our Terms and Privacy Policy.