GATE ME common mistakes
GATE ME · Understand the traps and how to avoid them.
Spending the first 40–50 minutes grinding through the 1-mark numerical questions in order, then discovering at minute 150 that the 2-mark questions and the General Aptitude section are untouched.
GATE ME has 65 questions in 180 minutes, so the average budget is under 3 minutes per question, but 2-mark numericals take 5–8 minutes each. Candidates who work front-to-back without a triage pass run out of clock on the highest-value questions, and GA (15 marks, roughly 23% of the paper) is easy to defer and then never reach.
What to do instead
Do a 5-minute read-only pass: mark each question A (under 2 min), B (2–5 min), or C (long/uncertain). Solve all A-tier questions first across the whole paper, then GA, then B, and only then C. Practise this on at least three full-length timed mocks so the triage itself becomes fast.
Losing 2–5 marks per paper to sign, unit, and rounding slips: reporting stress in MPa when the question asks for N/mm², giving a natural frequency in rad/s when Hz is asked, forgetting the negative sign on a compressive stress, or leaving an answer to 2 decimals when 3 are required.
GATE numerical-answer questions (NAT) have no options to sanity-check against, and many candidates derive the correct expression but rush the final substitution. Mixed units (mm vs m, kN vs N, rpm vs rad/s) are deliberately embedded in machine-design and fluid problems.
What to do instead
Adopt a fixed final-90-seconds routine for every NAT: write the unit of the target quantity, convert all inputs to a single consistent system (SI base or N-mm), re-substitute, and check order of magnitude against a physical estimate. Keep a personal 'unit trap' list (rpm→rad/s multiply by 2π/60; 1 bar = 10⁵ Pa; 1 MPa = 1 N/mm²) and review it before each mock.
Applying the general bending formula σ = My/I or τ = Tr/J to brittle materials or to sections where the failure criterion is maximum principal stress, and skipping the stress-transformation step entirely.
Most textbook practice is on ductile steels where von Mises or Tresca and the simple bending formula coincide closely enough. When a cast-iron shaft or a combined bending-plus-torsion problem appears, candidates plug into τ_max = 16T/πd³ and σ = 32M/πd³ separately and either add them arithmetically or ignore the principal-stress combination.
What to do instead
For every combined-loading problem, first compute σ_x, σ_y, τ_xy at the critical point, then draw Mohr's circle or use σ_1,2 = (σ_x+σ_y)/2 ± √[((σ_x−σ_y)/2)² + τ_xy²]. Then select the criterion: maximum principal stress for brittle, von Mises/Tresca for ductile. Drill 15 mixed combined-loading problems until the sequence is automatic.
Treating all thermodynamic processes as if they obey pV^n = C with the same n, or using the steady-flow energy equation for a closed system (or vice versa), and mis-signing work and heat.
The work formulas for isothermal (pV ln(V₂/V₁)), isentropic (pV^n with n = γ), and polytropic processes look similar, and candidates memorise the polytropic formula and apply it everywhere. Similarly, closed-system Q − W = ΔU and steady-flow Q − W = Δh get swapped, especially for nozzles and diffusers where ΔKE matters.
What to do instead
Before writing any equation, classify the system (closed vs control volume) and the process (specify what is constant). Build a one-page process table (isochoric, isobaric, isothermal, isentropic, polytropic) listing p-V-T relations, work, heat, and Δs, and reproduce it from memory weekly. For nozzles/diffusers, always start from h₁ + V₁²/2 = h₂ + V₂²/2.
Using the wrong Reynolds-number or friction-factor correlation regime: applying the laminar f = 64/Re to turbulent pipe flow, or using the Darcy friction factor where the Fanning factor is intended (a factor-of-4 error).
Both friction factors are denoted f in different textbooks, and the Moody chart and Colebrook equation use Darcy. Candidates also forget that the critical Reynolds number for pipe flow is about 2300, not 5×10⁵ (the flat-plate boundary-layer value).
What to do instead
Standardise on Darcy–Weisbach, h_f = f (L/D)(V²/2g), and always check Re against 2300 before choosing f. For turbulent flow, use the Colebrook or Haaland correlation rather than reading a chart. Add a one-line note in your formula sheet: 'Fanning f = Darcy f / 4'.
In metal-cutting numericals, mixing up the three velocities (cutting velocity V, chip velocity V_c, shear-plane velocity V_s) and using the wrong rake-angle sign in the shear-angle relation tan φ = r cos α / (1 − r sin α).
Merchant's circle geometry is easy to sketch but easy to mislabel under time pressure, and the chip-thickness ratio r = t₁/t₂ is sometimes inverted. Candidates also forget that increasing rake angle increases φ and reduces cutting force, so qualitative questions get answered backwards.
What to do instead
Draw the Merchant circle for every cutting problem before substituting, label V, V_c, V_s and the angles φ, α, and β explicitly, and compute r = t₁/t₂ from the given chip thicknesses. Sanity-check: φ should be between 0° and 90°, and larger α must give larger φ.
In engineering mathematics, computing eigenvalues or a matrix inverse by hand and making an arithmetic slip, or misapplying the Laplace transform of a derivative (forgetting initial conditions) and the convolution theorem.
These are multi-step procedural calculations with no partial credit in a NAT format, and candidates attempt them under time pressure without a verification step. Initial-condition terms (sY(s) − y(0)) are the most commonly dropped.
What to do instead
For 2×2 and 3×3 matrices, verify with trace = sum of eigenvalues and determinant = product of eigenvalues; this catches most errors in seconds. For Laplace problems, write the transform of each derivative with its initial condition explicitly before substituting, and use partial fractions rather than memorised inverse pairs.
Confusing the phase diagram labels and the eutectoid/eutectic compositions and temperatures in the Fe–Fe₃C system, then misidentifying the microstructure (pearlite, bainite, martensite) for a given cooling rate.
The diagram has many labelled points (0.76% C, 727 °C eutectoid; 4.3% C, 1147 °C eutectic) and candidates memorise the numbers without the lever rule. Rapid-cooling questions are answered by name rather than by the TTT diagram.
What to do instead
Reconstruct the Fe–Fe₃C diagram from memory weekly: label austenite, ferrite, cementite, pearlite, the eutectoid point, and the eutectic point. Practise the lever rule on at least five compositions. For cooling-rate questions, always draw the cooling curve on the TTT diagram and read off which transformation region it crosses.
In industrial engineering, computing the economic order quantity or a CPM/PERT network but reporting the wrong quantity: EOQ when total cost is asked, expected project duration when the critical path length is asked, or crashing cost per unit time without checking which activity has slack.
These problems have several plausible outputs, and candidates stop at the first computed number. In CPM crashing, the cheapest activity on the critical path is not always the one with the least crash cost per day once parallel paths become critical.
What to do instead
Before computing, underline the exact quantity requested and its unit. For EOQ problems, compute inventory holding cost and ordering cost separately and check they are equal at the optimum. For crashing, recompute the critical path after every crash step and stop when a new path becomes critical.
Treating General Aptitude as a reading-comprehension-only section and skipping the numerical-aptitude and data-interpretation practice, then losing 5–8 marks on ratio, percentage, and series questions that are quick wins.
GA is only 15 marks but is the most time-efficient section in the paper, and many ME candidates deprioritise it because it feels unrelated to engineering. Under exam pressure, word problems with percentages and ratios are misread.
What to do instead
Allocate 20–25 minutes of every mock to GA and aim for near-full marks there. Practise 10 numerical-aptitude and 2 data-interpretation sets per week, and for verbal questions always eliminate options that overstate or add information not in the passage.