Mechanical Engineer interview questions and answers

The first law is conservation of energy: energy can change form but cannot be created or destroyed. For a closed system, Q − W = ΔU: the heat added minus the work done by the system equals the change in internal energy. In a car engine, the chemical energy of fuel becomes heat, part of which becomes work at the crankshaft.

The second law says which way energy flows. The Kelvin-Planck statement: no engine working in a cycle can turn all the heat it receives into work; some must be rejected to a colder body, which is why the engine needs a radiator. The Clausius statement: heat cannot flow from a cold body to a hot one without work, which is why a refrigerator needs a compressor.

The Otto cycle, the ideal cycle for petrol engines, adds heat at constant volume, and its efficiency is 1 − 1/r^(γ−1), so it depends only on the compression ratio r. The Diesel cycle adds heat at constant pressure, and its efficiency also depends on the cut-off ratio.

For the same compression ratio, the Otto cycle is more efficient. But diesel engines compress only air, so there is no risk of knocking, and they run at much higher compression ratios than petrol engines. That is why real diesel engines are more efficient. For the same maximum pressure and temperature, the Diesel cycle is the more efficient one. Interviewers like it when you state both comparisons.

Entropy is a property that measures how much of a system's energy is unavailable for doing work, or the degree of disorder. For a reversible process, the change in entropy is dS = δQ/T. The second law says the total entropy of a system and its surroundings stays the same in a reversible process and increases in every real one.

It increases because real processes have irreversibilities: friction, heat transfer across a finite temperature difference, unrestrained expansion, mixing and throttling. In a refrigerator's expansion valve, for example, enthalpy stays the same but entropy rises, and that lost potential is why real machines fall short of the Carnot efficiency. Reducing these losses is what most efficiency improvements are about.

Three ways: raise the average temperature at which heat is added, lower the temperature at which it is rejected, and reduce losses. Raising boiler pressure and superheating the steam raise the average heat-addition temperature, limited by the strength of boiler tubes and turbine materials at high temperature. Lowering condenser pressure lowers the rejection temperature, limited by the cooling water temperature.

Higher pressure without enough superheat makes the steam wetter at the turbine exit, which erodes the last-stage blades. Reheating the steam between turbine stages solves that and adds some efficiency. Regenerative feed heating, using bled steam to heat the feed water, raises efficiency further. Supercritical plants combine these methods.

Its COP, the cooling delivered divided by the power input, has fallen, so I'd check the readings first: suction and discharge pressures, superheat and subcooling, and chilled and condenser water temperatures.

High discharge pressure usually means poor heat rejection: fouled condenser tubes, a cooling tower with a large approach, low condenser water flow, or air in the system. Low suction pressure and high superheat point to a low refrigerant charge, a leak or a faulty expansion valve. A fouled evaporator or low chilled water flow also cuts capacity. Then check the compressor itself for wear. Compare with the commissioning data or the last good log sheet; that tells you quickly what has changed.

At first stress is proportional to strain, which is Hooke's law, up to the proportional limit; the slope is Young's modulus, about 200 GPa for steel. Slightly beyond is the elastic limit, up to which the specimen returns to its original length. Then mild steel yields: the stress drops from an upper to a lower yield point and the strain increases with little extra load.

After yielding, strain hardening raises the stress to the ultimate tensile strength, the highest point of the curve. Then necking starts, the cross-section shrinks locally, and the specimen fractures. Brittle materials like cast iron show no clear yield point and break with little elongation. Mention that designs normally use the yield strength with a factor of safety, not the ultimate strength.

The bending equation is M/I = σ/y = E/R, where M is the bending moment, I the second moment of area about the neutral axis, σ the bending stress at distance y from the neutral axis, E Young's modulus and R the radius of curvature. Bending stress is zero at the neutral axis and largest at the outer fibres, σ = M/Z, where Z = I/ymax is the section modulus.

Because the material near the neutral axis carries very little stress, an I-section puts most of its material in the flanges, far from the axis. That gives a high section modulus for its weight, so it resists bending economically. The web mainly carries the shear force. The same reasoning explains why hollow sections are efficient.

The torsion equation is T/J = τ/r = Gθ/L, where T is the torque, J the polar moment of inertia, τ the shear stress at radius r, G the modulus of rigidity, and θ the angle of twist over length L. For a solid shaft J = πd⁴/32, so the maximum shear stress is 16T/πd³.

Shear stress is zero at the centre and highest at the surface, so the material near the centre does little work. A hollow shaft removes that material: for the same weight it is stronger and stiffer in torsion than a solid one, or for the same torque it is lighter. That's why vehicle propeller shafts are hollow. Solid shafts are still used where cost and simple manufacture matter more than weight.

Euler's crippling load is P = π²EI/Le², where Le is the effective length. Le depends on the end conditions: L for both ends hinged, L/2 for both ends fixed, 2L for one end fixed and the other free, and L/√2 for one end fixed and the other hinged. Buckling happens about the axis with the smaller I.

The formula assumes a perfectly straight, long, slender column of uniform material with the load exactly along its axis. It is valid only when the crippling stress stays below the proportional limit, that is for a high slenderness ratio. For short and intermediate columns it overestimates the load, so the Rankine-Gordon formula, which combines crushing and buckling, is used instead.

Ductile materials like steel fail by yielding, which is governed by shear, so we use the maximum shear stress theory (Tresca) or the distortion energy theory (von Mises). Von Mises matches test results for ductile materials best; Tresca is simpler and slightly conservative. For a shaft under bending moment M and torque T, Tresca gives an equivalent torque of √(M² + T²), and the diameter comes from 16Te/πd³ equal to the allowable shear stress.

Brittle materials like cast iron fail by fracture under normal stress, so the maximum principal stress theory (Rankine) is used, with care because they are much stronger in compression than in tension. In an interview, state the material first, then pick the theory.

Spur gears have straight teeth and connect parallel shafts; they are simple and efficient but noisy at high speed. Helical gears also connect parallel shafts, but their teeth are inclined, so they engage gradually, run quieter and carry more load; the drawback is an axial thrust on the bearings, which double helical gears cancel. Bevel gears connect intersecting shafts, usually at 90°, as in a vehicle differential.

Worm gears connect shafts at right angles that do not intersect and give a large speed reduction in one stage, which suits hoists and indexing drives. They can be self-locking, but the sliding contact makes them less efficient and they need good lubrication.

For a flat belt, T1/T2 = e^(μθ), where T1 is the tension on the tight side, T2 on the slack side, μ the coefficient of friction and θ the angle of contact in radians on the smaller pulley. The power transmitted is (T1 − T2) × v. For a V-belt the ratio becomes e^(μθ/sin β), where β is half the groove angle, so it grips much better.

A belt slips when the power demanded needs a larger tension difference than friction can provide: because of overload, low initial tension, a small angle of contact or a glazed, oily belt. This is different from creep, the small relative movement caused by the belt stretching unequally on the tight and slack sides, which always happens.

Grashof's law says that in a four-bar chain, if the sum of the shortest and longest links is less than or equal to the sum of the other two (s + l ≤ p + q), at least one link can rotate fully relative to the others. If not, no link can make a full turn and every inversion is a double rocker.

For a Grashof chain, the result depends on which link is fixed. Fixing a link next to the shortest gives a crank-rocker, used in beam engines and windshield wipers. Fixing the shortest link gives a double crank, like the coupling rod of locomotive wheels. Fixing the link opposite the shortest gives a double rocker. Interviewers often follow up with the inversions of a slider-crank chain.

Start from the loads and the speed. Deep groove ball bearings suit moderate radial and some axial load at high speed; cylindrical roller bearings take heavy radial load; taper roller bearings take combined radial and axial load, as in wheel hubs; thrust bearings take mainly axial load.

Then check life. The rated life is L10 = (C/P)^p million revolutions, with p = 3 for ball bearings and 10/3 for roller bearings, where C is the dynamic load rating from the catalogue and P = XFr + YFa is the equivalent load. Choose a bearing that meets the required life in hours at the operating speed, then consider shaft size, lubrication, sealing against dust and how it will be mounted and removed.

A failure after months of running, not at once, suggests fatigue. I'd look at the fracture surface: beach marks spreading from a start point at the keyway corner and a small final rough zone confirm fatigue, and a small final zone means a low nominal stress, so stress concentration is the real cause.

Then check the design against fluctuating stress, using the endurance limit corrected for surface finish, size and reliability and a criterion such as Goodman or Soderberg. Also check what's adding stress: misalignment, over-tensioned belts, shock loads, a sharp keyway corner. Fixes include a larger fillet radius, a better-finished keyway, a larger diameter at that section, shot peening, and correct alignment.

In first-angle projection the object is imagined in the first quadrant, between the observer and the plane, so the top view is drawn below the front view and the left side view is drawn on the right. In third-angle projection the plane is between the observer and the object, so the top view goes above the front view and each side view sits on its own side.

BIS drawing practice in India uses first-angle projection, as does most of Europe, while the USA uses third-angle. Always check the projection symbol, a truncated cone, in the title block: many Indian engineering service companies work on US clients' drawings in third angle, and misreading it mirrors the part.

A part is a single component modelled from sketches and features. An assembly brings parts together and positions them relative to each other. A drawing is the 2D manufacturing document, with views, dimensions, tolerances and a bill of materials, generated from the part or assembly. They are linked, so a change in the part updates the assembly and the drawing.

Mates remove degrees of freedom between components: coincident, concentric, parallel, perpendicular, distance and angle are the common ones. A shaft in a bearing needs a concentric mate and a coincident mate on its shoulder, leaving rotation free if it should spin. Avoid over-defining with conflicting mates, which gives errors when the design changes.

Tolerance is the permitted variation in one dimension: the difference between its upper and lower limits. Allowance is the intended difference between mating parts, the minimum clearance or the maximum interference between a hole and a shaft.

Fits are written in the ISO system with a hole tolerance and a shaft tolerance. In the hole-basis system, which is preferred because standard drills and reamers make holes, H is the hole with a lower deviation of zero. H7/g6 is a close clearance fit, used for sliding parts like a locating pin or a gear on a shaft that must move. H7/k6 is a transition fit for location, and H7/p6 an interference fit for pressing a bush or bearing in.

First confirm what was made against the drawing, then go through my own side. Was the drawing revision the shop used the same as the model? Did a tolerance stack-up through several parts use up the clearance, even though each part is within its limits? Are the datums on the drawing the ones the part actually locates from in the assembly, and did I use GD&T, such as position tolerances for hole patterns, where plain plus-minus dimensions aren't enough?

Also look for missing or double dimensions, the wrong projection or units, and features that can't be made as drawn. Then fix the drawing, not just the part, and add the check to your review list. Interviewers like candidates who own the problem.

By capturing design intent. Fully define every sketch with dimensions and relations instead of leaving it dragging. Reference the origin, base planes and stable faces instead of edges and fillets that may disappear. Keep a sensible feature order: main body first, then cuts and holes, and cosmetic fillets and chamfers last.

Drive key sizes with global variables or equations, so changing a diameter updates every related feature, and use configurations or design tables for families of similar parts. Name important features. Then test the model by changing a main dimension to see if it rebuilds cleanly. A model that can be changed safely saves far more time than one that was quick to build.

Blowholes and gas porosity are cavities caused by gas trapped in the metal, from moist sand, low sand permeability or poor venting. Shrinkage cavities form when the metal contracts as it solidifies and the riser can't feed it. A cold shut is a visible line where two streams of metal met but didn't fuse, from low pouring temperature or slow pouring; a misrun is a casting that isn't completely filled for the same reasons.

Hot tears are cracks from contraction being resisted by the mould or core while the metal is still weak. Sand inclusions come from loose sand washed in by the metal. Name the remedy for each: control moisture, improve venting, design risers properly and control the pouring temperature.

Manual metal arc welding, or stick welding, uses a flux-coated consumable electrode; it's cheap and portable, good for site and structural work. MIG welding feeds a continuous wire electrode through a gun with shielding gas, argon or an argon-CO2 mix, and is fast for fabrication and production lines. TIG welding uses a non-consumable tungsten electrode with argon shielding and a separate filler rod; it's slower but gives clean, precise welds on thin sheet, stainless steel and aluminium.

Resistance spot welding joins overlapping sheets by passing a heavy current through them between two electrodes; car bodies are assembled this way. Be ready to name defects too, such as porosity, undercut, lack of fusion and cracking.

A drift over hours points to something that changes with time, not a wrong program. The first suspect is tool wear: check the insert's edge and how many parts it has made, and set a tool life limit or regular wear offset correction. The second is thermal growth of the spindle and machine as they warm up; a warm-up cycle and in-process gauging help.

Also check coolant concentration and temperature, chips building up on the chuck or locating faces, worn chuck jaws, changes in the raw material's hardness, and whether anyone has edited offsets. Plot the measured dimension against time on a control chart: a steady drift and a sudden jump have different causes.

The seven basic quality tools are the check sheet, histogram, Pareto chart, cause-and-effect (fishbone or Ishikawa) diagram, scatter diagram, control chart, and stratification (some lists use a flowchart instead). They are simple enough for the whole shop floor to use.

An example: rejections on a machining line are rising. A check sheet records the defects by type for a week, and a Pareto chart shows that two defects, say oversize bores and burrs, cause about 80% of the rejections. A fishbone diagram for the oversize bores lists possible causes under man, machine, method, material, measurement and environment, and a control chart then shows whether the fix worked. Show that you know how the tools connect.

PDCA, Plan-Do-Check-Act, is a cycle for improvement: plan a change based on the problem's cause, do it on a small scale, check the results against the target, and act by making it the standard or trying again. It's the base of kaizen.

The 5 Whys means asking why until you reach a cause you can fix. For example: the machine stopped. Why? A fuse blew. Why? The motor was overloaded. Why? The bearing was jammed. Why? It wasn't lubricated. Why? There was no lubrication schedule for that machine. The fix is the schedule, not a new fuse. Mention that five is a guide, not a rule, and that the answers must be checked with facts on the shop floor.

Both compare the spread of a stable process with the tolerance. Cp = (USL − LSL)/6σ shows whether the process could fit within the limits if it were centred. Cpk = the smaller of (USL − mean)/3σ and (mean − LSL)/3σ also accounts for where the process is centred. Cpk is never more than Cp; a big gap between them means the process is off-centre.

A Cpk of 0.8 means some parts are already outside the tolerance, so I'd start 100% inspection of that dimension to protect the customer, then check whether Cp is fine and the process just needs centring, or whether the variation itself is too large and needs root cause work. Many automotive customers expect at least 1.33 for critical dimensions.

Give a genuine reason backed by something you've done. "I enjoy seeing how physical things are designed and made. During my internship at [company], I worked on [line or project], and solving a real [machining, assembly or design] problem there was the most satisfying work I've done. I want to build depth in [manufacturing, design or maintenance], and I know that takes years on the shop floor."

Don't criticise IT careers or say you couldn't get an IT job. If you also applied to IT companies, there's no need to hide it, but make clear why this role is your first choice.

Plants run round the clock and are often in industrial areas, so they need people who will stay. If you can, say so and show you understand it: "Yes. I know production engineers usually start in shifts, and that's where you learn the process, the machines and the people. I'm comfortable relocating to [plant location]."

If you have a real limit, such as a family responsibility, say it honestly and say what you can do. Agreeing now and asking for a transfer in three months hurts you more than being clear today.

Expect to be questioned deeply on it, so choose the part you know best. Use problem, approach, your role, result: "We designed and fabricated a [machine or system] to [purpose]. I did the design calculations for [shaft, frame or gear] and modelled it in SolidWorks; we tested it and achieved [result]. If I did it again, I'd [one improvement]."

Know the numbers: dimensions, materials, loads and cost. If it was a group project, be clear about your own part, and never claim work you didn't do, because one technical follow-up exposes it.

Use STAR and give figures. "On our machining line, cycle time on one operation was causing a bottleneck (situation). I was asked to raise output without new machines (task). I studied the cycle with a stopwatch and found that loading and clamping took 40% of it, so I designed a quick-clamp fixture and changed the tool path to remove air cuts (action). Output went up about 18%, and the fixture is now used on two other machines (result)."

Show that you involved the operators, because shop floor improvements only last when the people running the machine accept them.

This question checks whether you prepared, so research beforehand: the company's main products, its customers or industries, its plants, and any recent news such as a new plant, product launch or export order. Then connect it to yourself: "You make [products] for [customers], and you've recently [expanded or launched something]. My project on [related topic] and my interest in [process] fit the work at your [plant or department]."

Don't recite the website's history page, and never guess facts you aren't sure of. One accurate, relevant point is better than a long list.