Practice free

Mechanical Engineering Interview Questions for Freshers (2026): Core Roles, Not the IT Switch

Updated August 2026

Most placement advice a mechanical student receives is written for someone else. It is about aptitude sections, coding rounds and IT service companies, because that is where the campus volume is — and the quiet effect of four years of it is that a great many mechanical graduates arrive at final year having prepared for an IT job and never seriously prepared for a mechanical one. This page is for the students who want the core route, and it treats that as a legitimate first choice rather than as what you settle for.

The good news is that core interviews are more predictable than IT ones. There is no aptitude marathon and no coding round. What you get is a conversation about your subjects — thermodynamics, strength of materials, manufacturing processes, machine design, fluid mechanics — plus your project, your drawing and tool skills, and whether you can reason about a physical problem out loud. The syllabus is your own degree, which sounds obvious until you notice how many candidates cannot state the difference between stress and strain in an interview after passing an examination on it.

The regional picture matters too, because core mechanical employment in the Telugu states is real but concentrated in specific places rather than spread evenly: the Hyderabad public-sector and aerospace cluster, the automotive and industrial investments in Andhra including the Sri City belt near Tirupati and the plant at Anantapur, the Visakhapatnam steel and port economy, engineering-services firms that hire mechanical graduates into design and analysis work, and the pharma and process industries that need mechanical people for utilities and maintenance. The questions below are at fresher depth across those role types, followed by the honest version of the core-versus-IT question — including a route out if you decide the switch is what you actually want.

Frequently asked questions

How is a core mechanical interview different from an IT placement test?

Almost entirely, and preparing for one does little for the other. An IT services drive front-loads a written test — quantitative aptitude, logical reasoning, verbal ability, sometimes coding — and the technical conversation afterwards is often shallow. A core interview usually has a smaller written component, if any, and puts the weight on a technical discussion about your subjects and your project, conducted by an engineer who does the work. It is closer to a viva than to an examination: they will follow up on your answer, ask why, and keep going until they find the edge of what you understand. The practical implication is that months spent on aptitude are not preparation for this. Your syllabus is, and depth in a handful of core subjects beats familiarity across all of them.

Explain stress and strain, and the difference between ductile and brittle materials.

Stress is the internal resisting force per unit area developed in a material when a load is applied, expressed in pascals or newtons per square millimetre; strain is the resulting deformation expressed as a ratio — change in dimension divided by original dimension — and so it is dimensionless. In the elastic region they are related by Young's modulus, which is the slope of the initial straight portion of the stress-strain curve. A ductile material such as mild steel undergoes significant plastic deformation before fracture, shows a clear yield point, necks visibly, and gives warning before failing. A brittle material such as cast iron or glass fractures with very little plastic deformation and often without warning, which is why brittle failure is treated as more dangerous in design even when the material is strong. If you can sketch and label a stress-strain curve for mild steel while answering, do it — interviewers ask for it constantly.

What is factor of safety, and how is it decided?

Factor of safety is the ratio of a material's strength to the working or design stress actually expected in service, and it exists because neither the loads nor the material are known perfectly. It absorbs variation in material properties, manufacturing tolerances, unexpected loads, corrosion and wear, errors in the analysis, and the consequences of failure. That last one is why the value is a judgement rather than a formula: a bracket whose failure inconveniences someone justifies a smaller factor than a lifting hook whose failure kills someone, and pressure vessels, aerospace and structural work carry their own design codes and standards specifying what is acceptable. A good answer says which strength you are dividing by — yield strength for ductile materials, ultimate strength for brittle ones — and adds that a very high factor is not automatically good engineering, since it means excess weight, material and cost.

State the first and second laws of thermodynamics, and why the second one matters.

The first law is conservation of energy applied to a thermodynamic system: energy is neither created nor destroyed, so for a closed system the heat added equals the change in internal energy plus the work done. It tells you the bookkeeping must balance. The second law says that balance is not sufficient — it introduces direction and quality. Heat flows spontaneously from hot to cold and not the reverse, no heat engine can convert all the heat it receives into work, and entropy of an isolated system tends to increase. Why it matters is the practical part interviewers want: the second law is why every engine, refrigerator and power plant has a ceiling on efficiency that no amount of engineering can exceed, with the Carnot efficiency setting that limit between two temperatures. The first law tells you energy is conserved; the second tells you how much of it you can actually use.

What is the Carnot cycle, and what does its efficiency depend on?

The Carnot cycle is a theoretical, fully reversible cycle consisting of two isothermal processes and two adiabatic processes, and it represents the maximum efficiency any heat engine can achieve operating between two given temperatures. Its efficiency depends only on those two absolute temperatures — one minus the ratio of the sink temperature to the source temperature, both in kelvin — and not on the working fluid, which is the result that surprises students and which interviewers like to probe. Two consequences worth stating: efficiency improves by raising the source temperature or lowering the sink temperature, which is precisely why power plants push turbine inlet temperatures up and why condenser cooling matters; and no real engine reaches it, because real processes involve friction, finite temperature differences and other irreversibilities.

Explain Bernoulli's equation and Reynolds number.

Bernoulli's equation states that for steady, incompressible, inviscid flow along a streamline, the sum of pressure energy, kinetic energy and potential energy per unit volume stays constant — so where velocity rises, pressure falls. Its assumptions are the part to mention, because it is applied carelessly: steady flow, incompressible fluid, negligible viscosity, and along a streamline. Reynolds number is the dimensionless ratio of inertial to viscous forces, computed from density, velocity, characteristic length and dynamic viscosity, and it predicts the flow regime. For internal pipe flow the commonly quoted transition is laminar below roughly 2000 and turbulent above roughly 4000, with a transitional band between — cite those as conventional values rather than exact physical constants. Laminar flow moves in orderly layers with a parabolic velocity profile; turbulent flow is chaotic and mixes strongly, which increases both pressure loss and heat transfer.

What are the common casting defects, and what causes them?

The ones to know by name are blowholes and porosity, cavities caused by trapped gas from moisture, inadequate venting or gas dissolved in the melt; shrinkage cavities, caused by the metal contracting as it solidifies without sufficient feed metal from a riser; cold shut, where two streams of metal meet without fusing properly because the pouring temperature was too low or the flow too slow; misrun, where the metal solidifies before filling the cavity; hot tears, cracks from restrained contraction during cooling; and inclusions, foreign matter such as sand or slag caught in the casting. Being able to attach a cause to each is the actual test, because it shows you understand the process rather than having memorised a list. It also helps to know the standard remedies — proper gating and risering, venting, controlled pouring temperature, mould preparation — since the natural follow-up question is what you would do about it.

What is the difference between annealing, normalising, hardening and tempering?

All four are heat treatments and the differences lie in the cooling and the purpose. Annealing means heating to the appropriate temperature and cooling very slowly, usually in the furnace, to soften the material, relieve internal stresses, refine grain structure and improve machinability. Normalising heats to a similar range but cools in still air, which is faster — giving a finer grain and somewhat higher strength and hardness than annealing, and a more uniform structure. Hardening heats the steel and then quenches it rapidly in water, oil or brine to form martensite, which is very hard and also brittle and highly stressed. Tempering is what follows hardening: reheating to a lower temperature and cooling, which sacrifices some hardness to recover toughness and relieve the quenching stresses. The relationship between the last two is the point worth making — hardening alone leaves a component too brittle to use, so tempering is not optional.

What is GD&T, and why is it used instead of plus-minus tolerances?

Geometric Dimensioning and Tolerancing is a symbolic language on engineering drawings for specifying the permissible variation in the form, orientation, location and runout of features, relative to defined datums. It is used because coordinate plus-minus tolerancing describes the size of a feature but not adequately its geometry or relationship to other features — a hole can be exactly the right diameter and still be in the wrong place, at the wrong angle, or in a part whose face is not flat. GD&T controls those independently, which means the drawing communicates the designer's actual functional intent rather than an approximation of it. At fresher level you are expected to recognise the common controls — flatness, straightness, circularity, perpendicularity, parallelism, position, concentricity, profile and runout — to understand what a datum is and why datum selection matters, and to know that a feature control frame reads as symbol, tolerance value, then datum references. Deep expertise is not expected; blank unfamiliarity is noticed.

Explain limits, fits and tolerance, and the types of fit.

Tolerance is the permissible variation in a dimension — the difference between the upper and lower limits. Limits are those maximum and minimum permissible sizes. A fit describes the relationship between a mating shaft and hole and follows from their respective tolerances. There are three families. A clearance fit always leaves space, so the shaft is always smaller than the hole and the parts move or assemble freely — a shaft rotating in a bush. An interference fit means the shaft is always larger than the hole, so assembly requires force, heating or cooling and the joint transmits load by friction — a bearing race pressed onto a shaft. A transition fit sits between them and may produce either a small clearance or a small interference depending on where the actual sizes fall within tolerance, and is used for accurate location where the part is also secured by other means. It is worth adding that the hole-basis system is more commonly used in practice, because it is easier to machine a shaft to size than to vary a hole.

What is the difference between a pump and a compressor, and between a centrifugal and a reciprocating pump?

A pump moves liquid, which is essentially incompressible, so it raises pressure and moves the fluid without changing its density appreciably. A compressor works on gas and raises its pressure by reducing volume, which also raises its temperature — which is why compressors need intercooling or aftercooling and pumps generally do not. On pump types, a centrifugal pump is a rotodynamic machine: an impeller adds kinetic energy to the liquid which is then converted to pressure, giving smooth continuous flow, few moving parts and suitability for high flow at moderate head, but it needs priming and its delivery varies with the system. A reciprocating pump is a positive displacement machine using a piston or plunger to deliver a fixed volume per stroke, which produces high pressure and near-constant flow regardless of head, is self-priming, and delivers pulsating flow requiring an air vessel to smooth it. The general rule an interviewer wants: centrifugal for high flow and low to moderate head, positive displacement for high head and low flow, or for viscous fluids.

Compare two-stroke and four-stroke engines.

A four-stroke engine completes its cycle in four piston strokes and two crankshaft revolutions — intake, compression, power and exhaust — with one power stroke every two revolutions, valves controlling the gas exchange, and separate lubrication. A two-stroke completes the cycle in two strokes and one revolution, combining processes so that there is a power stroke every revolution, usually using ports rather than valves and scavenging to displace exhaust with the incoming charge. The trade-offs are the answer: the two-stroke is mechanically simpler, lighter and produces more power for a given size, but it is less fuel-efficient because some fresh charge escapes with the exhaust during scavenging, and it burns oil mixed with fuel, so emissions are considerably worse. That last point explains what an interviewer is usually driving at — why emissions regulations pushed two-strokes out of most road vehicles while they persist where light weight and simplicity dominate.

What is COP in refrigeration, and why can it exceed one?

Coefficient of performance is the ratio of the useful effect to the work input. For a refrigerator it is the heat removed from the cold space divided by the work supplied; for a heat pump it is the heat delivered to the warm space divided by the work supplied, which makes the heat pump COP exactly one greater than the refrigerator COP for the same cycle. It can and routinely does exceed one, and the reason is the part worth saying out loud: the machine is not creating energy, it is moving heat from one place to another, and the work input is only what it costs to pump that heat against the temperature gradient. Calling it efficiency is why students find it confusing, which is precisely why the term efficiency is not used here. COP improves as the temperature difference the machine works across gets smaller.

Which manufacturing processes should I be able to discuss, and at what depth?

Know the families and be able to say what each is good for and where its limits lie: casting for complex shapes in volume; forming and forging for strength through grain flow; machining — turning, milling, drilling, grinding — for accuracy and surface finish; joining, principally welding, brazing and soldering, with the distinction being that welding melts the parent metal while brazing and soldering do not; sheet metal work; and additive manufacturing, where you should at least be able to discuss where it is genuinely useful rather than repeating enthusiasm. On welding specifically, know the difference between arc processes such as shielded metal arc, MIG and TIG, and resistance welding, and know common weld defects like undercut, porosity, incomplete penetration and cracking. On machining, know the basic operations, what cutting speed and feed mean, and why coolant is used. Depth beyond that is welcome but the ability to place a process correctly is what is being tested.

What CAD and analysis tools should a fresher know, and how honest should I be about them?

Know one modelling package genuinely rather than five superficially — commonly SolidWorks, CATIA, Creo, NX or Fusion — meaning you can produce a part, assemble components, apply constraints and generate a proper drawing with dimensions, sections, tolerances and a title block. AutoCAD for 2D drafting is still widely expected. On the analysis side, an awareness of FEA and CFD concepts is valuable: what meshing is, why boundary conditions determine whether your result means anything, and the crucial point that software will always return a colourful result whether or not the model was sensible. Be honest about your level, because it is trivially exposed. "I have modelled and assembled parts in SolidWorks and produced drawings; I have done a static structural analysis in a course but I would not call myself an FEA person" is a strong answer that no interviewer punishes. An inflated claim followed by an inability to describe how you set up your boundary conditions is remembered.

How should I present my final-year project in a core interview?

As an engineering decision rather than as a description. Start with the problem and why it mattered, then what you personally designed, built or analysed, then the choices you made and why — the material, the mechanism, the process, the assumptions — and then what went wrong and what you did about it. That last part carries the most weight in a core interview and is the part students omit, because a project where nothing failed is either untrue or trivial. Bring your drawings, and be ready for the follow-ups that come from them: what tolerance did you specify and why, what material and why that one, what load did you assume, what would you change now. If your project was primarily fabricated by a workshop, say so honestly and be clear about what was yours — a candidate who claims machining they did not do is caught by one question about the setup.

What are the actual core job routes for a mechanical fresher in AP and Telangana?

Several, and they need different preparation. Manufacturing and production roles at plants — automotive and ancillary units around the Sri City belt near Tirupati and the plant at Anantapur, plus industrial units across both states — involve shifts, process control and shop-floor problem-solving. Design and analysis work sits largely with engineering-services firms and captive engineering centres, where CAD and FEA skills and drawing discipline matter most. Quality and maintenance functions exist across every manufacturing employer, including pharma and process plants that need mechanical people for utilities, equipment and maintenance. The Visakhapatnam steel and port economy and the Hyderabad public-sector and aerospace cluster are the other significant concentrations. Then there are the routes that need a separate exam or qualification: GATE for public-sector undertakings and for an MTech, which is a genuine and well-trodden path for core mechanical students. Indicative fresher pay in core roles commonly runs around ₹2–4.5 LPA and varies widely by employer and location.

Core pays less than IT at entry. Is it still worth it?

Sometimes, and the honest version of this needs both halves. It is broadly true that entry-level core salaries are lower than IT services offers and much lower than product-company offers, that core openings are fewer in absolute number, and that shift work at a plant in an industrial area is a different life from an office job in the city. Those are real costs and nobody should pretend otherwise. Against that: core work uses what you studied, the domain knowledge compounds in a way generic IT support work does not, experienced mechanical engineers in design, manufacturing and project roles do well, the competition for good core roles is far thinner than the queue for software, and the manufacturing investment in this region has been increasing rather than shrinking. The right decision depends on which of those you actually weight — and the wrong way to decide is by drifting, which is what happens to most students who never prepared for either route properly.

I want the IT or data switch instead. Is that a bad decision?

No, it is a completely legitimate one, and this page would be dishonest if it argued otherwise — many mechanical graduates build excellent careers in software, analytics and product roles, and a mechanical background is genuinely useful in domains like manufacturing analytics, simulation, robotics and engineering software. What matters is that it is a decision rather than a default. If you are switching because you find computing more interesting or the opportunities better, prepare for it properly: aptitude, one programming language to real depth, data structures, databases, and a project you can defend. If you are switching only because nobody told you what a core interview looks like, that is the situation this page exists to fix. And if the route you want is a masters that formalises the switch, our mechanical-to-data-science guide covers what that path actually requires and where it leads.

I am in the 2027 batch and want a core job. What should I do now?

Four things, and they are cheaper than they sound. Pick three or four core subjects — thermodynamics, strength of materials, manufacturing processes, and either fluid mechanics or machine design — and take them to interview depth rather than examination depth, meaning you can explain, sketch and be questioned past your first answer. Learn one CAD package properly, to the point where you can model, assemble and produce a correct drawing with tolerances. Get an internship or industrial training at an actual plant, however modest, because having stood on a shop floor changes what you can say in every interview afterwards. And decide about GATE early, since the public-sector and MTech routes need a preparation cycle that cannot be assembled in a final semester. Do those and you will be better prepared for a core season than most of your batch, who will have spent the same period preparing for a different job entirely.

Don't just read Mechanical core interviews questions — get asked them

Phiny's AI interviews you on exactly these topics, follows up on weak answers, and tells you what a stronger answer looks like. Text interviews are free and unlimited.

Start a free AI mock interview

How to prepare

Where these questions get asked

Related guides