Designing an engine means making hundreds of interdependent decisions about thermodynamics, materials, geometry, and airflow, then validating those choices through simulation and physical testing. Whether you’re a mechanical engineering student tackling a capstone project or a motorsport enthusiast planning a custom build, the process follows a common sequence: choose a thermodynamic cycle, size the core components, select materials that survive the thermal and mechanical loads, design the airflow and fuel delivery systems, and then simulate everything before cutting metal.
Choosing a Thermodynamic Cycle
Every engine converts heat into mechanical work, and the thermodynamic cycle you select determines the theoretical ceiling of that conversion. The most familiar choice is the Otto cycle, used in conventional gasoline engines, which reaches a theoretical brake thermal efficiency around 42%. The Diesel cycle compresses air alone before injecting fuel, allowing higher compression ratios and, in many configurations, better fuel economy under load. Hybrid cycles push efficiency further. A Dual-Miller cycle, which combines elements of both Otto and Diesel operation with a modified intake valve strategy, can reach roughly 46% brake thermal efficiency in ideal conditions, the highest among common cycle configurations.
Your choice of cycle shapes nearly every downstream decision. A higher compression ratio (typical of diesel designs) demands stronger, heavier components. A Miller or Atkinson cycle trades peak power for efficiency by manipulating how long the intake valve stays open, which changes the effective compression ratio without changing the physical one. For most first-time designers, starting with a conventional Otto or Diesel cycle and understanding its limits is more productive than jumping to exotic configurations.
Sizing the Core Geometry
The bore (cylinder diameter), stroke (piston travel distance), and number of cylinders define your engine’s displacement and strongly influence its character. A “square” engine where bore equals stroke balances power and torque. A long-stroke design (stroke larger than bore) favors low-end torque. A short-stroke, large-bore design revs higher and makes peak power at elevated RPM.
Compression ratio ties directly to your cycle choice and fuel type. Gasoline engines typically run between 9:1 and 13:1. Diesel engines operate much higher, often 15:1 to 22:1, because they rely on compression alone to ignite fuel. Setting the compression ratio too high for a given fuel causes uncontrolled detonation (knock), which destroys pistons and bearings. Too low, and you leave efficiency on the table. Once you fix bore, stroke, compression ratio, and cylinder count, you can calculate theoretical displacement and estimate power output using volumetric efficiency assumptions.
Selecting Materials for Extreme Conditions
Engine internals endure punishing combinations of heat, pressure, and repetitive stress. The piston crown sees temperatures of 350 to 380°C under pressures around 11 atmospheres in a typical aluminum or magnesium piston. High-performance and diesel applications push conditions even further, requiring materials that survive 500°C and pressures up to 120 bar.
Forged aluminum alloys remain the most common piston material because they’re light and conduct heat well. Aluminum-silicon composites reinforced with ultra-fine silicon carbide particles offer better wear resistance. Magnesium alloys like WE43A weigh even less and actually outperform aluminum-silicon pistons at elevated temperatures, though they’re weaker at room temperature and more expensive. For engine blocks, the decision usually comes down to cast iron (heavy but stiff and cheap) or aluminum alloy (lighter but requiring iron or steel cylinder liners). Composite materials are gaining ground: an aluminum 6061 matrix reinforced with boron carbide and silicon carbide particles shows 31 to 76% less deformation under load compared to conventional options, with lower mechanical stress as well.
Crankshafts are forged from steel in performance applications or cast from iron in lower-stress designs. The crankshaft must resist not only bending loads but torsional vibration, where the repeated firing impulses twist the shaft back and forth at specific frequencies.
Managing Vibration and Structural Loads
Torsional vibration is one of the most destructive forces in a running engine. Each combustion event delivers a sharp rotational impulse to the crankshaft, and at certain RPM ranges, these impulses resonate with the shaft’s natural frequency. Left unchecked, this can fatigue and crack the crankshaft.
Engineers model the crankshaft as a series of lumped masses connected by flexible segments, sometimes using twelve or more degrees of freedom to capture the system’s behavior accurately. The excitation torque from both gas pressure and reciprocating inertia forces is broken into harmonic components using Fourier analysis, revealing which engine speeds produce the most dangerous vibrations. A torsional vibration damper, mounted at the front of the crankshaft, absorbs this energy. Hybrid damper designs combining rubber and viscous fluid elements can reduce torsional vibrations by over 50% compared to unoptimized single-type dampers. Getting this right early in the design prevents catastrophic failures later.
Designing the Valvetrain and Airflow
The camshaft controls when and how much the intake and exhaust valves open, and these parameters have an outsized effect on where in the RPM range your engine makes power. Three variables matter most: lift (how far the valve opens), duration (how long it stays open), and overlap (the brief period when both intake and exhaust valves are open simultaneously).
Higher valve lift allows more air and exhaust to flow through the engine, improving peak power. Longer duration keeps the valve open for more crankshaft rotation, which helps at high RPM but hurts low-RPM torque and creates a rough idle. The lobe separation angle on the camshaft controls overlap. A narrower angle increases overlap, raising cylinder pressure and improving mid-range power. A wider angle reduces overlap, producing a smoother idle, better vacuum signal for accessories, and broader torque delivery across the RPM range.
For a street engine, you’d typically choose moderate lift, short to medium duration, and a wider lobe separation angle. A race engine reverses those priorities: aggressive lift, long duration, and tighter lobe separation to pack as much air as possible into the cylinder at high RPM, accepting a lumpy idle and narrow powerband as tradeoffs.
Fuel Delivery and Air-Fuel Ratio
For a gasoline engine, the stoichiometric air-fuel ratio is 14.7:1 by mass, meaning 14.7 grams of air for every gram of fuel. This is the chemically ideal ratio where all fuel and all oxygen are consumed. Anything below 14.7:1 is a “rich” mixture (more fuel than needed), and anything above is “lean” (excess air).
In practice, engines rarely run at exactly 14.7:1 across all conditions. Most operate slightly lean of stoichiometric for better fuel economy and lower hydrocarbon and carbon monoxide emissions. Under heavy load or high RPM, a slightly rich mixture is often used deliberately. The extra fuel cools the combustion chamber and reduces the formation of nitrogen oxide emissions, which spike when combustion temperatures climb. The mixture for best fuel efficiency sits just lean of stoichiometric, but the mixture for maximum power is slightly rich. Your fuel system design, whether carburetor, port injection, or direct injection, needs to deliver the right ratio across the entire operating map of speed and load.
Diesel engines work differently. They always run lean because fuel is injected directly into compressed air and ignition happens spontaneously. The air-fuel ratio varies with load, dropping closer to stoichiometric as the engine works harder, but different fuel blends tend to converge to similar ratios above about 50% load.
Simulation Before Fabrication
Modern engine design relies heavily on computational simulation before any physical prototype is built. Two categories of software dominate: computational fluid dynamics (CFD) for modeling airflow, combustion, and heat transfer inside the cylinder, and finite element analysis (FEA) for predicting stress, deformation, and fatigue life in structural components.
On the CFD side, tools like CONVERGE model the combustion process in three dimensions, tracking flame propagation, turbulence, and emissions chemistry. These can be coupled with one-dimensional engine simulation tools like GT-SUITE, which model the entire engine system including intake and exhaust piping, turbocharger behavior, and cooling circuits. The 3D solver handles the detailed combustion chamber physics while the 1D solver captures system-level effects, and data flows between them in real time during the simulation.
For structural analysis, FEA solvers like Abaqus predict how components deform and where stress concentrates under thermal and mechanical loads. Coupling FEA with CFD allows you to solve fluid-structure interaction problems: how combustion pressure deforms the piston, how that deformation changes the combustion chamber shape, and how that change feeds back into the next combustion event. This iterative approach catches design problems that would otherwise only appear during expensive physical testing.
If you’re working without access to commercial software, open-source alternatives exist. OpenFOAM handles CFD, and CalculiX or Code_Aster can perform structural FEA. They require more setup effort but can get you meaningful results for initial design validation.
Prototyping and Iteration
Even after extensive simulation, physical testing reveals problems that models miss. Friction between the piston rings and cylinder wall, oil film behavior under transient loads, and acoustic resonance in the intake manifold are difficult to predict precisely. A typical development cycle involves building a single-cylinder test engine first, instrumenting it with pressure transducers and thermocouples, and running it on a dynamometer to compare measured performance against simulated predictions.
Discrepancies between simulation and dyno results point to areas where your model assumptions need refinement. Maybe heat transfer through the cylinder wall is higher than predicted, or ring seal isn’t as good as assumed. Each round of testing tightens the correlation between your digital model and reality, making the model more trustworthy for the next design change. Professional engine programs go through dozens of these loops. For a student or independent project, even two or three iterations dramatically improve the final product.