article
Flywheels: storing energy in a spinning disc
A beginner-friendly engineering guide to flywheels, explaining rotational energy storage, speed smoothing, inertia sizing, failure modes, and practical machine-design judgment.
Published Jul 10, 2026
A single-cylinder engine keeps rotating between power strokes. A punch press needs a short violent burst of torque for only a fraction of each cycle. A workshop grinder feels steadier after a heavy wheel is fitted. All three situations use the same machine element: a rotating mass that stores energy when the machine has extra power and gives some of it back when the machine momentarily needs more.
That rotating energy bank is the flywheel. Beginners often think of a flywheel as just a heavy disc bolted to a shaft. That is true physically, but incomplete as engineering. A flywheel is really an inertia device that smooths speed fluctuation, stores short-term energy, and makes a machine less twitchy. The useful question is not "how heavy is it?" but "how much mass moment of inertia does it provide at the working radius, and how much speed variation are we willing to tolerate?"
The plain-language picture
Imagine pushing a child's roundabout. If it is very light, one extra shove makes it speed up sharply and a small opposing force slows it quickly. If it is heavy and the mass sits far from the center, the motion changes more reluctantly. That reluctance to change rotational speed is rotational inertia. A flywheel uses that reluctance on purpose.
When the driving torque from the engine or motor is greater than the machine's resisting torque, the flywheel speeds up slightly and stores more rotational energy. When the machine suddenly demands more torque than the prime mover can supply at that instant, the flywheel slows slightly and releases part of that stored energy. The speed does not stay perfectly constant; it swings a little above and below the average value. Good flywheel design is about keeping that swing small enough for the machine to work properly without making the rotating assembly unnecessarily heavy, costly, or dangerous.
Figure 1: A flywheel may look like a simple disc on the bench, but the useful engineering is hidden in its rim radius, hub connection, balance quality, and the speed range it must survive.
What a flywheel really does
A flywheel does not create energy. It only stores some of the energy already supplied by the driver. That sounds obvious, but it matters because beginners sometimes expect a flywheel to "add power." It cannot do that. What it can do is redistribute power over time. A motor that is too small for a repeated short torque peak might still run the machine if the average power is adequate and the flywheel can cover the brief peaks.
That is why flywheels are common in reciprocating engines, presses, shears, rolling machinery, indexing drives, grinders, and energy-recovery devices. They reduce cyclic speed variation, soften shock loads seen by the motor, and sometimes help a machine coast through a dead zone. In vehicles, the same idea also smooths clutch engagement and makes engine idle less jerky. The cost is slower acceleration, extra bearing load, extra stored hazard if the rotor bursts, and higher shaft stress during transient torque changes.
The governing physics
The most important flywheel equation is the rotational kinetic energy relation:
E = 0.5 x J x omega^2
where E is stored energy in J, J is mass moment of inertia in kg.m^2, and omega is angular speed in rad/s. The equation shows why speed matters so much. Doubling speed multiplies stored energy by four.
The second idea is how inertia depends on geometry. For a solid disc of mass m and radius r:
J = 0.5 x m x r^2
For a thin rim where most of the mass sits near one radius:
J ~= m x r^2
This is why many flywheels put extra material toward the rim instead of near the hub. Mass close to the center adds weight, but it contributes much less inertia because the radius term is squared. If you want energy storage without wasting material, rim radius is extremely valuable.
The third relation connects torque and angular acceleration:
T = J x alpha
where T is torque in N.m and alpha is angular acceleration in rad/s^2. This tells you how strongly the flywheel resists speed change. A large J smooths operation, but it also demands more torque during startup, braking, reversing, or emergency stops.
Flywheel sizing often uses the allowable coefficient of speed fluctuation:
Cs = (omega_max - omega_min) / omega_avg
If the speed variation is modest, the useful energy swing the flywheel can exchange over a cycle is approximated by:
Delta E ~= J x omega_avg^2 x Cs
This relation comes from taking the difference between the high-speed and low-speed values of 0.5 x J x omega^2 and simplifying around the average speed. It is one of the most practical beginner formulas because it converts a machine's energy fluctuation problem into a required inertia.
Figure 2: In a press drive, the motor feeds energy into the flywheel gradually, and the flywheel returns part of that energy quickly during the short high-force portion of the crank cycle.
Worked example 1: how much energy a workshop flywheel stores
A steel flywheel on a small test rig can be approximated as a solid disc of mass 85 kg and radius 0.32 m. Estimate its mass moment of inertia and the rotational energy stored at 600 rpm.
First find the inertia:
J = 0.5 x m x r^2 = 0.5 x 85 x 0.32^2 = 4.35 kg.m^2
Now convert speed to angular speed:
omega = 2 x pi x 600 / 60 = 62.8 rad/s
Stored energy is therefore:
E = 0.5 x 4.35 x 62.8^2 = 8,580 J
So the flywheel stores about 8.6 kJ. That is enough to matter in a small machine, but it is not magic. If one cycle needs more than that energy swing, this flywheel alone cannot keep speed variation small.
Worked example 2: sizing inertia from energy fluctuation
A crank press needs an extra 18 kJ of energy during the working part of each stroke beyond what the motor can deliver instantly. The average flywheel speed is 300 rpm, and the allowable coefficient of speed fluctuation is 0.08. Estimate the required flywheel inertia. If most of the mass can be concentrated near a mean rim radius of 0.75 m, estimate the required rim-dominated mass.
First convert average speed:
omega_avg = 2 x pi x 300 / 60 = 31.4 rad/s
Now rearrange the energy swing formula:
J = Delta E / (omega_avg^2 x Cs)
J = 18,000 / (31.4^2 x 0.08) = 228 kg.m^2
That is a large inertia, which is normal for impact-style machinery.
If the flywheel behaves roughly like a rim, then J ~= m x r^2, so:
m = J / r^2 = 228 / 0.75^2 = 405 kg
The rim-dominated mass would be about 405 kg. The interpretation is important: once the allowable speed fluctuation is tight and the energy pulse is large, flywheels become heavy very quickly. The machine designer may then reconsider process timing, motor size, or acceptable speed swing instead of blindly adding more iron.
Worked example 3: startup torque penalty
A fabricated flywheel and shaft assembly has a combined inertia of 9.5 kg.m^2. The drive must accelerate it from rest to 450 rpm in 4 s. Ignoring friction and load torque for the moment, estimate the average accelerating torque.
Convert final speed:
omega = 2 x pi x 450 / 60 = 47.1 rad/s
Angular acceleration is:
alpha = 47.1 / 4 = 11.8 rad/s^2
Required torque is:
T = J x alpha = 9.5 x 11.8 = 112 N.m
The drive therefore needs about 112 N.m just to spin up the inertia, before overcoming any process load. This is the other side of flywheel design: a smoother machine in steady operation can become harder to start and stop.
Where beginners underestimate flywheels
The first beginner mistake is confusing mass with inertia. Two flywheels of equal mass can behave very differently if one keeps more material near the rim. The second is checking stored energy at one maximum speed and forgetting the usable energy swing between the highest and lowest allowed speeds. The third is ignoring the shaft, key, taper lock, or shrink-fit connection. A flywheel that stores energy safely in theory can still fail at the hub because torque reversals and overspeed stresses concentrate there.
Another common oversight is forgetting that flywheel energy rises with the square of speed, but so do many damaging stresses. Running "just a little faster" may put the rim, spokes, or bolted joint into a very different risk zone. That is why overspeed protection and balancing are not optional housekeeping tasks. They are core design controls.
Assumptions and their limits
The simple formulas above assume the flywheel behaves like a rigid body, that the speed range is modest, and that the inertia value is known accurately about the real axis of rotation. Real assemblies include shafts, hubs, bolts, couplings, and occasionally flexible spokes. If the rotor is large or runs fast, stress analysis and dynamic balance become just as important as energy arithmetic.
The energy-swing formula also assumes the machine cycle can be represented by an average speed with a small fluctuation about it. If the speed dip is large, or if the torque pulse is highly irregular, the approximation becomes less reliable and a full torque-versus-crank-angle or time-domain model is better. For engine flywheels, combustion variation, misfire, and torsional vibration can make the real behavior harsher than a clean average-power calculation suggests.
Common failure modes and what they look like
- Overspeed burst risk: cracks, casting defects, or excessive hoop stress can cause catastrophic rim or disc failure at high speed.
- Poor balance: the machine shakes, bearings run hot, fasteners loosen, and surface finish worsens in connected equipment.
- Hub or keyway distress: fretting, ovality, loose fits, or cracked keyways show that the torque transfer path is being abused.
- Guarding damage after rub: shaft misalignment or bearing wear lets the flywheel run out and contact nearby covers or guards.
- Fatigue at spokes or web transitions: repeated acceleration cycles and stress raisers produce cracks where section changes are abrupt.
- Braking overload: emergency stops overheat the brake or twist the shaft because the stored flywheel energy was underestimated.
Figure 3: A flywheel is never only an energy calculation. Safe service depends on balance, guarding, shaft support, and confidence that the rotating mass can survive its full speed range without crack growth or contact rub.
Practical rules of thumb
- Before adding flywheel mass, decide what speed variation the machine can actually tolerate. That target drives the inertia.
- If extra inertia is needed, material near the rim is usually more effective than material near the hub.
- Check startup torque, stopping time, and emergency braking whenever flywheel inertia increases.
- Do not treat balancing as a finishing step. For high-speed rotors, balance quality is part of the design itself.
- Any machine with a large flywheel deserves robust guarding because the hazard remains even after power is cut.
- If the duty cycle contains repeated torque pulses, inspect hubs, keys, and shaft shoulders as carefully as the rim.
How standards and design practice treat flywheels
In practice, standards do not look at flywheels only as energy devices. Rotating-element guidance usually separates four concerns: rotor integrity at speed, acceptable balance quality, safe guarding and containment, and dependable shaft-to-hub torque transmission. Balancing standards such as the ISO 21940 family shape how engineers specify residual unbalance for rotors. General machinery safety standards push designers toward guards, overspeed protection, and safe run-down behavior. Internal company standards often add mandatory overspeed margins, inspection intervals, and non-destructive checks for large cast or fabricated flywheels.
The standards mindset is therefore broader than one inertia equation. Designers want evidence that the flywheel will not burst at maximum credible speed, that bearing loads and vibration stay acceptable, that the key or interference fit survives the torque pulses, and that a person cannot accidentally reach the rotating mass during coast-down. A technically correct energy calculation is only the beginning of that review.
Engineering judgment
A flywheel is best thought of as a time-shifting device for mechanical energy. It smooths short-term mismatches between what the driver can supply and what the machine demands. That makes it extremely useful, but only when the designer respects the penalty: more stored energy also means more startup torque, more coast-down hazard, and more consequence if balance or overspeed control is poor.
When you evaluate a flywheel as a beginner, ask five questions in order. How much energy must be exchanged over one cycle? What speed fluctuation is acceptable? Where is the mass located relative to the axis? What torque path carries acceleration, deceleration, and shock loads into the shaft? And what happens if the rotor overspeeds, cracks, or keeps spinning after power is removed? Those questions usually separate a sensible flywheel design from a heavy spinning guess.
If you want the companion article on the machine element that must absorb this stored energy during a stop, continue with Brakes: turning motion into heat on purpose, or browse the full EnggTools engineering article library.