How to Choose Flywheel Inertia: Load Selection for Power vs Strength

You bought a flywheel unit, or your gym did, and now there are three or four inertia cones sitting in a bag with nothing on them but a number stamped into the plastic — 0.025, 0.050, 0.075, 0.100. Most coaches load whatever cone was on the shaft when the box was opened and never touch it again. That means half the athletes in the room end up training a power quality on a setting built for eccentric overload, and the other half grind through a strength block on resistance so light the muscle never gets loaded hard enough to adapt. The inertia you select is the single biggest variable in flywheel training — bigger than sets, reps, or exercise choice — and almost nobody explains it beyond the manufacturer's one-page quick-start card.

How Flywheel Resistance Is Actually Generated

A flywheel device has no external weight stack. A strap winds around a rotating shaft attached to one or more discs; pulling the strap (the concentric phase) spins the discs and unwinds the strap fully, and the kinetic energy stored in the spinning discs then rewinds the strap back toward you, creating the eccentric resistance. There is no fixed number of kilograms anywhere in this system — the load you feel is a direct product of how hard and how fast you pulled during the concentric phase. Pull harder, the flywheel spins faster, and more energy comes back at you on the way down. This is the entire basis of isoinertial training, and it is also why two athletes on the identical inertia setting, doing the identical exercise, can report completely different subjective loads: the person who accelerates aggressively through the concentric phase creates their own overload, while a hesitant, submaximal pull barely spins the disc and the eccentric phase feels almost weightless.

The physics reduce to one equation coaches should actually know: torque required equals moment of inertia (I, in kg·m²) multiplied by angular acceleration. Raise the inertia number and you raise the torque needed to accelerate the disc at any given rate — which is why a heavier disc setting demands more force from the first inch of the pull. It also means a heavier disc stores more kinetic energy per rep, which is returned as a larger eccentric brake at the top of the strap's travel. Low inertia does the opposite on both counts: less force needed to spin it up, and less energy stored to return.

Why the Inertia Number Doesn't Map to a 1RM Percentage

Coaches coming from barbell training want to translate 0.050 kg·m² into something like 70% of squat 1RM, and that mapping doesn't exist in a stable, transferable way. Inertia interacts with how fast the athlete moves, so the same disc can feel like a moderate strength stimulus to an athlete who pulls conservatively and a maximal-velocity power stimulus to an athlete who attacks the concentric phase. Sabido, Hernández-Davó and Pereyra-Gerber (2018) tested three inertia loads (0.025, 0.05 and 0.075 kg·m²) during the flywheel squat and found that as inertia increased, peak and mean power tended to decrease while peak force and eccentric overload increased — the classic force-velocity trade-off playing out on a single device. The effect sizes reported were moderate and the sample was small (a single-session design in trained but non-elite lifters), so the exact cutoffs shouldn't be read as universal prescriptions, but the direction of the relationship — lower inertia biases the movement toward velocity and power, higher inertia biases it toward force and eccentric braking demand — has held up across the flywheel literature broadly.

Choosing Inertia When the Goal Is Power

When the target quality is power output — sprinters in-season, jumpers, field-sport athletes doing a speed-strength block — the goal is to let the athlete reach high angular velocity almost immediately after initiating the pull. That means erring toward the lower end of the available range, typically 0.010 to 0.030 kg·m² depending on the exercise and the athlete's training age. A 68 kg sprinter and a 110 kg rugby prop will not use the same disc for the same exercise even if both are chasing power; body mass and limb length change how quickly a given inertia can be spun up, so the number on the cone is a starting point, not a prescription. In practice, most athletes performing a flywheel half-squat or lunge for power reach their peak power output somewhere in the lower third of the inertia range available on the device, and power drops off measurably once inertia climbs high enough that the athlete has to slow down to control the eccentric return. If an athlete can't complete the concentric phase in well under a second, the inertia is almost certainly too high for a power-focused session regardless of what the label says.

Choosing Inertia for Strength, Eccentric Overload and Hypertrophy

Strength and hypertrophy blocks move in the opposite direction — the coach wants the athlete fighting the disc through a fuller range at a controlled speed, with an eccentric brake substantial enough to create genuine muscle damage and mechanical tension. This is where flywheel training earns its reputation: because the device can generate eccentric force above what the athlete produced concentrically, a well-executed set at 0.075–0.100 kg·m² creates an eccentric overload stimulus that free weights can't replicate without a spotter or weight releasers. Maroto-Izquierdo and colleagues' 2017 systematic review and meta-analysis of eccentric-overload flywheel training pooled data across roughly 15 studies and reported a moderate-to-large effect on lower-body muscle hypertrophy compared with traditional weight training at matched perceived effort, with the strongest signal in quadriceps cross-sectional area after training blocks of 4–10 weeks. The review's authors noted wide variability in dosing across the included studies and a lack of long-term (beyond 10-week) data, so treat the hypertrophy advantage as real but not unlimited — it doesn't mean higher inertia is always better, only that it's the more efficient tool once eccentric overload is specifically the target.

For tendon and rehab applications, the priority shifts again: moderate inertia (0.075–0.150 kg·m² on isolation exercises like knee extension or calf raise) combined with a deliberately slow, controlled eccentric — actively resisting the return rather than just absorbing it — is what produces the collagen loading stimulus, not maximal inertia.

Testing Individual Response Instead of Guessing

The most reliable way to select inertia is a short ramping protocol run once at the start of a training block and re-checked every 4–6 weeks as the athlete adapts. Have the athlete perform 3–4 reps at each of three or four available inertia settings, resting 90 seconds to 2 minutes between settings, while a velocity sensor records peak power and peak concentric velocity on every rep. Plot power against inertia for that athlete on that specific exercise; most athletes show a power output that rises through the low-to-moderate inertia settings and then plateaus or falls once inertia gets heavy enough that velocity drops faster than force rises. The setting nearest that peak is the appropriate choice for a power-oriented session; a setting one or two steps above it, chosen deliberately for its lower velocity and higher force demand, is the appropriate choice for a strength-oriented session on the same exercise. This individualizes the prescription in a way a printed chart never can, because limb length, tendon stiffness and technique efficiency all shift where that peak sits from one athlete to the next.

Common Mistakes When Choosing Inertia

Coaches who are new to flywheel training tend to make the same handful of errors. The most common is treating the kg·m² number as standardized across brands and exercises the way a kilogram plate is standardized across every barbell — it isn't; the same nominal inertia produces different demand on a squat versus a leg curl because the lever arm and limb mass involved are different, so a 0.05 setting that feels moderate on a squat can feel heavy on an isolated knee extension. A second common error is picking one inertia and running it for an entire 8-week block without progression, which stalls adaptation the same way a fixed barbell weight would in a traditional program — inertia needs to increase gradually as an athlete's peak power point on the ramping test shifts upward. A third, more physical mistake shows up in braking technique: athletes told to use a heavier disc without being taught to actively decelerate the strap end up getting yanked into the bottom position by the returning energy, which is both a poor training stimulus and a real injury risk at the knee and lower back. Cue athletes explicitly to fight the eccentric phase rather than simply survive it, and drop the inertia a step if that control breaks down before technical breakdown becomes the norm.