How this instrument works
Acceleration is a vector: it has a size and a direction, and when two causes act on the same object at right angles — gravity pulling down while a turn pushes sideways, say — the total is not the sum of the two numbers. It is the length of the diagonal they form together. That length comes from the Pythagorean theorem: |a| = √(ax² + ay²), the same relation that gives the hypotenuse of any right triangle its length from the two legs. Squaring removes the sign of each component, the addition combines their separate contributions, and the square root undoes the squaring to hand back a length in the original units.
This right-angle case is a shortcut, not a special law. The general rule for combining two vectors at any angle θ between them is the law of cosines, |a| = √(ax² + ay² − 2·ax·ay·cos θ). At θ = 90°, cos θ is zero, the cross term vanishes, and what remains is exactly the formula above — which is why perpendicular axes are the default choice for describing motion: they are the one arrangement where the components do not interfere with each other's contribution to the total.
The result is always a positive magnitude, even when one or both components are negative. An x-component of −3 m/s² and a y-component of 4 m/s² still combine to a magnitude of 5 m/s², because squaring erases the sign before the square root is taken. Direction is a separate question, answered by the angle atan2(ay, ax), not by this instrument — treat the magnitude here as the size of the vector, with the bearing tracked alongside it.
- Enter Acceleration, x-component — the horizontal or first-axis reading, in m/s² or ft/s², positive or negative as measured.
- Enter Acceleration, y-component — the perpendicular reading along the second axis, in the same or a different supported unit.
- Confirm the two axes are genuinely at right angles; the formula assumes 90° between them and will understate the true magnitude otherwise.
- Read Acceleration magnitude — the single positive value both components combine to, switchable between m/s² and ft/s².
Worked example — the 3-4-5 acceleration triangle
Take an x-component of 3 m/s² and a y-component of 4 m/s² — perhaps a data logger on a cornering vehicle reading 3 m/s² of forward acceleration and 4 m/s² of lateral acceleration at the same instant. The formula gives |a| = √(3² + 4²) = √(9 + 16) = √25 = 5.0 m/s² exactly, with no rounding at any step, which is precisely why this pair is the standard check for a Pythagorean calculator: every intermediate value lands on a whole number.
Double both inputs to 6 m/s² and 8 m/s² and the magnitude scales to exactly 10 m/s², not 14 — confirmation that the combination is not linear addition. Halve them instead, to 1.5 and 2, and the magnitude halves in turn to 2.5 m/s², the same 3-4-5 ratio holding at any scale because the Pythagorean relation is homogeneous of degree one.
Questions
Why isn't the magnitude just ax plus ay?
Because the two components point in different directions, and lengths at an angle to each other do not add like plain numbers. Adding 3 m/s² and 4 m/s² directly gives 7, but the actual resultant is 5 m/s² — the diagonal of the rectangle those two components form, not the sum of its sides. Only when one component is zero does simple addition happen to match the Pythagorean result.
What if ax and ay are not actually perpendicular?
The formula |a| = √(ax² + ay²) is only exact at 90° between the two axes. At any other angle the true magnitude follows the law of cosines, |a| = √(ax² + ay² − 2·ax·ay·cos θ), and using the simple square-root-of-squares formula anyway will overstate the resultant, since the missing cosine term is normally negative for the small non-perpendicular offsets that occur in practice.
Does a negative component change the result?
No — squaring a negative number gives the same positive value as squaring its positive counterpart, so the sign of ax or ay never changes the magnitude. An x-component of −3 m/s² and a y-component of 4 m/s² still return 5.0 m/s², identical to +3 and +4. The sign matters for direction, tracked separately as an angle, not for this instrument's output.
How does this relate to acceleration measured by an accelerometer?
A two-axis accelerometer reports exactly ax and ay separately, each along its own sensing axis, and this formula is the standard first step in reducing that raw pair to a single magnitude — used, for instance, when logging cornering and braking data from a vehicle's inertial sensor before analyzing the combined load on a traction circle.
What happens if one component is zero?
The magnitude simply equals the other component. With ax = 0 and ay = 4 m/s², |a| = √(0² + 4²) = 4 m/s² exactly — the acceleration is purely one-dimensional, so the vector sum collapses back to ordinary arithmetic. This is the edge case that confirms the formula behaves sensibly when only one axis is doing any work.
Can this formula extend to three components?
Yes — a third, perpendicular z-component simply adds another squared term inside the root: |a| = √(ax² + ay² + az²), the three-dimensional Pythagorean theorem used for full inertial-sensor readings. This instrument covers the two-axis case; a third axis is added the same way, one more square under the same root.