Normal Acceleration from Velocity and Acceleration

Why Separate Acceleration into Components?

Acceleration describes how velocity changes, but its effect on motion can take different forms. One part changes the particle’s speed; another changes the direction in which it is moving. The magnitude of the second part is called the normal acceleration.

This tutorial focuses on calculating that magnitude directly from the velocity and acceleration vectors. The method applies at an instant when the velocity is nonzero. It is useful even when the motion is nonuniform, meaning the speed may change as well as the direction.

Definition and Formula

Let a particle have velocity \(\mathbf{v}\) and acceleration \(\mathbf{a}\) at a given instant, with \(\mathbf{v}\ne\mathbf{0}\). The unit vector in the direction of motion is \(\mathbf{T}=\mathbf{v}/|\mathbf{v}|\). The component of acceleration parallel to the velocity has signed value \(\mathbf{a}\cdot\mathbf{T}\). Subtracting that parallel component from \(\mathbf{a}\) leaves the component perpendicular to the velocity.

Definition: When \(\mathbf{v}\ne\mathbf{0}\), the normal acceleration vector is \(\mathbf{a}_{N}=\mathbf{a}-(\mathbf{a}\cdot\mathbf{T})\mathbf{T}\), where \(\mathbf{T}=\mathbf{v}/|\mathbf{v}|\). The magnitude of normal acceleration is \(a_N=|\mathbf{a}_N|\).

In three dimensions, this magnitude can be computed without first finding the unit vector or the normal component vector:

Formula: If \(\mathbf{v}\ne\mathbf{0}\), then \(a_N=\dfrac{|\mathbf{v}\times\mathbf{a}|}{|\mathbf{v}|}\).

The numerator is the magnitude of the cross product, and the denominator is the speed. The formula gives a nonnegative magnitude. It does not give the direction of the normal acceleration vector.

Why the Cross-Product Formula Works

The definition makes clear that the normal component is what remains after removing the part parallel to \(\mathbf{v}\). To connect that definition to the formula, first square the magnitude of the normal component. Since \(\mathbf{T}\) is a unit vector, \(\mathbf{a}\cdot\mathbf{T}\) is the signed scalar component of \(\mathbf{a}\) along \(\mathbf{T}\).

\begin{aligned} a_N^2 &=|\mathbf{a}-(\mathbf{a}\cdot\mathbf{T})\mathbf{T}|^2\\ &=|\mathbf{a}|^2-2(\mathbf{a}\cdot\mathbf{T})^2 +(\mathbf{a}\cdot\mathbf{T})^2|\mathbf{T}|^2\\ &=|\mathbf{a}|^2-(\mathbf{a}\cdot\mathbf{T})^2\\ &=|\mathbf{a}|^2-\frac{(\mathbf{v}\cdot\mathbf{a})^2}{|\mathbf{v}|^2}. \end{aligned}

Here the third line uses \(|\mathbf{T}|=1\), and the last uses \(\mathbf{T}=\mathbf{v}/|\mathbf{v}|\). The cross-product identity \(|\mathbf{v}\times\mathbf{a}|^2=|\mathbf{v}|^2|\mathbf{a}|^2-(\mathbf{v}\cdot\mathbf{a})^2\) then gives the desired result after dividing by \(|\mathbf{v}|^2\) and taking the nonnegative square root:

a_N^2 =\frac{|\mathbf{v}|^2|\mathbf{a}|^2-(\mathbf{v}\cdot\mathbf{a})^2}{|\mathbf{v}|^2} =\frac{|\mathbf{v}\times\mathbf{a}|^2}{|\mathbf{v}|^2}, \qquad a_N=\frac{|\mathbf{v}\times\mathbf{a}|}{|\mathbf{v}|}.

This also shows why the method works for nonuniform motion: it removes the component of acceleration along the instantaneous velocity, whether that component speeds the particle up, slows it down, or is zero.

A Reliable Calculation Process

When \(\mathbf{v}\) and \(\mathbf{a}\) depend on time, evaluate both at the same instant before using the formula. If they come from a position function, differentiate the position to obtain velocity and differentiate velocity to obtain acceleration. Those are the standard definitions of velocity and acceleration for a differentiable trajectory.

Find the instantaneous vectors.
Evaluate \(\mathbf{v}\) and \(\mathbf{a}\) at the specified time, and check that \(\mathbf{v}\ne\mathbf{0}\).

Compute the cross product and speed.
Find \(\mathbf{v}\times\mathbf{a}\), its magnitude, and \(|\mathbf{v}|\).

Divide and check.
Calculate \(a_N=|\mathbf{v}\times\mathbf{a}|/|\mathbf{v}|\). When useful, confirm it by removing the squared tangential component from \(|\mathbf{a}|^2\).

Worked Examples

Worked Example: Compute Normal Acceleration from Two Vectors

At an instant, suppose \(\mathbf{v}=\langle 1,2,2\rangle\) metres per second and \(\mathbf{a}=\langle 2,-1,0\rangle\) metres per second squared. Find the magnitude of normal acceleration.

First find the speed:

|\mathbf{v}|=\sqrt{1^2+2^2+2^2}=\sqrt{9}=3\ \text{m/s}.

Next compute the cross product:

\mathbf{v}\times\mathbf{a} =\langle 2(0)-2(-1),\ 2(2)-1(0),\ 1(-1)-2(2)\rangle =\langle 2,4,-5\rangle.

Its magnitude is \(\sqrt{2^2+4^2+(-5)^2}=\sqrt{45}=3\sqrt{5}\) metres per second squared. Therefore,

a_N=\frac{3\sqrt{5}\ \text{m/s}^2}{3\ \text{m/s}} =\sqrt{5}\ \text{m/s}^2.

Check: \(\mathbf{v}\cdot\mathbf{a}=1(2)+2(-1)+2(0)=0\), so the acceleration has no component parallel to the velocity. The entire acceleration is normal. Indeed, \(|\mathbf{a}|=\sqrt{2^2+(-1)^2+0^2}=\sqrt{5}\ \text{m/s}^2\), agreeing with the result.

Worked Example: Find Normal Acceleration from a Position Function

A particle follows \(\mathbf{r}(t)=\langle t^2,t^3,t\rangle\) metres. Find its normal acceleration at \(t=1\) second.

Differentiate the position to obtain the velocity, then differentiate again to obtain acceleration:

\mathbf{v}(t)=\langle 2t,3t^2,1\rangle\ \text{m/s}, \qquad \mathbf{a}(t)=\langle 2,6t,0\rangle\ \text{m/s}^2.

At \(t=1\), these are \(\mathbf{v}=\langle 2,3,1\rangle\) metres per second and \(\mathbf{a}=\langle 2,6,0\rangle\) metres per second squared. The speed is \(|\mathbf{v}|=\sqrt{2^2+3^2+1^2}=\sqrt{14}\) metres per second, which is nonzero.

\mathbf{v}\times\mathbf{a} =\langle 3(0)-1(6),\ 1(2)-2(0),\ 2(6)-3(2)\rangle =\langle -6,2,6\rangle.

Thus \(|\mathbf{v}\times\mathbf{a}|=\sqrt{(-6)^2+2^2+6^2}=\sqrt{76}=2\sqrt{19}\) metres per second squared. The formula gives

a_N=\frac{2\sqrt{19}}{\sqrt{14}}\ \text{m/s}^2 =\sqrt{\frac{38}{7}}\ \text{m/s}^2.

Check using the parallel component: \(\mathbf{v}\cdot\mathbf{a}=2(2)+3(6)+1(0)=22\), and \(|\mathbf{a}|^2=2^2+6^2+0^2=40\). Therefore, \(a_N^2=|\mathbf{a}|^2-(\mathbf{v}\cdot\mathbf{a})^2/|\mathbf{v}|^2 =40-22^2/14=40-242/7=38/7\). This agrees with the squared result from the cross product.

Worked Example: Separate a Nonzero Tangential Component

At an instant, let \(\mathbf{v}=\langle 3,0,4\rangle\) metres per second and \(\mathbf{a}=\langle 1,2,2\rangle\) metres per second squared. Compute the normal acceleration even though the acceleration also has a component along the velocity.

The speed is \(|\mathbf{v}|=\sqrt{3^2+0^2+4^2}=5\) metres per second. The cross product is

\mathbf{v}\times\mathbf{a} =\langle 0(2)-4(2),\ 4(1)-3(2),\ 3(2)-0(1)\rangle =\langle -8,-2,6\rangle.

Its magnitude is \(\sqrt{(-8)^2+(-2)^2+6^2}=\sqrt{104}=2\sqrt{26}\) metres per second squared. Hence

a_N=\frac{2\sqrt{26}}{5}\ \text{m/s}^2.

Check: \(\mathbf{v}\cdot\mathbf{a}=3(1)+0(2)+4(2)=11\), so the signed tangential component is \(11/5\) metres per second squared. Also \(|\mathbf{a}|^2=1^2+2^2+2^2=9\). The squared normal magnitude is \(9-(11/5)^2=9-121/25=104/25\), whose nonnegative square root is \(2\sqrt{26}/5\) metres per second squared.

When the Formula Applies—and When It Does Not

The denominator is the speed, so the formula requires \(|\mathbf{v}|>0\). At an instant when velocity is zero, there is no defined direction of motion \(\mathbf{T}=\mathbf{v}/|\mathbf{v}|\), and this formula cannot determine a normal component there. Do not substitute zero into the denominator or treat the resulting expression as a normal acceleration.

If \(\mathbf{v}\times\mathbf{a}=\mathbf{0}\) while \(\mathbf{v}\ne\mathbf{0}\), then the normal acceleration is zero. This means acceleration is parallel to velocity (or is the zero vector); it does not necessarily mean that acceleration itself is zero. For example, acceleration parallel to velocity can change the speed while leaving no perpendicular component.

The formula gives a magnitude, not a signed quantity. Reversing the order in the cross product changes its direction, since \(\mathbf{a}\times\mathbf{v}=-(\mathbf{v}\times\mathbf{a})\), but does not change its magnitude. Thus either order gives the same value after taking the magnitude.

Common Mistake: Dividing by the Wrong Quantity

For the vectors in the third example, the speed is \(5\) metres per second and the cross-product magnitude is \(2\sqrt{26}\) metres per second squared. A common incorrect calculation divides by the square of the speed:

\text{Incorrect:}\qquad \frac{|\mathbf{v}\times\mathbf{a}|}{|\mathbf{v}|^2} =\frac{2\sqrt{26}}{25}.

The formula calls for division by speed, not speed squared. The incorrect expression also has units of inverse seconds, rather than acceleration units. The correct calculation is \(\frac{2\sqrt{26}\ \text{m/s}^2}{5\ \text{m/s}} =\frac{2\sqrt{26}}{5}\ \text{m/s}^2\). Checking units is a useful way to catch this particular error.

Why Normal Acceleration Is Useful

The total acceleration combines its parallel and perpendicular effects. The tangential component describes change in speed, while the normal component describes change in direction. Their magnitudes satisfy \(|\mathbf{a}|^2=a_T^2+a_N^2\), where \(a_T=(\mathbf{v}\cdot\mathbf{a})/|\mathbf{v}|\) is the signed tangential component. Consequently, \(a_N\) cannot exceed \(|\mathbf{a}|\). This relationship provides a second check on a calculation and helps distinguish turning from speeding up or slowing down.

The cross-product formula is particularly convenient when velocity and acceleration are already available as components: it avoids separately constructing the perpendicular direction. It remains an instantaneous calculation, and it says nothing by itself about the particle’s entire path.

Check Your Understanding

Use the definitions and formula above. Give units where appropriate, and explain any condition needed for your calculation.

If \(\mathbf{v}=\langle 0,2,0\rangle\) and \(\mathbf{a}=\langle 3,0,4\rangle\), what is the magnitude of normal acceleration?
Why must the velocity be nonzero to use the formula?
If \(\mathbf{v}\ne\mathbf{0}\) and \(\mathbf{v}\times\mathbf{a}=\mathbf{0}\), what can be concluded about normal acceleration?
For \(\mathbf{v}=\langle 1,0,0\rangle\) and \(\mathbf{a}=\langle 2,3,0\rangle\), verify the result using both the cross product and the parallel-component calculation.
What is the units error in dividing \(|\mathbf{v}\times\mathbf{a}|\) by \(|\mathbf{v}|^2\)?