Calculus › Section L · Vectors and the geometry of space › Topic L.3
The dot product
Vectors can be multiplied, and the first way, the dot product, turns two vectors into a single number that measures how strongly they point the same way. From that one number come angles, perpendicularity tests, shadows and work.
Multiplying vectors into a number
There are two useful ways to multiply vectors, and they produce different things: the dot product produces a number, and the cross product, covered next, produces a new vector. The dot product of \(\mathbf{u}\) and \(\mathbf{v}\) multiplies matching components and adds the results:
For \(\langle 2, -1, 3\rangle \cdot \langle 1, 4, 2\rangle = 2 - 4 + 6 = 4\). The result is a scalar, which is why the dot product is also called the scalar product. It behaves like ordinary multiplication where it can: it is commutative, it distributes over addition, and \(\mathbf{v} \cdot \mathbf{v} = v_1^2 + v_2^2 + v_3^2 = \|\mathbf{v}\|^2\), the squared length.
The geometric meaning
The component formula hides the dot product's real significance. If \(\theta\) is the angle between two nonzero vectors, then
A vector has its full length in its own direction and a shortened shadow in any other direction, and the shadow is scaled by \(\cos\theta\). The dot product multiplies one vector's length by the other's length along its direction. So it measures alignment: pointing the same way gives a large positive value, perpendicular gives zero, and opposite directions give a negative value. It also gives a quick length bound, since \(|\mathbf{u} \cdot \mathbf{v}|\) never exceeds \(\|\mathbf{u}\|\,\|\mathbf{v}\|\).
Worked example: the angle between vectors
Find the angle between \(\mathbf{u} = \langle 2, 0, 2\rangle\) and \(\mathbf{v} = \langle 3, 3, 0\rangle\). Compute the ingredients: \(\mathbf{u} \cdot \mathbf{v} = 6 + 0 + 0 = 6\); \(\|\mathbf{u}\| = \sqrt{8} = 2\sqrt{2}\); \(\|\mathbf{v}\| = \sqrt{18} = 3\sqrt{2}\). Then
so \(\theta = 60°\). Both vectors are face diagonals of a cube, one slanting up in the \(xz\)-plane and one in the \(xy\)-plane; the components were chosen so the arithmetic comes out clean, and \(\cos\theta = \frac{1}{2}\) pins the angle at exactly \(60°\).
Perpendicularity
Setting \(\theta = 90°\) in the geometric formula makes \(\cos\theta = 0\), so the dot product vanishes. This gives the cleanest perpendicular test in mathematics: two nonzero vectors are orthogonal, meaning perpendicular, exactly when their dot product is \(0\). For instance \(\langle 1, 2, -1\rangle \cdot \langle 3, -1, 1\rangle = 3 - 2 - 1 = 0\), so those vectors meet at a right angle. The test also works in reverse to build a perpendicular: to find a vector orthogonal to both \(\mathbf{u}\) and \(\mathbf{v}\), solve \(\mathbf{w} \cdot \mathbf{u} = 0\) and \(\mathbf{w} \cdot \mathbf{v} = 0\), two linear equations in the unknown components.
Worked example: building an orthogonal vector
Find a value of \(c\) making \(\langle 2, 1, 2\rangle\) orthogonal to \(\langle 3, 4, c\rangle\). The dot product must be zero: \(2 \cdot 3 + 1 \cdot 4 + 2c = 10 + 2c = 0\), so \(c = -5\). Check: \(\langle 2, 1, 2\rangle \cdot \langle 3, 4, -5\rangle = 6 + 4 - 10 = 0\). Geometrically, the tip of \(\langle 3, 4, c\rangle\) slides along a vertical line as \(c\) changes, and \(c = -5\) is the unique height where the sliding vector stands at a right angle to \(\langle 2, 1, 2\rangle\). One equation, one unknown: orthogonal directions are cheap to find.
Projections: one vector's shadow on another
The dot product measures how much of \(\mathbf{u}\) lies along \(\mathbf{v}\). Made precise, this is the projection of \(\mathbf{u}\) onto \(\mathbf{v}\):
Picture sunlight shining perpendicular to \(\mathbf{v}\): the projection is the shadow \(\mathbf{u}\) casts on the line of \(\mathbf{v}\). The fraction \(\frac{\mathbf{u} \cdot \mathbf{v}}{\|\mathbf{v}\|^2}\) says how many copies of \(\mathbf{v}\) the shadow spans, so the shadow is always some scalar multiple of \(\mathbf{v}\). The related number \(\frac{\mathbf{u} \cdot \mathbf{v}}{\|\mathbf{v}\|}\), without the final vector, is the signed length of the shadow, negative when \(\mathbf{u}\) points against \(\mathbf{v}\).
Worked example: force and work
Work is the dot product in its most physical disguise. A constant force \(\mathbf{F}\) acting through a displacement \(\mathbf{d}\) does work \(W = \mathbf{F} \cdot \mathbf{d}\): only the part of the force along the motion counts. A gardener drags a lawnmower with force \(\mathbf{F} = \langle 40, 30\rangle\) newtons (a 50-newton push) while it moves \(\mathbf{d} = \langle 10, 0\rangle\) metres. The work is \(40 \cdot 10 + 30 \cdot 0 = 400\) joules. The 30 newtons of upward pull does no work on the horizontally moving mower, though it does lighten the load on the wheels. Had the gardener dragged with \(\mathbf{F} = \langle 30, 40\rangle\) instead — still exactly 50 newtons of force — the work would drop to 300 joules: direction, not just strength, gets the job done.
Direction angles and direction cosines
A unit vector's components have a neat reading: if \(\mathbf{u}\) makes angles \(\alpha\), \(\beta\) and \(\gamma\) with the \(x\)-, \(y\)- and \(z\)-axes, then the dot product with each standard unit vector gives \(\cos\alpha = \frac{u_1}{\|\mathbf{u}\|}\), \(\cos\beta = \frac{u_2}{\|\mathbf{u}\|}\) and \(\cos\gamma = \frac{u_3}{\|\mathbf{u}\|}\). These direction cosines are just the components of the unit vector in the direction of \(\mathbf{u}\), and they satisfy \(\cos^2\alpha + \cos^2\beta + \cos^2\gamma = 1\), since the unit vector has length 1. For \(\langle 3, 4, 12\rangle\), with magnitude 13, the direction cosines are \(\frac{3}{13}, \frac{4}{13}, \frac{12}{13}\), so the vector makes angles of about \(77°\), \(72°\) and \(23°\) with the \(x\)-, \(y\)- and \(z\)-axes: mostly vertical, as the large \(z\)-component suggests.
Common mistakes to avoid
The most common error is treating the dot product's result as a vector: it is a number, and writing it with angle brackets is a category mistake. The second is dividing by \(\|\mathbf{v}\|\) instead of \(\|\mathbf{v}\|^2\) in the projection formula; the square in the denominator pairs with the vector \(\mathbf{v}\) that carries its own length. The third is concluding that a zero dot product means one of the vectors is zero: it means they are perpendicular, or one of them is the zero vector.
Key takeaways
The dot product \(\mathbf{u} \cdot \mathbf{v} = u_1v_1 + u_2v_2 + u_3v_3\) is a scalar equal to \(\|\mathbf{u}\|\|\mathbf{v}\|\cos\theta\), so it measures alignment. It finds angles through \(\cos\theta = \frac{\mathbf{u} \cdot \mathbf{v}}{\|\mathbf{u}\|\|\mathbf{v}\|}\) and tests perpendicularity through \(\mathbf{u} \cdot \mathbf{v} = 0\). The projection \(\frac{\mathbf{u} \cdot \mathbf{v}}{\|\mathbf{v}\|^2}\mathbf{v}\) is the shadow of \(\mathbf{u}\) on \(\mathbf{v}\), and the physical shadow of the idea is work: \(W = \mathbf{F} \cdot \mathbf{d}\). Direction cosines read a vector's tilt against each axis.
Flashcards
Flashcard
u · v = u₁v₁ + u₂v₂ + u₃v₃
Test your understanding
Question 1 of 5
What is ⟨3, −2, 5⟩ · ⟨1, 4, −1⟩?
Question 2 of 5
What is the angle between ⟨2, 0, 2⟩ and ⟨3, 3, 0⟩?
Question 3 of 5
For which c is ⟨2, 1, 2⟩ orthogonal to ⟨3, 4, c⟩?
Question 4 of 5
What is the projection of ⟨6, 2⟩ onto ⟨3, 4⟩?
Question 5 of 5
A force ⟨40, 30⟩ N acts through a displacement ⟨10, 0⟩ m. How much work is done?
Standards: Beyond AP
