The last two sections have introduced some basic algebraic operations on vectorsβaddition, scalar multiplication, and the dot productβwith useful geometric interpretations. In this section, we will meet a final algebraic operation, the cross product, which again conveys important geometric information.
To begin, we must emphasize that the cross product is only defined for vectors \(\vu\) and \(\vv\) in \(\R^3\text{.}\) Also, remember that we use a right-hand coordinate system, as described in SectionΒ 9.1. In particular, recall that the vectors \(\vi\text{,}\)\(\vj\text{,}\) and \(\vk\) are oriented as shown below in FigureΒ 9.4.1. Earlier, we noticed that if we point the index finger of our right hand in the direction of \(\vi\) and our middle finger in the direction of \(\vj\text{,}\) then our thumb points in the direction of \(\vk\text{.}\)
The cross product of two vectors, \(\vu\) and \(\vv\text{,}\) will itself be a vector denoted \(\vu\times\vv\text{.}\) The direction of \(\vu\times\vv\) is determined by the right-hand rule: if we point the index finger of our right hand in the direction of \(\vu\) and our middle finger in the direction of \(\vv\text{,}\) then our thumb points in the direction of \(\vu\times\vv\text{.}\)
We begin by defining the cross products using the vectors \(\vi\text{,}\)\(\vj\text{,}\) and \(\vk\text{.}\) Referring to FigureΒ 9.4.1, explain why \(\vi\text{,}\)\(\vj\text{,}\)\(\vk\) in that order form a right-hand system. We then define \(\vi \times \vj\) to be \(\vk\) β that is \(\vi\times\vj = \vk\text{.}\)
Now explain why \(\vi\text{,}\)\(\vk\text{,}\) and \(-\vj\) in that order form a right-hand system. We then define \(\vi \times \vk\) to be \(-\vj\) β that is \(\vi\times\vk=-\vj\text{.}\)
Up to this point, the products you have seen, such as the product of real numbers and the dot product of vectors, have been commutative, meaning that the product does not depend on the order of the terms. For instance, \(2\cdot5 = 5\cdot 2\text{.}\) The table above suggests, however, that the cross product is anti-commutative: for any vectors \(\vu\) and \(\vv\) in \(\R^3\text{,}\)\(\vu\times\vv =
-\vv\times\vu\text{.}\) If we consider the case when \(\vu=\vv\text{,}\) this shows that \(\vv\times\vv = -(\vv\times\vv)\text{.}\) What does this tell us about \(\vv\times\vv\text{;}\) in particular, what vector is unchanged by scalar multiplication by \(-1\text{?}\)
Explain why \(\vu = |\vu|\vi\) and \(\vv = |\vv|\cos(\theta) \vi +
|\vv|\sin(\theta) \vj\text{.}\) Then compute the length of \(|\vu\times\vv|\text{.}\)
As we have seen in Preview ActivityΒ 9.4.1, the cross product \(\vu\times\vv\) is defined for two vectors \(\vu\) and \(\vv\) in \(\R^3\) and produces another vector in \(\R^3\text{.}\) Using the right-hand rule, we saw that
If, in addition, we assume the cross product behaves like we think a product should (e.g., the cross product distributes over vector addition), we can compute the cross product in terms of the components of general vectors to find a formula for the cross product. Doing so we see that
(Like the dot product, the cross product arises in physical applications, e.g., torque, but it is more convenient mathematically to begin from an algebraic perspective.)
At first, this may look intimidating and difficult to remember. However, if we rewrite the expression in EquationΒ (9.4.1) using determinants, important structure emerges. The determinant of a \(2\times2\) matrix is
\begin{equation*}
\left|
\begin{array}{cc}
a \amp b \\
c \amp d
\end{array}
\right|
=ad - bc.
\end{equation*}
Evaluate the dot products \(\vu\cdot(\vu\times\vv)\) and \(\vv\cdot(\vu\times\vv)\text{.}\) What does this tell you about the geometric relationship among \(\vu\text{,}\)\(\vv\text{,}\) and \(\vu\times\vv\text{?}\)
Multiplication of real numbers is associative, which means, for instance, that \((2\cdot 5)\cdot 3 = 2\cdot(5\cdot 3)\text{.}\) Is it true that the cross product of vectors is associative? For instance, is it true that \((\vu\times\vv)\times\vi = \vu\times(\vv\times\vi)\text{?}\)
The cross product satisfies the following properties, some of which were illustrated in Preview ActivityΒ 9.4.1 and may be easily verified from the definition (9.4.1).
Just as we found for the dot product, the cross product provides us with useful geometric information. In particular, both the length and direction of the cross product \(\vu\times\vv\) encode information about the geometric relationship between \(\vu\) and \(\vv\text{.}\)
We may ask whether the length \(|\vu\times\vv|\) has any relationship to the lengths of \(\vu\) and \(\vv\text{.}\) To investigate, we will compute the square of the length \(|\vu\times\vv|^2\) and denote by \(\theta\) the angle between \(\vu\) and \(\vv\text{,}\) as in SectionΒ 9.3. Doing so, we find through some significant algebra that
Note that the third property stated above says that \(\vu\times\vv = \vzero\) if \(\vu\) and \(\vv\) are parallel. This is reflected in EquationΒ (9.4.2) since \(\sin(\theta)=0\) if \(\vu\) and \(\vv\) are parallel, which implies that \(\vu\times\vv = \vzero\text{.}\)
Equation (9.4.2) also has a geometric implication. Consider the parallelogram formed by two vectors \(\vu\) and \(\vv\text{,}\) as shown in FigureΒ 9.4.5.
Remember that the area of a parallelogram is the product of its base and height. As shown in the figure, we may consider the base of the parallelogram to be \(|\vu|\) and the height to be \(|\vv|\sin(\theta)\text{.}\) This means that the area of the parallelogram formed by \(\vu\) and \(\vv\) is
The length, \(|\vu \times \vv|\text{,}\) of the cross product of vectors \(\vu\) and \(\vv\) is the area of the parallelogram determined by \(\vu\) and \(\vv\text{.}\)
Note also that if \(\vu = u_1\vi + u_2\vj + 0\vk\) and \(\vv = v_1\vi + v_2\vj + 0\vk\) are vectors in the \(xy\)-plane, then EquationΒ (9.4.1) shows that the area of the parallelogram determined by \(\vu\) and \(\vv\) is \(|\vu \times \vv| = |u_1v_2-u_2v_1|\) is the absolute value of the \(2 \times 2\) determinant \(\left|
\begin{array}{cc}
u_1 \amp u_2 \\
v_1 \amp v_2
\end{array}
\right|.\) So the absolute value of a determinant of a \(2 \times 2 \) matrix is also the area of a parallelogram.
Find the area of the parallelogram in \(\R^3\) whose vertices are \((1,0,1)\text{,}\)\((0,0,1)\text{,}\)\((2,1,0)\text{,}\) and \((1,1,0)\text{.}\) (Hint: It might be helpful to draw a picture to see how the vertices are arranged so you can determine which vectors you might use.)
Now that we understand the length of \(\vu\times\vv\text{,}\) we will investigate its direction. Remember from Preview ActivityΒ 9.4.1 that cross products involving the vectors \(\vi\text{,}\)\(\vj\text{,}\) and \(\vk\) resulted in vectors that are orthogonal to the two terms. We will see that this holds more generally.
To summarize, we have \(\vu\cdot(\vu\times\vv) = 0\text{,}\) which implies that \(\vu\) is orthogonal to \(\vu\times\vv\text{.}\) In the same way, we can show that \(\vv\) is orthogonal to \(\vu\times\vv\text{.}\) The net effect is that \(\vu\times\vv\) is a vector that is perpendicular to both \(\vu\) and \(\vv\text{,}\) and hence \(\vu\times\vv\) is perpendicular to the plane determined by \(\vu\) and \(\vv\text{.}\) Moreover, the direction of \(\vu\times\vv\) is determined by applying the right-hand rule to \(\vu\) and \(\vv\text{,}\) as we saw in Preview ActivityΒ 9.4.1. In light of our earlier work that showed \(|\vu||\vv|\sin(\theta) = |\vu\times\vv|.\text{,}\) we may now express \(\vu \times \vv\) in the following different way.
Suppose that \(\vu\) and \(\vv\) are not parallel and that \(\vn\) is the unit vector perpendicular to the plane containing \(\vu\) and \(\vv\) determined by the right-hand rule. Then
There is yet one more geometric implication we may draw from this result. Suppose \(\vu\text{,}\)\(\vv\text{,}\) and \(\vw\) are vectors in \(\R^3\) that are not coplanar and that form a three-dimension parallelepiped as shown in FigureΒ 9.4.6.
The volume of the parallelepiped is determined by multiplying \(A\text{,}\) the area of the base, by the height \(h\text{.}\) As we have just seen, the area of the base is \(|\vu\times\vv|\text{.}\) Moreover, the height \(h=|\vw|\cos(\alpha)\) where \(\alpha\) is the angle between \(\vw\) and the vector \(\vn\text{,}\) which is orthogonal to the plane formed by \(\vu\) and \(\vv\text{.}\) Since \(\vn\) is parallel to \(\vu\times\vv\text{,}\) the angle between \(\vw\) and \(\vu\times\vv\) is also \(\alpha\text{.}\) This shows that
Do the vectors \(\va = \langle 1,3,-2\rangle\text{,}\)\(\vb=\langle2,1,-4\rangle\text{,}\) and \(\vc=\langle 0, 1, 0\rangle\) in standard position lie in the same plane? Use the concepts from this section to explain.
Subsection9.4.4Torque is measured by a cross product
We have seen that the cross product enables us to produce a vector perpendicular to two given vectors, to measure the area of a parallelogram, and to measure the volume of a parallelepiped. Besides these geometric applications, the cross product also enables us to describe a physical quantity called torque.
Suppose that we would like to turn a bolt using a wrench as shown in FigureΒ 9.4.8. If a force \(\vF\) is applied to the wrench and \(\vr\) is the vector from the position on the wrench at which the force is applied to center of the bolt, we define the torque, \(\tau\text{,}\) to be
When a force is applied to an object, Newtonβs Second Law tells us that the force is equal to the rate of change of the objectβs linear momentum. Similarly, the torque applied to an object is equal to the rate of change of the objectβs angular momentum. In other words, torque will cause the bolt to rotate.
In many industrial applications, bolts are required to be tightened using a specified torque. Of course, the magnitude of the torque is \(|\tau| =|\vF\times\vr|=|\vF||\vr||\sin(\theta)\text{.}\) Thus, to produce a larger torque, we can increase either \(|\vF|\) or \(|\vr|\text{,}\) which you may know if you have ever removed lug nuts when changing a flat tire. The ancient Greek mathematician Archimedes said: βGive me a lever long enough and a fulcrum on which to place it, and I shall move the world.β A modern spin on this statement is: βAllow me to make \(|\vr|\) large enough, and I shall produce a torque large enough to move the world.β
The cross product is defined only for vectors in \(\R^3\text{.}\) The cross product of vectors \(\vu = u_1 \vi + u_2 \vj + u_3 \vk\) and \(\vv = v_1 \vi + v_2 \vj + v_3 \vk\) in \(\R^3\) is the vector
where \(\theta\) is the angle between \(\vu\) and \(\vv\) and \(\vn\) is a unit vector perpendicular to both \(\vu\) and \(\vv\) as determined by the right-hand rule.
The magnitude \(|\vu \times \vv|\) of the cross product of the vectors \(\vu\) and \(\vv\) gives the area of the parallelogram determined by \(\vu\) and \(\vv\text{.}\) Also, the scalar triple product \(|(\vu \times
\vv) \cdot \vw|\) gives the volume of the parallelepiped determined by \(\vu\text{,}\)\(\vv\text{,}\) and \(\vw\text{.}\)
You are looking down at a map. A vector \(\bf{u}\) with \(\left| \mathbf{u} \right|\) = 9 points north and a vector \(\mathbf{v}\) with \(\left| \mathbf{v} \right|\) = 9 points northeast. The crossproduct \(\mathbf{u} \times \mathbf{v}\) points:
If \(\mathbf{a} = \mathbf{i} + 10 \mathbf{j} + \mathbf{k}\) and \(\mathbf{b} = \mathbf{i} + 15 \mathbf{j} + \mathbf{k}\text{,}\) find a unit vector with positive first coordinate orthogonal to both \(\mathbf{a}\) and \(\mathbf{b}\text{.}\)
Let \(A = \left(-3,3,-2\right)\text{,}\)\(B = \left(-4,-2,-3\right)\text{,}\) and \(P = (k,k,k)\text{.}\) The vector from \(A\) to \(B\) is perpendicular to the vector from \(A\) to \(P\) when \(k\) = .
A bicycle pedal is pushed straight downwards by a foot with a 36 Newton force. The shaft of the pedal is 20 cm long. If the shaft is \(\pi / 5\) radians past horizontal, what is the magnitude of the torque about the point where the shaft is attached to the bicycle? Nm
Let \(\vu = 2\vi + \vj\) and \(\vv = \vi + 2\vj\) be vectors in \(\R^3\text{.}\)
Without doing any computations, find a unit vector that is orthogonal to both \(\vu\) and \(\vv\text{.}\) What does this tell you about the formula for \(\vu \times \vv\text{?}\)
Using the properties of the cross product and what you know about cross products involving the fundamental vectors \(\vi\) and \(\vj\text{,}\) compute \(\vu \times \vv\text{.}\)
Next, use the determinant version of EquationΒ (9.4.1) to compute \(\vu \times \vv\text{.}\) Write one sentence that compares your results in (a), (b), and (c).
Observe that the area of \(\triangle PQR\) is half of the area of the parallelogram formed by \(\overrightarrow{PQ}\) and \(\overrightarrow{PR}\text{.}\) Hence find the area of \(\triangle PQR\text{.}\)
One of the properties of the cross product is that \((\vu+\vv) \times \vw = (\vu \times \vw) + (\vv \times \vw)\text{.}\) That is, the cross product distributes over vector addition on the right. Here we investigate whether the cross product distributes over vector addition on the left.
Let \(\vu = \langle 1,2,-1 \rangle\text{,}\)\(\vv = \langle 4,-3,6 \rangle\text{,}\) and \(\vv = \langle 4,7,2 \rangle\text{.}\) Calculate
Let \(\vu = \langle u_1, u_2, u_3 \rangle\text{,}\)\(\vv = \langle v_1, v_2, v_3 \rangle\text{,}\) and \(\vw = \langle w_1, w_2, w_3 \rangle\) be vectors in \(\R^3\text{.}\) In this exercise we investigate properties of the triple scalar product \((\vu \times \vv) \cdot \vw\text{.}\)
Show that \(\left|\begin{array}{ccc}
u_1 \amp u_2 \amp u_3 \\
v_1 \amp v_2 \amp v_3 \\
w_1 \amp w_2 \amp w_3
\end{array} \right| = -\left|\begin{array}{ccc}
v_1 \amp v_2 \amp v_3 \\
u_1 \amp u_2 \amp u_3 \\
w_1 \amp w_2 \amp w_3
\end{array} \right|\text{.}\) Conclude that interchanging the first two rows in a \(3 \times 3\) matrix changes the sign of the determinant. In general (although we wonβt show it here), interchanging any two rows in a \(3 \times 3\) matrix changes the sign of the determinant.
Now suppose that \(\vu\text{,}\)\(\vv\text{,}\) and \(\vw\) do not lie in a plane when they eminate from a common initial point.
Given that the parallepiped determined by \(\vu\text{,}\)\(\vv\text{,}\) and \(\vw\) must have positive volume, what can we say about \((\vu \times \vv) \cdot \vw\text{?}\)
Explain how (i.) and (ii.) show that if \(\vu\text{,}\)\(\vv\text{,}\) and \(\vw\) all eminate from the same initial point, then \(\vu\text{,}\)\(\vv\text{,}\) and \(\vw\) lie in the same plane if and only if \((\vu \times \vv) \cdot \vw = 0\text{.}\) This provides a straightforward computational method for determining when three vectors are co-planar.