Tutorials › Calculus III › Orienting a Boundary Curve by the Right-Hand Rule

Home  /  Tutorials  /  Stokes’ Theorem

Stokes’ Theorem · Tutorial 926 of 999

Orienting a Boundary Curve by the Right-Hand Rule

Use the chosen surface normal and the boundary’s outward conormal to determine and check the positive direction of traversal.

Advanced 10 min read

What You'll Learn

  • Relate a chosen surface normal to the positive direction along a smooth boundary.
  • Apply the right-hand rule to surfaces oriented upward or downward.
  • Use the outward conormal to find the positive boundary tangent.
  • Check an orientation with cross products and explicit circle parametrizations.
  • Avoid assuming a boundary’s outward direction without knowing the surface’s geometry.

Why Boundary Direction Depends on the Surface Normal

A boundary curve is a geometric set of points, but it can be traversed in either of two directions. When using Stokes’ theorem, that direction is not chosen independently: it must agree with the orientation of the surface. Reversing the surface normal reverses the positive direction around the boundary.

The right-hand rule gives a practical way to make this choice. It also provides a precise geometric test: at a smooth boundary point, compare the chosen normal with the surface’s outward direction across the boundary. Their relationship determines the positive tangent direction.

Key idea: Choose a unit normal \(\mathbf n\) to orient the surface. At a smooth boundary point, let \(\boldsymbol{\nu}\) be the outward conormal: the unit vector tangent to the surface and perpendicular to the boundary, pointing outward. The positive unit boundary tangent is \(\mathbf T=\mathbf n\times\boldsymbol{\nu}\).

The Right-Hand Rule and the Boundary Tangent

Imagine placing your right hand so that your thumb points in the direction of the chosen surface normal. Your curled fingers indicate the positive direction around the boundary. For a horizontal disk with upward normal, this means counterclockwise when viewed from above. With downward normal, the positive direction is clockwise from above.

The cross-product formula makes this rule usable even when the surface is tilted. The vector \(\boldsymbol{\nu}\) points outward across the boundary but remains tangent to the surface. The positive boundary tangent \(\mathbf T\) is perpendicular to both \(\mathbf n\) and \(\boldsymbol{\nu}\), and points in the direction required by the chosen orientation.

Definition: For a smooth oriented surface with smooth boundary, the outward conormal at a boundary point is the unit vector tangent to the surface, perpendicular to the boundary curve, and pointing out of the surface patch. The positive boundary direction is the direction of \(\mathbf T=\mathbf n\times\boldsymbol{\nu}\), where \(\mathbf n\) is the chosen unit normal.

Do not confuse the outward conormal with a normal vector to the surface. Both \(\mathbf n\) and \(\boldsymbol{\nu}\) are perpendicular to the boundary tangent, but \(\mathbf n\) is perpendicular to the surface, whereas \(\boldsymbol{\nu}\) lies in the surface.

Why the Cross-Product Formula Works

Here is the geometric justification. At a smooth boundary point, \(\mathbf n\) and \(\boldsymbol{\nu}\) are perpendicular unit vectors. Their cross product is therefore a unit vector perpendicular to both, hence tangent to the boundary. Its direction is the one compatible with the surface orientation.

More explicitly, the vector triple-product identity gives

$$ \boldsymbol{\nu}\times(\mathbf n\times\boldsymbol{\nu}) = \mathbf n(\boldsymbol{\nu}\cdot\boldsymbol{\nu}) - \boldsymbol{\nu}(\boldsymbol{\nu}\cdot\mathbf n) = \mathbf n(1)-\boldsymbol{\nu}(0) = \mathbf n. $$

Thus the ordered pair consisting of the outward conormal followed by the positive boundary tangent has the chosen surface normal as its cross product: \(\boldsymbol{\nu}\times\mathbf T=\mathbf n\). This is the orientation convention that the right-hand rule encodes. It also shows why reversing \(\mathbf n\) reverses \(\mathbf T\): replacing \(\mathbf n\) with \(-\mathbf n\) changes \(\mathbf n\times\boldsymbol{\nu}\) to its negative.

Worked Example: Upward-Oriented Disk

Worked Example: Upward-Oriented Disk

Consider the disk \(x^2+y^2\le 9\) in the plane \(z=0\), oriented by the upward unit normal \(\mathbf n=(0,0,1)\). Determine the positive direction at the boundary point \(\mathbf p=(3,0,0)\).

The disk is centered at the origin. At \(\mathbf p\), the vector from the center to the boundary point is \((3,0,0)\), so the outward unit conormal is

$$ \boldsymbol{\nu} = \frac{(3,0,0)}{\sqrt{3^2+0^2}} = (1,0,0). $$

The positive boundary tangent is

$$ \mathbf T = \mathbf n\times\boldsymbol{\nu} = (0,0,1)\times(1,0,0) = (0,1,0). $$

So at the rightmost point of the disk, the positive direction is toward increasing \(y\), which is counterclockwise when viewed from above. For a direct check, use the circle \(\mathbf r(t)=(3\cos t,3\sin t,0)\). It starts at \(\mathbf r(0)=(3,0,0)\), and

$$ \mathbf r'(t)=(-3\sin t,3\cos t,0), \qquad \mathbf r'(0)=(0,3,0). $$

This derivative points in the same direction as \(\mathbf T=(0,1,0)\). The path is therefore positively oriented at that point.

Worked Example: Downward-Oriented Disk

Worked Example: Downward-Oriented Disk

Take the same disk \(x^2+y^2\le 4\) in the plane \(z=0\), but now choose the downward unit normal \(\mathbf n=(0,0,-1)\). Find the positive direction at \(\mathbf p=(2,0,0)\).

Since the disk is centered at the origin, the outward conormal at this point is

$$ \boldsymbol{\nu} = \frac{(2,0,0)}{\sqrt{2^2+0^2}} = (1,0,0). $$

Then

$$ \mathbf T = \mathbf n\times\boldsymbol{\nu} = (0,0,-1)\times(1,0,0) = (0,-1,0). $$

Thus the positive direction at the rightmost point is toward decreasing \(y\), which is clockwise when viewed from above. Check this using \(\mathbf r(t)=(2\cos t,-2\sin t,0)\):

$$ \mathbf r(0)=(2,0,0), \qquad \mathbf r'(t)=(-2\sin t,-2\cos t,0), \qquad \mathbf r'(0)=(0,-2,0). $$

The derivative points in the same direction as \(\mathbf T\), confirming that this path agrees with the downward orientation.

Worked Example: A Tilted Disk

Worked Example: A Tilted Disk

Let the disk of radius \(2\), centered at the origin, lie in the plane \(x+z=0\). Choose the unit normal \(\mathbf n=(1,0,1)/\sqrt{2}\). At the boundary point \(\mathbf p=(0,2,0)\), find the positive direction.

The point is in the plane because \(0+0=0\), and its distance from the center is \(\sqrt{0^2+2^2+0^2}=2\), so it is on the disk’s boundary. Because the disk is centered at the origin, the outward conormal is the unit radial direction in the plane:

$$ \boldsymbol{\nu} = \frac{\mathbf p}{\|\mathbf p\|} = \frac{(0,2,0)}{2} = (0,1,0). $$

Now take the cross product:

$$ \mathbf T = \mathbf n\times\boldsymbol{\nu} = \frac{1}{\sqrt{2}}(1,0,1)\times(0,1,0) = \frac{1}{\sqrt{2}}(-1,0,1). $$

For a check, set \(\mathbf a=(0,1,0)\) and \(\mathbf b=(-1,0,1)/\sqrt{2}\). Both are unit vectors in the disk’s plane, they are perpendicular, and \(\mathbf a\times\mathbf b=(1,0,1)/\sqrt{2}=\mathbf n\). The circle \(\mathbf r(t)=2(\mathbf a\cos t+\mathbf b\sin t)\) starts at \(\mathbf r(0)=(0,2,0)\), and

$$ \mathbf r'(0)=2\mathbf b = \sqrt{2}(-1,0,1). $$

This derivative has the same direction as \(\mathbf T\), so the circle traverses the boundary positively there. The calculation demonstrates that the rule works without relying on the surface being horizontal.

A Common Pitfall: Assuming the Outward Direction

The boundary point and the plane alone do not always determine the outward conormal. The location of the disk’s center matters. For a disk centered at the origin in the plane \(x+z=0\), the point \(\mathbf p=(0,2,0)\) has outward conormal \((0,1,0)\). But that conclusion cannot be drawn from the plane and point alone.

For example, let a radius-\(2\) disk in the same plane have center \(\mathbf c=(-\sqrt{2},2,\sqrt{2})\). The center lies in the plane because \(-\sqrt{2}+\sqrt{2}=0\). The same point \(\mathbf p=(0,2,0)\) lies on its boundary, since

$$ \mathbf p-\mathbf c = (\sqrt{2},0,-\sqrt{2}), \qquad \|\mathbf p-\mathbf c\| = \sqrt{2+2} = 2. $$

Its outward conormal there is instead

$$ \boldsymbol{\nu} = \frac{\mathbf p-\mathbf c}{\|\mathbf p-\mathbf c\|} = \left(\frac{1}{\sqrt{2}},0,-\frac{1}{\sqrt{2}}\right). $$

So assuming \(\boldsymbol{\nu}=(0,1,0)\) without specifying the center gives the wrong tangent direction for this disk. The correction is to establish the surface’s outward direction from its geometry first, then apply \(\mathbf T=\mathbf n\times\boldsymbol{\nu}\).

Practical check: Before using the right-hand rule, identify both the chosen surface normal and the outward conormal at the boundary point. The normal alone does not tell you which way is outward across the boundary.

Using the Orientation in a Line Integral

In Stokes’ theorem, the boundary line integral must use the direction induced by the chosen surface normal. If you traverse the same geometric curve in the opposite direction, the line integral changes sign. Therefore, an orientation error can change the sign of the circulation even when the curve and vector field have been identified correctly.

A reliable procedure at a smooth boundary point is:

1
Record the chosen normal.
Use the specified surface orientation; for a graph, distinguish upward from downward orientation.
2
Find the outward conormal.
Determine which direction points out of the surface patch while remaining tangent to the surface and perpendicular to the boundary.
3
Apply the cross product.
Compute \(\mathbf T=\mathbf n\times\boldsymbol{\nu}\) to obtain the positive boundary tangent.
4
Check the direction.
Use the right-hand rule or compare with a local parametrized traversal. Reversing the normal should reverse the tangent.

This check is especially useful when a surface is tilted or when a problem specifies the normal in words rather than giving a picture. A sketch may help, but the cross product provides a definite direction once the normal and outward conormal are known.

Check Your Understanding

Use the chosen surface normal and the outward conormal to answer each question.

  1. For an upward-oriented disk in the \(xy\)-plane, what is the positive direction when viewed from above?
  2. At a boundary point, suppose \(\mathbf n=(0,0,1)\) and \(\boldsymbol{\nu}=(0,1,0)\). Compute the positive unit tangent.
  3. What happens to the positive boundary direction if the surface normal is replaced by its negative?
  4. Why might knowing a boundary point and its plane be insufficient to determine the outward conormal of a disk?
  5. For a downward-oriented disk in the \(xy\)-plane, is the positive traversal clockwise or counterclockwise when viewed from above?

Previous

Next

Back to Tutorials