Rotations of Coordinate System
Transcribed Lecture Notes (Complete Derivations & Diagrams)
1. Change of Coordinate System (2D Case)
For simplicity, let us start with a two-dimensional case. We establish a coordinate system with two mutually perpendicular axes, and .
Consider a point with position vector , where is the coordinate origin. When we rotate the reference frame counter-clockwise by an angle , we obtain our new coordinate axes, and .
Fundamental Invariance Principle
The physical position of the vector in physical space does not change; only the reference coordinate system used to measure its coordinates rotates.
Let be the unit basis vectors along the unprimed axes , and let be the unit basis vectors along the rotated primed axes .
Resolving the rotated unit vectors along the original axes gives:
Projection and Derivation of Coordinates
Each coordinate is obtained by projecting the position vector onto the respective basis vector:
For the transformed coordinate :
In the same way, for :
Thus, the 2D coordinate transformation relations are:
2. 3D Rotation and Direction Cosines
We can scale this derivation up to three-dimensional space. The position vector expanded in both unprimed and primed orthonormal coordinate systems is:
Taking the dot product of with :
Direction Cosines Notation
In 3D rotations, instead of a single angle , we define direction cosines between the primed and unprimed unit vectors:
where is the angle between the -th new axis and the -th old axis.
Therefore, the direction cosine coefficients for the first row are:
Similarly, for the second coordinate :
In the exact same way for :
In matrix notation, the full 3D coordinate transformation becomes:
3. General Transformation Matrix & Index Notation
The full direction cosine matrix in terms of unit vector scalar products is:
We can write each component compactly using summation notation:
Einstein Summation Convention
According to Einstein's summation convention, whenever an index is repeated in a single term, a summation over all valid values (1 to 3) is automatically implied:
Because both basis sets are orthonormal (), the rotation matrix is orthogonal, satisfying , which corresponds to and .
4. Multiple Rotations & Euler Angles
We can handle any complex 3D orientation in multiple stages. Before marching to classic Euler angles, let us start with a simpler and widely used formulation: Tait-Bryan angles ( sequence).
- Tait-Bryan (): First rotate about the original axis, then about the intermediate axis, and finally about .
- Classic Euler Angles (): Rotations performed about , then intermediate , and finally about .
First Rotation: About the Axis by Angle
Since the rotation is about , the coordinate and its basis vector remain unchanged:
The generalized transformation matrix becomes:
Even though our transformation occurs in a 3D space, the rotation happens within the 2D plane formed by :
Hence, the transformation matrix for rotation about is:
5. Successive Rotations and Total Matrix
Second Rotation: About Newly Formed by Angle
Evaluating the direction cosines between the newly generated basis vectors and the intermediate primed vectors :
| Dot Product | Angular Separation | Evaluated Value |
|---|---|---|
| Angle | ||
| Angle (Perpendicular) | 0 | |
| Angle | ||
| Rotation axis unchanged | ||
| Angle | ||
| Angle | 0 | |
| Angle |
Therefore, the rotation matrix is:
Third Rotation: About Newly Formed by Angle
In the exact same way, executing a rotation about by angle leaves the first coordinate unchanged:
Total Composite Transformation Matrix
The complete transformation matrix describing the entire sequence of rotations is obtained by multiplying the individual rotation matrices in sequence:
Each rotation matrix acts sequentially on the position vector coordinates from right to left.