education/spatial-mechanics/spatial-rotations-transformations/

[giscus[bot]]

education/spatial-mechanics/spatial-rotations-transformations/

https://siliconwit.com/education/spatial-mechanics/spatial-rotations-transformations/

[SamMachariaPhD]

🚀 PRACTICAL INSIGHT

3D rotations are not just math details, they shape how real systems behave.
In three dimensions, rotation order is non-commutative, so changing the sequence of rotations can lead to completely different orientations. Homogeneous transformation matrices make it possible to chain rotations and translations consistently, but they also hide pitfalls such as singularities and gimbal lock when using Euler angles.

Design tip: choose your rotation representation early. Euler angles are intuitive, but rotation matrices and quaternions are often safer for simulation, control, and interpolation.

[kaelokun]

Regarding the solution for the 3-DOF Robot Manipulator problem, I noticed that the composite rotation matrix was calculated using pre-multiplication. Since the problem specifies that the second and third rotations occur about the 'resulting Y-axis' and 'resulting X-axis' (mobile frame), shouldn't we use post-multiplication R_Total = R_z. R_y. R_x ? Pre-multiplication is usually reserved for rotations relative to the fixed base frame.

[SamMachariaPhD]

The problem statement has been corrected to specify "fixed Z-axis", "fixed Y-axis", and "fixed X-axis" to match the pre-multiplication solution. The extrinsic (fixed-frame) convention uses $R_{composite} = R_x \cdot R_y \cdot R_z$, applying rotations right-to-left about the original axes. For intrinsic (body-frame) rotations about successive "resulting" axes, post-multiplication $R_z \cdot R_y \cdot R_x$ would be correct.