6.1 Matrix Algebra, Rank & Systems of Linear Equations

Why Linear Algebra Sits First in the CIL Mechanical Syllabus

The official Syllabus of Mechanical Discipline for Advt. No. 03/2026 lists Engineering Mathematics as Section 1, before Applied Mechanics and Design. That ordering is not decorative. Every statically indeterminate truss, every pipe-network balance in a mine dewatering circuit, and every stiffness-matrix problem in a coal handling plant structure reduces to solving $Ax = b$. Paper-II carries 100 marks and no negative marking, so the mathematics items are among the fastest marks on the paper: they are short, purely mechanical, and require no engineering judgement.

Expect direct, computational questions — find the rank, find the determinant, state the condition for a non-trivial solution — rather than proofs.

Matrix Fundamentals

A matrix $A$ of order $m \times n$ has $m$ rows and $n$ columns. The operations you must be fluent in:

Special matrices worth memorising:

• Symmetric: $A^T = A$. Stiffness and inertia matrices in mechanics are symmetric.
• Skew-symmetric: $A^T = -A$; all diagonal entries are zero, and the determinant of an odd-order skew-symmetric matrix is always $0$.
• Orthogonal: $A^T A = I$, so $A^{-1} = A^T$ and $\det A = \pm 1$. Rotation matrices are orthogonal.
• Idempotent: $A^2 = A$. Nilpotent: $A^k = 0$ for some $k$.

Determinants and Their Properties

For a $3 \times 3$ matrix, expansion along the first row gives

The properties that generate most exam questions:

1. Interchanging two rows changes the sign of the determinant.
2. Multiplying one row by $k$ multiplies the determinant by $k$; hence $\det(kA) = k^n \det A$ for an $n \times n$ matrix. This is a classic trap — candidates write $k \det A$.
3. If two rows are identical or proportional, $\det A = 0$.
4. Adding a multiple of one row to another leaves the determinant unchanged. This is what makes Gauss elimination safe.
5. $\det(AB) = \det A \cdot \det B$ and $\det(A^{-1}) = 1/\det A$.
6. For a triangular matrix, the determinant is simply the product of the diagonal entries.

Rank of a Matrix

The rank $r$ of a matrix is the order of its largest non-vanishing minor, equivalently the number of non-zero rows when the matrix is reduced to row-echelon form. Because elementary row operations preserve rank, the practical method is always the same: reduce and count.

For an $m \times n$ matrix, $r \leq \min(m, n)$. A square matrix of order $n$ with $r = n$ is called full rank or non-singular, and only then does $A^{-1}$ exist.

The Rouche-Capelli Consistency Test

Given $Ax = b$ with $n$ unknowns, form the augmented matrix $[A \mid b]$ and compare ranks:

For the homogeneous system $Ax = 0$ the augmented column is zero, so the ranks always agree and the system is never inconsistent. The only question is whether a non-trivial solution exists:

This single equivalence is the most frequently examined statement in the whole linear-algebra sub-topic, and it is also the mathematical basis of buckling and vibration eigenproblems you will meet later in Paper-II.

Worked Example: Consistency of a Three-Equation System

A CIL workshop blends three scrap streams. The mass balance gives

Row-reduce the augmented matrix. Row 2 is exactly $2 \times$ Row 1, so it contributes nothing and reduces to a zero row. What remains is

Here $\text{rank}(A) = \text{rank}([A\mid b]) = 2$ but $n = 3$. The system is consistent with infinitely many solutions, carrying $n - r = 1$ free parameter. Setting $z = t$ gives $y = 3 - 2t$ and $x = t$. The physical reading is that the second stream is not independent of the first, so the blend is under-determined.

Worked Example: Determinant Scaling

If $A$ is $3 \times 3$ with $\det A = 5$, find $\det(2A^T A^{-1})$.

Work property by property. $\det(2A^T A^{-1}) = 2^3 \det(A^T)\det(A^{-1}) = 8 \times 5 \times \tfrac{1}{5} = 8$. The $2^3$ — not $2$ — is where marks are lost.