Simple (?) tensor index notation; When do the indices mean inner product and in what order?

57 Views Asked by At

In index notation, does the term $σ_{ik}x_{j}n_{k}$ mean $\bf{σx}\cdot\bf{n}$ or $\bf{xσ}\cdot\bf{n}$? Here $σ$ is a second-order tensor and $x,n$ are vectors.

On the same note, is $$\frac{\partialσ_{ik}}{\partial x_{k}}x_{j}$$ equivalent to $\nabla(\bf{xσ})$ or $\nabla(\bf{σx})$? For some reason there is an index notation rule that eludes me.

Pardon me for the fundamendality or even stupidity of my questions! Long time lurker, first time poster.

1

There are 1 best solutions below

0
On

In index notation (where summation over a repeated index is implied), matrix-vector multiplication looks like $\left[\mathbf A \mathbf x\right]_i = A_{ij}x_j$, while the dot product looks like $\mathbf x \boldsymbol \cdot \mathbf y = x_iy_i$. Thus, \begin{align*} \sigma_{ik}x_j n_k =\left[\boldsymbol\sigma \mathbf n\right]_ix_j. \end{align*} If, as one of the comments suggests, the current expression has a typo, then the desired result is \begin{align*} \sigma_{ik}x_i n_k &= \left[\boldsymbol \sigma \mathbf n\right]_i x_i\\ &= (\boldsymbol \sigma \mathbf n)\boldsymbol \cdot \mathbf x. \end{align*}