2
$\begingroup$

I was experimenting with the triple scalar product and forces in equilibrium when I came to this result:

Consider 4 forces $ \pmb{F_i}$ for $i=1,2,3,4$. $\pmb{F_i}=F_i\hat{e_i}$ where $\hat{e_i}$ is the unit vector in the direction of the corresponding force and $F_i$ is the magnitude.

$$F_1\hat{e_1}+F_2\hat{e_2}+F_3\hat{e_3}+F_4\hat{e_4}=0$$

Take the scalar product of the system with $e_3 \times e_4$ (the vector product of $e_3,e_4$).

$$F_1\hat{e_1}\cdot (e_3 \times e_4)+F_2\hat{e_2}\cdot (e_3 \times e_4)+F_3\hat{e_3}\cdot (e_3 \times e_4)+F_4\hat{e_4}\cdot (e_3 \times e_4)=0$$

$$\Rightarrow F_1\hat{e_1}\cdot (e_3 \times e_4)+F_2\hat{e_2}\cdot (e_3 \times e_4)+0+0=0$$

$$\Rightarrow \frac{F_1}{\hat{e_2}\cdot (e_3 \times e_4)}=\frac{F_2}{-\hat{e_1}\cdot (e_3 \times e_4)}$$

$$\Rightarrow \frac{F_1}{V_1}=\frac{F_2}{V_2}$$

Where $V_1$ is the volume of the parallelepiped with edges $\hat{e_{2}},\hat{e_{3}},\hat{e_{4}}$.

This can be extended to any other pair of forces.

If one expresses Lami's theorem as

If one has 3 forces in equilibrium acting at a single point in 2D then the magnitude of each force is proportional to the area of the parallelogram made from the unit vectors of the other 2 forces.

Then the similarities between my result and this are pretty clear. This also leads me to conjecture for that for two forces in 1 dimension, the forces are proportional to the magnitude of the direction of the other force.

Firstly, is my result correct and valid? Does this theorem have a name, or are there any resources on it I could read? Can this formulation and result be extended into other (n?) dimensions?

$\endgroup$
3
$\begingroup$

The statement generalizes to all dimensions.

Given a vector space $\mathbb{R}^n$ with the usual Euclidean metric, we represent a $k$-dimensional subplane spanned by the vectors $v_1,\dots,v_k$ with $v_1 \wedge v_2 \wedge \dots \wedge v_k$, where the wedge $\wedge$ is the antisymmetric linear product of the exterior algebra. Also, we have that the volume of an $n$-dimensional "parallelogramm" spanned by $v_1,\dots,v_n$ is just the $\lvert V \rvert$ in

$$ v_1 \wedge v_2 \wedge \dots \wedge v_n = V e_1 \wedge \dots \wedge e_n $$

where $e_i$ is the standard basis. Additionally, we have the Hodge star $\star$, which, in particular, sends a $n$-plane $v_1 \wedge \dots \wedge v_n$ to its volume $\mathrm{vol}(v_1 \wedge \dots \wedge v_n)$.The $n$-dimensional version of Lami's theorem is now:

Given $n+1$ vectors $v_i$ whose sum is zero, $$ \frac{\lvert v_i \rvert}{\mathrm{vol}(\hat{v}_{j_1} \wedge \dots \hat{v}_{j_n} )} = \frac{\lvert v_k\rvert}{\mathrm{vol}(\hat{v}_{l_1} \wedge \dots \wedge \hat{v}_{l_n})}$$

where the $j_{\circ}$ are a permutation of the indices except $i$, and likewise the $l_\circ$ a permutation of the indices except $k$.

Proof: Choose any two indices $i,j$, let them w.l.o.g. be $1,2$, and then apply $\wedge v_3 \wedge v_4 \wedge \dots \wedge v_{n+1}$ to the equation $\sum_i v_i = 0$. Antisymmetry of the wedge implies that a summand becomes zero as soon as one of the $v_i$ appears twice. Only the summands $v_1$ and $v_2$ survive this process, and we get

$$ v_1 \wedge v_3 \wedge \dots \wedge v_n = - v_2 \wedge v_3 \wedge \dots \wedge v_n $$

Now, write $v_i = \lvert v_i \rvert \hat{v}_i$ where $\hat{v_i}$ is a unit vector. Linearity of the wedge implies $v_1 \wedge v_3 \wedge \dots \wedge v_n = \lvert v_1 v_3 \dots v_n \rvert \hat{v}_1 \wedge \hat{v}_3 \wedge \dots \wedge \hat{v}_n$, and thus we arrive at

$$ \lvert v_1 \rvert \hat{v}_1 \wedge \hat{v}_3 \wedge \dots \wedge \hat{v}_n = \lvert v_2 \rvert \hat{v}_2 \wedge \hat{v}_3 \wedge \dots \wedge \hat{v}_n$$

and applying the Hodge star to both sides yields the desired result.

$\endgroup$
  • $\begingroup$ Is this a (well) known result in physics, does it have a name if I wanted to look into it? Also can you think of any applications of the result in 4 dimensions or higher? Thanks. $\endgroup$ – Gridley Quayle Feb 3 '15 at 2:04
  • $\begingroup$ @GridleyQuayle: I don't know of any name for this result, and (physical) applications in higher dimensions are not immediately obvious to me. You could call it "well-known" though, since it follows from linear algebra. It's not terribly interesting from a mathematial viewpoint, though, so it has not been given a name. $\endgroup$ – ACuriousMind Feb 3 '15 at 12:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.