Analyse vectorielle - Définition

Source: Wikipédia sous licence CC-BY-SA 3.0.
La liste des auteurs de cet article est disponible ici.

Expressions des opérateurs en différentes coordonnées

Coordonnées cylindriques

\vec{\mathrm{grad}}f=\frac{\partial f}{\partial r}\vec{u_r}+\frac{1}{r}\frac{\partial f}{\partial \theta}\vec{u_\theta}+\frac{\partial f}{\partial z}\vec{u_z}
\mathrm{div}\vec{A}=\frac{1}{r}\frac{\partial}{\partial r}\left(rA_r \right)+\frac{1}{r}\frac{\partial A_\theta}{\partial \theta}+\frac{\partial A_z}{\partial z}
\vec{\mathrm{rot}}\vec{A}=\left(\frac{1}{r}\frac{\partial A_z}{\partial \theta}-\frac{\partial A_\theta}{\partial z}\right)\vec{u_r} + \left(\frac{\partial A_r}{\partial z}-\frac{\partial A_z}{\partial r}\right)\vec{u_\theta} + \frac{1}{r}\left(\frac{\partial}{\partial r}(rA_\theta)-\frac{\partial A_r}{\partial \theta}\right)\vec{u_z}
\Delta f=\frac{1}{r}\frac{\partial}{\partial r}\left(r\frac{\partial f}{\partial r}\right) + \frac{1}{r^2}\frac{\partial^2 f}{\partial \theta^2} + \frac{\partial^2 f}{\partial z^2}

Coordonnées sphériques

\vec{\mathrm{grad}}f  =   \frac{\partial f}{\partial r}\vec{u_r}    + \frac{1}{r}\frac{\partial f}{\partial \theta}\vec{u_\theta}    + \frac{1}{r \sin\theta}\frac{\partial f}{\partial \varphi} \vec{u_\varphi}
\mathrm{div}\vec{A}  =   \frac{1}{r^2}\frac{\partial}{\partial r}(r^2A_r)    + \frac{1}{r\sin\theta}\frac{\partial} {\partial \theta}(\sin\theta A_\theta)    + \frac{1}{r\sin\theta}\frac{\partial A_\varphi}{\partial \varphi}
\vec{\mathrm{rot}}\vec{A}  =   \frac{1}{r\sin\theta}\left(\frac{\partial}{\partial \theta}(\sin\theta A_\varphi)-\frac{\partial A_\theta}{\partial \varphi}\right)\vec{u_r}    + \left(\frac{1}{r\sin\theta}\frac{\partial A_r}{\partial \varphi}-\frac{1}{r}\frac{\partial}{\partial r}(rA_\varphi)\right)\vec{u_\theta}    + \frac{1}{r}\left(\frac{\partial}{\partial r}(rA_\theta)-\frac{\partial A_r}{\partial \theta}\right)\vec{u_\varphi}
\Delta f  =   \frac{1}{r^2}\frac{\partial}{\partial r}\left(r^2\frac{\partial f}{\partial r}\right)    + \frac{1}{r^2 \sin\theta}\frac{\partial}{\partial \theta}\left(\sin \theta\frac{\partial f}{\partial \theta}\right)    + \frac{1}{r^2\sin^2\theta}\frac{\partial^2 f}{\partial \varphi^2}

Quelques formules différentielles

Attention : les formules suivantes sont valables à condition que certaines hypothèses soient vérifiées ! (la fonction scalaire dans la première formule doit être \mathcal{C}^2(\Omega) , où \Omega \subset \mathbb{R} , par exemple. De même, si \vec f désigne la fonction vectorielle concernée dans la seconde formule, il faut vérifier \vec f \in \mathcal{C}^2(\Omega) , \Omega \subset \mathbb{R}^n .)

  • \vec{\mathrm{rot}}(\vec{\mathrm{grad}})=\vec{0}
  • \mathrm{div}(\vec{\mathrm{rot}})=0
  • \vec{\mathrm{rot}}(\vec{\mathrm{rot}})=\vec{\mathrm{grad}}(\mathrm{div})-\vec{\Delta} (appliqué à un vecteur) (rotationnel du rotationnel)
  • \Delta = \mathrm{div}(\vec{\mathrm{grad}}) (appliqué à un scalaire)

Formules dites de Leibniz pour les produits

  • \vec{\mathrm{grad}}(\vec{X_0}\cdot \vec{B} ) = (\vec{X_0} \cdot \vec{\mathrm{grad}})\vec{B} + \vec{X_0} \wedge \vec{\mathrm{rot}}\vec{B} (où \vec{X_0} est un vecteur uniforme) et évidemment :
  • \vec{\mathrm{grad}}(\vec{A}\cdot \vec{B} ) = (\vec{A} \cdot \vec{\mathrm{grad}})\vec{B} + \vec{A} \wedge \vec{\mathrm{rot}}\vec{B} + (\vec{B} \cdot \vec{\mathrm{grad}})\vec{A} + \vec{B} \wedge \vec{\mathrm{rot}}\vec{A}
  • \nabla(\vec{F}\cdot \vec{F}) = 2 (\vec{F}\cdot\nabla)\vec{F} + 2 \vec{F} \wedge (\vec{\mathrm{rot}}\vec{F}) (dite de Bernoulli, en mécanique des fluides)
  • \mathrm{div}(\vec{X_0} \wedge \vec{B})= - \vec{X_0} \cdot \vec{\mathrm{rot}}\vec{B} (où \vec{X_0} est un vecteur uniforme, définition intrinsèque du rotationnel)
  • \mathrm{div}(\vec{A} \wedge \vec{B})= - \vec{A} \cdot \vec{\mathrm{rot}}\vec{B} + \vec{B} \cdot \vec{\mathrm{rot}}\vec{A}
  • (où \vec{X_0} est un vecteur uniforme, par définition de l'application linéaire tangente)
  • \vec{\mathrm{rot}}( \vec{A}\wedge \vec{B}) = \vec{A}\cdot \mathrm{div}\vec{B}     - (\vec{A}\cdot \vec{\mathrm{grad}})\vec{B} - \vec{B}\cdot \mathrm{div}\vec{A}     + (\vec{B}\cdot \vec{\mathrm{grad}})\vec{A}
  • \vec{\mathrm{grad}}(fg) = f \cdot\vec{\mathrm{grad}}(g)  +  g\cdot\vec{\mathrm{grad}}(f) (symétrique en f et g)
  • \mathrm{div}(\rho \cdot \vec{V}) = \rho \cdot \mathrm{div} \vec{V}+  \vec{\mathrm{grad}}(\rho)\cdot  \vec{V}
  • \vec{\mathrm{rot}}(\rho \cdot \vec{V}) = \rho \cdot \vec{\mathrm{rot}} \vec{V}+  \vec{\mathrm{grad}}(\rho)\wedge  \vec{V}
  • \Delta (f\cdot g) = f\cdot \Delta g + 2 \vec{\mathrm{grad}}(f) \cdot \vec{\mathrm{grad}}(g)  +g\cdot \Delta f
  • \mathrm{div} ( f \cdot \vec{\mathrm{grad}}(g) -g \cdot \vec{\mathrm{grad}}(f)) = f \Delta g - g \Delta f

Quelques formules utiles

  • Soient f(M) et g(M) deux champs scalaires,il existe un champ de vecteurs \vec{A}(M) tel que :

\vec{\mathrm{rot}}\vec{A} = \vec{\mathrm{grad}}f \wedge \vec{\mathrm{grad}}\,g

  • Le champ central \vec{OM}=\vec{r} joue un rôle très important en physique. Aussi convient-il de mémoriser ces quelques évidences :


son application linéaire tangente est la matrice identité (cf. la définition !),

donc \mathrm{div}\vec{r}=3 et \vec{\mathrm{rot}}(\vec{X_0}\wedge\vec{r})=2\vec{X_0} (où \vec{X_0} est un vecteur uniforme) et \vec{\mathrm{rot}}(\vec{r})=\vec{0}

  • D'autre part -mg\vec{k}=-\vec{\mathrm{grad}}(mgz)  ; soit \vec{X_0}=\vec{\mathrm{grad}}(\vec{X_0}\cdot\vec{r}) (où \vec{X_0} est un vecteur uniforme). Et aussi:
  • \vec{\mathrm{grad}}f(r)=f'(r)\vec{u} avec \vec{u}=\frac{\vec{r}}{r}

en particulier \vec{\mathrm{grad}}(r^2)=2\vec{r} (évident car d(\vec{r}\cdot\vec{r})=d(r^2) )

  • \Delta f(r)=f''(r)+\frac{2}{r}\cdot f'(r) , sauf en r = 0
  • Le champ newtonien, soit \frac{\vec{r}}{r^3} , est très souvent étudié ,

car c'est le seul champ central à divergence nulle (évident si l'on pense en termes de flux)sauf pour r = 0, où elle vaut 4\pi\cdot\delta(r)  ; ce résultat est le théorème de Gauss pour l'angle solide)

Il en résulte que \Delta(1/r) = - 4 \pi \cdot \delta(r)

  • Donc \Delta(\vec{X_0}/r) = - 4 \pi \cdot \vec{X_0}\cdot \delta(r)

(où \vec{X_0} est un vecteur uniforme)

qui se décompose en :

\vec{\mathrm{grad}}(\mathrm{div})(\vec{X_0}/r) = - 4 \pi \cdot \vec{X_0}\cdot \delta(r)\cdot(1/3) (où \vec{X_0} est un vecteur uniforme), et

\vec{\mathrm{rot}}(\vec{\mathrm{rot}})(\vec{X_0}/r) = + 4 \pi \cdot \vec{X_0}\cdot \delta(r)\cdot(2/3) (où \vec{X_0} est un vecteur uniforme)

ce qui est moins évident (cf. moment magnétique).

  • Les formules précédentes sont dites de calcul différentiel. Il convient de les associer aux formules de calcul intégral : formule de Stokes, théorème d'Ostrogradski, etc.
  • Enfin, il convient de ne pas perdre de vue le caractère axial ou polaire des champ de vecteurs étudiés. Ce ne sont absolument pas les mêmes entités mathématiques !
Page générée en 0.144 seconde(s) - site hébergé chez Contabo
Ce site fait l'objet d'une déclaration à la CNIL sous le numéro de dossier 1037632
A propos - Informations légales
Version anglaise | Version allemande | Version espagnole | Version portugaise