The equation for Hydrostatic Equlibrium in galaxies and galaxy clusters is quoted as: $$\frac{dP}{dr}\times\frac{1}{\rho} = -\frac{GM}{r^2}$$The minus sign on the right hand side indicates that the direction of the force of gravity is towards the center of the object (as it should be). However, the left hand side of the term also resolves to a negative acceleration. Is the physical interpretation that the pressure gradient as well is directed inward? I've read that the force of pressure is directed along a vector from the highest density to the lowest density, but that is with an unbalanced force (such as wind). What is the right way to interpret the vectors in this equation?
1 Answer
The pressure gradient is actually a vector, written $\nabla P$. Its direction is that in which the pressure gradient is steepest. In a spherically symmetric situation, the gradient in pressure is radially inwards and can be written as $(dP/dr) \hat{r}$. i.e. $dP/dr$ written as a scalar, is a negative number. The force per unit area on an object due to a pressure gradient is $-\nabla P$.
In a spherically symmetric "ball of gas", hydrostatic equilibrium can be written $$\frac{dP}{dr}\, \hat{r} = -\rho(r) g(r) \hat{r}\, $$ where $g(r)$ is the magnitude of the gravitational acceleration at radius $r$. The right hand side is negative because the force due to gravity is inwards.
Perhaps it's clearer if you write it as $$\frac{dP}{dr}\, \hat{r} + \rho(r) g(r) \hat{r} = 0 $$
The first term has negative magnitude and the second term has positive magnitude.
-
$\begingroup$ The higher pressure is in the center, the lower pressure in the outer annulus. The force due to a pressure gradient is from the highest pressure to the lowest. Is that the right way to view the forces? $\endgroup$– user32023Commented Nov 2, 2015 at 12:38
-
$\begingroup$ @DonaldRoyAirey Yes - opposite to the pressure gradient. $\endgroup$– ProfRobCommented Nov 2, 2015 at 12:45
-
$\begingroup$ Is this a valid way to rearrange the equation so that each term represents an acceleration:$$-g(r)\hat{r}-\frac{dP}{dr}\times\frac{1}{\rho(r)}\hat{r} = 0$$ $\endgroup$– user32023Commented Nov 2, 2015 at 13:00
-
$\begingroup$ @DonaldRoyAirey Yes I suppose so. $\endgroup$– ProfRobCommented Nov 2, 2015 at 13:48