Take the 2-minute tour ×
MathOverflow is a question and answer site for professional mathematicians. It's 100% free, no registration required.

From the energy functional, we can derive the Euler-Lagrange equation and its corresponding gradient flow equation. My question is, what is the physical unit for ``time'' in the gradient flow equation?

For example, the Oseen-Frank energy for liquid crystal is given by $$\int_{\Omega} \frac{1}{2} K |\nabla u|^2,$$ where $u$ is the angle of the director with the $x$-axis, and $K$ is an elastic constant with unit Newton. The corresponding gradient flow equation is $$\frac{\partial u}{\partial t} = K \triangle u.$$ But then what is the unit of time $t$? It seems that it does not have the unit of time. What am I missing here?

Thanks!

share|improve this question
1  
By the gradient flow equation, $[K] = L^2 T^{-1}$, where length and time are indicated. Meanwhile the Oseen-Frank "energy" (call it $\mathcal{E}$) has unit $[K] L^{-2} = T^{-1}$, not the physical energy unit $E = M L^2 T^{-2}$. This sort of elision of units is fairly typical of mathematical practice. Note that the unit of action is $[\hbar] = E T$, so $[\hbar \mathcal{E}] = E$. –  Steve Huntsman Aug 17 '11 at 15:47
    
That's a good point. Thanks! –  Chong Luo Aug 18 '11 at 18:32

1 Answer 1

up vote 4 down vote accepted

If you wish to interpret the gradient flow equation as an equation of physics (rather than mathematics), then you need to introduce a friction coefficient into your problem, which tells you how rapidly energy is dissipated in order to reach an equilibrium condition. The energy functional itself only tells you what the equilibrium state is, not how fast you will reach it if you start away from equilibrium. So just multiply the left-hand-side of your gradient flow equation by a friction coefficient $\gamma$ (units $Nsm^{-2}$) and integrate. You will then flow towards the equilibrium state having $\Delta u=0$, which minimizes the energy regardless of the value you chose for $\gamma$.

If you wish to learn more, see for example

http://wwwx.cs.unc.edu/~mn/classes/comp775/doc/calculusOfVariations.pdf

share|improve this answer
    
But according to en.wikipedia.org/wiki/Friction#Coefficient_of_friction, the friction coefficient is a dimensionless scalar. Obviously your $\gamma$ and their "coefficient of friction" are different. Is there another name for physics quantity with units $N s m^{-2}$? –  Chong Luo Aug 18 '11 at 20:12
    
the friction coefficient $γ$ can also be called a viscosity (or, more precisely, dynamic viscosity); the unit $Nsm^{-2}$ is called a "Poise" in the hydrodynamical context (after the French physicist Poiseuille). en.wikipedia.org/wiki/Viscosity#Dynamic_viscosity –  Carlo Beenakker Aug 19 '11 at 0:38
    
That helps a lot. Many thanks! –  Chong Luo Aug 30 '11 at 10:07
    
You're welcome. Question: you just "unaccepted" my answer; was that intentional? –  Carlo Beenakker Aug 30 '11 at 13:46
    
Sorry, I clicked by accident. –  Chong Luo Sep 7 '11 at 16:19

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

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