Tidal Torque for Inclined and Eccentric Orbits
Starting from eq. 20 of Lai (2012):
\[\begin{split}\mathbf{T}&=&-\int d^3 x \delta\rho(\mathbf{r}, t) \mathbf{r}\times
\nabla U*(\mathbf{r}, t)\\
&=&-\left(\frac{G M'}{a^3 \omega_0}\right)^2
\int d^3 x
\left\{
\sum_{m,m'}\mathcal{U}_{m,m'}
\delta\bar{\rho}_{m,m'}(\mathbf{r})
\exp(-im'\Omega t + i\Delta_{m,m'})
\right\}
\mathbf{r}\times
\left\{
\sum_{\mu,\mu'}\mathcal{U}_{\mu,\mu'}
\exp(i \mu'\Omega t)
\nabla \left[r^2 Y_{2,\mu}^*(\theta, \phi)\right]
\right\}\end{split}\]
Since \(\mathbf{r}\times\nabla r^2=0\):
\[\begin{split}\mathbf{T}&=&-\left(\frac{G M'}{a^3 \omega_0}\right)^2
\int d^3 x
\left\{
\sum_{m,m'}\mathcal{U}_{m,m'}
\delta\bar{\rho}_{m,m'}(\mathbf{r})
\exp(-im'\Omega t + i\Delta_{m,m'})
\right\}
\mathbf{r}\times
\left\{
\sum_{\mu,\mu'}\mathcal{U}_{\mu,\mu'}
\exp(i \mu'\Omega t) r^2
\nabla Y_{2,\mu}^*(\theta, \phi)
\right\}\\
&=& -\left(\frac{G M'}{a^3 \omega_0}\right)^2
\sum_{m,m',\mu,\mu'} \mathcal{U}_{m,m'} \mathcal{U}_{\mu,\mu'}
\exp(i(\mu'-m')\Omega t + i\Delta_{m,m'})
\int d^3 x
\delta\bar{\rho}_{m,m'}(\mathbf{r})
r^2 \mathbf{r}\times
\nabla Y_{2,\mu}^*(\theta, \phi)\end{split}\]
The torque in the z direction
Taking the dot product of the torque with \(\hat{z}=\cos\theta \hat{r} -
\sin\theta \hat{\theta}\), and noting that the \(\hat{r}\) part of
\(\hat{z}\) does not contribute, since at any point it is orthogonal to
\(\mathbf{r}\times \nabla Y_{2,\mu}^*(\theta, \phi)\), we get:
\[\begin{split}T_z&=& -\left(\frac{G M'}{a^3 \omega_0}\right)^2
\sum_{m,m',\mu,\mu'} \mathcal{U}_{m,m'} \mathcal{U}_{\mu,\mu'}
\exp(i(\mu'-m')\Omega t + i\Delta_{m,m'})
\int d^3 x
\delta\bar{\rho}_{m,m'}(\mathbf{r})
r^2 \sin\theta \hat{\theta}\cdot \left[
\mathbf{r}\times
\nabla Y_{2,\mu}^*(\theta, \phi)
\right]\\
&=& \left(\frac{G M'}{a^3 \omega_0}\right)^2
\sum_{m,m',\mu,\mu'} \mathcal{U}_{m,m'} \mathcal{U}_{\mu,\mu'}
\exp(i(\mu'-m')\Omega t + i\Delta_{m,m'})
\int d^3 x
\delta\bar{\rho}_{m,m'}(\mathbf{r})
r^2 \sin \theta (\mathbf{r}\times\hat{\theta})\cdot
\nabla Y_{2,\mu}^*(\theta, \phi)\\
&=& -\left(\frac{G M'}{a^3 \omega_0}\right)^2
\sum_{m,m',\mu,\mu'} \mathcal{U}_{m,m'} \mathcal{U}_{\mu,\mu'}
\exp(i(\mu'-m')\Omega t + i\Delta_{m,m'})
\int d^3 x
\delta\bar{\rho}_{m,m'}(\mathbf{r})
r^3 \sin\theta
\hat{\phi}\cdot\nabla Y_{2,\mu}^*(\theta, \phi)\\
&=& -\left(\frac{G M'}{a^3 \omega_0}\right)^2
\sum_{m,m',\mu,\mu'} \mathcal{U}_{m,m'} \mathcal{U}_{\mu,\mu'}
\exp(i(\mu'-m')\Omega t + i\Delta_{m,m'})
\int d^3 x
\delta\bar{\rho}_{m,m'}(\mathbf{r})
r^2
\frac{\partial Y_{2,\mu}^*(\theta, \phi)}{\partial\phi}\\
&=& -\left(\frac{G M'}{a^3 \omega_0}\right)^2
\sum_{m,m',\mu,\mu'} i\mu\mathcal{U}_{m,m'} \mathcal{U}_{\mu,\mu'}
\exp(i(\mu'-m')\Omega t + i\Delta_{m,m'})
\int d^3 x
\delta\bar{\rho}_{m,m'}(\mathbf{r})
r^2 Y_{2,\mu}^*(\theta, \phi)\end{split}\]
Using the fact that
\(\delta\bar{\rho}_{m,m'}(\mathbf{r})\propto\exp(im\phi)\), we must have
\(m=\mu\), further if we average over over an orbit we must have
\(m'=\mu'\):
\[T_z= -\left(\frac{G M'}{a^3 \omega_0}\right)^2
\sum_{m,m'} i m \mathcal{U}_{m,m'}^2
\exp(i\Delta_{m,m'})
\int d^3 x \delta\bar{\rho}_{m,m'}(\mathbf{r}) r^2
Y_{2,m}^*(\theta, \phi)\]
The real part of which is:
\[\begin{split}T_z&=&\frac{G R^3}{M}\left(\frac{M'}{a^3}\right)^2
\sum_{m,m'} m \mathcal{U}_{m,m'}^2
\sin(\Delta_{m,m'})
\int d^3 x \delta\bar{\rho}_{m,m'}(\mathbf{r}) r^2
Y_{2,m}^*(\theta, \phi)\\
T_z&=&T_0 \sum_{m,m'} \mathcal{U}_{m,m'}^2 m \kappa_{m,m'}
\sin(\Delta_{m,m'})\\
T_0&\equiv& G R^5\left(\frac{M'}{a^3}\right)^2\\
\kappa&\equiv&\frac{1}{MR^2}
\int d^3 x \delta\bar{\rho}_{m,m'}(\mathbf{r}) r^2
Y_{2,m}^*(\theta, \phi)\end{split}\]
The torque in the x direction
Now we dot with \(\hat{x}=\sin\theta\cos\phi\hat{r} +
\cos\theta\cos\phi\hat{\theta} - \sin\phi \hat{\phi}\). Again, the
\(\hat{r}\) term does not contribute:
\[\begin{split}T_x &=& -\left(\frac{G M'}{a^3 \omega_0}\right)^2
\sum_{m,m',\mu,\mu'} \mathcal{U}_{m,m'} \mathcal{U}_{\mu,\mu'}
\exp(i(\mu'-m')\Omega t + i\Delta_{m,m'})
\int d^3 x
\delta\bar{\rho}_{m,m'}(\mathbf{r})
r^2 (\cos\theta\cos\phi\hat{\theta} -
\sin\phi\hat{\phi}) \left[
\mathbf{r}\times
\nabla Y_{2,\mu}^*(\theta, \phi)
\right]\\
&=& -\left(\frac{G M'}{a^3 \omega_0}\right)^2
\sum_{m,m',\mu,\mu'} \mathcal{U}_{m,m'} \mathcal{U}_{\mu,\mu'}
\exp(i(\mu'-m')\Omega t + i\Delta_{m,m'})
\int d^3 x
\delta\bar{\rho}_{m,m'}(\mathbf{r})
r^3 (\cos\theta\cos\phi\hat{\phi} +
\sin\phi\hat{\theta})\cdot
\nabla Y_{2,\mu}^*(\theta, \phi)\\
&=& -\left(\frac{G M'}{a^3 \omega_0}\right)^2
\sum_{m,m',\mu,\mu'} \mathcal{U}_{m,m'} \mathcal{U}_{\mu,\mu'}
\exp(i(\mu'-m')\Omega t + i\Delta_{m,m'})
\int d^3 x
\delta\bar{\rho}_{m,m'}(\mathbf{r})
r^2 \left(
\cot\theta\cos\phi\frac{\partial}{\partial \phi} +
\sin\phi\frac{\partial}{\partial \theta}
\right) Y_{2,\mu}^*(\theta, \phi)\end{split}\]
Averaging over an orbit:
\[\begin{split}T_x &=& \left(\frac{G M'}{a^3 \omega_0}\right)^2
\sum_{m,m',\mu} \mathcal{U}_{m,m'} \mathcal{U}_{\mu,m'}
\exp(i\Delta_{m,m'})
\int d^3 x
\delta\bar{\rho}_{m,m'}(\mathbf{r})
r^2 \left(
i\mu\cot\theta\cos\phi Y_{2,\mu}^*(\theta, \phi) -
\sin\phi\frac{\partial Y_{2,\mu}^*(\theta, \phi)}
{\partial \theta}
\right)\\\end{split}\]
Now we use:
\[\frac{\partial Y_{2,\mu}^*}{\partial \theta}=\mu\cot\theta Y_{2,\mu}^* +
\sqrt{(2-\mu)(3+\mu)}\exp(i\phi) Y_{2,\mu+1}^*\]
To get:
\[\begin{split}T_x &=& \frac{T_0}{MR^2}
\sum_{m,m',\mu} \mathcal{U}_{m,m'} \mathcal{U}_{\mu,m'}
i\exp(i\Delta_{m,m'})
\int d^3 x
\delta\bar{\rho}_{m,m'}(\mathbf{r})
r^2 \left(
\mu\cot\theta\exp(i\phi) Y_{2,\mu}^*(\theta, \phi) +
i\sin\phi\sqrt{(2-\mu)(3+\mu)}\exp(i\phi) Y_{2,\mu+1}^*
\right)\\
&=& \frac{T_0}{MR^2}
\sum_{m,m',\mu} \mathcal{U}_{m,m'} \mathcal{U}_{\mu,m'}
i\exp(i\Delta_{m,m'})
\int d^3 x
\delta\bar{\rho}_{m,m'}(\mathbf{r})
r^2 \left(
\mu\cot\theta\exp(i\phi) Y_{2,\mu}^*(\theta, \phi) +
\sqrt{(2-\mu)(3+\mu)}[\exp(2i\phi)-1] Y_{2,\mu+1}^*
\right)\\\end{split}\]
Since \(\delta\bar{\rho}_{m,m'}(\mathbf{r})\propto\exp(im\phi)\) the real
part of the above expression is:
\[\begin{split}T_x &=& -\frac{T_0}{MR^2}
\sum_{m,m',\mu} \mathcal{U}_{m,m'} \mathcal{U}_{\mu,m'}
\sin(\Delta_{m,m'})
\int d^3 x
\delta\bar{\rho}_{m,m'}(\mathbf{r})
r^2 \left(
\mu\cot\theta\exp(i\phi) Y_{2,\mu}^*(\theta, \phi) +
\frac{\sqrt{(2-\mu)(3+\mu)}}{2}[\exp(2i\phi)-1] Y_{2,\mu+1}^*
\right)\\
&=& T_0
\sum_{m,m'} \mathcal{U}_{m,m'} \sin(\Delta_{m,m'})(
\kappa_{m,m'}^-\mathcal{U}_{m-1,m'}+
\kappa_{m,m'}^+\mathcal{U}_{m+1,m'})\\
\kappa_{m,m'}^-&\equiv& \frac{\sqrt{(3-m)(2+m)}}{2MR^2}
\int d^3 x \delta\bar{\rho}_{m,m'}(\mathbf{r})
r^2 Y_{2,m}^*=\frac{\sqrt{(3-m)(2+m)}}{2}\kappa_{m,m'}\\
\kappa_{m,m'}^+&\equiv&-\frac{1}{MR^2} \int d^3 x
\delta\bar{\rho}_{m,m'}(\mathbf{r}) r^2\left[
(m+1)\cot\theta\exp(i\phi) Y_{2,m+1}^*(\theta,\phi)+
\frac{\sqrt{(1-m)(4+m)}}{2}\exp(2i\phi) Y_{2,m+2}^*
\right]\end{split}\]
We have already expressed \(\kappa_{m,m'}^-\) in terms of
\(\kappa_{m,m'}\), and just as in Lai (2012), we only need
\(\kappa_{m,m'}^+\) for \(m=0, \pm 1, 2\).
\[\begin{split}\kappa_{0,m'}^+&=&-\frac{1}{MR^2} \int d^3 x
\delta\bar{\rho}_{0,m'}(\mathbf{r}) r^2\left[
\cot\theta\exp(i\phi) Y_{2,1}^*(\theta,\phi)+
\exp(2i\phi) Y_{2,2}^*(\theta,\phi)
\right]\\
&=&-\frac{1}{MR^2} \int d^3 x
\delta\bar{\rho}_{0,m'}(\mathbf{r}) r^2\left[
-\frac{1}{2}\sqrt{\frac{15}{2\pi}}\cos^2\theta+
\frac{1}{4}\sqrt{\frac{15}{2\pi}}\sin^2\theta
\right]\\
&=&\frac{1}{MR^2} \int d^3 x
\delta\bar{\rho}_{0,m'}(\mathbf{r}) r^2
\frac{1}{4}\sqrt{\frac{15}{2\pi}}(3\cos^2\theta-1)\\
&=&\frac{\sqrt{3/2}}{MR^2}\int d^3 x
\delta\bar{\rho}_{0,m'}(\mathbf{r}) r^2 Y_{2,0}^*(\theta, \phi)\\
\Rightarrow \kappa_{0,m'}^+&=&\sqrt{3/2}\kappa_{0,m'}\\
\kappa_{-1,m'}^+&=&-\frac{1}{MR^2} \int d^3 x
\delta\bar{\rho}_{-1,m'}(\mathbf{r}) r^2
\frac{\sqrt{6}}{2}\exp(2i\phi) Y_{2,1}^*\\
&=&\frac{1}{MR^2} \int d^3 x
\delta\bar{\rho}_{-1,m'}(\mathbf{r}) r^2
\frac{\sqrt{6}}{2}\exp(2i\phi) Y_{2,-1}^*\\
\Rightarrow\kappa_{-1,m'}^+&=&\sqrt{3/2}\kappa_{-1,m'}\\
\kappa_{1,m'}^+&=&-\frac{1}{MR^2} \int d^3 x
\delta\bar{\rho}_{1,m'}(\mathbf{r}) r^2
2\cot\theta\exp(i\phi) Y_{2,2}^*(\theta,\phi)\\
&=&-\frac{1}{MR^2} \int d^3 x
\delta\bar{\rho}_{1,m'}(\mathbf{r}) r^2
\frac{1}{2}\sqrt{\frac{15}{2\pi}}\sin\theta\cos\theta\exp(-i\phi)\\
&=&\frac{1}{MR^2} \int d^3 x
\delta\bar{\rho}_{1,m'}(\mathbf{r}) r^2 Y_{2,1}^*(\theta,\phi)\\
\Rightarrow\kappa_{1,m'}^+&=&\kappa_{1,m'}\\
\kappa_{-2,m'}^+&=&-\frac{1}{MR^2} \int d^3 x
\delta\bar{\rho}_{-2,m'}(\mathbf{r}) r^2\left[
-\cot\theta\exp(i\phi) Y_{2,-1}^*(\theta,\phi)+
\sqrt{\frac{3}{2}}\exp(2i\phi) Y_{2,0}^*
\right]\\
&=&-\frac{1}{MR^2} \int d^3 x
\delta\bar{\rho}_{-2,m'}(\mathbf{r}) r^2\left[
-\frac{1}{2}\sqrt{\frac{15}{2\pi}}\cos^2\theta\exp(2i\phi)+
\frac{1}{4}\sqrt{\frac{15}{2\pi}}\exp(2i\phi)(3\cos^2\theta-1)
\right]\\
&=&-\frac{1}{MR^2} \int d^3 x
\delta\bar{\rho}_{-2,m'}(\mathbf{r}) r^2
\frac{1}{4}\sqrt{\frac{15}{2\pi}}\exp(2i\phi)\left[
-\cos^2\theta+3\cos^2\theta-1
\right]\\
&=&\frac{1}{MR^2} \int d^3 x
\delta\bar{\rho}_{-2,m'}(\mathbf{r}) r^2 Y_{-2,m}^*(\theta, phi)\\
\Rightarrow \kappa_{-2,m'}^+&=&\kappa_{-2,m'}\\\end{split}\]
The torque in the y direction
Now we dot with \(\hat{y}=\sin\theta\sin\phi\hat{r} +
\cos\theta\sin\phi\hat{\theta} - \cos\phi \hat{\phi}\). Again, the
\(\hat{r}\) term does not contribute:
\[\begin{split}T_y &=& -\left(\frac{G M'}{a^3 \omega_0}\right)^2
\sum_{m,m',\mu,\mu'} \mathcal{U}_{m,m'} \mathcal{U}_{\mu,\mu'}
\exp(i(\mu'-m')\Omega t + i\Delta_{m,m'})
\int d^3 x
\delta\bar{\rho}_{m,m'}(\mathbf{r})
r^2 (\cos\theta\sin\phi\hat{\theta} -
\cos\phi\hat{\phi}) \left[
\mathbf{r}\times
\nabla Y_{2,\mu}^*(\theta, \phi)
\right]\\
&=& -\left(\frac{G M'}{a^3 \omega_0}\right)^2
\sum_{m,m',\mu,\mu'} \mathcal{U}_{m,m'} \mathcal{U}_{\mu,\mu'}
\exp(i(\mu'-m')\Omega t + i\Delta_{m,m'})
\int d^3 x
\delta\bar{\rho}_{m,m'}(\mathbf{r})
r^2 \left(
\cot\theta\sin\phi\frac{\partial}{\partial \phi} +
\cos\phi\frac{\partial}{\partial \theta}
\right) Y_{2,\mu}^*(\theta, \phi)\end{split}\]
Averaging over an orbit:
\[\begin{split}T_y &=& -\left(\frac{G M'}{a^3 \omega_0}\right)^2
\sum_{m,m',\mu} \mathcal{U}_{m,m'} \mathcal{U}_{\mu,m'}
\exp(i\Delta_{m,m'})
\int d^3 x
\delta\bar{\rho}_{m,m'}(\mathbf{r})
r^2 \left(
i\mu\cot\theta\sin\phi Y_{2,\mu}^*(\theta, \phi) +
\cos\phi\frac{\partial Y_{2,\mu}^*(\theta, \phi)}
{\partial \theta}
\right)\\
&=& -\left(\frac{G M'}{a^3 \omega_0}\right)^2
\sum_{m,m',\mu} \mathcal{U}_{m,m'} \mathcal{U}_{\mu,m'}
\exp(i\Delta_{m,m'})
\int d^3 x
\delta\bar{\rho}_{m,m'}(\mathbf{r})
r^2 \left\{
i\mu\cot\theta\sin\phi Y_{2,\mu}^*(\theta, \phi) +
\cos\phi\left[\mu\cot\theta Y_{2,\mu}^* +
\sqrt{(2-\mu)(3+\mu)}\exp(i\phi) Y_{2,\mu+1}^*
\right]
\right\}\\
&=& -\left(\frac{G M'}{a^3 \omega_0}\right)^2
\sum_{m,m',\mu} \mathcal{U}_{m,m'} \mathcal{U}_{\mu,m'}
\exp(i\Delta_{m,m'})
\int d^3 x
\delta\bar{\rho}_{m,m'}(\mathbf{r})
r^2 \left[
\mu\cot\theta\exp(i\phi) Y_{2,\mu}^*(\theta, \phi) +
\cos\phi\sqrt{(2-\mu)(3+\mu)}\exp(i\phi) Y_{2,\mu+1}^*
\right]\\
&=& -\left(\frac{G M'}{a^3 \omega_0}\right)^2
\sum_{m,m',\mu} \mathcal{U}_{m,m'} \mathcal{U}_{\mu,m'}
\exp(i\Delta_{m,m'})
\int d^3 x
\delta\bar{\rho}_{m,m'}(\mathbf{r})
r^2 \left[
\mu\cot\theta\exp(i\phi) Y_{2,\mu}^*(\theta, \phi) +
\frac{\sqrt{(2-\mu)(3+\mu)}}{2}(\exp(2i\phi)+1)
Y_{2,\mu+1}^*
\right]\\
&=& -\frac{T_0}{MR^2}
\sum_{m,m'} \mathcal{U}_{m,m'}
\left(\kappa^-_{m,m'}\mathcal{U}_{m-1,m'}+
\kappa^+_{m,m'}\mathcal{U}_{m+1,m'}\right)
\exp(i\Delta_{m,m'})\end{split}\]
With the same \(\kappa^+_{m,m'}\) and \(\kappa^-_{m,m'}\) as for
\(T_x\). So taking the real part gives:
\[T_y = -T_0
\sum_{m,m'} \mathcal{U}_{m,m'}
\left(\kappa^-_{m,m'}\mathcal{U}_{m-1,m'}+
\kappa^+_{m,m'}\mathcal{U}_{m+1,m'}\right)
\cos(\Delta_{m,m'})\]
So the real part is proportional to \(\cos\Delta_{m,m'}\) and so for small
tidal dissipation it is independent of the dissipation, as expected since
this term is responsible for the precession.