I asked a question over on Math.stackexchange a few days ago, but it didn't get much activity. Hopefully this question isn't considered too elementary by the standards of Mathoverflow. Here is what I am wondering, but first some setup.
Let $M$ be a tracial von Neumann algebra acting on a Hilbert space $\mathcal{H}$. A cocycle action of a discrete group $G$ on $M$ is a pair $(\sigma, v)$ with $\sigma : G \to \mathrm{Aut}(N)$ and $v : G \times G \to \mathcal{U}(M)$ satisfying the following three conditions for all $k,l,m \in G$:
(1) $\sigma_{k} \sigma_{l} = \mathrm{ad}(v(k,l))\sigma_{kl}$
(2) $v(k,l)v(kl,m) = \sigma_{k}(v(l,m)) v(k,lm)$
(3) $v(1,l) = v(l,1) = 1$
The twisted crossed product of $M$ by $G$, denoted by $M \rtimes_{(\sigma,v)} G$, is defined as the von Neumann algebra acting on $\ell^2(G,\mathcal{H})$ generated by $\pi_{\sigma}(M)$ and $\lambda_{v}(G)$, where $\pi_{\sigma}$ is the faithful normal representation of $M$ on $\ell^2(G,\mathcal{H})$ defined via
$$(\pi_{\sigma}(x) \xi)(l) = \sigma_{l^{-1}}(x) \xi (l)$$
and, for each $g \in G$, $\lambda_{v}(g)$ is the unitary operator defined by
$$(\lambda_{v}(k)(\xi))(l) = v(l^{-1},k) \xi (k^{-1}l)$$
for all $x \in M$, $\xi \in \ell^2(G,\mathcal{H})$, and $l \in G$.
My question is,
are there necessary and sufficient conditions for when the twisted crossed product $M \rtimes_{(\sigma,v)} G$ is a $II_1$ factor?
