1: \begin{abstract}

2: 	Computing stationary states is an important topic for phase

3: 	field crystal (PFC) models. Great efforts have been made for

4: 	energy dissipation of the numerical schemes when using

5: 	gradient flows. However, it is always time-consuming due to

6: 	the requirement of small effective time steps. In this paper, we

7: 	propose an adaptive accelerated proximal gradient 

8: 	method for finding the stationary states of PFC models. The

9: 	energy dissipation is guaranteed and the convergence property

10: 	is established for the discretized energy functional.

11: 	Moreover, the connections between generalized proximal operator with classical (semi-)implicit and

12: 	explicit schemes for gradient flow are given. Extensive

13: 	numerical experiments, including two three dimensional

14: 	periodic crystals in Landau-Brazovskii (LB) model and a two

15: 	dimensional quasicrystal in Lifshitz-Petrich (LP) model,

16: 	demonstrate that our approach has adaptive time steps which

17: 	lead to significant acceleration over semi-implicit methods

18: 	for computing complex structures. Furthermore, our result

19: 	reveals a deep physical mechanism of the simple LB model via

20: 	which the sigma phase is first discovered.

21: 

22: 	% REQUIRED

23: 	\begin{description}

24: 	\item[Keywords]

25: 	Phase field crystal models, Stationary states, 

26: 	Spectral collocation method,

27: 	Semi-implicit scheme, Adaptive time step

28: 	\item[AMS subject classifications] 35J60, 35Q74, 65N35

29: 	\item[DOI]

30: 	\end{description}

31: \end{abstract}

32: