1: \begin{abstract}

2:     We present an efficient and robust numerical algorithm for solving

3:     the two-dimensional linear elasticity problem that combines the

4:     Quantized Tensor Train format and a domain partitioning strategy.

5:     %%

6:     This approach makes it possible to solve the linear elasticity

7:     problem on a computational domain that is more general than a

8:     square.

9:     %%

10:     %% By integrating Z-ordering and subdomain concatenation, our

11:     Our method substantially decreases memory usage and achieves a

12:     notable reduction in rank compared to established Finite Element

13:     implementations like the FEniCS platform.

14:     %%

15:     %% This efficiency is maintained while still guaranteeing exponential

16:     %% convergence with respect to the number of degrees of freedom.

17:     %%

18:     This performance gain, however, requires a fundamental rethinking

19:     of how core finite element operations are implemented, which

20:     includes changes to mesh discretization, node and degree of

21:     freedom ordering, stiffness matrix and internal nodal force

22:     assembly, and the execution of algebraic matrix-vector operations.

23:     %%

24:     In this work, we discuss all these aspects in detail and assess

25:     the method's performance in the numerical approximation of three

26:     representative test cases.

27:   \end{abstract}

28: