粘度是液体的一种基本输运和流变学性质，用于量化分子层之间相对运动的阻力，在理解材料行为中起着关键作用。传统方法如Green-Kubo（GK）方法依赖于相关函数的时间积分，由于在玻璃化转变附近相关函数衰减缓慢，计算成本变得非常高昂。近期提出的一种基于非谐晶格动力学（NALD）和瞬时正则模分析的方法为估算粘度提供了一种有前景的替代方案。本研究应用NALD方法计算了Kremer-Grest聚合物系统的粘度，并将其结果与GK方法和非平衡分子动力学模拟的结果进行了比较。我们的研究表明，包括瞬时正则模在内的所有振动模式都对粘度有贡献。这项工作为在各种系统中高效计算粘度提供了框架，特别是在GK方法不再适用的玻璃化转变附近。此外，它还为进一步理解与结构相关的不同振动模式的作用开辟了道路，有助于设计具有可调流变特性的材料。

Viscosity, a fundamental transport and rheological property of liquids, quantifies the resistance to relative motion between molecular layers and plays a critical role in understanding material behavior. Conventional methods, such as the Green-Kubo (GK) approach, rely on time integration of correlation functions, which becomes computationally intensive near the glass transition due to slow correlation decay. A recently proposed method based on non-affine lattice dynamics (NALD) and instantaneous normal mode analysis offers a promising alternative for estimating the viscosity. In this study, we apply the NALD approach to compute the viscosity of the Kremer-Grest polymer system over a range of temperatures and compare these results with those from the GK method and non-equilibrium molecular dynamics simulations. Our findings reveal that all vibration modes, including the instantaneous normal modes, contribute to the viscosity. This work presents an efficient framework for calculating viscosity across diverse systems, including near the glass transition where the GK method is no longer applicable. Also, it opens the avenue to understanding the role of different vibrational modes linked with structure, facilitating the design of materials with tunable rheological properties.