這個叫做界面剛度，等效于零厚度的內聚力單元， 這個值理論上講無窮大，但是太高結果不收斂出現非物理結果！

不過有很多文獻有相關的報道

It has been verified that the stiffness of the interface influences the accuracy of prediction [13]. High values of stiffness can prevent interpenetration of the crack faces but will lead to numerical problems. An effective method for normalization of the interface stiffness in terms of small thickness t in the resin rich zone was developed by Daudeville et al. [36]. Proposed values for interfacial stiffness include 107 N/mm3 [37], and 108 N/mm3 [38]. Zou et al. [39] proposed a value for the stiffness between 104 and 107 times the value of the strength of the interface per unit length. The zero-thickness CEs used in this study cannot be normalized by the thickness. The same stiffness value of 106 N/mm3 was taken for each fracture mode, which has been verified as accurate for graphite/epoxy laminates in Ref. [40]. However, the interface stiffness of CEs used to simulate the matrix cracking was defined as a value of 107 N/mm3 based on Shi’s study [20].

Cao D, Duan Q, Hu H, Zhong Y, Li S. Computational investigation of both intra-laminar matrix cracking and inter-laminar delamination of curved composite components with cohesive elements. Composite Structures. 2018;192:300-9.

這個是膠材的彈性模量，必須做實驗，如果沒有實驗參數，就自己假設，一個是正向的 另外兩個是切線方向。