Hello, certainly a sognificant difference is that GT consider both surfaces as rough, while in GW only one is.

 Hello, Giuseppe Pintaude.

I agree with you. The significant difference is that the GT model was developed for the contact between two rough surfaces with the misalignment of asperity in consideration, while the GW model  is suitable for the contact between one smooth surface and one rough surface. However,  the contact between two rough surfaces can be  equivalently treated as the contact between a smooth surface and a rough surface. In this way, it seems that the GW model can be also used for two rough surface.  

Certainly, It can be found that the GT model may be a advanced version of the GW model.  I have read a lot of article. It seems that the GW model ( or other advanced models, ie: considering the elasto-plastic deformation of asperity) is more popular in the study involving the wear, the GT model is used to predict the friction behavior. At some time, the two models are used to study the same research object.  Therefore, I am confused the application scope of these two model. 

Hello Chunxing Gu, your point of view with respect to the application is very interesting.

A brazilian colleague, Eduardo Tomanik, published a lot of papers discussing this issue, he is on RG platform as well, please check his profile here.

Another issue is related to the elasto-plastic behavior, which many others researches have criticized GW and GT models, such as CEB (Chang-Etsion-Bogy) model. In this sense, probably you have read the papers from research group of Bari-Italy, I think they are an important people reviewing the mechanical behavior considerations.

Good luck in this topic, I like it very much.

Cheers, Giuseppe