2017年03月工作周报

03月09日 - 03月29日

• 撰写大论文
• 学习机器学习

03月01日 - 03月08日

• 撰写大论文
• 学习机器学习
• CASA被拒，review：

Overall evaluation: 1 (weak accept)
———– Overall evaluation ———–
1) Short description of the paper
The paper builds up on the previous work “Nonlinear Material Design Using Principal Stretches” [Xu et al. 2015] with discussions on stability conditions and how to incorporate them in the intuitive material design by editing the curves of the derivatives of the separable energy density functions (f’, g’, h’ in the paper).
2) Clarity of exposition, and suggestions for improvement. I found one error - in eq (3), first F^hat should be just F, I believe. The exposition of the paper is very clear.
3) Quality of references.
Suggest additional relevant references, if any. The references are adequate.
4) Explanation of your rating.
The discussion on the stability conditions is more thorough compared to [Xu et al. 2015], which will surely lead to intuitive yet stable material editing. The contribution on top of [Xu et al. 2015] is relatively incremental and thin, but the idea is definitely worth sharing. The paper also was not able to solve the problem of intuitive and stable editing of orthotropic materials. Nonetheless, the experimental result showing its difficulty is another point worth sharing. For these reasons, I do recommend accepting this paper, but only weakly.

Overall evaluation: -2 (reject)
———– Overall evaluation ———–
The paper concerns the modeling of hyperelastic materials. The discussed modeling approach offers a richer design space than the tuning the parameters for classical linear or neo-hookean materials.

The paper very closely follows the approach that was introduced in [7]. Though the presentation in this paper is not identical to [7], all important concepts, the separable energy model, the modeling approach for isotropic hyperelastic material and the extension to orthotropic materials, are very similar to the approach introduced in [7]. This leaves me with the question whether or not there are any novel contributions that go beyond what was introduced in [7]. Unfortunately, the paper does not properly address this question, and, when directly comparing the two papers, I do not see any substantial novelty as well. Before the paper could be considered for publication, the proposed concepts need to be put in context with [7] and novel concepts, if there are any, need to be highlighted, discussed and positioned against [7].

Overall evaluation: -2 (reject)
———– Overall evaluation ———–
This paper presents an intuitive method for designing stable material models for nonlinear finite element applications. It is based on the strain energy function parameterized by the principal stretches. The paper first discusses necessary stability conditions to prevent instable simulations. Based on these, in the material modeling step, analytic functions that meet these conditions are used.
The paper is well written and technically sound and relevant related work is listed. However, I had troubles identifying the technical contribution. It seems the paper is very similar to:
Hongyi Xu, Funshing Sin, Yufeng Zhu, and Jernej Barbič. Nonlinear material design using principal stretches. ACM Trans. Graph., 34(4):75:1–75:11, July 2015.
In its current form it is unclear if the paper contains any novelty in comparison to Xu et al. 2015. This paper can only be considered for publication in my opinion if the differences to Xu et al. 2015 are clearly stated.

• 没有直接说出15年论文的问题，没有和15年对比，让评审没有看到文章的改进点。计划改写论文，并且添加正交异性能量函数Hessian正定条件部分，重新投递…