A New and Direct R-Value Measurement Method of Sheet Metal Based on Multi-Camera DIC System
Round 1
Reviewer 1 Report
The described principle for direct r-value measurement is a good idea since if we can measure thickness strain directly not with the usual assumption as general using the volume constancy approach would be a good step ahead in sheet metal forming.
However, the description of the double-sided (front and back side) DIC measuring method is not quite clearly given in the paper to be suitable for full repeatability. You may extend your paper with a more detailed description to be easily repeatable by anyone.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
In the present paper a stereo Digital Image Correlation (DIC) setup consisting of two pairs of cameras is used to measure the displacement field of a 1mm thick DP980 steel sheet uniaxial tensile specimen during a tensile test and calculate the corresponding strain tensor components. Using a Random Sample Consensus (RANSAC) algorithm the Lankford coefficient (r-Value) of the material is calculated. The r-value calculation based on the assumption of isochoric conditions is checked against the full 3D measurement.
The reviewer believes that the topic investigated by the authors is of high interest to the community. However, the analysis and discussion is very limited and lacks scientific depth. It needs to be strengthened significantly before it can be recommended for publication.
General comments:
- The reviewer believes that 3D DIC measurement techniques have significant more merit than solely calculating the r-value. Overall a much deeper analysis (and application) of the very promising measurements would be appreciated.
- Limiting the analysis to a single UT test on a single material is a bit shallow. Have the authors considered assessing other material (classes)?
Detailed comments:
- Using the width to thickness strain ratio (the original r-value) goes back to Lankford (1950). This should be referenced in the paper.
- 122: sheet metal thickness can be on the order of several mm. Stating that it is around 1mm is an unnecessary limitation/assumption.
- The reviewer believes that Figs. 3, 4, 5 and 6 partially show redundant information and should be combined.
- Have the authors thought about working on (and plotting) the strain increment ratio d_eps_w/d_eps_th, i.e. the incremental r-value (see. Fig 12)? The reviewer believes that here the RANSAC approach might lead to promising results. In addition, non-linearities in the r-value evolution could be better observed. After all, the linear relation of the width to thickness strain is only an assumption, too (see Fig. 12). You have the means to actually measure it.
- More details on the experimental setup would be highly appreciated. How do the authors achieve the same focal distance with all 4 zoom lenses? What is the spatial resolution (object size - mm/pixels ratio) achieved? What is the acquisition rate? What are the cameras/sensors used? What is the DIC noise floor?
- 241ff: The conclusion drawn here is superficial. Why should anyone choose the more complex setup with 4 cameras, if they can obtain the same result with 1 camera? A better discussion should be carried out and a clear conclusion drawn.
- 13 a,b could be combined in a single one plot with two colors (perhaps with a primary and secondary axis).
- The reviewer disagrees with the conclusion that “large space requirement” is detrimental (l. 292). How about using lenses with a shorter focal distance?
- The reviewer believes that the test results might have significance for many more materials than composites (l.285). Have the authors ever considered looking at polymers? What about the isochoric assumption in that case?
- Some editorial comments
- 35 “crystal plastic” should read ”crystal plasticity”
- The reviewer recommends proofreading for the use of English. Some minor mistakes still exist.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Interesting manuscript, however, requires improvements. Some minor things, like the use of tensile speed instead of strain rate, indicate the need for reviewing the existing text by the Authors. My major objections are as follows.
No explanation is given to selecting DP980 steel. No information provided about the microstructure/heat treatment.
No critical analysis is offered of physical meaning of the differences in R values measured. What precision is required in determining R values?
Conclusion with regard to non-constant volume in the region of necking needs to be elaborated. Is this finding material specific? Is it related to void formation? If yes - it can be confirmed experimentally.
Finał recommendation - shorten descriptive parts of the manuscript and extend discussion of the results.
Author Response
Please see the attachment
Author Response File: Author Response.pdf
