Influence of the Miniaturisation Effect on the Effective Stiffness of Lattice Structures in Additive Manufacturing

Round 1

Reviewer 1 Report

The present work deals with a multi-scale analysis of metal lattice structures produced by additive manufacturing processes. The study aims to correlate mechanical performances of lattice structures by type of lattice involved and by the different grain orientations and textures, respectively.

The topic is very interesting but the performed experimental analysis should be better detailed because it appears not exhaustive. In general, it is not clear how the authors conducted the experimental tests, how many specimens for each configuration were tested, Etc…

Due to this reason the referee is positive toward this submission, nevertheless, in referee’s opinion there is still room for some improvements, and the following minor revisions are requested:

1. There are some typos to be fixed (g. line 192 dot is missing at the end of the sentence; line 509 wherein should be replaced by where in…)
2. Line 125 and line 143 appear like a commercial slogan for the EIGA and SLM Solutions AG For the scientific purposes of the work it seems unnecessary information. Please delete the first one and replace the second one with a generally description of the SLM process.
3. Line 151 the authors refer to previous studies: they should specify which ones and improve the references.
4. Line 158 the authors should better explain what they mean by “partly recompiled from [35]” in caption of figure 3.
5. More details are desirable for the testing procedure. For example, line 178 on the basis of which considerations a load rate of 01 mm s^-1 was chosen?
6. Line 297 the authors affirm that the beams diameters were determined based on the raw images recorded for reference purposes for DIC prior to loading, but the authors should specify with more details, how many measurements were carried out? In how many and which points. The authors should indicate the uncertainty of provided measures.
7. Line 355 the authors should better explain the results in fig.9. It is a representative curve of the family trend or an average trend? How many tests were carried out for each family?
8. Line 373 The paragraph 4 (Discussion) lacks of quantitative considerations. Authors should improve this by enriching with quantitative data.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The manuscript is well-written and organized.

The subject of the manuscript is interesting and worth of investigation.

According to the reviewer, the novelty of the research should be more highlighted.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Congratulations to the authors. This work covers interesting ground, providing useful information for future work. I have the following comments:

1. Line 52: "10-20%" this difference seems high and it is not clear from the reference where this is extracted. Additionally, peer-reviewed literature (example EOS SS316L: https://www.ijrte.org/wp-content/uploads/papers/v8i4/D5199118419.pdf) would suggest this difference is much lower (with substantial variability only being related to strain at break). All other key data (modulus, yield, UTS) appears to have far less variability. Could the authors justify/expand on how 10-20% was derived? Whilst material properties are of course a concern for AM components, the challenge for performance-based manufacture lies in geometrical accuracy and porosity (as highlighted on line 67); particularly for cellular structures. This can be referenced widely and would form a more defendable argument here.
2. Line 53: "...not acceptable..." reference? Metal AM for application in these industries is widely published.
3. Line 58: Fig 1. has two lines. One is stated as AM, whilst the other is un-categorized. It would be beneficial to state what this secondary line is indicative of. Conventional manufacturing? Injection moulding? 
4. Line 59: would suggest "still not widely implemented". There are examples of the application of metallic cellular structure to loading in the literature and industry.
5. line 165: inclusion of the heat treatment curve here would be beneficial for replication of the study. I.e. duration and type of loading temp (e.g. ramp) and similarly for unloading. 
6. line 209-210: "derived from conventionally manufactured materials" and the data presented in table 4. Reference for these values?
7. Line 286: Fig 8, images are low resolution and quite small. Due to the detail in these images, they would benefit from enlarging and higher resolution to ensure information is not lost in printing/photocopying.
8. Line 363: Please provide descriptive text to support equation 1, in particular for "a" and "t".
9. Line 374: Another issue that has significant potential to affect these discrepancies would be the values used in the analytical calculations. I assume from the statement "conventionally manufactured" on line 209 that the values used in the analytical calculations were not for additively manufactured 316L or Ti6AL4V. It is well known, and can be observed in the literature, that AM materials and conventionally manufactured materials have different material properties (this can be seen in the example paper included in point 1). Could the authors comment on this?
10. line 404: "However, these arguments solely do not explain the high variance in the deviations observe between experimental and analytical results in the present work" is this comment referring only to the difference in Eexp and Eana for 316L - f2ccz in table 7? The agreement between Eexp and Eana is otherwise commendable for prediction by an analytical model. Particularly for one reliant on the assumptions behind beam theory. This may explain why the f2ccz prediction is less accurate (increased connectivity at joints, etc.).
11. Discussion section: This section covers the potential issues/errors due to manufacturing well; however, it does not cover the potential sources of error in the use of the analytical model. For example, as covered in point 9, if there are sources of error in the values (e.g. material data) used in calculations, this will be reflected in the calculated values. Additionally, whilst this section covers limitations well, it does not highlight the surprising agreement between the analytical model and experiments. Whilst the agreement does not align with the message of the paper (i.e. analytical models do not account for manufacturing data and are therefore limited), the absence of commentary around it in the discussion is notable.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf