Thermal Layer Design in Fused Filament Fabrication

Round 1

Reviewer 1 Report

This study is important; however, many important contexts are not investigated and discussed. Therefore, Authors need to revise the study carefully to address the following concerns.

Which ASTM standard were used to prepare Tensile testing samples?

What is is criteria of selecting the parameters shown in table 1, in particular Printing speed?

Environmental effect and temperature is a significant parameter which might effect the layer temperature while determining the effect of printing speed on layer temperature. How did authors protect the experimental setup to get the accurate results?

Figure 7 shows that the printing quality is not acceptable at these temperature. What was the criteria of selecting these parameters when they cannot printing the part in an acceptable shape? 146.4 degee temperature gives better shape, what is the mechanical property at this temperature? And how about 183 degree?

Ra value is not affected significantly as shown in Fig. 8. Why is it important to be included in this manuscript. As per reviewers understanding, and Fig 7 result, Ra significantly affected by layer temperature. So, there is an opposite result here, authors need to explain their results with more facts in this context.

Fig. 9 also shows the non-significant results. Authors need to explain clearly why this parameter study was important even the effect is non-significant?

 

Most of the hyperlinks are broken and shown errors. It is difficult to understand the figures and their discussion in the presence of aforementioned errors.

Author Response

"Please see the attachment."

Author Response File: Author Response.pdf

Reviewer 2 Report

This work shows that layer temperature influences FDM parts' mechanical and geometrical characteristics. Bone samples are used to study the relation between printing speed and temperature. Dimensions, roughness, and UTS are represented with respect to temperature. Additionally, the authors present a redesign of a backrest component for better heat distribution. The study is easy to follow, and it is clear, but there are a few recommendations for the authors:

- The message "Error! Reference source not found" appears instead of Figure labels on Pages: 3 to 7.

- The authors use a one-at-a-time experimental approach focused on printing speed. Is printing speed the most influential parameter in layer temperature? Do other parameters such as the extrusion temperature and the chamber temperature influence it?

- Figures 6 and 8. Despite the authors' claim, it is not clear that UTS and roughness peaks are at 170 ºC; please clarify this point. Also, are the Ra of the three specimens printed for each speed the same? Finally, there is no apparent Ra decrement from 175 ºC: Figure 8 does not substantiate the comment "... then decreases gradually for higher temperatures..."

Author Response

"Please see the attachment."

Author Response File: Author Response.pdf

Reviewer 3 Report

In this paper, the authors have considered the effects of layer temperature on strength, surface roughness, and dimensional accuracy of PA6-CF, and introduced a concept of Thermal Layer Design (TLD). The authors claimed that the layer temperature within a certain range has a positive effect on the mechanical properties. And the authors aim to modify the layer temperature by making the printing speed a parameter. Overall, the presented work gives readers a sense of the effect of the printing speed on the investigation. 

 

And there are several doubts need to be clarified: 

 

1.     For example, the ABS melt was cooled from the extrusion temperature to its glass-transition temperature within 2 s. In order to reduce the increase in the viscosity as well as to increase the bonding time, travel time is supposed to be with order of 0.1 s. How fast the nozzle moves away from the specimen? What is the cooling rate and effect in the current research? And for the layer temperature measurement, how to judge the layer temperature is the real (instant) temperature of the layer? Heat lose effect when measuring? 

2.     For a layer temperature profile, how the authors claim a point temperature? For example, the temperature of the left part is supposed to be lower than the right part for a nozzle moving from left to right. Temperature distribution is a function of location and time. 

3.     Fixing other parameters, the layer temperature profile is supposed to be a result of varying the printing speed. So instead of claiming the effect of the printing speed, why the authors introduce the concept of TLD? 

4.     For Figure 1, what is the corresponding relations of items “1”, “2”, “3” and “E”, “C”, “B”? 

5.     For Figure 2, the reduced cross-section area of 50 mm^2 should be removed in the cross-section view, which is confused with 5x18 mm^2.

6.     Besides presenting the work, the authors should explain why. 

-       For example, for a higher layer temperature, why the width and thickness decrease reversely? 

-       Higher layer temperature may lead to more diffusion of polymer, why more voids between the outer perimeter and rectilinear infill for higher speeds is observed? Please explain. 

7.     For thermal images, without ambient temperature and scale bar, it is difficult to judge the difference. And in addition, why point temperature is applied? What is the accuracy of the point temperature? 

8.     Since the printed part is supposed to cool down very quicky, how the heat loses over 90 s observed? Heat transfer from the building plate or chamber considered? 

9.   For the comparison of original and mass added components, the first layer temperature is distributed with almost no difference, while mass added component has more evenly temperature distribution. It is reasonable for a large mass to maintain less temperature lose. What is the point of the mass added component? It is more or less for the optimization of the component. 

10.  The English should be edited very deeply.

I do not suggest that the work can be published in its current form.

Author Response

"Please see the attachment."

Author Response File: Author Response.pdf

Reviewer 4 Report

This work describes a phenomenological study about the correlation between printing parameters, layer temperature and eventual properties of parts fabricated using an additive manufacturing technology. The work appears to be interesting. Nevertheless there are some major flaws I would like to present authors to increase comprehension of the manuscript as well as consistency improvement between observations and conclusions. Unfortunately, I believe the article need to be rewritten in some parts and some extra experimentation is needed. I cannot recommend the article to be published in the actual form and hope to be of help with my comments to improve the quality and ease a resubmission.

1. Abstract. Please avoid non defined acronyms. The unit for pressure should be MPa instead of Mpa. Please check this error through all the manuscript.

2. Introduction.

2.1. Overall. Since the journal is devoted to a general scientific reader, it would be of interest to extend the description concerning AM and particularly FFF techniques. Enumerate the most used materials in FFF as well as describing PA, CF and some related references in literature would enrich the introduction as well.

2.2. Please consider eliminating the sentence "The key enabling [...] and metrology" (page 1, lines 31-32). I believe the meaning is repeated in the next sentence.

2.3. Please explain better the sentence in page 1, lines 42-43 (Molecular diffusion [...] associated times).

2.4. In page 2, lines 63-64, do the authors mean that layer temperature eventually influence geometrical changes rather than the opposite as it is written?

3. Experimental part

3.1. When authors refer to layer time, this parameter is the same than printing speed? And when they refer to PA6-CF speed the refer to the flow speed?

3.2. In page 2, line 89, is "wielding" a mistake? If authors actually mean "welding", I would prefer bonding temperature among material deposited roads.

3.3. In this section and all throughout the manuscript, please correct the references in text to Figures. It seems there has been a problem with your text processor when uploading your manuscript to the journal for peer review. Maybe converting all your original document to plain text would be of help.

3.4. References 15 and 17 should be ammended. Authors would better reference the webpage where they accesssed to the referenced documents (including the last date they consulted the web). Is PA6-CF a commercial grade or did the authors prepare the composite? Was said "room temperature" controlled?

3.5. Please improve the quality of Figure 2. Increase size fonts or image size to enhance legibility. How did authors design test specimens? Please indicate the used software.

3.6. Could authors please include flow speed in Table 1?

3.7 In page 4, line146 there is a reference to something in the text I could not guess.

4. Results and discussion

I usually prefer to merge both sections into only one where results are discussed at the time they are referred in the text. I understand in this case it would be tedious to do so, however, I believe there are some aspects in the results that are not enough discussed later. I will point out some of them hereafter. 

4.1. Overall. Please increase size fonts or size images of Figures 6, 8, 9, 11, and especially, 10 and 12. The last two are barely legible.

4.2. In the presentation of Figure 6 authors state that the deviation is more extensive for 180 ºC and higher temperatures. Later on they discuss that this is in correlation with the higher amount of pores and voids observed in the cross-sections. Nevertheless, for 150 ºC the deviation in UTS is extensive too. Could authors discuss why? And in case, which explanation might be ascribed to the fact that lesser deviations in UTS are observed precisely in the better mechanical behaviour at 165-175 ºC?

4.3. Figure 8 and Figure 6 show a trend not exactly overlapping as authors state. The surface roughness does not vary significantly below 170 ºC with a maximun at 173 ºC. I believe this point would need further discussion and comparison with the UTS behaviour.

4.4. The discussion section is interesting itself but I believe authors should emphasise the relation between the properties of the material by means of test specimens characterisation and the properties of the component.

4.5. In page 8, lines 243-244 the statement is an assumption. Can authors support this with a reference or further experimental work? As referred previously by the authors a misread of 36 ºC is to high to have enough confidence in thermal measurements.

4.6. In page 8, lines 265-266. As far as I know tensile test experiments must be performed five times for each sample rather than triply. If it is not the case, authors should ammend the data accordingly, especially taking into account the high deviation exhibited by many samples.

4.7. I believe DSC experiments should be carried on to demonstrate that the crystallinity of the specimens is comparable. Otherwise it is another undiscussed uncertainty factor.

4.8. When authors refer discarding a number of samples due to "unsatisfying failures" I believe a key aspect not discussed is missed. The orientation of test specimens plays a crucial role in the subsequent tensile test. When specimen test are oriented in Z-axis, the deposited roads grow perpendicular to the specimen axis. Thus, when specimens undergo tensile load, the breakage may be due to debonding rather than to the actual tensile strength of the material. It would be of help to extend the study to characterise at least X-axis oriented specimens, where the deposited roads grow parallel to the specimen axis and the debonding effect is minimised. Indeed, when printing an end-use part, just like the component described in this work, the result is a combination of three-axis oriented zones all throughout the volume of the part.

5. Conclusions

Please explain if there is any way to correlate the thermal profiles measured in specimens with the printing parameters of end-use parts. How did authors re-engineer the original component in relation to the characterisation performed in test specimens? Is there any relation between the added mass and TLD? In that case, do authors think it would be possible to elaborate an algorithm to optimise end-use parts design, based in the material characterisations and printing parameters upon previous experimentation?

The answer to these questions may help readers to better understand the impact of your work.

Author Response

"Please see the attachment."

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Thank you.

Author Response

Dear referee

Thank you for your valuable feedback and the opportunity to review our manuscript in a minor revision. We have now thoroughly investigated the initial review letter and manuscript and feel that, for the most part, the reviewer’s concerns are addressed. However, in our initial review, we realised that we had replied to the first referee without integrating all the feedback in the revised manuscript. Also, the previous letter to referee 1 lacked important information as similar concerns were already addressed to other referees. Therefore, we have now carefully read and compared the comments with our manuscript and edited our manuscript accordingly. We believe that most of the concerns were already addressed, but important details were missing in some places. To our knowledge, concerns from the first referee are now addressed in this minor revision. In addition, we also re-made figures 8 and 9 with error bars as we believe this will strengthen the response to the referee and improve the quality of our manuscript. Thank you again for the opportunity, and we hope our revised document is satisfactory and befitting the standards of applied sciences.

Best regards

Sindre Wold Eikevåg

Reviewer 3 Report

The authors revised the manuscript very well, and the responds are correct generally. It is recommended for publishing in the current form. 

Author Response

Dear referee,

This is now an additional round between the journal editor and the authors, but for system purposes, we have to add a response to you before we are able to re-submit. Please excuse this additional message.

Best regards

Sindre Wold Eikevåg

Reviewer 4 Report

Authors have rewritten significantly the paper performing an important effort to address the reviewers comments. The explanations given are enough and hence the paper should be considered for publication.

Author Response

Dear referee,

This is now an additional round between the journal editor and the authors, but for system purposes, we have to add a response to you before we are able to re-submit. Please excuse this additional message.

Best regards

Sindre Wold Eikevåg

Author Response File: Author Response.pdf