Experimental Investigation of Deposition Pattern on the Temperature and Distortion of Direct Energy Deposition-Based Additive Manufactured Part

Round 1

Reviewer 1 Report

This paper focuses on the effect of the deposition pattern (6 different toolpaths as a combination of layers with long or short raster pattern) on the temperature distribution and longitudinal-angular deflection for the DMT (Direct Metal Tooling) based AM process.

It shows the measurements and analysis of the temperature at the centre of the substrate as well as of the distortion as a function of the 6 patterns.

The original part of this work is the one related to the study of the effect of multi-layer pattern of temperature distribution and global distortion. This subject is presented only based on an experimental point of view discussing only a single geometry.

The paper is substantially well structured and written. The author should separate the discussion from the conclusions

Regarding the content, I suggest the authors to address the generalization of their conclusions to other sliced shapes (especially for those without length much greater than the width) and identify the characteristics of the trajectory to master the maximum temperature and global distortion. 

Moreover, if the authors should better explain: “(line 168) […] the same heat input is applied to the same area for the same time […]”. Is the same area related to the beam diameter independently of the position on the substrate or to a fixed region on the substrate (for instance a neighborhood of the TC2)? How this affects the conclusions and how to predict the maximum temperature?

In the light of what is set out above, it is my considered opinion that, the paper requires a major revision in order to be publishable.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

This work investigates the effect of deposition pattern on the temperature and global distortion of Direct Metal Tooling process-DMT. Six combinations of short and long raster are used in this investigation.

The title fits with the contents of the work. The references list is appropriate and include all the recent work.

The document is well structured and the sections are well designed. The mechanism of observed results are appropriate, and the results tabulated and analysed extensively. Figures, tables and diagrams are all well-constructed and easy to read. The quality of English is good.

Authors should replace the ‘Measurement’ with ‘Measuring’ in Lines: 17, 75, 132, 138 (correct is: CMM-Coordinate Measuring Machine).

I propose the Authors to proofread, one more time, the whole paper.

Author Response

First of all, we deeply appreciate the Reviewer`s recommendation. We feel that different eyes lead better results. Their kind reviews led us to a significant improvement in our work. We tried to take all of the comments and reflect them to our revised manuscript. In this document, we present a list of detailed replies and changes that were made in accordance with their comments.

General Comment

This work investigates the effect of deposition pattern on the temperature and global distortion of Direct Metal Tooling process-DMT. Six combinations of short and long raster are used in this investigation. The title fits with the contents of the work. The references list is appropriate and include all the recent work. The document is well structured and the sections are well designed. The mechanism of observed results are appropriate, and the results tabulated and analysed extensively. Figures, tables and diagrams are all well-constructed and easy to read. The quality of English is good.

Response to Reviewer 2 Comments

Point 1: Authors should replace the ‘Measurement’ with ‘Measuring’ in Lines: 17, 75, 132, 138 (correct is: CMM-Coordinate Measuring Machine).

Response 1: Thank you for the comment. Considering the reviewer’s comment, we have corrected the words.

Reviewer 3 Report

The manuscript “Experimental investigation of deposition pattern on the temperature and distortion of direct metal tooling-based additive manufactured part,” by J. Lee and H. Chung reports temperature and coordinate measurements for six parts fabricated by direct metal tooling using combinations of four different laser scanning patterns.  The data have been scientifically collected and should be of general interest to the AM community.  However, less than ideal reporting mars the usefulness of the data, and thus, this reviewer would recommend a conditional acceptance with mandatory changes.

1) The paper is generally well organized, however the numerous English grammar mistakes sometimes make the text hard to interpret.

2) The choice of two different steels for the substrate and the deposit, with two different coefficients of thermal expansion, almost guarantees residual stress and some distortion.  Can the authors justify their choice?

3) The text (line 92) states the deposit is 4.6 mm wide and 60.6 mm long.  Yet, Figure 2 which shows the sample dimensions, indicates 4 mm by 60 mm.  Obviously, the last digit has been truncated, but these should be reconciled as it might lead to some confusion.

4) The text (starting in line 109 and going to 111) states, “Pattern 3 ~ 6 are designed to apply two patterns uniformly in four layer deposition by applying short raster pattern and long raster pattern twice (Table 2).”  This is not true and disagrees with Table 2.  Patterns 1 through 4 apply two raster patterns, while Patterns 5 & 6 apply four raster patterns.  The authors should devote some text to justifying their selection of these particular layering patterns.  In particular, unlike the other patterns, Patterns 3 & 4 require switching from the previous layer’s ending point to the current layer’s staring point.  Assuming the laser raster moves at the same speed as during deposition, this extra movement allows extra cooling time between these layers and increases the total deposition time by ≈4.7% and ≈8.5%, respectively.

5) The dimensions in Figure 4 may suffer from the same issues as Figure 2. Why were TCs 4 & 5 placed symmetrically on either side of the center line?  One would think the temperatures about the center line are symmetric and placing the TCs at two different distances would yield more information instead of essentially replicate data. 

6) Table 3 is essentially redundant and the information it contains is better reported as text. The confusing part is that some of the “failed” measurements are apparently reported in Figure 8.  For example, in Figure 8C showing the results from TC 3, four patterns are reported as expected since this TC failed in two runs.  However, in Figures 8A and 8D, all six patterns are reported when each had one TC failure.  Why are the authors presenting data they think have “failed?”

7) The subplots in Figure 8 are poorly constructed and explained.  If the measured temperatures are approximately 20 to 600˚C, why is the ordinate y-axis offset and from 0 to 1000˚C?  This only serves to overlay the data one on top of the other making distinguishing the curves difficult.  Likewise, the abscissa is presented from 0 to 200 seconds.  Given that at the stated laser travel speed, deposition times are ≈100 seconds, almost half of the graph is devoted to cooling after deposition which gets little consideration in the text.  Expanding the graphs will allow each TC trace to be followed more easily.  Choosing better colors and line styles for each pattern may also help discriminate the curves.

8) Given that most of the paper focuses on the data in Figure 8B ( TC 2) it may warrant a separate figure.  The authors might consider explaining the overall features of these curves.  There is evidence for a general linear heating of the substrate during deposition followed by cooling after deposition ends.  There is a pattern of six small peaks corresponding to the individual tracks of the A & B patterns.  There is a single large peak evident in the C & D patterns as the laser beam approaches and recedes from the TC position.  Note that the TC records when the patterns change from A/B types to C/D types.  (Figure 11 also makes this point nicely.) Furthermore, note the red curves (Patterns 3 & 4) trail in time just a bit after the others.  This should be expected give these patterns take longer to deposit (see comment 4 above).  Also note the amplitude of the six small peaks deceases as the number of layers increase.  This would be expected as there is more material present to modulate the heat input.

9) The Figure 9 caption should make clear that this is data from TC2.  Is the initial temperature really 0˚C as indicated, in contrast to Figure 8B where it appears to be 20˚C?  Also, if the authors choose to expand and new Figure 8B, this Figure 9 may not be necessary.

10) Table 4 is not required if Figure 11 is included.  The point appears to be that Patterns 2, 3, and 5, have higher final temperatures than Patterns 1, 4, and 6.  Is layer 0 really 0˚C?

11) Since the graphs in Figure 12 display coordinates in X & Y, one assumes the “out of plane” measurements are in the Z axis.  This should be made clear in the caption.  Were any distortions in the X or Y directions noted?

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 4 Report

General comment:

The article investigates the effect of deposition pattern on the temperature and distortion of the directed energy deposition technique. The article is well structured and written. It can be further improved by addressing the following concerns.

Specific comments:

1. Suggest using the ASTM standard naming for the technique, it should be classified as the directed energy deposition technique.
2. Direct energy deposition technique, similar to fused deposition technique, selectively deposits materials layer by layer to form the parts. It is known that the deposition pattern such as the raster angle, print orientation would affect the mechanical properties of the resulting parts. There are a few review articles that specifically discuss about the process-mechanical properties relationship. Suggest citing:
   16. Brenken, B., Barocio, E., Favaloro, A., Kunc, V., & Pipes, R. B. (2018). Fused filament fabrication of fiber-reinforced polymers: A review. Additive Manufacturing, 21, 1-16.
   17. Mohamed, Omar A., Syed H. Masood, and Jahar L. Bhowmik. "Optimization of fused deposition modeling process parameters: a review of current research and future prospects." Advances in Manufacturing1 (2015): 42-53.
   18. Goh, G. D., et al. "Process–Structure–Properties in Polymer Additive Manufacturing via Material Extrusion: A Review." Critical Reviews in Solid State and Materials Sciences (2019): 1-21
3. What is the melting point of the SM45C? from what I found, the processing temperature for this material should be in 850-1050 degree Celsius. However, I noticed that the highest temperature measured was around 600 degree Celsius. What could be the reason for the lower measured temperature?
4. Authors used K-type thermocouples with 2.3mm diameter for the experiments. What is the measurable temperature range? What is the response time of the thermocouple to reach 90% of the actual value? Do the author think the response time is sufficient for the fast heating and cooling nature of the AM process?
5. Figure 4, the figure should be self-explanatory. Suggest labelling the dash line where the specimen will be printed or indicate in the caption what the dash line represents.
6. Figure 5, suggest using another line type to differentiate the path of CMM and specimen position. Describe what each line represents.
7. Figure 8, the results are poorly presented. the lines are too close to each other, it is hard to distinguish the lines. Suggest making it clearer.
8. What is the relationship among the deposition pattern, temperature and the distortion? Especially between the temperature and the distortion.
9. For all the figures, suggest using the same font size for all the axis title and label.
10. The author could add a conclusion section discussing how the result of this study can be used to improved the direct energy deposition process in the future.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The paper in the present form can be published

Author Response

Thank you so much for the good comments.

Reviewer 3 Report

Re-Review of Manuscript applsci-943096, “Experimental investigation of deposition pattern on the temperature and distortion of direct metal tooling-based additive manufactured part,” by J. Lee and H. Chung

As stated in the previous review, this paper should be of general interest.  This is primarily due to the thermal/distortion data developed on a generic thins sheet part. However, to be of use (especially to future modelers), the data collection conditions must be thoroughly reported and some analysis of the meaning of the data is helpful since the authors understand the work most likely better than anyone else ever will.

Point 1, concerning thermal expansion: The authors attempt to cover this point by appealing to industry practice.  I think the issue goes deeper to the heart of this experiment, primarily since the substrate is not cutoff and discarded in this experiment. Consider:

Powder:  SM45C steel has a CTE of ≈ 13 mm/m/˚C (0-300˚C)

Substrate: SUS304 steel has a CTE of ≈ 17.8 mm/m/˚C (20-300˚C)

Assume a length of deposited powder material 60 mm long, firmly attached to a baseplate of similar dimension, undergoing a cooling of 300˚C.  The part from the powder material wants to shrink ≈220 mm while the baseplate wants to shrink ≈300 mm.  If the materials were reversed, would the measured distortions have been in the same direction? Probably not.  Thus, it is important to document the materials and the coordinate measuring portions of this experiment.

Figure 5 implies no addition substrate heating was applied.  This should be stated.  Also, if the grips warmed significantly during the course of the experiments, that might alter the test-to-test initial conditions.

The authors have kindly added information on the Z-axis measurement (Figure 12), but the text still remains unclear concerning direction. It is assumed the measurements are taken on the bottom.  The text only says the positions were selected.  Since the reported numbers are (Post-deposition Height) - (Pre-deposition Height), one assumes positive numbers mean the substrate deflected upwards (away from the deposit), and negative numbers means downwards (towards the deposit).  Positive axis identification should be communicated to prevent confusion.

Note: The authors have edited some occurrences of CMM which was Coordinate Measurement Machine to Coordinate Measuring Machine (eg., line 18).  However, not all instances have been changed.  See: lines 144 and 150.

Point 3, on combining patterns:  The authors have attempted to correct their description of the experiment but have overlooked the topic sentence of the paragraph starting on page 4 which is still factually incorrect.  I would suggest, “Six patterns were designed using long raster patterns, short raster patterns, and combinations of both.”

Point 6, on failed experimental data: The authors wish to present all their data in the name of openness and transparency is commendable, but it is this reviewer’s opinion it is still confusing.  I would suggest reporting these as supplemental data in a separate addendum, leaving the important data in the main body of the paper.

Point 9, on analysis of the data: The authors have improved their temperature-time plots which improves the clarity.  However, they still do not discuss any of the cooling curves (post ≈120 seconds).  Furthermore, they seem concerned that they may be faulted for generalizing from their specific experiments.  I would implore the authors to do further analysis, and speaking only for myself, do not see how interpreting the data from this experiment speaks to any other situation.

If we consider TC 2 as a sensor receiving signals from two sources, the substrate plate and the deposit, we can separate the baseline (long term behavior) from the high frequency (rater pattern) behavior much like one does in differential scanning calorimetry (DSC).  (I had to digitize your data for the following analysis.  I apologize for any digitizing errors and assume your data would produce better results.) Fitting a quadratic equation to the histories of Patterns 1 & 2 defines a baseline very well (R^2 > 0.997).  The baseline signals are important when considering the CTE-caused distortion in Point 1 (above).

The plot below shows the raster signals with the quadratic baseline removed.

(Image file attached)

Pattern 1:

1) Each of the six passes is registered in each of the four layers.

2) The patterns of the first two passes are different from the last two passes.

3) The maximum temperature in each subsequent layer decreases except on the fourth pass (which may be a digitization or baseline artifact).

Pattern 2:

1) The maximum temperature in each layer decreases continuously.

Pattern Comparisons:

1) It is clear the maximum temperatures for Pattern 1 & 2 coincide in time during the first layer.

2) In the second layer the maximum temperature for Pattern 1 comes before that of Pattern 2.

3) In the third and fourth layers, the maximum temperatures for Pattern 1 comes after those of Pattern 2.

Similar analyses of the other TCs might yield additional insights.

These phenomena and data all together should be of interest to modelers.  The authors may wish to look at the works of Dr. Sam Lambrakos at the Naval Research Laboratory. At recent paper may be found here: 

DOI: 10.1615/IntJMultCompEng.v7.i1.60

(http://www.dl.begellhouse.com/journals/61fd1b191cf7e96f,031156b85a696ca4,3a7b07cf2f910bc7.html)

Comments for author File: Comments.pdf

Author Response

Thank you again for the detailed review and recommendation.

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 4 Report

the replies are satisfactory. i recommend accepting in its current form.

Author Response

Thank you so much for the good comments.