Density-Based Optimization of the Laser Powder Bed Fusion Process Based on a Modelling Framework

Round 1

Reviewer 1 Report

The article has validated a modelling framework in Laser-based Powder Bed Fusion (L-PBF) to optimize the density-based defects from the fabrication of defect-free parts. However, a similar modelling framework has been proposed by other authors. The terms and the characters of L-PBF have been widely defined and published from related articles. Many of the fundamental theories have been presented in the following literature, so please add these references in this manuscript, compare the difference of this manuscript with the following literature, and strengthen your novelty. This comment is very important. Thanks!

1)     doi.org/10.3390/app112412053

2)     doi.org/10.1016/j.addma.2016.12.001

3)     doi.org/10.1007/s00339-019-3092-9

4)     doi.org/10.1007/s00170-021-08276-9

5)     doi.org/10.1016/j.matdes.2018.06.037

The title of the article “Density-Based Optimization of the Laser Powder Bed Fusion Process Based on a Modelling Framework” promises to deliver a broader objective on multiple factors related to density-based optimization. But the experiment only focuses primarily on scan strategy. 

In general, the article is helpful to verify the scan strategy with different speeds on 316L stainless steel and Ti-6Al-4V metal alloys to prove the geometry characterization and the melt pool characterization of L-PBF modelling framework. From the results, the article has proposed the dense-optimum region for L-PBF processing on 316L stainless steel and Ti-6Al-4V metal alloys.

The following are detailed comments and suggestions:

1.      Figure 1 (line 180 and line 181): The flowchart is very helpful which shows the detailed modelling framework in the characters of: melt pool’s width, depth, and length; the optimum zone surrounded by keyhole, balling and lack of fusion zones; porosity simulations; and process map design.

2.      Figure 1 (line 214, 215, 216, 217): The figure name should be Figure 2. The schematic description is a very good illustration of the melt pool width, depth, and length.

3.      Figure 2 (line 252, 253): The figure name should be Figure 3. The lack of fusion porosity formation mechanism is well presented with unmelted powder layer.

4.      Figure 3 (line 265, 266): The figure name should be Figure 4. The figure well shows the impact of recoil pressure on the pores under laser power.

5.      Figure 4 (line 276, 277): The figure name should be Figure 5. The figure is a little confused which how the processed layer is broken into two pieces.

6.      Table 1 (line 284): This is very good summary of the proposed criteria by giving the inequalities of L, D, W and t in related to type of porosity. The author may a note to show: L, D, W are the length, the depth, and the width of the melt pool, and t is the layer thickness.

7.      Figure 5 (line 299, 300, 301, 302): The figure name should be Figure 6. The figure well shows the hatch spacing, the scanning speed direction, the raster pattern, and the sample dimensions.

8.      Figure 6 (line 371, 372, 373): The figure name should be Figure 7. The results from 316L stainless steel seems aligned with the process map design in Figure 1.

9.      Figure 7 (line 374, 375, 376): The figure name should be Figure 8. The results from Ti-6Al-4V metal alloys are different from 316L stainless steel, but it still seems aligned with the process map design in Figure 1.

10.   Table 2 (line 378): The table shows ok comparison between the results of 316L stainless steel and Ti-6Al-4V metal alloys.

11.   Figure 8 (line 432, 433): The figure name should be Figure 9. There is a typo at line 432 which number 4 is missing. The figure well compares the dense-optimum regions between single laser scan and sample porosity on 316L stainless steel.

12.   Figure 9 (line 434, 435): The figure name should be Figure 10. The figure well compares the dense-optimum regions between single laser scan and sample porosity on Ti-6Al-4V metal alloys.

13.   Table 3 (line 444): The table shows ok comparison between the experimental and the predicted porosity simulations of 316L stainless steel and Ti-6Al-4V metal alloys.

14.   Figure 10 (line 458, 459): The figure name should be Figure 11. The figure well shows the changes of optimum regions based on the changes of hatch spacing on 316L stainless steel.

15.   Figure 11 (line 462, 463): The figure name should be Figure 12. The figure well shows the changes of optimum regions based on the changes of hatch spacing on Ti-6Al-4V metal alloys.

16.   In summary, the modelling framework in scan strategy of 316L stainless steel and Ti-6Al-4V metal alloys have been validated.

17.   316L stainless steel is the least conservative criterion while Ti-6Al-4V metal alloys is the most conservative criterion.

18.   The experiment is still limited by the restriction of boundary side and the laser power levels where the balling does not appear.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

There are a number of grammatical errors in the work that should be addressed.  A comprehensive review should be completed to address these issues.  In a couple of cases these issues potentially impact the meaning of the statements:

Line 28.  The statement reads "expect from" and should read "except for"

Line 64:  the "former" defects are references, which could indicate either the systematic or stochastic defects since both are mentioned previously.  please clarify.  Also, does it make sense to say that defects from either of these categories can fall into both categories?  The last sentence in this paragraph seems to be saying that, and such a statement would essentially nullify the use of the categories.

Table 1:  It is unclear for each criteria presented if the the range identified is where defects occur or are prevented (in fact it seems like there is a mixture).  Lack of fusion defects obviously occur when the melt pool depth is smaller than powder thickness, and this is what is presented.  However, balling occurs when the melt pool track becomes too long, which appears to be the opposite of what is presented.  Line 400 also appears to state that the criteria is for balling prevention.  Please unify the approach used to identify only criteria for defect formation OR prevention.  It appears that all of the models were categorized correctly despite this mixture of rules.

Line 407-408: the same criteria are listed in parenthesis, one of them should likely be D<1.5t.

Discussion and Conclusions.  Overall, the discussion of the limitations of the approach are sound.  However, additional attention should be paid to the fact that different materials will have different properties, and thus it makes sense that they have numerically different criteria for each kind of flaw.  please address the need for future work to incorporate this effect.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

The manuscript titled “Density-Based Optimization of the Laser Powder Bed Fusion 2 Process Based on a Modelling Framework” has been reviewed.

The introduction is too long, please summarize it.

Please add more data about single-track simulations. Add some figures about their results and dimensioning process.

Double-check figure 4.

Add some information about the optimization procedure.

The following papers are suggested for introduction sections:

The effect of absorption ratio on meltpool features in laser-based powder bed fusion of IN718

Microstructure simulation and experimental evaluation of the anisotropy of 316 L stainless steel manufactured by laser powder bed fusion

Residual stresses in additively manufactured parts: predictive simulation and experimental verification

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 4 Report

The authors present an interesting work on the prediction of process window for LPBF based on different simulation steps from single tracks to bulk.

The English language is very good. The paper is well-structured and the topic is of interest for a broad readership in the field of Additive Manufacturing of metallic materials.

My comments are as follows:

11.  The introduction is very detailed and can serve as reference for everybody, how wants well structured and filtered information about different defects in LPBF and their formation. Well done!

22.  In my opinion, key hole pores are less irregularly shaped, but large in size. This is also supported by novel defect classification algorithms based on size and shape of individual defects; see e.g. 10.3390/met11121912

33. I can not really follow the last sentence in sec. 2.1.1, regarding the susceptibility of 316 stainless steel for solidification cracks and porosity.

44. Please correct the numbering of the figures.

55. In my optinion, the authors overestimate the importance of a single melting track, this here chemical issues (material of build plate), heat management and powder layer thickness are quite different from the inside of a LPBF-inside.

66. Page 7, line 254: key hole porosity is known from laser welding, but it originates not from it.

77. The authors use different criteria to classify the susceptibility of different defects based on literature date. It would be helpful, if they add the materials, for which this criteria were determined. This also helps later on in the discussion of the results.

88. The authors compare their results with an experimental work (ref. [50]). It is not clear, how this reference determined the different defects (metallographic cross sections?) and what is the criterion for almost dense?

99. Fig 7: key hole porosity is displayed by a filled circle, but by an open circle in the legend.

110.   The occurrence of key hole porosity in experiments at 500 mm/s and 200W is explained by stochastics. This statement is very weak and difficult to follow, since it is not clear, if just one pore is present for this parameter set (supports stochastic) or if a remarkable number of key hole pores have been formed. Please check ref. 50 again

111. The used literature data (ref. 50) may not provide enough data to support conclusions regarding the boundary for key hole formation in 316L.

112.   It would be helpful for the reader, if the authors add some representative images to show the results for melt pool simulations for both materials (at least one set of parameter for each) to clarify, how the melt pool characteristics have been determined.

113. I do not understand the differences of Fig. 7/8 and Fig. 9/10. Sec. 3.2 is not clear.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Thanks very much!

Reviewer 4 Report

The previous comments are adapted by the authors. The manuscript is suited for publication