Mechanism to Reduce the Porosity during Argon Arc Welding of Aluminum Alloys by Changing the Arc Angle

Round 1

Reviewer 1 Report

The study is interesting and while not taking into account the complex flow pattern that may occurs in the weld bath give keys to understand the rôle of the tilt angle on the pore release.  

I may suggest only minor correction to increase the overall quality of the paper :  

Figure 3. : Could you please either in the text discuss , nor in the figure give an "error" bar corresponding to the measurements ?  

Also, can you please correct the caption of the figure,  (I think that the old "Influence" is lacking - Influence of arc tilt on... ) and I suggest to write it as : 

Influence of arc tilt on welding pores for number of pores (a), radius of pores (b), volume of pores (c) and superficial area of pores (d)    (take care you are using two times (c) 

Line 173 : What do you mean by "the dendrite formation was severe ???? during crystallization.  ?  Also you should rather replace "crystallization" by "solidification" .  (the same in  line 254 and 259 ). 

Line 274  : 

replace the sentence by  :  The following assumptions are made for the convenience of the analysis : 

Line 356 - 357 : delete them (already in line 353 - 354)

Thanks to you. 

Figure 3 : it would be easier to read if you can increase the font size of the axis  - please. ! 

Author Response

Quesion1:Figure 3. : Could you please either in the text discuss , nor in the figure give an "error" bar corresponding to the measurements ?  

Answer: the text discuss about Figure3 is  in3.3, 3.4 and 3.5, now the contents of 3.4 and 3.5 are merged into 3.3. A note to explained the error problem is added under figure 3 as follow: 

Note: The method used in the research can only identify pores with a diameter of 0.1mm, and pores below 0.1mm are regarded as errors of the system in Figure 3.

Quesion2:Also, can you please correct the caption of the figure,  (I think that the old "Influence" is lacking - Influence of arc tilt on... ) and I suggest to write it as : 

Quesion3:Line 173 : What do you mean by "the dendrite formation was severe ???? during crystallization.  ?  

Quesion4:Also you should rather replace "crystallization" by "solidification" .  (the same in  line 254 and 259 ). 

Quesion5:Line 274  : replace the sentence by  :  The following assumptions are made for the convenience of the analysis : 

Answer: all questions above have revised the text according the reviewer’s comments.

Quesion6:Line 356 - 357 : delete them (already in line 353 - 354)

Answer: content in Line 356 - 357 is the average value, and  content in line 353 - 354 is the sum value, so the total value has changed  the sum value.

Quesion7:Figure 3 : it would be easier to read if you can increase the font size of the axis  - please. ! 

Answer: Figure 3 has been reworked according the reviewer’s comments.

Reviewer 2 Report

The authors report on a study of the porosity in TIG welding of Aluminium and the impact of the torch inclination angle. The main interesting facts are the systematic x-ray tomographic analysis of pores in dependence of the inclination angle including not only the number but also the size of the pores. Although, similar studies can be found in the literature, the systematic results are beneficial. In a second part of the paper, the authors try to give an explanation of the dependencies by considering the movement of the bubbles under different forces. However, this part includes many trivial, well known or redundant information with low scientific merit.

Therefore, I ask for a substantial revision of the paper before a publication considering the following points:

1. The description in subsections 3.4 and 3.5 should be substantially shortened and summarized in subsection 3.3 because obviously pore volume and surface scale with the radius and show a similar behaviour.
2. Please, complete the caption of Fig. 3 like “Impact of arc tilt…”.
3. The authors mention in subsection 4.3 the impact of droplets. But they should explain what droplets are considered and what is their origin? I am afraid that a systematic impact of droplets occurring spontaneous in TIG processes can hardly by expected.
4. What is the plasma force (different from arc force) with a “direction perpendicular to the molten pool height” in row 193? The authors mean a drag force of the gas flow in direction parallel to the work piece surface?
5. The authors should explain why gravity force on a pore (obviously lower density than the surrounding melt) acts downward – I guess that the hydrodynamic lift has to be considered.
6. It remains completely unclear how the theoretical number of pores in Fig. 6 was calculated. The authors cannot quantify the resulting force on a pore without quantitative estimations of the different contributions. Even if the force would be known, the estimation of the number of pores remaining in the weld pool requires the description of the weld pool flow and of the  movement of the bubbles relative to the weld pool flow and in the direction of the surface.
7. The arc force perpendicular to the weld pool surface as considered in subsection 4.5 causes a pressure increase in the weld pool and also in the bubbles, but not a force to the bubble. Instead hydrodynamic lifting may change (see simple hydrodynamic lifting experiments).
8. What is the resistance force mentioned in row 239? There should be no change of the friction force when temperature differences in the weld pool are not considered.
9. The tilting of the torch considered in subsection 4.6 obviously leads to a strong flow above the weld pool surface accompanied with a drag force directed to the weld pool end. The formation of an elliptical weld pool surface and of a flow near the surface driven by the gas flow is well known as long as it cannot be compensated by the Marangoni convection. The resulting stirring of the weld pool seems me a key issue for the bubble formation and escape as described in this section. In comparison with this effect, the other effects, in particular the change of forces on the bubble and their movement relative to the melt flow maid be of low importance. The authors should carefully check if the section 4 can be considerably shortened. Trivial explanations like the elliptic heating zone of the tilted arc (subsection 4.7) should be avoided.
10. The analysis of the weld pool surface as partly presented in subsection 4.8 should be considerably extended to extent the scientific merit of the paper. Infrared camera images like in Table 3 can be used to characterize the weld pool surface. But here, the starting behaviour of the arc process is of less importance, and the focus should be on the established welding process (a couple of seconds after start). A method should be applied to avoid the arc radiation impact on the surface imaging (e.g. a fast switch-off of the arc).
11. The paper would benefit from a study of the flow behaviour of the weldpool to confirm the theoretical explanations. This can be done by high-speed imaging of tracer particles at the weld pool surface.
12. From my experience, untypical phrases are frequently used, and the language has to be checked by a native speaker (e.g. “computed tomography”, “surfacing weld”, “it was only a variable in the experiment”).

Author Response

Question1:The description in subsections 3.4 and 3.5 should be substantially shortened and summarized in subsection 3.3 because obviously pore volume and surface scale with the radius and show a similar behaviour.

Answer: The contents of 3.4 and 3.5 have been compressed and merged into 3.3 based on the reviewer's comments. More than half of the content is deleted.

Question2: Please, complete the caption of Fig. 3 like “Impact of arc tilt…”.

Answer: We hav modified the caption based on the review's suggestion as follow:

Figure 3. Influence of arc tilt on welding pores for number of pores (a) and radius of pores (b) and volume of pores (c) and superficial area of pores (d).  

Question3:The authors mention in subsection 4.3 the impact of droplets. But they should explain what droplets are considered and what is their origin? I am afraid that a systematic impact of droplets occurring spontaneous in TIG processes can hardly by expected.

Answer: MIG is used in this work, so add the "during gas metal arc welding process in this research." in content.

Question4:What is the plasma force (different from arc force) with a “direction perpendicular to the molten pool height” in row 193? The authors mean a drag force of the gas flow in direction parallel to the work piece surface?

Answer: add the explained content into the content as follow 

Plasma shear is due to the arc plasma moving radially outwards on the weld pool surface, causing a drag force, and is therefore a function of the dynamic viscosity and radial velocity of the shielding gas[23].

Question5:The authors should explain why gravity force on a pore (obviously lower density than the surrounding melt) acts downward – I guess that the hydrodynamic lift has to be considered.

Answer: The author just lists all the forces acting on the bubble, regardless of the magnitude of the force. So add the content " for all the forces acting on the bubble."

Question6:It remains completely unclear how the theoretical number of pores in Fig. 6 was calculated. The authors cannot quantify the resulting force on a pore without quantitative estimations of the different contributions. Even if the force would be known, the estimation of the number of pores remaining in the weld pool requires the description of the weld pool flow and of the  movement of the bubbles relative to the weld pool flow and in the direction of the surface.

Answer:the author’s original intention was to find the relationship between the pore change and the arc inclination angle for Fig.6, so add the content "What is the relationship between the trend of weld porosity change and arc inclination angle?"

Question7:The arc force perpendicular to the weld pool surface as considered in subsection 4.5 causes a pressure increase in the weld pool and also in the bubbles, but not a force to the bubble. Instead hydrodynamic lifting may change (see simple hydrodynamic lifting experiments).

Answer:The author believes that the vertical component of the arc force is also one of the resistances preventing bubbles from rising. So the most important factor has been changed the one important factor.

Question8:What is the resistance force mentioned in row 239? There should be no change of the friction force when temperature differences in the weld pool are not considered.

Answer: The resistance force are the forces that prevents bubbles from rising,including arc force, droplet force, gravity and internal friction force. This study mainly discusses the force change caused by the arc inclination angle.

Question9:The tilting of the torch considered in subsection 4.6 obviously leads to a strong flow above the weld pool surface accompanied with a drag force directed to the weld pool end. The formation of an elliptical weld pool surface and of a flow near the surface driven by the gas flow is well known as long as it cannot be compensated by the Marangoni convection. The resulting stirring of the weld pool seems me a key issue for the bubble formation and escape as described in this section. In comparison with this effect, the other effects, in particular the change of forces on the bubble and their movement relative to the melt flow maid be of low importance. The authors should carefully check if the section 4 can be considerably shortened. Trivial explanations like the elliptic heating zone of the tilted arc (subsection 4.7) should be avoided.

Answer:A large section was deleted based on the review's suggestion. Most of the content is still retained, because the author newly proposed the concept of melting arc column, and the retained content is also related to the arc inclination.

Question10:The analysis of the weld pool surface as partly presented in subsection 4.8 should be considerably extended to extent the scientific merit of the paper. Infrared camera images like in Table 3 can be used to characterize the weld pool surface. But here, the starting behaviour of the arc process is of less importance, and the focus should be on the established welding process (a couple of seconds after start). A method should be applied to avoid the arc radiation impact on the surface imaging (e.g. a fast switch-off of the arc).

Answer:The image of the molten pool at 2 second point in table 3 is added based on the review's suggestion, and we explain in more detail the law of molten pool length change with the arc inclination change and time change in text. We found that the length of the molten pool does not change very significantly after 3 second point. We only use light-filtered glass during welding, and we hace no equipment recommended by the reviewer(e.g. a fast switch-off of the arc).

Question11:The paper would benefit from a study of the flow behaviour of the weldpool to confirm the theoretical explanations. This can be done by high-speed imaging of tracer particles at the weld pool surface.

Answer:Sorry that we do not have the equipment recommended by the reviewer. In addition,  the flow of liquid on the surface of the molten pool can be only observed by this equipment, and it is used to guess the force acting onthe bubbles in the molten pool, and cannot be directly used to analyze the forces on the bubbles in the molten metal. 

Question12:From my experience, untypical phrases are frequently used, and the language has to be checked by a native speaker (e.g. “computed tomography”, “surfacing weld”, “it was only a variable in the experiment”).

Answer: Thanks the review's suggestion. I need to continue to improve my english and writing skills. But the paper was revised by AJE, and "computed tomography" and "surfacing weld" are often used in the professional literature i have read.

Round 2

Reviewer 2 Report

I thank the authors for making changes of the paper and for the response to my comments.

However, a large part of the comments has not been answered sufficiently, and the paper is still not on a level appropriate for publication.

My suggestions to improve the scientific merit have been rejected with the reason that specific equipment for these studies is missing. However, if the scientific content cannot be extended by these or other additional studies, then I would propose a rejection of the paper.

An additional point is that the authors should clearly state already in the beginning of the paper that a MIG and not a TIG process is considered. As a consequence, the action of droplet transfer on the weld pool flow behaviour must be considered much more in detail.

In detail, my questions 5 and 6 have not been answered by the response of the authors (e.g. it is still completely unclear, how the theoretical values in Fig. 6 have been calculated). Response to comment 7 is unsatisfactory. The suggestions in comment 10 and 11 to include a more sophisticated study of the weld pool surface temperature and flow behaviour have been rejected without proposal of an alternative study.

Author Response

Thanks to the reviewers for hard work and suggestions for improving the scientific value of the paper. Perhaps my last reply has not clear or complete, and also inadequate consideration. Now I am replying to the following again. Please correct me.

Question 5 :The authors should explain why gravity force on a pore (obviously lower density than the surrounding melt) acts downward – I guess that the hydrodynamic lift has to be considered.

Answer: I add the explaination of gravity force based on reciewer adevice, revicing content is " f_g is the attraction of the earth to the gas in the bubble, and very small."  The hydrodynamic lift is "the liquid buoyancy f_b", so no further explanation.

Question 6: It remains completely unclear how the theoretical number of pores in Fig. 6 was calculated. The authors cannot quantify the resulting force on a pore without quantitative estimations of the different contributions. Even if the force would be known, the estimation of the number of pores remaining in the weld pool requires the description of the weld pool flow and of the  movement of the bubbles relative to the weld pool flow and in the direction of the surface.

Answer: I’m sorry, i did not explain clearly the purpose of calculating for Figure 6. Now we add a paragraph to explain the original meaning of calculating result in Figure 6 as follow.

In fact, the forces acting on the bubble in the molten pool are complex, and are difficult to be tested and observed directly, so it is difficult to obtain the law of bubble change from the quantitative analysis of bubble forces. The influence of arc tilt for the pores change is more difficult to carry out quantitative analysis. In order to be able to analyze the influence of the arc tilt on the bubbles, from the known results, that is, the results of the pores number are observed to look for the law, and then speculate the trend of the force acting on bubble.

After the analysis in Figure 6, the following conclusions are given.

Therefore, it can be inferred that the vertical component of the arc force is one of the main factors affecting the pores.

Question 7: The arc force perpendicular to the weld pool surface as considered in subsection 4.5 causes a pressure increase in the weld pool and also in the bubbles, but not a force to the bubble. Instead hydrodynamic lifting may change (see simple hydrodynamic lifting experiments).

Answer: The reviewers is right that a pressure on the weld pool surface is not to act directly the bubble, but the arc force perpendicular to the weld pool surface causes a convection of the molten metal, the direction of its movement is as shown in Figure 4 (a), convection can cause that the direction of a force acting on the bubble is downward, and prevent the bubbles from rising.

So add the following content in paper:

The arc force perpendicular to the surface of the weld pool,  causes a convection of the molten metal, its direction of the movement is as shown in Figure 4 (a). The convection can cause that the direction of a force acting on the bubble is downward, and prevents the bubbles from rising.

Question10:The analysis of the weld pool surface as partly presented in subsection 4.8 should be considerably extended to extent the scientific merit of the paper. Infrared camera images like in Table 3 can be used to characterize the weld pool surface. But here, the starting behaviour of the arc process is of less importance, and the focus should be on the established welding process (a couple of seconds after start). A method should be applied to avoid the arc radiation impact on the surface imaging (e.g. a fast switch-off of the arc).

Answer:Thanks reviewer’s suggestion. According to the reviewer's suggestion, the welding process with the fast turn-off method was photographed by the far infrared thermometer, and the images in Table 3 were remade. The time for the editor to submit the manuscript is today, but the accuracy of the turn-off needs to be improved.

Question11:The paper would benefit from a study of the flow behaviour of the weldpool to confirm the theoretical explanations. This can be done by high-speed imaging of tracer particles at the weld pool surface.

Answer: The reviewer gave us a good idea, but very sorry that we have not the equipment recommended, we asked 7 universities and research institutions, only there are high-speed photography, no a high-speed imaging of tracer particles. If looking for further, time is too late. According to the reviewer's thinking, i use theoretical analysis instead of experiment, please guide,so add new content is 4.9 Analysis of molten metal flow. In Section 4.9, the influence of the arc force on the flow of liquid metal in the molten pool is carefully analyzed when the arc force is vertical and inclined.  

New Question:An additional point is that the authors should clearly state already in the beginning of the paper that a MIG and not a TIG process is considered. As a consequence, the action of droplet transfer on the weld pool flow behaviour must be considered much more in detail.

Answer: In the first revised draft, I have clarified this research in 4.3 and added the content was “ during gas metal arc welding process in this research”. According to the reviewer’s new opinion, I wrote a new section 4.3. Influnce of the droplet impact in this revision, and the influence of the droplet transition on the bubbles in the molten pool has been carefully analyzed when the arc is vertical and inclined.

The two opinions of the reviewer guided me to think deeply about some problems, which not only provided hints to the scientific value of this paper(of course, this is only my opinion), but also i believe to promoted our future research work. For this reason, I thank very much for the reviewers again.