Highly Molybdenum-Alloyed Materials Hastelloy BC-1 (2.4708) and B3 (2.4600): Diffusion Bonding Experiments and Evaluation of both Mechanical Behavior and Corrosion Resistance in Hot 70% Sulfuric Acid

Round 1

Reviewer 1 Report

In general, the paper is well structured and the language is appropriate (some improvements can be made). The topic is interesting, and seems that the experiments were carried out carefully. Although many minor changes in the figures labelling that should be performed before publishing.

Therefore, the reviewer suggests some revisions before the publishing the paper:

• abstract is to long, some parts are already from the Introduction
• Page 1(line 39) and Page 2(line 61) "within a AiF project"; a small description of the AiF project will be helpful to the reader.
• Table 1 - a average value of the Weight loss % with the deviation could be introduced and calculated to the samples from the same type. 
•  Figure 1 - a white background can be introduce in the graph caption to better reading.
• Figure 2 - dimensions are missing in the figure. mm?
• Figure 3 - the white arrows are too small and not well visible in the figure.
• Figure 3 - please label the figure with (a), (b) and (c) because is easier for the reader to identify.
• Figure 4 - the bond line is still visible. How the interface looks like at higher magnification? 
• Page 13 (line260) "De facto," should be changed to "In fact"
• Table 6 and Table 7 - A graph with the results will be easier to compare the results.
• Conclusions are to long and the Table 8 should be in the Discussion part.

Author Response

Dear Sir/Madam,

 

Thank you very much for your valuable contributions

• The abstract was revised and shortened
• AiF: In the abstract, AiF was cut since it is not appropriate to explain abbreviations in the abstract. In the Introduction (line 61) however, the abbreviation AiF is explained now.
• 1: Grids have been added to make it easier for the reader to find cooling gradients for different temperature ranges. However, it has been removed now.
• Fig 2: Unless otherwise indicated, all dimensions in mechanical engineering are in millimeters.
• 3: Size of arrows were increased, Figures were labeled a-c
• 4: Indeed, the impact of temperature for Hastelloy BC-1 in terms of grain growth across bonding layers is not clear. Whereas for T=1100°C grain growth for the bimodal grains is found at some spots, it is not for T=1150°C. Fort T=1200°C, however, there are some spots where grain growth occurred across the bonding plane (indicated for Fig. 4a and Fig. 4c now). In fact, all micrographs at this section of the paper were taken to compare bonding results in terms of grain size for different temperatures for first bonding experiments performed at samples consisting of ten layers. Since it turned out, the deformation was rather low. In our opinion, if bonding planes are clearly visible, without pronounced grain growth across bonding layers, mechanical properties cannot be satisfactory. Hence, it was no closer look taken to theses bonding planes at all. For the tensile test, for samples possessing a higher aspect ratio, the bearing pressure was increased. For these samples, reasonable mechanical properties and strain to rupture vales with dimples at fracture surface were found.
• Page 13, line 260: changed to “In fact”
• 6 & 7: In fact, we were undecided how to present the results. You are right, from a graph it is obvious at first glance. However, we felt, from the tables it is easier to compare strength- and elongation to rupture values to these given in the data sheets of the material, as well as standard deviation of the five tests for eachcondition. In all tables, min- & max-values are now cut to safe some space.
• Conclusions were shortened and Tab. 8 moved to Discussion.

Reviewer 2 Report

This paper reports the effects of diffusion bonding parameters on mechanical as well as corrosion properties of two high molybdenum alloys, it presents several interesting results but several claims need to be clarified before its publication in Metals Journal and the comments are listed below:

1. The abstract is too long and needs to be shortened.
2. In fig. 3, the authors state several pores are observed but they are not marked in the figure.
3. What are the black dots in figure 4 (right image)? Are they impurities or precipitates? Please clarify this.
4. I would suggest the authors to replace the ‘right’, middle“ and left“ with letters such as a, b, c, d …“ in the figures.
5. In table 6, the authors used average+dev to characterize the yield strength, in this case, min and max are not required.
6. The manuscript needs a proper polishing to improve its readability.
7. The conclusion section is too long.

Author Response

Dear Sir/Madam,

thank you very much for your valuable contributions. The truth is, that writing this paper with contributions from several authors decreased readability. We hope we could fix several issues by your hints.

1. Indeed, another reviewer gave the same hint. The abstract was revised and shortened
2. Fig:3 : Pores were marked
3. Unfortunately, the black spots are from etching. Since the pictures were only taken to assess the grain growth across the joining plane, no care was taken to accomplish identical etching results. Mr. Messerschmidt assured to me, he did all metallographic micrograph in the same way so after etching he went instantly to the microscope taking the pictures.
4. Figures left/middle/right => a/b/c: already done (Indeed, another reviewer gave the same hint)
5. Min/max.-Values cut off to save some space…
6. We did our best. Indeed, several sections were written by different contributors and we felt this is the issue you described….
7. Conclusions were shortened and Tab. 8 moved to Discussion (Indeed, another reviewer gave the same hint)

Reviewer 3 Report

The paper reports and discusses both the solid-state diffusion bonding and corrosion resistance of two nickel-based alloys. Approach to optimize the diffusion bonding condition is quite interesting and enough to capture the attention. The subject is topical and of interest to potential readers of journal Metals. However, it is necessary to clarify the following.

1. General
   1. Terminology you used for corrosion testing is not consistent. Use only one word “sulfuric acid” instead of “sulphuric acid”, “sulfuric acid”, and “H2SO4”.
   2. Space between the numbers and units, except the “%”.
2. Section: Introduction
   1. Table 1. Write alloy name (e.g., Alloy 22 for 2.4602) and UNS No. with the DIN No. you listed.
   2. Table 1. Check the typo in the table. That is, “condition”.
3. Section: Materials and Design of Experiments
   1. Table 2. Write the meaning of Ts. That is, “Melting Temperature Ts”.
   2. Line 88. In Ref. [8], the tensile strength ranges from 862 to 883 MPa.
   3. Line 109. Designation for the standard testing method is “ASTM G28-02 (2015)”.
4. Section: Initial Diffusion Bonding Experiments Made of Ten Sheets
   1. Line 131. Use the upper case for the first letters except proposition.
   2. Line 135. Define the aspect ratio to deliver the information more clearly.
   3. Table 3. Deformation looks too small. Did you keep the surface pressure during the whole diffusion bonding process for 4 hours? Or does surface pressure mean the initial value? I have tried the similar testing for Alloy 617 (UNS N06617) at 1150 ^oC under 15 MPa and found out that the reduction in thickness is -1.91, -4.52, and -4.26% after 1, 3, and 5 hours, respectively (reviewer’s unpublished result).
   4. Line 156 and Fig. 4 (left). Do you have any evidence (e.g., EBSD analysis) for the bimodal grain size distribution? It is well known that discontinuous (or abnormal) grain growth which leads the bimodal grain size distribution is not usually encountered without the presence of secondary phases in the isotropic matrix. I’m so confused about that.
   5. Line 167. The passivation layer (e.g., oxide layer) may be broken if one increases the compressive pressure from 6 to 10 MPa at 1200 oC. However, where would the broken passivation layer go? I think that the passivation layer will still remain at the interface and interrupt the movement of interface. So, it could not be ultimate method for successful diffusion bonding.
5. Section: Corrosion Experiment in Sulfuric Acid at 100 oC
   1. Line 173. Check the typo in the title. That is, “at”.
6. Section: Corrosion Test at Coupons aged in 70% Sulfuric Acid
   1. Line 175. Check the grammatical error. That is, “the specimen was”.
   2. Table 4. Check the typo in the table. That is, “condition”.
   3. Table 5. Check the typo in the table. That is, “condition”.
7. Section: Tensile Specimen Geometry and Test Method
   1. Line 223. Check the grammatical error. That is, “a universal testing machine”.
   2. Line 235. One usually uses “tempering” for heat treatment of ferritic/martensitic steels. The materials you used in this study is nickel-based alloy, so I don’t catch the exact meaning you want here.
   3. Figure 8 and Table 6. Mechanical properties of heat-treated 1200 oC/4 h with contact pressure are 339 MPa of YS, 797 MPa of TS, and 100% of elongation. In ASME Sec. II Part D, mechanical properties of N10675 (Hastelloy B3) are 51 ksi (351 MPa) of YS and 110 ksi (758 MPa) of TS. While the TS meets the minimum requirement, YS does not meet the minimum requirement. It means that the mechanical properties of the diffusion bonding would never exceed the minimum requirement even though the satisfactory grain growth occurs at the interface. As you mentioned in Section 2, the recommended annealing temperature for this alloy is 1066 oC (Line 89). I guess that temperature of 1200 oC and dwell time 4 h you applied in this study is too severe for diffusion bonding. This could be applicable for Hastelloy BC-1.
8. Section: Results of Tensile Tests on Hastelloy B3
   1. Table 6. Specimen information is missed at the upper part of this table.
9. Section: Results of Tensile Tests on Hastelloy BC-1
   1. Line 290. One usually uses “tempering” for heat treatment of ferritic/martensitic steels. The materials you used in this study is nickel-based alloy, so I don’t catch the exact meaning you want here.
10. Section: Conclusions
    1. The conclusion must summarize the whole paper and explain its main purpose. But, unfortunately, I found lots of new information which was not covered in the Body Section. This paper is not acceptable in this current status.
    2. Too long. Make your conclusion more effective.
11. Section: References
    1. Line 380. Check the name of the author in Ref. [3]. That is, “L. Paul”.
    2. Line 401. Check the published year of Ref. [14]. That is, “2017”.
    3. Among 14 references you listed, three comes from the data sheet provided by the vendors and one comes from ASTM. I recommend you to find more papers with regard to corrosion and diffusion bonding of nickel-based alloys/superalloys and Hastelloy family to enhance your paper.
    4. Among 14 references you listed, four references are written in German. Could you list some more articles written in English for the potential readers all around the world?
    5. Please follow the general submission guidelines to be published in our journal.

Others, nice and neat paper.

 

Author Response

Dear Sir/Madam,

we thank you very much for your very carefully reading and valuable contributions! Due to writing this paper with contributions from several authors, indeed the readability had to be improved.

We recognize that this is the best review of all those involved and you have extensive knowledge in diffusion bonding. I hope that we will meet at a conference in the future.

To my excuse I can only say that after several versions and frequent reading you won't notice even simple errors like the different spelling of sulfuric acid. We apologize for this. We hope we could fix several issues by your hints.

 

1. General
   1. Issues concerning varying spellings sulfuric acid and space before units were fixed.
   2. Space issues between values and units were fixed.
2. Introduction
3. UNS-No`s were added to Tab. 1. Actually, in our mine, DIN-No`s are the most appropriate identifyer since danger of confusion is low. Unfortunately, for some alloys no DIN-No exists (maraging steel “Corrax” by “Uddeholm”) or several DIN- as well as UNS-No`s exist (Alloy 800). In general, we try to avoid manufacturer designations as “Alloy 22” because many others, e.g. by VMD metals, exist.
4. Spelling of “condition” in Tab. 1 was fixed, Thank you very much for very carefully reading! We didn`t find such issues anymore.
5. Materials and Design of Experiments
   1. Ts => Melting Range T_M fixed
   2. Range of Rm for Hastelloy B3 was fixed to 862-883 MPa
   3. ASTM G-28 A changed to ASTM G28-02 (2015)

 

4. Initial Diffusion Bonding Experiments Made of Ten Sheets
   1. Line 131 (125): fixed
   2. Line 135 (129): aspect ratio (d/h) added
   3. 3: The bearing pressure was kept constant all the time (4 h). Indeed, we agree that the deformation for T = 1100/1150/1200 °C is too small. The decreasing bearing pressures with increasing bonding temperatures were chosen according to our experience for other alloys. Furthermore, due to the low aspect ratio deformation is small and due to the low overall-heights, measuring error may be larger than for higher parts.
      For diffusion bonding samples for tensile test samples, exhibiting an aspect ratio of about 1.5, the bearing pressure for T=12000 °C was raised from 6 to 10 MPa, what is a raise of 66%. Nevertheless, the deformation as little more than doubles.
      Alloy 617 (2.4663) was designed for high temperature application under mech. Load and therefore exhibits a good creep resistance due to mixed crystal hardening. The content of molybdenum is much lower compared to Hastelloy B3 & BC-1. The bearing pressure of 15 MPa you used at 1150°C is 50% higher than the 10 MPa we used in our experiments
      This emphasizes that bonding parameters cannot be transferred to different alloys, giving comparable deformations.
      Deformation depends on the alloys composition as well as the geometry (in terms of the absolute cross section AND the aspect ratio) of the test samples.
      In our opinion, too little attention is generally paid to deformation in diffusion bonding tests. Most authors only consider the joining surface with regard to residual pores. Internal structures contribute significantly to the deformation or may change the deformation behavior considerably.
   4. Line 156 and Fig. 4 (left): Unfortunately, we have no EBSD available. We could not clarify this.
   5. Line 167: This is a serious issue for diffusion bonding of corrosion resistant alloys at all! Since corrosion resistance is a system property in the interaction of alloy composition and environmental conditions (velocity, flow pattern, chemical composition of the medium), passive layers vary greatly in their resistance depending on the alloy and previous heat treatment. There is some literature that claims that asperities can penetrate passive layers and that the fragments can form spherical residues due to surface tension effects, depending on time and temperature. In our opinion, these processes are very complex, difficult to quantify and not yet sufficiently investigated. Again, different alloys exhibit different behavior.

 

5. Corrosion Experiment in Sulfuric Acid at 100 oC
   1. Line 173 (168): fixed: at => on

 

6. Corrosion Test at Coupons aged in 70% Sulfuric Acid
   1. Line 175. Check the grammatical error. That is, “the specimen was”. Fixed, sentence was divided for better readability.
   2. Table 4. Check the typo in the table. That is, “condition”. Fixed
   3. Table 5. Check the typo in the table. That is, “condition”. Fixed

 

7. Tensile Specimen Geometry and Test Method
   1. Line 223. Check the grammatical error. That is, “a universal testing machine”. Fixed
   2. Line 235. One usually uses “tempering” for heat treatment of ferritic/martensitic steels. The materials you used in this study is nickel-based alloy, so I don’t catch the exact meaning you want here.
      Answer: From Fig. 8 it can be seen, that the mech. strength is decreased for diffusion bonded and heat treated samples, compared to the delivery condition. Heat treatment over 1000°C will first lead to a decrease in dislocation density by recovery (rearrangement of dislocations) and recrystallization (formation of a new grain structure with grains low in dislocation density). Furthermore, due to long dwell time at high temperatures, gain growth will occur, decreasing mech. strength according Hall-Petch-relationship.
      Exponential increase of dislocation density is the reason for hardening effect. At the same time, remaining deformation reserve decreases with increasing dislocation density.
   3. properties after 1200°C/4h:We agree. This is a severe treatment. Despite we did not check the materials properties according ASME Sec. II Part D, it seems that the mech. values obtained are just a little below the specifications. There are two possible answers now:
      First: There is no technical reason imaginable why a material should be subjected to a heat treatment as for diffusion bonding!
      Second: In general, diffusion bonding is an exotic joining technology, which is not widespread and is only used if no other process is applicable. In the literature, there are hardly any strength values after diffusion bonding that come close to those of the starting material in the heat-treated state. Diffusion bonded joints almost always have significantly lower strength and elongation at break values. This is due to the application of high temperatures and long dwell times. Unless polymorphic materials are involved, considerable grain growth is always to be expected. Without grain growth there is no grain growth above joining levels! If diffusion bonding is used, there is no way to avoid corresponding preliminary tests for optimization. It may be necessary to derive lower yield strength values for dimensioning of devices from these tests.

Bonding temperature and bearing pressure acts in opposite directions. Since both show strong nonlinearity, systematic investigations are necessary Experiments on stainless steel (to be published) showed that comparable mech. prpopertiescan be obtained at lower temperatures and higher bearing pressures, combined with higher tensile strength due to less pronounced grain growth.

8. Results of Tensile Tests on Hastelloy B3
   1. Table 6. Specimen information is missed: Just realized & fixed.

 

9. Results of Tensile Tests on Hastelloy BC-1
   1. Line 290. One usually uses “tempering” for heat treatment of ferritic/martensitic steels. Answer: In our opinion, heat-treatment is more appropriate here. For ferritic/martensitic steels, a temperature treatment may be used to release intrinsic distortion of the lattice without changing the materials microstructure considerably. Here, the heat-treatment has a strong effect on dislocation density as well as on grain size.

 

10. Conclusions
    1. The Conclusion have already been revised and streamlined as suggested by another reviewer. The heading was extended to “Conclusion & Outlook”.

 

11. References
    1. Line 380 (372): Author L. Paul corrected
    2. Line 401 (393): 2016=> 2017 (print date)
    3. & 4: We have added some literature in English. However, since the editor had removed the cross reference function, it was very difficult to update and very error-prone. The editor should reconsider this.

5. E.g. reference numbers in the manuscript were adapted to rules given by metals.

Round 2

Reviewer 2 Report

I am happy with the authors' corrections and recommend its publication to Metals Journal.