Impact of Thermomechanical Fatigue on Microstructure Evolution of a Ferritic-Martensitic 9 Cr and a Ferritic, Stainless 22 Cr Steel

Round 1

Reviewer 1 Report

This manuscript presents the benefit of the 22 Cr steel against the damage by thermal cycles. The target materials were carefully chosen from the viewpoint of engineering application, and the difference is very clearly exhibited. However, the scientific result are difficult to be found and the explanation for the microstructural observation is not suitable. These should be revised for the academic publication. The followings might have to be considered.

(1) Materials. The comparison between grade 91 (9Cr) and Crofer(R)22 h (22Cr) was conducted in this work. The commercial reason for this choice is really understandable. However, the metallurgical microstructures of these steels have several big differences, and this leads difficulty for the metallurgical consideration. It should be needed to describe previous work on the ferritic steels with laves phase (if possible) and all the microstructural factors having the difference between these two steels such as difference of the solute atoms in the matrix.

(2)P.2, Subsection 2.1

The solidification process is important for inclusion-control. If it is possible, please describe the solidification and refinement processes of these materials which should be conducted before the heat treatment.

(3)p.3 EBSD (microstructural observation)

Please identify the observation condition for SEM/EBSD. Acc. voltage for SEM, step size, possible phase list for EBSD analysis at the data collection should be documented.

(4)p.4, Fig.3

The description for the notations, ε_mech and  ε_th cannot be found in the text. In addition, is this clear difference of apparent elastic modulus (linier slope of the stress-strain curves) between these two samples and among the cycles? (The microstructures in both the steels changed during the test.)

(5)Fig.4 EBSD

It is difficult to find the band contrast images in Fig.4 (a-e). These should be just line maps indication the position of high-angle grain boundaries. In addition, some oxides were illustrated in (f), however, no information of the key phases with EBSD analysis.

(5') Fig.9 high angle boundaries 

Why was the definition of the HAGB set as 3deg? 

 

 

 

Author Response

Response to reviewer 1:

(1) Materials. The comparison between grade 91 (9Cr) and Crofer(R)22 h (22Cr) was conducted in this work. The commercial reason for this choice is really understandable. However, the metallurgical microstructures of these steels have several big differences, and this leads difficulty for the metallurgical consideration. It should be needed to describe previous work on the ferritic steels with laves phase (if possible)…

Re: Previous and current, parallel work (development, microstructure, mechanical properties and application) on the Laves phase strengthened, ferritic, stainless steel is covered in detail in the cited references [15, 16, 20-34, 36, 38, 39].

…  and all the microstructural factors

Re: … which are:

• tempered martensitic structure (P91): already covered in the manuscript
• fully ferritic structure (Crofer22 H): covered in the manuscript
• MX and M23C6 precipitation (P91): covered in the manuscript
• Laves phase precipitation (Crofer22 H): covered in the manuscript

… having the difference between these two steels such as difference of the solute atoms in the matrix.

Re: The differences in chemical compositions are covered in Table 1. For sure, there is a difference in the solid solution content of the two materials. The materials were chosen, because they yield almost the same stress range in the stable regime of the fatigue curve (which is stated in the manuscript at the top of subsection 3.1). For this reason, they are most comparable from an application point of view.

Following your suggestion and to emphasize the difference in solid solution content the sentence “From a metallurgical point of view the main difference between the two materials besides microstructure and strengthening precipitate species are the differing solid solution contents of chromium, tungsten and molybdenum.” has been added in subsection 2.1.

 

(2)P.2, Subsection 2.1

The solidification process is important for inclusion-control. If it is possible, please describe the solidification and refinement processes of these materials which should be conducted before the heat treatment.

Re: Some modification on available details on the production routes of the materials, like “… by vacuum induction melting (VIM)” in case of Crofer22 H has been implemented into subsection 2.1.

In case of the P91 material used for our study unfortunately, no information on the production route is available. Nevertheless, the material obeyed ASTM A335 specifications and thus the production route applied guarantees the specified properties. An according statement “…, manufactured by Vitkovice Steel, CZ according to ASTM A335 specifications.” has been implemented into subsection 2.1.

(3)p.3 EBSD (microstructural observation)

Please identify the observation condition for SEM/EBSD. Acc. voltage for SEM

Re: The missing information was implemented into subsection 2.3: …(acceleration voltage: 10 kV)

…, step size (EBSD)

Re: The missing information was implemented into the figure captions of Fig. 4 (“step size: 0.2191 mm”) and 9 (“step size: 0.438 mm”)

, possible phase list for EBSD analysis at the data collection should be documented.

Re: A statement concerning the missing information was implemented into subsection 2.3: “Electron backscatter diffraction (EBSD) was utilized to characterize the orientation relations of individual grains of the Fe_bcc matrix and the analysis of localized deformation.”

(4)p.4, Fig.3

The description for the notations, ε_mech and  ε_th cannot be found in the text.

Re: Descriptions have been implemented into subsection 2.2.

In addition, is this clear difference of apparent elastic modulus (linier slope of the stress-strain curves) between these two samples and among the cycles? (The microstructures in both the steels changed during the test.)

Re: There are no real differences in elastic modulus between the materials and cycle numbers. At a first glance it may look like that, but:

The stress-strain hysteresis loops in thermomechanical fatigue testing are not recorded isothermally (in case of our experiments at dT/dt = 10Ks^-1), i.e. the linear slope parts of the individual loops are thus composed from changing elastic modulus with temperature.

Furthermore, the only recognizable difference in the linear elastic slope sections is between the entry part (the branch starting from strain = 0 / stress = 0 and propagating towards lower strain and stress) of cycle 1 and the following cycles for both the materials.

This is caused by the fact, that the temperature profile (distribution) of the material volume within the gauge length of the specimen in the entry part of cycle 1 inevitably differs from all the following cycles, because:

Cycle 1 is started from more or less “isothermal” condition, i.e. the specimen is heated to the lower temperature of the cycle (250 °C) by the induction heating system of the testing machine, applying a comparably slow heating rate (0.5 Ks^-1) to prevent overshooting.

The experiment is then started by initializing the temperature and strain controllers, i.e. the homogenously warm specimen gauge length is further heated (and thermal expansion is automatically obstructed) until the maximum temperature (650 °C) of the cycle is reached, where the heating system is switched off.

Specimen cooling (or better say: cooling of the gauge length) is then forced by blowing compressed air onto the specimen surface, until the surface temperature of the gauge length (recorded by a sling thermocouple located in the middle of the gauge length) reaches the minimum temperature of the cycle, where the heating system is switched on again.

Under the applied rapid cooling condition (dT/dt = 10Ks^-1, used to keep creep at the hot end of the cycle to a minimum), the gauge length material volume faces small axial and lateral temperature inhomogeneity (less than 5 °C).

These little deviations, together with the elastic modulus being a function of changing temperature cause the deviating slope of the entry part of cycle 1.

(5)Fig.4 EBSD

It is difficult to find the band contrast images in Fig.4 (a-e). These should be just line maps indication the position of high-angle grain boundaries. In addition, some oxides were illustrated in (f), however, no information of the key phases with EBSD analysis.

Re: The original figure caption was misleading or imprecise in this instance. It has been changed to “Binarized images (a - e) from post-processed electron backscatter diffraction band contrast images (misorientation angle used to define fatigue induced sub-grain boundaries: > 3°, step size: 0.2191 mm) for the quantitative analysis of martensite lath degradation and complete polygonization of ferritic-martensitic grade 91 steel after 0 (a), 10 (b), 250 (c), 1250 (d) and 3500 (e) TMF cycles. (f): sub-grain orientation added to (e).” to clearly indicate, that the images displayed are not the original EBSD band contrast images, but the ones processed for quantitative image analysis of martensite lath polygonization.

The oxides within the crack in subfigures (e) and (f) were not in the focus of our examination, because they do not play a role in microstructural degradation of the bulk material, which governs fatigue crack initiation. For this reason, the oxide phases were not given in the original manuscript. We now changed the description from just “oxide” to “FeCr-mixed oxide” and added the sentences “The iron and chromium containing mixed oxide within the crack do not play a role in microstructural degradation of the bulk material and the initiation of short fatigue cracks. A possible effect on the propagation of long fatigue cracks and thus residual lifetime is subject of ongoing research.” into the test.  

(5') Fig.9 high angle boundaries 

Why was the definition of the HAGB set as 3deg? 

Re: The misorientation angle set for the evaluation of grain boundaries, newly formed by fatigue induced dislocation activity, was set to > 3°. This must not be mistaken with pre-existing high angle grain boundaries (HAGBs). For this reason the three pre-existing high angle boundaries, visible in Fig. 9, where denoted with “HABG” text boxes.

Various phrases have been implemented to clarify this issue. For example in the captions of Fig. 4 (… images (misorientation angle used to define fatigue induced sub-grain boundaries: > 3° …) and Fig. 9 (… with fatigue induced sub-grain boundaries…), as well as in subsection 2.3: …  A misorientation angle of > 3° was applied for the definition of fatigue induced sub-grain boundaries.  …

The definition value of > 3° misorientation for fatigue induced sub-grain boundaries was chosen, because it allowed the best illustration of the encountered polygonization phenomenon.

 

Re: Thank you very much for your valuable comments, which helped to increase the quality of the manuscript.

 

Reviewer 2 Report

The present submission reports microstructural evolution, damage, and failure of two steels under thermomechanical fatigue loading.  It is a good quality of work, of particular interest to the readers who works in the field. However, there are several minor issues to be checked for better readability. I personally appreciate the very detailed explanation throughout the manuscript.

 

1) Readability. Yet, the results and discussion are reasonably structured. I believe that there are still many rooms to improve the readability of the manuscript. In addition, too many parentheses and hyphens are unnecessarily used, which further breaks the flow of the paper including abstract.

2) [Introduction] The current version is acceptable. Yet, it would be much beneficial to potential readers including more details about the literature that have been reported so far with reference(s) in the last paragraph.

3) minor comments. Some stylistic comments.

Please check all the stylistic things carefully, before submitting a manuscript. For example…  

-No space between the number, the symbol, and the Latin letters “C”.

-Please double-check the stylistic guideline for the abbreviation of weight percent. Is this right, “wt.-%”?

-Variables are italicized.

-I personally am familiar with using “,” as a decimal separator. Yet, please double-check these decimal points. Please keep consistency in style. E.g. “<0,01” and “0.07” in table1.

-Leave a space after the sign. “<0” to “< 0” in table1.

-Is it common to use underlines when using abbreviations for example “Electron backscatter diffraction (EBSD)”?

-Please use abbreviations if the authors started to use them. [line91] “out-of-phase” (oop) cycles > [line95] oop cycles > [line122] out-of-phase cycles
-many others.

 

Overall, the reviewer believes that the key experimental results are interesting and certainly publishable. However, as argued in the review, some parts are not sufficient with the presentation.

Author Response

Response to reviewer 2:

1) Readability. Yet, the results and discussion are reasonably structured. I believe that there are still many rooms to improve the readability of the manuscript. In addition, too many parentheses and hyphens are unnecessarily used, which further breaks the flow of the paper including abstract.

Re: We rearranged parts of the manuscript for a couple of times during preparation of the original manuscript and felt that the result is the best way of comprehensively reporting the differing mechanisms and not losing connection between the two materials. For this reason we´d like to keep the manuscript as it is and not rearrange the structure again. We sincerely hope that this o.k. for you.

We avoided the usage of parentheses and hyphens in the revised manuscript.

2) [Introduction] The current version is acceptable. Yet, it would be much beneficial to potential readers including more details about the literature that have been reported so far with reference(s) in the last paragraph.

Re: The related literature is extensively utilized for discussion of the outlined results and cited in the context of the respective microstructural mechanisms in section 3. Nevertheless we implemented the literature citations into the introduction and changed the list of references accordingly.

3) minor comments. Some stylistic comments.

Please check all the stylistic things carefully, before submitting a manuscript. For example…  

-No space between the number, the symbol, and the Latin letters “C”.

Re: Done

-Please double-check the stylistic guideline for the abbreviation of weight percent. Is this right, “wt.-%”?

Re: Done

-Variables are italicized.

Re: Done

-I personally am familiar with using “,” as a decimal separator. Yet, please double-check these decimal points. Please keep consistency in style. E.g. “<0,01” and “0.07” in table1.

Re: Done

-Leave a space after the sign. “<0” to “< 0” in table1.

Re: Done

-Is it common to use underlines when using abbreviations for example “Electron backscatter diffraction (EBSD)”?

Re: Done intuitively. Has not been changed in submissions to other MDPI journals.

-Please use abbreviations if the authors started to use them. [line91] “out-of-phase” (oop) cycles > [line95] oop cycles > [line122] out-of-phase cycles
-many others.

Re: Done

Overall, the reviewer believes that the key experimental results are interesting and certainly publishable. However, as argued in the review, some parts are not sufficient with the presentation.

Re: Thank you very much for your valuable comments, which helped to increase the quality of the manuscript.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

n/a