Review Reports - Oxidation Effects on Short-Term Creep Response in Air of Commercially Pure Titanium (CP-2 Ti)

Round 1

Reviewer 1 Report

1. It has been shown in many works that the creep rate depends on the structural state of the material. There is no data at least on the average grain size before and after creep tests, clarification is required. Remarks about structural data (at least the grain size) also apply to samples with preliminary annealing for 70 hours (before and after annealing).

2. It is desirable to give information (or at least a link to other researchers) what happens to the structure at temperatures of 550, 600 and 650 (study of the thermal stability of the structure).

3. Creep tests carried out before rupture? Provide data on the magnitude of breaking strain.

4. Page 2, lines 75-77 ((«An interesting feature is that after load-changes in VLEs, the minimum creep rates are usually well below those recorded under similar loads in CLEs.»). But in fig. 1, such a regularity is obvious only for Fig. 1с. For fig. 1a can only speculate. For fig. 1b it is difficult to state this (44 MPa is significantly higher than 40 MPa).

5. For the convenience of perceiving the graphs, I recommend highlighting the dependencies for CLE and VLE separately (by color or by lines of different types).

6. Page 2 lines 93-94. What is the reason for the change in the exponent n?

7. Page 3 lines 106-109. («This effect can hardly be explained in terms of microstructural mechanisms in this pure metal unless the effect of oxidation in considered. Thus, a more detailed and quantitative analysis of the hardening effect of oxidation was required»). It is important to understand what happens to the structure during the creep tests at given temperatures (at least the grain size). Oxidation certainly affects hardening and creep rate. But is this the predominant strengthening mechanism in this case?

8. There is some contradiction:

page 1 line 25-27 «Detailed studies, such as [1], just to quote a single example, showed that oxidation reaction is very fast during the initial stage of the process and leads to the formation of oxide layers composed of TiO2 structures» and page 11 line 351-352 «The curves for exposures shorter than 5 h were calculated by assuming that oxidation did not occur. In general term, the accuracy of the model is more than acceptable.» An explanation of this assumption is required.

9. Page 11 lines 353-354 “Figure 8d reports the experimental data from [14]. The authors did not specify if a protective atmosphere was used, which led one to suppose that testing was performed in air." The peer-reviewed article deals with the influence of the atmosphere (oxygen), so when placing links it is important to indicate whether the source is relevant to the issue under consideration.

10. Page 13 lines 385-386. "This fact implies that oxygen-rich pure-Ti is assimilable, in terms of mechanical properties, to more complex industrial alloys."

Different kinds of impurities can differently affect short-term (for example, microhardness) and long-term (for example, tension in creep mode) mechanical characteristics. And the increase in microhardness does not always indicate an increase in the durability of the material.

11. Page 8 lines 229-230

"In particular, the oxide layer largely detaches from the substrate during creep (Figure 4), and for this reason its strengthening contribution can be thought to be negligible"

And page 11 lines 330-331 "As a result, load is transferred from the soft to the hard portions of the sample, giving...".

Is there a contradiction here?

Author Response

It has been shown in many works that the creep rate depends on the structural state of the material. There is no data at least on the average grain size before and after creep tests, clarification is required. Remarks about structural data (at least the grain size) also apply to samples with preliminary annealing for 70 hours (before and after annealing).

The plate was provided in annealed state; we clarified this aspect in the revised text. An annealed material should be not significantly affected by further permanence at high temperature, although a certain grain growth could occur. Figure 4 clearly shows that this was not the case since the grain size remained substantially unchanged.

It is desirable to give information (or at least a link to other researchers) what happens to the structure at temperatures of 550, 600 and 650 (study of the thermal stability of the structure).

This point could be of great significance in cold-worked materials. Yet, the material of the present study was already provided in annealed state. Nor the grain size, nor the HV values (150-160 HV) were substantially affected by further high-temperature exposure in the unstressed portions of the samples. Once this aspect has been clarified (as done in the revised text), in our opinion the reader should be not “overloaded” with details which are not significant for the study.

Creep tests carried out before rupture? Provide data on the magnitude of breaking strain.

The tests were interrupted before fracture. Typical strains at test interruption reached 30-40%. We included a sentence to clarify this point.

Page 2, lines 75-77 ((«An interesting feature is that after load-changes in VLEs, the minimum creep rates are usually well below those recorded under similar loads in CLEs.»). But in fig. 1, such a regularity is obvious only for Fig. 1с. For fig. 1a can only speculate. For fig. 1b it is difficult to state this (44 MPa is significantly higher than 40 MPa).

In a way, this comment is correct. The strengthening effect due to high-T exposure should be obtained by multiple evidence. We re-wrote this paragraph to better clarify this point.

For the convenience of perceiving the graphs, I recommend highlighting the dependencies for CLE and VLE separately (by color or by lines of different types).

The Figures have been modified following this suggestion.

Page 2 lines 93-94. What is the reason for the change in the exponent n?

The higher creep exponent at 650°C is an effect of oxidation, namely, for long time of exposure, the thickness of the hard-creep resistant layer increases, and the creep rate decreases. This fact was not commented in the location indicated by the reviewer, since this conclusion was premature at that stage of the analysis. Yet, we modified the draft by including a comment on the mechanisms that lead to high stress exponents.

Page 3 lines 106-109. («This effect can hardly be explained in terms of microstructural mechanisms in this pure metal unless the effect of oxidation in considered. Thus, a more detailed and quantitative analysis of the hardening effect of oxidation was required»). It is important to understand what happens to the structure during the creep tests at given temperatures (at least the grain size). Oxidation certainly affects hardening and creep rate. But is this the predominant strengthening mechanism in this case?

As already stated under comments 1 and 2, the grain size does not change. An increase in grain size would lower the hardness and UTS. The effect on creep response of an increase in grain size is usually negligible, at least for these “normal” grain sizes. No other strengthening mechanisms could occur in a pure metal as an effect of high-T exposure. Thus, we did not modify the text to consider this comment.

There is some contradiction:

The comment is correct. We modified the sentence in: The curves for exposures shorter than 5 h were calculated by assuming that the hard oxygen-rich layer (zone II) was too thin to produce any significant effects.

Page 11 lines 353-354 “Figure 8d reports the experimental data from [14]. The authors did not specify if a protective atmosphere was used, which led one to suppose that testing was performed in air." The peer-reviewed article deals with the influence of the atmosphere (oxygen), so when placing links it is important to indicate whether the source is relevant to the issue under consideration.

We did not indeed understand this comment. The mentioned dataset is VERY relevant, because is just one of the very few recent studies dealing on creep of pure Ti. The subject was investigated (many) decades ago, so high-quality results are now scarcely available. The analysis of data from [14] remains an interesting point to be considered. We did not take any actions on this regard.

Page 13 lines 385-386. "This fact implies that oxygen-rich pure-Ti is assimilable, in terms of mechanical properties, to more complex industrial alloys."

Also, in this case we did not fully understand the comment. We were dealing with mechanical properties – UTS- at room temperature. This point has been considered at length in literature studies on Ti-O alloys [13,19,20,21]. We slightly modified the text to clearly explain that the sentence was considering tensile properties.

Page 8 lines 229-230

"In particular, the oxide layer largely detaches from the substrate during creep (Figure 4), and for this reason its strengthening contribution can be thought to be negligible"

And page 11 lines 330-331 "As a result, load is transferred from the soft to the hard portions of the sample, giving...".

Is there a contradiction here?

No, there is no contradiction here. The load, in our model, is transferred to zone II, that is, the portion of oxygen-rich layer not cracked. We more clearly analyzed the point by modifying the sentence.

Reviewer 2 Report

1. The authors write An interesting feature is that after load-changes in VLEs, the minimum creep rates are usually well below those recorded under similar loads in CLEs. This phenomenon is well appar¬ent at 600°C and at 650°C, see, for example, the VLE at the highest temperature, where the strain rate under 40 MPa (nominal; the true stress value was 42.9 MPa) is much lower than that measured in a CLE under the same nominal stress. However, this conclusion can only be drawn for the results obtained at 650°C.

2. Figures 3 should have line numbers to distinguish between hardness and oxygen curves.

3. The authors allow not very correct statements: “ii. an intermediate zone (Ib) of almost constant and very high (roughly 700 HV) microhardness; these HV values attest a brittle nature". Brittleness should be indicated not so much by high hardness as by microbrittleness, determined by the area occupied by cracks coming from the corners of the indenter impression.

4. The authors presented a very simplified model of titanium creep after oxidation, which can be used with significant errors.

5. The authors, unfortunately, practically did not analyze the actual distribution of oxygen over the thickness of the hardened layer and the relationship between the oxygen content in titanium, changes in microhardness and creep.

Author Response

The authors write An interesting feature is that after load-changes in VLEs, the minimum creep rates are usually well below those recorded under similar loads in CLEs. This phenomenon is well apparent at 600°C and at 650°C, see, for example, the VLE at the highest temperature, where the strain rate under 40 MPa (nominal; the true stress value was 42.9 MPa) is much lower than that measured in a CLE under the same nominal stress. However, this conclusion can only be drawn for the results obtained at 650°C.

In a way, this comment is correct. The strengthening effect due to high-T exposure should be obtained by multiple evidence. We re-wrote this paragraph to better clarify this point.

Figures 3 should have line numbers to distinguish between hardness and oxygen curves.

The Figures have been modified following this suggestion.

The authors allow not very correct statements: “ii. an intermediate zone (Ib) of almost constant and very high (roughly 700 HV) microhardness; these HV values attest a brittle nature". Brittleness should be indicated not so much by high hardness as by microbrittleness, determined by the area occupied by cracks coming from the corners of the indenter impression.

The sentence was indeed based on literature evidence on O-rich pure Ti, not on a specific experimental finding in this paper. The point was better clarified in the revised text.

The authors presented a very simplified model of titanium creep after oxidation, which can be used with significant errors.

This comment is substantially correct but does not imply the necessity to modify the draft. Nevertheless, we included the new sentence “In addition, a simple, easy to handle and reasonably accurate model for quantification of the effects of oxidation on creep response (minimum creep rate) will be proposed.” To better identify our major goal.

The authors, unfortunately, practically did not analyze the actual distribution of oxygen over the thickness of the hardened layer and the relationship between the oxygen content in titanium, changes in microhardness and creep.

We never stated that our goal was to investigate the oxygen distribution and its correlation with hardness. Specific investigations on these regards are already available in the literature, and some of them have been quoted and commented in the revised text. As for the last comment, to identify the effect of oxygen on creep of pure titanium is impossible, unless Ti-O alloys, as those discussed in [13,19,20,21], are creep-tested. As obvious, this goes well beyond the aims (and possibilities) of the authors of the present study.

Reviewer 3 Report

This manuscript focused on the effect of oxidation on the short-term creep response of commercially pure Ti at 550-650 ℃. Both the experiment and related simulation had successfully demonstrated the occurrence of a marked decrease of the creep rate. The experiment has been well-designed and the results are sounding. This manuscript should be published after minor revisions, which are shown below:

1. The introduction is too short. It is recommended to have a more in-depth review on this research topic.

2. It is recommended to add some SEM-EDX results to show the chemical composition distribution around the near-surface region.

3. In Fig. 4d, it is clear to see that the microstructure of the near-surface matrix is very different from the matrix away from the surface. It is recommended to add EBSD results to show their differences.

Author Response

The introduction is too short. It is recommended to have a more in-depth review on this research topic.

The comment is not so clear. In any case, the reference list includes ALL the works we were able to find on the effect of oxygen on creep response of Ti-alloys (the effect on pure Ti was never tested, at least in recent times). We nevertheless slightly modified the Introduction giving additional information of the effect of oxidation in pure Ti.

It is recommended to add some SEM-EDX results to show the chemical composition distribution around the near-surface region.

We never stated that our goal was to investigate the oxygen distribution. Specific investigations on these regards are already available in the literature, and some of them have been quoted and commented in the text. What we needed was to understand what the thickness of the hardened zone was, and this can be done only with hardness measurements. The specific value of the oxygen content at a given location is an interesting datum but does not provide additional information that can be used to model the creep response. Thus, in our opinion the text does not require amendments on this regard.

In Fig. 4d, it is clear to see that the microstructure of the near-surface matrix is very different from the matrix away from the surface. It is recommended to add EBSD results to show their differences.

The difference is due to a degree of cold work introduced in the softer zones during metallographic preparation. The plate was received in annealed state and presented an equiaxed structure clearly shown in the previous paper (Regev, M.; Almoznino, B.; Spigarelli, S. A Study of the Metallurgical and Mechanical Properties of Friction-Stir-Welded Pure Titanium. Metals 2023, 13, 524). The present Figure shows the microstructure in the sample heads, i.e. in almost unstressed portions of the material. The HV values do not change after long exposure and the grain size remains substantially the same. The “cold work effect” of Figure 4d is just the result of metallographic preparation of the sample. The effect has been explained in the revised text.