Submit to this Journal Review for this Journal Propose a Special Issue

Article Menu

Share Help Cite Discuss in SciProfiles

Open AccessFeature PaperArticle

Peer-Review Record

Phase Stress Measurement of Centrifugally Cast Duplex Stainless Steel by Neutron Diffraction

Quantum Beam Sci. 2020, 4(3), 28; https://0-doi-org.brum.beds.ac.uk/10.3390/qubs4030028

by Yun Wang

Reviewer 1: Anonymous

Reviewer 2:

Joe Kelleher

Reviewer 3: Anonymous

Quantum Beam Sci. 2020, 4(3), 28; https://0-doi-org.brum.beds.ac.uk/10.3390/qubs4030028

Submission received: 10 June 2020 / Revised: 21 July 2020 / Accepted: 23 July 2020 / Published: 29 July 2020

(This article belongs to the Special Issue Analysis of Strain, Stress and Texture with Quantum Beams)

Round 1

Reviewer 1 Report

Reviewer comments on

Phase Stress Measurement of Centrifugally Cast Duplex Stainless Steel by Neutron Diffraction

Yun Wang

General comments

The paper describes the application of the well established stress/strain determination by means of neutron diffraction. The power of this method is shown. The paper is clearly written. The conclusions are derived evidently from the results and discussion. The English should be improved. More references should be given.

The reviewer recommends accepting with minor revisions.

Detailed comments

Abstract:

Improve the English.

Introduction:

At the beginning of the century, a lot of work on the field of stress distribution between the phases in austenitic/ferritic steels was conducted and published by M. Daymond et al.. Similar results were obtained here. These should be mentioned in the introduction and used in the discussion.

Section 2.1.

The volume concentrations of the two phases were determined by EBSD. In the Rietveld refinement, the volume concentrations has to be refined too. Please give the volume concentrations and compare them with the results of the EBSD.

Section 2.3 and 2.4.1

The author describes the determination of the stress-free lattice parameter. A rotating bunch of four foils with a thickness of 60 – 90 µm were measured with a neutron beam having a width of 2 mm. This is a setup which can result easily in systematic errors: If the beam extends the sample dimensions in the scattering plane and the sample is asymmetrical positioned in the beam, the peaks shift slightly. How was it prevented or are the slightly differences detected by the two detectors caused by this reason?

2 mm ^x 2 mm is not a volume.

Author Response

General comments

→Thank you very much for your comments. I have done some corrections on the English and added more references.

The reviewer recommends accepting with minor revisions.

Detailed comments

Abstract:

Improve the English.

→I have done some corrections on English.

Introduction:

→I have rewritten the introduction. Not only M. Daymond et al. but also some other researchers’ work related to this study were introduced.

Section 2.1.

→I have added the phase volume fractions estimated by Rietveld refinement and compared them to the EBSD results (p.7 Line16-p.8 Line 5). Considering the Rietveld refinement results should be more reliable than EBSD, I have chosen the phase volume fractions obtained by Rietveld refinement to calculate the macro-stress and thus Figure 10 and 13 have also been revised. However, the conclusions haven’t changed anyway.

Section 2.3 and 2.4.1

→The key point is how to make a good alignment among (1) the vertical rotation axis of stage (2) the vertical center axis of gauge volume and (3) the vertical rotation axis of the strain-free specimens. It was confirmed by the equipment group that a perfect alignment between (1) and (2) had been realized during the initial calibration of the equipment. In this experiment, the alignment between (1) and (3) had been carefully carried out with laser level measurement devices within a rotation range from 0 to 90 degree to limit the deviation between (1) and (3) as low as possible. As you have pointed out, this deviation might cause a slight difference between the two detectors. Considering the width of the specimens is 3 mm>2×√2mm (the largest width inside the gauge volume)，when you observe them in the side view direction the specimens should fill up the gauge volume from its edge to edge at each rotation angle, which means the part of the specimens inside the gauge volume keeps symmetrical to (2) all the time during the rotation even though the specimens are asymmetrical to (3). According to lattice constants obtained by the two detectors, which is shown in Figure 9, the phase stress errors due to the difference in lattice constants are approximately 10 MPa for a phase and 5 MPa for g phase (simply estimated by using Young’s modulus in p.4 Line 12 and supposing the average value of the two detectors is the true lattice constant).

2 mm x 2 mm is not a volume.

→I have corrected the gauge volume as 2 mm x 2 mm x 2 mm.

※ I have highlighted the revised contents in red color for your convenience (please see the attached PDF file)

Author Response File: Author Response.pdf

Reviewer 2 Report

Interesting paper. Residual stress measurement with neutrons is widely applied to industrial components, and normally for a paper like this I'd hope to see some modelling of the quenching process that you only discuss qualitatively. However this is the most rigorous treatment of stress measurement in two-phase material that I've seen so far. While many authors have used the approach you've used here (weighted volume fractions etc.), they have lamely appealed to 'common sense' while you actually show some experimental justification for your approach e.g. the in-situ tests and d0 samples. Accordingly, this paper is valuable reference for others measuring two-phase materials. (Incidentally it would be good to have something analogous for hexagonal materials). The technical points I'd make are as follows:

1) Your procedure of averaging together angles from 0 degrees to 180 or 160 degrees would only average away the anisotropy in the material itself ( texture) , but not necessarily the pseudostrain caused by the sample not completely filling the gauge volume. A better way to do that would be to average together a pair of measurements at angles 180 degrees apart from each other (many such pairs would be even better, but your approach is like averaging together only one of each 180-degree pair without including the other one. Did you average together the raw spectra before fitting? If so what lattice spacing do you get as a function of angle if you fit each spectrum separately? A misalignment would make this plot look like part of a sine wave, which you could fit to extract the average value maybe. Happy to help with this - please get in touch through editor if needed.

2) Having said that, the Rietveld fits look good, and the agreement between banks in Fig 9 is encouraging with regard to possible sample positioning and pseudostrain errors. Still, please be clear about what Fig 8 is – one particular angle, or a sum of all angles? If the latter, how does the fitted lattice parameter vary when banks and directions are fitted separately?

3) You quote and discuss literature values for stress-free lattice spacings in single phase (pure elemental?) material. Your C, N, Ni and Mo alloying additions will certainly affect the stress-free lattice parameter compared to single phase material, unless you know the alloy composition for the cited single phase material matches the individual phase components of your material. Otherwise, it’s no surprise that the d0 values don’t agree.

Apart from that, my only other comments concern grammar and presentation:

Introduction

Neutron diffraction method -> neutron diffraction

A diffraction device -> an instrument

under the applied -> under applied

was discussed -> ‘was deduced’ or ’is discussed’, as appropriate

Material and methods

× 27 mm (wall thickness). -> with a wall thickness of 27 mm. [Not really correct to use multiplication of diameter and wall thickness]

× 16 mm (wall thickness) -> [similarly…]

tensile test.-> tensile testing

as TP16. -> as sample set TP16 [Your TPxx sample numbers refer to sets of identical samples with the same purpose I think, rather than literal individual samples, yes? No need to keep repeating the words ‘sample set’ (or whatever phrase you choose) after this though].

for tensile test -> for tensile testing

strait -> straight. Also the ‘straight parallel part’ of a tensile sample is normally called the ‘gauge length’.

phase volume rate f -> phase volume fraction f [rate implies something changing with time]

(EBSD) method -> (EBSD)

(OIM) -> (OIM) module [I’m assuming OIM refers to the hardware while EBSD refers to the technique. If so it’s not really necessary to mention OIM as EBSD is mainly used to measure orientation images as far as I’m aware]

inverse pole figure (IPF) -> inverse pole figure (IPF) image

is 2 mm×2 mm… is 4um -> was 2mm×2 mm… was 4um.

phase volume rate, -> phase volume fraction, or phase volume ratio

the phase partition -> the area identified as a given phase [incidentally I think it’s safe to assume the ratio of areas is the ratio of volumes for this microstructure ("Delesse principle"), though there are other cases where this is not necessarily true].

of neutron spending on the -> for the neutron to traverse the

of diffraction -> of the diffraction

stress directions.-> strain directions [or even just ‘sample directions’]

then can be determined -> then can be determined for each phase [for clarity, this helps emphasise you’re still talking about stresses in the individual phases, since you’re not using a subscript to specifically refer to the phase]

to the internal -> to perform internal

triaxial -> the triaxial

that the lattice structure -> that this lattice spacing [you’re talking specifically about the a0 value here yes? Not wholesale FCC-BCC phase changes]

is uneasy -> requires caution

carefully fabricated in -> their surfaces prepared using [also no real need to use quotes and arrows, could just say emery paper, diamond paste and electropolishing.]

hardening layer -> work-hardened layer

longitudinal direction -> diffraction vector for the North detector [is this what you mean?]

till -> until

nine diffraction intensity -> Why not ten? (0 to 180 degrees inclusive?) Was the last measurement at 160 degrees? If so change text to quote this value.

Ground surface -> do you mean the surface of the earth, or the surface of one of your samples that you ground? Clearer to just say ‘floor’ if the latter.

discussing data scatter.-> investigating data scatter

strait -> straight

“loading-unloading” -> no need for quotes here

What’s the relevance of reference [8] in the figure 6 and 7 captions?

Results

Similar tendency -> a similar tendency

In rdirection, -> In the r direction

in h direction -> In the h direction

equal biaxial -> equibiaxial

good balance -> realistic stress balance

no obvious distribution of macro-stress -> no obvious distribution of radial macro-stress

original phase stress sigma^alpha lower than epsilon0^gamma -> you’re comparing stress with strain in that statement, approach may be valid but please fix phrasing

Discussion

in h and z direction -> in the h and z directions

with SHT temperature -> at SHT temperature

barricaded instantaneously from emitting outside -> briefly contained and prevented from leaving

inner surface is retarded. -> inner surface is retarded in comparison to the outside surface.

temperature gap -> temperature difference

Author Response

Comments and Suggestions for Authors

→Thank you very much for your comments. Yes, if I had rotated them till 360 deg, then I could have obtained every complete pair of measurements at angles 180 degrees apart from each other. I didn’t average together the raw spectra separately before fitting. Instead, I summed up all the raw spectra obtained at all the rotation angles into one spectrum. Then I conducted the Rietveld refinement on that spectrum only.

→If the alignment was perfect among (1) the vertical rotation axis of stage (2) the vertical center axis of gauge volume and (3) the vertical rotation axis of the strain-free specimens, then the pseudostrain should be small enough, even though the samples couldn’t completely fill the gauge volume. However, exactly as you have pointed out, the misalignment still exists and is possibly reflected into the slight difference of lattice constants obtained from the two detectors, which is indicated in Figure 9. Actually I estimated the effect on the phase stress errors due to this difference between the two detectors. They are approximately 10 MPa for Alpha phase and 5 MPa for Gamma phase (simply estimated by using Young’s modulus in 2.2 and supposing the average value of the two detectors is the true lattice constant).

Fig 8 is a sum of all angles from the same detector. I didn’t do the fitting on the spectrum at every angel separately because from the raw spectrum I noticed that the peak heights varied in a wide range due to textured microstructure, especially in Alpha phase. I think it is also related to the coarse grains in a phase.

→Yes. The composition of minor elements could also affect the lattice constants. I just would like to suggest that carful attention should be payed even if we have similar lattice information in hand.

Apart from that, my only other comments concern grammar and presentation:

→ I deeply appreciate you for such a careful review. I have reflected all the corrections that you suggested in the revised paper.

※ I highlighted the revised contents in red color for your convenience (see the attached PDF file). Please be aware that the revised parts also include the corrections responding to other reviewers’ comments.

1.Introduction

Neutron diffraction method -> neutron diffraction

→Corrected in p. 2-Line 9 (and similar cases in other places).

A diffraction device -> an instrument

→Corrected in p. 2-Line 11

under the applied -> under applied

→Corrected in p. 2-Line 16. I corrected as "under the uniaxial tensile loading"

was discussed -> ‘was deduced’ or ’is discussed’, as appropriate

→Corrected in p. 2-Line 15

2.Material and methods

× 27 mm (wall thickness). -> with a wall thickness of 27 mm. [Not really correct to use multiplication of diameter and wall thickness]

→Corrected in p. 2-Line 23

× 16 mm (wall thickness) -> [similarly…]

→Corrected in p. 2-Line 30

tensile test.-> tensile testing

→Corrected in p. 2-Line 33

→Corrected in p. 2-Line 31, 34

for tensile test -> for tensile testing

→Corrected in p. 2-Line 33

strait -> straight. Also the ‘straight parallel part’ of a tensile sample is normally called the ‘gauge length’.

→Corrected in p. 2-Line 35

phase volume rate f -> phase volume fraction f [rate implies something changing with time]

→Corrected in p. 2-Line 37 (and the same cases in other places)

(EBSD) method -> (EBSD)

→Corrected in p. 2-Line 37

→Corrected in p. 2-Line 38

inverse pole figure (IPF) -> inverse pole figure (IPF) image

→Corrected in p. 3-Line 2 and Fig. 2 (b)

is 2 mm×2 mm… is 4um -> was 2mm×2 mm… was 4um.

→Corrected in p. 3-Line 3

phase volume rate, -> phase volume fraction, or phase volume ratio

→Corrected in p. 3-Line 5

→Corrected in p. 3-Line 6

For responding to the comments from other reviewers, I have added the phase volume fractions estimated by Rietveld refinement and compared them to the EBSD results (p.7 Line16-p.8 Line 5). Considering the Rietveld refinement results should be more reliable than EBSD, I have chosen the phase volume fractions obtained by Rietveld refinement to calculate the macro-stress and thus Figure 10 and 13 have also been revised. However, the conclusions haven’t changed anyway.

of neutron spending on the -> for the neutron to traverse the

→Corrected in p. 3-Line 17

of diffraction -> of the diffraction

→Corrected in p. 4-Line 3

stress directions.-> strain directions [or even just ‘sample directions’]

→Corrected in p. 4-Line 8

→Corrected in p. 4-Line 11

to the internal -> to perform internal

→Corrected in p. 4-Line 25

triaxial -> the triaxial

→Corrected in p. 4-Line 25

that the lattice structure -> that this lattice spacing [you’re talking specifically about the a0 value here yes? Not wholesale FCC-BCC phase changes]

→Corrected in p. 4-Line 28

is uneasy -> requires caution

→Corrected in p. 4-Line 30

carefully fabricated in -> their surfaces prepared using [also no real need to use quotes and arrows, could just say emery paper, diamond paste and electropolishing.]

→Corrected in p. 4-Line 44-45

hardening layer -> work-hardened layer

→Corrected in p. 4-Line 47

longitudinal direction -> diffraction vector for the North detector [is this what you mean?]

→Corrected in p. 5-Line 5

Here the longitudinal direction refers to 8 mm-side of the samples in Fig. 1. In order to avoid misunderstanding, I corrected this phrase to “longitudinal direction of the plates”.

till -> until

→Corrected in p. 5-Line 9

nine diffraction intensity -> Why not ten? (0 to 180 degrees inclusive?) Was the last measurement at 160 degrees? If so change text to quote this value.

→Corrected in p. 5-Line 9 and Fig. 3 (b)

I mistyped the step of rotation angle. The correct rotation step is 22.5 deg. I corrected it in the article.

Ground surface -> do you mean the surface of the earth, or the surface of one of your samples that you ground? Clearer to just say ‘floor’ if the latter.

→Corrected in p. 5-Line 23-24

In order to avoid misunderstanding, I corrected it to “stage surface”

discussing data scatter.-> investigating data scatter

→Corrected in p. 6-Line 2

strait -> straight

→Corrected in p. 7-Line 6

“loading-unloading” -> no need for quotes here

→Corrected in p. 7-Line 7

What’s the relevance of reference [8] in the figure 6 and 7 captions?

→Same experiments have been reported in my another article.

3.Results

Similar tendency -> a similar tendency

→Corrected in p. 9-Line 6

In r direction, -> In the r direction

→Corrected in p. 9-Line 7

in h direction -> In the h direction

→Corrected in p. 9-Line 11

equal biaxial -> equibiaxial

→Corrected in p. 9-Line 13

good balance -> realistic stress balance

→Corrected in p. 9-Line 15

no obvious distribution of macro-stress -> no obvious distribution of radial macro-stress

→Corrected in p. 9-Line 10

original phase stress sigma^alpha lower than epsilon0^gamma -> you’re comparing stress with strain in that statement, approach may be valid but please fix phrasing

→Corrected in p. 10-Line 14

4.Discussion

in h and z direction -> in the h and z directions

→Corrected in p. 10-Line 26

with SHT temperature -> at SHT temperature

→Corrected in p. 10-Line 28

barricaded instantaneously from emitting outside -> briefly contained and prevented from leaving

→Corrected in p. 10-Line 30

inner surface is retarded. -> inner surface is retarded in comparison to the outside surface.

→Corrected in p. 10-Line 31

temperature gap -> temperature difference

→Corrected in p. 11-Line 4

Author Response File: Author Response.pdf

Reviewer 3 Report

This paper studies stress of centrifugally cast duplex stainless steel by neutron diffraction (TOF) technique. The manuscript is very well prepared. There are few comments which can improve the quality of the work:

Section 1: The introduction section is too short, especially when the second para is more about neutron methodology. Please could you elaborate more and provide more literature where such testing/measurements has been carried.
Section 2.3: For strain free specimens, is the method proposed unique with this experiment, or is it a well-established method. If yes, please provide reference or discuss with additional information and limitations. How many such specimens were prepared?
Section 2.4.1: Detectors (designated as North and South) may not be appropriate for the manuscript. May be mention as detection 1 and 2. What is the third dimension in the Gauge volume (2 mm x 2 mm x ?). Why the holding time was set at 600 s (provide rational or share experience).
Section 2.4.3: Why tested under held at 180 s and 2.6% axial strain?
Section 4: Can the limitations of neutron measurements and significance about the stresses in the proposed geometry be expanded more. Could you enhance the discussion little bit more? Say where/how the results can be compared qualitatively from other experimental study.

Author Response

→Thank you very much for your comments. I have done some corrections on the English and added more references.

Section 1: The introduction section is too short, especially when the second para is more about neutron methodology. Please could you elaborate more and provide more literature where such testing/measurements has been carried.

→I have rewritten the introduction. More related previous studies were introduced.

Section 2.3: For strain free specimens, is the method proposed unique with this experiment, or is it a well-established method. If yes, please provide reference or discuss with additional information and limitations. How many such specimens were prepared?

→I have added the more literature and compared the method in this work to the conventional ones (p.4 Line 26-37).

Section 2.4.1: Detectors (designated as North and South) may not be appropriate for the manuscript. May be mention as detection 1 and 2.

→I have corrected as Detector 1 and Detector 2 .

What is the third dimension in the Gauge volume (2 mm x 2 mm x ?).

→I have corrected as 2 mm x 2 mm x 2mm.

Why the holding time was set at 600 s (provide rational or share experience).

→Considering the diffraction intensity related to the output power of neutron beam, which I have added in p.2 Line 13, the proper holding time was set as 600s.

Section 2.4.3: Why tested under held at 180 s and 2.6% axial strain?

→For the purpose of continuous measurement during plastic deformation, a large gauge volume in size of 5 mm × 5 mm × 3 mm was utilized to acquire high diffraction intensity (I have added this explanation in p.6 Line 10-p.7 Line 1). Therefore a relatively short holding time was applied. The tensile test was conducted until the maximum applied stress σ_A had approximately reached 500 MPa, just exceeding its proof stress sy by 20-30 MPa. Correspondingly the maximum applied strain ε_A in the axial direction was nearly 2.6 % (p.7 Line 8-9).

Section 4: Can the limitations of neutron measurements and significance about the stresses in the proposed geometry be expanded more. Could you enhance the discussion little bit more? Say where/how the results can be compared qualitatively from other experimental study.

→I have added more comparisons with the previous studies in p.9 Line 4-6, p.10 Line 16-18. I have discussed the prospect about the relief of residual stress and optimization of the component shape based on the precedent research in p.11 Line 8-16.

※ I have highlighted the revised contents in red color for your convenience (please see the attached PDF file)

Author Response File: Author Response.pdf

Article Menu

Phase Stress Measurement of Centrifugally Cast Duplex Stainless Steel by Neutron Diffraction

Further Information

Guidelines

MDPI Initiatives

Follow MDPI