the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 16 Jan 2023
The influence of siliciclastic content on the strength and deformation behavior of rock salt – Constraints from thermomechanical experiments
Abstract. Halite forms the main constituent of rock salt, which is regarded as a possible host rock for nuclear waste repositories and storage caverns. The deformation behavior of pure halite and rock salt has been revealed by microfabric analyses of naturally and experimentally deformed samples. Such studies, however, are rare for rock salt with significant amount of secondary phases (e.g., siliciclastics, clay, anhydrite), although these are also common in nature. In order to determine the influence of siliciclastic material on the deformation behaviour of rock salt, we performed deformation experiments using rock salt samples with variable siliciclastic content (1, 7, 38 and 53 vol. %). The experiments were conducted under bulk flattening strain, elevated temperature (345 °C), low differential stress (< 4.6 MPa), and a strain rate of 10-7 s-1. To gain inside in the 3D distribution of the siliciclastic material and in the deformation mechanisms of the constituent minerals, computer tomographic (CT), microstructural and electron backscatter diffraction (EBSD) analyses were applied to both initial and experimentally deformed samples. The EBSD and microstructural data suggest that, independent of the amount of siliciclastic content, the deformation of the halite matrix was largely accommodated by subgrain formation without subgrain rotation recrystallization. The deformation of the siliciclastic domains, on the other hand, was entirely brittle. CT images show open fractures, oriented sub-perpendicular to the least principal stress, σ3. An increase in the siliciclastic content leads to an increase in differential stress of the halite matrix. The new results suggest that the barrier properties of rock salt is significantly reduced by larger content of siliciclastic material, particularly in cases where the siliciclastic parts and their fractures are interconnected making pervasive ascendant fluid transport possible in these lithological units. Future thermomechanical experiments of impure rock salt should focus on the effect of confining pressure, which is expected to reduce the number and width of open fractures in the siliciclastic domains.
- Preprint
(2278 KB) - Metadata XML
- BibTeX
- EndNote
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2022-1432', Anonymous Referee #1, 10 Feb 2023
The manuscript by Linckens et al. investigates the influence of siliciclastic second phase on the strength and deformation behaviour of rock salt. Especially in the context of the safety and suitability of geological repositories in salt formations, a valuable and interesting question to address.
Four natural samples of variable siliciclastic content, ranging from 1 to 53 vol.%, were deformed in thermomechanical experiments. 2- and 3-dimensional analysis techniques (electron microscopy and CT scanning) were combined to investigate the dominant deformation mechanisms in different domains of the samples. While the halite matrix in all samples was dominated by ductile deformation as evident from the formation of subgrains, the siliciclastic domains showed brittle deformation features such as fractures. On this basis, a positive correlation between the siliciclastic content of the samples and their strength was established.
The experimental approach seems largely sound. Though relatively simple in nature, it seems appropriate to address the underlying hypothesis. The given detail of information, however, does not allow for the unambiguous reproduction of the results as information on acquisition parameter, image analysis protocols etc. are missing. This lack of information should be addressed by the authors during revision. In addition to that, the presented results are largely sufficient to support the conclusions. However, here and there they seem to leave space for non-evidence-based speculation (see scientific comments). The interpretation of the results is not always clear to the reader and large variations in parameters that can affect the deformation behaviour (e.g. grain size and sample composition) do not allow to fully correlate the behaviour of the samples to the amount of second phases.
Further, the authors argue that knowledge about the deformation behaviour of impure rock salt is crucial to estimate the integrity of potential host rocks for nuclear waste. In this context open fractures would represent pathways for the migration of radio nuclides. However, neither was an interconnected fracture network observed that transects the halite nor were experiments conducted under confined conditions which may suppress the formation of open fractures. In this sense the applicability of the presented results to engineering problems and their merit are questionable. In its current form the main message is that samples with a higher amount of siliciclastic phases are harder to deform. It would be crucial to point out what the consequences are for the formation and density of brittle deformation features.
Nonetheless, I acknowledge that the concept of this study is a valuable addition to recent studies on layered sequences of rock salt and suggest major revisions.
Scientific comments:
Line 44: Spiers et al. 1990 establishes that pressure solution in halite is diffusion controlled and formulates a constitutive rate law. It does not conclude that a fluid film along the grain boundaries is prerequisite for pressure solution in halite. Additional references are needed e.g. Urai et al. 1986 (for the comparison of dry & wet halite); Raj 1982, Rutter 1983 & Gratier et al. 1987 (for a more general take on fluid films as prerequisite for pressure solution)
Line 44 - 47: The connection between a fluid film as a prerequisite for pressure solution and fluid inclusions along halite grain boundaries seems unclear.
Line 47-51: Experimental studies have shown, that pressure solution also occurs in closed systems and fluid flow/advection is not prerequisite for pressure solution (e.g. Renard et al. 2001, Renard et al. 2004, Macente et al. 2018, Schwichtenberg et al. 2022…)
The conclusion that stylolites do not occur in rock salt is based on the false assumption that rock salt in deeper structural levels acts as a closed system in which pressure solution is absent. I agree that a lack of open and connected porosity that allows fluid flow may play a crucial role in prohibiting the formation of stylolites, however, we know too little about the actual formation of stylolites to reduce it purely to the presence or absence of pressure solution and fluid flow. The main question in this context is what causes strain to localize?Line 94: To validate this statement, it needs to be compared to the deformation behaviour of a pure salt sample (see also Results section).
Figure 1: In a) the label “South Permian Basin” is very difficult to read and the relevant structure, the Krempe salt wall, hard to identify. A zoom-in of the Glückstadt-Graben & including the Krempe salt wall would add clarity to the figure and improve its accessibility. What is the location of the cross-section c)?
In b) The lithology column contains “Rotliegend rock salt”. That however is self-explanatory as is belongs to the Rotliegend Group. In addition to that, DSK (2016) is missing in the references nor is explained what it stands for.
What is the location of the borehole?Line 114: As far as I’m aware the viscosity was not determined in the presented manuscript.
Line 115: How many samples were used? Why was no pure halite reference sample investigated? In order to quantify the effect of the siliciclastic material this should be done.
Table 1: What is the difference between the starting samples and the experimentally deformed samples is. From the information given one would expect that the starting sample is either the sample before deformation or a reference/ archive sample that was taken from the same core. However, it is very confusing that neither sample names can be matched nor the depth range.
Figure 2: I recommend to include the amount of siliciclastica in the figure as well as a representative length scale. What is the orientation of the cubes within the drill core?
Line 118-119: What information is given in Table 1 in context of the thick sections?
Line 121: Please provide more detailed and relevant information on the data acquisition or alternatively give a reference that includes the scan parameters (e.g. voxel size, voltage, current, beam width…).
Line 123-124: How was this done? Please include details about the image analysis workflow including the segmentation procedure, software used, were the images processed in any way (filtering, noise reduction...), considering the partial volume effect in CT-scanning, were "grains" below a certain size excluded from the analysis?
Line 125-126: This is a very rough estimate of the error. Depending on the method used to analyse the 3D distribution of phases there are more elegant ways to estimate a more precise error (e.g. via the probability of the segmentation).
Line 129: A sketch of the experimental set-up would add clarity to the description. This could be included in the appendix.
Line 131-132: How much thermal expansion was expected; hence, how much space was left between the sample and the plates? Did this have an impact on the heating of the samples? What about the surfaces that were open to the atmosphere? How do you ensure that heat was not lost via these surfaces and that no temperature gradient evolved? Did you monitor the temperature of the actual sample?
Line 141: Was the strain rate constant for all samples? Also, the unit is missing!
Line 143: When was the second CT scan acquired? If it was after deformation, what is the effect of the cooling and removal of the confining plates on the microstructure?
Line 146: The strain rate used in the experiments 2x10-7 s-1 is an order of magnitude higher than the fastest natural strain rates.
Line 147-148: The reasoning behind choosing 340°C remains unclear. Also, if it is not relevant to engineering problems you are trying to address, why going that high? How do you scale your observations to natural processes?
Line 149-151: This should go into the discussion.
Line 152: Did you mean one thick section per sample? Please clarify!
Line 155-157: Which tools did you use to trace the subgrain boundaries and calculate the subgrain areas. What is the error of this method? Is it reproducible?
Regarding ImageJ, an appropriate reference would be Schindelin et al. 2012.Line 157-158: I assume you used D=k*σ-m. What did you use for k and m? Please make this calculation more comprehensible!
Line 165-166: CT data allow for a more precise determination of the volume fraction than an “approximation”. Please make use of that. Simple approximations can be done by e.g. point counting or comparison to a chart. You are using a more advanced technique that allows you to give an error. So within this error your determination is quite accurate.
Line 167: What defines the heterogeneous distribution. Is it layers or clusters? From the images it looks like a heterogeneity due to differences in the grain size. Please specify!
Figure 2: Please give the amount of second phases for each sample (can be included next to the cube). What was the orientation of the axes during the experiment? Would a different orientation cause a difference in the deformation behaviour?
Figure 3: What is the orientation of the samples? Were the scans acquired before or after deformation?
Line 169: What is the composition of the individual samples? Do they all have the same proportions of different phases? What is its influence on the strength and deformation behaviour of the samples? Does a quartz dominated sample deform differently than a illite dominated sample? I would assume so!
Line 170: That is quite a large range. Is it the same for all samples? How does the grain size of the second phase affect the strength of the bulk sample? Do you see differences in deformation behaviour between domains dominated by large grains vs. domains dominated by smaller grains?
Line 171-172: Again, quite a large grain size distribution. Is it the same for all samples? How does it affect the deformation of the sample?
Line 174-176: Does this all refer to sample GM-HT1? What about the other samples? Did you observe a change in shape from the undeformed to the deformed sample?
Line 177-179: This could be nicely shown in an image.
Line 180: reference to Fig. 4a?
Figure 4: Please include in the caption which sample is shown. Also, it’s not clear if this is before & after deformation or 2 different samples or just 2 different views from the same sample. In a) include arrow pointing to subgrain boundaries and what does gb stand for (explanation in caption). In b) what does subgb stand for (caption)? Please clarify that there are 2 “types” of subgrain boundaries visible. Further, the scale can not be read!
Figure 5: Which sample is shown? Is it representative for all 4 samples?
Line 183: I do not understand why the standard deviation is used as the average grain size for paleopiezometry. Intuitively I would use d(sub) from your table. This needs to be clarified! Please also see the comment on Table 2.
Line 190-192: What is the maximum amount of siliciclastica that allows for a good measurement? What are the limits of this set up?
Figure 6: Please explain the meaning of the star also in the captions. In order to really see an influence of the secondary phase it is necessary to have a pure halite sample plotted as well. The legend should be sorted by sample names 1 to 4. What happened at the zigzag point of GM-HT4? Why do not all samples all start in the origin? If you have a constant strain rate, it’s worth mentioning it here.
Figure 7: Increase contrast of images to make the distinction between the phases easier. The fractures seem to follow old discontinuities so, instead of newly formed fractures these maybe reactivated ones at the interface of halite to the siliciclastic phase. What does the fracture network look like in 3D?
Please also identify the silicilastic phase in both samples. They appear rather different; do they also have a strength contrast?
What is the strain of the deformed sample? Is it at final stage?
What is the influence of the different composition/ grain sizes of the siliciclastic material on the deformation behaviour?
How do you explain the formation of fractures of different orientation? How do the pores form? Do the fractures propagate into the salt or are they limited to the siliciclastic phase? Are they interconnected and influence the permeability of the sample or are they isolated fractures?
Please use coloured arrows to make them more distinguishable!Line 206: Where is this visible? How do you distinguish them from subgrains already existing in the sample before deformation?
Figure 9: In a) the arrow is very hard to see in the black and white image. In d) What is the fluid inclusion here?
Figure 11: What does IPF stand for? Please clarify! Also, the difference between the 5-10° and >10° line is not clear from the EBSD map.
Table 2: Please clarify. What is st. dev.? And what is d(sub)? Which one did you use for the paleopiezometry? I assume you used D=k*σ-m, then D should be the average subgrain size d(sub)? This was done for the deformed samples but not for the starting samples.
Line 222: What about grain size & type of secondary material? In order to make reliable statements on the pure effect of the amount of second phase, other parameters need to be carefully controlled and comparable to each other. This is very difficult with natural samples!
Line 222-223: Needs comparison to a pure NaCl! Otherwise, this statement is not tenable.
Line 224-225: reads “the stress increases with increasing second phase content.” This is misleading as it also increases with increasing strain. What you are comparing is the strength of the sample via the measured stress that is needed to deform the sample.
Line 240-241: Needs support by references.
Line 246: What was the initial SPO of the grains? Was there one at all? What was the grain shape?
Line 248: What is the grain size of the individual samples? That needs to be clarified. The range of 50µm to 1cm does not allow to simply draw this conclusion. Another reason for the absence of PS might be the absence of a connected fluid phase?!
Line 251: What is the effect of anhydrite? Why was there no weakening?
Line 270: The question is, when do the fractures form and how long does it take for the salt to migrate into these fractures?
Line 270-275: I don’t understand how these observations are similar when you say that the halite was able to flow into the boudinage necks.
Line 285: What are possibilities to explain the pores? Please start discussion here.
Line 309-311: Is there a correlation between the siliciclastic content and the fracture density or density of possible pathways for hazardous material?
Technical comments:
Line 22: To gain insight…
Line 38: …make rock salt…
Line 46: …salt show…
Line 55: …by the formation of…
Line 78-83: The citation is valid for the entire text block. It does not need to be referenced in every single sentence.
Line 88: Needs support by citations. What are examples of experimental studies focused on elastic parameters & small strains?
Line 119: In order to visualize…
Line 128: …the cubes were confined…
Line 129: Two? The? … experiments
Line 138: …by the motion of the right plate. The sample was not confined in the…
Line 145: Talbot and Jackson…
Line 156: …by fitting an ellipse…
Line 224: strength instead of stress
Line 228: show
Line 273: …in the present study
Line 130 onwards: Please check that the methods are reported consistently in past tense.
Tabel 2: 4th column: MPa
Line 244: … our experiments?
Line 249-250: Sound like these experiments were part of the present study/ larger project Maybe better “During deformation experiments…”
Line 253-254: Liang et al. (2007) also observed… composite samples decreased.
Line 278 & 281: wrong citation style, should be Martin-Clave et al. (2021)
Citation: https://doi.org/10.5194/egusphere-2022-1432-RC1 -
RC2: 'Comment on egusphere-2022-1432', Janos Urai, 15 Mar 2023
Review of:
The influence of siliciclastic content on the strength and deformation behavior of rock salt – Constraints from thermomechanical experiments by Linckens et al.,
J.L. Urai
This paper attempts to put some constraints on the effect of siliciclastic impurities on the mechanical properties and deformation mechanisms of rock salt with application to the evolution of heat generating nuclear waste repositories. They do this by deforming samples with different fraction of siliciclastic impurities, in the Zulauf machine, unconfined at 345 C.
In my opinion, there are a number of major issues which need to be addressed before the paper can be published.
First, the temperature used - there are many experimental data published at these high temperature (Heard, Carter, Senseny etc) and the results here need to be compared with these. The stress-strain-data in this unique instrument need to be corrected for friction effects to be compared. In addition, the temperatures used here are higher than allowed in a nuclear waste repository and the Authors need to explain why and how their results are relevant.
Second, the unconfined experiments - it has been shown by several studies (papers by Peach) that these conditions allow grain boundary brine to escape and stop grain boundary migration. So in my opinion if the authors would have preserved the in-situ brine content of their samples and did confined, non-dilatant tests, the samples would have completely recrystallized. But maybe these experiments are (only?) relevant to the unconfined, high temperature EDZ around a repository - can I ask the Authjors' opinion on this? In general, in my opinion these experiments are not relevant to the far field, non dilatant conditions around a repository because there dislocation creep and pressure solution are both important processes and the rock salt is not dry.
Third, the effect of impurities: in the dislocation creep regime of these experiments the effect of impurities can be described by the law of mixtures with the Voigt and Reuss bounds (see papers by Bons and many others). The authors need to include analysis of their results along these lines, otherwise the differences between the samples' stress-strain behaviour is very difficult to understand.
Fourth, the Representative Elementary volume. In my opinion, the distribution of impurities, especially in samples HT1 and HT4, is such, that it does not represent the behaviour of a larger mass. I would like to invite the Authors to consider this, and discuss its consequences.
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2022-1432', Anonymous Referee #1, 10 Feb 2023
The manuscript by Linckens et al. investigates the influence of siliciclastic second phase on the strength and deformation behaviour of rock salt. Especially in the context of the safety and suitability of geological repositories in salt formations, a valuable and interesting question to address.
Four natural samples of variable siliciclastic content, ranging from 1 to 53 vol.%, were deformed in thermomechanical experiments. 2- and 3-dimensional analysis techniques (electron microscopy and CT scanning) were combined to investigate the dominant deformation mechanisms in different domains of the samples. While the halite matrix in all samples was dominated by ductile deformation as evident from the formation of subgrains, the siliciclastic domains showed brittle deformation features such as fractures. On this basis, a positive correlation between the siliciclastic content of the samples and their strength was established.
The experimental approach seems largely sound. Though relatively simple in nature, it seems appropriate to address the underlying hypothesis. The given detail of information, however, does not allow for the unambiguous reproduction of the results as information on acquisition parameter, image analysis protocols etc. are missing. This lack of information should be addressed by the authors during revision. In addition to that, the presented results are largely sufficient to support the conclusions. However, here and there they seem to leave space for non-evidence-based speculation (see scientific comments). The interpretation of the results is not always clear to the reader and large variations in parameters that can affect the deformation behaviour (e.g. grain size and sample composition) do not allow to fully correlate the behaviour of the samples to the amount of second phases.
Further, the authors argue that knowledge about the deformation behaviour of impure rock salt is crucial to estimate the integrity of potential host rocks for nuclear waste. In this context open fractures would represent pathways for the migration of radio nuclides. However, neither was an interconnected fracture network observed that transects the halite nor were experiments conducted under confined conditions which may suppress the formation of open fractures. In this sense the applicability of the presented results to engineering problems and their merit are questionable. In its current form the main message is that samples with a higher amount of siliciclastic phases are harder to deform. It would be crucial to point out what the consequences are for the formation and density of brittle deformation features.
Nonetheless, I acknowledge that the concept of this study is a valuable addition to recent studies on layered sequences of rock salt and suggest major revisions.
Scientific comments:
Line 44: Spiers et al. 1990 establishes that pressure solution in halite is diffusion controlled and formulates a constitutive rate law. It does not conclude that a fluid film along the grain boundaries is prerequisite for pressure solution in halite. Additional references are needed e.g. Urai et al. 1986 (for the comparison of dry & wet halite); Raj 1982, Rutter 1983 & Gratier et al. 1987 (for a more general take on fluid films as prerequisite for pressure solution)
Line 44 - 47: The connection between a fluid film as a prerequisite for pressure solution and fluid inclusions along halite grain boundaries seems unclear.
Line 47-51: Experimental studies have shown, that pressure solution also occurs in closed systems and fluid flow/advection is not prerequisite for pressure solution (e.g. Renard et al. 2001, Renard et al. 2004, Macente et al. 2018, Schwichtenberg et al. 2022…)
The conclusion that stylolites do not occur in rock salt is based on the false assumption that rock salt in deeper structural levels acts as a closed system in which pressure solution is absent. I agree that a lack of open and connected porosity that allows fluid flow may play a crucial role in prohibiting the formation of stylolites, however, we know too little about the actual formation of stylolites to reduce it purely to the presence or absence of pressure solution and fluid flow. The main question in this context is what causes strain to localize?Line 94: To validate this statement, it needs to be compared to the deformation behaviour of a pure salt sample (see also Results section).
Figure 1: In a) the label “South Permian Basin” is very difficult to read and the relevant structure, the Krempe salt wall, hard to identify. A zoom-in of the Glückstadt-Graben & including the Krempe salt wall would add clarity to the figure and improve its accessibility. What is the location of the cross-section c)?
In b) The lithology column contains “Rotliegend rock salt”. That however is self-explanatory as is belongs to the Rotliegend Group. In addition to that, DSK (2016) is missing in the references nor is explained what it stands for.
What is the location of the borehole?Line 114: As far as I’m aware the viscosity was not determined in the presented manuscript.
Line 115: How many samples were used? Why was no pure halite reference sample investigated? In order to quantify the effect of the siliciclastic material this should be done.
Table 1: What is the difference between the starting samples and the experimentally deformed samples is. From the information given one would expect that the starting sample is either the sample before deformation or a reference/ archive sample that was taken from the same core. However, it is very confusing that neither sample names can be matched nor the depth range.
Figure 2: I recommend to include the amount of siliciclastica in the figure as well as a representative length scale. What is the orientation of the cubes within the drill core?
Line 118-119: What information is given in Table 1 in context of the thick sections?
Line 121: Please provide more detailed and relevant information on the data acquisition or alternatively give a reference that includes the scan parameters (e.g. voxel size, voltage, current, beam width…).
Line 123-124: How was this done? Please include details about the image analysis workflow including the segmentation procedure, software used, were the images processed in any way (filtering, noise reduction...), considering the partial volume effect in CT-scanning, were "grains" below a certain size excluded from the analysis?
Line 125-126: This is a very rough estimate of the error. Depending on the method used to analyse the 3D distribution of phases there are more elegant ways to estimate a more precise error (e.g. via the probability of the segmentation).
Line 129: A sketch of the experimental set-up would add clarity to the description. This could be included in the appendix.
Line 131-132: How much thermal expansion was expected; hence, how much space was left between the sample and the plates? Did this have an impact on the heating of the samples? What about the surfaces that were open to the atmosphere? How do you ensure that heat was not lost via these surfaces and that no temperature gradient evolved? Did you monitor the temperature of the actual sample?
Line 141: Was the strain rate constant for all samples? Also, the unit is missing!
Line 143: When was the second CT scan acquired? If it was after deformation, what is the effect of the cooling and removal of the confining plates on the microstructure?
Line 146: The strain rate used in the experiments 2x10-7 s-1 is an order of magnitude higher than the fastest natural strain rates.
Line 147-148: The reasoning behind choosing 340°C remains unclear. Also, if it is not relevant to engineering problems you are trying to address, why going that high? How do you scale your observations to natural processes?
Line 149-151: This should go into the discussion.
Line 152: Did you mean one thick section per sample? Please clarify!
Line 155-157: Which tools did you use to trace the subgrain boundaries and calculate the subgrain areas. What is the error of this method? Is it reproducible?
Regarding ImageJ, an appropriate reference would be Schindelin et al. 2012.Line 157-158: I assume you used D=k*σ-m. What did you use for k and m? Please make this calculation more comprehensible!
Line 165-166: CT data allow for a more precise determination of the volume fraction than an “approximation”. Please make use of that. Simple approximations can be done by e.g. point counting or comparison to a chart. You are using a more advanced technique that allows you to give an error. So within this error your determination is quite accurate.
Line 167: What defines the heterogeneous distribution. Is it layers or clusters? From the images it looks like a heterogeneity due to differences in the grain size. Please specify!
Figure 2: Please give the amount of second phases for each sample (can be included next to the cube). What was the orientation of the axes during the experiment? Would a different orientation cause a difference in the deformation behaviour?
Figure 3: What is the orientation of the samples? Were the scans acquired before or after deformation?
Line 169: What is the composition of the individual samples? Do they all have the same proportions of different phases? What is its influence on the strength and deformation behaviour of the samples? Does a quartz dominated sample deform differently than a illite dominated sample? I would assume so!
Line 170: That is quite a large range. Is it the same for all samples? How does the grain size of the second phase affect the strength of the bulk sample? Do you see differences in deformation behaviour between domains dominated by large grains vs. domains dominated by smaller grains?
Line 171-172: Again, quite a large grain size distribution. Is it the same for all samples? How does it affect the deformation of the sample?
Line 174-176: Does this all refer to sample GM-HT1? What about the other samples? Did you observe a change in shape from the undeformed to the deformed sample?
Line 177-179: This could be nicely shown in an image.
Line 180: reference to Fig. 4a?
Figure 4: Please include in the caption which sample is shown. Also, it’s not clear if this is before & after deformation or 2 different samples or just 2 different views from the same sample. In a) include arrow pointing to subgrain boundaries and what does gb stand for (explanation in caption). In b) what does subgb stand for (caption)? Please clarify that there are 2 “types” of subgrain boundaries visible. Further, the scale can not be read!
Figure 5: Which sample is shown? Is it representative for all 4 samples?
Line 183: I do not understand why the standard deviation is used as the average grain size for paleopiezometry. Intuitively I would use d(sub) from your table. This needs to be clarified! Please also see the comment on Table 2.
Line 190-192: What is the maximum amount of siliciclastica that allows for a good measurement? What are the limits of this set up?
Figure 6: Please explain the meaning of the star also in the captions. In order to really see an influence of the secondary phase it is necessary to have a pure halite sample plotted as well. The legend should be sorted by sample names 1 to 4. What happened at the zigzag point of GM-HT4? Why do not all samples all start in the origin? If you have a constant strain rate, it’s worth mentioning it here.
Figure 7: Increase contrast of images to make the distinction between the phases easier. The fractures seem to follow old discontinuities so, instead of newly formed fractures these maybe reactivated ones at the interface of halite to the siliciclastic phase. What does the fracture network look like in 3D?
Please also identify the silicilastic phase in both samples. They appear rather different; do they also have a strength contrast?
What is the strain of the deformed sample? Is it at final stage?
What is the influence of the different composition/ grain sizes of the siliciclastic material on the deformation behaviour?
How do you explain the formation of fractures of different orientation? How do the pores form? Do the fractures propagate into the salt or are they limited to the siliciclastic phase? Are they interconnected and influence the permeability of the sample or are they isolated fractures?
Please use coloured arrows to make them more distinguishable!Line 206: Where is this visible? How do you distinguish them from subgrains already existing in the sample before deformation?
Figure 9: In a) the arrow is very hard to see in the black and white image. In d) What is the fluid inclusion here?
Figure 11: What does IPF stand for? Please clarify! Also, the difference between the 5-10° and >10° line is not clear from the EBSD map.
Table 2: Please clarify. What is st. dev.? And what is d(sub)? Which one did you use for the paleopiezometry? I assume you used D=k*σ-m, then D should be the average subgrain size d(sub)? This was done for the deformed samples but not for the starting samples.
Line 222: What about grain size & type of secondary material? In order to make reliable statements on the pure effect of the amount of second phase, other parameters need to be carefully controlled and comparable to each other. This is very difficult with natural samples!
Line 222-223: Needs comparison to a pure NaCl! Otherwise, this statement is not tenable.
Line 224-225: reads “the stress increases with increasing second phase content.” This is misleading as it also increases with increasing strain. What you are comparing is the strength of the sample via the measured stress that is needed to deform the sample.
Line 240-241: Needs support by references.
Line 246: What was the initial SPO of the grains? Was there one at all? What was the grain shape?
Line 248: What is the grain size of the individual samples? That needs to be clarified. The range of 50µm to 1cm does not allow to simply draw this conclusion. Another reason for the absence of PS might be the absence of a connected fluid phase?!
Line 251: What is the effect of anhydrite? Why was there no weakening?
Line 270: The question is, when do the fractures form and how long does it take for the salt to migrate into these fractures?
Line 270-275: I don’t understand how these observations are similar when you say that the halite was able to flow into the boudinage necks.
Line 285: What are possibilities to explain the pores? Please start discussion here.
Line 309-311: Is there a correlation between the siliciclastic content and the fracture density or density of possible pathways for hazardous material?
Technical comments:
Line 22: To gain insight…
Line 38: …make rock salt…
Line 46: …salt show…
Line 55: …by the formation of…
Line 78-83: The citation is valid for the entire text block. It does not need to be referenced in every single sentence.
Line 88: Needs support by citations. What are examples of experimental studies focused on elastic parameters & small strains?
Line 119: In order to visualize…
Line 128: …the cubes were confined…
Line 129: Two? The? … experiments
Line 138: …by the motion of the right plate. The sample was not confined in the…
Line 145: Talbot and Jackson…
Line 156: …by fitting an ellipse…
Line 224: strength instead of stress
Line 228: show
Line 273: …in the present study
Line 130 onwards: Please check that the methods are reported consistently in past tense.
Tabel 2: 4th column: MPa
Line 244: … our experiments?
Line 249-250: Sound like these experiments were part of the present study/ larger project Maybe better “During deformation experiments…”
Line 253-254: Liang et al. (2007) also observed… composite samples decreased.
Line 278 & 281: wrong citation style, should be Martin-Clave et al. (2021)
Citation: https://doi.org/10.5194/egusphere-2022-1432-RC1 -
RC2: 'Comment on egusphere-2022-1432', Janos Urai, 15 Mar 2023
Review of:
The influence of siliciclastic content on the strength and deformation behavior of rock salt – Constraints from thermomechanical experiments by Linckens et al.,
J.L. Urai
This paper attempts to put some constraints on the effect of siliciclastic impurities on the mechanical properties and deformation mechanisms of rock salt with application to the evolution of heat generating nuclear waste repositories. They do this by deforming samples with different fraction of siliciclastic impurities, in the Zulauf machine, unconfined at 345 C.
In my opinion, there are a number of major issues which need to be addressed before the paper can be published.
First, the temperature used - there are many experimental data published at these high temperature (Heard, Carter, Senseny etc) and the results here need to be compared with these. The stress-strain-data in this unique instrument need to be corrected for friction effects to be compared. In addition, the temperatures used here are higher than allowed in a nuclear waste repository and the Authors need to explain why and how their results are relevant.
Second, the unconfined experiments - it has been shown by several studies (papers by Peach) that these conditions allow grain boundary brine to escape and stop grain boundary migration. So in my opinion if the authors would have preserved the in-situ brine content of their samples and did confined, non-dilatant tests, the samples would have completely recrystallized. But maybe these experiments are (only?) relevant to the unconfined, high temperature EDZ around a repository - can I ask the Authjors' opinion on this? In general, in my opinion these experiments are not relevant to the far field, non dilatant conditions around a repository because there dislocation creep and pressure solution are both important processes and the rock salt is not dry.
Third, the effect of impurities: in the dislocation creep regime of these experiments the effect of impurities can be described by the law of mixtures with the Voigt and Reuss bounds (see papers by Bons and many others). The authors need to include analysis of their results along these lines, otherwise the differences between the samples' stress-strain behaviour is very difficult to understand.
Fourth, the Representative Elementary volume. In my opinion, the distribution of impurities, especially in samples HT1 and HT4, is such, that it does not represent the behaviour of a larger mass. I would like to invite the Authors to consider this, and discuss its consequences.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 275 | 94 | 15 | 384 | 4 | 9 |
- HTML: 275
- PDF: 94
- XML: 15
- Total: 384
- BibTeX: 4
- EndNote: 9
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1