Reliability-Based Design of Driven Piles Considering Setup Effects

Round 1

Reviewer 1 Report

This paper deals with a reliability-based approach on evaluating the setup effect on resistance capacity of driven piles. The presented introduction is insufficient.  Methodology, discussion and conclusions should be completely revised.  Accordingly, the following comments are offered.

Please remove some unnecessary sections as follows:

Line 10: “Featured Application: Authors are encouraged to provide a concise description of the specific 10

application or a potential application of the work. This section is not mandatory.”

Lines 324, 332, 340, and 345 should be completed removed from the manuscript and be replaced with the relevant descriptions.

Please clearly define the term “set up effect.” What are the main criteria to consider such an effect. The effect of soil profile, pile slenderness ratio, and the time history should be considered for discussing such an effect.

The abstract should be revised. A brief introduction of the work and the novelty should be clearly stated.

The presented introduction is pretty modest. Please add a comprehensive litterer review on pile setup techniques and application of different piles. It is recommended to add literature specifically from the published journals in the last five years.

Line 40: Please include a brief summary of the following references regarding the static load tests on micropiles and drilled shafts:

Effect of competent caliche layers on measuring the capacity of axially loaded drilled shafts using the Osterberg test." Applied Sciences 10.18 (2020): 6169.

CPT-Based Design Method for Axial Capacities of Drilled Shafts and Auger Cast-in-Place Piles." Journal of Geotechnical and Geoenvironmental Engineering 147.8 (2021): 04021077

Pile Base and Shaft Capacity under Various Types of Loading. Appl. Sci. 2021, 11, 3396

Effect of micropiles on clean sand liquefaction risk based on CPT and SPT." Applied Sciences 10.9 (2020): 3111.

Drilled Shaft Load Testing during Design." IFCEE 2021. 257-266.

The presented parametric analyses merely review the already available studies. A comprehensive analysis should be conducted for each of the analyzed factors. I recommend including examples, graphs and statistical analysis. A comparative analysis is also required to evaluate the effectiveness of the presented parameters.

Based on the following recommendation, a comprehensive discussion should be added to the manuscript. It is the main weak point of this work.

The conclusion should include a quantity-based approach. Merely indicating adjectives such as “significantly” “almost no effect” “slight influence” has no scientific value.

Author Response

Reviewer #1

This paper deals with a reliability-based approach on evaluating the setup effect on resistance capacity of driven piles. The presented introduction is insufficient.  Methodology, discussion and conclusions should be completely revised.  Accordingly, the following comments are offered.

Please remove some unnecessary sections as follows:

1. Lines 324, 332, 340, and 345 should be completed removed from the manuscript and be replaced with the relevant descriptions.

The authors have replaced with the relevant descriptions in the text. (See page 10, lines 336-341. in the revised manuscript)

2. Please clearly define the term “set up effect.” What are the main criteria to consider such an effect. The effect of soil profile, pile slenderness ratio, and the time history should be considered for discussing such an effect.

The authors have modified in the text. (See page 8,10. in the revised manuscript)

3. The abstract should be revised. A brief introduction of the work and the novelty should be clearly stated.

The authors have revised in the text. (See page 1, line12-20. in the revised manuscript)

4. The presented introduction is pretty modest. Please add a comprehensive litterer review on pile setup techniques and application of different piles. It is recommended to add literature specifically from the published journals in the last five years.

The authors have added literature from the published journals in the last five years. (See page 1,2, line34-45. in the revised manuscript)

5. Line 40: Please include a brief summary of the following references regarding the static load tests on micropiles and drilled shafts:

Effect of competent caliche layers on measuring the capacity of axially loaded drilled shafts using the Osterberg test." Applied Sciences 10.18 (2020): 6169.

CPT-Based Design Method for Axial Capacities of Drilled Shafts and Auger Cast-in-Place Piles." Journal of Geotechnical and Geoenvironmental Engineering 147.8 (2021): 04021077

Pile Base and Shaft Capacity under Various Types of Loading. Appl. Sci. 2021, 11, 3396

Effect of micropiles on clean sand liquefaction risk based on CPT and SPT." Applied Sciences 10.9 (2020): 3111.

Drilled Shaft Load Testing during Design." IFCEE 2021. 257-266.

The authors have summarized the above references in the text. (See page 2, line46-54. in the revised manuscript)

6. The presented parametric analyses merely review the already available studies. A comprehensive analysis should be conducted for each of the analyzed factors. I recommend including examples, graphs and statistical analysis. A comparative analysis is also required to evaluate the effectiveness of the presented parameters.

The authors have revised in the text. (See page 5-9. in the revised manuscript)

7. The conclusion should include a quantity-based approach. Merely indicating adjectives such as “significantly” “almost no effect” “slight influence” has no scientific value

The authors have revised in the text. (See page 9, line309-335. in the revised manuscript)

Author Response File: Author Response.pdf

Reviewer 2 Report

First of all, the work should be distinguished as a kind of deduction regarding the consequences of mathematical formulas in engineering practice in the field of geotechnics.

General comments:

1. Abstract – some errors. See them in the attachment.

2. Introduction

It is OK but should be much better. The problem is discussed on the background on old and modern literature positions. What I would suggest is to refer to the regulations as ISO (ISO 22477-4:2018) or EN (Eurocode) standards as well. Certainly it will raise the scientific level of the text and it will enhance the practical aspect of the work.

3. The text

To help with the language of the text I am passing my comments or underlined errors. So the levele of the language is pretty good. There are errors related mostly to missed “the” (Caret “^”), commas to be removed (underscore “_”).These errors are often accepted in scientific jargon but please maintain good standards. It will be also a kind of respect for the quality of the journal you are going to publish you paper.

4. Conclusions

Your “conclusions to be obtained..” are clearly pointed! Very good. But you have to add some summary of the work you made and please give general comments on practical application or expected effects. It will be appreciated some comments on your results in the context of the mentioned before standards or your country regulations on the discussed issue.
The title of the journal “Applied...” obliges you to expand the practical context.

Note: some suggestions and indications in "some correctios.pdf" file are not obligatory. 

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Dear Authors

I have to apologize for taking all of my time credits, but I needed to be sure that my primary impression was right.

I have some fundamental concerns about your study:

1. I checked references for inappropriate self-citations (which is a part of reviewing process) and I found that presented study is very similar to work [17]. Authors changed some symbols For example L (load) was replaced by Q. But in the text (line 97), it is still L.
2. I have some experience on reliability based design. I cannot understand, how can dead loads (gravity) and live loads be considered the same way? Considering dead load we may analyse its variability around the assumed one (self weight and log term storage weight). Considering live loads, we may only consider extreme values (appearing the same time with various probability). If I understood it well, you just treated both loads Q_D and Q_L the same way (The value of η generally is obtained as 1).
3. Generally, I have an impression that you do not have any experience, considering pile testing, In your paper you do not distinguish between physical phenomena and mathematical description. The whole analysis is based on week assumptions that form the basis for strong conclusions. For example: In my opinion, splitting the pile capacity into EOD (end of driving) capacity and setup capacity makes no sense. For design purposes, piles are usually tested (by means of SLT and DLT - static and dynamic load testing) after the time necessary for reconsolidation of soil along pile shaft. In my country, it is usually 7 days for piles in sand, 20 days for piles in saturated fine sand, and 30 days for piles in clay. That takes into account the setup effect reported by Skov [18] and Svinkin [20]. Any reliability analysis make sense only if we consider the final capacity and its variability. Any other ideas with separate evaluation of initial and final capacity make no sense when initial (EOD) value is usually unknown. 
4. I have an impression that you just transform equations and draw figures that are useless for design and control of foundation piles. Maybe I'm not fair but my opinion wouldn't be so tough if a single example of application were provided.
5. Presented study will never have any chances for application as building (geotechnical) practice will never provide data required for an analysis.  Authors gave examples with data assumed on the basis of references (slightly random). There is no way to use (comply) this methodology in the real life situation.
6. Results seem to be trivial and do not need a complicated analysis to be drawn. Concerning Fig. 1 - it is obvious that if you increase live load 4 times, FS will decrease significantly. Concerning Fig. 2. - higher target reliability index must impose higher Factor of Safety. Concerning Fig. 3. no matter what is M_setup, if only the final value counts it's no surprise that FS remains constant. Concerning Fig. 4. - I'd appreciate an example, some kind of physical description of negative correlation between R and R_setup. 

Last points make me opt for rejection or very serious revision of this paper (considering publication in Applied Sciences). In my eyes it does not have to do with Applications and scientific merit seems also dubious to me.  The work desperately needs a connection to "real world practice".

Best regards

Author Response

Dear Editor,

We would like to thank the editor and reviewers for their time and comprehensive comments. The manuscript has been carefully revised according to the review comments. The major revisions are marked in red in the revised manuscript. The point-by-point response to the review comments are listed below.

1. I checked references for inappropriate self-citations (which is a part of reviewing process) and I found that presented study is very similar to work [17]. Authors changed some symbols. For example L (load) was replaced by Q. But in the text (line 97), it is still L.

The author has modified some of the contents in the text. (See page 3, line 107-112. in the revised manuscript).  And, the work [17] in original manuscript and this paper are using LRFD method to calculate the resistance factor, so there are some formulas are similar.  But, the work [17] is obviously different from the study in this paper.  In the work [17] in original manuscript, the setup effects of driven piles is analyzed by dividing the ultimate resistance into the ultimate shaft resistance and ultimate base resistance.  However, this paper considers the setup effects of driven piles through the overall resistance and then designs the reliability of the pile.  The two are obviously different.

2. I have some experience on reliability based design. I cannot understand, how can dead loads (gravity) and live loads be considered the same way? Considering dead load we may analyse its variability around the assumed one (self weight and log term storage weight). Considering live loads, we may only consider extreme values (appearing the same time with various probability). If I understood it well, you just treated both loads Q_D and Q_L the same way (The value of η generally is obtained as 1)

In the specification AASHTO,  in which: For loads for which a maximum value of is appropriate: . For loads for which a minimum value of is appropriate: . In which, for the strength limit state:  for nonductile components and connections,  for conventional designs and details complying with these Specifications, for all other limit states:.  For the strength limit state:  for nonredundant members,  for conventional level of redundancy, for all other limit states:.  For the strength limit state:  for important bridges,  for typical bridge, for all other limit states:.  Therefore, the value of η is 1.0 in this paper.

3. Generally, I have an impression that you do not have any experience, considering pile testing, In your paper you do not distinguish between physical phenomena and mathematical description. The whole analysis is based on week assumptions that form the basis for strong conclusions. For example: In my opinion, splitting the pile capacity into EOD (end of driving) capacity and setup capacity makes no sense. For design purposes, piles are usually tested (by means of SLT and DLT - static and dynamic load testing) after the time necessary for reconsolidation of soil along pile shaft. In my country, it is usually 7 days for piles in sand, 20 days for piles in saturated fine sand, and 30 days for piles in clay. That takes into account the setup effect reported by Skov [18] and Svinkin [20]. Any reliability analysis make sense only if we consider the final capacity and its variability. Any other ideas with separate evaluation of initial and final capacity make no sense when initial (EOD) value is usually unknown.

This article first carries on the mathematical analysis, and describes its change, finally summarizes some physical phenomena.

Yang et al. establish a statistical database is developed to describe the increase in pile axial capacity with time, known as set-up, when piles are driven into clay. Based on the collected pile testing data, pile set-up is significant and continues to develop for a long time after pile installation.  So, I think splitting the pile capacity into EOD (end of driving) capacity and setup capacity makes sense.

Researchers, such as Svinkin et al. (1994), Axelsson (1998), McV ay et al. (1999), and Bullock et al. (2005), have proposed the use of t0= 1 day. Chow et al. (1998) reported values of A ranging from 0.25 to 0.75. Studies by Axelsson (1998) yielded values of A ranging from 0.2 to 0.8. The observations from this compiled database seem to conform with the previous experience. Therefore, the reference time t0= 1 day and factor A = 0.5 are used for the subsequent statistical analyses in this paper. A couple of cases show that pile capacity increases linearly with the logarithm of time until 100 days after driving. After an elapsed time of 100 days, the increase in pile capacity with time becomes less pronounced compared with that during the first 100 days after driving. The compiled database supports the general notion that 100 days after EOD can be taken as the time after which the set-up effect would be minimal.  In this paper.

In this paper, the bearing capacity of pile is divided into the initial bearing capacity and the setup bearing capacity, in order to better reflect the change of bearing capacity of piles in the process of driving, and evaluate the ultimate bearing capacity through the initial bearing capacity and the setup bearing capacity.  In this process, more consideration can be given to the uncertainty of driven piles.

4. I have an impression that you just transform equations and draw figures that are useless for design and control of foundation piles. Maybe I'm not fair but my opinion wouldn't be so tough if a single example of application were provided.

The author proposes a key parameter (M_setup) to modify the resistance factor and factor of safety calculation formula in accordance with the load and resistance factor design (LFRD) principle.  And the four key parameters are analyzed in detail and some useful conclusions are obtained (for example for ρ=2.0, the recommended resistance factor of driven piles in clay and sand are estimated as 0.26 and 0.20 respectively. the value of M_setup is determined by the type of soil, so it is concluded that the size of M_setup is adjusted by selecting different soil, and then the size of R_setup is controlled). Therefore, this paper can well evaluate the setup effects of driven piles and design the reliability of driven piles more effectively.

5. Presented study will never have any chances for application as building (geotechnical) practice will never provide data required for an analysis. Authors gave examples with data assumed on the basis of references (slightly random). There is no way to use (comply) this methodology in the real life situation.

In this paper, the reliability-based design of driven piles considering setup effects is studied in sand and clay, λ_QL, λ_QD, COV_QL and COV_QD have been given in AASHTO specification, and λ_R0, λ_Rsetup, COV_R0 and COV_Rsetup have also been reported in relevant literature, and these data are reasonable. This paper aims to provide a reference case when the actual project encounters the same or similar conditions.  Therefore, this paper uses these data to analyze the influence of four key parameters on the resistance coefficient and safety factor, and concludes that the resistance coefficient of β_T and ρ_R0,Rsetup has a great influence, so more attention should be paid to β_T and ρ_R0,Rsetup in practical engineering.

6. Results seem to be trivial and do not need a complicated analysis to be drawn. Concerning Fig. 1 - it is obvious that if you increase live load 4 times, FS will decrease significantly. Concerning Fig. 2. - higher target reliability index must impose higher Factor of Safety. Concerning Fig. 3. no matter what is M_setup, if only the final value counts it's no surprise that FS remains constant. Concerning Fig. 4. - I'd appreciate an example, some kind of physical description of negative correlation between R and R_setup

The authors have revised in the text. (See page 5-9, in the revised manuscript)

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

N/A

Reviewer 3 Report

Dear Authors.

You responded to my comments and introduced relevant changes to the manuscript. I understand your point of view and I respect the fact that you may make some assumptions based on literature review and suggestions taken for existing codes of practice.

I still believe that splitting pile capacity into the "End of Driving" and "setup" arts is slightly artificial in reliability based analysis but it is allowable. 

I noticed that other reviewers were in favour to publish this manuscript after necessary corrections and I also tend to accept it in current form for publication.

Best regards