Concrete Structures for RLG Containment

1

Committee 376

Concrete Structures for RLG Containment

Agenda ACI 376-C Analysis Sub-Committee Meeting,

Sunday, March 23, 2014, 3 PM – 5 PM

Grand Sierra Resort, Reno, NV

Room: N-12

1. CALL TO ORDER and APPROVAL OF AGENDA

2. INTRODUCTIONS

3. ANNOUNCEMENTS

4. DISCUSSION: TN on Design and Analysis of the TCP Embedment Zone Resolution of remaining negative votes and comments.

5. OTHER BUSINESS / INFORMAL DISCUSSION OF PROJECTS

6. ADJOURNMENT

ACI 376 - Technical Note on TCP 2nd ACI 376-C Ballot on TCP Technical Note Last Update: 2 / 14 / 2014 - 376-D webinar Balloted from 1 /2 0 to 3 / 10 / 2013 2012

Page 1 of 16

Technical Note on TCP (TCP – TN) Approved Sections Section Approved with Comments to be resolved Negative Vote 2013.04.13 376-C Minneapolis 2014.02.14 376-Webinar

Members that Voted Voting Members that did not vote 2013.04.13 – 376 D Minneapolis 2014.02.14 – 376 D Webinar

Voting Members that voted (79%) 1. Hjorteset, Kare 2. Brannan, Mike S 3. Allen, Junius 4. Ballard, Thomas A 5. Garrison,Jeffrey 6. Hatfield,Alan 7. Hoff, George C 8. Hoffmann, Richard A

9. Hoptay,Joseph 10. Howe, Thomas R 11. Jiang,Dajiu 12. Khalifa, Jameel U 13. Krstulovic-Opara, Neven 14. Legatos, Nicholas A 15. Malhotra,Praveen 16. Mash,Keith 17. Moncarz, Piotr D

18. Nussmeier, Robert W 19. Oliver, William H 20. Pawski, Rolf P 21. Rushing, William E 22. Wu,Sheng-Chi

Associate Members

Roetzer, Josef Widianto, Widianto

23. Hanskat, Charles S 24. Douglas,Hamish 25. Hashmi, Humayun 26. Rajan, Ramanujam S 27. Sward,Robert 28. Thompson,Eric

Members Attending: 1. Brannan, Mike 2. Hjorteset, Kare 3. Ballard, Thomas 4. Garrison. Jeffrey 5. Hoff, George 6. Hoptay, Joseph 7. Howe, Thomas 8. Khalifa, Jameel 9. Pawski, Rolf 10. Hanskat, Charles 11. Roetzer, Josef

Participating: 1. Allen, Junius 2. Brannan, Mike 3. Douglas, Hamish 4. Garrison. Jeffrey 5. Hjorteset, Kare 6. Hoptay, Joseph 7. Howe, Thomas 8. Jiang,Dajiu 9. Legatos, Nick 10. Krstulovic-Opara, Neven 11. Mash, Keith 12. Mehta, Sanjay 13. Oliver, Billy 14. Pawski, Rolf 15. Roetzer, Josef 16. Vossoughi, Fariborz

1st Ballot Results (Approved w. Comment) 2nd Ballot Results

Ref No.

Text Author Comments 2ND BALLOT:

RESPONSE being BALLOTED

Voting Member’s VOTE &

COMMENTS

Suggested Change

1

Brannan

I am not sure the TCP can be constructed as detailed because it would be impossible to weld both sides of the TCP plate to the embedment plate. It would also be better to radius the sharp corner shown on the TCP plate. Would it be possible to simply lap weld the TCP plate to the embedment plate? Should the embedment zone also extend 2X the wall thickness below the embedment plate?

The comment found partially persuasive. It is not intended that the TN covers TCP construction details. Furthermore, welding requirements should be as per API 620. Address the comment by revising text on page 1 lines 11-14: INTRODUCTION ACI 376 provides requirements and some basic recommendations for the design of the Thermal Corner Protection (TCP) embedment zone. The purpose of this Technical Note document is to provide further guidance for on (a) design and analysis of the TCP embedment anchorage detailing and (b) TCP liquid-tightness analysis (i.e., crack width calculations) in the TCP embedment zone (i.e., crack width calculations).

Agree – Jiang, Moncarz, Hoff, Khalifa, Rushing, Humayum, Howe, Wu, Malhotra, Garrison, Mash, Douglas, Allen, Legatos, Widianto, Hoptay, Roetzer, Oliver, Pawski, Hatfield, Brannan, Krstulovic

Agree with comment - Hoffman

With Respect to M. Brannan’s Comments, it is my opinion this can be welded from both sides, by making this attachment weld first and then going on with the construction. There will be a “fill-in plate that is required to build up the TCP side wall.

UPDATE NOTES (2013.01.08) a) 2013.10.21 Phoenix discussion items are on pg 14 of 15. b) See the following Ref. no. for NEGATIVES REQUIRING FURTHER DISCUSSION: 1, 18, 32, 37, 41, 44 & 47, 49, 63, 74. c) Pg. 16 of 16 has Legatos negative that requires further discussion. WEBINAR NOTES (2014.02.14) ~ no quorum a) Resolved Ref.No. 1, 32, 41, 44 & 47. Ref. No. 63 changed to editorial; RP to propose moving text. b) Ref. No. 18 and 49 to be discussed further in Reno. Ref. No. 37 and 74 need further discussion – missing notes of resolution.


Page 2 of 16


Ref No.




COMMENTS

Suggested Change

Update title to read: Thermal Corner Protection Design & Crack Analysis Design and Analysis of the TCP Embedment Zone

I have no objection to the proposed changes

Disagree – Ballard

I believe the embedment zone should extend at least to the bottom of the TCP plate since migration of cracks in this zone would compromise the leak tightness of the detail.

2013.04.13 – 376-C Minneapolis: Agreed and that Tom Ballard to propose wording. From TB email dated 2013.04.15: “Although the code and this TN do not explicitly define leak tightness criteria for the anchorage and embeddment, the design and construction of these components should be carried out with the clear objective of protecting the corner joint from product temperatures.” 2014.02.14 Webinar Ballard wording ok

8 Page 5, line 2, item 3

Mash

Suggest rewording and removing 'substantially' link to definitive acceptance criterion to avoid confusion. Suggest exact acceptance criterion words such as - under the spillage condition the crack widths on the inside face shall be limited to xx, compressive stresses in both the meridional and vertical directions shall be minimum y, max z,

The comment found partially persuasive. Limiting values are covered in the Code (e.g., compressive zone size, crack width limits, etc.) and do not need to be repeated here. Revise text for clarity to read: “The concrete wall above the TCP, directly exposed to spilled product, should remain liquid and substantially vapor tight during the spill condition. This is assured by maintaining a compression zone in the wall exposed to the spill and by and limiting the crack width cracking on the inside face of the wall in the vicinity of the TCP embedment plate zone during a spill.”

Agree – Jiang, Moncarz, Hoff, Rushing, Hoffman, Humayum, Howe, Wu, Ballard, Malhotra, Garrison, Mash, Douglas, Allen, Legatos, Widianto, Hoptay, Oliver, Pawski, Hatfield, Brannan, Krstulovic

Agree with editorial changes – Khalifa, Roetzer, Krstulovic

“The concrete wall above the TCP, directly exposed to spilled product, should remain liquid tight and substantially vapor tight during the spill condition. 2013.04.13 – 376-C Minneapolis: Agree with comment.

10 Page 5 Line 10

Hoptay

The material selection is temperature dependent and to state that the TCP is 9% Ni infers an LNG tank and not RLG tanks in general.

The comment found persuasive. Change text to read: “Material selection and material requirements used in the

Agree – Jiang, Moncarz, Hoff, Khalifa, Rushing, Hoffman, Humayum,


Page 3 of 16


Ref No.




COMMENTS

Suggested Change

design of the 9% nickel-steel TCP and secondary bottom plates may be performed in accordance with API 620.”

Howe, Wu, Ballard, Malhotra, Garrison, Mash, Allen, Legatos, Widianto, Hoptay, Roetzer, Oliver, Pawski, Hatfield, Brannan, Krstulovic Disagree - Douglas This may still be ambiguous as API 620 does not refer to

TCP

2013.04.13 – 376-C Minneapolis: Change last line to read: “… should be in accordance with API 620.”


Mash

Assurance is through cracks + compression zones not through cracks alone. Crack width limitation in the vicinity of the embed is to prevent liquid migration behind the TCP area.

No action. Covered in previous response.

Agree – Jiang, Moncarz, Hoff, Khalifa, Rushing, Hoffman, Humayum, Howe, Wu, Ballard, Malhotra, Garrison, Mash, Douglas, Allen, Legatos, Widianto, Hoptay, Roetzer, Oliver, Pawski, Hatfield, Brannan, Krstulovic

13 P. 5, Line 15

Wu

add “seismic Aftershock load”. The comment found persuasive. Change text to read:

“The TCP embedment plate should be designed to resist radial and vertical loads and moments resulting from both (a) thermal gradients, as well as (b) mechanical forces that develop during a spill condition including SSEAFT. The TCP design consists of the following steps:”



Mash

Starting temperatures and conditions should be selected so as to maximise the demand under consideration. For instance for the design of the anchors the critical condition is likely to be warm wall (summer) and cold shocking to the embed. Whereby creating maximum straining across the section.

The comment found persuasive. Add proposed text to line 22: 1.2.1 – STEP 1: TCP Load Definition Finite element analyses is performed first to determine forces resulting from thermal gradients applied to the TCP, TCP embedment plate and outer concrete wall. Loads should include all loads that the TCP and the embedment will be subjected to including thermal, prestress, creep, concrete shrinkage, internal

Agree – Jiang, Moncarz, Khalifa, Rushing, Hoffman, Howe, Wu, Ballard, Malhotra, Garrison, Mash, Legatos, Widianto, Hoptay, Roetzer, Hatfield, Krstulovic

2013.04.13 – 376-C Minneapolis: To address all comments, replace last three sentences with: “Initial temperatures and conditions should be selected to obtain the most severe design conditions. For instance the most severe condition for design of the anchors is likely to be a warm wall (summer) with a cold thermal shock to the embedment.”


Page 4 of 16


Ref No.




COMMENTS

Suggested Change

pressure, hydrostatic pressure and hydrodynamic pressure due to seismic aftershock. Starting temperatures and conditions should be selected so as to maximize the demand under consideration. For instance for the design of the anchors the critical condition is likely to be warm wall (summer) and cold shocking to the embed. Whereby creating maximum straining across the section.

Agree with editorial changes - Hoff

For instance, for the design of the anchors, the critical condition is likely to be a warm wall (summer) and with a cold shocking to the embedment, thereby creating maximum straining across the section.

Agree with minor comment - Humayum

last sentence - straining or strain?

Agree with minor comment - Brannan

consider using “Design” instead of “Starting”.

Agree with editorial changes - Pawski

Starting temperatures and conditions should be selected so as to maximize the demand under consideration. For instance for the design of the anchors the critical condition is likely to be warm wall (summer) and cold shocking thermal shock to the embed. Whereby creating maximum straining across the section.

Agree with editorial changes - Douglas

Pawski 2013.04.23 – Allen comment below shown incorrectly as Douglas and deleted here. Correct comment is: "…....... cold shock to the embed which results in the maximum strain across the section " Douglas additional comment1 at end of Ballot Summary: "1. , Para 1.2.1 Line 24: Also Comment LE27: What is the advantage of the use of the term "demand" as opposed to load? If a new term is to be introduced then it's use must be defined: When is a load a demand?" Pawski 2013.04.23 – this is addressed by 376-C Minneapolis response above.

Agree with editorial changes - Allen

For instance for the design of the anchors the critical condition is likely to be warm wall (summer) and cold shocking to the embed. Whereby Thereby creating maximum straining across the section.

Agree with changes - Oliver

Agree with suggested grammatical revision: Starting temperatures and conditions should be selected so as to maximize the resulting stresses in the component under consideration. For example for the design of the anchors, the critical condition is likely to be warm wall (summer) and cold shocking to the embed, creating maximum straining across the section. Comment: What does demand mean in this context? If not revised, add “demand” to definitions.

16 P. 5, Line 22, at the end of the paragraph

Wu

Add “and hydrodynamic pressure due to seismic aftershock”.

The comment found persuasive. Change text to read:

“Finite element analyses is performed first to determine forces resulting from thermal gradients applied to the TCP, TCP embedment plate and outer concrete wall. Loads should include all loads that the TCP and the embedment

Agree – Jiang, Moncarz, Hoff, Rushing, Hoffman, Humayum, Howe, Wu, Ballard, Malhotra, Garrison, Mash, Allen, Legatos,


Page 5 of 16


Ref No.




COMMENTS

Suggested Change

will be subjected to including thermal, prestress, creep, concrete shrinkage, internal pressure, hydrostatic pressure and hydrodynamic pressure due to seismic aftershock (SSEAFT).”

Widianto, Hoptay, Oliver, Pawski, Hatfield, Brannan, Krstulovic

Agree with editorial change – Khalifa, Douglas, Roetzer, Krstulovic

“Finite element analyses is are performed first to determine forces resulting from thermal gradients applied to the TCP, TCP embedment plate and outer concrete wall.


Mash

Suggest inclusion of the words 'in the absence of a defined leak rate all intermediate levels shall be investigated. In practice this may be achieved by running a number spill levels to so as adequately bound the demands. Typically 5 levels may be used dividing the spill height into 4 equal units. In addition levels in close proximity to the TCP must be investigated.

Agree in principle with comment. Revise as noted and address spill levels in a future edition of the Code.

“Various spill levels should be considered to determine the most severe effect on the embedment and concrete wall. This should include a spill levels located just below the top of the TCP which will place the largest temperature differential between the TCP and the concrete wall. as well as various Also a minimum of three levels should be considered for spill levels above the TCP up to, at top of the TCP embedment zone, at mid-height, and full spill height. “

Agree – Jiang, Moncarz, Rushing, Hoffman, Humayum, Howe, Wu, Ballard, Malhotra, Garrison, Allen, Legatos, Widianto, Roetzer, Hatfield, Brannan, Krstulovic

2013.04.13 – 376-C Minneapolis: Mash comment non-persuasive. Three is a minimum and does not preclude specifier from requiring or the designer choosing more.

Agree with suggested change -Khalifa

This should include a spill levels located just below the top of the TCP which will place generate the largest temperature differential between the TCP and the concrete wall.

Agree with editorial changes – Hoff, Pawski

This should include a spill levels located just below the top of the TCP which that will place the largest temperature differential between the TCP and the concrete wall Also A minimum of three levels should be considered for spill levels above the TCP up to,1) at top of the TCP embedment zone, 2) at mid-height, and 3) at full spill height.


This should include a spill levels located just below the top of the TCP which will place result in the largest temperature differential between the TCP and the concrete wall

Agree with editorial changes - Hoptay

“…at the top of the TCP embedment zone, at mid-height, and full spill height”

Agree with editorial comment - Oliver

Change “place” to “produce”. Observation: The proposed text requires four levels, not so far from five.

Disagree - Mash Statement of 3 levels does not reflect the significance of the analysis. This is the most important analysis carried out for the outer tank and it is essential that an appropriate amount of levels be considered. Suggest that the text is amended to the following: Also a minimum of three five levels should be considered 2014.02.14 Webinar Discuss in Reno the issue of safety. Mash insists 5 minimum is a good starting point and contractor must demonstrate they have identified critical


Page 6 of 16


Ref No.




COMMENTS

Suggested Change

condition.

20 Page 5, line32, item 14

Mash

Must be achieved through mechanical anchoring! The comment found persuasive. Revise to read: “The TCP embedment plate should be cast in place and mechanically anchored to the concrete wall or by other suitable means (e.g., ???). Anchor bolts (studs) should be placed in continuous (circumferential) rows along the concrete wall.”

Agree – Jiang, Moncarz, Hoff, Rushing, Hoffman, Humayum, Howe, Wu, Ballard, Malhotra, Mash, Douglas, Allen, Legatos, Widianto, Oliver, Pawski, Hatfield, Brannan, Krstulovic

Agree with suggested change -Khalifa, Roetzer

Delete second sentence completely as the same is essentially said in the next sentence. “The TCP embedment plate should be cast in place and mechanically anchored to the concrete wall. Anchor studs should be placed in continuous (circumferential) rows along the concrete wall.” 2013.04.13 – 376-C Minneapolis: After discussion it was agreed delete second sentence as suggested.

Agree with suggested change -Garrison

Modify last sentence: “Anchor studs should be welded in continuous circumferential rows along the embed plate.”

2013.04.13 – 376-C Minneapolis: Withdrawn after discussion on deleting second sentence as per Khalifa, Roetzer, and Hoptay.

Disagree -Hoptay Delete second sentence because next sentence covers the same information

2013.04.13 – 376-C Minneapolis: After discussion it was agreed delete second sentence as suggested.


Mash

When? Embedment must be anchored with studs? The comment found persuasive. Revise to read as noted below. Address questions on spacing and gap size in future: “When the The embedment plate isshould be anchored with

circumferential studs, a minimum of two rows of circumferential studs should be used to anchor the embedment plate. The spacing between mechanical anchors (studs) has significant impact on the TCP emebedment

Agree – Jiang, Moncarz, Hoff, Rushing, Hoffman, Humayum, Howe, Wu, Ballard, Malhotra, Garrison, Mash, Douglas, Allen, Legatos, Widianto,


Page 7 of 16


Ref No.




COMMENTS

Suggested Change

performance and thus should be considered in the design. should be limited to ensure that the axial, bending and shear stresses in the embedment plate are within the allowable stresses.”

Hoptay, Roetzer, Oliver, Hatfield, Krstulovic

Agree with comments -Pawski

Bullet points 1 & 2 should be combined Use “studs” or “bolt studs” consistently

Agree - Brannan Note that our Figure 1.1a shows four rows of circumferential studs. Should it show two rows of studs in solid line ink and the other two in shaded or dashed ink to indicate the potential for using more than two rows? 2013.04.13 – 376-C Minneapolis: Agree, annotate sketch to state “2-rows minimum.”.

Agree with editorial changes - Khalifa

Also recommend deleting bolts from first line under section 1.2.3

2013.04.13 – 376-C Minneapolis: “Anchors” to be used throughout. Edit TN and replace bolts & studs with “anchors.”

Disagree - Hoptay Disagree – studs are not always used for consistency suggest the following rewording The embedment plate should be anchored with a minimum of two circumferential rows of anchors. The spacing between anchors has significant impact on the TCP embedment performance and thus should be considered in the design. 2013.04.13 – 376-C Minneapolis: Persuasive, use proposed wording.


Mash

Nothing mentioned regarding detailing? Suggest inclusion of words such as. Ideally crack control is best achieved through the introduction of smaller bars at closer spacings as opposed to larger bars at wider spacings. In this regard it is suggested that the bar spacing be selected so as to enable the subsequent introduction of lacer bars to the system. Therefore adoption of bar spacings of between 150 and 200mm is recommended in the vicinity of the TCP whereby allowing additional bars top be slotted in. Anchor spacings should be selected to as to be multiples of the bar spacing to enable a clash free design.

The comment found non-persuasive regarding suggested inclusion of bar detailing because this is normal everyday concrete detailing. Add the following after line 7 on page 6. “Anchor spacings should be selected to as to be multiples

of the bar spacing to enable a clash free design. “ Delete lines 6 and 7 on page 6. “Allowable stresses i.e., material selection and material

requirements used in the design of the embedment plates, may be selected in accordance with API 620.”

Agree – Jiang, Moncarz, Khalifa, Rushing, Hoffman, Humayum, Howe, Wu, Ballard, Hoff, Hoptay, Malhotra, Garrison, Mash, Douglas, Allen, Legatos, Widianto, Roetzer, Oliver, Pawski, Hatfield, Brannan, Krstulovic


Page 8 of 16


Ref No.




COMMENTS

Suggested Change

NEW Page 6 Para 1.2.4

This comment inserted by Pawski 2013.04.23 Douglas additional comment 2 at end of Ballot Summary: The liquid tightness criteria are specified in Para 2.1, items 1 & 2. Additional tightness criteria on “embedment plates” should not require to be defined by owner. I am not in favour of definitions to be provided by owners. The statement is vague and ultimately provides no guidance.The code is to provide guidance to all, including owners.

Douglas

28 Page 7, line 1 4, item 1

Mash

Suggest using nomenclature such as demand as opposed to load.

The comment found persuasive in part. Revise to read. “When subjected to During the spill condition, the concrete wall will be loaded with subjected to high forces due to mechanical loads and thermal effects and mechanical forces. These forces that will lead to significant wall cracking. During the spill condition, integrity Integrity of the concrete wall must be maintained. As specified in the ACI 376 Code, the wall should remain liquid and substantially vapor tight. More specifically:”

Agree – Jiang, Moncarz, Khalifa, Rushing, Hoffman, Humayum, Howe, Wu, Ballard, Malhotra, Garrison, Mash, Douglas, Allen, Legatos, Widianto, Roetzer, Oliver, Pawski, Hatfield, Brannan, Krstulovic

Agree with editorial changes - Hoff, Krstulovic

During the spill condition, the concrete wall will be subjected to high forces due to mechanical loads and thermal effects that will can lead to significant wall cracking. 2013.04.13 – 376-C Minneapolis: Discussed and agree with comment.

Disagree - Hoptay Suggest rewording sentence as follows to cover both the transient and steady state condition. The integrity of the concrete wall must be maintained during the entire spill. 2013.04.13 – 376-C Minneapolis: Withdrawn.


Mash

EN only calculates characteristic widths. Historically US codes have cocooned the crack width requirements into the detailing rules, These were based on the Gergely Lutz expressions which used mean widths. Replace 'as per' with 'in accordance with'

The comment found persuasive. Revise page 7 lines 20-21 to read. “3) ACI 376 section 8.1.1.8 - requires that the cracks widths should be calculated as characteristic and not mean calculated crack widths. as per EN 1992-1-1.”

Agree – Jiang, Moncarz, Hoff, Khalifa, Rushing, Hoffman, Humayum, Howe, Ballard, Malhotra, Douglas, Allen, Legatos, Widianto, Hoptay, Roetzer, Oliver, Pawski, Hatfield, Brannan, Krstulovic

2013.04.13 – 376-C Minneapolis: Agree with Wu. NEW BUSINESS for CODE - JG write the new business item on code change form Q1.

Disagree - Wu Suggest that: a) EN 1992-1-1 should stay, as stated in ACI 376 to delete “as characteristic and not mean”


Page 9 of 16


Ref No.




COMMENTS

Suggested Change

2013.04.13 – 376-C Minneapolis: Persuasive since crack widths calculated by EN 1992-1-1 will be characteristic. Proposed wording to be adopted.

Disagree - Garrison b) Committee needs to review the ability of the contractor / designer to meet the requirement of 0.008” max crack width using EN equations for the spill condition (2 cryogenic temperatures). Experience shows that this is a difficult requirement to meet. Instead, perhaps ACI 376 should specify minimum rebar size and spacing within the TCP embedment zone.

2013.04.13 – 376-C Minneapolis: Withdrawn pending future action by main committee. Jeff Garrison to prepare New Business (code change form Q1) proposal to address crack width values and method of calculation.

Disagree - Mash By removing the design code there is potential for use of codes that do not generate characteristic levels. In particular Model Code 90 is not prohibited, also Ven der Veen’s approach generates mean expectation of crack widths. There are some Contractors still purporting to MC90 and reviewer has a concern this will continue if left unchecked. Without guidance the Client is without a Captain and I uninformed. Note the NL Analysis is in accordance with EN so there should be no objection to using EN for crack widths. 2014.02.14 Webinar Mash ok as long as Charcateristic is the basis. Discuss Wu comment in Reno concerning reference to EN 1992-1-1. Webinar consensus was to include reference to EN 1992-1-1 for clarity.

34 Section 2.2.1 Page 7, Line # 31

1) Modeling

Jiang

Suggest to change to: 1) Modeling including modeling geometry and

material thermal properties

The comment found persuasive. Revise page 7 lines 20-21 to read.

1) “Modeling including specific tank geometry and material thermal properties”



Page 10 of 16


Ref No.




COMMENTS

Suggested Change

37 Section 2.2.1 Page 8, Line # 2 4) Continue analysis until steady state final temperatures are reached

Jiang

Suggest to change to: 4) Continue analysis until steady state final temperatures are reached and the temperature distribution is obtained

The comment found persuasive. Revise page 7 lines 20-21 to read. 4) “Continue analysis until steady state final temperatures are reached temperature distribution is obtained”

Agree – Jiang, Moncarz, Hoff, Khalifa, Rushing, Hoffman, Humayum, Howe, Wu, Ballard, Malhotra, Mash, Douglas, Allen, Legatos, Widianto, Hoptay, Roetzer, Oliver, Pawski, Hatfield, Brannan, Krstulovic

Disagree - Garrison Committee needs to define analysis / design requirements. Considering the variability in the possible loading conditions, variability in material properties and variability in modeling techniques it seems that a steady state analysis would be sufficient. 2013.04.13 – 376-C Minneapolis: Withdrawn. 2014.02.14 Webinar ??????

41 Section 2.2.2 Page 8, Line # 8 “Apply dead load, prestressing and any other steady state load, iterate to equilibrium”

Jiang

Suggest to change to: “Apply dead, prestressing and any other steady state load”

The comment found persuasive. Revise page 7 lines 20-21 to read. “To establish initial conditions, aApply dead load, prestressing, normal-operating thermal and any other steady state load, then iterate to equilibrium”


Disagree - Garrison Committee needs to define analysis / design requirements. 2014.02.14 Webinar Mash delete "then iterate to equilibrium" is not losing anything Pg 8 line 11 & 13 delete as suggested.


Mash

Determine strains in section, check for Residual Compressive Zone (RCZ), check for max comp stress etc.

The comment found persuasive. Change text to read: 5) “Beginning with the initial condition defined in step 2, cCompute crack widths at time steps times throughout the transient where reinforcing strains are the greatest. and exceed the material yield stress Furthermore, at each time step determine the adequacy of the compression zone.”

Agree – Jiang, Moncarz, Hoff, Khalifa, Rushing, Hoffman, Humayum, Howe, Wu, Ballard, Malhotra, Mash, Douglas, Allen, Legatos, Widianto, Hoptay, Roetzer, Oliver, Pawski, Hatfield, Brannan,


Page 11 of 16


Ref No.




COMMENTS

Suggested Change

Krstulovic

Disagree - Garrison Considering the variability in the possible loading conditions, variability in material properties and variability in modeling techniques it seems that a steady state analysis would be sufficient.

47 Section 2.2.2 Page 8, Line # 11 5) Computer crack widths at times throughout the transient where reinforcing strains are the greatest and exceed the material yield stress

Jiang

Suggest to change to: 6) Computer crack widths at times throughout the transient where reinforcing strains are the greatest and the stresses exceed the material yield stress

2014.02.14 Webina straw poll Is Garrison negative persuasive? Yes - 0 No - 9 Abstain - 1


Mash

What does this mean? The comment found persuasive. Change Page 8, lines 17-18 to read: “Calculation of cracking and crack widths for the RLG LNG spill condition and/or external fire condition shall be performed using a rational engineering analysis. Both linear Linear or non-linear finite element (FE) analysis should be used as discussed in 2.3.1 and 2.3.2.” 2014.02.14 Webinar Mash withdrawn Ballard comment : NKO comment is not relevant & discuss with Ballard in Reno. As written spill condition does not preclude spill + fire as per referenced code. Canadian code.

Agree – Jiang, Moncarz, Hoff, Khalifa, Rushing, Hoffman, Humayum, Howe, Wu, Malhotra, Garrison, Douglas, Allen, Legatos, Widianto, Hoptay, Roetzer, Oliver, Pawski, Hatfield, Brannan, Krstulovic

Disagree - Ballard CSA Z276 requires that fire and spill be considered concurrently. I believe that this needs to be addressed in this paragraph.

Disagree - Mash Rational to one uninformed Engineer could be irrational to informed Engineer. Rational requires defining. 2014.02.14 Webinar straw poll Mash withdrawn NKO regarding Ballard comment: It is not relevant & discuss with Ballard in Reno. As written spill condition does not preclude spill + fire as per referenced code, Canadian CSA Z276.

50 Section 2.3.1 Page 8, Lines # 21 to 23 “In this case a non-linear finite element program is used in conjunction with

Non-linear FE material constitutive models for

Jiang

Suggest to change to: “In this case a non-linear finite element program is used by:

Implementing the non-linear material constitutive relationship for concrete and steel

Carrying out FE analysis to obtain stress counters, and

Performing FE analysis post process to calculate crack width as per Eurocode 2 1992-1-1

The comment found persuasive. Change text to read: “In this case, a non-linear finite element program is used in conjunction with by:

Non-linear FE material constitutive models Implementing the non-linear material constitutive relationship for concrete and steel and

Carrying out FE analysis to obtain stress/strain contours, and

Agree – Jiang, Moncarz, Hoff, Rushing, Hoffman, Humayum, Howe, Wu, Malhotra, Garrison, Mash, Douglas, Allen, Legatos, Hoptay, Roetzer, Oliver, Pawski, Hatfield, Brannan, Krstulovic


Page 12 of 16


Ref No.




COMMENTS

Suggested Change

concrete and steel, and

Crack width calculations as per E

Performing FE analysis post-process to calculate characteristic crack width, calculations following a method, such as as per Eurocode 2 EN 1992-1-1 (see section 2.5 in this TN).”

Agree with minor changes - Khalifa

Performing FE analysis post-processing to calculate characteristic crack width, following a method, such as Eurocode 2 EN 1992-1-1 (see section 2.5 in this TN).”

Agree with minor changes - Ballard

Change word “contours” to “distribution” or change sentence to “Carrying out FE analysis to obtain stresses and strains.”

Agree with changes - Widianto

To be consistent with Sections 2.3.1 and 2.3.2, suggest the sentence is re-worded as follows: "Non-linear or simplified finite element (FE) analysis should be used as discussed in 2.3.1 and 2.3.2" To be consistent with the last sentence in Section 2.3, I suggest that we switch the order between the current Sections 2.3.1 and 2.3.2 (i.e., present Simplified FE analysis in Section 2.3.1 and present non-linear FE analysis in Section 2.3.2. This will also be consistent with the progression of the actual design, where Simplified FE analysis is done in the preliminary design, before non-linear FE analysis.

53 Section 2.3.2 Page 9, Lines # 2 & 3 “However, it should be noted that this is a very time consuming process and is thus usually used in the preliminary design.”

Jiang

Suggest to change to: “However, it should be noted that this is a very time consuming process and is thus usually used for focusing on more critical load cases after a preliminary screen for load cases has been performed in the preliminary design.”

The comment found persuasive. Remove the whole sentence. “However, it should be noted that this is a very time consuming process and is thus usually used for focusing on more critical load cases after a preliminary screen for load cases has been performed in the preliminary design.”



Page 13 of 16


Ref No.




COMMENTS

Suggested Change

56 Section 2.3.2 Page 9, Lines # 4 & 5 “Moment-Curvature Approach: The moment curvature response is obtained using a concrete section analysis program in conjunction with:”

Jiang

Suggest to change to: “Moment-Curvature Approach: The moment curvature response is obtained using a concrete section analysis program by implementing:” Comment: Weather should this approach be considered as “Linear Analysis” or “Non-linear Analysis”? May be this approach could be considered as a simplified non-linear analysis approach and should be used for preliminary only.

The comment found persuasive. Change text to read: “2.3.2 Simplified Linear FE Analysis Simplified Linear FE analysis is usually used only in the preliminary design. …… “Moment-Curvature Approach: The moment curvature properties are response is obtained using a concrete section analysis program by implementing:” non-linear stress-strain material properties, and a non-linear shell formulation that accounts for the cracking

induced stiffness redistribution. This moment-curvature approach would only be suitable for

consideration of the steady state linear thermal gradient case and should also be used only for preliminary design.”



Mash

Seems out of place in terms of the flow of the document. 2014.02.14 Webinar Oliver withdraw negative to editorial RP review if moving to another location

The comment found persuasive. Edit text as shown below: “2.3.4 Loading The RLG spill case should …… Nodal temperature transients will typically be computed in a separate heat transfer finite element or finite difference solution. The thermal model used for determining nodal temperatures This thermal model is also a high density model which should include concrete foundation, soil thermal boundary condition, walls, roof and outside or inside ambient temperature boundary conditions. This model should also include the base slab insulation and the thermal corner protection insulation. “

Agree – Jiang, Moncarz, Hoff, Khalifa, Rushing, Hoffman, Humayum, Howe, Wu, Ballard, Malhotra, Garrison, Mash, Douglas, Allen, Legatos, Widianto, Hoptay, Roetzer, Pawski, Hatfield, Brannan, Krstulovic

Disagree - Oliver This response does not address the basic comment that this whole paragraph doesn’t below under “Loading”. Suggest moving it to Section 2.3.3 “Mesh Type and Size”, or inserting a new section before “Loading” called “Thermal Model”. 2014.02.14 Webinar straw poll Oliver negative withdrawn. Consider editorial. RP to review if another location is more appropriate considering concern.

66 A Page 10 Line 10

Hoptay

On the code side of ACI 376 ASTM A706 is not referred to as low temperature reinforcing, only in the commentary. Can it be referred to as low temperature reinforcing in the technical note?

The comment found persuasive. Change text to read: “The concrete wall internal temperatures from the thermal model are also typically used for selection and to determine the placement of normal (A615), cold-temperature (A706) and cryogenic reinforcing complying with the requirements of ACI 376.” Furthermore, in a future Coed revision, revisit the use of A706 materials in ACI 376 section 4.7.2.



Page 14 of 16


Ref No.




COMMENTS

Suggested Change

74

Garrison

1) Code and TN need to define conditions for transient thermal analysis for the spill condition (leak rate).

2) Why does TN reference withdrawn version of Eurocode

2 (DD ENV 1992-1-1:1992)? Current version is BS EN 1992-1-1:2004.

The comment found partially persuasive. 1) Definition of the spill loading is beyond the scope of the current TN

and will thus be addressed in a separate document defining LNG spill conditions.

2) Update and change the model to the latest 2004 edition of the Eurocode,

Agree – Jiang, Moncarz, Hoff, Khalifa, Rushing, Hoffman, Humayum, Howe, Wu, Malhotra, Mash, Douglas, Allen, Legatos, Widianto, Hoptay, Roetzer, Oliver, Pawski, Hatfield, Brannan, Krstulovic

Agree with comment - Ballard

Garrison was going to forward a copy of EN 1992-1-1:2004 to Ballard for updating the code equations in the TN.

???? - Garrison Agreed, loading and consideration of need for transient analyses needs to be reviewed. 2014.02.14 Webinar ??????

75 Appendix A Pages 15-16

Pawski

It is redundant because it repeats paragraphs 2 through 4 in section 2.1 (page 7, lines 9-21)

The comment found non-persuasive. The comment found to be an editorial comment. Appendix A is a “copy” of what is in the Code. It is provided for reader’s convenience.


76 Page 16, last paragraph in section R8.1.1.7

Thompson

Remove the last sentence from the Code Commentary: “Due to the non-homogeneous nature of concrete, calculated crack width values measured in the field will vary from calculated values, and therefore they cannot be directly compared to calculated crack widths.”

Section R8.1.1.7 is text copied over from the Code and provided herein for the user’s benefit. Text must remain as-is to properly reflect the Code requirement. Changes to be addressed only if/once the Code has been changed accordingly.



Page 15 of 16

ADDITIONAL COMMENTS included with the ballots

Ref No.




COMMENTS Page 4, line 8; Section 1.1

Oliver EDITORIAL: Add “the” “The primary function of TCP is to protect the already highly…”

Page 6, Lines 28 - 30 Hoff Can we provide some guidance to the owner in this section? 10-21: Hoff comment resolution: The crack width requirements in section 6 are provided to for this purpose. DELETE 1.2.4

Hjorteset Wu Biily Brannan Pawski Howe Garrison Hoff Hanskat

Wu

The liquid tightness criteria are specified in Section 2.1, Items 1 & 2. Why do we need additional tightness criteria on “embedment plates” to be defined by owner?

Page 7, line 8 Pawski

Referenced section does not match final 376-10 1) ACI 376 section 6.3.3 6.3.2”



77 Page 9, section 2.3.3, lines 19, Khalifa Add ‘separate’ before ‘model’ for clarity. 78 Page 8, section 2.3.1, line32, 33

Khalifa Revise to: Either a proprietary or one of several commercially available FE programs can be used for performing this calculation.

Page 9, line 24 Hoff

Change this sentence to read: “If Liners are considered as structural members, they should be included if they are considered as structural members too.”

10-21-13 editorial


Use “should” in a non-mandatory TN “Recognized concrete constitutive models shall should be used in the analysis.”

10-21-13 editorial

79 Page 10, section 2.4.1, line 30 Khalifa

Revise to: …account for reduction in tensile stiffness after cracking of concrete including the effects of tension stiffening.

10-21-13 Account for reduction in tensile stiffness after cracking of concrete .

Page 11, line 7 Hoff, Widianto

The stress symbols are missing 10-21-13 editorial

Page 12, lines 10, 11

Hoff

Change to read: “Since For the RLG spill case, it is critical that cryogenic liquid be prevented from migrating behind the TCP, and conservative assumptions should be used when selecting material properties.”

10-21-13 editorial

Page 13, Section 2.5

Wu

The specified Code, EN2 part 1- 4.4.2.4 does not contain the equations for crack width calculation. The crack width calculation is shown in Section 7.3.4 of EN 1992-1-1 (2004 edition). Also, the equations shown here are not the same as the equations shown in EN 1992-1-1 (2004 edition). Please provide clarification.

10-21-13 Modify to address to match code and equations. Same as item 74


Page 16 of 16

To be addressed as future business:

Author Comments Legatos Conspicuous by its absence in the TCP-TN write-up is any reference to outer-wall designs that incorporate an impervious, continuous liquid-tight and vapor-tight liner.

This omission overlooks the crucial role such a liner plays in the formulation of criteria and parameters as they pertain to (a) the serviceability of the outer wall following a spill and/or thermal shock; (b) the specified limits on crack-width, residual wall compression, etc., etc.; and (c) the function, scope and design of the TCP itself. Moreover, this omission overlooks the differences between inner-versus-outer, and metallic-versus-non-metallic liners; and the role these differences similarly play in (a) through (c) above. If I am not mistaken, the only Code section that fleetingly mentions this subject with respect to the TCP is Section 6.3.3. This article provides guidance on what to do if a “leak-tight membrane/liner” is not used, but does not say anything about what to do in cases where such a liner is in fact used. The subject is also covered in 6.3.16.1 1nd 16.3.16.2, but those sections do not relate to the TCP. While I have not thoroughly perused the rest of the Code itself to see exactly where else this subject “belongs” and how it should be covered, I believe that TCP-TN coverage might fit in the following sections:

Section 1.1 (add a fifth paragraph) Fig. 1.1 (add Fig. 1.1 c) Section 1.2.4 Section 2.1

2014.02.14 Webinar Discuss NAL negative in Reno. Note - subsequent to webinar NAL staed he would provide draft wording.

ACI 376 TCP Design & Analysis

1 11 / 16 / 2011

ACI 376 Technical Note on 1

Thermal Corner Protection Design & Crack Analysis 2

3

Reported by ACI Committee 376 4

5

Concrete Structures for Refrigerated Liquefied Gas Containment 6

7

* Contributing Member 8

9

INTRODUCTION 10

ACI 376 provides requirements and some basic recommendations for the design of the Thermal Corner 11

Protection (TCP). The purpose of this document is to provide further guidance on (a) TCP detailing and 12

(b) TCP liquid-tightness analysis (i.e., crack width calculations). The document therefore consists of 13

two parts: 14 15

1. Part 1 – Detailing & Construction Issues: TCP detailing and related design issues are presented in 16

the first part of this document 17 18

Neven Krstulovic-Opara*

Chair

Piotr D. Moncarz

Secretary

Junius Allen

Thomas A. Ballard*

Dale Berner

Mike S. Brannan

Hamish Douglas*

Jeffrey Garrison*

Charles S. Hanskat

Alan D. Hatfield

Humayun Hashmi

Kare Hjorteset

George C. Hoff*

Richard Hoffmann

John Holleyoak

Joseph M. Hoptay*

Thomas R. Howe

Dajiu Jiang

Jameel U. Khalifa

Nicholas A. Legatos

Praveen K. Malhotra

Keith A. Mash*

Stephen W. Meier

Robert W. Nussmeier

Rolf P. Pawski*

Ramanujam S. Rajan

William E. Rushing

Robert W. Sward

Eric S. Thompson

Sheng-Chi Wu


2 11 / 16 / 2011

2. Part 2 – Crack Width for Liquid Tightness: One of the key TCP-performance criteria is related to 1

assuring adequate liquid tightness during a spill condition. ACI 376 Code specifies liquid tightness 2

limits in terms of limiting concrete crack widths. The liquid tightness / crack width analysis 3

procedure is presented in the second part of this document, followed by an example outlining 4

proposed analysis steps that would meet ACI 376 Code requirements. 5

6


3 11 / 16 / 2011

CONTENTS 1

2

INTRODUCTION 3

4

CHAPTER 1 � TCP DESIGN 5

1.1—General 6

1.2—Embedment Plate and Anchor Studs 7

8

CHAPTER 2 � LIQUID TIGHTNESS AND CRACK WIDTH ANALYSIS 9

2.1—General 10

2.2—Analysis Procedure 11

2.3—Numerical Model 12

2.4—Material Constitutive Models 13

2.5—Crack Width Calculations 14

15

APENDICES 16

17

APENDIX A � ACI 376 TCP CODE ERQUIREMENTS 18

19

REFERENCES 20

21

A

1

2

3

4

5

16

T7

s8

s9

b10

c11

12

T13

F14

a15

e16

d17

ACI 376

1.1—Genera

The thermal c

secondary con

stressed wall

both (a) a liq

cryogenic tem

The TCP is c

Figure 1.1 a.

as shown in

embedment w

designed to re

al

corner protec

ncrete wall, a

to slab joint f

quid and (b) a

mperatures dur

connected to

The TCP is a

Figure 1.1 b

will be subje

emain intact a

C

a)

Figure 1

ction (TCP) i

as shown in

from addition

a thermal bar

ring a spill co

a secondary

attached to th

b. During th

ected to signi

and in place d

CHAPTER

.1: Thermal C

s an insulated

Figure 1.1 a

nal thermal lo

rrier at the w

ondition.

bottom plate

e concrete wa

he spill condi

ificant therm

during the spil

4

R 1 – TCP

Corner Protec

d and liquid

a. The primar

oads during sp

wall to slab jo

e that provid

all through an

ition, the con

mal and mech

ll condition.

P Design

ction (TCP) D

tight system

ry function o

pill condition

oint, thus pro

es a liquid b

n embedment

ncrete wall a

hanical loads

b)

Detail

located at th

of TCP is to

ns. The TCP

otecting the j

barrier above

t plate located

above the TC

. The TCP e

TCP Design

11 / 16 / 20

he bottom of

protect alrea

is designed t

joint from ex

the slab, as

d at the top of

CP, the TCP

embedment s

& Analysis

011

f the outer

ady highly

to provide

xposure to

shown in

f the TCP,

and TCP

should be


5 11 / 16 / 2011

1

The concrete wall above the TCP, directly exposed to spilled product, should remain liquid and substantially 2

vapor tight during the spill condition. This is assured by limiting cracking in the vicinity of the TCP embedment 3

plate during a spill. 4

5

The scope of this section is to provide guidelines and best practices for the detailing of the TCP region, i.e., the 6

TCP, the TCP embedment plate and TCP embedment plate anchorage to the outer concrete wall. Information is 7

provided for the evaluation of the embedment plate, design of the anchor studs and attachment to the concrete 8

wall, and for the evaluation of cracking in the embedment zone. Material selection and material requirements 9

used in the design of the 9% nickel-steel TCP and secondary bottom plates may be performed in accordance with 10

API 620. 11

12

1.2—Embedment Plate and Anchor Studs 13

The TCP embedment plate should be designed to resist radial and vertical loads and moments resulting from both 14

(a) thermal gradients, as well as (b) mechanical forces that develop during a spill condition. The TCP design 15

consists of the following steps: 16

17

1.2.1 - STEP 1: TCP Load Definition 18

Finite element analyses is performed first to determine forces resulting from thermal gradients applied to the 19

TCP, TCP embedment plate and outer concrete wall. Loads should include all loads that the TCP and the 20

embedment will be subjected to including thermal, prestress, creep, concrete shrinkage, internal pressure, 21

hydrostatic pressure. 22

Various spill levels should be considered to determine the most severe effect on the embedment and concrete 23

wall. This should include spill levels below the top of the TCP which will place the largest temperature 24

differential between the TCP and the concrete wall as well as various spill levels above the TCP up to the full 25

spill level. 26

The analysis should be based on both steady state and transient conditions, as specified in ACI 376 Code section 27

8.4.8. R8.4.8 specifically indicates that during transient conditions an embedment such as the TCP will frequently 28

experience the greatest self-straining forces. 29

30

1.2.2 - STEP 2: Embedment Plate Design 31

The TCP embedment plate should be cast in place and mechanically anchored to the concrete wall or by other 32

suitable means (e.g., ???). Anchor bolts (studs) should be placed in continuous (circumferential) rows along 33

the concrete wall. 34


6 11 / 16 / 2011

When the embedment plate is anchored with circumferential studs, a minimum of two rows of studs should be 1

used to anchor the embedment plate. The spacing between mechanical anchors (studs) should be limited to 2

ensure that the axial, bending and shear stresses in the embedment plate are within the allowable stresses. 3

Embedment plate and mechanical anchor material as well as the TCP material should be suitable for the 4

product temperature. 5

Allowable stresses i.e., material selection and material requirements used in the design of the embedment 6

plates, may be selected in accordance with API 620. 7

8

1.2.3 - STEP 3: Anchor Bolt Design 9

The mechanical anchors bolts (studs) and concrete at the anchor location should meet the requirement for 10

anchoring to the concrete wall defined in the ACI 350 Appendix D. 11

The steel strength of the anchor in tension should be as per ACI 350 section D.5.1. 12

The concrete breakout strength of the anchor in tension should be checked per ACI 350 section D.5.2. 13

The pullout strength of the anchor in tension should be checked per ACI 350 section D.5.3. 14

The steel strength of the anchor in shear should be checked per ACI 350 section D.6.1. 15

The concrete breakout strength of the anchor in shear should be checked per ACI 350 section D.6.2. 16

The concrete pry-out strength of the anchor in shear should be checked per ACI 350 section D.6.3. 17

Interaction of tensile and shear forces should be checked per ACI 350 section D.7. 18

19

20


7 11 / 16 / 2011

CHAPTER 2 – PART 2: Liquid Tightness and Crack Width Analysis 1

2

2.1—General 3

When subjected to the spill condition, the concrete wall will be loaded with high thermal and mechanical forces. 4

These forces will lead to significant wall cracking. During the spill condition, integrity of the concrete wall must 5

be maintained. As specified in the ACI 376 Code, the wall should remain liquid and vapor tight. More 6

specifically: 7

8

1) ACI 376 section 6.3.3: “Unless a leak-tight membrane/liner has been used, a minimum portion of the concrete 9

should remain in compression in accordance with the following: 10

a) A compressive zone of either 10% of the section thickness or 3.5 in., whichever is greater should be provided; and 11

b) A minimum average compressive stress within the compressive zone of 145 lb/in2 should be maintained.” 12

13

2) ACI 376 section 6.3.4: “Calculated crack widths should be considered at TCP embedment when cracking would 14

result in liquid product migrating behind the TCP and compromising its effectiveness. The embedment zone should 15

extend a minimum of two times the wall thickness above the TCP anchorage. Calculated crack widths should not 16

exceed 0.008 in. within the TCP embedment zone. Recommendations for crack width calculations are provided in 17

8.1.1.8.” 18

19

3) ACI 376 section 8.1.1.8 - requires that the cracks widths should be calculated as characteristic and not mean 20

calculated crack widths as per EN 1992-1-1. 21

22

Non-linear finite element analysis should be performed to ensure that these requirements are met. 23

24

25


The analysis procedure consists of the following steps: 27

28

2.2.1 - STEP 1: Thermal Analysis 29

Thermal analysis is performed using Finite Element or Finite Difference approach. Steps include: 30

1) Modeling 31

2) Thermal boundary condition definition/application to equilibrium 32

3) Application of LNG spill or external fire temperature loads obtained by considering rate of leakage into 33

annular space and exposed to transient temperatures 34


8 11 / 16 / 2011

1

4) Continue analysis until steady state final temperatures are reached 2

5) Repeat for all thermal load cases 3

4

2.2.2 - STEP 2: Structural Analysis 5

Structural Analysis is performed using the Finite Element approach. Steps include 6

1) Modeling 7

2) Apply dead load, prestressing and any other steady state load, iterate to equilibrium 8

3) Check crack widths 9

4) Apply LNG spill temperature transients and iterate to equilibrium 10

5) Compute crack widths at times throughout the transient where reinforcing strains are the greatest and 11

exceed the material yield stress 12


14

15


Calculation of cracking and crack widths for the LNG spill condition and/or external fire condition shall be 17

performed using a rational engineering analysis. Both linear or non-linear finite element (FE) analysis 18

19

2.3.1 Non-Linear FE Analysis 20

In this case a non-linear finite element program is used in conjunction with 21

non-linear FE material constitutive models for concrete and steel, and 22

crack width calculations as per Eurocode 2 1992-1-1. 23

Either a proprietary or a commercially available FE programs can be used. There are several commercially 24

available FE programs that are adequate for performing this calculation. A few of more widely used non-linear 25

FE programs available in the United States include ANSYS1, ABACUS2, ADINA3 and NX NASTRAN4. There 26

are also many proprietary FE programs that have been designed specifically to consider the cracking problem for 27

reinforced concrete containment structures. 28

29

2.3.2 Linear FE Analysis 30

Linear FE analysis is usually used only in the preliminary design. Two possible approaches include: 31


9 11 / 16 / 2011

a) Reduced-Stiffness / Cracked-Section Approach: The calculation is performed using a linear finite element 1

analysis where the reduction in stiffness due to cracking is considered. However, it should be noted that this 2

is a very time consuming process and is thus usually only used in the preliminary design. 3

b) Moment-Curvature Approach: The moment curvature response is obtained using a concrete section 4

analysis program in conjunction with: 5

non-linear stress-strain material properties, and 6

a non-linear shell formulation that accounts for the cracking induced stiffness redistribution. 7

This moment-curvature approach would only be suitable for consideration of the steady state linear 8

thermal gradient case and should also be used only for preliminary design. 9

10

2.3.3 Mesh Type and Size 11

The linear or non-linear FE model will usually be axisymmetric or a small slice 3D model of the containment for 12

consideration of the general case of a symmetric LNG spill loads and a larger sector model for consideration of 13

the LNG spill case where non-axisymmetric discontinuities are considered. 14

15

Typically, the structural model contains a fine-mesh or a high density of solution points. The model includes the 16

base slab, foundation springs, containment wall, roof structure and local discontinuities, such as the prestressing 17

buttresses. If the liners are considered as structural members, they should be included too. 18

19

A more detailed substructure model of the region above the TCP anchorage might be considered in order to 20

reduce the computational demand, provided it can be demonstrated that the boundary conditions and loadings 21

correctly represent the global model. 22

23

The reinforcing and prestressing steel may also be modeled and the prestressing forces applied as part of the 24

loading or construction sequence. 25

26

2.3.4 Loading 27

The LNG spill case should include hydrostatic pressure from the LNG fluid in the annular space. Dead load, live 28

load and steady state temperatures coincident prior to the LNG spill, shall be applied prior application of the 29

temperature transients. 30

31

32

Nodal temperature transients will typically be computed in a separate heat transfer finite element or finite 33

difference solution. This thermal model is also a high density model which should include concrete foundation, 34


10 11 / 16 / 2011

soil thermal boundary condition, walls, roof and outside or inside ambient temperature boundary conditions. This 1

model should also include the base slab insulation and the thermal corner protection insulation. 2

3

2.3.5 Heat Analysis 4

The heat transfer analyses should consider both maximum (95th percentile) and minimum (5th percentile) ambient 5

temperatures. The LNG spill is applied internally to the containment and the adjacent tank fire case is applied 6

externally. The LNG spill load case should also include the vapor temperature generated from the spill above the 7

liquid level. Nodal temperature time histories are computed and mapped to the structural finite element model. 8

The concrete wall internal temperatures from the thermal model are also typically used to determine the 9

placement of normal (A615), cold-temperature (A706) and cryogenic reinforcing complying with the 10

requirements of ACI 376. 11

12

The LNG hydrostatic pressure and temperature time histories are applied gradually and the iterative solution 13

computes concrete stresses and strains, crack locations and reinforcing stresses. 14

15

16


18

2.4.1 Concrete Constitutive Model 19

Recognized concrete constitutive models shall be used in the analysis. Examples include material models 20

described by Eurocode 25 and shown in the Figure 2.1 below. The concrete constitutive model should also 21

account for linear tension stiffness effects with tension failure or reduction in stiffness at the cracking strain of the 22

concrete. 23

24

Figure 2.1: Stress-Strain Diagram for Uniaxial Compression from Eurocode 2 1992-1-1 25

A

1

T2

3

F4

w5

6

T7

w8

9

I10

n11

12

213

ACI 376

The c – c m

For short term

c/fc

where:

fc

cu

n

c1

k

Ec,nom

c

Ecm

The tension si

c

c

where:

fctm

In reality, the

numerical stab

2.4.2 Steel C

material relatio

m loading the

= (k.n-n2)/[1 +

concrete

ultimate s

= c/c1 (b

= -0.0022

= (1.1Ec,n

= Ecm/c (

partial sa

= 9.5(fc +

ide of the �c –

= Ecmc (N/m

= 0 for c > o

= 0.30fc2/3 (N

tension softe

bility in the so

Constitutive

onship shown

relationship i

+ n.(k-2)]

compressive

strain < 0.0

both < 0.0)

2 (strain at the

nom).c1/fc

kN/mm2)

fety factor = 1

8)1/3 (Ecm in k

– �c material

mm2) for c < f

or = fctm

/mm2)

ning is usuall

olution.

Models

n in Figure 2.1

is defined as:

strength at 2

e peak stress

1.5 unless jus

kN/mm2; fc in

l relationship

fctm

ly introduced

11

1 is taken from

28 days (N/mm

fc)

stified by adeq

N/mm2)

may be taken

following rup

m Eurocode 2

m2) < 0.0

quate control

n as:

upture of the c

2 Part 1, 4.2.1

procedures.

concrete in or

TCP Design

11 / 16 / 20

1.3.1 (4.1).

(1

(2

(3

(4

(5

(6

(7

(8

(9

(10

(11

(13

der to provide

& Analysis

011

1)

2)

3)

4)

5)

6)

7)

8)

9)

0)

)

3)

e for

A

F1

s2

3

4

5

O6

a7

8

29

S10

c11

12

213

I14

T15

m16

17

218

F19

w20

p21

o22

ACI 376

Figure 2.2 sho

strain curve fo

Figure 2.2:

Reinforc

Other concret

accurately cha

2.4.3 Selecti

Since for the

conservative a

2.4.3.1 - Com

It should be n

Therefore, the

modulus of el

2.4.3.2 - Tens

For tension pr

width. Also,

percentile con

of the concret

ows a typical

or prestressing

Typical Unia

cing Steel from

te and steel c

aracterize mat

on of Mater

LNG spill c

assumptions s

mpressive Res

noted that the

e use of the m

asticity of the

sile Propertie

roperties, the

, if tensile st

ncrete tensile

te in the analy

uniaxial stres

g steel.

axial Stress-St

m Eurocode 2

constitutive m

terial behavio

rial Parame

case, it is crit

should be use

sponse

e extent of th

mean compres

e reinforcing

es

minimum yie

trength of th

strength limit

ysis.

ss strain curve

train Curve fo

2 1992-1-1

models may b

or.

eter Values

tical that cryo

ed when selec

e compressio

ssive concrete

and prestress

eld strength o

he concrete is

ts. A more co

12

e for reinforci

or Figur

Pr

be used prov

ogenic liquid

ting material

on zone is not

e strength is a

ing steel and

of the reinforc

s considered,

onservative ap

ing steel and

re 2.3: Typica

restressing St

vided it can b

d be prevente

properties.

t as critical a

adequate for c

concrete is al

cing should b

, the analysi

pproach wou

Figure 2.3 a t

al Uniaxial St

teel from Euro

be demonstra

ed from migr

as the size an

crack width c

lso considere

be considered

is should eva

uld be to consi

TCP Design

11 / 16 / 20

typical uniaxi

tress-Strain C

ocode 2 1992

ated that thes

ating behind

nd extent of th

calculations.

d adequate.

to maximize

aluate the 95

ider no tensile

& Analysis

011

ial stress-

Curve for

2-1-1

se models

the TCP,

he cracks.

The mean

e the crack

5th and 5th

e capacity


13 11 / 16 / 2011

1

2.4.3.3 - Temperature Effects on Material Properties 2

Improved material properties at low temperature, should not be used unless adequate cryogenic testing has been 3

carried out to confirm these material properties. The variation of the mean coefficient of thermal expansion with 4

temperature should also be considered in the solution. 5

6

7


The Code suggests that Eurocode 2 Part 1 4.4.2.4, (4.80) be the basis for calculation of crack widths. In this 9

approach crack width is determined from steel strains using the following relationship: 10

wk = srmsm (14)

where wk is the 95th percentile crack width, sm is the mean steel strain, srm is the average final crack spacing and 11

is a design value that can be taken as: 12

1.7 for load induced cracking and for restraint cracking in section with minimum dimension in excess of 800 13

mm, or 14

1.3 for restraint cracking in sections with a minimum dimension depth, breadth or thickness, (whichever is the 15

lesser) of 300 mm or less. 16

17

The average final crack spacing, srm, is computed from Eurocode 2 Part 1, 4.4.2.4 (4.82) as: 18

srm = 50 + 0.25k1kk2/r (15)

where: 19

is the barsize or average bar size in mm (16)

k1 = 0.8 for high bond bars and 1.6 for plain bars (17)

k = 0.8 for tensile stresses that are due to intrinsic deformations (18)

k2 is a coefficient which accounts for the form of the strain distribution and is between

0.5 for bending and 1.0 for pure tension.

(19)

r is the effective reinforcing ratio, As/Ac,eff, where As is the area of reinforcement

contained within the effective tension area, Ac,eff

(20)

20

The mean strain value, sm, should be computed directly using the non-linear finite element model and considering 21

all loads existing at the time of cracking. More specifically, the mean strain is obtained by averaging calculated 22

strain over the length of steel that has exceeded the yield limit. In other words, for a single piece of reinforcing 23

steel that has exceeded yield, sm would be computed as the integrated strain over the length of the plastic zone of 24

the reinforcing divided by the length of the plastic zone. 25


14 11 / 16 / 2011

REFERNCES 1

1. ANSYS Inc., Canonsburg, PA 2

2. ABACUS, Inc., Providence, RI 3

3. ADINA R & D, Inc., Watertown, MA 4

4. Siemens PLM Software, Plano TX 5

5. EUROCODE 2 Design of concrete structures – Part 1 General rules and rules for buildings 6

7

8


15 11 / 16 / 2011

1

Appendix A: ACI 376 Code TCP Requirements 2

3

CHAPTER 2—NOTATION AND TERMINOLOGY

thermal corner protection (TCP)—insulated and liquid tight system to protect the outer

tank from thermal shock.

calculated crack width—crack width calculated using a concrete constitutive model

defined in 8.1.1.7.

CHAPTER 6—MINIMUM PERFORMANCE REQUIREMENTS

6.3.2—Under spill conditions, the concrete above the thermal corner protection (TCP)

should remain liquid tight, based upon minimum depths of compression and

precompression.

6.3.4—Calculated crack widths should be considered at TCP embedment when cracking

would result in liquid product migrating behind the TCP and compromising its

effectiveness.

The embedment zone should extend a minimum of two times the wall thickness above the

TCP anchorage. Calculated crack widths should not exceed 0.004 in. within the TCP

embedment

zone.

CHAPTER 8—ANALYSIS & DESIGN

8.1.1.7—Cracking and tension stiffening

should be included by appropriate

modification of the material stress strain

relationship or by the use of finite elements

R8.1.1.7—The Eurocode 2 is recommended

for determining calculated crack widths. In

this case, the calculated crack widths are

characteristic and not mean calculated crack


16 11 / 16 / 2011

that have the capability of cracking under

tension, and crushing under compression as

well as the ability to include reinforcing

steel.

Unless otherwise specified, the concrete

constitutive mode from European Code

should be used for determining calculated

crack widths.

widths.

Extensive field experience with liquid

retaining and offshore concrete structures

has demonstrated that satisfactory liquid

tightness and liquid retaining performance

can be achieved by imposing calculated

crack width limits.

However, it must be noted that calculated

crack width represents an average numerical

value that is not directly equivalent to

calculated crack width measured in the field.

Due to the non-homogeneous nature of

concrete, calculated crack width values

measured in the field will vary from

calculated values, and therefore they cannot

be directly compared to calculated crack

widths.

1

A

1

2

ACI 376

17

TCP Design

11 / 16 / 20

& Analysis

011

ACI 376 (markup update 2014.01.08) TCP Design & Analysis

March 21, 2013 Phoenix – cumulative markups 1 11 / 16 / 2011 (2nd ballot)


Design and Analysis of the TCP Embedment Zone 2

3


5


7


9

INTRODUCTION 10


Protection (TCP) embedment zone. The purpose of this Technical Note is to provide further guidance 12

for (a) design and analysis of theTCP anchorage and (b) liquid-tightness analysis in the TCP embedment 13

zone (i.e., crack width calculations). The document therefore consists of two parts: 14

15


the first part of this document 17

18


Chair

Piotr D. Moncarz

Secretary

Junius Allen

Thomas A. Ballard*

Dale Berner

Mike S. Brannan

Hamish Douglas*

Jeffrey Garrison*

Charles S. Hanskat

Alan D. Hatfield

Humayun Hashmi

Kare Hjorteset

George C. Hoff*

Richard Hoffmann

John Holleyoak

Joseph M. Hoptay*

Thomas R. Howe

Dajiu Jiang

Jameel U. Khalifa

Nicholas A. Legatos

Praveen K. Malhotra

Keith A. Mash*

Stephen W. Meier

Robert W. Nussmeier

Rolf P. Pawski*

Ramanujam S. Rajan

William E. Rushing

Robert W. Sward

Eric S. Thompson

Sheng-Chi Wu








6

Although the code and this TN do not explicitly define leak tightness criteria for the anchorage and 7

embeddment, the design and construction of these components should be carried out with the clear 8

objective of protecting the corner joint from product temperatures. 9

10



CONTENTS 1

2

INTRODUCTION 3

4

CHAPTER 1 ———— TCP DESIGN 5

1.1—General 6

1.2—Embedment Plate and Anchorage 7

8

CHAPTER 2 ———— LIQUID TIGHTNESS AND CRACK WIDTH ANALYSIS 9

2.1—General 10





15

APENDICES 16

17

APENDIX A ———— ACI 376 TCP CODE ERQUIREMENTS 18

19

REFERENCES 20

21



CHAPTER 1 – TCP Design 1

2

a) b)

3

Figure 1.1: Thermal Corner Protection (TCP) Detail 4

5

1.1—General 6

The thermal corner protection (TCP) is an insulated and liquid tight system located at the bottom of the outer 7

secondary concrete wall, as shown in Figure 1.1 a. The primary function of TCP is to protect the already highly 8

stressed wall to slab joint from additional thermal loads during spill conditions. The TCP is designed to provide 9

both (a) a liquid and (b) a thermal barrier at the wall to slab joint, thus protecting the joint from exposure to 10

cryogenic temperatures during a spill condition. 11

12

The TCP is connected to a secondary bottom plate that provides a liquid barrier above the slab, as shown in 13

Figure 1.1 a. The TCP is attached to the concrete wall through an embedment plate located at the top of the TCP, 14

as shown in Figure 1.1 b. During the spill condition, the concrete wall above the TCP, the TCP and TCP 15

embedment will be subjected to significant thermal and mechanical loads. The TCP embedment should be 16

designed to remain functional and in place during the spill condition. 17



1

The concrete wall above the TCP, directly exposed to spilled product, should remain liquid tight and substantially 2

vapor tight during the spill condition. This is assured by maintaining a compression zone in the wall exposed to 3

the spill and by limiting crack width on the inside face of the wall in the TCP embedment zone plate during a spill. 4

5



provided for the evaluation of the embedment plate, design of anchors and attachment to the concrete wall, and 8

for the evaluation of cracking in the embedment zone. 9

10

1.2—Embedment Plate and Anchors 11


(a) thermal gradients, as well as (b) mechanical forces that develop during a spill condition including SSEAFT. 13

The TCP design consists of the following steps: 14

15


• Finite element analyses are performed first to determine forces resulting from thermal gradients applied to the 17



hydrostatic pressure and hydrodynamic pressure due to seismic aftershock (SSEAFT). Initial temperatures and 20

conditions should be selected so as to maximize the demand under consideration. For instance for the design 21

of the anchors the critical condition is likely to be warm wall (summer) a cold thermal shock to the 22

embedment, thereby creating maximum strain across the section. 23

• Various spill levels should be considered to determine the most severe effect on the embedment and concrete 24

wall. This should include a spill level located just below the top of the TCP which will place the largest 25

temperature differential between the TCP and the concrete. A minimum of three levels should be considered 26

for spill levels above the TCP, 1) at top of the TCP embedment zone, 2) at mid-height, and 3) at full spill 27

level. 28

• The analysis should be based on both steady state and transient conditions, as specified in ACI 376 Code section 29



32


• The TCP embedment plate should be cast in place and mechanically anchored to the concrete wall. 34



• The embedment plate should be anchored with a minimum of two rows of circumferential anchors. The 1

spacing between anchors has significant impact on the TCP embedment performance and thus should be 2

considered in design. 3

• Embedment plate and mechanical anchor material as well as the TCP material should be suitable for the 4


• Allowable stresses i.e., material selection and material requirements used in the design of the embedment 6


• Anchor spacings should be selected so as to be multiples of the bar spacing to enable a clash-free design. 8

9

1.2.3 - STEP 3: Anchorage Design 10

The mechanical anchorage (anchors bolts or studs) and concrete at the anchor location should meet the 11

requirement for anchoring to the concrete wall defined in the ACI 350 Appendix D. 12

13

1.2.4 - STEP 4: Li quid Tightness 14

Liquid tightness at the embedment plate should be considered in the details and design of the anchorage as 15

defined by the owner. 16

17

Note - consesus at 2013.10.21 Phoenix meeting was new "1.2.4 Step 4" is not needed and

should be deleted in its entirety. Shown for information and is to be deleted.




2

2.1—General 3

During the spill condition, the secondary concrete wall will be subjected to high large forces due to mechanical 4

loads and thermal effects that will lead to significant wall cracking. Integrity of the concrete wall must be 5

maintained. As specified in the ACI 376 Code, the wall should remain liquid and substantially vapor tight. More 6

specifically: 7



a) A compressive zone of either 10% of the section thickness or 3.5 in., whichever is greater shall be provided; and 10


12





8.1.1.8.” 17

18


crack widths. 20

21


23

24



27


Thermal analysis is performed based on a thermal model using Finite Element or Finite Difference approach. 29

Steps include: 30

1) Modeling including specific tank geometry and material thermal properties 31


3) Application of RLG spill temperature loads obtained by considering rate of leakage into annular space 33

and exposed to transient temperatures 34



1

4) Continue analysis until steady state temperature distribution is obtained 2


4


Structural Analysis is performed based on the stress model using the Finite Element approach. Steps include 6

1) Modeling 7

2) To establish intiial conditions, apply dead load, prestressing, normal opearting thermal and any other 8

steady state load, then iterate to equilibrium 9

3) Apply RLG spill temperature transients and iterate to equilibrium 10

4) Beginning with the intial condition defined in step 2, compute crack widths at time-steps throughout the 11

transient where reinforcing strains are the greatest . Furthermore, at each time step determine the adequacy of 12

the compression zone. 13


15

16


Calculation of cracking and crack widths for the RLG spill condition shall be performed using a rational 18

engineering analysis. Linear or non-linear finite element (FE) analysis should be used as discussed in 2.3.1 and 19

2.3.2. 20

21


In this case a non-linear finite element program is used by: 23

• Implementing the non-linear material constitutive relationship for concrete and steel, 24

• Carrying out FE analysis to obtainstress/strain contours, and 25

• Performing FE analysis post process to calculate characteristic crack width, following a method such as 26

Eurocode EN 2 1992-1-1 (see section 2.5 in this TN). 27

Either a proprietary or one of several commercially available FE programs can be used for performing this 28

calculation. A few of more widely used non-linear FE programs available in the United States include ANSYS1, 29

ABACUS2, ADINA 3 and NX NASTRAN4. There are also many proprietary FE programs that have been 30

designed specifically to consider the cracking problem for reinforced concrete containment structures. 31

32

2.3.2 SimplifiedFE Analysis 33



Simplified FE analysis is usually used only in the preliminary design. Two possible approaches include: 1


analysis where the reduction in stiffness due to cracking is considered. 3

b) Moment-Curvature Approach: The moment curvature properties are obtained using a concrete section 4

analysis by implementing in conjunction with: 5

• non-linear stress-strain material properties, and 6

• a non-linear shell formulation that accounts for the cracking induced stiffness redistribution. 7

8


The linear or non-linear FE model will usually be axisymmetric or a 3D model of the containment for 10

consideration of the general case of a symmetric RLG spill loads and a separate model for consideration of the 11

RLG spill case where non-axisymmetric discontinuities are considered. 12

13



buttresses. Liners should be included if they are considered as structural members. 16

17




21

The reinforcing and prestressing steel should may also be modeled and the prestressing forces applied as part of 22

the loading or construction sequence. 23

24

2.3.4 Loading 25

The RLG spill case should include hydrostatic pressure from the RLG fluid in the annular space. Dead load, live 26

load and steady state temperatures coincident prior to the RLG spill, shall be applied prior application of the 27


29

30


difference solution. The thermal model used for determining nodal temperatures This thermal model is also a 32

high density model which should include concrete foundation, soil thermal boundary condition, walls, roof and 33



outside or inside ambient temperature boundary conditions. This model should also include the base slab 1

insulation and the thermal corner protection insulation. 2

3

2.3.5 Thermal Analysis 4


temperatures. The RLG spill is applied internally to the containment y. The RLG spill load case should also 6

include the vapor temperature generated from the spill above the liquid level. Nodal temperature time histories 7

are computed and mapped to the structural finite element model. The concrete wall internal temperatures from 8

the thermal model are also typically used for selection and placement of reinforcement complying with the 9


11

The RLG hydrostatic pressure and temperature time histories are applied gradually and the iterative solution 12


14

15


17


Recognized concrete constitutive models should be used in the analysis. Examples include material models 19


account for reduction in tensile stiffness after cracking of concrete including the effects of tension stiffening. 21

22

Figure 2.1: Stress-Strain Diagram for Uniaxial Compression from Eurocode EN 2 1992-1-1 23

24

The σc – εc material relationship shown in Figure 2.1 is taken from Eurocode EN 2 Part 1, 4.2.1.3.1 (4.1). 25



1

For short term loading the relationship is defined as: 2

σc/fc = (k.n-n2)/[1 + n.(k-2)] (1)

where: 3

fc concrete compressive strength at 28 days (N/mm2) < 0.0 (2)

εcu ultimate strain < 0.0 (3)

n = εc/εc1 (both < 0.0) (4)

εc1 = -0.0022 (strain at the peak stress fc) (5)

k = (1.1Ec,nom).εc1/fc (6)

Ec,nom = Ecm/γc (kN/mm2) (7)

γc partial safety factor = 1.5 unless justified by adequate control procedures. (8)

Ecm = 9.5(fc + 8)1/3 (Ecm in kN/mm2; fc in N/mm2) (9)

4

The tension side of the fc - εc material relationship may be taken as: 5

σc = Ecmεc (N/mm2) for σc < fctm (10)

σc = 0 for σc > or = fctm (11)

where: 6

fctm = 0.30fc2/3 (N/mm2) (13)

7

Tension softening is usually introduced following rupture of the concrete in order to provide for numerical 8

stability in the solution. 9

10

2.4.2 Steel Constitutive Models 11

Figure 2.2 shows a typical uniaxial stress strain curve for reinforcing steel and Figure 2.3 a typical uniaxial stress-12

strain curve for prestressing steel. 13

14

Figure 2.2: Typical Uniaxial Stress-Strain Curve for

Reinforcing Steel from Eurocode EN 2 1992-1-1


Prestressing Steel from Eurocode EN 2 1992-1-1



1

2

Other concrete and steel constitutive models may be used provided it can be demonstrated that these models 3

accurately characterize material behavior. 4

5

2.4.3 Selection of Material Parameter Values 6

For the RLG spill case, it is critical that cryogenic liquid be prevented from migrating behind the TCP, and 7

conservative assumptions should be used when selecting material properties. 8

9

2.4.3.1 - Compressive Response 10

It should be noted that the extent of the compression zone is not as critical as the size and extent of the cracks. 11

Therefore, the use of the mean compressive concrete strength is adequate for crack width calculations. The mean 12

modulus of elasticity of the reinforcing and prestressing steel and concrete is also considered adequate. 13

14

2.4.3.2 - Tensile Properties 15

For tension properties, the minimum yield strength of the reinforcing should be considered to maximize the crack 16

width. Also, if tensile strength of the concrete is considered, the analysis should evaluate the 95th and 5th 17

percentile concrete tensile strength limits. A more conservative approach would be to consider no tensile capacity 18

of the concrete in the analysis. 19

20





25




The Code suggests that Eurocode EN 2 Part 1 4.4.2.4, (4.80) be the basis for calculation of crack widths. In this 2


wk = βsrmεsm (14)

where wk is the 95th percentile crack width, εsm is the mean steel strain, srm is the average final crack spacing and β 4


• 1.7 for load induced cracking and for restraint cracking in section with minimum dimension in excess of 800 6

mm, or 7

• 1.3 for restraint cracking in sections with a minimum dimension depth, breadth or thickness, (whichever is the 8


10


srm = 50 + 0.25k1kk2φ/ρr (15)

where: 12

φ is the barsize or average bar size in mm (16)





(19)

ρr is the effective reinforcing ratio, As/Ac,eff, where As is the area of reinforcement


(20)

13

The mean strain value, εsm, should be computed directly using the non-linear finite element model and considering 14



steel that has exceeded yield, εsm would be computed as the integrated strain over the length of the plastic zone of 17


19

REFERNCES 20









2





defined in 8.1.1.7.




precompression.



effectiveness.



embedment

zone.











widths.





steel.




crack widths.






crack width limits.










widths.

1



1

2




Design and Analysis of the TCP Embedment Zone 2

3


5


7


9

INTRODUCTION 10


Protection (TCP) embedment zone. The purpose of this Technical Note is to provide further guidance 12

for (a) design and analysis of theTCP anchorage and (b) liquid-tightness analysis in the TCP embedment 13

zone (i.e., crack width calculations). The document therefore consists of two parts: 14

15



18


Chair

Piotr D. Moncarz

Secretary

Junius Allen

Thomas A. Ballard*

Dale Berner

Mike S. Brannan

Hamish Douglas*

Jeffrey Garrison*

Charles S. Hanskat

Alan D. Hatfield

Humayun Hashmi

Kare Hjorteset

George C. Hoff*

Richard Hoffmann

John Holleyoak

Joseph M. Hoptay*

Thomas R. Howe

Dajiu Jiang

Jameel U. Khalifa

Nicholas A. Legatos

Praveen K. Malhotra

Keith A. Mash*

Stephen W. Meier

Robert W. Nussmeier

Rolf P. Pawski*

Ramanujam S. Rajan

William E. Rushing

Robert W. Sward

Eric S. Thompson

Sheng-Chi Wu

Deleted: Thermal Corner Protection Design & Crack Analysis

Comment [LE1]: 001 Brannan part 1 Revise title to “Design and Analysis of the TCP Embedment Zone.” Monday 3-19: (9/9) Wu, Pawski, Howe, Hoff, Hoptay, Garrison, Ballard, Brannan, NKO) 002 Pawski Revise title to read “Liquid tightness Anaysis at the TCP.” Saturday 17 - (8/8) Brannan, Ballard, Hoptay, Howe, Juang, NKO, Pawski, Wu Note superceded by Monday vote.

Comment [LE2]: 003 Hoff comment not addressed. “Useful but seems to be a little bit too long for a TechNote”

Comment [LE3]: 004 Mash comment not addressed. “Comments included in document. There are couple of blind spots which we ought to address via telecon”

Comment [LE4]: 005 Pawski withdrawn Revisit sugested organization at time of final editing. (Sunday 3-18: present - Wu, Pawski, Hoff, Berner, Howe, Hjorteset, Hoptay, Garrison, Ballard, Brannan, NKO)

Comment [LE5]: 001 – Brannan part 2. Page 1 lines 11-14. Revise to read as shown. (9/9) Wu, Pawski, Howe, Hoff, Hoptay, Garrison, Ballard, Brannan, NKO)

Deleted: document

Deleted: on

Deleted: detailing

Deleted: TCP








6




10

Comment [LE6]: Part of Ref. 001: T.Ballard negative . Draft of new paragraph for committee consideration agreed to at Minneapolis – 2013.04.13.



CONTENTS 1

2

INTRODUCTION 3

4


1.1—General 6

1.2—Embedment Plate and Anchorage 7

8


2.1—General 10





15

APENDICES 16

17


19

REFERENCES 20

21

Deleted: Anchor Studs

Comment [LE7]: Part of Ref. no. 021 - 2013.04.13 Minneapolis. Edit TN to use "anchors" rather than "bolts" or "studs" throughout.




2

a) b)

3


5

1.1—General 6






12





designed to remain functional and in place during the spill condition. 17

Comment [LE8]: 006 Wu Comment is “Need to include the secondary bottom plate in the model.” On Figure 1a add a detail circle to identify where Figure 1b is located.

Comment [LE9]: Part of 001 Ballard negative : I believe the embedment zone should extend at least to the bottom of the TCP plate since migration of cracks in this zone would compromise the leak tightness of the detail.

Comment [LE10]: Part of 021 Hoptay negative 2013 04.13 Minneapolis: annotate sketch to state "2 rows of anchors minimum."

Comment [LE11]: Oliver – editorial Pg 4 line 8: addd "the" as shown. 2013.10.21 Phoenix

Deleted: intact

Comment [LE12]: 007 Mash Commetn is “Suggest replacement of 'intact' with functional.” Agree - (8/8) Garrison, Howe, Hoff, Ballard, Pawski, Brannan, Hoptay, Wu



1



the spill and by limiting crack width on the inside face of the wall in the TCP embedment zone plate during a spill. 4

5



provided for the evaluation of the embedment plate, design of anchors and attachment to the concrete wall, and 8

for the evaluation of cracking in the embedment zone. 9

10

1.2—Embedment Plate and Anchors 11




15


• Finite element analyses are performed first to determine forces resulting from thermal gradients applied to the 17



hydrostatic pressure and hydrodynamic pressure due to seismic aftershock (SSEAFT). Initial temperatures and 20

conditions should be selected so as to maximize the demand under consideration. For instance for the design 21

of the anchors the critical condition is likely to be warm wall (summer) a cold thermal shock to the 22

embedment, thereby creating maximum strain across the section. 23


wall. This should include a spill level located just below the top of the TCP which will place the largest 25

temperature differential between the TCP and the concrete. A minimum of three levels should be considered 26

for spill levels above the TCP, 1) at top of the TCP embedment zone, 2) at mid-height, and 3) at full spill 27

level. 28




32


• The TCP embedment plate should be cast in place and mechanically anchored to the concrete wall. 34

Comment [LE13]: 009 Hoptay Withdrawn

Comment [LE14]: 2013.04.13 Minneapolis – Pg 5 line 2 change "liquid" to "liquid tight."

Comment [LE15]: 008 Mash Response: “Limiting values are covered elsewhere and do not need to be repeated here.”Revise to read for clarity as shown. (8/8) Garrison, Howe, Hoff, Ballard, Pawski, Brannan, Hoptay, Wu

Deleted: cracking

Comment [LE16]: 011 Mash Page 5 line 4 - no action, included with response 010

Deleted:

Deleted: the vicinity of

Deleted: the anchor studs


Deleted: Material selection and material ... [1]

Comment [LE18]: 010 Hoptay Delete last sentence. ... [2]

Deleted:

Deleted: Anchor Studs

Comment [LE19]: Part of Ref no. 021 - ... [3]

Comment [LE20]: 013 Wu ... [4]

Comment [LE21]: 014 Mash ... [5]

Deleted: is

Comment [LE22]: 015 Mash

Comment [LE23]: 016 Wu ... [6]

Deleted: Starting

Comment [LE24]: 2013.04.13 Minneapolis. ... [7]

Deleted: and

Deleted: shocking

Deleted: embed

Deleted: . Whereby

Deleted: straining

Deleted: spill levels

Deleted: wall as well as various

Deleted: up to the

Comment [LE25]: 017 and 018 Mash ... [8]

Comment [LE26]: 019 Mash. ... [9]

Comment [LE27]: 020 Mash negative - ... [10]

Deleted: or by other suitable means (e.g., ???)

Deleted: Anchor bolts (studs) should be placed in ... [11]

Comment [LE28]: 020 Hoptay negative - ... [12]



• The embedment plate should be anchored with a minimum of two rows of circumferential anchors. The 1

spacing between anchors has significant impact on the TCP embedment performance and thus should be 2

considered in design. 3






9

1.2.3 - STEP 3: Anchorage Design 10

The mechanical anchorage (anchors bolts or studs) and concrete at the anchor location should meet the 11


13




17



Deleted: When the embedment

Deleted: is

Deleted: circumferential studs,

Deleted: studs

Deleted: should be used to anchor the embedment plate

Deleted: mechanical anchors (studs) should be limited to ensure that the axial, bending and shear stresses in the embedment plate are within the allowable stresses

Comment [LE29]: 021 Mash negative - resolved Comment is “When? Embedment must be anchored with studs?” Revise as shown, and consider further discussion in future TN. Tues 3-20: (5/5) Garrison, Ballard, Pawski, Brannan, NKO 021 Hoptay negative - resolved – addressed here & Fig. 1.1(b)

Comment [LE30]: 022 Mash Comment is “Consider use of ACI 349? More stringent with respect to anchors.” Non-persuasive. Tues 3-20: (4/4) Garrison, Ballard, Pawski, Brannan

Comment [NKO31]: 026 Mash Non-persuasive, but edit bullet point as shown. Tuesday 3-20: (5/5) Garrison, Ballard, Pawski, Brannan, NKO

Deleted: Anchor Bolt

Deleted: (

Comment [LE32]: 023 Mash Comment is “Consider use of ACI 349? More stringent with respect to anchors.” Non-persuasive. . Discussed March 19 and decided that ACI 350 is appropriate. Also, there does not appear to be any difference between ACI 350 and ACI 349 on this topic; Garrison to verify. Tues 3-20: (4/4) Garrison, Ballard, Pawski, Brannan.

Deleted: <#>The steel strength of the anchor in tension should be as per ACI 350 section D.5.1.

Comment [NKO33]: 024 Hoptay Withdrawn, already addressed.

Deleted: <#>¶<#>The concrete breakout strength of the anchor in tension should be checked per ACI 350 section D.5.2.¶<#>The pullout strength of the anchor in tension should be checked per ACI 350 section D.5.3.¶<#>The steel strength of the anchor in shear should be checked per ACI 350 section D.6.1.¶... [13]

Comment [LE34]: 027 Wu Comment is to add bullet point to read “In moderate or high seismic site, the design strength of anchors should be reduced to 75% of the design strength per Section D.3.3.3.” Withdrawn, Monday 3-19..

Comment [LE35]: 025 Wu Revise 1.2.3 as shown and as new section 1.2.4 as shown. ... [14]




2

2.1—General 3

During the spill condition, the secondary concrete wall will be subjected to high large forces due to mechanical 4

loads and thermal effects that will lead to significant wall cracking. Integrity of the concrete wall must be 5

maintained. As specified in the ACI 376 Code, the wall should remain liquid and substantially vapor tight. More 6

specifically: 7



a) A compressive zone of either 10% of the section thickness or 3.5 in., whichever is greater shall be provided; and 10


12





8.1.1.8.” 17

18


crack widths. 20

21


23

24



27



Steps include: 30



3) Application of RLG spill temperature loads obtained by considering rate of leakage into annular space 33

and exposed to transient temperatures 34

Deleted: When subjected to

Comment [LE36]: 028 Mash negative - resolved

Comment [LE37]: 029 Pawski Comment is mostly superceded by 028 changes, but add “secondary” from Sat 9-17 discussion. (9/9 – see Saturday attendance)

Deleted: loaded with

Deleted: thermally inducedand mechanical forces. These forces

Deleted: During the spill condition, integrity

Comment [LE38]: 028 Hoptay negative - resolved Withdrawn 2013.04.13 Minneapolis

Comment [LE39]: 028 Mash negative - resolved 030 & Comment is “Suggest using nomenclature such as demand as opposed to load.” Agree in part, revise to read as shown Tues 3-20 - (4/4) Garrison, Ballard, Pawski, Brannan

Deleted: 6.3.3

Comment [LE40]: Pawski – editorial. Pg 7 line 8, referenced section does not match published 376-11. Change to 6.3.2. 21-Oct-2013 Phoenix

Deleted: should

Comment [LE41]: 031 Mash Tues 3-20 - (4/4) Garrison, Ballard, Pawski, Brannan

Formatted: Superscript

Deleted: 6.3.4


Deleted: calculated

Deleted: as per EN 1992-1-1

Comment [LE43]: 032 Mash Tues 3-20 - (4/4) Garrison, Ballard, Pawski, Brannan ... [15]

Deleted:

Comment [LE44]: 033 Jiang Agree, revise as shown. ... [16]

Comment [LE45]: 034 Jiang Agree, revise as shown. ... [17]

Deleted: LNG

Comment [LE46]: 035, 046, 048 Jiang Agree to change LNG to RLG throughout. ... [18]

Deleted: or external fire

Comment [LE47]: 036 Mash ... [19]



1

4) Continue analysis until steady state temperature distribution is obtained 2


4



1) Modeling 7

2) To establish intiial conditions, apply dead load, prestressing, normal opearting thermal and any other 8

steady state load, then iterate to equilibrium 9

3) Apply RLG spill temperature transients and iterate to equilibrium 10

4) Beginning with the intial condition defined in step 2, compute crack widths at time-steps throughout the 11

transient where reinforcing strains are the greatest . Furthermore, at each time step determine the adequacy of 12

the compression zone. 13


15

16


Calculation of cracking and crack widths for the RLG spill condition shall be performed using a rational 18

engineering analysis. Linear or non-linear finite element (FE) analysis should be used as discussed in 2.3.1 and 19

2.3.2. 20

21


In this case a non-linear finite element program is used by: 23

• Implementing the non-linear material constitutive relationship for concrete and steel, 24


• Performing FE analysis post process to calculate characteristic crack width, following a method such as 26

Eurocode EN 2 1992-1-1 (see section 2.5 in this TN). 27

Either a proprietary or one of several commercially available FE programs can be used for performing this 28

calculation. A few of more widely used non-linear FE programs available in the United States include ANSYS1, 29

ABACUS2, ADINA 3 and NX NASTRAN4. There are also many proprietary FE programs that have been 30

designed specifically to consider the cracking problem for reinforced concrete containment structures. 31

32

2.3.2 SimplifiedFE Analysis 33

Comment [LE48]: 037 Mash & 037 Garrison negative – not resolved Comment is “More explanation required regarding semi coupled analysis (fully coupled not needed)” Respons is that comment is not clear. (5/5) Garrison, Ballard, Pawski, Brannan, Hjorteset

Deleted: final temperatures are reached

Comment [LE49]: 037 Jiang Agree, revise as noted. (11/11 Wu, Pawski, Berner, Howe, Hanskat Hoptay, Garrison, Ballard, DJ, Brannan, NKO)

Comment [LE50]: 039 Jiang Agree, revise as shown. (11/11 Wu, Pawski, Berner, Howe, Hanskat Hoptay, Garrison, Ballard, DJ, Brannan, NKO)

Comment [LE51]: 040 Mash Comment is “Thermal loads from spillage need porting across to structural model.” Response is that no change is needed because the comment is implicit in the step-by-step procedure. (5/5) Garrison, Ballard, Pawski, Brannan, Hjorteset

Deleted: Apply

Comment [LE52]: 041 Jiang & 041 Garrison negative – not resolved Agree, revise as shown. ... [20]

Deleted: <#>Check crack widths¶

Comment [LE53]: 042 Pawski – Sat ... [21]

Deleted: LNG

Deleted: Compute

Deleted: times

Deleted: and exceed the material yield stress

Comment [LE54]: 044 & 047 Garrison ... [22]

Deleted: LNG

Deleted: and/or external fire condition

Deleted: Both linear

Comment [LE55]: 049 Mash & Ballard ... [23]

Deleted: in conjunction with

Deleted: non-linear FE material constitutive ... [24]

Deleted: and

Deleted: calculations as per

Comment [LE56]: 050 Jiang ... [25]

Deleted: a

Deleted: . There are several commercially ... [26]

Comment [LE57]: 078 Khalifa: Combine 1st ... [27]

Deleted: Linear

Comment [LE58]: Part of 056 Jiang



Simplified FE analysis is usually used only in the preliminary design. Two possible approaches include: 1


analysis where the reduction in stiffness due to cracking is considered. 3

b) Moment-Curvature Approach: The moment curvature properties are obtained using a concrete section 4

analysis by implementing in conjunction with: 5



8


The linear or non-linear FE model will usually be axisymmetric or a 3D model of the containment for 10

consideration of the general case of a symmetric RLG spill loads and a separate model for consideration of the 11

RLG spill case where non-axisymmetric discontinuities are considered. 12

13



buttresses. Liners should be included if they are considered as structural members. 16

17




21



24

2.3.4 Loading 25

The RLG spill case should include hydrostatic pressure from the RLG fluid in the annular space. Dead load, live 26

load and steady state temperatures coincident prior to the RLG spill, shall be applied prior application of the 27


29

30




Deleted: Linear

Deleted: However, it should be noted that this is a very time consuming process and is thus usually only used in the preliminary design.

Comment [LE59]: 051 Mash Comment is “We should exercise caution. A number of vendors have historically used this approach and we ought not to say it should not be used.” No change, non-persuasive. Tue 3-20 - (3/3) Pawski, Garrison, Ballard) 053 Jiang Deleted sentence in its entirety (11/11 Wu, Pawski, Hoff, Berner, Howe, Berner, Hjorteset, Garrison, Ballard, DJ, Brannan, NKO) 052 Hoptay Withdrawn Monday March 19. 054 Mash Not addressed, but deletion of the sentence should satisfy comment that “contractors should be able to choose design method.

Deleted: response is

Deleted: program

Deleted: <#>This moment-curvature approach would only be suitable for consideration of the ... [28]

Comment [LE60]: 055 Wu Withdrawn Monday March 19. 056 Jiang Revised as shown. ... [29]

Deleted: small slice

Deleted: LNG

Deleted: larger sector

Comment [LE61]: 077 Khalifa: Pg 9 line 19 add "separate" before "model." ... [30]

Deleted: LNG

Comment [LE62]: 059 Jiang Revise as shown. ... [31]

Comment [LE63]: 060 Jiang No change. ... [32]

Deleted: If the liners are considered as structural members, they

Deleted: too

Comment [LE64]: Pg 9 line 24 Hoff: editorial Change last sentence to read ... [33]

Comment [LE65]: 061 Jiang - withdrawn

Comment [LE66]: 062 Wu ... [34]

Deleted: LNG

Deleted: LNG

Deleted: LNG

Comment [NKO67]: 063 Mash ... [35]





3

2.3.5 Thermal Analysis 4


temperatures. The RLG spill is applied internally to the containment y. The RLG spill load case should also 6

include the vapor temperature generated from the spill above the liquid level. Nodal temperature time histories 7

are computed and mapped to the structural finite element model. The concrete wall internal temperatures from 8

the thermal model are also typically used for selection and placement of reinforcement complying with the 9


11

The RLG hydrostatic pressure and temperature time histories are applied gradually and the iterative solution 12


14

15


17


Recognized concrete constitutive models should be used in the analysis. Examples include material models 19


account for reduction in tensile stiffness after cracking of concrete including the effects of tension stiffening. 21

22


24


Deleted: Heat

Comment [LE68]: 064 Pawski Change as shown. (9/9 – see Saturday attendance)

Deleted: LNG

Deleted: and the adjacent tank fire case is applied externall

Deleted: LNG

Comment [LE69]: 065 Mash Comment is “Vapour temperature? This will be affected by convection etc? How is this to be determined.” This is a general comment/question. No new text proposed. Leave as is and consider the issue in the future.

Deleted: to determine the

Deleted: normal (A615), cold-temperature (A706) and cryogenic reinforcing

Comment [LE70]: 066 Hoptay Revise as shown. FUTURE ACTION: revisit use of A706 materials in ACI 376 section 4.7.2. Mon 3-19 -(9/9) Wu, Pawski, Howe, Hoff, Hoptay, Garrison, Ballard, Brannan, NKO)

Deleted: LNG

Deleted: shal

Comment [LE71]: Pg 10 line 28 Hoff: editorial Use "should" in non-mandatory TN 21-Oct-2013 Phoenix

Deleted: l

Deleted: linear tension stiffness effects with tension failure or reduction in stiffness at the cracking strain of the concrete

Comment [LE72]: 079 Khalifa: Revise pg 10 line 30 to read: 21-Oct-2013 Phoenix

Comment [LE73]: Typo (see page 8)

Comment [LE74]: 067 Wu Comment is “Need to specify the value for ɛcu = 0.3%” Insert in Figure 7 if editing is possible.

Comment [LE75]: Typo



1


σc/fc = (k.n-n2)/[1 + n.(k-2)] (1)

where: 3



n = εc/εc1 (both < 0.0) (4)


k = (1.1Ec,nom).εc1/fc (6)




4

The tension side of the fc - εc material relationship may be taken as: 5



where: 6

fctm = 0.30fc2/3 (N/mm2) (13)

7

Tension softening is usually introduced following rupture of the concrete in order to provide for numerical 8

stability in the solution. 9

10




14





Comment [LE76]: 068 Pawski - withdrawn Add a tension “tail” to the figure. (Sunday: present - Wu, Pawski, Hoff, Berner, Howe, Hjorteset, Hoptay, Garrison, Ballard, Brannan, NKO)

Deleted: c – c

Comment [LE77]: Pg 11 line 7 Hoff & Widianto: editorial Stress symbols are missing 21-Oct-2013 Phoenix

Deleted: In reality, the tension

Comment [LE78]: 069 Pawski - withdrawn Editorial, change as noted. (Sunday: present - Wu, Pawski, Hoff, Berner, Howe, Hjorteset, Hoptay, Garrison, Ballard, Brannan, NKO)

Comment [LE79]: 070 Wu Comment is “Need to specify the limiting ultimate steel strain, ɛu.

(Note that the ultimate strains for steel are not discussed in the Code. Committee should discuss, and set the allowable limits. ) CONSIDER FOR FUTURE ACTION




1

2



5


For the RLG spill case, it is critical that cryogenic liquid be prevented from migrating behind the TCP, and 7

conservative assumptions should be used when selecting material properties. 8

9





14






20





25

Deleted: Since for

Comment [LE81]: Pg 12 lines 10 & 11 Hoff: editorial. Change to read 21-Oct-2013 Phoenix

Deleted: LNG

Comment [LE82]: 072 Mash Comment is “Need to look into this area in more detail and will advise through discussion.” This is a general comment.. No new text proposed and no additional input provided sine the last ballot. Leave as is and consider the issue in the future.






wk = βsrmεsm (14)




mm, or 7



10


srm = 50 + 0.25k1kk2φ/ρr (15)

where: 12






(19)



(20)

13






19

REFERNCES 20






Comment [LE83]: 071 Wu Monday march 19 withdrawn because 2004 edition equations will be used. 074 Garrison negative and 2012.10.13 Phoenix Wu negative – not addressed . To update equations to 2004 edition.





2





defined in 8.1.1.7.




precompression.



effectiveness.



embedment

zone.











widths.

Comment [LE85]: 075 Pawski Resolution is: This is an editorial comment. Appendix A is a “copy” of what is in the Code. It is provided for reader’s convenience. Reconcile duplication between 2.1 and Appendix A, such as removing Appendix A

Comment [LE86]: 074 Garrison Response is: 1)Definition of the spill loading is beyond the scope of the current TN and will thus be addressed in a separate document defining LNG spill conditions. 2)Update and change the model to the latest 2004 edition of the Eurocode,

(Sunday: present - Wu, Pawski, Hoff, Berner, Howe, Brnnan, Hjorteset, Hoptay, Garrison, Ballard, NKO)

Deleted: 0.004

Comment [LE87]: 073 Wu Comment is “Change the crack width of 0.004 in to 0.008 in.” Section 6.3.4 is text copied over from the Code and provided herein for the user’s benefit. Text must remain as-is to properly reflect the Code requirement. Changes to be addressed only if/once the Code has been changed accordingly.





steel.




crack widths.






crack width limits.










widths.

1

Comment [LE88]: 076 Thompson Comment is “Remove the last sentence from the Code Commentary: “Due to the non-homogeneous nature of concrete, calculated crack width values measured in the field will vary from calculated values, and therefore they cannot be directly compared to calculated crack widths.”” Section R8.1.1.7 is text copied over from the Code and provided herein for the user’s benefit. Text must remain as-is to properly reflect the Code requirement. Changes to be addressed only if/once the Code has been changed accordingly.



1

2

[1] Deleted rpawski 3/31/2012 1:57:00 PM

Material selection and material requirements used in the design of the 9% nickel-steel TCP and secondary

bottom plates may be performed in accordance with API 620.

Page 5: [2] Comment [LE18] rpawski 1/14/2014 3:42:00 PM

010 Hoptay Delete last sentence. 012 Hoptay Comment found non-persuasive. Dealt with already elsewhere by deleting last sentence (Hoptay comment). (8/8) Garrison, Howe, Hoff, Ballard, Pawski, Brannan, Hoptay, Wu


Part of Ref no. 021 - 2013.04.13 Minneapolis. Edit TN to use "anchors" rather than "bolts" or "studs" throughout.


013 Wu Agree, revise as shown. (9/9) Wu, Pawski, Howe, Hoff, Hoptay, Garrison, Ballard, Brannan, NKO)


014 Mash Agree in principle, but this is already addressed by requirement of analysis for thermal gradient. Future TN will define design conditions. (6/6) Garrison, Howe, Ballard, Pawski, Hoptay, Wu


016 Wu Agree, revise as shown. (9/9) Wu, Pawski, Howe, Hoff, Hoptay, Garrison, Ballard, Brannan, NKO)


2013.04.13 Minneapolis. Revise last three sentences as shown.


017 and 018 Mash negative . Should be 5 locations, not 3 as shown. Revise as shown. (6/6) Garrison, Howe, Ballard, Pawski, Hoptay, Wu


019 Mash. Comment is “Suggest inclusion of the words 'in the absence of a defined leak rate all intermediate levels shall be investigated. In practice this may be achieved by running a number spill levels to so as adequately bound the demands. Typically 5 levels may be used dividing the spill height into 4 equal units. In addition levels in close proximity to the TCP must be investigated.” No change. The response will be covered in the future TN on the design basis (DJ comment).


020 Mash negative - resolved Comment is “Must be achieved through mechanical anchoring.” Agree, revise as shown.

Anchor bolts (studs) should be placed in continuous (circumferential) rows along the concrete wall. •


020 Hoptay negative - resolved Delete 2nd sentence as it is essentially same as next sentnece. Agree 2013.04.13 Minneapolis.

Page 6: [13] Deleted rpawski 3/31/2012 2:47:00 PM

The concrete breakout strength of the anchor in tension should be checked per ACI 350 section

D.5.2. The pullout strength of the anchor in tension should be checked per ACI 350 section D.5.3. The steel strength of the anchor in shear should be checked per ACI 350 section D.6.1. The concrete breakout strength of the anchor in shear should be checked per ACI 350 section

D.6.2. The concrete pry-out strength of the anchor in shear should be checked per ACI 350 section D.6.3. Interaction of tensile and shear forces should be checked per ACI 350 section D.7.


025 Wu Revise 1.2.3 as shown and as new section 1.2.4 as shown. (9/9) Wu, Pawski, Howe, Hoff, Hoptay, Garrison, Ballard, Brannan, NKO) 2013.10.21 Phoenix – consensus was new section 1.2.4 is not needed and should be deleted in its entirtety.


032 Mash Tues 3-20 - (4/4) Garrison, Ballard, Pawski, Brannan 032 Mash, Wu & Garrison negatives – not resolved – Requires discussion as it is not clear what was resolved, if anything.


033 Jiang Agree, revise as shown. (9/9 Wu, Pawski, Howe, Hoptay, Garrison, Ballard, DJ, Brannan, NKO)


034 Jiang Agree, revise as shown. (10/10 Wu, Pawski, Berner, Howe, Hoptay, Garrison, Ballard, DJ, Brannan, NKO)


035, 046, 048 Jiang Agree to change LNG to RLG throughout. (11/11 Wu, Pawski, Berner, Howe, Hanskat Hoptay, Garrison, Ballard, DJ, Brannan, NKO)


036 Mash Comment is “This needs careful though to ensure alignment and is the key.

Does the analyst knock out the insulation in the 'wet' zone and apply a rapid temperature gradient? or do something more sensible.” Response is no change, this is topic of a future TN for RLG spill load defintion. (5/5) Garrison, Ballard, Pawski, Brannan, Hjorteset

Agree, revise as shown. (13/13 Wu, Pawski, Hoff, Berner, Howe, Hanskat, Hjorteset, Hoptay, Garrison, Ballard, DJ, Brannan, NKO)


042 Pawski – Sat (superceded) 043 Jiang – delete as shown. Sunday 3-18 - (11/11 Wu, Pawski, Hoff, Berner, Howe, Hjorteset, Hoptay, Garrison, Ballard, DJ, NKO)


044 & 047 Garrison negative – not resolved 044 Mash & 047 Jiang - Revise as shown. Tue 3-20 - (3/3) Garrison, Ballard, Pawski & Sun 3-19 - (12/12 Wu, Pawski, Hoff, Berner, Howe, Hjorteset, Hoptay, Garrison, Ballard, DJ, Brannan, NKO) 045 Hoptay – withdrawn


049 Mash & Ballard negatives – not resolved Agree, revise as shown. Tue 3-20 - (3/3) Pawski, Garrison, Ballard)


non-linear FE material constitutive models •


050 Jiang Agree, revise as shown. (11/11 Wu, Pawski, Hoff, Howe, Hjorteset, Hoptay, Garrison, Ballard, DJ, Brannan, NKO)


. There are several commercially available FE programs that are adequate


078 Khalifa: Combine 1st two sentences pg 8 lines 32, 33 to read: 21-Oct-2013 Phoenix


This moment-curvature approach would only be suitable for consideration of the steady state

linear thermal gradient case and should also be used only for preliminary design.


055 Wu Withdrawn Monday March 19. 056 Jiang Revised as shown. (13/13 Wu, Pawski, Hoff, Berner, Howe, Hanskat, Hjorteset, Hoptay, Garrison, Ballard, DJ, Brannan, NKO) 057 & 058 Mash not addressed. Leave as is.

21-Oct-2013 Phoenix


059 Jiang Revise as shown. (12/12 Wu, Pawski, Berner, Howe, Hanskat, Hjorteset, Hoptay, Garrison, Ballard, DJ, Brannan, NKO)


060 Jiang No change. (13/13 Wu, Pawski, Hoff, Berner, Howe, Hanskat, Hjorteset, Hoptay, Garrison, Ballard, DJ, Brannan, NKO)


Pg 9 line 24 Hoff: editorial Change last sentence to read 21-Oct-2013 Phoenix


062 Wu Comment is “Change the word, “may” to “should”, since the reinforcing and pre-stressing steel need to be included in the nonlinear FEM cracking analysis in the liquid spill case.”

Page 9: [35] Comment [NKO67] krstulovic 1/14/2014 3:42:00 PM

063 Mash Comment is “Seems out of place in terms of the flow of the document.” 063 Oliver negative - not addressed




Thermal Corner Protection Design & Crack AnalysisDesign and Analysis 2

of the TCP Embedment Zone 3

4


6


8


10

INTRODUCTION 11


Protection (TCP) embedment zone. The purpose of this document Technical Note is to provide further 13

guidance on for (a) design and analysis of theTCP detailing anchorage and (b) TCP liquid-tightness 14

analysis in the TCP embedment zone (i.e., crack width calculations). The document therefore consists 15

of two parts: 16

17


Chair

Piotr D. Moncarz

Secretary

Junius Allen

Thomas A. Ballard*

Dale Berner

Mike S. Brannan

Hamish Douglas*

Jeffrey Garrison*

Charles S. Hanskat

Alan D. Hatfield

Humayun Hashmi

Kare Hjorteset

George C. Hoff*

Richard Hoffmann

John Holleyoak

Joseph M. Hoptay*

Thomas R. Howe

Dajiu Jiang

Jameel U. Khalifa

Nicholas A. Legatos

Praveen K. Malhotra

Keith A. Mash*

Stephen W. Meier

Robert W. Nussmeier

Rolf P. Pawski*

Ramanujam S. Rajan

William E. Rushing

Robert W. Sward

Eric S. Thompson

Sheng-Chi Wu

Comment [LE1]: 001 Brannan part 1 Revise title to “Design and Analysis of the TCP Embedment Zone.” Monday 3-19: (9/9) Wu, Pawski, Howe, Hoff, Hoptay, Garrison, Ballard, Brannan, NKO) 002 Pawski Revise title to read “Liquid tightness Anaysis at the TCP.” Saturday 17 - (8/8) Brannan, Ballard, Hoptay, Howe, Juang, NKO, Pawski, Wu Note superceded by Monday vote.

Comment [LE2]: 003 Hoff comment not addressed. “Useful but seems to be a little bit too long for a TechNote”

Comment [LE3]: 004 Mash comment not addressed. “Comments included in document. There are couple of blind spots which we ought to address via telecon”

Comment [LE4]: 005 Pawski withdrawn Revisit sugested organization at time of final editing. (Sunday 3-18: present - Wu, Pawski, Hoff, Berner, Howe, Hjorteset, Hoptay, Garrison, Ballard, Brannan, NKO)

Comment [LE5]: 001 – Brannan part 2. Page 1 lines 11-14. Revise to read as shown. (9/9) Wu, Pawski, Howe, Hoff, Hoptay, Garrison, Ballard, Brannan, NKO)





3






9




13

Comment [LE6]: Part of Ref. 001: T.Ballard negative . Draft of new paragraph for committee consideration agreed to at Minneapolis – 2013.04.13.



CONTENTS 1

2

INTRODUCTION 3

4


1.1—General 6

1.2—Embedment Plate and Anchor StudsAnchorage 7

8


2.1—General 10





15

APENDICES 16

17


19

REFERENCES 20

21





2

a) b)

3


5

1.1—General 6






12





designed to remain intact functional and in place during the spill condition. 17

Comment [LE8]: 006 Wu Comment is “Need to include the secondary bottom plate in the model.” On Figure 1a add a detail circle to identify where Figure 1b is located.

Comment [LE9]: Part of 001 Ballard negative : I believe the embedment zone should extend at least to the bottom of the TCP plate since migration of cracks in this zone would compromise the leak tightness of the detail.

Comment [LE10]: Part of 021 Hoptay negative 2013 04.13 Minneapolis: annotate sketch to state "2 rows of anchors minimum."

Comment [LE11]: Oliver – editorial Pg 4 line 8: addd "the" as shown. 2013.10.21 Phoenix

Comment [LE12]: 007 Mash Commetn is “Suggest replacement of 'intact' with functional.” Agree - (8/8) Garrison, Howe, Hoff, Ballard, Pawski, Brannan, Hoptay, Wu



1



the spill and by limiting cracking crack width on the inside face of the wall in the vicinity of the TCP embedment 4

zone plate during a spill. 5

6



provided for the evaluation of the embedment plate, design of the anchor studsanchors and attachment to the 9

concrete wall, and for the evaluation of cracking in the embedment zone. Material selection and material 10

requirements used in the design of the 9% nickel-steel TCP and secondary bottom plates may be performed in 11

accordance with API 620. 12

13

1.2—Embedment Plate and Anchor StudsAnchors 14




18


• Finite element analyses is are performed first to determine forces resulting from thermal gradients applied to 20

the TCP, TCP embedment plate and outer concrete wall. Loads should include all loads that the TCP and the 21


hydrostatic pressure and hydrodynamic pressure due to seismic aftershock (SSEAFT). Starting Initial 23

temperatures and conditions should be selected so as to maximize the demand under consideration. For 24

instance for the design of the anchors the critical condition is likely to be warm wall (summer) and a cold 25

shocking thermal shock to the embedembedment. Whereby, thereby creating maximum straining strain across 26

the section. 27


wall. This should include spill levels a spill level located just below the top of the TCP which will place the 29

largest temperature differential between the TCP and the concrete. wall as well as various A minimum of 30

three levels should be considered for spill levels above the TCP, up to the1) at top of the TCP embedment 31

zone, 2) at mid-height, and 3) at full spill level. 32

Comment [LE13]: 009 Hoptay Withdrawn

Comment [LE14]: 2013.04.13 Minneapolis – Pg 5 line 2 change "liquid" to "liquid tight."

Comment [LE15]: 008 Mash Response: “Limiting values are covered elsewhere and do not need to be repeated here.”Revise to read for clarity as shown. (8/8) Garrison, Howe, Hoff, Ballard, Pawski, Brannan, Hoptay, Wu

Comment [LE16]: 011 Mash Page 5 line 4 - no action, included with response 010


Comment [LE18]: 010 Hoptay Delete last sentence. 012 Hoptay Comment found non-persuasive. Dealt with already elsewhere by deleting last sentence (Hoptay comment). (8/8) Garrison, Howe, Hoff, Ballard, Pawski, Brannan, Hoptay, Wu

Comment [LE19]: Part of Ref no. 021 - 2013.04.13 Minneapolis. Edit TN to use "anchors" rather than "bolts" or "studs" throughout.

Comment [LE20]: 013 Wu Agree, revise as shown. (9/9) Wu, Pawski, Howe, Hoff, Hoptay, Garrison, Ballard, Brannan, NKO)

Comment [LE21]: 014 Mash Agree in principle, but this is already addressed by requirement of analysis for thermal gradient. Future TN will define design conditions. (6/6) Garrison, Howe, Ballard, Pawski, Hoptay, Wu

Comment [LE22]: 015 Mash

Comment [LE23]: 016 Wu Agree, revise as shown. (9/9) Wu, Pawski, Howe, Hoff, Hoptay, Garrison, Ballard, Brannan, NKO)

Comment [LE24]: 2013.04.13 Minneapolis. Revise last three sentences as shown.

Comment [LE25]: 017 and 018 Mash negative . Should be 5 locations, not 3 as shown. Revise as shown. (6/6) Garrison, Howe, Ballard, Pawski, Hoptay, Wu






4


• The TCP embedment plate should be cast in place and mechanically anchored to the concrete wall or by other 6

suitable means (e.g., ???). Anchor bolts (studs) should be placed in continuous (circumferential) rows along 7

the concrete wall. 8

• When the embedmentThe embedment plate is should be anchored with circumferential studs, a minimum of 9

two rows of circumferential studsanchors should be used to anchor the embedment plate. The spacing 10

between mechanical anchors (studs) should be limited to ensure that the axial, bending and shear stresses in 11

the embedment plate are within the allowable stressesanchors has significant impact on the TCP embedment 12

performance and thus should be considered in design. 13






19

1.2.3 - STEP 3: Anchor BoltAnchorage Design 20

The mechanical anchorage (anchors bolts (or studs) and concrete at the anchor location should meet the 21


• The steel strength of the anchor in tension should be as per ACI 350 section D.5.1. 23

• The concrete breakout strength of the anchor in tension should be checked per ACI 350 section D.5.2. 24

• The pullout strength of the anchor in tension should be checked per ACI 350 section D.5.3. 25

• The steel strength of the anchor in shear should be checked per ACI 350 section D.6.1. 26

• The concrete breakout strength of the anchor in shear should be checked per ACI 350 section D.6.2. 27

• The concrete pry-out strength of the anchor in shear should be checked per ACI 350 section D.6.3. 28

• Interaction of tensile and shear forces should be checked per ACI 350 section D.7. 29

30




34



Comment [LE26]: 019 Mash. Comment is “Suggest inclusion of the words 'in the absence of a defined leak rate all intermediate levels shall be investigated. In practice this may be achieved by running a number spill levels to so as adequately bound the demands. Typically 5 levels may be used dividing the spill height into 4 equal units. In addition levels in close proximity to the TCP must be investigated.” No change. The response will be covered in the future TN on the design basis (DJ comment).

Comment [LE27]: 020 Mash negative - resolved Comment is “Must be achieved through mechanical anchoring.” Agree, revise as shown. Tues 3-20: (5/5) Garrison, Ballard, Pawski, Brannan, NKO

Comment [LE28]: 020 Hoptay negative - resolved Delete 2nd sentence as it is essentially same as next sentnece. Agree 2013.04.13 Minneapolis.

Comment [LE29]: 021 Mash negative - resolved Comment is “When? Embedment must be anchored with studs?” Revise as shown, and consider further discussion in future TN. Tues 3-20: (5/5) Garrison, Ballard, Pawski, Brannan, NKO 021 Hoptay negative - resolved – addressed here & Fig. 1.1(b)

Comment [LE30]: 022 Mash Comment is “Consider use of ACI 349? More stringent with respect to anchors.” Non-persuasive. Tues 3-20: (4/4) Garrison, Ballard, Pawski, Brannan

Comment [NKO31]: 026 Mash Non-persuasive, but edit bullet point as shown. Tuesday 3-20: (5/5) Garrison, Ballard, Pawski, Brannan, NKO

Comment [LE32]: 023 Mash Comment is “Consider use of ACI 349? More stringent with respect to anchors.” Non-persuasive. . Discussed March 19 and decided that ACI 350 is appropriate. Also, there does not appear to be any difference between ACI 350 and ACI 349 on this topic; Garrison to verify. Tues 3-20: (4/4) Garrison, Ballard, Pawski, Brannan.

Comment [NKO33]: 024 Hoptay Withdrawn, already addressed.

Comment [LE34]: 027 Wu Comment is to add bullet point to read “In moderate or high seismic site, the design strength of anchors should be reduced to 75% of the design strength per Section D.3.3.3.” Withdrawn, Monday 3-19..

Comment [LE35]: 025 Wu Revise 1.2.3 as shown and as new section 1.2.4 as shown. (9/9) Wu, Pawski, Howe, Hoff, Hoptay, Garrison, Ballard, Brannan, NKO) ... [1]




2

2.1—General 3

When subjected toDuring the spill condition, the secondary concrete wall will be loaded withsubjected to high 4

large forces thermally inducedand mechanical forces. These forcesdue to mechanical loads and thermal effects 5

that will lead to significant wall cracking. During the spill condition, integrity Integrity of the concrete wall must 6

be maintained. As specified in the ACI 376 Code, the wall should remain liquid and substantially vapor tight. 7

More specifically: 8

1) ACI 376 section 6.3.36.3.2: “Unless a leak-tight membrane/liner has been used, a minimum portion of the concrete 9


a) A compressive zone of either 10% of the section thickness or 3.5 in., whichever is greater should shall be 11

provided; and 12


14

2) ACI 376 section 6.3.46.3.3: “Calculated crack widths should be considered at TCP embedment when cracking 15

would result in liquid product migrating behind the TCP and compromising its effectiveness. The embedment zone 16

should extend a minimum of two times the wall thickness above the TCP anchorage. Calculated crack widths should 17

not exceed 0.008 in. within the TCP embedment zone. Recommendations for crack width calculations are provided in 18

8.1.1.8.” 19

20


calculated crack widths as per EN 1992-1-1. 22

23


25

26



29



Steps include: 32



Comment [LE36]: 028 Mash negative - resolved

Comment [LE37]: 029 Pawski Comment is mostly superceded by 028 changes, but add “secondary” from Sat 9-17 discussion. (9/9 – see Saturday attendance)

Comment [LE38]: 028 Hoptay negative - resolved Withdrawn 2013.04.13 Minneapolis

Comment [LE39]: 028 Mash negative - resolved 030 & Comment is “Suggest using nomenclature such as demand as opposed to load.” Agree in part, revise to read as shown Tues 3-20 - (4/4) Garrison, Ballard, Pawski, Brannan


Comment [LE41]: 031 Mash Tues 3-20 - (4/4) Garrison, Ballard, Pawski, Brannan

Formatted: Superscript


Comment [LE43]: 032 Mash Tues 3-20 - (4/4) Garrison, Ballard, Pawski, Brannan 032 Mash, Wu & Garrison negatives – not resolved – Requires discussion as it is not clear what was resolved, if anything.

Comment [LE44]: 033 Jiang Agree, revise as shown. (9/9 Wu, Pawski, Howe, Hoptay, Garrison, Ballard, DJ, Brannan, NKO)

Comment [LE45]: 034 Jiang Agree, revise as shown. (10/10 Wu, Pawski, Berner, Howe, Hoptay, Garrison, Ballard, DJ, Brannan, NKO)



3) Application of LNG RLG spill or external fire temperature loads obtained by considering rate of leakage 1

into annular space and exposed to transient temperatures 2

3

4) Continue analysis until steady state final temperatures are reachedtemperature distribution is obtained 4


6



1) Modeling 9

2) Apply To establish intiial conditions, apply dead load, prestressing, normal opearting thermal and any 10

other steady state load, then iterate to equilibrium 11

3) Check crack widths 12

4)3) Apply LNG RLG spill temperature transients and iterate to equilibrium 13

5)4) Compute Beginning with the intial condition defined in step 2, compute crack widths at times 14

time-steps throughout the transient where reinforcing strains are the greatest and exceed the material yield 15

stress. Furthermore, at each time step determine the adequacy of the compression zone. 16

6)5) Repeat for all thermal load cases 17

18

19


Calculation of cracking and crack widths for the LNG RLG spill condition and/or external fire condition shall be 21

performed using a rational engineering analysis. Both linearLinear or non-linear finite element (FE) analysis 22

should be used as discussed in 2.3.1 and 2.3.2. 23

24


In this case a non-linear finite element program is used in conjunction withby: 26

• non-linear FE material constitutive modelsImplementing the non-linear material constitutive relationship 27

for concrete and steel, and 28


• Performing FE analysis post process to calculate characteristic crack width, calculations as per following 30

a method such as Eurocode EN 2 1992-1-1 (see section 2.5 in this TN). 31

Either a proprietary or a one of several commercially available FE programs can be used. There are several 32

commercially available FE programs that are adequate for performing this calculation. A few of more widely 33

Comment [LE46]: 035, 046, 048 Jiang Agree to change LNG to RLG throughout. (11/11 Wu, Pawski, Berner, Howe, Hanskat Hoptay, Garrison, Ballard, DJ, Brannan, NKO)

Comment [LE47]: 036 Mash Comment is “This needs careful though to ensure alignment and is the key. Does the analyst knock out the insulation in the 'wet' zone and apply a rapid temperature gradient? or do something more sensible.” Response is no change, this is topic of a future TN for RLG spill load defintion. (5/5) Garrison, Ballard, Pawski, Brannan, Hjorteset

Comment [LE48]: 037 Mash & 037 Garrison negative – not resolved Comment is “More explanation required regarding semi coupled analysis (fully coupled not needed)” Respons is that comment is not clear. (5/5) Garrison, Ballard, Pawski, Brannan, Hjorteset

Comment [LE49]: 037 Jiang Agree, revise as noted. (11/11 Wu, Pawski, Berner, Howe, Hanskat Hoptay, Garrison, Ballard, DJ, Brannan, NKO)

Comment [LE50]: 039 Jiang Agree, revise as shown. (11/11 Wu, Pawski, Berner, Howe, Hanskat Hoptay, Garrison, Ballard, DJ, Brannan, NKO)

Comment [LE51]: 040 Mash Comment is “Thermal loads from spillage need porting across to structural model.” Response is that no change is needed because the comment is implicit in the step-by-step procedure. (5/5) Garrison, Ballard, Pawski, Brannan, Hjorteset

Comment [LE52]: 041 Jiang & 041 Garrison negative – not resolved Agree, revise as shown. (13/13 Wu, Pawski, Hoff, Berner, Howe, Hanskat, Hjorteset, Hoptay, Garrison, Ballard, DJ, Brannan, NKO)

Comment [LE53]: 042 Pawski – Sat (superceded) 043 Jiang – delete as shown. ... [2]

Comment [LE54]: 044 & 047 Garrison negative – not resolved 044 Mash & 047 Jiang - Revise as shown. Tue 3-20 - (3/3) Garrison, Ballard, Pawski ... [3]

Comment [LE55]: 049 Mash & Ballard negatives – not resolved Agree, revise as shown. Tue 3-20 - (3/3) Pawski, Garrison, Ballard)

Comment [LE56]: 050 Jiang Agree, revise as shown. (11/11 Wu, Pawski, Hoff, Howe, Hjorteset, Hoptay, Garrison, Ballard, DJ, Brannan, NKO)

Comment [LE57]: 078 Khalifa: Combine 1st two sentences pg 8 lines 32, 33 to read: 21-Oct-2013 Phoenix



used non-linear FE programs available in the United States include ANSYS1, ABACUS2, ADINA 3 and NX 1

NASTRAN4. There are also many proprietary FE programs that have been designed specifically to consider the 2

cracking problem for reinforced concrete containment structures. 3

4

2.3.2 Linear SimplifiedFE Analysis 5

Linear Simplified FE analysis is usually used only in the preliminary design. Two possible approaches include: 6


analysis where the reduction in stiffness due to cracking is considered. However, it should be noted that this 8

is a very time consuming process and is thus usually only used in the preliminary design. 9

b) Moment-Curvature Approach: The moment curvature response isproperties are obtained using a concrete 10

section analysis program by implementing in conjunction with: 11



• This moment-curvature approach would only be suitable for consideration of the steady state linear 14

thermal gradient case and should also be used only for preliminary design. 15

16


The linear or non-linear FE model will usually be axisymmetric or a small slice 3D model of the containment for 18

consideration of the general case of a symmetric LNG RLG spill loads and a larger sector separate model for 19

consideration of the LNG RLG spill case where non-axisymmetric discontinuities are considered. 20

21



buttresses. If the liners are considered as structural members, theyLiners should be included tooif they are 24

considered as structural members. 25

26




30



33

2.3.4 Loading 34

Comment [LE58]: Part of 056 Jiang

Comment [LE59]: 051 Mash Comment is “We should exercise caution. A number of vendors have historically used this approach and we ought not to say it should not be used.” No change, non-persuasive. Tue 3-20 - (3/3) Pawski, Garrison, Ballard) 053 Jiang Deleted sentence in its entirety (11/11 Wu, Pawski, Hoff, Berner, Howe, Berner, Hjorteset, Garrison, Ballard, DJ, Brannan, NKO) 052 Hoptay Withdrawn Monday March 19. 054 Mash Not addressed, but deletion of the sentence should satisfy comment that “contractors should be able to choose design method.

Comment [LE60]: 055 Wu Withdrawn Monday March 19. 056 Jiang Revised as shown. (13/13 Wu, Pawski, Hoff, Berner, Howe, Hanskat, Hjorteset, Hoptay, Garrison, Ballard, DJ, Brannan, NKO) 057 & 058 Mash not addressed. Leave as is.

Comment [LE61]: 077 Khalifa: Pg 9 line 19 add "separate" before "model." 21-Oct-2013 Phoenix

Comment [LE62]: 059 Jiang Revise as shown. (12/12 Wu, Pawski, Berner, Howe, Hanskat, Hjorteset, Hoptay, Garrison, Ballard, DJ, Brannan, NKO)

Comment [LE63]: 060 Jiang No change. (13/13 Wu, Pawski, Hoff, Berner, Howe, Hanskat, Hjorteset, Hoptay, Garrison, Ballard, DJ, Brannan, NKO)

Comment [LE64]: Pg 9 line 24 Hoff: editorial Change last sentence to read 21-Oct-2013 Phoenix

Comment [LE65]: 061 Jiang - withdrawn

Comment [LE66]: 062 Wu Comment is “Change the word, “may” to “should”, since the reinforcing and pre-stressing steel need to be included in the nonlinear FEM cracking analysis in the liquid spill case.”



The LNG RLG spill case should include hydrostatic pressure from the LNG RLG fluid in the annular space. 1

Dead load, live load and steady state temperatures coincident prior to the LNG RLG spill, shall be applied prior 2

application of the temperature transients. 3

4

5






11

2.3.5 Heat Thermal Analysis 12


temperatures. The LNG RLG spill is applied internally to the containment and the adjacent tank fire case is 14

applied externally. The LNG RLG spill load case should also include the vapor temperature generated from the 15

spill above the liquid level. Nodal temperature time histories are computed and mapped to the structural finite 16

element model. The concrete wall internal temperatures from the thermal model are also typically used to 17

determine the for selection and placement of normal (A615), cold-temperature (A706) and cryogenic 18

reinforcingreinforcement complying with the requirements of ACI 376. 19

20

The LNG RLG hydrostatic pressure and temperature time histories are applied gradually and the iterative solution 21


23

24


26


Recognized concrete constitutive models shall should be used in the analysis. Examples include material models 28


account for reduction in tensile stiffness after cracking of concrete including the effects of tension stiffeninglinear 30

tension stiffness effects with tension failure or reduction in stiffness at the cracking strain of the concrete. 31

32


Comment [NKO67]: 063 Mash Comment is “Seems out of place in terms of the flow of the document.” 063 Oliver negative - not addressed

Comment [LE68]: 064 Pawski Change as shown. (9/9 – see Saturday attendance)

Comment [LE69]: 065 Mash Comment is “Vapour temperature? This will be affected by convection etc? How is this to be determined.” This is a general comment/question. No new text proposed. Leave as is and consider the issue in the future.

Comment [LE70]: 066 Hoptay Revise as shown. FUTURE ACTION: revisit use of A706 materials in ACI 376 section 4.7.2. Mon 3-19 -(9/9) Wu, Pawski, Howe, Hoff, Hoptay, Garrison, Ballard, Brannan, NKO)

Comment [LE71]: Pg 10 line 28 Hoff: editorial Use "should" in non-mandatory TN 21-Oct-2013 Phoenix

Comment [LE72]: 079 Khalifa: Revise pg 10 line 30 to read: 21-Oct-2013 Phoenix

Comment [LE73]: Typo (see page 8)



1


3


σc/fc = (k.n-n2)/[1 + n.(k-2)] (1)

where: 5



n = εc/εc1 (both < 0.0) (4)


k = (1.1Ec,nom).εc1/fc (6)




6

The tension side of the �c – �cfc - εc material relationship may be taken as: 7



where: 8

fctm = 0.30fc2/3 (N/mm2) (13)

9

In reality, the tensionTension softening is usually introduced following rupture of the concrete in order to provide 10

for numerical stability in the solution. 11

12


Comment [LE74]: 067 Wu Comment is “Need to specify the value for ɛcu = 0.3%” Insert in Figure 7 if editing is possible.


Comment [LE76]: 068 Pawski - withdrawn Add a tension “tail” to the figure. (Sunday: present - Wu, Pawski, Hoff, Berner, Howe, Hjorteset, Hoptay, Garrison, Ballard, Brannan, NKO)

Comment [LE77]: Pg 11 line 7 Hoff & Widianto: editorial Stress symbols are missing 21-Oct-2013 Phoenix

Comment [LE78]: 069 Pawski - withdrawn Editorial, change as noted. (Sunday: present - Wu, Pawski, Hoff, Berner, Howe, Hjorteset, Hoptay, Garrison, Ballard, Brannan, NKO)





3





4

5



8


Since forFor the LNG RLG spill case, it is critical that cryogenic liquid be prevented from migrating behind the 10

TCP, and conservative assumptions should be used when selecting material properties. 11

12





17






Comment [LE79]: 070 Wu Comment is “Need to specify the limiting ultimate steel strain, ɛu.

(Note that the ultimate strains for steel are not discussed in the Code. Committee should discuss, and set the allowable limits. ) CONSIDER FOR FUTURE ACTION


Comment [LE81]: Pg 12 lines 10 & 11 Hoff: editorial. Change to read 21-Oct-2013 Phoenix



1





6




wk = βsrmεsm (14)




mm, or 13



16


srm = 50 + 0.25k1kk2φ/ρr (15)

where: 18






(19)



(20)

19






25

Comment [LE82]: 072 Mash Comment is “Need to look into this area in more detail and will advise through discussion.” This is a general comment.. No new text proposed and no additional input provided sine the last ballot. Leave as is and consider the issue in the future.

Comment [LE83]: 071 Wu Monday march 19 withdrawn because 2004 edition equations will be used. 074 Garrison negative and 2012.10.13 Phoenix Wu negative – not addressed . To update equations to 2004 edition.




REFERNCES 1









2





defined in 8.1.1.7.




precompression.



effectiveness.


TCP anchorage. Calculated crack widths should not exceed 0.0040.008 in. within the TCP

embedment

zone.











widths.

Comment [LE85]: 075 Pawski Resolution is: This is an editorial comment. Appendix A is a “copy” of what is in the Code. It is provided for reader’s convenience. Reconcile duplication between 2.1 and Appendix A, such as removing Appendix A

Comment [LE86]: 074 Garrison Response is: 1)Definition of the spill loading is beyond the scope of the current TN and will thus be addressed in a separate document defining LNG spill conditions. 2)Update and change the model to the latest 2004 edition of the Eurocode,

(Sunday: present - Wu, Pawski, Hoff, Berner, Howe, Brnnan, Hjorteset, Hoptay, Garrison, Ballard, NKO)

Comment [LE87]: 073 Wu Comment is “Change the crack width of 0.004 in to 0.008 in.” Section 6.3.4 is text copied over from the Code and provided herein for the user’s benefit. Text must remain as-is to properly reflect the Code requirement. Changes to be addressed only if/once the Code has been changed accordingly.





steel.




crack widths.






crack width limits.










widths.

1

Comment [LE88]: 076 Thompson Comment is “Remove the last sentence from the Code Commentary: “Due to the non-homogeneous nature of concrete, calculated crack width values measured in the field will vary from calculated values, and therefore they cannot be directly compared to calculated crack widths.”” Section R8.1.1.7 is text copied over from the Code and provided herein for the user’s benefit. Text must remain as-is to properly reflect the Code requirement. Changes to be addressed only if/once the Code has been changed accordingly.



1

2

[1] Comment [LE35] rpawski 1/14/2014 3:42:00 PM

025 Wu Revise 1.2.3 as shown and as new section 1.2.4 as shown. (9/9) Wu, Pawski, Howe, Hoff, Hoptay, Garrison, Ballard, Brannan, NKO) 2013.10.21 Phoenix – consensus was new section 1.2.4 is not needed and should be deleted in its entirtety.


042 Pawski – Sat (superceded) 043 Jiang – delete as shown. Sunday 3-18 - (11/11 Wu, Pawski, Hoff, Berner, Howe, Hjorteset, Hoptay, Garrison, Ballard, DJ, NKO)


044 & 047 Garrison negative – not resolved 044 Mash & 047 Jiang - Revise as shown. Tue 3-20 - (3/3) Garrison, Ballard, Pawski & Sun 3-19 - (12/12 Wu, Pawski, Hoff, Berner, Howe, Hjorteset, Hoptay, Garrison, Ballard, DJ, Brannan, NKO) 045 Hoptay – withdrawn

ACI 376 - Technical Note on TCP 2nd ACI 376-C Ballot on TCP Technical Note Last Update: 4 / 13 / 2013 - 376-D meeting in Minne apolis Balloted from 1 /2 0 to 3 / 10 / 2013 2012

Page 1 of 15

Technical Note on TCP (TCP – TN) Approved Sections Section Approved with Comments to be resolved Negative Vote 2013.04.13 376-C Minneapolis

Members that Voted Voting Members that did not vote 2013.04.13 – 376 D Minneapolis

Voting Members that voted (79%) 1. Hjorteset, Kare 2. Brannan, Mike S 3. Allen, Junius 4. Ballard, Thomas A 5. Garrison,Jeffrey 6. Hatfield,Alan 7. Hoff, George C 8. Hoffmann, Richard A

9. Hoptay,Joseph 10. Howe, Thomas R 11. Jiang,Dajiu 12. Khalifa, Jameel U 13. Krstulovic-Opara, Neven 14. Legatos, Nicholas A 15. Malhotra,Praveen 16. Mash,Keith 17. Moncarz, Piotr D

18. Nussmeier, Robert W 19. Oliver, William H 20. Pawski, Rolf P 21. Rushing, William E 22. Wu,Sheng-Chi

Associate Members

� Roetzer, Josef � Widianto, Widianto

23. Hanskat, Charles S 24. Douglas,Hamish 25. Hashmi, Humayun 26. Rajan, Ramanujam S 27. Sward,Robert 28. Thompson,Eric

Members Attending: 1. Brannan, Mike 2. Hjorteset, Kare 3. Ballard, Thomas 4. Garrison. Jeffrey 5. Hoff, George 6. Hoptay, Joseph 7. Howe, Thomas 8. Khalifa, Jameel

9. Pawski, Rolf 10. Hanskat, Charles 11. Roetzer, Josef


Ref No. Text Author Comments

2ND BALLOT: RESPONSE being BALLOTED


COMMENTS

Suggested Change

1

Brannan

I am not sure the TCP can be constructed as detailed because it would be impossible to weld both sides of the TCP plate to the embedment plate. It would also be better to radius the sharp corner shown on the TCP plate. Would it be possible to simply lap weld the TCP plate to the embedment plate? Should the embedment zone also extend 2X the wall thickness below the embedment plate?

The comment found partially persuasive. It is not intended that the TN covers TCP construction details. Furthermore, welding requirements should be as per API 620. Address the comment by revising text on page 1 lines 11-14: INTRODUCTION ACI 376 provides requirements and some basic recommendations for the design of the Thermal Corner Protection (TCP) embedment zone. The purpose of this Technical Note document is to provide further guidance for on (a) design and analysis of the TCP embedment anchorage detailing and (b) TCP liquid-tightness analysis (i.e., crack width calculations) in the TCP embedment zone (i.e., crack width calculations). Update title to read: Thermal Corner Protection Design & Crack Analysis

Agree – Jiang, Moncarz, Hoff, Khalifa, Rushing, Humayum, Howe, Wu, Malhotra, Garrison, Mash, Douglas, Allen, Legatos, Widianto, Hoptay, Roetzer, Oliver, Pawski, Hatfield, Brannan, Krstulovic

2014.01.08 TCP text negatives requiring discussion: ~ Ref. no. 1, Comment 6 - Ballard proposed addition to Introduction ~Ref. no. 1, Comments 8, 9, 10 - edit Fig. 1.1

Agree with comment - Hoffman

With Respect to M. Brannan’s Comments, it is my opinion this can be welded from both sides, by making this attachment weld first and then going on with the construction. There will be a “fill-in plate that is required to build up the TCP side wall. I have no objection to the proposed changes

UPDATE NOTES (2013.01.08) a) 2013.10.21 Phoenix discussion items are on pg 14 of 15. b) See the following Ref. no. for NEGATIVES REQUIRING FURTHER DISCUSSION:

1, 18, 32, 37, 41, 44 & 47, 49, 63, 74. c) Pg. 15 of 15 has Legatos negative that requires further discussion.


Page 2 of 15





COMMENTS

Suggested Change

Design and Analysis of the TCP Embedment Zone

Disagree - Ballard I believe the embedment zone should extend at least to the bottom of the TCP plate since migration of cracks in this zone would compromise the leak tightness of the detail. NOTED in Fig 1.1(b).

2013.04.13 – 376-C Minneapolis: Agreed and that Tom Ballard to propose wording. From TB email dated 2013.04.15: “Although the code and this TN do not explicitly define leak tightness criteria for the anchorage and embeddment, the design and construction of these components should be carried out with the clear objective of protecting the corner joint from product temperatures.” Note – added to Introduction on pg 2.


Mash

Suggest rewording and removing 'substantially' link to definitive acceptance criterion to avoid confusion. Suggest exact acceptance criterion words such as - under the spillage condition the crack widths on the inside face shall be limited to xx, compressive stresses in both the meridional and vertical directions shall be minimum y, max z,

The comment found partially persuasive. Limiting values are covered in the Code (e.g., compressive zone size, crack width limits, etc.) and do not need to be repeated here. Revise text for clarity to read: “The concrete wall above the TCP, directly exposed to spilled product, should remain liquid and substantially vapor tight during the spill condition. This is assured by maintaining a compression zone in the wall exposed to the spill and by and limiting the crack width cracking on the inside face of the wall in the vicinity of the TCP embedment plate zone during a spill.”

Agree – Jiang, Moncarz, Hoff, Rushing, Hoffman, Humayum, Howe, Wu, Ballard, Malhotra, Garrison, Mash, Douglas, Allen, Legatos, Widianto, Hoptay, Oliver, Pawski, Hatfield, Brannan, Krstulovic

Agree with editorial changes – Khalifa, Roetzer, Krstulovic

“The concrete wall above the TCP, directly exposed to spilled product, should remain liquid tight and substantially vapor tight during the spill condition. 2013.04.13 – 376-C Minneapolis: Agree with comment.

10 Page 5 Line 10

Hoptay

The material selection is temperature dependent and to state that the TCP is 9% Ni infers an LNG tank and not RLG tanks in general.

The comment found persuasive. Change text to read: “Material selection and material requirements used in the design of the 9% nickel-steel TCP and secondary bottom plates may be performed in accordance with API 620.”

Agree – Jiang, Moncarz, Hoff, Khalifa, Rushing, Hoffman, Humayum, Howe, Wu, Ballard, Malhotra, Garrison, Mash, Allen, Legatos, Widianto, Hoptay, Roetzer, Oliver, Pawski, Hatfield, Brannan, Krstulovic

Note – This sentence is deleted based on Comment LE15 in text markup.


Page 3 of 15





COMMENTS

Suggested Change

Disagree - Douglas This may still be ambiguous as API 620 does not refer to TCP

2013.04.13 – 376-C Minneapolis: Change last line to read: “… should be in accordance with API 620.”


Mash

Assurance is through cracks + compression zones not through cracks alone. Crack width limitation in the vicinity of the embed is to prevent liquid migration behind the TCP area.

No action. Covered in previous response.


13 P. 5, Line 15

Wu

add “seismic Aftershock load”. The comment found persuasive. Change text to read:

“The TCP embedment plate should be designed to resist radial and vertical loads and moments resulting from both (a) thermal gradients, as well as (b) mechanical forces that develop during a spill condition including SSEAFT. The TCP design consists of the following steps:”



Mash

Starting temperatures and conditions should be selected so as to maximise the demand under consideration. For instance for the design of the anchors the critical condition is likely to be warm wall (summer) and cold shocking to the embed. Whereby creating maximum straining across the section.

The comment found persuasive. Add proposed text to line 22: 1.2.1 – STEP 1: TCP Load Definition Finite element analyses is performed first to determine forces resulting from thermal gradients applied to the TCP, TCP embedment plate and outer concrete wall. Loads should include all loads that the TCP and the embedment will be subjected to including thermal, prestress, creep, concrete shrinkage, internal pressure, hydrostatic pressure and hydrodynamic pressure due to seismic aftershock. Starting temperatures and conditions should be selected so as to maximize the demand under consideration. For instance for the design of the anchors the critical condition is likely to be warm wall (summer) and cold shocking to the embed. Whereby creating maximum straining across the section.

Agree – Jiang, Moncarz, Khalifa, Rushing, Hoffman, Howe, Wu, Ballard, Malhotra, Garrison, Mash, Legatos, Widianto, Hoptay, Roetzer, Hatfield, Krstulovic

2013.04.13 – 376-C Minneapolis: To address all comments, replace last three sentences with: “ Initial temperatures and conditions should be selected to obtain the most severe design conditions. For instance the most severe condition for design of the anchors is likely to be a warm wall (summer) with a cold thermal shock to the embedment.”

Agree with editorial changes - Hoff

For instance, for the design of the anchors, the critical condition is likely to be a warm wall (summer) and with a cold shocking to the embedment, thereby creating maximum straining across the section.

Agree with minor comment - Humayum

last sentence - straining or strain?

Agree with minor comment - Brannan

consider using “Design” instead of “Starting”.


Page 4 of 15





COMMENTS

Suggested Change

Agree with editorial changes - Pawski

Starting temperatures and conditions should be selected so as to maximize the demand under consideration. For instance for the design of the anchors the critical condition is likely to be warm wall (summer) and cold shocking thermal shock to the embed. Whereby creating maximum straining across the section.


Pawski 2013.04.23 – Allen comment below shown incorrectly as Douglas and deleted here. Correct comment is: "…....... cold shock to the embed which results in the maximum strain across the section " Douglas additional comment1 at end of Ballot Summary: "1. Page 5, Para 1.2.1 Line 24: Also Comment LE27: What is the advantage of the use of the term "demand" as opposed to load? If a new term is to be introduced then it's use must be defined: When is a load a demand?" Pawski 2013.04.23 – this is addressed by 376-C Minneapolis response above.

Agree with editorial changes - Allen

For instance for the design of the anchors the critical condition is likely to be warm wall (summer) and cold shocking to the embed. Whereby Thereby creating maximum straining across the section.

Agree with changes - Oliver

Agree with suggested grammatical revision: Starting temperatures and conditions should be selected so as to maximize the resulting stresses in the component under consideration. For example for the design of the anchors, the critical condition is likely to be warm wall (summer) and cold shocking to the embed, creating maximum straining across the section. Comment: What does demand mean in this context? If not revised, add “demand” to definitions.

16 P. 5, Line 22, at the end of the paragraph

Wu

Add “and hydrodynamic pressure due to seismic aftershock”.

The comment found persuasive. Change text to read:

• “Finite element analyses is performed first to determine forces resulting from thermal gradients applied to the TCP, TCP embedment plate and outer concrete wall. Loads should include all loads that the TCP and the embedment will be subjected to including thermal, prestress, creep, concrete shrinkage, internal pressure, hydrostatic pressure and hydrodynamic pressure due to seismic aftershock (SSEAFT).”

Agree – Jiang, Moncarz, Hoff, Rushing, Hoffman, Humayum, Howe, Wu, Ballard, Malhotra, Garrison, Mash, Allen, Legatos, Widianto, Hoptay, Oliver, Pawski, Hatfield, Brannan, Krstulovic

Agree with editorial change – Khalifa, Douglas, Roetzer, Krstulovic

“Finite element analyses is are performed first to determine forces resulting from thermal gradients applied to the TCP, TCP embedment plate and outer concrete wall.


Page 5 of 15





COMMENTS

Suggested Change


Mash

Suggest inclusion of the words 'in the absence of a defined leak rate all intermediate levels shall be investigated. In practice this may be achieved by running a number spill levels to so as adequately bound the demands. Typically 5 levels may be used dividing the spill height into 4 equal units. In addition levels in close proximity to the TCP must be investigated.

Agree in principle with comment. Revise as noted and address spill levels in a future edition of the Code.

• “Various spill levels should be considered to determine the most severe effect on the embedment and concrete wall. This should include a spill levels located just below the top of the TCP which will place the largest temperature differential between the TCP and the concrete wall. as well as various Also a minimum of three levels should be considered for spill levels above the TCP up to, at top of the TCP embedment zone, at mid-height, and full spill height. “

Agree – Jiang, Moncarz, Rushing, Hoffman, Humayum, Howe, Wu, Ballard, Malhotra, Garrison, Allen, Legatos, Widianto, Roetzer, Hatfield, Brannan, Krstulovic

2014.01.08 TCP text negative requiring discussion: ~ Ref. no. 18, Comment 25 – number of locations 5 or 3 2013.04.13 – 376-C Minneapolis: Mash comment non-persuasive. Three is a minimum and does not preclude specifier from requiring or the designer choosing more.

Agree with suggested change -Khalifa

This should include a spill levels located just below the top of the TCP which will place generate the largest temperature differential between the TCP and the concrete wall.

Agree with editorial changes – Hoff, Pawski

This should include a spill levels located just below the top of the TCP which that will place the largest temperature differential between the TCP and the concrete wall Also A minimum of three levels should be considered for spill levels above the TCP up to,1) at top of the TCP embedment zone, 2) at mid-height, and 3) at full spill height.


This should include a spill levels located just below the top of the TCP which will place result in the largest temperature differential between the TCP and the concrete wall

Agree with editorial changes - Hoptay

“…at the top of the TCP embedment zone, at mid-height, and full spill height”

Agree with editorial comment - Oliver

Change “place” to “produce”. Observation: The proposed text requires four levels, not so far from five.

Disagree - Mash Statement of 3 levels does not reflect the significance of the analysis. This is the most important analysis carried out for the outer tank and it is essential that an appropriate amount of levels be considered. Suggest that the text is amended to the following: Also a minimum of three five levels should be considered

20 Page 5, line32, item 14

Mash

Must be achieved through mechanical anchoring! The comment found persuasive. Revise to read: “The TCP embedment plate should be cast in place and mechanically anchored to the concrete wall or by other suitable means (e.g., ???). Anchor bolts (studs) should be placed in continuous (circumferential) rows along the concrete wall.”

Agree – Jiang, Moncarz, Hoff, Rushing, Hoffman, Humayum, Howe, Wu, Ballard, Malhotra, Mash, Douglas, Allen, Legatos, Widianto, Oliver, Pawski, Hatfield, Brannan,


Page 6 of 15





COMMENTS

Suggested Change

Krstulovic

Agree with suggested change -Khalifa, Roetzer

Delete second sentence completely as the same is essentially said in the next sentence. “The TCP embedmPartent plate should be cast in place and mechanically anchored to the concrete wall. Anchor studs should be placed in continuous (circumferential) rows along the concrete wall.” 2013.04.13 – 376-C Minneapolis: After discussion it was agreed delete second sentence as suggested.

Agree with suggested change -Garrison

Modify last sentence: “ Anchor studs should be welded in continuous circumferential rows along the embed plate.”

2013.04.13 – 376-C Minneapolis: Withdrawn after discussion on deleting second sentence as per Khalifa, Roetzer, and Hoptay.

Disagree -Hoptay 2014.01.08 TCP text - negative is resolved

Delete second sentence because next sentence covers the same information

2013.04.13 – 376-C Minneapolis: After discussion it was agreed delete second sentence as suggested.


Mash

When? Embedment must be anchored with studs? The comment found persuasive. Revise to read as noted below. Address questions on spacing and gap size in future: • “When the The embedment plate isshould be anchored with

circumferential studs, a minimum of two rows of circumferential studs should be used to anchor the embedment plate. The spacing between mechanical anchors (studs) has significant impact on the TCP emebedment performance and thus should be considered in the design. should be limited to ensure that the axial, bending and shear stresses in the embedment plate are within the allowable stresses.”

2014.01.08 TCP text negative requiring discussion: ~ Comment 27 – number of locations 5 or 3

Agree – Jiang, Moncarz, Hoff, Rushing, Hoffman, Humayum, Howe, Wu, Ballard, Malhotra, Garrison, Mash, Douglas, Allen, Legatos, Widianto, Hoptay, Roetzer, Oliver, Hatfield, Krstulovic

Agree with comments -Pawski

Bullet points 1 & 2 should be combined Use “studs” or “bolt studs” consistently

Agree - Brannan Note that our Figure 1.1a shows four rows of circumferential studs. Should it show two rows of studs in solid line ink and the other two in shaded or dashed ink to indicate the potential for using more than two rows?


Page 7 of 15





COMMENTS

Suggested Change

2013.04.13 – 376-C Minneapolis: Agree, annotate sketch to state “2-rows minimum.”.

Agree with editorial changes - Khalifa

Also recommend deleting bolts from first line under section 1.2.3

2013.04.13 – 376-C Minneapolis: “Anchors” to be used throughout. Edit TN and replace bolts & studs with “anchors.”

Disagree – Hoptay 2014.01.08 TCP text - negative is resolved

Disagree – studs are not always used for consistency suggest the following rewording The embedment plate should be anchored with a minimum of two circumferential rows of anchors. The spacing between anchors has significant impact on the TCP embedment performance and thus should be considered in the design. 2013.04.13 – 376-C Minneapolis: Persuasive, use proposed wording.


Mash

Nothing mentioned regarding detailing? Suggest inclusion of words such as. Ideally crack control is best achieved through the introduction of smaller bars at closer spacings as opposed to larger bars at wider spacings. In this regard it is suggested that the bar spacing be selected so as to enable the subsequent introduction of lacer bars to the system. Therefore adoption of bar spacings of between 150 and 200mm is recommended in the vicinity of the TCP whereby allowing additional bars top be slotted in. Anchor spacings should be selected to as to be multiples of the bar spacing to enable a clash free design.

The comment found non-persuasive regarding suggested inclusion of bar detailing because this is normal everyday concrete detailing. Add the following after line 7 on page 6.

• “Anchor spacings should be selected to as to be multiples of the bar spacing to enable a clash free design. “

Delete lines 6 and 7 on page 6.

• “ Allowable stresses i.e., material selection and material requirements used in the design of the embedment plates, may be selected in accordance with API 620.”

Agree – Jiang, Moncarz, Khalifa, Rushing, Hoffman, Humayum, Howe, Wu, Ballard, Hoff, Hoptay, Malhotra, Garrison, Mash, Douglas, Allen, Legatos, Widianto, Roetzer, Oliver, Pawski, Hatfield, Brannan, Krstulovic

NEW Page 6 Para 1.2.4

This comment inserted by Pawski 2013.04.23 Douglas additional comment 2 at end of Ballot Summary: The liquid tightness criteria are specified in Para 2.1, items 1 & 2. Additional tightness criteria on “embedment plates” should not require to be defined by owner. I am not in favour of definitions to be provided by owners. The statement is vague and ultimately provides no guidance.The code is to provide guidance to all, including owners.

Douglas Note – consensus at 2013.10.21 Phoenix meeting was to delete the new section 1.2.4

28 Page 7, line 1 4, item 1

Mash

Suggest using nomenclature such as demand as opposed to load.

The comment found persuasive in part. Revise to read. “When subjected to During the spill condition, the concrete wall will be loaded with subjected to high forces due to mechanical loads and thermal effects and mechanical forces. These forces that will lead to significant wall cracking. During the spill condition, integrity Integrity of the concrete wall must be maintained. As specified in the ACI 376 Code, the wall should

Agree – Jiang, Moncarz, Khalifa, Rushing, Hoffman, Humayum, Howe, Wu, Ballard, Malhotra, Garrison, Mash, Douglas, Allen, Legatos, Widianto, Roetzer, Oliver,

Note – Mash negative resolved . Wording incorporated into 2014.01.08 TCP markup is essentially as suggested by Mash, as follows: During the spill condition, the secondary concrete wall will be subjected to high large forces due to mechanical loads and thermal effects that will lead to significant wall cracking. Integrity of the concrete wall must be maintained. As specified in the ACI 376 Code, the wall should remain liquid


Page 8 of 15





COMMENTS

Suggested Change

remain liquid and substantially vapor tight. More specifically:”

Pawski, Hatfield, Brannan, Krstulovic

and substantially vapor tight.

Agree with editorial changes - Hoff, Krstulovic

During the spill condition, the concrete wall will be subjected to high forces due to mechanical loads and thermal effects that will can lead to significant wall cracking.

2013.04.13 – 376-C Minneapolis: Discussed and agree with comment.

Disagree – Hoptay 2014.01.08 TCP text - negative is resolved

Suggest rewording sentence as follows to cover both the transient and steady state condition. The integrity of the concrete wall must be maintained during the entire spill. 2013.04.13 – 376-C Minneapolis: Withdrawn.


Mash

EN only calculates characteristic widths. Historically US codes have cocooned the crack width requirements into the detailing rules, These were based on the Gergely Lutz expressions which used mean widths. Replace 'as per' with 'in accordance with'

The comment found persuasive. Revise page 7 lines 20-21 to read. “3) ACI 376 section 8.1.1.8 - requires that the cracks widths should be calculated as characteristic and not mean calculated crack widths. as per EN 1992-1-1.”

Agree – Jiang, Moncarz, Hoff, Khalifa, Rushing, Hoffman, Humayum, Howe, Ballard, Malhotra, Douglas, Allen, Legatos, Widianto, Hoptay, Roetzer, Oliver, Pawski, Hatfield, Brannan, Krstulovic

2014.01.08 TCP text negatives requiring discussion: ~ Ref. no. 32, Comment 48 – Mash, Wu & Garrison negatives . 2013.04.13 – 376-C Minneapolis: Agree with Wu. NEW BUSINESS for CODE - JG write the new business item on code change form Q1.

Disagree - Wu Suggest that: a) EN 1992-1-1 should stay, as stated in ACI 376 to delete “as characteristic and not mean”

2013.04.13 – 376-C Minneapolis: Persuasive since crack widths calculated by EN 1992-1-1 will be characteristic. Proposed wording to be adopted.

Disagree - Garrison b) Committee needs to review the ability of the contractor / designer to meet the requirement of 0.008” max crack width using EN equations for the spill condition (2 cryogenic temperatures). Experience shows that this is a difficult requirement to meet. Instead, perhaps ACI 376 should specify minimum rebar size and spacing within the TCP embedment zone.


Page 9 of 15





COMMENTS

Suggested Change

2013.04.13 – 376-C Minneapolis: Withdrawn pending future action by main committee. Jeff Garrison to prepare New Business (code change form Q1) proposal to address crack width values and method of calculation.

Disagree - Mash By removing the design code there is potential for use of codes that do not generate characteristic levels. In particular Model Code 90 is not prohibited, also Ven der Veen’s approach generates mean expectation of crack widths. There are some Contractors still purporting to MC90 and reviewer has a concern this will continue if left unchecked. Without guidance the Client is without a Captain and I uninformed. Note the NL Analysis is in accordance with EN so there should be no objection to using EN for crack widths. 2013.04.13 – 376-C Minneapolis: Considered as part of action item in the Garrison response.

34 Section 2.2.1 Page 7, Line # 31

1) Modeling

Jiang

Suggest to change to: 1) Modeling including modeling geometry and

material thermal properties

The comment found persuasive. Revise page 7 lines 20-21 to read.

1) “Modeling including specific tank geometry and material thermal properties”


37 Section 2.2.1 Page 8, Line # 2 4) Continue analysis until steady state final temperatures are reached

Jiang

Suggest to change to: 4) Continue analysis until steady state final temperatures are reached and the temperature distribution is obtained

The comment found persuasive. Revise page 7 lines 20-21 to read. 4) “Continue analysis until steady state final temperatures are reached temperature distribution is obtained”


2014.01.08 TCP text negative requiring discussion: ~ Ref. no. 37, Comment 48 – Garrison negative


Page 10 of 15





COMMENTS

Suggested Change

Disagree - Garrison Committee needs to define analysis / design requirements. Considering the variability in the possible loading conditions, variability in material properties and variability in modeling techniques it seems that a steady state analysis would be sufficient. 2013.04.13 – 376-C Minneapolis: Withdrawn.

41 Section 2.2.2 Page 8, Line # 8 “Apply dead load, prestressing and any other steady state load, iterate to equilibrium”

Jiang

Suggest to change to: “Apply dead, prestressing and any other steady state load”

The comment found persuasive. Revise page 7 lines 20-21 to read. “To establish initial conditions, aApply dead load, prestressing, normal-operating thermal and any other steady state load, then iterate to equilibrium”


2014.01.08 TCP text negative requiring discussion: ~ Ref no. 41, Comment 52 – Garrison negative

Disagree - Garrison Committee needs to define analysis / design requirements.


Mash

Determine strains in section, check for Residual Compressive Zone (RCZ), check for max comp stress etc.

The comment found persuasive. Change text to read: 5) “Beginning with the initial condition defined in step 2, cCompute crack widths at time steps times throughout the transient where reinforcing strains are the greatest. and exceed the material yield stress Furthermore, at each time step determine the adequacy of the compression zone.”


2014.01.08 TCP text negatives requiring discussion: ~ Ref no. 44 & 47, Comment 54 – Garrison negative

Disagree - Garrison Considering the variability in the possible loading conditions, variability in material properties and variability in modeling techniques it seems that a steady state analysis would be sufficient.

47 Section 2.2.2 Page 8, Line # 11 5) Computer crack widths at times throughout the transient where reinforcing strains are the greatest and exceed the material yield stress

Jiang

Suggest to change to: 6) Computer crack widths at times throughout the transient where reinforcing strains are the greatest and the stresses exceed the material yield stress


Mash

What does this mean? The comment found persuasive. Change Page 8, lines 17-18 to read: “Calculation of cracking and crack widths for the RLG LNG spill condition and/or external fire condition shall be performed using a rational engineering analysis. Both linear Linear or non-linear finite element (FE) analysis should be used as

Agree – Jiang, Moncarz, Hoff, Khalifa, Rushing, Hoffman, Humayum, Howe, Wu, Malhotra, Garrison, Douglas, Allen, Legatos,

2014.01.08 TCP text negative requiring discussion: ~ Ref no. 49, Comment 55 – Mash & Ballard negatives


Page 11 of 15





COMMENTS

Suggested Change

discussed in 2.3.1 and 2.3.2.”

Widianto, Hoptay, Roetzer, Oliver, Pawski, Hatfield, Brannan, Krstulovic Disagree - Ballard CSA Z276 requires that fire and spill be considered

concurrently. I believe that this needs to be addressed in this paragraph.

Disagree - Mash Rational to one uninformed Engineer could be irrational to informed Engineer. Rational requires defining.

50 Section 2.3.1 Page 8, Lines # 21 to 23 “In this case a non-linear finite element program is used in conjunction with

• Non-linear FE material constitutive models for concrete and steel, and

• Crack width calculations as per E

Jiang

Suggest to change to: “In this case a non-linear finite element program is used by:

• Implementing the non-linear material constitutive relationship for concrete and steel

• Carrying out FE analysis to obtain stress counters, and

• Performing FE analysis post process to calculate crack width as per Eurocode 2 1992-1-1

The comment found persuasive. Change text to read: “In this case, a non-linear finite element program is used in conjunction with by:

• Non-linear FE material constitutive models Implementing the non-linear material constitutive relationship for concrete and steel and

• Carrying out FE analysis to obtain stress/strain contours, and

• Performing FE analysis post-process to calculate characteristic crack width, calculations following a method, such as as per Eurocode 2 EN 1992-1-1 (see section 2.5 in this TN).”

Agree – Jiang, Moncarz, Hoff, Rushing, Hoffman, Humayum, Howe, Wu, Malhotra, Garrison, Mash, Douglas, Allen, Legatos, Hoptay, Roetzer, Oliver, Pawski, Hatfield, Brannan, Krstulovic

Agree with minor changes - Khalifa

Performing FE analysis post-processing to calculate characteristic crack width, following a method, such as Eurocode 2 EN 1992-1-1 (see section 2.5 in this TN).”

Agree with minor changes - Ballard

Change word “contours” to “distribution” or change sentence to “Carrying out FE analysis to obtain stresses and strains.”

Agree with changes - Widianto

To be consistent with Sections 2.3.1 and 2.3.2, suggest the sentence is re-worded as follows: "Non-linear or simplified finite element (FE) analysis should be used as discussed in 2.3.1 and 2.3.2" To be consistent with the last sentence in Section 2.3, I suggest that we switch the order between the current Sections 2.3.1 and 2.3.2 (i.e., present Simplified FE analysis in Section 2.3.1 and present non-linear FE analysis in Section 2.3.2. This will also be consistent with the progression of the actual design, where Simplified FE analysis is done in the preliminary design, before non-linear FE analysis.


Page 12 of 15





COMMENTS

Suggested Change

53 Section 2.3.2 Page 9, Lines # 2 & 3 “However, it should be noted that this is a very time consuming process and is thus usually used in the preliminary design.”

Jiang

Suggest to change to: “However, it should be noted that this is a very time consuming process and is thus usually used for focusing on more critical load cases after a preliminary screen for load cases has been performed in the preliminary design.”

The comment found persuasive. Remove the whole sentence. “However, it should be noted that this is a very time consuming process and is thus usually used for focusing on more critical load cases after a preliminary screen for load cases has been performed in the preliminary design.”


Proposed wording incorporated into 2014.01.08 TCP text. .

56 Section 2.3.2 Page 9, Lines # 4 & 5 “Moment-Curvature Approach: The moment curvature response is obtained using a concrete section analysis program in conjunction with:”

Jiang

Suggest to change to: “Moment-Curvature Approach: The moment curvature response is obtained using a concrete section analysis program by implementing:” Comment: Weather should this approach be considered as “Linear Analysis” or “Non-linear Analysis”? May be this approach could be considered as a simplified non-linear analysis approach and should be used for preliminary only.

The comment found persuasive. Change text to read: “ 2.3.2 Simplified Linear FE Analysis Simplified Linear FE analysis is usually used only in the preliminary design. …… “ Moment-Curvature Approach: The moment curvature properties are response is obtained using a concrete section analysis program by implementing:”

• non-linear stress-strain material properties, and • a non-linear shell formulation that accounts for the cracking

induced stiffness redistribution. • This moment-curvature approach would only be suitable for

consideration of the steady state linear thermal gradient case and should also be used only for preliminary design.”


Proposed wording incorporated into 2014.01.08 TCP text. .


Mash

Seems out of place in terms of the flow of the document. The comment found persuasive. Edit text as shown below: “2.3.4 Loading The RLG spill case should …… Nodal temperature transients will typically be computed in a separate heat transfer finite element or finite difference solution. The thermal model used for determining nodal temperatures This thermal model is also a high density model which should include concrete foundation, soil thermal boundary condition, walls, roof and outside or inside ambient temperature boundary conditions. This model should also include the base slab insulation and the thermal corner protection insulation. “

Agree – Jiang, Moncarz, Hoff, Khalifa, Rushing, Hoffman, Humayum, Howe, Wu, Ballard, Malhotra, Garrison, Mash, Douglas, Allen, Legatos, Widianto, Hoptay, Roetzer, Pawski, Hatfield, Brannan, Krstulovic

Proposed wording incorporated into 2014.01.08 TCP text. . 2014.01.08 TCP text negative requiring discussion: ~ Ref no. 63, Comment 67 – Oliver negative not addressed

Disagree - Oliver This response does not address the basic comment that this whole paragraph doesn’t below under “Loading”. Suggest moving it to Section 2.3.3 “Mesh Type and Size”, or inserting a new section before “Loading” called “Thermal Model”.


Page 13 of 15





COMMENTS

Suggested Change

66 A Page 10 Line 10

Hoptay

On the code side of ACI 376 ASTM A706 is not referred to as low temperature reinforcing, only in the commentary. Can it be referred to as low temperature reinforcing in the technical note?

The comment found persuasive. Change text to read: “The concrete wall internal temperatures from the thermal model are also typically used for selection and to determine the placement of normal (A615), cold-temperature (A706) and cryogenic reinforcing complying with the requirements of ACI 376.” Furthermore, in a future Code revision, revisit the use of A706 materials in ACI 376 section 4.7.2.


Proposed wording incorporated into 2014.01.08 TCP text. . 2014.01.08 TCP text negatives requiring discussion: ~ Ref no. 74, Comment 83 – Garrison & Wu negatives

74

Garrison

1) Code and TN need to define conditions for transient thermal analysis for the spill condition (leak rate).

2) Why does TN reference withdrawn version of Eurocode

2 (DD ENV 1992-1-1:1992)? Current version is BS EN 1992-1-1:2004.

The comment found partially persuasive. 1) Definition of the spill loading is beyond the scope of the current TN

and will thus be addressed in a separate document defining LNG spill conditions.

2) Update and change the model to the latest 2004 edition of the Eurocode,

Agree – Jiang, Moncarz, Hoff, Khalifa, Rushing, Hoffman, Humayum, Howe, Wu, Malhotra, Mash, Douglas, Allen, Legatos, Widianto, Hoptay, Roetzer, Oliver, Pawski, Hatfield, Brannan, Krstulovic

Agree with comment - Ballard

Garrison was going to forward a copy of EN 1992-1-1:2004 to Ballard for updating the code equations in the TN.

???? - Garrison Agreed, loading and consideration of need for transient analyses needs to be reviewed.

75 Appendix A Pages 15-16

Pawski

It is redundant because it repeats paragraphs 2 through 4 in section 2.1 (page 7, lines 9-21)

The comment found non-persuasive. The comment found to be an editorial comment. Appendix A is a “copy” of what is in the Code. It is provided for reader’s convenience.


76 Page 16, last paragraph in section R8.1.1.7

Thompson

Remove the last sentence from the Code Commentary: “Due to the non-homogeneous nature of concrete, calculated crack width values measured in the field will vary from calculated values, and therefore they cannot be directly compared to calculated crack widths.”

Section R8.1.1.7 is text copied over from the Code and provided herein for the user’s benefit. Text must remain as-is to properly reflect the Code requirement. Changes to be addressed only if/once the Code has been changed accordingly.

Agree – Jiang, Moncarz, Hoff, Khalifa, Rushing, Hoffman, Humayum, Howe, Wu, Ballard, Malhotra, Garrison, Mash, Douglas, Allen, Legatos, Widianto, Hoptay, Roetzer, Oliver, Pawski, Hatfield, Brannan,


Page 14 of 15





COMMENTS

Suggested Change

Krstulovic ADDITIONAL COMMENTS included with the ballots




COMMENTS Page 4, line 8; Section 1.1

Oliver EDITORIAL: Add “the” “The primary function of TCP is to protect the already highly…”

Note - Added to TCP-TN markup (2014.01.08)

Page 6, Lines 28 - 30 Hoff Can we provide some guidance to the owner in this section? 10-21: Hoff comment resolution: The crack width requirements in section 6 are provided to for this purpose. DELETE 1.2.4 Note added to TCP-TN markup (2014.01.08) that new "1.2.4 Step 4" is to be deleted, and is shown for information.

Hjorteset Wu Billy Oliver Brannan Pawski Howe Garrison Hoff Hanskat

Wu

The liquid tightness criteria are specified in Section 2.1, Items 1 & 2. Why do we need additional tightness criteria on “embedment plates ” to be defined by owner?







77 Page 9, section 2.3.3, lines 19, Khalifa Add ‘separate’ before ‘model’ for clarity. Note - Added to TCP-TN markup (2014.01.08) 78 Page 8, section 2.3.1, line32, 33

Khalifa Revise to: Either a proprietary or one of several commercially available FE programs can be used for performing this calculation.

From hand notes: "Concur, combine first two sentences" Note - Added to TCP-TN markup (2014.01.08)

Page 9, line 24 Hoff

Change this sentence to read: “If Liners are considered as structural members, they should be included if they are considered as structural members too.”

10-21-13 Editorial Note - Added to TCP-TN markup (2014.01.08)


Use “should” in a non-mandatory TN “Recognized concrete constitutive models shall should be used in the analysis.”


79 Page 10, section 2.4.1, line 30 Khalifa

Revise to: …account for reduction in tensile stiffness after cracking of concrete including the effects of tension stiffening.

10-21-13 Account for reduction in tensile stiffness after cracking of concrete . Note - Added to TCP-TN markup (2014.01.08)

Page 11, line 7 Hoff,

Widianto

The stress symbols are missing 10-21-13 Editorial Note - Added to TCP-TN markup (2014.01.08)

Page 12, lines 10, 11

Hoff

Change to read: “Since For the RLG spill case, it is critical that cryogenic liquid be prevented from migrating behind the TCP, and conservative assumptions should be used when selecting material properties.”


Page 13, Section 2.5

Wu

The specified Code, EN2 part 1- 4.4.2.4 does not contain the equations for crack width calculation. The crack width calculation is shown in Section 7.3.4 of EN 1992-1-1 (2004 edition). Also, the equations shown here are not the same as the equations shown in EN 1992-1-1 (2004 edition). Please provide clarification.

10-21-13 Modify to address to match code and equations. Same as item 74


Page 15 of 15

To be addressed as future business:

Author Comments Legatos Conspicuous by its absence in the TCP-TN write-up is any reference to outer-wall designs that incorporate an impervious, continuous liquid-tight and vapor-tight liner.

This omission overlooks the crucial role such a liner plays in the formulation of criteria and parameters as they pertain to (a) the serviceability of the outer wall following a spill and/or thermal shock; (b) the specified limits on crack-width, residual wall compression, etc., etc.; and (c) the function, scope and design of the TCP itself. Moreover, this omission overlooks the differences between inner-versus-outer, and metallic-versus-non-metallic liners; and the role these differences similarly play in (a) through (c) above. If I am not mistaken, the only Code section that fleetingly mentions this subject with respect to the TCP is Section 6.3.3. This article provides guidance on what to do if a “leak-tight membrane/liner” is not used, but does not say anything about what to do in cases where such a liner is in fact used. The subject is also covered in 6.3.16.1 1nd 16.3.16.2, but those sections do not relate to the TCP. While I have not thoroughly perused the rest of the Code itself to see exactly where else this subject “belongs” and how it should be covered, I believe that TCP-TN coverage might fit in the following sections:

Section 1.1 (add a fifth paragraph) Fig. 1.1 (add Fig. 1.1 c) Section 1.2.4 Section 2.1

Concrete Structures for RLG Containment

Documents

Transcript of Concrete Structures for RLG Containment