Transcript
Hvordan oppnås ønsket pålitelighet for konstruksjoner
Steinar Leivestad, Standard Norge 2015-12-09
In the Deck mating situation for the Statfjord C concrete (back ground picture, concrete Condeep structure is submerged to 6m platform
freeboard, having full characteristic loads governing design of all wall thicknesses of cells and domes etc. with, γF = 1,2, γC = 1,25, γS = 1,15 )
adequate safety was obtained by Adequately precise prediction models for: •Loads •Load effects •Resistance models
•This was achieved by use of best practice and up to date standard And Adequate confidence that design is actually correct •Achieved by extensive control by designer and client Adequate confidence that execution is actually correct •Achieved by extensive control by contractor and client And •Achieved by having personnel with adequate qualifications and experience in all positions
In the Deck mating situation for the Statfjord C concrete platform (back ground picture, concrete Condeep structure is submerged to 6m
freeboard, having full characteristic loads governing design of all wall thicknesses of cells and domes
etc. with, γF = 1,2, γC = 1,25, γS = 1,15 )
adequate safety was obtained by
Adequately precise prediction models for: •Loads •Load effects •Resistance models
•This was achieved by use of best practice and up to date standards And
Adequate confidence that design is actually correct •Achieved by extensive control by designer and client
Adequate confidence that execution is actually correct •Achieved by extensive control by contractor and client And •Achieved by having personnel with adequate qualifications and experience in all positions
IN FUTURE PROJECTS THE SAME IS ACHIEVED BY Adequately precise prediction models in design standards for: - Loads - Load effects - Resistance models
Adequate methods for the execution in execution standards and Adequate confidence that design is actually correct Adequate confidence that execution is actually correct •This is achieved by a Quality Scheme for control (scheme to be developed in EN 1990) • Control of design (EN 1990 + design standards) • Control of execution (Design + Execution standards) • Qualifications and experience of personnel
ISO 2394 Levels of design “sophistication” 4.4.2.1 Risk-informed decisions concerning design and assessment In a risk-informed design and/or assessment, the decisions shall be optimized with due consideration of the total risks, considering loss of lives and injuries, damages to the qualities of the environment, and monetary losses. The time horizon to be considered in the assessment of the total risks shall be determined on the basis of the duration of the functionality which the structure shall provide.
4.4.2.2 Reliability-based design and assessment As an alternative to risk-based design and assessment of structures, a reliability-based approach can be chosen. This approach shall utilize an assessment and minimization of costs and/or minimization of committed resource usage subject to given reliability requirements for the structure. The reliability requirements shall be assessed on the basis of a full risk-informed assessment as described in 4.4.2.1and will thus facilitate reliability differentiation in dependency of consequences of failure and costs of reliability improvements.
4.4.3 Semi-probabilistic approaches For structures for which the consequences of failure and damage are well understood and the failure modes can be categorized and modelled in a standardized manner, semi-probabilistic codes are appropriate as basis for design and assessment. Standards shall serve to ensure the quality of analysis, design, materials, production, construction, operation and maintenance, and documentation, and thereby explicitly or implicitly account for the uncertainties which influence the performance of the structures. The specifications given in standards should be developed such that they quantify all known uncertainties.
8.4 Target failure probabilities The target failure probabilities, i.e. pft, should be chosen taking into account the consequence and the nature of failure, the economic losses, the social inconvenience, effects to the environment, sustainable use of natural resources, and the amount of expense and effort required to reduce the probability of failure. - What is acceptable failure probability; Is a probability of 1% of the fatality risk of a 16 years old boy ok today? - What is the actual failure probability for normal “on-land” structures in Norway and Europa today. Would the public accept a larger risk than today?? - How do you convert failure into consequence if it is not total collapse???
- What would the annual number of fatalities per million be if the actual reliability index was 3,8???? - How to enter the effect of flaws and errors in design and execution into the estimated reliability?????
ISO 2394 Annex G Informative This is the only place in ISO 2394 where numerical values are indicated, it could be questioned if these values would be accepted by society today?
EN 1990 Capter 2 opens for reliability differentiation, details in Annex B (informative) proposed rev. B2.1 Consequences classes (1) For the purpose of reliability differentiation, consequences classes (CC) may be established by considering the consequences of failure or malfunction of the structure as given in Table B2.
Table B2 - Definition of consequences classes Consequences Class CC3
Description High consequence for loss of human life, or economic, social or environmental consequences very great
CC2
Medium consequence for loss of human life, economic, social or environmental consequences considerable
CC1
negligible consequence for loss of human life, and economic, social or environmental consequences small
Examples of buildings and civil engineering works Grandstands, public buildings where consequences of failure are high (e.g. a concert hall) and public buildings where the function is critical (e.g. hospitals, firestations), large bridges etc. Residential and office buildings, public buildings where consequences of failure are medium (e.g. an office building), small and medium size bridges etc. Agricultural buildings, buildings where people do not normally enter (e.g. storage buildings), greenhouses etc.
(2) The criterion for classification of consequences is the importance, in terms of consequences of failure, of the structure or structural member concerned. See B2.3 (3) Depending on the structural form and decisions made during design, particular members of the structure may be designated in the same, higher or lower consequences class than for the entire structure. NOTE The requirements for reliability are related to the structural members of the construction works, the system reliability should for reasons of robustness be higher than for the individual members.
EN 1990 Annex B, proposed rev B2.2 Differentiation by β values (1) The reliability classes (RC) may be defined by the β reliability index concept. (2) Three reliability classes RC1, RC2 and RC3 may be associated with the three consequences classes CC1, CC2 and CC3. (3) Table B3 gives recommended target values for the reliability index β for new structures associated with reliability classes (see also Annex C) using a 50-year reference period. Table B3- Recommended target values for reliability index β for new structures (ultimate limit states) Target values for β
Reliability Class Annual value
Value for a 50 years reference period
RC3
5,2
4,3
RC2
4,7
3,8
RC1
4,2
3,3
Note 1: The corresponding failure probabilities for the reference period of 50 years are equal to 50 times the annual values, which makes the two requirements in principle equivalent. The difference is that the 50 years requirement allows temporary higher annual failure probabilities for some periods, if compensated by lower ones for others. NOTE 2 A design using EN 1990 with the partial factors given in annex A1 and EN 1991 to EN 1999 is considered generally to lead to a structure in RC2. NOTE 3 Reliability classes for members of the structure above RC3 are not further considered in this Annex, since these construction works and their members each require individual consideration.
Sikkerhetsnivå i dagens standarder • I diskusjonene i EN 1990 Expert Group er det enighet om; • Sikkerhetsnivået i dagens byggeri er nok betydelig høyere enn det måltallene for sikkerhetsindeksen skulle tilsi • Det er liten grunn til å tro at samfunnet ville være villig til å akseptere vesentlig økning i faktisk konstruksjonssvikt. • Dagens sikkerhetsnivå og partialfaktorer er basert på kalibrering mot tidligere praksis IKKE de måltall som refereres • Måltall for sikkerhetsindeks er primært egnet for å vurdere oppnådd sikkerhet ved ulike alternative løsninger opp mot hverandre, ikke som absolutte måltall. • Bruk av “global sikkerhet” konsepter er en “pådriver” for absolutt verdier for måltall.
Nye bestemmelser for eksisterende konstruksjoner introduserer nye problemstillinger
Existing structures, indicative target values
Note the significant increase of target reliability index when the area involved in the collapse Acol increases, This could correspond to a - system failure versus a - member failure.
System of European standards for Construction Products Directive (CPD) WORKS + National legislation
Interface Society / construction project For a construction project use of Eurocodes with the underlying standards for execution and materials provide confirmation at the interface to society that structures are safe, assuming design and execution are without flaws.
Eurocode - 1990 Basis of structural design TC250
This figure is drawn to illustrate the situation for Concrete Works, similar figures could be drawn for
Eurocode - 1991 Actions on structures TC250/SC1
• • •
Steel structures Composite steel/concrete Aluminum structures
While not for
Eurocode - 1992 Design of concrete structures TC250/SC2
• Timber structures • Masonry structures As these are not having satisfactory standards for the Execution of the Works.
EN 13670 Execution of concrete structures TC104/SC2
EN 206-1 Concrete TC104/SC1
ISO 6934 or ETA Tendons & PT kits
EN 10080 reinforcement
EN 13369 - xx or ETA Prefabricated elements TC229
Product and testing standards TC104/SCs and WGs
Product and testing standards
Product and testing standards
Product and testing standards
NS-EN 1990/NA NA.A1.3.1 Design values of actions in persistent and transient design situations NA.A1.3.1(1) For structures in reliability classes 1, 2, 3 (see Table NA.A1 (901)), the values ascribed to γF in tables NA.A1.2(A), NA.A1.2(B) and NA.A.1.2(C) may be used. This is on the condition that the requirements for a quality system, design supervision and inspection during execution specified in the following text and in tables NA.A1(902) and NA.A1 (903) are complied with. For structures in reliability class 1 (CC1/RC1), the partial factor γF for variable actions may be reduced by the factor kFi = 0.9. Note, we can not say anything about reliability if we don’t have control over the quality of design and execution!
We do not only need adequately precise design and construction methods, we need design and execution without flaws, and standards to provide for that.
Illustration of the effect of control on the β value from reduced variability on load-effects and resistance. Compared to use of load factors γFI = 0,9-1,0-1,1. Diagrams from Milan Holicky.
6
5
4
3
2
0
0.2
6
Strict, (illustratio 0.6 n) 0.8
0.4
β
5
Normal
A B β = 3,8
4
C
χ 3
0
6
5
5
RC3 β = 4,3
4
RC2 β = 3,8 RC1
3
β = 3,3 3
0
0. 4
0. 6
0. 8
6
β
4
0. 2
χ 0.2
0.4
0.6
. 0.8
2
Low, (illustration)
6
Comparison of basic and increased production quality keeping γc=1,5 and γm=1,15. Figure 1 is taken from Milan Holicky, Figure 2 shows new results by Holicky assuming reduced variability as a result of improved execution control. 6
β
5
A B
4
β = 3,8
β
C
χ 3
5
A
β = 3,8
0.2
0.4
β
0
C
0.2
0.4
0.8
A
5
χ 3
0.6
6
B 4
0
0.6
B 0.8
Figure 1. Variation of the reliability index β of the beam with the load ratio χ for reinforcement ratio ρ = 1 %, basic production quality, γc=1,5, γm=1,15 and load combinations A, B and C.
.4
3 0
C
χ 0.2
0.4
0.6
0.8
Figure 2. Variation of the reliability index β of the beam with the load ratio χ for reinforcement ratio ρ = 1 %, increased production quality, upper diagram γc=1,35, γs=1,05 lower diagram γc=1,5, γs=1,15
Structural failures are due to a set of reasons: • risk inherent in our design procedures (β’s) • errors in design • errors in execution • errors in use • acts of God
Society have no acceptance or forgiveness for the first four, hardly even the last Failures are normally due to a combination of errors, hardly ever only the first Target β -values are low, real values are ~ok!
When designing steel structures according to EN 1993-1-1 [10] for buckling of slender compression members in combined compression and bending you need the interaction factor kyy, for calculating expression (6.61) M y , Ed + ∆M y , Ed M z , Ed + ∆M z , Ed N Ed + k yy + k yz ≤1 χ y N Rk M y , Rk M z , Rk χ LT γ M1 γ M1 γ M1
for buckling around the y-y axis, in the for a plastic design according to method 1 you need to calculate the following expressions given in Annex A and shown in figure 5; Calculate kyy=
and with
and with
C my C mLT
µy 1 N C 1 − Ed yy N cr , y
with
N Ed N cr , y = N Ed 1− χy N cr , y 1−
µy
Wel , y 1,6 2 1,6 2 2 C yy = 1 + (wy − 1) 2 − Cmy λ max − Cmy λ max n pl − bLT ≥ w w W pl , y y y 2
bLT = 0,5 aLT λ 0
M y , Ed
M z , Ed
χ LT M pl , y , Rd M pl , z , Rd
Figure 5 The interaction factor kyy for use in expression (6.61) WHAT IS PROBABILITY OF AN ERROR AND WHAT HAPPENS TO β (10-6) IF CALCULATION IS ERRONEOUS????
WHO USE WHAT TECHNICAL STANDARD
Standard etc.
Designer
Constructor
Concrete producer
Material producer.
Building law and regulations PU Interface between society and construction project PU* EN 1990 Basis of design PU* EN 1991 Actions PU* EN 1992 Design of concrete National Standards for quantity, cost and bidding
PU*Interface
PU Interface
Design/Constructor
Design/Constructor
EN 13670 Execution of concrete str., incl Execution spec
PU Interface
PU* Interface
Design/Constructor
Design/Constructor
SU
PU Interface
PU*Interface
Constructor/producer
Constructor/producer
SU
PU Interface
PU* Interface
Producer/material
Producer/materialpro
PU*
PU
SU
PU*
EN's for special tasks, bored piles, diaphragm walls etc.
EN 206-1+ NA Concrete
EN for concrete constituents
SU
EN 12620 EN 197, NS 3086 EN 1080
EN for testing concrete
SU
SU
EN 12350 Fresh concrete EN 12390 Hardend concrete
EN for testing constituents
PU* = Primary user, who the standard is "written for" PU = Primary user, one who needs to know the standard in detail SU = Secondary user, one who needs to be aware of the standard Interface standard are standards that are used for communication between the parties who are both primary users (PU)
There are three main pillars • DESIGN • Eurocodes provide potentially safe design, if only it is without flaws…....
• EXECUTION • EN 13670 and EN 1090 provide potentially safe execution, if only it is without flaws ………
• QUALITY MANAGEMENT • ISO 9000/9001 common for bakers, hairdressers, designers and contractors provides adequate quality management if only content is appropriate……. There must be adequate procedures relevant to the work i.e. design, execution of concrete, steel etc. This is a task for the EUROCODES to ensure.
We have in EN 1990 the building blocks for a reliability management system, but we have nowhere established a coherent system, or invited the member states to do so, this is area of MS competence
FROM EN 1990 Annex B
EN 1090 EN 13670
CC1
RC1
DSL1
IL1
EXC1
CC2
RC2
DSL2
IL2
EXC2
CC3
RC3
DSL3
IL3
EXC3
Proposal in EN 1990 Annex B for a Quality Management system and differentiation dependant on Consequence of failure.
NA.A1.3.1(902) Kvalitetssystem NA.A1(902.1) Ved prosjektering, utførelse og kontroll av konstruksjoner i pålitelighetsklasse 2, 3 og 4 skal et kvalitetssystem være tilgjengelig. For konstruksjoner i pålitelighetsklasse 4 skal kvalitetssystemet tilfredsstille kravene i NS-EN ISO 9000-serien. MERKNAD For marine konstruksjoner for petroleumsindustrien i pålitelighetsklasse 3 krever myndighetene (Petroleumstilsynet) at kvalitetssystemet skal tilfredsstille NS-EN ISO 9000serien. NA.A1(902.2) Kvalitetssystemet skal spesifisere krav for: − organisasjon; − personell; − prosjektering, omfang og dokumentasjon av beregninger; − programvare benyttet i prosjekteringen; − prosjekteringskontroll; − utførelse (arbeidsutførelse og arbeidsledelse); − kontroll av materialer og komponenter; − kontroll av utførelse; − kontroll under bruk; − system for håndtering av avvik; dokumentasjon av: prosjekteringskontroll, kontroll av utførelse og kontroll under bruk
NA.A1.3.1(902) Kvalitetssystem NA.A1(902.1) Ved prosjektering, utførelse og kontroll av konstruksjoner i pålitelighetsklasse 2, 3 og 4 skal et kvalitetssystem være tilgjengelig. For konstruksjoner i pålitelighetsklasse 4 skal kvalitetssystemet tilfredsstille kravene i NS-EN ISO 9000-serien. NA.A1(902.2) Kvalitetssystemet skal spesifisere krav for: − organisasjon; − personell; − prosjektering, omfang og dokumentasjon av beregninger; − programvare benyttet i prosjekteringen; − prosjekteringskontroll; − utførelse (arbeidsutførelse og arbeidsledelse); − kontroll av materialer og komponenter; − kontroll av utførelse; − kontroll under bruk; − system for håndtering av avvik; dokumentasjon av: prosjekteringskontroll, kontroll av utførelse og kontroll under bruk
Tabell NA.A1(903) – Krav til kontrollform ved prosjektering og ved utførelse, avhengig av kontrollklasse Kontrollform Prosjektering
Utførelse
Egenkontroll
Intern systematisk kontroll
(DSL 1)1)
Utvidet kontroll 2) + Uavhengig kontroll 5)
(DSL 2)1)
PKK1 / UKK1
kreves
PKK2 / UKK23)
PKK3 / UKK3
Kontrollklasse
1)
Egenkontroll
Intern systematisk kontroll
Utvidet kontroll 2) + Uavhengig kontroll 5)
(DSL 3)1)
(IL 1)1)
(IL 2)1)
(IL 3)1)
kreves ikke
kreves ikke
kreves
kreves ikke
kreves ikke
kreves
kreves
enkel utvidet kontroll kreves
kreves
kreves
enkel utvidet kontroll kreves 3)
kreves
kreves
normal utvidet kontroll kreves
kreves
kreves
normal utvidet kontroll kreves 4)
Se punktene B4 og B5 (informativt tillegg B) for parallelle betegnelser og bestemmelser, DSL og IL. Utvidet kontroll utføres i byggherrens regi enten av byggherrens egen organisasjon eller et annet foretak som er uavhengig av foretaket som utførte arbeidene 3) Der de løsningene som benyttes gjør at bæreevnen er særlig avhengig av utførelsen, for eksempel for: materialer med høy fasthet (stålsort S460 eller høyere, betong trykkfasthetsklasse B55 eller høyere), sveisesoner i utmattingspåkjente konstruksjoner, konstruksjonsdeler med etteroppspent armering, samt i eventuelle energiabsorberende soner i seismisk påkjente konstruksjoner (se NS-EN 1998-1) utføres og kontrolleres arbeidene i overensstemmelse med kravene for utførelseskontrollklasse UKK3 (utvidet kontroll). 4) Ved prefabrikkerte produkter som skal beregnes i overensstemmelse med Eurokodene, kan forutsetningen om uavhengig kontroll av utførelsen ansees tilfredsstilt dersom produktet er produsert i henhold til en harmonisert standard og underlagt samsvarskontroll under en sertifiseringsordning med sertifisert kontrollsystem, med et ekstra kontrollelement ivaretatt internt for eksempel av egen prosjekteringsavdeling. 5) MERKNAD Denne standarden forutsetter at det, som et tillegg til utvidet kontroll, utføres uavhengig kontroll i henhold til byggesaksforskriften SAK10 § 14-2 siste ledd, i form av bekreftelse av at utvidet kontroll er gjennomført og dokumentert. 2)
Control pyramid in a project, with interface to authorities Basics open for national choice Standards need a system to get things right Authorities needs a system to confirm that things are right
Execution class 3 [CC3/RC3 + special technology] Clients Quality System / Authorities
Authority control if req. INDEPENDENT DSL3 / IL3
Execution class 2 [CC2/RC2] Constructors Quality System
INTERNAL
Execution class 1 [CC1/RC1]
DSL2
SYSTEMATIC IL2
Interface between ”project” and building authorities •Documentation •Audit
Constructors Quality System
SELF DSL1
CHECKING / IL1
Quality in a project should come from below, as “good quality work” from the very start. Not as “corrections” from above.
Control shall help us manage our projects not give us the catastrophe with the pyramid turned upside down
OUR AMBITION !!
