コンクリート中の鉄筋のマクロセル腐食機構(梗概) : 塩化物濃度の異なる電池

(1)

NII-Electronic Library Service

{ut

1]

Journal

of

St[uctural

and

Construction

EngLneering

UDC :6gl. 32 ;666. g7 :62o. Ig4

(Transactiens

ef

AIJ)

No.

407,

January,

lggo

H"reM7ftutrekdeRvakes

ig4O7e

･199O.G

1 fi

MACROCELL

CORROSION

MECHANISM

OF

STEEL

IN

CONCRETE

(Differential

Salt

Concentration

Cells)

by

KAZUO

SUZUKI*,

YOSHITERU

OHNO**,

SOMNUKE

PRAPARNTANATORN"'

and

HIROSHI

TAMURA"'",

Members

ef

A.I.J.

1. Introduction

The

most

common

_problem

concerning

the

durability

of reinforcecl

concrete

structures

is

reinforcing steel

coirosion.

The

steel corrosien causes

cracking

and eventual spalling of

the

concrete

cover

to

reinforcement

'

consequently,

_the

loss

_{of seFviceability}

_or,

in

_extreme _{case, structural} _collapse'i,

Naturally,

any

corrosion

in

aqueous

solution must

involve

simultaneous anodic and cathodic reactionsi),

For

the

causes

of corrosion of steel

in

eoncrete,

it

has

been

well reported

that

the

intrusion

of aggressive substances such as chloride

ions3]

_{or caibon}

dioxide`)

_{can substantially}

initiate

_anodic _action,

_i.e.

_Fe-Fe"+2eT

･J････(1).

_The

controlLing _cathodic _reactiop

in

_concrete

is

_oxygen _reduction5},

i.e.

O,+2H,O+4e--4OHT････..(2>.

_When

_the

sites of anodic and cathodic reactions are coincident,

it

is

called

microcell corrosion.

If

the

sites

are

formed

some

distance

apart,

it

is

known

as, macrocell corrosion.

Lewis

and

CopenhagenS}

first

raised

the

concept

that,

the

heterogeneity

of

environment

over

structllres

and

the

heterogeneity

of concrete

itself

can activate macrocell corrosion.

Until

recently,

there

has

been

little

information

published

on

the

macroceil

action

of

which

steel anode and steel cathode

are

both

embedded

in

actual

concrete.

The

macrocell

characteristics of embedded

anode

and cathode

have

not

been

clarified

_yet.

Some

reseaTchers7)･S) studied steel

corrosion

in

some simulated concrete environments,

_particularly

in

alkaline solutions.

Although

numerous

impressiye

and

informative

data

have

been

carried out, corrgsion

gharacteristics

of steel

in

alkaline

solutions

may

be

not

_absolutely _similar

to

those

in

real

concrete.

.Recently,

Yonezawa

et

a19).

have

shown

that

mortar

_provides

better

corrosion

_protection

to

_steel

than

alkaline _solutions,

For

degree

_of

_macrocell

_corrosion,

Vennesland

_and

_GjorviO}

suggested

that

_once _{an anode}

has

been

formed,

the

rate of corrosion would

_primarily

be

controlled

by

the

rate of

oxygen

diffusion

through

the

concrete and

the

cathode-to-anode area ratio.

However,

_their

_immersion

_test

_results _of

galvanic

cells, which composed of a

bare

stainless steel

_plate

and

an embedded steel

in

_precracked

speciinens,

showed

that

_even

for

_{a very}

large

_{cathode-to-anode} _{area ratio}

_steel

_corrosion _{at a narrow}

_crack

_can

_be

_finally

inhibited.

One

of

the

main objective of

this

study

is

to

investigate

the

macrocell

corrosion mechanism of

differential

salt concentration cells which are

commonly

found

in

real concrete environment.

Embedded

steel

in

a multi-crack member,

for

_example,

_is

_easy

to

_undergo

differential

_{saLt concentration} _cells

_as

_chloride

_ions

_can

_penetrate

_through

_a wider

crack

easier

than

through

a narrower one,

In

_previous

_paperM,

the

_authors

have

_proved

_that

_corrosion

_{of steel}

in

multi-crack

_specimens _will

be

first

initiated

at

the

major cracks

6f

the

cracked

specimens and

_proposed

that,

by

macrocell action, corrosion at

the

major

cracks

will

delay

and suppress corrosien

of

steel at

the

other cracks

(minor

cracks).

The

second

objective

is

to

clarify

the

_proporsal

_{qualitatively.}

'

The

present

report shows not only mechanism of

the

common

two-salt

concentration

cell,

that

was usually used

in

the

investigation7)･i2)･]3),

but

also

a

three-salt

concentration

cell,

The

three-salt

concentration cell approaches some

i

Professor

of

Osaka

University,

Dr.Eng.

'*

_Associate

_Professor

of

Osaka

University,

D[,Eng.

#i

_Graduate

_Student

of

Osaka

Uniyersity

##

_Chief

_Research

_Engineer,

_General

_Building

_Research

Cerpo[ation,

Osaka

(Manuscript

recelyed

July

10, 1989;

Paper

Accepted

October

31, 19S9)

(2)

-1-Architectural Institute of Japan

NII-Electronic Library Service

ArchitecturalInstitute of Japan

real conditions

of

macrocell

corrosion

in

concrete structures, such as

that

in

multi-crack members, more

than

the

two-salt

concentration

cell.

PaTarneters

of

the

corrosiion cells were

total

length

of

the

coupled steels, which relates

directly

to

cathode-to-anode area ratios, and

distance

between

the

embedded anodes and cathodes.

The

macrocell

currents, electrochemical

characteristics

and

corrosion

deterioration

of

the

corrosion cells were observed.

2. Experimental

Procedures

2,1

Specimens

and

Differential

Salt

Concentration

Cells

Specimens

composed of

three

compartments made of concrete

containing

1,5

%

and

_O.5

%

chloride

ions,

by

weight of cement, and

_plain

concrete

<1.

5 %

Cl',

O,

5 %

Cl'

and

O. O

%

Cl-

concrete, respectively).

Dimensional

and

ernbedded

steel

details

of

the

specimens are shown

in

Fig.

1. All

concretes

had

similar mix

_proportion

except chloride contents

and

were made

from

high

early strength cement

(Portland

Cement

Type

M

),

tap

water, iiver sand and

10 mm

maximum size crushed stone.

The

mix

_proportion

was cement=355

kglm3,

water-to-cement

ratio=O,

55 and

cement

:

sand

:

_gravel=1

:

2.

1 :

2,

6. Chloride

additions were made

by

dissolving

the

required

_quantites

of

Laboratory

Grade

NaCl

in

the

rnix water

to

yield

the

designed

chloride

ion

contents.

The

strength

properties

at

28 day

aTe

given

in

Table

1. The

embedded

steels were

bare

D

13

steel

bars,

The

specimens were

kept

wet under a

plastic

sheet

for

28days

after

demoulding.

The

exposure

test

started one week

later.

The

embedded

steels were

electrically

connected

just

prior

to

the

beginning

of

the

exposure

test.

The

electrical couplings

are

shown

in

Fig.

2. The

loop

coupling

of

three-salt

concentration cells should reveal

to

which

part,

anode or cathode,

that

steels

in

the

O.5 %

Cl-

concrete

beiong

in

the

corrosion

cells.

Parameters

were

total

length

of coupled steels and

distance

between

steels

in

the

1,5

%

and

O･5%

Cl'

concretes.

The

designation

and

_parameters

are

Tablel

Concrete

properties

presented

in

Table

2.

2.2 Exposure

Condition

and

Observation

The

specimens were

subjected

to

repeated cycles of one

day

wetting

in

650C,

3.1%

NaCl

solution

and one

day

drying

in

laboratory

environment

in

3.1%

NaCl

solution

is

Steeli DIJ withevt millscfle

L

.

_s Steel

length

.gF asting

ireetien

ordercloseto

accelerate

corrosion.

The

to

the

NaCl

level

found

in

Steel

Bars

Cl--ContentProperties{kgfcm2)

Concrete

Title

Fc

Ec

OiOZCI"O,07.3443.xlo5

o.s7.cl-O,5Z

_{'1.5ZCI'1.57.414J･7xlo5}

1"O13･7xlo5

"Percerit

by eer::ent ,,rei{:hL

ec:

Cemp;essive

strength

Ec:

Hodulus of elastieity

rza;ai

1

[=- a)

.

.-S3-+31+3B-]lt2"

Lsny-31HB-311

Plan

Epoxy

Coating

Table2

Gaivanic

couplings

_qnd

parameters

One

way coup!ing of

two-salt

eoncentrat'ion cell

/ll･

k

'

mT i.sz

cr1

Concretel

[I] P 1 O,OZ CIConcrete = InIO.5Z

CI'

TConcrete iJI

-53-Sections

-too-

50+L,-

Elevation

'

unit:

･mm

Fig.1

Details

ef specimen

PloebCetrOchem

±cal

Onetwo-saltvay

eovpling ef concentrat'i'on ce]l

CellTitleLecat

_Steels

±ansof

_{betwegnLengthof}

DistanceTotal(mm)

t.5ZCI'o.elCl"O.5ZCI-Steel{mm)CoupledSteels

TvJl1L･J12.pt

MI]Infinity uncoupled38each

ls'21

₂₀₀

₇₆

1･I22ll31

30o

76 W321･i41

50o

1,J42

76

`Ceter

to center of sZeels in 1.5Z and

--_H::.her'in

parenthesis indlcates ]ength

b)

Loop

coupling of

three-salt

cencentration

'O.51

Cl- concrete or stee] ine.O: Cl' conerete cel'1

c

.

g･

::::b

e:

Ammeter

/Locatians

ef ele,ct. measurement:1:1::

Fig.2

Loop･

coupling ef

three-salt

concentratien cell

Galyanic

couplings and

locations

of electrochemical measurernent

CellTitleLocattonsee

_{Steels1.5XCI-o.0ZCI-o,szCl-}

Distancebetween"Steel'

Cmm)

Ll-

EM]LInf ±nity uncoupled

L21

₂₀₀

L22L31

joo

L32L41

500 L42L5'

-Total

(mm)

Length or

Coupleg.Steels

38

each

126 C50)

226 (150)

426 {j50>

188

C150)

-2

(3)

NII-Electronic Library Service

seawater2).

The

elevated

temperature

of

650C

is

selected

from

the

temperature

effect

data

of

bare

steel

in

various solutions2}-i`).

The

one

day

interval

is

based

on

the

_preliminary

study.

Corrosion

currents were measured

twice

for

each

half

cycle

by

a zero shunt arnrneter.

The

first

was

taken

at

2 hours

after

beginning

of either wetting or

drying

and

the

seconcl one was obtained at

_the

end of

those

half

cycles,

Half

cetl

potential

(Ec)

and

_polarization

Tesistance

(Rp)

were measured against a

AglAgCl

reference electrode and a

_platinum

counter electrodei5).

These

electrochemical

characteristics were

periodically

observed at about

_3hours

before

ending of

the

drying

half

cycles.

At

the

end of

the

exposure

test,

weight

loss

and rust

area

were

examined.

3. Results

and

Discussion

3.1 Corrosion

Deterioration

All

steels

in

the

1.

5 %

and

O.

5 %

Cl-

concretes corroded,

The

average weight

losses

are shown

in

Fig.3.

In

cells

L41

and

L42,

rust was also

found

at

the

right

end

of

steel

bars

in

the

O.O

%

CIL

cencrete,

i,e.

near

the

O. s

%

Cl'

concrete.

Free

corrosion of uncoupled

steels

in

the

1,5

Corrosion

degree

of ceuplecl

steels

in

the

_1.

₅

_%

Cl'

concentration cerls,

the

increase

was

in

concrete, of

the

three-salt

concentration

cells

,

increased

from

38 to

₂₂₆

mm,

However,

the

'

length

from

226 to

426

mm.

conerete, of

both

types

of

corrosion

cells, concrete

increased

3,2

Electrochemical

Characteristics

and

The

half

cell

_potential

(Ec)

and

_polarization

logation

as

also shown

in

Fig,

_2,

obse:ved

before

the

beginning

of

the

exposure

3.2.1 Uncoupled

steels

The

typical

Ec

and

Rp-time

Fig,4.

The

Ec

the

nonhomogeneous corrosion on a steel

surface]6).

with

the

Ec

scatter.

Obviously,

the

Rp

characteristics

The

shifts of

Ec

and

Rp

towards

higher

val

the

declines

the

uncoupled

steels

in

the

1.5 %

Cl'

time

of about

5 to

12 cycles

to

be

depassivated.

than

that

of steels

in

_the

_O.5

%

Cl'

concrete.

It

is

worth noting

the

"tnLJutmo-P=tn,Has3

1.0

O.8 o.6

O.4O.2

o

-Ac: Aa:

1.0-tn

O.8vasg

o.6H:

O.4.neS

O.2 o

uneeupled

lOO

200 ₃₀₀

{o)

{1.6)

(4.3)

C6,9)

?X:91.LRgg,l:-,mm

Tetal area of steels

in

O.5Z

and O.OZ CI-Area of steel

in

1.5Z

CI-

eoncrete

40o(9.5)

cenerete

500(12.2)

%

concrete

was

increased

by

the

macrocell

action

dependent

of

the

distance.

increased

lncrease ln

There

was a certain

_tendency

that

corrosion

degree

of

the

coupl

increased

y

Distance,

mm

(uneoupled}

Fig

3 The

effect of total

_length

of ceupled steels and

distance

between

steels on coTTosion

deterioTation

concrete was

higher

_than

_that

_in

the

O.5 %

Cl'

concrete.

.

For

the

two-salt

The

corrosion

degree

of steels

in

the

_1.

_s

_%

proportionally

as

total

length

of

_the

coupled

steels

corrosion

degree

was not

_proportional

to

the

increase

in

_total

ed steels

in

the

O.

5 %

Cli

as

the

distance

between

them

and steels

in

the

_1.s

_%

Corrosion

Situations

resistance

(Rp)

were measured on

2

opposite

Both

values are

_presented

in

all

following

figures

of

this

section.

test.

faces

of

each

steel

The

first

data

were

curves _{of uncoupled} _steels

in

the

1,5

%

and

O,5

%

Cl-

concretes are compared

in

and

Rp

characteristics varied

between

opposite

faces.

The

scatter

of

Ec

from

face

te

face

indicated

The

scatter of

Rp

was _clearer _{and more}

_pronounced

_comparing

characteristics

revealed

corrosion

situations more clearly

than

the

Ec

ues

have

been

considered

to

reflect

the

passivation

phenomenon

whilst of

both

result

from

the

depassivation

on steel

surface')･S)･]]).

The

Ec

and

Rp

characteristics

showed

that

concrete

took

only one cycle while

those

in

the

O.

5 %

Cl'

concrete

took

longer

The

Rp

of steels

in

the

1.

5 %

Cl-

concrete was

almost

always

lower

The

steels

in

the

_1,

_s

_%

Cl-

concrete

corroded with

higher

degrees,

effect of steel

levels

on corrosion

(see

also

Fig.

1).

The

Ec

and

Rp

of opposite

faces

of

the

(4)

-3-Architectural Institute of Japan

NII-Electronic Library Service

ArchitecturalInstitute of Japan -Hotncxtn ¢

.>s

1L.-om

4003002001OO

o1000

800 6oO

400

200 o

m[

speei in i.sza)

Cell

Wll

Cl' aoncrete

x

ateel

tn O.5Z CI- concrete

f

Steel in 1,SZ

/1#

l:1]}･.･

NstRei

±n o･sz AutH=eL.-aat

b)Cl'

o

Cl-Cell

Ll ¢oncrete

-ts-l

coacrete

AHotn<xas<

->E

'vopt

400

300

200 1OO

o1OOO

80Q

6oO

a.sz clrr O.S:

"

N..=t-t<:

1.SZ

CI'

1.Sl

Twe oppesite reees

5

10

15

20

25 O

₅

le

15

20

25 Exposure

Time

(Cycles)

Exposure

Tirpe

_(C)rcles)

Fig.4

Typical

eLectrochemicai characteristics of uncoupled

steels

in

1.5%

and

O.5%

Cl'

concretes

Epm

a)

:el12ts'il..

e

s

b)

Cxe.11so//4k,,

l"

AanEcovam400200

o

O,5Z

CI-'-!frf<Xg'A-ti

Y-..e 1.5: Cl'

kr

Two appeslte i'aees

7pt--4r/"ICIil

i5s::>ta=sc':･cr ee=s

Fig.5

O

s

10

IS

₂₀

25 O

₅

Exposure

Time

(Cycles}

Exposure

Typicat

electrochemicat characteristics of steeLs

CL'

concretes of

two-salt

concentration ceLl

g--coc

LOO)tn

300S

200fi

i-100omo

Steel in 1.

1

s-)"

r

v

Asteel

in o.sz,cl' concrete a) Cell L21

5Z

CI- eo]crete

/

"..A-.x..

15

25 Tirlle

{CJrcles)

in

1.5%

and

O.5%

[ll!S!IEi!]

zSteel

in 1.sxb) Cell

L41

Cl- cancrete x

Steel in O,5Z CI'cencrete

1OOO

800-:t

6oo

9

n

4oo

ec

20:

O

5

10

15

20

25

O

₅

10

Exposure Time

(Cycles)

Exposure

Fig.6

Typical

electrocheniical characteristics of steeis

C]'

concretes of three-satt concentration ceLl

15

20

25 Time

_(C}tcles)

in

_1.5%

and

O.5%

upper steels scatter

less

than

the

lewer

steel$.

The

Rp

of

the

upper steels was also

lower,These

electrochemical

data

indi-cated

that

the

upper steels corroded easier

with

higher

degrees.

Concrete

at

bottorn

of

the

upper steels may

have

higher

Cl'

and

_porousityi7)

than

that

of

the

lower

steels

by

the

effect

of

bleeding.

3.2.2 Coupled

steels

The

typical

Ec

and

Rp-time

curves of

steels

in

the

1.5%

and

O.s%

Cl-

con-cretes are shown

in

Fig.5

and

_6.

The

typical

Ec

and

Rp

characteristics of steels

in

the

O,

O

%

Cl'

concrete

are

presented

in

Fig.7,

The

Ec

and

Rp

characteristics of

the

coupled steels

in

the

1,5

%

CIT

concrete,

ef

both

types

of corrosion

cells,

reveaLed

that

the

steels weTe all

depassivated

within

one cycle of

the

exposure

test

The

steels

always

had

the

lowest

values of

Ec

com-paring

with

the

others

in

their

corrosion

cells

indicating

that

they

were anodes while

the

others were cathodes.

Compar-ing

with

the

uncoirpled ones,

Rp

values

of

the

coupled

steels were

lower.

In

addi-tion,

the

scatter of

Rp

between

faces

was

less.

The

results suggested

that,

rnac-rocell action eased and

homogenized

iron

dissolution

of

the

steel anodes.

The

Ec

and

Rp

of

the

coupled steels

in

the

_o.s

%

Cl'

concrete, of

both

types

of coTrosion cells,

gradually

shifed

towards

higher

valuds after

the

first

drop

in

the

first

cycle of

the

exPosure

test.

The

declines

of

both

Ec

and

Rp

aPpeared

finally.

The

simultaneous

declines

of

both

Ec

and

Rp

as well as

the

Rp

scatter near

the

end of

the

test

ascertained

the

depassivation

of

the

steels.

Comparing

with

the

uncoupled ones,'

the

coupled

steels

took

longer

time

to

be

depassivated.

The

galvanic

action

delayed

corrosion

initiation

of

the

steel

cathodes.

The

distance,

from

the

steel anodes

in

the

1.

5 %

Cl'

concrete,

had

an effect on' corrosion

initiation

of

the

steel cathodes

in

the

O,

5 %

Cl'

concrete.

The

remotei steei cathodes started

to

corrode

(5)

NII-Electronic Library Service

CellL41i

i

x

AHotnaxtu<

->E

dvvm

400 30o

200 1oe

o1OOO

8006oO

400 ?oo

o

a) Point

1

Point1b)

AutH=oLiAec

Polnt2Point

2Peint

3eerree

±enc)

Point

3

nN

O

5

10

15

20

25 e

Exposure

Time

(Cycles)

Fig.7

Typical

electrochemical characteristics

earlier.

The

Rp

of

the

coupled steels

in

_the

_O.

₅

%

Cl'

concrete scattered

between

faces

more

than

the

uncoupled

ones

in

the

similar concrete.

The

mac-rocell action

dispersed

iron

dissolution

of

the

steel

cathodes.

Corrosion

was

found

on

the

right end of

the

steel

cathodes

in

o.O

%

Cl'

concrete

of

cells

C41

and

C

42,

as also shown

in

Fig,

_7.

The

Rp

characteris-tics

confirm

the

fact`')

that

Rp

of

the

uncoTroded spots

increases

as

time

increases.

The

Rp

scatter,

between

faces,

and

the

trend

of

the

decline

of

Rp

with

time

of

the

corroded

spots

were also similar

to

those

of

the

steel cathodes

in

the

O.5 %

Cl'

3.2.3 Relationship

between

Rp

and

'

The

relationship

between

Rp

and corrosion rate

B/Rp,

where

B=constant.

Hope

et al'B),

value of

the

constant

B

for

steel

specimens

'

corrosion was

_presented

in

terrn

of

Sum

llRp･

dT.

Ioss

of

all corroded steels

in

_the

_1.

₅

%

and

O.

5 %

Tue eppesitefaces /

5 -10

15

20

25 O

Exposure

Time

(Cycles)

of steels

in

_O.O

%

Cl'

cencrete

"mNsoxxophosvvevflecxT-Eptco

500

400 3oe

200･

1OO

o

OoA',ili'

oLA

JvJ.-.

"Ai,

5

10

15

20

25 Exposure

Time

(Cycles)

of

three-salt

concentration cell

eie ie ee eA i

Fig.8

LmmL-

ll-O

O.2

O.4

O.6

O.8

1.0 Gravimetric

1'ieighL

Loss

(g)

Rerationship

between

polarization

resistance

(Rp)

and

gravimetric

weight

less

concrete.

gravimetric

weight

loss

is

expressed

by

the

well-known

Stern-Geary

equation

:

ico..=

suggested

_that,

_the

actual corrosion rate

cannot

be

estimated

because

tlie

m concrete appears

to

vary widely,

Therefore,

acomparable

degree

of

The

relationship

between

Sum

11Rp

･dT

and

_gravimetric

weight

Cl-

concretes

is

shown

in

Fig.

8. The

linear

relationship clearly

indicated

that

Rp

could

be

used

to

determine

comparative

degrees

of steel corrosion

in

the

macrocell systems as

it

was

used

_effectively

in

some simpler _systems _such _{as corrosion} _{of one}

_steel

_electrode

_immersied

_in

_alkaline

solutions]9)

or embedded

in

mortar`)

3.3 Macrocell

Currents

and

Their

Effects

3,3.1

Two-salt

concentration

cell

The

currents of

the

two-salt

concentration cells

are

shown in

Fig.

_9.

The

anodic current,

_generally,

flowed

from

steels

in

the

O.5 %

Cl"

concrete

te

steels

in

the

_1.s

_%

Cl'

concrete, confirming

the

steels were

the

cathodes

and anodes of

the

corrosion cells, respectively.

The

current

during

the

wetting

half

cycles was

higher

than

during

the

drying

_periods

as

the

result of

the

test

taken

under

normal

temperature".

The

ross

of

electrons

that

caused anodic currents

to

the

steel anodes slowed

down

_the

cathodic reaction

<2>

around

the

_{steels2); consequently,}

_iron

_dissolution

_of

_the

_steel

_anodes

_in

_the

_1.5%

_Cl"

_concrete _was

eased and

(6)

Architectural Institute of Japan

NII-Electronic Library Service

ArchitecturalInstitute ofJapan A`oko-svprte k k :o co"e,Hsvp=okk=o

200

100

o

-100

200

100

-

I{+)

a? Cells(x

=V31,32joe

mm}

-10

5

10

IS

20

25

Exposure Time

_{Cycles)

During

wetting

halt

eycles

-<oko-EvD=digk=U

o

200 CellsW41,42Cx=500nm) .510Exposure15Time2e(Cycles)25

CIwet)

'

:etres±enSUPPresslencvrr

ttt---Corrosienprotectien

CfarsteelinO.SZCI-curreencTe

5101520

ExpesureTime{cyeles)2S

EL

'

5

10

15

Exposure

Time

20

25 (cycles)

b)

During

drylng

{15

(1

half

cyeles

(Id.y)

c)

Complete

cycles cuTrent-tlme curves mA!m2)tnAlm2)

100 o

-100

Fig.9Comparisonof

510

Expesure

C=Ivet+Idry)

of twe-salt concentration cell

15Time

20

25 {Cycles)

-4okv.HEL-p=okkso

200

1OO

o

ceuL21:

Totallength=

226

mm

-leo

Il{+}

I2(+)i

-

-ff

-qoke.HEwp"mkkpo

Cell

L31length

= v 200 1OO 20

25 (cycles}

:

Total

326

mmCellL41:Totallength=

426

mm

5

15 Exposure

Time

a)

Duping

wetttng

o

.100

'

S

10 ₁₅

2e

25

Exposure

Time

(Cycles)

Corres ±o" suppressien current

CIS

mAtm!} b] During

drying

rer steel in C.OZ CI- concrete

"'

cerres±en pr2teetien current

"

mAim2)

rer steel in O.O"h Cl' cencrete

Fig.10

Typical

cuirent-time curves of

half

cycles

tttt.tttttttt..tttttttttttt

510Expesure15Ttme20CCycles).25xl

half

cyales

three-salt

5Exposure

concentratloncell 10Timelc5

:.

lief?)

'Xl3'

6

(7)

NII-Electronic Library Service

homogenized.

The

comparison

of

currents

of complete

cycles,

shown

in

Fig.9c,

indicated

_that

the

200-500mm

distance

had

nearly no effect on

iron

dissolution

of

the

steel

anodes

in

the

1.5%

Cl'

concrete.

The

distance

also

had

no effect on

the

efficiency

of

cathodic

reaction

en

the

steel cathodes

in

the

O.5 %

Cl'

concrete according

to

the

fact

that

_both

anodic

and cathodic reactions of

the

corresion cells

simttltaneously

_proceed

at equal and opposite rate.

The

supply of

electrons

or cathodic current

to

a

cathode will suppress anodic reaction

{1)

of

the

cathodeZ).

Therefore,

iron

dissolution

of steel cathodes

in

the

O.

5 %

Cl-

concrete was

delayed

and

dispersed,

The

magnitude

of cathodic currents required

to

_protect

an embedded steel

from

corrosien

varies with

internal

and external environmental conditions.

The

current

density

may range

from

1

mAlrn2

for

corrosion

protec-tion

to

15rhAlm2

for

corrosion

suppression20}.

Both

'

currents were converted

and

_presented

by

dotted

and

solid

horizontal

lines

on

Fig,

9

a and

9 b

as arbitrarily

boundary

current

levels

for

comparison of corrosion

protection

degrees

of

the

steel

cathodes.

During

the

wetting

half

cycles,

the

cathodic currents

to

the

steel cathodes

in

the

O,

5 %

Cl'

concrete weTe much

higher

than

the

boundary

current

levels.

Corrosion

of

the

steel cathodes

should

not

be

initiated

and

_propagated

during

these

half

cycles.

During

the

drying

half

cycles,

the

cathodic

currents

iay

between

the

boundary

current

levels.

Corrosion

of

the

steel cathodes was,

therefore,

initiated

and

_propagated

during

the

drying

half

cycles,

The

comparison of

the

cathodic

currents

during

the

drying

half

cycles with

the

boundary

current

levels

indieated

a

trend

that

remoter steel cathodes

in

_the

_o.5

3.3.2 Three-salt

concentration

cell

The

typical

current-time relationships of

the

currents,

_generally,

flowed

from

steels

in

the

lower

_Cl'

former

steels were relatively

cathodes.

The

net

currents

to

each steel of

the

corresiQ'n

cells

diHt-.< e k o.HEvp:mk'"=oUH

5004003002001OO

o-100

E

iooiHSaOR.:

-1005p

-2oo:L

-3oo8

-400o--500

4Exposure

Time

812

oH-c oku･"evp:o"g=ooH

(cycles)

16

2024

100 o-100"200-300-40o-500

Fig.11

%

Cl-b)Netin

Ocurrent.OZ

CI

to

steel concrete

o:

Cell

L2

A:

Cell

L3

o:

Cell

L4

Comparison

of net current

to

concentration cell coneretewouldstart each

to

corrode steel

inthree-salt

earlier.

three-salt

concentration cells are shown

in

Fig.10.

The

anodic

concretes

to

_steels

in

the

higher

Cl'

concretes,

indicating

the

are compared

in

Fig.Il.

Totally,

steels

in

the

1.s

%

Cl'

concrete were

_the

anodes

while

the

other

were

the

cathodes

of

the

corrosion cells.

The

anodic

currents of

corrosion

cell

_groups

L

2

_ancl

L

3

_were _{approximately}

_proportional

_to

_cathode

_areas

_(see

_also

_Fig,

_3),

_On

_the

contrary,

the

current

of

corrosion

_cell

_group

L

4

was

not

proportional

from

the

beginning.

A

linear

relationship

between

anodic

dissolution

and cathode area can

be

clearly obtained when

both

steel anode and steel cathode

are

directly

subjected

to

solution' _containing

chtoride

ions2').

The

relationship

has

never

been

obtained

when concrete

involved

in

either

cathodic

_part]Z)･n)

_or

both

_anodic _{and cathodic}

_partsi3).

Because

of

the

fact

that

the

ZOO-500

mm

distance

had

no effect on

the

effeciency

of

the

cathodic reaction on

the

steel

cathodes

as

described

_previously,

the

deviation

of

the

relationship

in

this

study may

be

considered

to

be

due

to

the

restriction of

iron

dissolution

of steel

in

concrete.

In

anodic

_processes,

there

is

_a

type

of rate-limiting

behavior

when

the

solution

to

the

steel surfaces

becomes

saturated and crystallization _on

the

_surfaces _occuTs2),

The

_concrete _confinement _around _steels

_inevitably

affects

the

(8)

Architectural Institute of Japan

NII-Electronic Library Service

ArchitecturalInstitute of Japan

concentration of

ions

in

the

bulk

solution next

to

anode surfaces,

thus

the

iron

dissolution.

The

currents, shown

in

Fig.10,

were

higher

duTing

the

wetting

half

cycles

than

during

the

drying

_periods.

Therefore,

corrosion of

the

cathodes was

initiated

and

developed

during

the

dTying

_periods.

The

effect of

distance,

from

the

steel anodes

in

the

_1.5

_%

Cl'

concrete, on corrosion

initiation

and

propagation

of

the

steel

cathodes

in

the

O. s

%

Cl'

concrete can

be

explained as

those

in

the

two-salt

concentration

cells.

For

steel

cathodes

in

the

o.

o

%

Cl-concrete,

the

right end of

the

steels, neaT

the

O.5 %

Cl'

concrete,

of

cells

L41

and

L42

corroded,

The

corrosion

should

result

from

the

intrusion

of chloride

ions

from

external environment.

Corrosion

did

not

occur

on

the

steel cathodes

in

the

_O.O

_%

Cl'

concrete of

the

other cells

because

of

the

higher

cathodic current

density

to

those

steels.

Comparing

with

the

boundary

current

levels,

as

shown

in

Fig.

10 b,

the

steel cathode

in

the

_O.

_O

_%

Cl-

concrete of cell

L41

had

the

lowest

degree

of corrosion

protection.

'

A

significant

degree

of macrocell action was activated when

steels

in

the

1.5 %

CIL

concrete were

depassivated

(Fig.9

and

10).

For

steel

in

the

multi-crack

specimens,

steel

at

the

major cracks will

be

first

depassivated").

The

macrocell

action

should

take

_place

in

multi-crack specimens as

in

the

differential

salt

concentration

cells

presented

heTein.Therefore,

the

_present

corrosion cells may compare

to

steel

corrosion

in

mttlti-crack specimens

to

some

extent.

Corrosion

of steei at

the

major cracks will

delay

Fnd

suppress

steel

corrosion at

the

minor cracks

(Fig.4,

6,

9,

11).

The

corrosion

degree

of steel at

a

minor

crack

will

depend

on

the

distance

from

its

major crack

(Fig.

3).

If

the

iron

dissolution

at a major crack

is

limited

as

that

of

corrosion

cell

group

L

4

and chloride

ions

around steel

at

a

minor crack are accumulated sufficiently,

there

is

a

tendenpy

that

steel at

the

minor crack may at a

time

become

anew anode of

the

corrosion cell and share

a

_part

of

cathodic

_Ieaction

of

_passive

steel

in

uncracked concrete.

4. Conclusions

'

The,electrochemical

characteristics were effective

to

determine

corrosion situations

(4ctive

or

passive

state) and comparative

degree

of cerrosien of embedded

steels,

The

macrocell currents revealed

the

mechanism of corrosion cells.

The

_gravimetric

weight

losses

showed

the

over

all

dggrees

of

corrosion.

By

the

results of

these

data,

the

macrocell

coriosion

rnechanism and

its

effects can

be

concluded

as

follows

:

'

(

1 )

Macrecell

action was activated when steels

in

the

1,5

%

and

_O.5

%

Cl'

concrete or steels

in

the

_1.5

_%,

O,

O

%

and

o.

s

%

CIJ

concrete

were electrically connected.

Steels

in

the

highest

Cl-

concrete corroded

_,first

and

became

the

anodes

of

the

differential

salt concentration cells

(Fig.5-7).

This

resulted

in

the

flow

of significant macrocell currents within

the

corrosion cetls

(Fig.9

and

10).

(2)

The

macrocell

current

increased

corrosion

degree

of

the

steel

anedes

in

the

highest

Cl'

concrete

<Fig.

g,

ll

and

3a).

The

increase

was

_proportional

to

the

cathode area,

However,

there

was also a

limitation

of

the

linear

'

relationship

(Fig.

3

aand

11).

The

reason may

be

considered

to

be

due

to

the

restriction of

irop

dissolution

of steel

in

concrete.

(3)

The

macrocell

current

delayed

corrosion

initiation

and

suppressed

corrosion

_propagation

of

the

steel cathodes

in

the

lower

Cl-

concretes

(Fig.4-7,

9

and

11),

Corrosion

degree

of

the

ste.el cathodes

depended

on

the

distance

form

the

steel

anodes

(Fig,3b

and

7>.

References

1)

Isecke,

B. :Failure

Analysis

of

the

Collapse

of

the

B.erlin

Congress

Hall,

CoTrosion

of

Reinforcernent

in

Concrete

Construction,

Ellis

Horwood

Limited,

_pp.79-89,

_19S3

2)

Shreir,

L.L.(Ed,):Corrosion,

2'nd

edition,

Newnes-Butterworths,

London,

1976

3)

Spellman,

D.L,

and

Stratfull,

F.R.

:Chlorides

and

Bridge

Deck

Deterioration,

Highvvay

Research

Recerd,

No.328,

pp.38-49,

197e

'

4)

Gonzalez,

J.A.,

Algaba,

S.

and

AndTade,

C. :Corrosion

of

Reinfercing

Bars

in

Carbonated

Concrete,

Br.

Coiros.

J.,

Vol.15,

No.3,

pp.135-l39,

]980

5)

Cerrosion

of

Metals

in

Association

with

Concrete,

ASTM

Pub.818,

1984

6)

Lewis,

D.A.

and

Copenhagen,

W.J.

:

CoTTosion

of

ReinfoTcing

Steel

in

Conciete

in

MaTine

Atmespheres.

Corrosion,

Vol.15,

pp.382t-388t,

July

1959

7)

Gouda,

V.

K.

and

Mourad,

H.M,

:

Galvanic

Celts

Encountered

in

the

CoTrosion

of

Steel

Reinforcement-

ll

Differential

Salt

Concentration

Cells,

Corrosion

Science,

Vol.15,

pp.307-315,

1975

8)

Macias,

A.

andAndrade,

C. :

Corrosion

of

Galvanized

Steel

Reinforcements

inAlkalineSolutions,

Br.

Corros.

J,.

Vol.22,

(9)

NII-Electronic Library Service

9}

10)

11}

12)

13)

14)15}

16)

17)

l8)

19}

20}

21)

22)

Ne.22,

_pp.113-l18,

1987

Yonezawa,

T.

,

Ashworth,

V.'and

Procter,

R.

P.

M. :

Pore

Solutien

Composition

and

Chloride

Effects

on

the

CorTosion

_of

Steel

in

CencTete,

Corrosion-NACE,

Vol,44,

No.7,

pp.489-499,

July

1988

Vennesland,

O.

and

Gjorv,

O.E.

:

Effect

_of

_Cracks

_in

_Submerged

_Concrete

Sea

Structures

on

Steel

Corrosion,

_Material

Performance,

pp.49-51,

Aug

1981

Suzuki,

K.et

al.

:Influence

of

Flexural

Crack

on

Corrosion

of

Steel

in

Concrete.

J.of

StTuctllral

and

Construction

Engineering

(Transaction$

of

AIJ),

No.397,

pp.1-11,

March

l989

Okada,

K.

_and

_Miyagawa,

_T.

:

Chloride

Corresien

of

Reinforcing

Stee'1

in

Cracked

Concrete.

_ACI

_Publication

SP-6s,

pp.237-254,

1980

Suzuki,

K.

et al.

:

The

Effect

of

Cover

Thickness

_and

W!C

_ratioon

Cprrosion

_of

Steel

in

Concrete,

Review

_of

_the

_11th

GeneTal

Meeting,

Japan

Concrete

Institute,

1989

Foley,

R.T.

:

Role

of

the

Chleride

Ien

in

Iron

Corrosion,

CorrosiQn-NACE,

Vol.26,

No.2,

_pp,58-70,

Feb

197e

Tamura,

H,

and

Yoshida,

M,

:Non-Destructive

Method

of

Detecting

Cerresion

of

Reinforcing

_Steel

_in

_Concrete,

Transaction

of

the

Japan

Concrete

Institute,

Vol.6,

_pp.185-l92,

l984

Grimaldi,

G,

et aL

:

Factors

Influencing

ELectrode

Potential

of

Steel

in

Concrete,

Br.

Corros,J.

,

Vel.

_21,

No.

1,

pp.

5s-62,

l986Suzuki,

K.,

Ohno,

Y.and

Sornchai.

S. :

Experimental

Study

on

Internal

Cracking

ef

Partially

Prestressed

Concrete

Flexural

Members,

J.

of

Structural

and

Construction

Engineering

(Transactions

_of

A.I.J.

),

No.

_365,

_pp,

_9-17,

_July

_lgs6

Hope,

B.

,

Page,

J.

A.

and

Alan,

K.

C. :

Corrosion

Rates

of

Steel

in

ConcTete,

Cement

and

Concrete

Research,

Vol,

16,

pp.771-781,

1986

Andrade,

C.

and

Gonzalez,

J.A.

:

Quantitative

Measurements

of

Corrosion

Rate

ofReinforcing

Steels

Embedded

in

Concrete

Using

Polarization

Resistance

MeasuTement,

Werkstoffe

Korrosion,

Vol.29.

pp.515-519,

1978

Ashworth,

V.and

Geegan,

_C.G.

:Cathedic

Protection

of

Concrete

Reinforcernent,

CorTosion,

Preyention

&

Control,

pp,5-10,

Feb1987

Tsujino,

B.

and

Miyase,

S. :

On

Area

Ratie

of

Anode

to

Cathede

for

lronin

Neutral

Solutien,

Cerrosion-NACE,

_Vol.

_37,

No.9,

pp,541-545,

Sept

1981

Vrable,

J.B,

:

Galvanic

Corrosion

6f

Reinforcing

Steel

Exposed

to

Concrete

and

Water,

Material

Perfo[mance,

_pp.sl-s2,

March

1982

(10)

Architectural Institute of Japan

NII-Electronic Library Service

Arohiteotural エnstitute 　of 　Japan

【

論

文

l

UDC ；691

．

32 ：666

．

97

：

620 ．

194 日本建築学会構造系論文報告纂第 407 号

・

1990

年

1

月

コ

ン

クリ

ー

ト

中

の

鉄筋

の

マ

_ク

ロ

セ

ル

腐食機構

（

梗概

）

（

塩化物濃度

の

_異

な

る

電

池）

正会員正

会員

正会員正会員

鈴　　木　　計　　

夫

＊

大　　

野

　　義　　

照

＊＊

PRAPARNTANATORN

SOMNUKE

＊＊＊

田

　　

村

　　　　博

＊＊＊＊

　

1．

まえ

がき

　

鉄筋

コンクリ

ー

ト

構造

の

耐久性

に

関

して

最

も

重

要

な

問

題

は

鉄筋

の

腐食

である1，

_。

　

一

般

に_，

水溶液中

の

鋼材

の

腐食

は

，

アノ

ー

ド反応

（

Fe

→

Fe ＋

＋

2e

−

）

とカソ

ー

ド反応

（

O

_，＋

2

H

，

〇

＋

4 e

−

→

40H

−

）

_が

同

時

に

生

じるこ

と

によって

進行す

るz）

−

5，

_。

アノ

ー

ド域と

カソ

ー

ド域が

一

致

している

場合

の

腐食

はミクロセル

_腐

食

，

2

つの

領域

がいくらかで

も離

れている

場

合

の

_腐食

はマクロセル

_腐食

と

呼

ばれている

。

　

Lewis

と

Copenhage

皿6）_は

，1959

年

に_，

初め

てコンクリ

ー

ト

自身

の

不均

一

さ

等

が鉄筋にマクロセル

腐食を

生

じさせる

と

の

考

え

を示

した。その

後

，

多

くの

研究者

7 ）

・

S）_によってアルカ

リ溶液

のようなコンク

リ

ー

ト

の

擬似環境

において

鉄筋

のマクロセル

腐食

が研

究

されているが

，

アルカリ

溶液中

の

腐食特性

は

，

コンクリ

ー

ト

中

のそれとは

必

ず

し

も同

じでは

なく

9）

_，

コンクリ

ー

ト

中

のマクロセル

腐

食

におけるアノ

ー

ド

とカソ

ー

ド

特性

は十分には

把握

されていない10）

。

　

本

研究

の

主

な

目的

の

一

っは

，

異

なる

量

の

Cl一

を含有

するコンクリ

ー

ト

中

の

鉄筋

からなる電

池

におけるマクロセル

腐食

のメカニズムを

調

べ _ることである

。

そのような

電池

は，

複数

の

大小

のひび

割

れが

生

じている

場合

の

よう

に

実際

の

建物

やコンクリ

ー

ト

環境

においてしばしば

認

められる。

既報

］1）_{にお}_い _て，

筆

者

らは，

複数

のひび割れが

生

じた

試験体

における

鉄筋

の

腐食

は

，

ひび

割

れ

箇

所

に生じ

，

最

も

早

く腐

食

が

始

まる

箇所

のひび

割

れ

（

主

ひ

び割

れ

）

は

各

試験

体

における

最

も

大

きいひび

割

れであること

，

マ _クロセル

作用

に

よ

って

主

ひび

割

れ

箇所

の

腐食

はその

他

本論文の

一

部は

，

平成元年度日本建築学会大会学術講演梗概集に発表している