AIR ENTRAINfoIENT OF MORTAR NEIL BENINGFIELD RMC

Transcription

AIR ENTRAINfoIENT OF MORTAR NEIL BENINGFIELD RMC
..
118
AIR ENTRA I NfoIENT OF MORTAR
NEIL BENINGFIELD
RMC Mortars Limited ,
Holly House , 74 Upper Holly Walk, Royal Leamington Spa , Warwickshire , UK
ABSTRACT
Th e mechanism of air entrainment is considered.
Research work on many of the factors affecting the air content of
mortars is discussed . The effect of mix consistence , mixing time , sand
composition and grading , clay content , temperature , cement particle size
and cement chemistry a r e quantified.
It is seen that the relationships
found were not always in accord with existing literature although th is was
often predominately arientated to concrete rather than mortar .
The effects of entrained air on the plastic and hardened mortar
properties are discussed .
INTRODUCTION
An earlier literature survey by the Author(l) showed that the factors
affecting the amount and type of air entrained and the effect of this air
on the properties of mortar are rarely quantified in the literature.
Much air entrained mortar is produced without due regard to changes
in the factors that can alter the total amount of air or the stability af
that air.
The majority o f these influences are addressed and relationships
shown linking the effects of mix design and other factors on the amount af
air entrained .
THE FACTORS AFFECTING AIR ENTRAINMENT
Mixing Time
For the most widely used generic type of air entrainment , that based on
vinsol resin , air content increases to a peak quite rap idly and then
119
declines as shown in Graph 1.
Mix Consistency and Mixing Time
Graph 1 also shows that a mix of much gr e ater co nsistence, i.e . much
wetter , loses air at about the same rat e as a drier mix o
Sand Grading:
Graph 2 shows that fine sand loses air much more
rapidly than coarse when subjected to prolonged mixing.
In order to obtain the same air content however , for both mixes , much
more agent was required for the fine than for the coarse and so t he
comparison could not be made on the basis of identica l original mixes.
It may be therefor e that air loss relates directly to the amount of
agent used and this relationship rather than a grading related one is the
primary one .
Graph 2.
The effect of change in
Graph l . The effect o f prolonged
mlxlng a1: t'NO àifferent consistences. sand graàing on t he 'Tlixing cime
effect: .
13
14
A
12
?ine Sanà
Coarse Sanà
12
d.b . 9.4
~l
::::
~O
<::
*'
9
8
a
7
d. b.
a
i6
24
12.75
32
TIME =N ,HNUTES
ô
2.0
i2
24
36
TIME IN ,üNUTES
FACTORS CAUSING AIR LOSS WITH TIME
It had been suggested that the loss of air was caused by the mi x drying
but careful weighing of a mix, complete with mi xer bowl and blade for
accuracy , showed that only small amounts of water were lost.
Restoring
accurately that loss did not restore the air contento
Various authors have reported particle attrition, leading to
increased fines, as a reason for air loss with time. ( 2)
....
ela
~
120
Mixes were therefore made using a soft oolitic material and a hard
siliceous sand, mixed for 30 minutes and the air content loss and grading
change measured . This showed a slightly greater increase in fines for the
oolite , as would be expected, but the air loss was in the same order for
each , as shown in Graph 3.
Graph 3.
~ardness
The effect o f aggregace
o n the mixing
~ime
effec~ .
16
14
12
~
o\
x
Silceous sand
\0
!O::
y
<:
Ool ite
X
lO
"-
"-
"-
"-
3
"o
",, ,
~
"-
6
,,
"
'""
4
o
4
8
12
16
20
24
28
,':ME :N :.1INUTES
It is believed that this demonstrates without doubt that sand
attrition is not a factor in causing air loss . Water loss also having
earlier been ruled out this leaves two further p oss ible influences ,
chemical reactions associated with the commencement of hydration and the
breaking apart of agglomerates of cement and/or clay , material too fine to
be meaningfully identified by the ordinary sieve analysis used .
In order to investigate whether breakdown/attrition of clay was a
factor a sample of sand was washed clean of alI less than 75~m particles,
such that alI clay had been removed.
This showed exactly the same characteristics of loss with time as the
untreated product containing clay and showed the clay not to be a
causative factor in the air loss.
Graph 4 shows this with the clean sand
plotted together with an untreated contraI test for comparison.
It was concluded therefore that , as neither aggregate attrition,
water loss nor the reduction of clay agglomeration caused the air loss
121
it was associated with cement hydration or particle size/agglomeration
effects .
Graph
4.
The mixing time effect fo r sand 'IJi th the minus 75\JlTI fraction
16
removed.
10
o:::
--;
~.
x control sand untreated
sand all:> 75~m
<
~
8
Ô
4
OL-____
~
___________________
a
12
::'6
20
24
28
TIME IN MINUTES
In order to investigate the inf1uence of cement on air loss with time
a series of mixes were made with varying cement contents and subjected to
pro1onged mixing using a fixed amount of air entraining agent . These
clearly showed the influence of cement content in depressing the air.
In
addition , analysis of the form of the graphica1 re1ationship, rep1icated
many times , a1ways revea1ed the same characteristic forms, especial1y at
higher air contents .
The form of the relationship, shown in Graph 5 , may be seen to
comprise three interconnecting phases.
These can be considered as propagation , initial 10ss which is rapid
at first and then lessens and final 10ss, the period at which the rate of
10ss increases, marked1y so if considered as a function of initia1 air
contento
This 1ast phase is particu1ar1y noticeab1e at high air contents,
1ess so or perhaps even absent at low air contents.
It was seen that there was a re1ationship between cement content and
mixing time in two specific areas, the propagation phase and the loss
phase.
...
----------------------------122
Both the time to reach a maximum , i.e . the length of the pro pagation
ph as e and the total air gained increased with an increas e in cemen t
content o Indeed , at low cement contents there was of cours e a i r loss and
not ai r ga in.
It is b el iev ed that these ph enom en a ar e caused by the adsorption of
AEA on to the ce ment . This adsorption appears preferential , in that the
AEA reacts first with the c em e nt such that it doe s not e ntrain air.
It
then appears to progressively desorb and b egin to entra i n air . The reason
behind this phenomenon and its reversal is unclear .
Graph S .
The
effec~
o.oge
of c emen~ co ntent o n the mixing time effect us in g
.-'lEA on to caI dry mix .
36
32
O O . P.C
c:::
~
""
:6
':t:==
===
-----......... 33% O . P.C
50% O .P. C
ôô% O.P. C
d
O
3
16
2d
32
dO
TIME IN MINUTES
d8
Repeat in g this work at dosages of AEA th at were
cement weight gave the form of relat i onship shown i n
in ter e sting to no te that the final a ir con tent after
as desorption has taken place was virtually the same
11 . 2% f or a l I four mixes .
a fixed percentage of
Graph 6 .
It is
what is hypothesised
at between 9 . 2% and
The final f actor affecting air loss with time that was invest i gated
was the initial air content o It was f oun d that the amount lost was in the
same o rd er at between 3 & 4 . 7% , for mixes of i nitial air contents that
varied widely from b etween 27% down to only 6% .
Mix Consistence
It is generally accepted in the fieId of Concrete Technology that air
content inc r eases with slump (consistence) (3 , 4,5,6)
but some source s
qualify this statement by restricting it to a maximum slump of between 150
123
and 200mm (7,8,9 , 10)
One source , the PCA (11) reports a decrease at very
high slumps.
33% op c
50% opc
õõ% :>pc
14% opc
z
6u
8
16
24
TI~~E
Graph 3.
~he
e! '~ ec~
of
cemen~
[N
32
40
48
'lINUTES
concent fo r 0 . 1% AEA on cemenc
~eight .
The Author ' s work showed a gradual and slight reduction in air
content as consistence increased and this showed a linear relationship
with air loss directly proportional to increase in water contento
It is suggested that this is due merely to an expansion of the total
volume of aqueous phase, the actual amount of air bubbles r emaining the
same, t hus reducing in percentage terms .
Temperature
An empirical relationship developed by RMC Mortars Limited is in agreement
with data reported by Hercules Inc (12) and is shown in Graph 7 .
Graph
7
~he
~ela~~onship
af
tem pera~ ure
~o
e n t~ained .
sne amount of air
.----2
~ mp iri~a l
- - - - - Lla 1:a du e
OL-____
~8
...
~~la~lonsnip
developed jy
2.0
TEMPERATURE
~
20
____
~
22
____
~
°C
____
24
~
':0
____
26
~MC
Lt d
~or~ars
t-i ercules =nc.
~
____
28
~
____
30
~
32
124
This has proved satisfactory in practice for admixtures based both on
vinso l resin and on synthetic surfactant formulations.
Air Entraining Agent Concentration : A great deal o f work ha s been
carried out in this area , and accordingly a comprehensive treatment is not
presented here , although the general form of the relationship is
considered .
There are grounds for treating the relationship between AEA addition
and air content as linear for constant cement content o At modest air
contents it is commonplace so to do as a convenience when marketing or
usi n g the admi x ture ; thus the common re l iance on admixtu r e dosage
expressed as a percenta ge of the cement content o
Although widely used t his approach is incorrect , the correct
relationship having been àiscussed by, amongst others , Backstrom
et al (13)
anà Lauer (14)
The present research showed that the assumption o f linearitv is
barely justified at lower addition levels . This region is the straight or
nearly straigh t part o f the hyperbolic or parabolic relationship shown in
Graph 8 .
Graph 8 .
The eff e c~ of AEA co n centration on air conten~ .
l30 ~
40
x
<
~4%
OPC
? / 33% OPC
20
~
10
O
4
2
6
8
10
l4
12
AMOUNT OF ..l,EA AS % OF CEMENT .I/EIGHT
The non l inear form of the relationship , with an increasing falling
off of the effectiveness of the AEA until a plateau is reached , is
probably due to adsorption and to the formation of AEA rich micelles in
the aqueous phase .
CEMENT
P A fi~I CLE
SIZE
Previous workers have ground a single specimen of cement progressively
more finely (15,16,17) or useà fine and coars e cemen t from different
works (18 ) .
125
It was thought worthwhile to carry out a brief investigation to
confirm the effect using a range of specific surfaces that could in
practice be used on site .
The cements used were all from the same works, they were not
subjected to laboratory grinding and were therefore in every respect
ordinary production cements . Graph 9 shows that increase in cement
fineness depressed the amount of air entrained and that this effect held
for widely varying air contents .
This data shows that increasing the cement specific surface from 280
to 440m 2 /kg reduced the air content by a virtually constant amount, 2.4%
on average , regardless of the amount actually present.
Graph 9.
~he
effect of cement fineness 8n air
le11el s .
20
con~ent
for 5 dif[erent AEA
____ 0 .2% AEA
lã
12
"-------------
O. 0 16%"EA
~ 0 . J1%
."EA
•.:r:;
<
3
õi'.
4
o
~
-----
:].004% AEA
0 . 002% AEA
280 320 360 400 440
CEMENT SIZE AS S?ECIFIC SURFACE m' Ik g
CEMENT CHEMISTRY
It is alfftcult to obtai n a cement of different chemistry without other
diffe rences in e.g. particle size , but it proved possible , by using a
foreign cement, to obtain a material with exactly the same specific
surface as one of those already tested but with very different composition .
This material was a Danish white cement with a specific surface of
430m 2 /kg , virtually identical to the 436 of the rapid hardening Portland
cement previously used . The difference in composition is shown in Table 2 .
Although Chatterji (19) reported that white cement produced the same
air content as Portland cement , his materiais varied in specific surface
thus making a comparison of less value , in the writer's view.
In the present investigation there was a defini te decrease in the
amount of air entrained with the grey cement , such that the white cement
mi x entrained 22% more, as shown in Graph 10.
L
126
TABL::: 2
Comoarison of Lhe composi tion of th e , ,h i te c ement and th e RHPC.
CEMENT TYPE
?ROPSRTY
'I/HITE
CE MENT
5PSC:FIC SüRFACE
(mlk g )
5iO,
CaO
50,
C, S
C, S
C1
430
24
69
1 .9
.-\
C.1 AF
Na, O
Grap h 10.
RHPC
TYE'ICAL OPC ( FOR
COMPARISON ONLY )
.136
365
20
64
2.3
20
63
80
3.4
:59
9
13
~4
4
10
1
10
6 .6
S.3
0 .2
O. S
0.7
59
The effecc o f c hange in cemenc com posi tion on a ir co ntento
20
x
16
'"hi te c emenc
::lHPC
12
,...,
<
8
*4
O
,J04 , 008 , 0 12 ,016 .020
AEA% DF TOTAL ORY ~T .
SAND PARTICLE SIZE
In or der to i nve stigate the effect of individual sand particle sizes,
without interactive effects of a complete range of particles , a ' standard'
sand was s ie ved in to s eparate sizes .
Each size was tested on it s own to investigate the a ir entrained for
a cement:sand mix of totally single sized aggregate and the results are
shown in Graph 11 .
127
~raph
~l.
The
effec~
OI sana
~ar~~cle
size 8n air concent :or single sized
sand.
30
20
<
~o
150 300 600
2 . 36
1. 18
8 . S. SEIVE SIZE
( ~m / mm )
The results confirmed that sand particles in the so called
inte rmed iate range promot e air entrainme~t bu t the present work gave the
maximum air content with 1 . 18mm particles , compared with the 300 ~m found
most effective in other work ( 20) .
Graph 12 shows fur ther work on th is topic and it is seen that the
co arser sand can attain much higher air co ntents than the fin e compar i son .
Graph 12.
The effec1: o f sand ;Jar'::icl e si ze on air com:en t .
24
20
15
:::
<
/, 36
I
iTlffi
sanà
I
o L -__________
. 02 . 0 4
~
150 11m sand
____________
. 06
. 08
% AEA ON !OTAL JRY
0 .1
~T .
. 12
0lIlIIII
128
Sand Composition and Particle Shape
Although previous work has shown tha( ~and composition, e.g . whether polar
or non - polar can affect air content 14
the writer failed to validate
this thesis and obtained the same air contents with both limestone and
silica sand.
With respect to particle shape, sands artificially graded to the same
grading but with rounded particles and angular particles were shown to
entrain similar amounts of air.
THE AFFECT OF ENTRAINED AIR ON MORTAR
There are two classes of effect of the air . These being the effect on the
plastic or wet properties of the mortar and the hardened or dry
properties.
Each is considered hereafter.
The Plastic Properties
The ease af use perceived by the brick or blocklayer is enhanced as is the
homogeneity/resistance to segregation and bleeding and to some extent the
cohesion and adhesion.
An obvious effect is reduction in unit mass again leading to easier
usage.
These effects increase with increasing air contents up to leveis tha t
are grossly excessive for proper usage, this by virtue of the concomitant
reduction in desired hardened properties as the excessively high leveis
are reached.
Thus the maximum is imposed, not by the plastic properties but by the
hardened ones.
The Hardened Properties
As air content increases the resistance to freeze thaw cycling generally
increases.
Although sometimes thought to relate directly, the air content
is itself only a secondary function and it is the bubble spacing factor
that is directly related to enhancement of freeze thaw resistance.
The
air content is a function of this and of the size of the bubbles.
Thus small air contents can provide excellent freeze thaw resistance
provided that their bubble spacing factor is sufficiently low.
This will
be so for low air contents, if the bubbles are of very small size and this
approach is not practicable with the current air entraining agent
technology; the bubbles are much larger than desired for this purpose to
be fulfilled .
It is of interest that although it was long suggested that suffici ent
air was required to accommodate the we l l known volume changes of water at
01' about freezing point Li tvan (21)
amongst others has shown that this is
not really the case , a much more sophisticated m~chanism of diffusion
obtains .
With the bubble sizes generally available currently there is little
reduction of the desired hardened properties at up to perhaps 15 to 20% of
129
air . Indeed, sometimes a modest increase in strength is recorded at low
and medium leveIs due to water cement ratio reductions.
However, at leveIs of 20 to 25% air reduction in strength becomes
noticeable and it is suggested that the prudent maximum is at about the
20% leveI .
REFERENCES
1.
Beningfie l d, N.E.
Air Entrainment of Mortar - a bibliography.
Paper submitte d to the 8th International Brick/Block Masonry
Conference Ireland ' 88 .
2.
Burg, G.R .U . Slump loss, air loss and field performance of
concrete.
ACI Journal , Title No. 80 - 34, July - August 1983 , pp .
332 - 339.
3.
British Ready - Mixed Concrete Association.
Air entrainment and
ready- mi xed concrete . BRMCA, Middlesex, March 1972 . Technical
Report No. 2.
4.
Wr ight , P.J . F .
Entrained air in concrete .
Proc . Inst i tution
of Civil Engineers , Vol. 2 , Par t 1 , January - November 1953 .
Paper No. 5915 , pp . 337- 358 .
5.
Fulton , F . S. Concrete technology - A South African handbook.
Portland Cement Institute , Richmond , Johannesburg , South
Africa, 1969 .
6.
Bartel , F.F.
Air content and unit weight . ASTM STP 169B
' Significance of Tests & Properties of Concrete & Concrete
Making MateriaIs ', Chapter 10, pp . 122 - 131.
7.
Brown , B . V . Use of air entraining agents in concrete production.
RMC Technical Services Ltd., Mi ddlesex, Jan 1981 . Tech Note 116 .
8.
Brown, B.V.
Air entrainment - Part 2 .
IM/07/3 - No . 81 , pp . 45 - 46 .
9.
Suthe r land , A.
Air entrained concrete .
Concrete, January 1983,
C&CA 1974 .
10 .
Russell , P.
Concrete admixtures.
Margate , 1983.
Eyre & Spottiswoode Ltd. ,
11.
Portland Cement Association . Concrete Information Ser i es - Ai r
entrained concrete.
Concrete, July 1968 .
12.
Hercules Powder Company . Vinsol air entraining agents for
cement products. Hercules Powder Co ., Delaware , USA . Bulletin
PC - 178.
130
13.
Backstrom, J . E . et aI.
Origin, evolut ion and effects of the
air void system in concrete. Part 2 - Influence of type and
amount of air entraining agent.
Proc . ACI V. 55 , ACI Journal
Vol . 30 , No. 2 , Title No . 55- 16 , August 1958. pp. 26 1- 272.
14.
Lauer , K.R.
The mechanisms of air entrainment in mortars .
Thesis for Doctor of Philosophy, Purdue Unive rsi ty , August 1960
University of Mi crofilms Inc . Ann Arbor , Michigan, USA . Mie.
60 - 6119 .
15 .
Mayfield, B . & Moreton , A. J .
Effect of fineness of cement on
the air entraini ng proper t i es of concrete . Civil Engineering
Public Works Review , January 1969 , pp. 37- 41.
&
16 .
Scripture , E . W. et aI . Effect of temperature and surface area
of the cement on air entrainment.
Proc . ACI V.48, ACI Journal ,
Vol . 23 , No . 3 , Ti tle No . 48 - 15 , November 1951, pp. 205- 210.
17 .
Greening , N. R . Some causes for variation in required amount of
air entraining agent in Portland cement mortars . Journal of
PCA Res. & Dev . Labs., Vol 9 , No . 2 , PCA , USA , May 1967 .
18 .
Hall, W.K.
An investigation into the influence of the fineness
of current production cements and cement conten ts on air
entrained concrete . 4 th Convention, Association of Concrete
Technologists , 1 - 3 June 1976 .
19 .
Chatterji , S . et aI.
The characteristics of air - bubble systems
in hardened ......... air entraining agents - plasticiser
combination . Silicates Industriels, July - August 1978 , pp .
153 - 156 .
20 .
Mielenze, R . C . et aI . Origin , evolution and effects of the air
void system in concrete . Part I - Entrained air in unhardened
concrete . Proc. ACI V.55, ACI Journal Vol . 30 , No . 1 , Title
No. 55 - 5 , July 1958 . pp . 95- 121 .
21.
Li tvan , G. G.
1973 .
Frost activiti es in cement paste.
July - August