Part 3 of a 4-part series
Measuring
air in
fresh
and
hardened
concrete
By Ken Hover
For the most accurate
results, use only trained
technicians to conduct
air-content tests
Equipment for microscopic examina-
tion of hardened concrete should be
operated by a trained specialist.
revious articles have dis- concrete for two practical reasons. essential to a consistent yield. Air
P cussed why and how air
bubbles are intentionally in-
corporated in fresh concrete,
and how batching, mixing, transport,
placing, and consolidation influence
First, specifications may require a
minimum or maximum air content.
Testing shows if the concrete meets
specified requirements. Second, a
consistent air content indicates uni-
content tests can be important for
control of concrete production.
Typical locations for sampling con-
crete to measure air content are at the
batch plant, at the chute of the ready
the total volume and gradation of the form materials and batching, and is mix truck, and at the point of place-
air-void system in hardened con- ment. Air content is most often mea-
crete. The measurement of total air sured at the truck. Tests of samples
content in fresh concrete and of air A single microscopic taken from truck chutes determine
volume and air-void gradation in concrete quality as delivered to the
hardened concrete are the subjects of analysis will evaluate site by the concrete supplier. The
this article. about 1,000 air voids. proper location to sample air content
There may be 10 to 15 depends on the reason for testing
Air in fresh concrete and the construction specifications.
Current methods can measure billion air voids in a Concrete with an adequate vol-
only the total air content of fresh con- cubic yard of air- ume of properly sized, stable bubbles
crete, not the gradation of air bub- entrained concrete. at the truck chute can normally toler-
bles. Even so, it’s important to mea- ate the changes resulting from typi-
sure the total air content in fresh cal handling, placing, and consolida-
tion. Air content in-place, however, is
Equipment for Determining
determined only by testing samples
of hardened concrete. Total Air Content in Fresh Concrete
Measurement of air content is usu-
ally done along with slump tests and
molding of test cylinders. Testing
technicians under contract to the
owner often perform these tests, but
some quality-management programs
assign routine testing to the contrac-
tor. Producer quality-control person-
nel frequently are involved in this
testing to monitor product uniformi-
ty and consistency. Technicians or
others who conduct these tests
should be trained to conduct them
properly. Certification as a field test-
ing technician is one way to verify
this training.
Standard tests
Total air content in fresh concrete
is determined by one of three stan-
dard tests:
• ASTM C 138, “Unit Weight, Yield
and Air Content (Gravimetric) of Pressure meters for ASTM C 231 Volumetric air meter, or “roll-o-me-
Concrete” (Unit weight test) test. Type A meter is larger and ap- ter” for ASTM C 173 method. The
plies pressure gradually to the con- volumetric method should be used for
• ASTM C 173, “Air Content of crete sample. highly porous aggregates. The test
Freshly Mixed Concrete by the requires hard work by the operator.
Volumetric Method” (Roll-o-meter
test)
Measurement of unit weight
• ASTM C 231, “Air Content of by ASTM C 138 allows cal-
Freshly Mixed Concrete by the culation of fresh concrete air
Pressure Method” (Pressure meter content. Accuracy of results
test) depends on accurate batch
weights and information on
Equipment for each standard test properties of mix materials.
is shown. Each method is effective Significant changes in unit
and reliable when performed prop- weight indicate variability
erly, but erratic results can occur in air content.
when the operator is careless. Each
method requires that a sample of (Note: Concrete is consolidat-
fresh concrete be placed in the ed carefully in each method to
equipment and consolidated in a make results consistent but
standard way. This does not repre- not necessarily simulate in-
sent the actual consolidation of con- place conditions.)
crete in-place, but is intended to Photos Courtesy Of Construction Technology Laboratories Inc.
make test results more consistent.
Agreement among the methods air content, and no consensus as to dated in a unit weight bucket, then
depends on the concrete and on the which test is more reliable. Pressure weighed on a scale. The actual and
test operators. Some laboratory stud- meters and roll-o-meters are used for theoretical unit weights are then
ies show the three methods essential- most field testing. compared. A mix generally over-
ly measure the same thing. However, Unit weight test. This test esti- yields if the air content is higher than
in field studies under less controlled mates the volume of air present on planned, and under-yields if the air
conditions and with less experienced the basis of the weight/volume rela- content is lower. The air content is es-
operators, the three tests have pro- tionships of a sample of fresh con- timated by a calculation based on the
duced widely different estimates of crete. Concrete is placed and consoli- known quantities and properties of
the mix materials. This test is accu- with less experienced operators, • Evaluate the effectiveness of air-
rate if the batch weights and specific the results were scattered within entraining admixtures
gravities of mix ingredients also are about 25% of the average reading. • Determine compliance with speci-
accurate. The pressure applied during the fied requirements for characteris-
Roll-o-meter. In this method, test can force water into aggregate tics of the air-void system
water is added to a sample of con- pores. The meter then reads the vol-
• Estimate the frost resistance of the
crete and agitated to wash air bub- ume of air bubbles in the cement
concrete in-place, often when frost
bles out of the concrete. Water is paste plus air in the aggregate pores.
resistance is questioned based on
mixed into the concrete by shaking, To compensate for this behavior,
fresh concrete air content
rocking, and rolling the container. ASTM C 231 requires subtraction of
Air bubbles float into the neck of an “aggregate correction factor” from Microscopic analysis of hardened
the container where the water level the meter reading. In practice this concrete is done by a trained special-
drops. The method is simple but re- correction often is ignored. This con- ist under laboratory conditions—it is
quires hard work from the techni-
cian. Different operators can get
dramatically different results from
the same concrete since the amount
of air released from the concrete de-
pends on the operator’s effort in ag- Fine and coarse ag-
itating the meter. Relatively unsta- gregate, cement
ble and buoyant large bubbles wash paste, and air voids
out quickly and easily, but smaller are visible in a con-
bubbles may require more work crete sample after
and time to float to the top. Though magnification.
these small bubbles contribute the
least to total air content, they may
contribute the most to frost resis-
tance. In one test, air bubbles con-
tinued to accumulate 20 minutes af-
ter shaking the meter.
The volumetric method should be tributes to errors in the reported air not a field test, nor should it be done
used for highly porous (such as most content, but may not be significant by amateurs. As currently performed,
lightweight) aggregates. If the opera- for low-porosity aggregates. analyses of air voids in hardened con-
tor can easily pour the concrete from crete is tedious and time-consuming,
the container after the test, the con- Air in hardened concrete and sensitive to the operator’s tech-
crete has been mixed well enough to Current test methods cannot mea- nique, skill level, experience, and
wash out most of the bubbles. If stiff- sure total air content of concrete af- judgment. The experience of the per-
ened material is found in the bottom ter placement and consolidation and son behind the microscope is more
of the pot, the test results are invalid, provide no information about the important than the newness of the in-
but the test can be repeated. gradation of air bubbles in fresh strument or the sophistication of
Pressure meter. In the pressure concrete that become the air-void computer-aided analysis. The proce-
meter test a sample of fresh concrete is system in hardened concrete. Sam- dure also is expensive, often costing
squeezed by compressed air to about ples of hardened concrete can be ex- at least $300 to $500 per specimen.
twice normal atmospheric pressure. amined microscopically to evaluate Smaller testing laboratories may not
Since air is the only compressible com- these properties. However, expect perform these tests but can recom-
ponent of fresh concrete, the change in differences in air contents in fresh mend a firm that does.
sample volume indicates the total air and hardened concrete due to the ASTM C 457, “Microscopical De-
content. There are two variations of effects of placement and consolida- termination of Parameters of the Air-
the pressure meter. The “Type A” me- tion. ASTM C 231 states, “the air void System in Hardened Concrete,”
ter applies additional pressure gradu- content of hardened concrete may often is confused with the related
ally, and the “Type B” meter applies be either higher or lower than deter- ASTM C 856, “Petrographic Exami-
pressure suddenly. The Type B pres- mined by this method.” nation of Hardened Concrete,” since
sure meter is more common. Microscopic evaluation of the the same specimen may be used for
In controlled conditions with hardened concrete air-void system both tests. Both methods require a
skilled operators testing the same is performed to: sample of hardened concrete, often a
concrete, results of the pressure me- core extracted from an existing struc-
ter can be within 5% to 10% of • Evaluate placing, consolidating, or ture. A specimen is then cut and
the average meter reading. In tests finishing operations carefully polished for viewing with a
microscope that magnifies the image, specific surface, , is usually 600- mated, computerized methods may
typically 50 to 125 times the original 1,100 inches2/inches3, and the spac- eventually perform these tasks using
size. Some labs use higher or lower ing factor, L, is usually less than image analysis. For now, the human
magnifications. about 0.008 inch. observer is the key element.
When examined through the micro- Results of microscopic analysis are
scope, air voids appear as random- statistical estimates rather than direct References
sized circles and irregular shapes. The measurements. The results are based 1. K. Hover, “An Analytical Investiga-
sizes of these shapes on the two-di- on random sampling of a small per- tion of the Influence of Air Bubble Size
mensional plane don’t necessarily rep- centage of the voids and indirect de- on the Determination of the Air Con-
resent the size of the original three-di- termination of void sizes. A single mi- tent of Freshly Mixed Concrete,” Ce-
ment, Concrete and Aggregates,
mensional voids. A small circle could croscopic analysis will evaluate about
ASTM, Summer 1988, pp. 29-35.
be a random cut through the center of 1,000 air voids. There may be 10 to 15
2. L. W. Roberts and R. D. Gaynor,
a small spherical void, or a cut near billion air voids in a cubic yard of air-
“Discussion of ‘Comparison of the Air
the edge of a much larger void. Since it entrained concrete. Uncertainty or er- Contents of Freshly Mixed and Hard-
isn’t possible to measure the true di- ror is expected in such estimates. Vari- ened Concretes by H. C. Ozyildirim,’”
mensions of air voids, index parame- ations are caused by the statistical Cement, Concrete and Aggregates,
ters based on geometry, statistics, and nature of the procedures and calcula- ASTM, Summer 1991, pp. 16-17.
some simplifications are used to de- 3. D. Whiting and D. Stark, “Control of
scribe the air-void system. Simplified Air Content in Concrete,” NCHRP Re-
descriptions normally are useful for port 258, 1983.
The air content of
comparing a particular concrete with 4. K. Snyder, K. Hover, and K. Nate-
concrete known to have good, or bad, hardened concrete saiyer, “An Investigation of the Mini-
performance. may be either higher or mum Expected Uncertainty in the Lin-
ear Traverse Technique,” Cement,
ASTM C 457 describes two alter- lower than that Concrete and Aggregates, ASTM,
nate procedures for systematic exam-
ination of the specimens. These are
determined by this Summer 1991, pp. 3-10.
method — ASTM C 5. R. Pleau and M. Pigeon, “Precision
called the linear traverse procedure Statement for the ASTM C 457 Micro-
and modified point count procedure. 231, Air Content by scopical Determination of the Spacing
Both procedures can be used simply Pressure Method Factor and Other Characteristics of
to estimate the total air content, or al- the Air Void System in Hardened Con-
so determine other parameters of the crete,” Cement, Concrete and Aggre-
air-void system. The modified point gates, ASTM, Winter 1992, pp. 118-
126.
count procedure is faster and used tions. They also are caused by differ-
more often. Paste and fine- and ences in specimen preparation, test 6. M. J. Simon, R. B. Jenkins, and K.
Hover, “The Influence of Immersion Vi-
coarse-aggregate contents also can equipment, magnification, and opera- bration on the Air Void System of
be estimated. tor skill and judgment. Hardened Concrete,” Durability of
The parameters of the air-void sys- In one set of controlled tests, air con- Concrete, Proceedings of the Idorn
tem determined from microscopic tents reported for a single specimen, Symposium, ACI SP-131, 1992, pp.
analysis include the specific surface prepared by one laboratory and ana- 99-126.
and the spacing factor. The specific lyzed by several others, varied from
surface ( in test reports) represents 3.9% to 5.3%. Researchers have pro- Ken Hover is a professor of civil engi-
the total surface area of air voids, di- posed that results of the C 457 analysis neering at Cornell University, Ithaca,
vided by the total void volume. If include estimates of error (Refs. 4 and N.Y. He also is a Fellow of the Ameri-
two air void systems have the same 5). In addition to the uncertainties in can Concrete Institute and chairman
total air content, the system with a analyzing a specimen, air content and of ACI Committee 308 on Curing.
finer air-void gradation will have void gradation are not necessarily uni-
a larger specific surface. The spacing form throughout a concrete placement
factor (L in test reports) is related in because of handling, placing, com- This is the third article of Con-
some manner to the maximum dis- paction, and finishing. Results depend crete Construction’s four-part se-
tance from any point in the cement on the number of cores and their loca- ries on air in concrete. “Why is
paste to an air void (Ref. 3). The spac- tion in the placement. In one experi- There Air in Concrete?” appeared
ing factor is a useful characteristic for ment, the air content within a 4- in the January issue (pp. 11-15),
describing air-entrained concrete, square-foot slab varied from 2.5% to and “Air Bubbles in Fresh Con-
though it is based on simplifying as- 5.5%, and from 3.0% to 6.5% in a 16- crete” is in February (pp. 148-152).
sumptions that don’t represent real square-foot slab (Ref. 6). Part 4 of the series will suggest
air-void systems. Uncertainty can be reduced, but not practical specifications for air-en-
For conventional concretes pro- eliminated, in evaluating the air-void trained concrete.
portioned for frost resistance, the system of hardened concrete. Auto-
PUBLICATION #C930275, Copyright © 1993, The Aberdeen Group, All rights reserved
