Relationship Between Spray Erosion Tests and the
Performance of Test Specimens in the Field
K.A.Heathcote (University of Technology Sydney)
R. Sri Ravindrarajah (University of Technology Sydney)
Abstract
This paper presents a “Limit State” approach to the durability of earthwalls whereby
the climatic “loading” of a site is matched with the materials “resistance” in the field.
The suggestion is made that a modified spray test may be used as a measure of field
resistance. The paper presents field and laboratory tests on split face specimens in
order to justify this. The field specimens were subjected to the weather conditions
for a period of about 5 months. Hourly wind and rain records were recorded during
this time and the amount of water impacting the test specimens calculated using a
driving rain index. The laboratory specimens were subjected to a modified spray
test. A comparison was then made between the rates of erosion per litre of
impacting water and comments were then made on this relationship.
Keywords
Durability, earthbuilding, pressed earth, spray testing, soil-cement, wind driven rain,
accelerated durability tests
Introduction
The use of earth as a building material dates back to at least the Ubaid period in
ancient Mesopotamia (5000-4000 B.C.).
Mud bricks or Adobes as they are
sometimes called, were made by mixing a plastic mixture of clayish material with a
binding agent such as straw. With a rainfall rate less than about 200 mm per year
and little stone mud brick was the natural choice of building material and it is still
used extensively in the region today. Mounds formed by the accumulation of eroded
earth buildings are visible today but little else remains .
There are a few complete mud brick structures that have survived the ravages of time
through regular maintenance. Perhaps the oldest existing unprotected earth structure
is the Pueblo at Taos, in New Mexico, which is reportedly over 900 years old (Fig 1).
Here the combination of a dry climate with regular maintenance has led to the
survival of this three storey building.
Fig 1 Adobe Pueblo at Taos
Earthbuilding was an established form of construction in many parts of the world in
the 19th Century, with most of the buildings rendered for weather protection. With
the advent of the industrial revolution earthwall construction declined in
industrialised countries with very little activity in the first half of the 20th Century.
Shortages of material following the Second World War saw an upsurge in interest in
earthwall construction. This interest was stimulated by the work that had been done
on stabilised soil roads in the early twenties and thirties. Sheets and Catton [Sheets
and Catton, 1938] set out the basic principles of cement stabilised earth which were
established at that time.
Because of its suitability for mechanisation pressed earth bricks were the natural
development from the traditional mudbrick. In this method a dryish soil is placed in
a steel mould and compacted under high pressure. The moulds are typically 290 mm
long, 140 mm wide and 90 mm high yielding bricks weighing about 7 kg, although
in Australia larger bricks weighing around 20 kg are common. Soils suitable for
making pressed earth bricks are generally sandy loams with clay contents less than
20 % and low silt percentage , however a wide variety of soils can be used with clay
contents up to 35 %.
The extra density obtained by pressing bricks makes them stronger and more
resistant to erosion by rain. Typically densities of around 1750 kg/m3 are achieved
compared to around 1600 kg/m3 for traditional Adobe bricks. To improve durability
pressed earth bricks are generally stabilised with from 5 % to 12 % cement, with
around 8 % being generally suitable for most soils.
Durability and Climate
The late 80’s saw a significant increase in interest in cement stabilised pressed earth
blocks both in developed countries such as Australia as well as developing countries
such as Ghana.
Coupled with this increased interest was a trend towards
“performance based specification” of building materials. Such codes require proof
of the resistance of uncoated pressed earth bricks to wind-driven rain in various
climatic conditions.
The original development of mud brick buildings in Mesopotamia, Egypt, China and
North Africa was influenced largely by the climatic conditions. Annual rainfall in
these areas does not exceed 50 cm and in Mesopotamia it is as low as 20 cm.
Erosion by wind driven rain was therefore not a significant problem and
consequently wall coatings were not employed.
When this form of construction made its way to areas of higher rainfall such as the
European Continent surface coatings became necessary to protect the walls from
driving rain. Where such coatings are effective (such as the stucco coatings in New
Mexico) the question of the resistance of the earthwalls themselves to rain is
principally one of water permeability through rising damp. In these cases bitumen
has been effectively used to waterproof the walls.
Recently however, earthbuilding has been extended to climates with annual rainfalls
in excess of 100 cm or sometimes uncoated walls have been used in areas of annual
rainfall between 50 and 100 cm. In these cases the resistance of unprotected walls to
driving rain is a major problem and needs to be addressed in the selection of
materials and methods of construction. In earthbuildings with large protective eaves
this is not much of a problem but where the walls are more exposed the majority of
natural soils will in the long term suffer deterioration unless regular maintenance is
carried out (Fig 2).
Fig 2 Mud Brick Building Constructed in Sydney in 1949
Quite clearly earthbuildings are not as weather resistant as most of the "modern"
buildings constructed out of stone or fired clay brick. This does not mean however
that earthbuildings cannot be designed to resist the climatic conditions of the area in
which they are to be built. What is needed is an understanding of the relationship
between the climatic “loading” and the material “resistance” so that an adequate
factor of safety against material deterioration may be achieved.
This “Limit State” concept means that unlike other materials where the erosion
resistance is fixed (such as fired clay bricks) the design of earthwalls offers the
designer the opportunity to uniquely match the erosion resistance of the material with
the factored climatic loading.
Laboratory Testing and its Relationship to Field Testing
In order to predict the performance of earthwall specimens in the field (and to
provide a design methodology) it is necessary to correlate laboratory test results with
the in-situ performance of similar specimens. At the present time in Australia the
standard test for durability of earthwalls is the spray test as defined in Bulletin 5
[Bulletin 5, 1987] , which is referenced in the Building Code of Australia.
This spray test uses a standard nozzle with thirty five 1.3 mm diameter holes placed
470 mm from the specimen and with a water pressure of 50 kPa. In the study
conducted by the authors a modified spray test was used consisting of a
commercially available nozzle which has a rotating nozzle inside and which
produces drop sizes of the order of 2 mm at a velocity of around 12 m/sec (Fig 3).
The turbulent spray produced by this nozzle was considered to more realistically
simulate the turbulent effect of rainfall.
Fig 3 UTS Spray Test Rig with “Fulljet” Nozzle
Spray tests using this apparatus were performed in the laboratory on one face of split
specimens and these results were correlated with the field performance of the
opposite faces in the field. Both wind and rain records were monitored in the field
and correlated with the velocity and quantity of water delivered in the laboratory
spray tests.
Field Specimen Details
In order to correlate the field performance of specimens with wind and rain records it
was necessary to locate the specimens as close as possible to an automatic recording
station. There are only six Bureau of Meteorology weather stations in the Sydney
area that provide automatic hourly collection of wind and rain data. For ease of
access the station at the southern end of the main runway of Sydney’s International
Airport was chosen (Fig 4). This site has an unobstructed fetch to the south and as
the majority of the wind driven rain was expected to come from the south the test rig
was oriented to face south.
The site is operated by the Sydney Airport Meteorology Office of the Bureau of
Meteorology of the Commonwealth of Australia and is site number 066037 in their
network of sites. The observation station was opened in 1929 and is situated at
Latitude 3356’28”S and Longitude 15110’21”E. The wind frequency recorder is
situated along the runway extending into Botany Bay at a height of 10 metres. The
average annual rainfall for the site over this period is 1106 mm.
Rainfall records are logged at hourly intervals and include maximum and minimum
hourly temperatures, relative humidity and dewpoint in addition to the rainfall in mm
over the hourly period. The maximum hourly rainfall recorded during the two years
of records analysed was 28.8 mm on the 21st January 1999. This roughly translates
into a 6 minute rainfall intensity of 90 mm/hr .
Wind records indicate the 10-minute average wind speed and direction as well as the
hourly maximum wind speed and the time at which it occurred.
Fig 4 New Test Rack Placed at Airport in July 1999
The amount of driving rain impacting the specimens (Fig 5) was determined using
the formula for driving rain index proposed by Lacy [Lacy & Shellard, 1962].
Fig 5 Driving Rain Rose for Period of Exposure of Specimens
Performance of Initial Test Series
An initial series of tests was carried out in 1998 involving 6 test specimens made
from a sandy clay with cement contents of 3, 4, 5, 6, 7 & 8 %. Cylindrical specimens
were pressed using this soil to achieve a density of around 1750 kg/m3. Following
curing the specimens were split into two, with one of the split faces being exposed to
the weather and the other being subjected to the spray test. Prior to field and
laboratory testing both split specimens were dried and weighed, this process being
repeated following testing. The duration of the spray test was determined in each
case by the need to obtain measureable erosion.
The field specimens were subjected to the weather conditions for a period of around
5 months and their condition after that time is shown in Fig 6.
Fig 6 Specimens After Field Testing
The values of erosion measured following testing was divided by the volume of
water impacting on the specimens, either measured directly in the case of the spray
tests or through calculation using measured wind and rain records in the field testing.
For both the field and laboratory experiments the weight lost was inversely
proportional to the cement content as can be seen from Figs 7 & 8 .
Fig 7 Variation of Mass Loss with cement Content for Laboratory Tests
Fig 8 Variation of Mass Loss with Cement Content for Field Tests
Discussion of Results
For the higher cement contents (6, 7 & 8 % ) the correlation between weight loss in
the field and weight loss in the laboratory was poor (Fig 9). Two reasons for this can
be identified
1.
The amount of erosion in these cases was very small and inaccuracies in
measurement could have been critical.
2.
In order to get measureable erosion these specimens were sprayed for
much longer periods of time, of the order of 2 to 3 hours. Recent tests
have shown that the erosion rate drops off significantly with time, thus
introducing an error into the calculations of water volumes. This work is
ongoing and will be presented at a later date.
There was good correlation between weight loss in the field, compared to weight loss
in the laboratory, for cement contents 3, 4 & 5 % (Higher erosion values in Fig 9).
The ratio between the two in this case was approximately 4.5, i.e. the erosion in the
field was four and a half times that of the erosion in the laboratory per unit volume of
water impacting on the surface.
Some of the possible reasons why this ratio may be greater than one are
1.
Difference between incident wind velocity and spray velocity.
2.
Effect of angle of attack of rainfall.
3.
Difference between raindrop sizes and spray drop sizes.
4.
Effect of wetting and drying of the field specimens.
All of these factors are presently being examined by the authors and will be
presented elsewhere.
Fig 9 Variation Between Field and Laboratory Erosion
Conclusions
For effective prediction of the service life of earth buildings it is necessary to have an
accelerated durability test which is a reliable predictor of in-service performance.
The authors have shown that a modified spray test has the potential to provide such
evidence of field performance. This material “resistance” can then be related to the
climatic conditions (“loading”) of a site, thus providing a limit state methodology to
the design of earthwalls for durability.
References
1.
Sheets, F.T., and Catton, M.D., 1938, Basic principles of soil cement
mixtures, Eng. News Rec., June 23, p 869.
2.
Bulletin 5 - Earthwall Construction, 1987,National Building Technology
Centre, Chatswood, Sydney.
3.
Lacy, R.E., & Shellard, H.C., 1962, An Index of Driving Rain, Meteorol.
Mag., 91, , pp 177-184.