INTERNATIONAL SOCIETY FOR
SOIL MECHANICS AND
GEOTECHNICAL ENGINEERING
This paper was downloaded from the Online Library of
the International Society for Soil Mechanics and
Geotechnical Engineering (ISSMGE). The library is
available here:
[Link]
This is an open-access database that archives thousands
of papers published under the Auspices of the ISSMGE and
maintained by the Innovation and Development
Committee of ISSMGE.
The paper was published in the proceedings of the 17th
African Regional Conference on Soil Mechanics and
Geotechnical Engineering and was edited by Prof. Sw
Jacobsz. The conference was held in Cape Town, South
Africa, on October 07-09 2019.
Proceedings of the 17th African Regional Conference on Soil Mechanics and Geotechnical Engineering. 7, 8 & 9 October 2019 – Cape Town
The importance of plate load tests in geotechnical engineering practice
H.F.T. Barnard
Golder Associates Africa (Pty) Ltd, Midrand, South Africa
ABSTRACT: Stiffness and bearing capacity of soils are two of the most important parameters required for
geotechnical design. Plate load tests have been used extensively in the past for site characterisation to determine
both stiffness and bearing capacity of soils as well as other soil parameters, such as sub-grade reaction during
road construction. The aim of this paper is to demonstrate the importance of plate load tests in the field of
geotechnical engineering. In the last few years the author has been involved with research on plate load tests
and has developed an in depth understanding of the status quo of plate load tests by means of industry surveys.
Two of the main advantages of plate load tests are the simplistic test procedure and cost effectiveness of the
tests. However, there are numerous details that, if done incorrectly during testing, will result in poor quality test
data. The main objectives of this paper are to (1) provide a brief introduction to plate load tests and the factors
affecting plate load tests; (2) discuss the industry surveys that was conducted in recent years and (3) provide a
simple flow diagram for the design of plate load tests that will ensure reliable and useful results.
1 INTRODUCTION AND BACKGROUND conducted in recent years and (3) provide a simple
flow diagram for the design of plate load tests that
Plate load tests (PLTs) have been used extensively in will ensure reliable and useful results.
the past to determine the bearing capacity and the
stiffness of soil. Most geotechnical problems are gov-
erned by allowable settlement requirements. There- 2 APPLICATIONS OF PLATE LOAD TESTS
fore, geotechnical professionals are continuously
searching for improved test methods and new tech- Plate load tests have an important role in geotechnical
niques to determine soil stiffness more accurately and engineering, not only in the determination of soil
cost effectively. stiffness, but also in various other applications. The
Two of the main advantages of plate load tests are numerous applications of plate load tests include, but
the cost effectiveness of the test and the relative are not limited to the following:
straightforward test procedure. During conventional 1. To validate the compaction quality of soils during
plate load tests, the stiffness of the soil is determined construction;
by measuring the settlement of a plate placed on the 2. To estimate foundation settlements based on plate
soil surface subject to a known applied pressure. If the load test settlements;
contact between the plate and the ground is not 3. To determine the allowable bearing capacity of
smooth, bedding errors may occur during the test, soils;
where plastic strain occurs at the contacts between the 4. To calculate soil stiffness values at different ap-
plate and the ground. plied loads.
In Barnard (2013), the effect of bedding errors on It is very important that the application of the required
the accuracy of plate load tests and how bedding er- plate load tests will be agreed with the client, prior to
rors can be minimized for routine plate load tests was designing of the test methodology.
investigated. Since completing his studies, the author
has received numerous calls and emails regarding the
design and application of plate load tests. 3 COMPONENTS OF PLATE LOAD TESTS
In light of these considerations, the main objec-
tives of this paper are to (1) provide a brief introduc- The five main components of a conventional plate
tion to plate load tests and the factors affecting plate load test include (1) a reaction beam or force; (2) a
load tests; (2) discuss the industry surveys that were
345
17th ARC Conference 2019
hydraulic pump and jack system to provide the ap- 4 PLATE LOAD TEST SURVEYS
plied load; (3) a selection of loading plates; (4) instru-
ments to measure both the applied load and the plate A Plate load test survey was set-up to understand how
displacements; and (5) a reference beam system to at- plate load tests are being performed in South Africa.
tach the displacement instruments. Figure 1 shows the This survey was sent to over ninety (90 No.) geotech-
typical components of a PLT. nical consultants, contractors and laboratories in
The steel plates typically vary between 150 mm South Africa in order to understand the status of plate
and 600 mm, which are loaded using a hydraulic load testing in South Africa.
pump and jack system. It is important that the capac- The survey consisted of nine questions to investi-
ity of the hydraulic jack be large enough to safely sup- gate the various factors affecting plate load tests re-
ply the required load. The plate size selection might sults:
sometimes be governed by equipment limitation, but 1. Typical plate sizes use during testing;
preferably the required influence depth should be less 2. Type of reaction force used to apply loads on the
than 1.5 times the available plate diameter. plate;
This rule of thumb is based on the influence depths 3. Type of measuring system used to measure the ap-
for circular foundations from Boussinesq’s theory, plied loads;
where less than 20% of the applied stress occurs be- 4. Type of displacement instruments used to measure
low 1.5 times the plate diameter (Boussinesq, 1885). plate settlement during loading;
The displacement of the plate is typically meas- 5. The amount and arrangement of displacement in-
ured with three calibrated dial gauges placed 120 de- struments to measure plate settlement and tilt;
grees apart and approximately 25 mm from the plate 6. Surface preparation method used to achieve a level
perimeter, to accommodate any tilt that might occur and smooth test area;
during testing. Figure 2 shows a typical plate load test 7. Type of reference beam used to attach displace-
set-up. ment instrumentation (e.g. wood or steel);
8. Typical span/length of reference beam;
9. Standard test procedure implemented to perform
plate load tests.
In total, twenty (20 No.) completed surveys were re-
ceived that revealed very interesting results. This was
believed to be an acceptable representative sample
because there are most probably not many more insti-
tutes or companies that perform plate load tests in
South Africa on a routine basis. The following section
describes the most important findings obtained from
the survey results.
5 SURVEY RESULTS
5.1 Plate size
Figure 1. Components of Plate load tests
The results revealed the most popular plate sizes are
450 mm and 600 mm in diameter. Given that the in-
fluence of the soil depth is about 1.5 times the plate
diameter (Clayton et al. 1995), one can conclude that
depths of between 675 mm and 900 mm soil thickness
are typically being tested by plate load tests in South
Africa.
5.2 Applied load measurements
The options provided for the type of devices used to
measure the applied load have included pressure
gauges and load cells. The results have revealed that
17 out of 20 suppliers are using analogue pressure
gauges and only ten per cent are using load cells to
measure the applied loads. Another type of device
that was recorded in one case was a digital pressure
transducer linked to a laptop. These results are shown
in Figure 3.
Figure 2. Conventional Plate load test set-up
346
H.F.T. Barnard
5.5 Type and length of reference beam
The two material types of reference beams that it was
expected would be used were steel and wood, but sur-
prisingly the results have revealed that all respondents
are using steel reference beams. The reference beams
varied in length from the shortest at 1.8 m to a maxi-
mum of 10 m in one case. One supplier is using a
600 mm diameter steel ring beam to keep the dis-
placement instruments firmly in place. The average
length was 3.0 m and was recorded in 55 per cent of
all cases.
6 MINIMISING BEDDING ERRORS
Barnard (2013) showed that bedding errors can have
a significant effect on soil stiffness values when ex-
ternal measurements are used to measure the dis-
Figure 3. Preferred method to measure applied load placements of the plate. Such conventional plate load
tests can result in stiffness values of up to 50% lower
5.3 Plate settlement measurements
than stiffness values inferred from seismic tests (Bar-
The options provided for the type of displacement in- nard 2013).
struments used to measure the plate settlement in- This is of particular concern when the plate load
cluded dial gauges and Linear Variable Differential test stiffness values are used for foundation designs in
Transducers (LVDTs). The results revealed that 19 large construction projects. Plaster of Paris is the best
out of 20 of suppliers are using mechanical dial interface material to place between the plate and the
gauges and one is using LVDTs to measure the settle- soil, that best reduces the bedding errors, and is there-
ment of the plate. fore recommended that Plaster of Paris is used as in-
terface material when conducting plate load tests on
5.4 Surface preparation method
soils.
Three surface preparations methods were provided as Sometimes it is difficult or impractical to use Plas-
the possible options which would achieve a levelled ter of Paris; under such conditions well-graded sand
and smooth test area (see Fig. 4). The results revealed may be used. The reduction in bedding errors when
that approximately 60 per cent of respondents use using sand is not as much as with Plaster of Paris, but
well-graded sand in their attempt to achieve a levelled it is significantly better that not using any interface
test area and 20 per cent prefer plaster of Paris. material.
Ten per cent of respondents believe that using only
hand tools, an acceptable levelled and smooth test
area can be achieved. Other surface preparation meth- 7 INTERPRETATION OF PLATE LOAD TEST
ods recorded during the survey was the use of cement DATA
powder, fine fly-ash and quick set grout.
The average vertical displacement of the plate (ρ) ob-
tained from the three measuring devices can be used
in Equation 1 below (Timoshenko & Goodier 1951),
together with the plate diameter (D), and Poisson’s
ratio (v), to determine the secant stiffness values at
different applied loads. The contact stress (q) is deter-
mined as the applied load divided by the plate area,
which assumes a uniform pressure distribution across
the plate.
#∙%∙&∙(()* + )
𝐸= (1)
-∙.
Figure 4. Preferred surface preparation methods
347
17th ARC Conference 2019
8 HOW TO DESIGN AN ACCURATE PLATE the required influence depth criterion is met, the plate
LOAD TEST area can be calculated.
Step 4: The available load (kN) is calculated from
Step 1: The first step is to determine the required in- the lesser of the reaction force (RF) and the available
fluence depth to be tested. The required influence hydraulic jack system (HJ) as shown in Figure 5.
depth will typically be the layer thickness to be tested Step 5: The available applied pressure can now be
as specified by the client. For compaction control calculated with the plate area and available load as
tests, the influence depth is normally flexible and can determined in step 3 and 4, respectively. If the avail-
be changed to accommodate possible equipment lim- able pressure (P) is less than the required pressure
itations such as plate size availability. On the other (RP), determined in step 2, the only two options to
hand, if the aim of the plate load test is to calculate increase the available pressure, is either to decrease
the bearing capacity or stiffness value of a soil hori- the plate size (PZ) or to increase the available load
zon, it is important that the thickness of the relevant (AL) as indicated in Figure 5. Note that by decreasing
soil horizon is known. the plate size, the influence depth will also decrease
and therefore the required influence depth should be
Step 2: The second step is to determine the pressure reconsidered. However, it may be impossible to in-
required during testing. The required pressure is typ- crease either the reaction force or the hydraulic jack
ically governed by the working load expected during capacity due to equipment or reaction force limita-
operations. For compaction control tests, the required tions.
pressure should be at least three times the allowable Step 6: Once the equipment is selected such that
bearing capacity assumed during the design phase. the required pressure can be applied, one needs to de-
For stiffness tests, it may be acceptable to test up to termine the loading sequence. Once again, it all de-
1.5 times the working load during operations. pends on the application of the tests.
Step 3: In the next step, the plate area should be Step 7: The last step is to select the appropriate in-
calculated for the selected plate size. As previously strumentation for the plate load test. The decision to
mentioned, the plate size has a direct correlation to select a pressure gauge or load cell to measure the ap-
the influence depth (1.5 times plate diameter). If the plied load, as well as which displacement instruments
calculated influence depth is less than the required in- you select will depend mainly on the instrument’s
fluence depth, determined in step 1, the plate size ability to provide acceptable accuracy and repeatabil-
should be increased as shown in Figure 5, or the re- ity in the readings.
quired influence depth should be reconsidered. Once
Figure 5. Flow chart for Plate load test design
348
H.F.T. Barnard
9 CONCLUSIONS 10 REFERENCES
Are plate load tests still relevant in geotechnical en- ASTM D 1196—64 1987. Standard test method for non-repeti-
gineering? Absolutely! As presented in Figure 5, one tive static plate load tests of soil and flexible pavement com-
ponents, for use in evaluation and design of airport and high-
needs a good understanding of the different factors way pavements. American Society for Testing and Materials,
affecting plate load test results in order to ensure Philadelphia, USA.
sound readings, and a holistic approach should be fol- Barnard, H.F.T. 2013. The effect of bedding errors on the accu-
lowed when designing plate load tests. racy of Plate load tests. MEng Thesis. University of Pretoria.
The results from the industry research survey Barnard, H.F.T. & Heymann G. 2015. The effect of bedding er-
demonstrate that: rors on the accuracy of Plate load tests. Technical Paper. Pa-
per [Link]. Journal of the South African Institution of
- The typical plate sizes used for plate load tests con- Civil Engineering. 57(1): 67-76.
ducted in South Africa varies between 300 mm and Boussinesq, J. 1885. Application des pontentiels a laetude de
600 mm; l’equilibre et du movement de solides elastiques. Paris:
- More than 80 per cent of suppliers are using pres- Gauthier-Villard.
sure gauges to measure the applied loads; BS 5930: 1981. Code of Practice for Site Investigations (for-
merly CP 2001). British Standards Institution, London.
- Approximately 95 per cent of respondents are us- Clayton, C.R.I. Matthews, C. & Simons, N.E. 1995. Site Inves-
ing dial gauges to measure plate settlement during tigation. 2nd Edition, Chapter 9. Blackwell Science: Oxford.
loading, of these about 85 per cent are using three Timoshenko, S. & Goodier, J.N. 1951. Theory of elasticity. New
dial gauges spaced 120 degrees apart, to measure York: McGraw-Hill Book Co., Inc: 368-372.
any tilt of the plate;
- Approximately 60 per cent of suppliers are using
well-graded sand as the preferred surface prepara-
tion method to achieve a level and smooth test
area, where only 4 out of 20 suppliers prefer plas-
ter of Paris;
- All suppliers are using steel reference beams to
keep the displacement measuring devices firm on
the plate and most of the beams were recorded as
being at least 3 m in length;
- The two most preferred standard test procedures
adopted to perform plate load tests are BS 5930
(1980) and ASTM 1196/64.
349
17th ARC Conference 2019
350