Updating Bituminous Stabilized Materials Guidelines Mix Design Report Phase II

APPENDIX L
Technical Memorandum
Updating Bituminous Stabilized Materials
Guidelines: Mix Design Report, Phase II
Task 12: Laboratory Compaction
AUTHORS:
KJ Jenkins
RWC Kelfkens
1

1. INTRODUCTION
Background
The purpose of this report is to provide results and conclusion as well as any possible recommendations
with regard to an alternative method of compaction of granular material in the laboratory.
In the civil engineering practice, properties of granular materials are tested in various ways, including triaxial
testing, CBR (California Bearing Ratio) etc. Prior to these tests taking place, specimens of material
first have to be prepared. That is to say it has to be compacted into some or other form so that the tests
may be carried out. Samples are typically compacted into a cylindrical form for triaxial tests and into
block form for MMLS type tests i.e. accelerated testing. Laboratory results and experience have shown
that the results of material properties achieved in the laboratory are not as fair and consistent a
representation of the field results as the industry would like. This is believed to be as a result of
laboratory compaction methodologies currently being used.
In the laboratory, compaction procedures include methods such as Mod AASHTO compaction (the
densities achieved in the field are typical measured against the density achieved when using this
compaction method), Marshall Hammer, Gyratory compaction and Vibratory Table compaction etc.
These procedures all have their advantages and disadvantages. The factor affecting the final outcome of
the material properties on site is that on site high amplitude and frequency vibratory compaction is used
as apposed to impact compaction, which influences particle orientation, packing and other factors
Mod AASHTO compaction and Marshall Hammer compaction are both impact compaction methods. Site
compaction is done by means of vibratory compaction. This poses a problem when trying to compare
the properties of the site compacted material to the properties of the laboratory compacted material.
There are various differences in the sample structure when comparing impact and vibratory compaction:
• the arrangement of the material particles is different (i.e. a different skeleton structure),
• there may be a vary clear differences in the void contents of the two compaction methods, and
• the final densities on site may be much higher than that which is achieved in the laboratory
(e.g. site compaction may be as high as 104% of Mod AASHTO compacted density).
Gyratory compaction allows for the particles to be kneaded against each other, thus giving a different
skeletal structure and particle orientation to that of Mod AASHTO and Marshall Hammer compaction as
well as different voids content. This however still differs from the results on site and as a result also
yields different material properties from those achieved on site. According to results from research done
by the CSIR (HL Theyse, 2004), vibratory table compaction provides the best results in terms of
producing the same material properties in the laboratory as those which are obtained on site. The
2

skeleton structure and voids content are more similar than that of the other compaction methodologies
when compared to site compacted material.
As a result of the inconsistencies occurring between the field and laboratory properties of the compacted
material, it was proposed that an alternative compaction method be identified and researched, one that
would allow field compaction results to be simulated a closely as possible in the laboratory. Feedback
received during the launch of TG2 (Technical Guidelines 2) (2002) indicated that vibratory compaction
could provide a possible alternative. The research of the viability of the compaction method was to be
assigned to a student at master’s level.
Objectives
The objectives of the research were as follows:
1. Identify an appropriate refusal density compaction procedure, similar to the one currently used
internationally e.g. in the United Kingdom (UK) on asphalt material, that can possibly be adapted to
compaction of Bitumen Stabilised Materials (BSMs) in South Africa. At the same time identify the
different compaction requirements of BSMs relative to HMA.
2. Investigate refusal density compaction procedure for BSMs, i.e. bitumen emulsion and foamed
bitumen stabilised material.
3. Establish a correlation between the refusal density compaction procedure and Mod AASHTO in order
to provide a reliable link to field densities.
4. To enable the compaction of a specimen of 150mm diameter and 300mm height for use in triaxialtype
testing. Currently specimens of this diameter can only be produced to a height of 125mm.
5. Establish a compaction protocol for the refusal density compaction method
Layout of the report
The report begins with an executive summary of the research, followed by Section 1, the
Introduction. In the introduction, background to the research is given as well as providing the
objectives and giving a brief layout of the structure of the report. Section 2 follows which is the
Literature Study. Here a summary of various literatures that was studied for background on
materials and compaction methods is made. This is followed by Section 3, the Methodology.
This chapter explains how the various objectives will be achieved. Section 4 then provides the
results, which include statistical results, of the experiments performed and provides for the
interpretation of these results. Conclusion and Recommendations are then made in Sections 5
and 6 respectively, with Section 7 providing all references.
3

2. LITERATURE STUDY
This Section of the report will look at various pieces of literature regarding compaction.
Literature on foamed and emulsion mixes will also be addressed and the respective properties
of both bitumen emulsion and foamed bitumen i.e. the material properties of the bitumen
emulsion and foamed bitumen as well as the effect these two cold mixes have on the material
properties of granular materials.
Emulsion
There are three main types of emulsion (US Patent, 2002):
1. Cationic emulsions: surfactants are constituted by polar molecules; formula RNH3 + X - . R
is the hydrocarbon chain which constitutes the lipophilic portion of the molecule. NH3 + is
the hydrophilic portion and X - is any anion from a strong acid.
2. Anionic emulsions: the general formula is R--Y-- C + . R is the hydrocarbon chain which
constitutes the lipophilic portion. The hydrophilic chain is given by Y - C + . Y- is a
carboxylic, sulphonic, sulphuric, phsphonic or phosphoric group. C+ is a metal cat-ion,
often alkaline or ammonium.
3. Non-ionic emulsions: In this type of emulsion, surfactants are constituted by the R--
(EO) n –OH molecule type. R is again the hydrocarbon chain which constitutes the
lipophilic portion of the molecule. The hydrophilic portion is constituted by the (EO) n –
OH radical; EO represents Ethylene Oxide.
Typically either an anionic or cationic bitumen emulsion is used.
Emulsion mixes are used in the base layer of the pavement structure as well as for a surfacing
course (Miller Group et al, 2004). The application rates of the emulsion mix vary from 50mm to
200mm.
Binder within the BSM-Emulsion coats the smaller particles selectively and the stiffness of the
emulsion mix is less than that of the hot mix asphalt. For BSM-Emulsion voids range between
12% and 15% where in the open grade emulsion mixes it varies between 20% and 30%. The
surface of an emulsion mix is also relatively fragile when it is compared to the surface of Hot
mix Asphalt. Waterproofing is not provided for in unsealed emulsion mixes and the surface
cohesion may also not be sufficient to withstand tangent stresses. The sealing of the surface of
4

unsealed emulsion mixes provides waterproofing and the inherent performance of the emulsion
mix is not compromised in terms of premature stripping, oxidation and ravelling.
Foam
Mr. KM Muthen of the CSIR shows the following advantages of foam mixes in the Contract
Report CR-98/077 (1998).
Foam binder increases the shear strength properties of the treated material; it also reduces the
moisture susceptibility of granular materials. In foamed bitumen stabilized material (BSM-foam)
the strength properties approach those of cemented material, the BSM-foam is however flexible
and fatigue resistant.
When compared to other cold mix processes, foam treatment material can be used with a wider
range of aggregate types. There is a saving in time, because foamed mixes may be compacted
immediately and may carry traffic almost directly after compaction. Energy is also conserved
because the aggregates remain cold while only the bitumen is heated. Curing of foam mixes
does not result in the release of volatiles; therefore environmental side-effects are avoided.
Usual time constraints for achieving compaction, shaping and finishing of the layers are also
avoided. This is because foam mixes remain workable for extended periods of time. Foamed
mixes may also be stockpiled without the risk of binder runoff and leeching. The last advantage
listed is that the foam mixes may be constructed in adverse weather conditions such as cold
weather or light rain, this is because the workability or quality of the finished layer is not really
affected.
5

Compaction of Emulsion and Foamed BSMs
Some of the early investigations into the compactability of BSMs were carried out at
Stellenbosch University by Weston (1998), where he investigated Marshall, Mod AASHTO,
Gyratory and roller compaction of foamed mixes.
The CSIR also performed compaction experimentation on both BSM-foam and BSM-emulsion to
determine the compactability of the types of mixes (Theyse, 2004). They used three methods of
compaction i.e. Mod AASHTO, Gyratory and vibratory compaction. They found that there
appears to be some logic governing the compaction of material but the rules seem to change
from situation to situation. Factors such as the nature of the aggregate (grading and Atterberg
Limits) compaction method, the type of bituminous binder (foam or emulsion) as well as the
filler types used and filler contents all seem to have an effect. These factors, state the CSIR,
make it very difficult to formulate a set of consistent guidelines that will ensure that the most
appropriate compaction equipment, binder type, filler type and binder and filler content levels
are achieved and are used.
The CSIR also found that the difference between the grading and Atterberg Limits of the
materials used in the investigation resulted in different optimal compaction methods for the
different materials. One of the materials (Crushed hornfels) had a continuous grading and a low
PI, this material was more conducive to vibratory table compaction. The other material, the
decomposed granite, had a more uniform grading and it was more conducive to gyratory
compaction. It is postulated that it may be because the crushed hornfels may have insufficient
fines to form a paste in which the larger particles are suspended, orientated and moulded
during compaction.
During the CSIR’s investigation some observations were made. They are as follows:
Compaction of Crushed hornfels using Vibratory Table compaction
• When low to intermediate emulsion contents (below 1.5 %) were used without filler, there
was a strong positive effect on the compaction of the crushed hornfels.
• Where cement was used in combination with emulsion, there was a negative effect on the
compaction of the crushed hornfels.
• When cement was used in combination with foamed bitumen, there was no effect on the
compaction of the material.
Compaction of crushed Hornfels using Mod AASHTO compaction
• The use of cement and fly-ash alone had a negative effect on the compaction of the material.
Significant minimums occurred in the filled volume and volume of solids in the samples
containing either 1 percentage fly ash or 1 percent cement.
6

• Where the binder content was increased, there was a negative effect for both BSM-foam and
BSM-emulsion. This negative effect was on the volume of solids; here significant reductions
took place at the highest binder contents for both BSM-emulsion and BSM-foam.
• When filler was used in combination with foam, there was a reduction in the negative effect
of increasing the binder content mentioned above. The effect of adding filler to the BSMemulsion
was found to be insignificant, only the individual effects of the binder and filler
were reflected in the case of the BSM-emulsion.
Compaction of Gauteng granite (decomposed Granite) using Vibratory Table compaction
• When the binder contents were increased, using both foam and emulsion, there was a slight
negative effect on the compaction in terms of the volume of solids.
Compaction of Gauteng granite (decomposed Granite) using Mod AASHTO compaction
• In the case of both cement and fly-ash, there was no discernable influence on the
compaction;
• Emulsion and foam both had the same negative impact in terms of a decrease in volume of
solids with an increase in binder content.
When looking at the vibratory compaction of the crushed Hornfels, the CSIR noted that the
intermediate percentage of emulsified bitumen acted as a compaction lubricant, this was in
terms of both volume filled and volume of solids results.
This benefit of the lubricant was not found when using Mod AASHTO or Gyratory compaction.
The vibratory type of compaction is, however, the preferred compaction method for this type of
material; this in both the laboratory and in the field compaction. Emulsified bitumen at low to
intermediate binder content levels (

The CSIR stated that there may be other considerations in favour of adding bituminous binder.
Such considerations as to the improvement of the workability of an old crushed stone base
layer that is being recycled, the retention of the fines in the layer in the long-term, and/or
improving the water resistance of the material may be looked at when adding bituminous
binder.
The CSIR also noted that there was a definite benefit in using foam or emulsion in combination
with cement. They state that the UCS and ITS requirements of the TG2 document could be
achieved using these bituminous binders in combination with cement.
From the validation phase of their research the effect of aggregate grading on the compaction
of the material was again confirmed. They indicate that it is not only the deviation of the
grading from the maximum density grading curve that determines the level of density that is
achieved but also the grading of the material. The grading determines the preference of the
material in terms of the type of compaction that will result in the highest possible density being
achieved
Comparison of different compaction methods using X-ray Computer Tomography
(CT)
In report No. 113/12 (EMPA No. FE 840544), March 2002 EMPA performed research to compare
the difference in compaction methods using X-ray Computer Tomography (CT). The change
that takes place as well as the difference in homogeneity & isotropy in asphalt concrete was
examined, using the same mix but with different compaction methods.
Three methods of compaction were used during this research: a) Marshall compaction, b)
Gyratory compaction and c) LCPC Rolling wheel compaction. The investigation was carried out
using the standard Air Void content determination (AV) and X-ray CT.
Under the gyratory compaction it was found that the material loses heat in the centre (2 0 C) and
where it is in contact with the outside walls of the mould (10 0 C). Although heat is lost in the
centre of the specimen it was found that after the compaction the temperature in the centre
was 5 0 C higher than the temperature of the material near the top or bottom of the specimen.
The results of the gyratory compaction showed that the material had a tendency to flow radially
towards the side of the specimen. Under Marshall Compaction the pins used in the X-ray CT test
did not move significantly in the horizontal plane, however, a few of the pins showed a slight
tendency to move away radially, indicating that the material, during compaction is squeezed to
the side. The air void content under the gyratory compaction was in some cases found to be
8

negative where the Marshall compaction gives a rather even air void distribution in the vertical
direction, the bottom core pieces however have a slightly higher void content. EMPA noted the
after 20 blows the difference in air void content using Marshall Compaction almost vanishes.
The results from the three compaction methods showed that none of the compaction processes
was able to produce homogeneous asphalt concrete specimens. Structurally different specimens
are produced; therefore, state EMPA, there is no reason to expect these methods to produce
specimens comparable to real life or to be used on alternate bases.
Vibratory Compaction
In the Example Paper: One-Point Vibrating Hammer Compaction Test for Granular Soils, Adam
B. Prochaska and Vincent P. Drnevich (2005) show that there is great promise for the use of
vibratory hammer compaction for the preparation of samples in the laboratory. They state that
a One-Point Vibrating Hammer test on an oven dried sample will provide the maximum dry unit
weight of the material and moisture content range for effective field compaction of granular
soils. The results from this procedure were found to produce consistent and reproducible
results. The test method is also applicable to a broader range of material than current vibratory
table tests. The test results from the compaction of sandy soils indicated that the values for the
dry unit weight in the densest condition are comparable to that obtained from the vibratory
table tests.
Mr. Thorsten Frobel of Fulton Hogan Ltd. in New Zealand provided information on the vibratory
hammer compaction method used in New Zealand. In New Zealand the Mod AASHTO
compaction procedure has been replaced by the Vibratory Hammer compaction test. The reason
for this is that New Zealand has fairly soft aggregates, and the heavy dynamic compaction of
the Mod AASHTO compaction method causes a change in the grading and this intern influences
the Maximum Dry Density (MDD) and Optimum Moisture Content (OMC) significantly. The
Vibratory Hammer compaction is used to specify the MDD and this is used to specify the target
Dry Density for site. In the case of unbound granular materials no adjustment needs to be
made to the target Dry Density, but in the case of stabilized material a correction needs to be
made, therefore a sample is usually taken and compacted, either in the lab or on site at that
moisture content so as to get an idea of the shift; the unbound pavement layer specification of
New Zealand does however call for plateau testing. The New Zealand Specifications (TNZ,
2005) provide target compaction levels for two site types:
1. Greenfield sites
The target Dry Density is quoted as: “The Maximum Dry Density (MDD) for
construction shall be the higher of the maximum laboratory dry density at
optimum water content (OWC) and the plateau density at optimum water
content (OWC).”
9

Note that in New Zealand the term OWC is the same as the term OMC used in
South Africa.
2. Overlay sites
The target Dry Density is quoted as: “The Maximum Dry Density (MDD) for
construction shall be the maximum laboratory dry density at optimum water
content (OWC).”
These compaction levels are expressed as percentage of the MDD achieved using the vibrating
hammer, these levels are provided in table L. Lit 1 (TNZ, 2005).
Table L. Lit 1: Mean an Minimum Level of Compaction of Pavement Layers as %MDD
of Vibrating Hammer
Values Sub-basecourse Pavement
Layers
Basecourse Pavement
Layer
Mean Value ≥ 95% ≥ 98%
Minimum Value ≥ 92% ≥ 95%
These compaction levels in table L. Lit1 are checked by testing five (5) randomly selected areas
on site with a frequency of one (1) MDD per 5000m 3 of material laid. Should the tested areas
conform to the criterion in table L. Lit 1 the compaction levels are accepted (TNZ, 2005).
The samples prepared in New Zealand are also used for UCS testing; this is in the case of
stabilized material samples. The procedure followed by the New Zealanders is briefly outlined
below; this is taken from NZS 4402: 1986 Test 4.1.3.
• Hammer Specifications
o Frequency = 25 to 60 blows per minute
o Rating = 60 to 1200 Watt power consumption
o Mass of Loading Frame + hammer and Downward Force = 300N ± 50N.
The Hammer may also be operated manually by experienced personnel
provided the hammer is held in an upright position and that the total
downward force is also in the order of 300N ± 50N. For inexperienced
personnel the hammer may be placed on a scale and a downward force be
applied till the scale reads 30 or 40kg; this is done prior to compaction while
the machine is switched off.
o Foot Piece Diameter = 145mm
• Mould Specifications
10

o
o
o
Non-corrodible cylindrical metal mould = 152mm ± 0.5mm inside Diameter
A Metal Spacer = 150mm ± 0.5mm diameter. This is placed inside the
mould prior to compaction.
Final specimen height = 125 to 127mm high.
• Compaction Procedure
o Mass of Material used for a sample = 5.5 kg
o Assess the moisture contents required for the compaction. Adjust these
content so that there are different moisture contents across the samples
which span the OMC with in the required range.
o Compaction time = 180sec ± 10sec per layer.
o Number of layers compacted per sample = 2 layers
The procedure described above is the procedure used to determine the Dry Density vs.
Moisture content curve and the graphic image of the mounted hammer is shown on the
following page in figure L.A: Mounting of vibrating hammer for the New Zealand vibrating
hammer compaction procedure. There is also a procedure developed to determine whether
or not the hammer which is being used is adequate. This is as follows:
• A 10kg sample of Leighton Buzzard silica sand is taken, of which at least 75%
passes the 600μm test sieve. The coarse fraction is discarded. Sufficient water is
mixed with the sand finer than 600μm to raise the moisture content to 25% ± 5%.
The material is compacted according to the procedure described above, excluding
varying moisture contents, for a total of three samples, the mean Dry Density is
then determined. If the mean Dry Density of the sand exceeds 1.74 t/m 3 the
hammer may be considered suitable for the compaction procedure.
11

Figure L. Lit 1: Mounting of vibrating hammer for the New
Zealand vibrating hammer compaction procedure
12

In 2007 ASTM published a standard test method for the vibratory hammer, this is found in
ASTM D7328-07. The procedure followed differs from the procedure developed and followed by
the New Zealanders. The procedure described in ASTM D7328-07 has two methods, the first
(method A) uses a mould of 152.4mm diameter and the second procedure (method B) uses a
mould of 279.4mm diameter. For purposes of this research method B of this procedure may be
ignored. The procedure described in method A is as follows:
• Hammer Specifications
o Frequency = 3200 to 3500 beats per minute
o Impact energy (manufacturers rating) = 9.5 to 12 Joule
o Weight of hammer = 53 to 89N excluding the weight of the tamper. A list of
potential hammers and their characteristics are provided by the ASTM and is
shown in table L. Lit 1 below:
Table L. Lit 2: Potential vibratory hammers for ASTM vibratory hammer compaction
procedure
Bosch
11248EVS
Bosch
11318EVS
Milwaukee
5327-21
Milwaukee
5336-22
Volts 120 120 120 120
Amps 11 11 11 13
Beats/min 1700-3300 1300-3300 3400 1300-3450
Hz 28-55 22-55 57 22-58
Impact Energy (J) 10 12 11 12
Length (cm) 46 45 44 47
Weight (N) 14.4 12.5 12.9 15
13

• Hammer Frame Specifications
o The frame shall have a metal clamp assembly to firmly hold the vibrating
hammer that moves on guide rods that allows for free vertical movement.
The guide rods are fastened to a metal base so as to keep them vertical and
parallel.
o The frame is designed to securely hold the vibrating hammer and clamp
assembly in a vertical position during the removal and insertion of the
mould.
o The total applied surcharge of the clamp assembly + the vibrating hammer
and tamper shall be 19.3kPa ± 0.7 kPa.
The figures below were taken from ASTM D 7382 – 07 and the dimensions
shown are in inches.
Pins and/or clamps will be
needed to secure the
clamp assembly, vibrating
hammer, and tamper
above the mould to allow
inserting the mould,
adding soil to the mould,
and removing the mould.
Metal clamp Assembly
Figure L. Lit 2: Mounting frame for ASTM vibrating hammer compaction
14

Figure L. Lit 3: Metal clamp assembly to firmly hold the vibrating hammer
• Mould Specifications
o Cylindrical mould made of rigid metal
o Average inside diameter = 152.4mm ± 0.7mm.
o Height = 116.4mm ± 0.5mm.
o Volume = 2124 ± 25cm 3 .
• Compaction Procedure
o Material passing the 19mm sieve shall be used for the preparation of
samples.
o Samples are compacted in three layers.
o A compaction time of 60sec ± 5sec per layer is used to compact each layer.
The ASTM also has a procedure in order to check the suitability of a vibratory hammer for the compaction
process. This procedure is in some ways similar to the New Zealand procedure for checking the suitability of
a vibratory hammer. The procedure is as follows:
• Standard sand shall be tested and is to conform to the requirements for 20-30 sand. These
specifications are found in the ASTM specifications C778. Before the test is performed the
material should be stored in such a way that freezing and/or contamination does not occur,
if the material was previously used it should not be re-used. A required dry specimen mass
of 7kg is required and must have a moist mass of at least 9kg. A representative sample
meeting this specification is selected using a riffler or splitter or any such method quartering
included. The vibratory hammer and mould (152mm diameter mould) are then prepared.
The sand is then compacted according to method A described above. After compaction the
Dry Density is calculated and should the sample meet or exceed a dry density of 1.76
ton/m 3 (17.29 kN/m 3 ) then the vibratory hammer may accepted as having sufficient energy.
15

Critical comments on the Literature
In terms of the purpose of this research the procedure used in New Zealand is not an adequate procedure
from which to work. The compaction time of 180 seconds per layer is long, although the samples may be
reaching very high Dry Densities the coarse aggregate materials used in New Zealand are of a soft quality, it
is as a result of this that New Zealand has moved away from the Mod AASHTO compaction method and
adopted the vibrating hammer compaction method.
The compaction method used by New Zealand does serve well in that it shows that the vibratory hammer
compaction may be used to specify field densities as well as moisture levels in the field.
The ASTM method for vibratory hammer compaction appears to be a newly developed method as it was only
published toward the end of 2007. This procedure serves well to show what level of compaction time per
layer may be necessary when preparing samples. It also helps identify what the total mass (i.e. all
components of the set up; hammer, foot piece, mounting head etc.) of the set up should be. The ASTM
methods identify a total mass of ±34kg; the New Zealand method identifies a total mass of 30-40kg. Both
these masses are similar to the procedure followed in the British Standards (this method is discussed in sub
section 3.1.2 of the methodology section), which shows a mass also in the order of 35kg ± 5kg. This is a
good indication of what the order of the total mass of the vibratory hammer set up should be for our
research. The total masses of New Zealand and the United Kingdom compaction procedure include the force
applied by an operator, which shows what the mass of an applied dead load should be in order to achieve a
total mass of ± 30kg, this is beneficial as less physical labour is needed by the operator to compact the
material if a surcharge load is applied.
From the literature on the ASTM vibratory hammer compaction method and the New Zealand compaction
method a table analyzing the compaction energies was set up.
Table L.Lit 3: Comparison of Total Compaction Energy –
New Zealand vs. ASTM (Bosch 11248EVS)
Vibratory Hammer Lower Limit (kJ) Upper Limit (kJ)
New Zealand 63 151.2
Bosch 11248EVS® 51 99
Table L.Lit 4: Comparison of Average Compaction Energy per Layer –
New Zealand vs. ASTM (Bosch 11248EVS)
Vibratory Hammer Lower Limit (kJ) Upper Limit (kJ)
New Zealand 31.5 75.6
Bosch 11248EVS® 17 33
16

The New Zealand specifications do not clearly state what the order of the impact energy should be, however
by noting that the frequency of 25 to 60 blows per second (1500 to 3600 blows per minute) is similar to the
ASTM standard of 3200 to 3500 blows per minute, which provides an impact energy of 9.5 to 12 Joule, a
rough estimate of 7 Joule impact energy was made. The upper limit and lower limit indicate the range in
which the energy should be. The ASTM method does show a lower impact energy both in terms of the total
and per layer energy used, but this is due to the fact that compaction time is only 60 seconds, i.e. 1/3 of the
New Zealand procedure. This shows that a higher point energy (energy per blow) may require less total
energy to compact the sample, i.e. for a higher point energy less compaction time per layer is need to
compact a sample and hence less total energy.
17

3. METHODOLOGY
3.1 COMPACTION
This chapter provides a description of the methodologies used and followed in order to perform
the required experimentation so as to determine the viability of the vibratory hammer as a
means of compacting Granular materials in the laboratory. For this methodology either a clean
untreated G1 or G2 material could be used to perform the experiments, both these materials
are a good quality granular material used in the base course of pavements. Material would also
be collect form site and this site material would be used to establish a correlation between the
compaction of material in the laboratory and the compaction on site. A G2 material was finally
taken; the fact that fine material from a source other than the parent rock could be added to a
G2 quality material seemed more adequate because the material being taken from the site
would have material present that was not from the original material itself. Two and three tests
would be performed for a single experiment and the experiments were measured against the
Mod AASHTO compaction method. Three experiments were typically done as this accounted for
the variability of the results better than what two experiments would have. Two mix conditions
were studied i.e. BSM-emulsion and BSM-foam. Under both mix conditions the moisture content
of the material, the surcharge load on the vibratory hammer and the temperature of the
material were the conditions under which the compaction experimentation would take place.
These conditions were chosen as they are all factors which affect the compaction of granular
materials. Under the conditions of moisture variation and surcharge load variation experiments
were performed using time as measure to obtain both density equal to the Mod AASHTO
density of the material at specific moisture content (100% Mod AASHTO) and the material’s
refusal density. From these results a fixed moisture content and fixed dead load were decided
on and the time to 100%Mod AASHTO was taken and experimentation on the material under
varying temperature conditions was carried out as well as the correlation experiments. The
results obtained from the experimentation were then compared to the results of vibratory table
compaction, excluding the correlation experiment results. A full flow chart of the experimental
design is provided for under the results section of this report.
3.1.1 Mod AASHTO
Mod AASHTO compaction is the compaction method against which the vibratory
hammer will be referenced. That is to say all results will be measured relative to
Mod AASHTO densities.
18

3.1.2 Vibratory Hammer
3.1.2.1 Existing Procedures
United Kingdom (UK)
In the UK a compaction procedure for the vibratory hammer has already
been developed, this can be found in BSEN12697-32-2003. The procedure
is as follows:
1. The vibrating hammer is fitted with a circular steel tamper of 146
mm diameter
2. Each layer is compacted for a period of 60 seconds ± 2s.
3. During compaction a firm downward force is applied so that the
resulting force (which includes the mass of the hammer is 350N ±
50N.
4. It is recommended that for inexperienced persons, that the hammer
is placed on a scale and a downward force be applied till a reading of
35kg ± 5kg be achieved. This gives an indication of how force is
needed to be applied by the person during compaction
Comment: The tolerance of 50kN allowed for the applied downward force is
very lenient. The result is that there is a large variability of the achieved
densities.
Delft University of Technology (TU Delft)
TU Delft in the Netherlands also developed a compaction procedure regarding
the vibratory hammer, refer PhD. Student Patrick Muraya. The procedure
followed is long; therefore only a basic description of the procedure followed
at TU Delft is provided here:
1. The exact mass of material of each layer is weighed off.
2. The thickness to which the layer will be compacted is also
determined.
3. The mass of material is then poured into the mould, not spilling any
material.
4. The compacting unit is then set up accordingly. The height adjusting
rings are loosened and the compacting bar is lowered till it touches
the material. The material is first compacted by hand (8 blows). The
19

height adjusting rings are then fastened X mm down from the nylon
rings – X being the calculated thickness of the layer.
5. The hammer is then turned on and a small amount of pressure is
applied till the nylon rings meet the height adjusting rings - i.e. the
height of the layer is achieved.
6. The surface is then roughened (scarified) up using a bar with a
rounded head and the next layer is added and compacted in the same
way. This is done until compaction of the sample is completed.
Figure 1 below shows a visual of the compaction set up at TU Delft.
Vibratory
Hammer
Nylon Ring
Compacting
Bar
Adjusting Ring
Figure L.1: Compaction set up at TU Delft
20

3.1.2.2 Kango 637 ®
Figure L.2A: Kango 637®_Vibratory Hammer
The Kango 637 ® Vibratory Hammer was the first Vibratory Hammer that was tested in South
Africa. Its technical specifications are provided in the table below:
Table L.1: Technical details of Kango 637®
Hammer
Rated power
Impact
Impact rate at
Frequency
Weight
input
energy
rated speed
Kango 637 750 W uncertain 2750 1/min 45.83 Hz 7.5 kg
21

3.1.2.3 Bosch GSH 11E ®
During the course of experimentation, the Kango 637 ® experienced
technical difficulties and as a result became unusable. The replacement part
for the hammer could not be located in South Africa and hence a new
hammer was purchased, the Bosch GSH 11E. Figure 2 below shows the
Bosch GSH 11E
Figure L.2B: Bosch GSH 11E ® Vibratory Hammer
Replacement of the Kango Hammer 637 ® during the research became
necessary due to the fact that it is no longer supported with parts and back
up service in South Africa; therefore adjustments had to be made for the
new implement i.e. the Bosch Hammer.
A comparison of the Kango 637 ® and Bosch GSH 11E ® technical
specifications are provided below:
Table L.2: Comparison between Bosch GSH 11E® hammer and Kango 637®
Hammer
Rated power
input
Impact
energy
Impact rate
at rated speed
Frequency Weight
Kango
637
Bosch
GSH 11E
750 W uncertain 2750 1/min 45.83 Hz 7.5 kg
1500 W 6-25 J 900 – 1890
1/min
15 – 31.5
Hz
10.1 kg
22

3.1.3 Design of the mounting frame for the vibratory hammer
3.1.3.1 Kango 637 ®
With the aid of technical/mechanical support at the University of
Stellenbosch i.e. Mr Johan Muller, a mounting system for the Kango 637®
was developed. Below is given a schematic of the frame as well as figures
of the constructed frame
Figure L.3: Top view of mounting frame for Kango Hammer ®
23

Figure L.4: Front view of the mounting frame for Kango Hammer ®
Figure L.5: Front view of mounting frame for Kango Hammer ®
24

Figure L.6: Left view of
Figure L.7: Rear view of mounting
mounting frame frame for Kango Hammer ®
Figure L.8: Full view of the frame for Kango Hammer ®
25

3.1.3.2 Bosch GSH 11E ®
The mounting frame designed for the Bosch GSH 11E was effectively a
modification of the existing Kango frame. The mounting head which is
attached to the hammer was modified. This modification design was
developed and executed by the workshop at the University of Stellenbosch
Civil Engineering Department. This modification was necessary due to the
fact that the size and shape of the Bosch GSH 11E® differs from that of the
Kango 637®. The design appears a follows:
Positioning of dead
weight
Rubber fitting
127
Sleeve
113
Sleeve
112
870
64
50
Figure L.9: Schematic of the mounting head of the Bosch GSH 11E ®
In both the original design and the modified deign, rubber was placed between the areas
where the steel plates of the frame come into contact with the vibratory hammer. Due to the
vibratory effect, the steel is constantly vibrating against the hammer; this could easily result
in damages to the hammer. The rubber insertions protect the hammer from these damages.
26

Figure L.10: Front view of Bosch
mounting
Figure L.11: Rear view of Bosch
mounting
Figure L.12: Left view of Bosch mounting head
Figure L.13: Full view of
mounting head
3.1.3.3 Modifications to mounting system
Further modifications were later made to the rods and base. A frame was
designed by Mr. Dion Viljoen of Stellenbosch University Civil Engineering
workshop in order to better stabilize the entire system. A pulley system was
also fixed to the mounting head and frame to minimize the labour intensity
of the compaction operation.
During experimentation, questions regarding the mounting system arose.
Specifically, how accurately are measurements being taken and how
perpendicular is the footplate to the material mould when compacting. In
27

addition, after one layer of the specimen had been compacted, the hammer
would be physically raised and removed from the rods. When the next layer
was ready to be compacted, the hammer was then placed on the steel rods
and physically lowered into position. This raises an issue concerning the
amount of physical effort and time needed to raise and lower the hammer.
The modifications stated in the first paragraph of this subsection (3.1.3.3)
addressed these questions.
The pulley system allows the vibratory hammer to be raised with less
physical effort and allows it to be suspended once raised, so as to allow
preparation of the next layer. This results in the hammer never being
removed from the vertical shafts, whilst material for an additional layer is
added. The hammer may then be lowered onto the material and
compaction may commence. Furthermore, the stabilized frame results in
less “wobble effect” of the mounting head thus more consistent and
accurate readings may be taken and the hammer is more perpendicular
during compaction.
Figure L.14: Bottom of pulley
Figure L.15: Top of Pulley
Stability frame
Position where guides
are fastened to the
Figure L.16: Left view of Stabilizing Frame
Figure L.17: Front view of Stabilizing Frame
28

Mounted guide
Figure L.18: Pulley system Figure L.19: Wooden base and mounted Figure L.20:
guide
Suspended Bosch
Vibratory hammer
3.1.4 Vibratory table
For both BSM-foam and BSM-emulsion an experiment using the vibratory table was
to be done. The standard procedure for the vibratory table compaction method is
found in the TMH 1: Revised Addition (1990). The specifications of the TMH 1
vibratory table compaction method are amplitude of 0.5mm, a frequency of 50Hz, a
dead load surcharge of 50kg and a compaction time of 120sec (2min); for the
purposes of this research compaction was done until the layer being compacted
reached a layer thickness of 60mm. Measuring from the foot piece up, markings
were made at intervals of 60mm and were labelled layer 1 through to layer 5, these
markings are shown in figure L.21B. While the sample was compacted the operator
watched to see at what point in time the marking, for the layer being compacted,
become level with the top of the mould (not the extension piece). A steel rule was
then placed across the top of the mould with the surcharge still inside the mould to
verify that the marking for the respective layer was in fact level with the top of the
mould (image on the right in figure L.21B); this was done for layer 1 through 4. An
extension piece was fitted to the mould prior to compacting layer 5 but no marking
was made for layer 5 on the surcharge; this was because the operator physically
held the surcharge in place and could see when the footing of the surcharge
reached the top of the mould.
The compaction time to 100% Mod AASHTO density of the vibratory hammer was to
be measured against the compaction time to 100% Mod AASHTO density using the
vibratory table.
29

Three (3) samples were prepared for each mix (BSM-emulsion and BSM-foam) and
again the moisture content was decided on and the target dry density was obtained
from that; using the Mod AASHTO curves initially set up for each mix. The samples
were then compacted till a layer thickness of 60mm was achieved and the time
noted. The results from the experiment were then compared to the results of the
vibratory hammer.
50 kg dead load
300mm x 150mm
mould
Vibratory table
Figure L.21A: Vibratory table set up
30

50kg Surcharge
The left image shows
how and where the
markings were made
on the surcharge. The
image on the right
shows how the height
300 x 150 mm
Steel Mould
of a layer is checked
(in this case layer 2)
Layer 1
Layer 2
Layer 3
Layer 4
Foot
Piece
60mm
Steel Rule
Layer 2
Layer 1
after the compaction
of that layer was
completed i.e. a layer
thickness of 60mm
was achieved.
Figure L.21B: 60mm intervals marked on the 50kg Surcharge
Frequency
dial
On/Off
Neck extension
Figure L.22: Vibratory table
the final layer
during compaction of
31

3.2 Experimentation
3.2.1 Material Type and Properties
3.2.1.1 Material Type
The material type chosen to be used for experimentation purposes was a
G2 material. This was acquired from Lafarge at their Tygerberg Quarry.
For purposes of correlation to site compaction, material from a recycling
project taking place along the N7 (between Cape Town and Malmsbury)
was used. This material was acquired in two states.
1. Untreated milled material.
2. BSM-emulsion: milled material.
The N7 material is also a G2 material but due to the milling process RAP
(Recycled Asphalt Pavement) was present in the material make up.
After completion of the G2 and N7 experimentation, a G5 material was used
to perform repeatability experiments so as to establish whether or not the
compaction procedure developed from the G2 material would be compatible
with other granular materials, the G5 material was acquired from Lafarge at
their Eesrte River Quarry.
3.2.1.2 Material Properties
The following material properties for the G2, G5 and N7 Material were
obtained.
• Mod AASHTO curve of untreated material
Table L.3: Technical Information of untreated G2 Material
G2 Material: Untreated
OMC
Max Dry Density
6.15% 2260 kg/m 3
32

Table L.4: Technical Information of untreated G5 Material
G5 Material: Untreated
OMC Max Dry Density
6.7% 2228 kg/m 3
Table L.5: Technical Information of untreated N7 Material
N7 Material: Untreated
OMC
Max Dry Density
5.12% 2138 kg/m 3
• Mod AASHTO curve of Bitumen Stabilized Material (BSM)
Table L.6: Technical Information of BSM-emulsion G2 Material
G2 Material: BSM-emulsion
OMC
Max Dry Density
4.0% 2188 kg/m 3
Table L.7: Technical Information of BSM-emulsion G5 Material
G5 Material: BSM-emulsion
OMC Max Dry Density
6.8% 2217 kg/m 3
Table L.8: Technical Information of BSM-emulsion N7 Material
N7 Material: Material Treated with BSMemulsion
on Site
OMC
Max Dry Density
5.6% 2130 kg/m 3
33

Table L.9: Technical Information of BSM-foam G2 Material
G2 Material: BSM-foam
OMC
Max Dry Density
4.8% 2132 kg/m 3
Table L.10: Technical Information of BSM-foam G5 Material
G5 Material: BSM-foam
OMC Max Dry Density
6.95% 2149.5 kg/m 3
Atterberg limits of the G2, G5 and N7 material
Table L.11: Atterberg Limits of Granular Materials used for experimentation
Atterberg Limit G2 G5 N7
Linear Shrinkage Non Plastic 3.3% 2%
Liquid Limit Non Plastic 24% 20%
Plastic Limit Non Plastic 20.7% 16.3%
Plasticity Index Non Plastic 3.3% 3.7%
Grading curves: the first haul of G2 material was brought in bags, where
the second haul was a large stock pile of material. Therefore to check the
consistency of the grading 2 bags were selected and a grading done on
them and a grading was done on the stockpile material. The grading
between the selected bags and stockpile varied very little. The grading
curve for the N7 material was obtained from MSc. Student Mr. Percy Moloto.
34

Grading curve: G2 material
120.0
100.0
Sample 1
Sample 1 Stock Pile
Random Bag
80.0
% Passing
60.0
40.0
20.0
0.0
0.01 0.1 1 10 100
Seive Sizes (mm)
Figure L.23: Grading curve G2 material
N7 Graded Crushed Stone: Wet Grading Curve
120.0
UPPER LIMIT (Unsuitable: Too Fine)
LOWER LIMIT (Unsuitable: Too Coarse)
IDEAL (Suitable)
N7 GRADED CRUSHED STONE
% Passing
100.0
80.0
60.0
40.0
Upper Limit: Too Fine
Ideal: Suitable
20.0
Lower Limit: Too Coarse
0.0
0.0 0.1 1.0 10.0 100.0
Sieve Size (mm)
Figure L.24: N7 Grading Curve
35

From the grading curves, material from the G2 stock pile and N7 were sieved into fractions
and reconstituted. The N7 had a different grading to the G2 material and after
reconstitution the differences could be seen.
G2 Stock Pile & Quarry Grading Curve vs. N7 Material Grading Curve
120.0
100.0
G2 Stock Pile (US)
N7 Material
G2 Quarry Grading
80.0
% Passing
60.0
40.0
20.0
0.0
0.01 0.1 1 10 100
Sieve Size (mm)
Figure L.25: Comparison of G2 Stock Pile & Quarry Grading to N7 Material Grading
Figure L.26: Visual N7 Material
Figure L.27: Visual N7 Material < 13.2mm
≥13.2mm
36

Figure L.28: Visual G2 Material
Figure L.29: Visual G2 Material < 13.2mm
≥13.2mm
Similar to the way in which the G2 and N7 material samples were prepared so to were the G5
materail samples prepared from a grading curve, the curve used to reconstitute the G5 sieved
fractions into a sample is provided below.
100.0
90.0
80.0
Grading Curve G5 Material: Quarry Grading vs. US Adjusted Grading
Curve
Quarry Grading of G5 Material
US Adjusted Grading
% Passing
70.0
60.0
50.0
40.0
30.0
20.0
10.0
0.0
0.01 0.1 1 10 100
Sieve Size (mm)
Figure L.30: Grading Curve G5 Material: Quarry Grading vs. US Adjusted Grading Curve
Lafarge quarry provided a grading curve (the blue line in figure L.30), there curve however
included fractions of material retained on the 26mm sieve. For purposes carried out during
experimentation of the G5 material the grading curve was adjusted to include only material
retained on the 19mm sieve and smaller sieves (the pink line); the curve was adjusted using
a dry grading.
37

3.2.2 Technical Aspects of Experiments
3.2.2.1 Compaction procedure for experimentation performed
Based on the two existing methods stated in sub section 3.1.2.1 a design
procedure for the compaction experiments was developed.
1. A Mod AASHTO curve for the given material would be done. Form this
curve the OMC value of the material was determined.
2. A BSM-emulsion Mod AASHTO curve and a BSM-foam Mod AASHTO curve
were then done. From the GEMS manual it was found that for the
moisture contents of BSM-emulsion a target moisture content of 60%
OMC was used. This is due to the fact that the bitumen acts as a lubricant
during compaction. So the final fluids content during compaction is then
60%OMC added to the % bitumen binder in the BSM-emulsion.
For this Mod AASHTO curve the following moisture contents were used for
BSM-emulsion:
• 60%, 70%, 80%, 90% and 100% OMC
For the BSM-foam Mod AASHTO curve the same moisture contents
were used.
3. The targeted moisture content during compaction would then be decided
on, e.g. compact at 80% OMC. From this, referring to the Mod AASHTO
curve for the specific mix (BSM-emulsion or BSM-foam), the target dry
density was then determined.
4. From step 3 the final mass of material and the mass of material per layer
could then be determined. Only the moisture content needs to be taken
out of the target dry density because the Mod AASHTO curve of the
specific mix accounts for the presence of bitumen.
The mass of material for each individual layer is then determined and
weighed off. This is done because by controlling the mass of material per
layer the thickness of the layer which will yield the equivalent dry density
for 100% Mod AASHTO dry density is known and can be identified during
compaction.
38

The sample was compacted in 5 layers to a height of 300mm, therefore
the equivalent dry density for 100% Mod AASHTO dry density is at 60mm
layer thickness for each layer. The decision to use five layers was based
on ITT Report 18.1-1997 (van de Ven et al, 1997) where five layers were
used during vibratory table compaction.
Prior to compaction the vibratory hammer is placed on the steel rods and
allowed to rest on the base plate. Where the bottom of the sleeve rests
on the rod, a mark is made using masking tape. This is the zero line. The
hammer is then raised and 60mm is measured upward from the zero line
and another mark is made. When, during compaction the sleeve reaches
the 60mm marked off line then it is known that 100% Mod AASHTO dry
density has been reached. After compaction, a mark is made where the
sleeve is at its final resting place, the hammer is raised and 60mm is
measured up from that point again, and so it continues till the compaction
of the sample as whole is completed; figure L.31 provides a visual of how
this procedure looks.
Figure L.31 is a visual of the procedure used to perform the compaction
experiments on the G2, G5 and N7 material. The line marked 100% Mod
AASHTO (60mm) is the position of the foot piece at the point in time
when the sleeve reaches the 60mm marked off line on the guide rod. This
point, based on the relevant calculations, is the point when 100% Mod
AASHTO is achieved. The dashed line marked final position is the point
where the refusal density of the material is reached, in the case of
experiments performed where time was taken as a factor and not as a
fixed value. The experiments were at first run so as to determine the
compaction density plot over time, therefore compaction was done to the
point where it was decided that no further compaction was possible
without crushing the aggregate; refusal density, this point was decided on
when it appeared that the sleeve no longer moved down the guide rods.
After a layer had reached refusal density a mark was made on the guide
rod and 60mm was measured up from that point. This 60mm mark was
then the point at which the following layer reached 100% Mod AASHTO.
It should be noted that during the compaction time to refusal density, the
compaction was stopped at regular time intervals and the layer thickness
at that point in time was noted. This was done by making a mark on the
sleeve at that specific position and measuring the distance from that point
to either the zero line in the case of layer 1 or the refusal density line of
the previously compacted layer in the case of the subsequent layers.
39

In the case of the experiments where time was fixed and not taken as a
factor, the 60mm position of the first layer was determined. Compaction
of the layer took place as described previously but this time for an
allocated time. After compaction of the first layer the thickness was
measured by measuring from the final position of the sleeve, after the
allocated compaction time, to the zero line, a mark was also made and
termed final position. The hammer was then raised and 60mm was
measured up from the point marked off as final position. The subsequent
layers were then compacted according to their allocated times and the
final positions noted. These layer thicknesses were then determined by
measuring from their respective final positions to the final position of the
previously compacted layer. This compaction procedure allowed the
individual compacting to note whether or not the targeted 100% Mod
AASHTO or site achieved density could be achieved in the allocated time.
5. Between the compaction of layers, the surface of the previously
compacted layer is scarified (±10mm) using a chisel, this is done before
the next layer’s material is added. This is done so that interlocking of the
particles may take place.
6. The standard oven drying method is used to check the moisture content
of the compacted sample.
For each experiment a minimum number of two samples were used and a maximum
of three samples was used. Two samples were used in the event that the variability
of results was at a minimum (for e.g. Layer 1 of sample 1 takes 60 sec to compact
and layer 1 of sample 2 takes 61 sec to compact). Typically though, three samples
were used because this gave more accurate results and the variability of the results
could be seen more clearly.
40

Sleeve
Vibratory
Hammer
Zero Line
Side of Mould
Steal Rod
Base Plate
Wooden base
Initial Step:
determining
the Zero Line
60mm
100% Mod AASHTO
(60mm)
Final position
Zero Line
Next Step:
determining
the 60mm Line
for the first
layer
Final position
100% Mod AASHTO(60mm)
Final position
60mm
100% Mod AASHTO (60mm)
Final position
Zero Line
100% Mod AASHTO (60mm)
Remaining
Steps:
determining
the 60mm
Line for the
next layer
Figure L.31: Measurement of layer thicknesses
41

3.2.3 BSM-emulsion
On the N7 recycling site a 60/40 anionic bitumen emulsion (stable grade) was used
for the recycling. The content of emulsion added to treat the material was 3.3%
bitumen emulsion. Therefore, in order to correlate the laboratory compaction to the
site compaction as accurately as possible, the type of bitumen emulsion and
emulsion content used on site was also used for the laboratory experiments. The
bitumen emulsion was acquired from Colas.
3.2.4 BSM-foam
Rehabilitation of the N7 highway between Cape Town and Malmsbury using BSMfoam
took place around 2002/2003. 60/70 penetration bitumen was used with a
content of 2.3% (Theyse, 2003), the First Level Analysis Report: HVS Testing of
Foamed bitumen-treated crushed stone base on N7/1 near Cape Town. The material
type of the base layer was a G2 material.
For purposes of testing in the Laboratory an 80/100 penetration bitumen was used
with a binder content of 1.98% (2%). The binder content of 1.98% was used
because that was the binder content being used on site during rehabilitation using
BSM-emulsion. The result is that a comparison of the compaction of BSM-emulsion
to BSM-foam can be done with the binder content being the same. Below are
pictures of the foam plant WLB 10 at the University of Stellenbosch.
Figure L.32: Foam Nozzle Figure L. 33: Foam plant WLB 10
42

Figure L.34: Twin shaft pug mill mixer
Figure L.35: Twin shaft pug mill mixer with
dome
3.2.5 Further Experimentation
It is thought that this compaction method will also be used to compacted untreated
granular material. Therefore experimentation using clean granular material having
only water added will also be investigated.
43

3.3 Material Property Tests
3.3.1 C.T. Scanning
CT scanning was to be performed on a series of samples compacted under the
Vibratory hammer and Mod AASHTO compaction. The samples were two (2)
vibratory hammer compacted samples: 1 BSM-emulsion sample and 1 BSM-foam
sample. The other sample was a BSM-emulsion sample compacted using Mod
AASHTO compaction.
The purpose of this scanning was to determine what the particle orientation of the
samples was like and what the voids of the compacted sample looked like. TU Delft
in the Netherlands agreed to perform the tests for the University of Stellenbosch.
Transporting the samples oversees was made difficult due the mass of each sample.
A single sample compacted using the vibratory hammer had a mass of around 11kg.
Therefore the samples were cut into two sections of ±75mm thick and the Mod
AASHTO sample was section in half (this sample a massed to ±5kg). The sectioning
is indicated schematically below.
150mm
150mm
150mm
Top:
S1B
Middle:
S1A
75mm
75mm
Piece
Sent: S2
±60mm
Top:
S3B
Middle:
S3A
75mm
75mm
Emulsion Mix: Mod Aashto
BSM-emulsion: Mod
Compaction
AASHTO Compaction
BSM-emulsion:
Emulsion Mix:
Vibratory Vibratory
Hammer
Foam BSM-foam:
Mix: Vibratory
Vibratory Hammer Hammer
Figure L.36: Sectioned samples for CT Scanning
44

3.4 Statistical Analysis
Statistical analyses were performed on the N7 site results. This was done to determine the target
dry density to which the samples in the laboratory would be compacted when using the N7
material. The full statistical analysis performed on the N7 material can be viewed in Appendix A.
For the BSM-emulsion, compaction densities on site were reaching as high as 110% Mod AASHTO
and the lowest result found was at 101% Mod AASHTO. Therefore it was decided to use the 75 th
percentile to which compaction would be targeted. The 100% Mod AASHTO dry density was then
determined.
• 75 th percentile = 104.61% Mod AASHTO = 2255kg/m 3
• The 100%Mod AASHTO target dry density = 2159.05kg/m 3
From the CSIR report stated in sub section 3.2.4 it was found that the BSM-foam layer never
reached 100% Mod AASHTO compaction. Therefore to correlate the compacted laboratory sample to
the site results the mean density achieved on site was used as the target compaction dry density.
• Mean density achieved on site = 2177.33kg/m3
A statistical analysis will also be performed on the experimentation results of both N7 and G2
material.
45

4 Results and Interpretations
This section of the report presents the results of the experiments performed as well as
the interpretations there of. The moisture content provided in the figures in this section
of the report are expressed as a percentage of the OMC of the clean or untreated
material from the Mod AASHTO compaction (OMC (MOD-U)), e.g. 70% OMC is 70% of
the OMC of the untreated material from the Mod AASHTO compaction. Below is a flow
chart depicting the experimentation planning:
Untreated G2 Material: OMC - Mod AASHTO
Curve
BSM: Bitumen Stabilised Material
BSM-emulsion
BSM-foam
Mod AASHTO OMC Curve
Compaction type
Kango 637®/Bosch GSH 11E®
Surcharge
Vibratory Table
TMH 1 Procedure
10 kg 20 kg 15/30 kg
Moisture Content (%)
70%OMC
80%OMC
90%OMC
Choose Best Result
0
Temperature (
Celsius)
5
15 35
Compaction of N7 Material:
Field Correlation
Flow Chart L.1: Experimentation Structure for G2 Granular Materials
46

4.1 Mod AASHTO Curves
Prior to the commencement of the experiments, moisture curves were set up.
First a moisture curve was set up using the clean untreated G2 material. The
result is a follows:
Moisture Curve of Untreated G2 Material: Mod AASHTO
2280.0
2260.0
2240.0
Dry Density (kg/m3)
2220.0
2200.0
2180.0
2160.0
2140.0
2120.0
2100.0
OMC: 6.15%
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0
Moisture Content (%)
Figure L.37: Mod AASHTO Moisture Curve of Untreated G2 Material
From the figure it is found that the OMC (MOD-U) of the G2 material is 6.15% with
a maximum dry density of 2260kg/m 3 . From the OMC, two new figures were set up
using the G2 material. A moisture curve for material treated with emulsion and a
moisture curve for material treated with foamed bitumen. The BSM-emulsion was
prepared using a 60/40 bitumen emulsion with 3.3% emulsion added to the
material. The figures are as follows:
47

Moisture Curve of BSM-emulsion G2 Material: Mod AASHTO
2200
2186
2180
Dry Density (kg/m3)
2160
2140
2120
2100
2080
2060
OMC: 3.8%
0.00 2.00 4.00 6.00 8.00
Moisture Content (%)
Figure L.38: Moisture Curve of BSM-emulsion G2 Material – Mod AASHTO
Moisture Curve of BSM-foam G2 Material: Mod AASHTO
2140.00
2131.00
2120.00
Mod Curve of BSM-foam G2
Dry Density (kg/m3)
2100.00
2080.00
2060.00
2040.00
2020.00
2000.00
OMC: 4.8%
0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00
Moisture Content (%)
Figure L.39: Moisture Curve of BSM-foam G2 Material – Mod AASHTO
From the BSM-emulsion figure and the BSM-foam figure it is found that the
optimum moisture contents are 3.8% and 4.8% respectively. It is to be expected
that the moisture content of the BSM-emulsion is lower than that of the BSM-foam.
This is because in the BSM-emulsion the bitumen itself acts a compaction lubricant,
where in the BSM-foam it does not. Therefore more moisture may be required
48

during compaction of the BSM-foam. For the BSM-emulsion the total fluid content is
taken as the moisture content plus the bitumen binder content, in the BSM-foam it
is only the moisture that acts as a compacting lubricant. Therefore for moisture
content of 3.8% the bitumen emulsion will have a fluid content of 5.78%, much
higher than the foam mix moisture of 4.8%.
Comparative figures are provided below to show how the BSM-emulsion and BSMfoam
moisture curves look in comparison to the moisture curve of the untreated G2
material. Comparative curves for both BSM-emulsion and BSM-foam material were
also set up so as to compare the Mod AASHTO OMC moisture curve to the OMC
moisture curve of the Vibratory Hammer, these curves are provided and discussed
at a later stage in this section of the report.
Comparison between Mod AASHTO curve of Untreated G2 and
BSM-emulsion G2 Material - Mod AASHTO
2300
2260
2250
Mod AASTHO Untreated G2
BSM-emulsion G2
2.35%
Density kg/m3
2200
2186
2150
2100
2050
3.80 6.15
0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00
Moisture %
Figure L.40: Comparative Moisture Curve of Untreated G2 and BSM-emulsion G2 Material –
Mod AASHTO
49

Comparison between Mod AASHTO curve of Untreated G2 and BSMfoam
G2 Material - Mod AASHTO
Dry Density (kg/m3)
2300
2260
2250
2200
2150
2135
2100
Mod curve of untreated G2
Mod Curve of BSM-foam G2
1.35
2050
2000
0.00 2.00 4.00 4.80
6.15
6.00 8.00 10.00
Moisture Content (%)
Figure L.41: Comparative Moisture Curve of BSM-foam and Untreated G2 Material – Mod
AASHTO
50

4.2 Compaction of Site treated Material – BSM-emulsion: Kango 637
At the beginning of 2007, i.e. between February 2007 and April 2007 material was
obtained from the N7 site while the recycling project was underway. This material
had already been milled and treated with bitumen emulsion and was used to
produce samples for triaxial testing that needed to be performed on the material,
the triaxial set up at the University of Stellenbosch does not accommodate a
sample height in excess of 250mm, therefore the samples prepared were to have
dimensions of 250mm x Ф 150mm. The vibratory hammer was used to produce
these samples; therefore initial experimentation of the vibratory hammer
compaction set up took place on this material. For these experiments the Kango
637® was used.
Samples were compacted in five layers (van de Ven, 1997) and the target Dry
Density was known before hand (site compaction results were obtained from Soil
Lab and these were used to determine the target Dry Density of 100% Mod
AASHTO), therefore the mass of material per layer could be determined and the
time to the target Dry Density could be measured; once a layer reached a
thickness of 50mm the 100% Mod AASHTO target Dry Density was achieved. The
marking procedure described in the methodology section was not used for the
experiments on the site treated BSMs, rather at regular time intervals the hammer
was raised and the height of the sample was measured using a tape measure by
measuring from the surface of the compacted layer to the top of the mould and
subtracting that measurement from either 250mm (in the case of layer 1) or the
sum height of the previously compacted layers.
51

4.2.1 0kg Surcharge
120.0
%Mod AASHTO vs. Time: 0kg Surcharge Kango 637®
N7 Site Treated Material
100.0
% Mod AASHTO
80.0
60.0
40.0
Layer 1
Layer 2
Layer 3
Layer 4
Layer 5
20.0
0.0
0 20 40 60 80 100 120
Time (seconds)
Figure L.42: 0kg surcharge - %Mod AASHTO vs. Time of separate layers site treated material,
Kango 637®
From Figure L.42 it is found that with no surcharge (i.e. the hammer is compacting
under its own weight and that of the frame) it takes around 60 seconds to compact
the site treated material to 100% Mod AASHTO (from information provided in the
methodology chapter, 100% Mod AASHTO for this experiment is given as
2130kg/m 3 ). The compaction curve plateaus out between 101% and 103% Mod
AASHTO (depending on the layer). That plateau is taken as the refusal density.
The material compacted for layer one reached a higher refusal density than the
remaining four layers. This is possibly due to it being compacted against the steel
base plate. From the figure it appears that when the material is compacted against
other layers the density is not as high as layer one’s refusal density but they appear
more consistent, this is possibly due to the softer surface of the underlying layer
against which it is compacted when compared to the hard steel base plate.
52

4.2.2 10kg Surcharge
120.0
%Mod AASHTO vs. Time: 10kg surcharge Kango 637®
N7 Site Treated Material
100.0
% Mod AASHTO
80.0
60.0
40.0
Layer 1
Layer 2
Layer 3
Layer 4
Layer 5
20.0
0.0
0 20 40 60 80 100 120 140 160
Time (sec)
Figure L.43: 10kg Surcharge - %Mod AASHTO vs. Time of separate layers of sample of site
treated material, Kango 637®
Figure L.43 shows a slightly different trend from Figure L.42. Here the layers do not
all reach 100% Mod AASHTO in the same time. Layer 1, 3, and 5 all are at 100%
Mod AASHTO after ± 45 sec of compaction (although layer 5 reaches 100% Mod
AASHTO after only 30seconds). Layers 2 and 4 achieve 100% Mod AASHTO after 90
seconds of compaction. The final refusal densities are also more distributed varying
from 109.5% Mod AASHTO for layer 5 to 100.7% Mod AASHTO for layers 2 and 4.
It is not certain why layer 5 achieved such a high refusal density in comparison to
the remaining layers. Small amounts of material were lost when emptying the bag
of weighed of material for layer 5. therefore it is possible that the assumed mass of
material of layer 5, when performing calculations, is in fact less and could yield a
higher density calculation, this because, say for example 2000gm of material is
weighed off for the layer and when it is emptied into the mould assume 100gm is
lost to side of the mould, because there is a small space between the neck
extension and the top of the mould, when the material is compacted to 55mm the
actual density is less due the loss of material than what it would be because there is
less mass in that volume of the layer, but for the calculations the original mass per
layer is assumed.
53

4.3 Influence of Compaction over Time with Varying Moisture Content:
G2 Material
A factor that affects compaction quite significantly is moisture content. Moisture acts
as a lubricant between particles and therefore aids in the shifting and orientation of
particles during compaction. Therefore experiments were carried out using varying
moisture contents and varying the applied dead load so as to asses the outcome of
the compaction.
4.3.1 BSM-emulsion G2 Material
4.3.1.1 10kg Surcharge – Kango 637
70% OMC (MOD-U)
Experiments performed on the BSM-emulsion G2 material at 70% OMC (MOD-
U) were performed by civil engineering student Rojean Hanekom (Hanekom,
2007).
Time To 100% Mod AASHTO Compaction: BSM-emulsion G2
Material 70% OMC (Mod-U) 10kg Surcharge, Kango 637®
60
50
40
Time (Sec)
30
20
10
0
Layer 1 Layer 2 Layer 3 Layer 4 Layer 5
Layer
Figure L.44: Time to 100% Mod AASHTO compaction at 70% OMC (MOD-U) – 10kg Surcharge
BSM-emulsion G2 Material, Kango 637®
54

The results provided in Figure L.44 are the average result of three samples.
The figure shows a variation in the time it takes to compact to 100% Mod
AASHTO with layer 1 compacted the fastest to 100% Mod AASHTO (30
seconds). This is possibly due to the hard steel base against which the first
layer is compacted. Layers 2 and 4 took the longest with both taking 50
seconds to compact to 100% Mod AASHTO. Layer 3 and 5 were also very
close to each other at 40 seconds and 39 seconds respectively. These
results show that there is only a 10 second variation in time with regard to
the compaction of the final 4 layers.
110.0
%Mod AASHTO vs. Time: BSM-emulsion G2 Material 70% OMC
(Mod-U) 10kg Surcharge - Kango 637®
108.0
106.0
%Mod AASHTO
104.0
102.0
100.0
98.0
Layer 1
Layer 2
Layer 3
Layer 4
Layer 5
96.0
94.0
92.0
0 50 100 150 200
Time (Sec)
Figure L.45: Mod AASHTO Density over time of 70% OMC (MOD-U) BSM-emulsion G2 Material
– 10kg Kango 637®
Figure L.45 shows a relatively large variation in the achieved refusal density,
with densities ranging from 103.3 % Mod AASHTO to 109% Mod AASHTO.
Layers 1 through 4 all took 180 seconds (3 minutes) to compact to refusal
density, but layer 5 took the least amount of time, requiring 150 seconds
(2.5 minutes) to achieve refusal density.
55

80% OMC (MOD-U)
Time to 100% Mod AASHTO Compaction: BSM-emulsion G2
Material 80% OMC(Mod-U) 10kg Surcharge, Kango 637®
70.00
60.00
50.00
Time (S)
40.00
30.00
20.00
10.00
0.00
Layer1 Layer2 Layer3 Layer4 Layer5
Layers
Figure L.46: Time to 100% Mod AASHTO compaction at 80% OMC (MOD-U) – 10kg Surcharge
BSM-emulsion, Kango 637®
Figure L.46 shows a very consistent compaction profile for the time to 100% Mod
AASHTO result. Only layer 4 is significantly out when compared to the remaining
layers. When compared to Figure L.44, it appears that at 80% OMC (MOD-U) the
compaction of the material is more consistent.
56

112.0
%Mod AASHTO vs. Time: BSM-emulsion G2 Material 80% OMC
(Mod-U) 10kg Surcharge, Kango 637®
110.0
% Mod AASHTO
108.0
106.0
104.0
102.0
Layer1
Layer2
Layer3
Layer4
Layer5
100.0
98.0
0 50 100 150 200 250
Time (sec)
Figure L.47: Mod AASHTO Density over time of 80% OMC (MOD-U) BSM-emulsion– 10kg
Surcharge BSM-emulsion, Kango 637®
The profile for the refusal density (Figure L.47) shows a slightly different
result in terms of consistency of the results. The achieved refusal densities
vary considerably, ranging from 103.5% Mod AASHTO to about 111% Mod
AASHTO. Compaction times to refusal density also vary, ranging from 155
seconds to 240 seconds. The main reason for this could be the grading.
Although the grading of each sample is the same, the grading of the various
weighed off layers may differ from each other; this trend is seen in the
other experiments. The mass of material required for one sample is decided
before hand (Typically 13kg of material for a sample) and the separate
layers are weighed of from this mass. The grading of the 13kg of material
across the samples is consistent but because the layers are scooped out of
this source using either a scoop or small spade, the grading of each layer
may vary – some layers having more fines than others. If this is the case it
could explain the variability in the refusal density.
57

90% OMC (MOD-U)
Time to 100% Mod AASHTO Compaction: BSM-emulsion G2 ® Material
90% OMC(Mod-U) 10kg Surcharge, Kango 637
Time (s)
16.00
14.00
12.00
10.00
8.00
6.00
4.00
2.00
0.00
Layer1 Layer2 Layer3 Layer4 Layer5
Layers
Figure L.48: Time to 100% Mod AASHTO compaction at 90% OMC (MOD-U) – 10kg Surcharge
BSM-emulsion, Kango 637®
Figure L.48 also shows a relatively consistent compaction profile for the
time to 100% Mod AASHTO result. The times vary by no more than about 2
seconds across the sample. The compaction times range from 11 to 13
seconds, depending on the layer, this is however are too short to allow for
adequate particle orientation. The result is that quality of site representation
may not be adequate enough. The appearance of the material was very
‘slushy’ when compared to the 80% OMC (MOD-U) material.
58

116.0
%Mod AASHTO vs. Time : BSM-emulsion G2 Material 90% OMC
(Mod-U) 10kg Surcahrge - Kango 637®
% Mod AASHTO
114.0
112.0
110.0
108.0
106.0
104.0
102.0
100.0
Layer1
Layer2
Layer3
Layer4
Layer5
98.0
0 50 100 150 200 250 300 350
Time (sec)
Figure L.49: %Mod AASHTO Density over time of 90% OMC (MOD-U) BSM-emulsion G2
Material – 10kg Surcharge, Kango 637®
Once again the refusal density profile (Figure L.49) shows high variability
with densities ranging from 109% Mod AASHTO to about 115% Mod
AASHTO. Compaction times to refusal density also vary, ranging from 210
seconds to 300 seconds. A clear result is that moisture content has an affect
on the compaction of the material. From this result it was decided not to
investigate compaction at 90% OMC (MOD-U) any further. As with the 80%
OMC (MOD-U) the main reason for the variation in refusal density and time
to refusal density could be the grading of the individual layers.
4.3.1.2 20kg Surcharge – Kango 637
70% OMC (MOD-U)
Results are provided with reference to Rojean Hanekom (2007). The results
obtained by Mr. Hanekom indicate that for a 20kg surcharge the compaction
time to 100% Mod AASHTO took longer than that of the 10kg surcharge for
the same moisture content (Figure L.50 and Figure L.44). The result appears
interesting because it would be expected that for a higher surcharge the
compaction time would be less. It was noted during experimentation that the
Kango 637® did not rebound off the material (therefore having lower
amplitude) as well with the 20kg surcharge as it did with the 10kg surcharge,
59

i.e. there was less of impact/vibratory affect on the material and more of a
general vibratory affect. This could be a possible reason for this result.
Time To 100% Mod AASHTO: BSM-emulsion G2 Material 70%
OMC (Mod-U) 20kg Surcharge, Kango 637®
160
140
120
Time (Sec)
100
80
60
40
20
0
Layer 1 Layer 2 Layer 3 Layer 4 Layer 5
Layer
Figure L.50: Time to 100% Mod AASHTO compaction at 90% OMC (MOD-U) – 10kg Surcharge
BSM-emulsion G2 Material, Kango 637®
60

106.0
%Mod AASHTO vs. Time: BSM-emulsion G2 Material 70% OMC (Mod-U)
20kg Surcharge - Kango 637®
104.0
%Mod AASHTO
102.0
100.0
98.0
Layer 1
Layer 2
Layer 3
Layer 4
Layer 5
96.0
94.0
92.0
0 50 100 150 200 250
Time (Sec)
Figure L.51: Mod AASHTO Density over time of 70% OMC (MOD-U) BSM-emulsion– 20kg
Surcharge BSM-emulsion G2 Material, Kango 637®
The refusal density profile of Figure L.51 also shows variability in the refusal
densities achieved, with densities ranging from 101% Mod AASHTO to
about 105% Mod AASHTO. The figure indicates that the time to refusal
density is consistent. As with the 80% OMC (MOD-U) 10kg surcharge the
main reason for the variation in refusal density could be the grading of the
individual layers but interestingly the time to refusal density is consistent.
61

80% OMC (MOD-U)
Time To 100% Mod AASHTO: BSM-emulsion G2 Material 80%OMC
(Mod-U) 20kg Surcharge, Kango 637®
30
25
20
Time (sec)
15
10
5
0
Layer1 Layer2 Layer3 Layer4 Layer5
Layer
Figure L.52: Time to 100% Mod AASHTO compaction at 80% OMC (MOD-U) – 20kg Surcharge
BSM-emulsion G2 Material, Kango 637®
The result of 80% OMC (MOD-U) BSM-emulsion with a 20kg surcharge is a very
consistent time to 100% Mod AASHTO. The 70% OMC (MOD-U) result for the
same criteria (Figure L.50) showed a more variable result; Layers 1 and 4 taking
150 and 90 seconds respectively while layers 2, 3 and 5 took 60 seconds each).
The assumption that a higher surcharge would result in less compaction time
proved to be correct in the experiment at 80% OMC (MOD-U) (Mod-U) (taking
less than half the time of 70% OMC (MOD-U) with the 10kg surcharge) but for the
moisture content of 70% OMC (MOD-U) (figureL.50) it proved to be incorrect.
62

%Mod AASHTO vs. Time : BSM-emulsion G2 Material 80% OMC
(Mod-U) 20kg Surcahrge - Kango 637 ®
112.0
110.0
%Mod AASHTO
108.0
106.0
104.0
102.0
Layer1
Layer2
Layer3
Layer4
Layer5
100.0
98.0
0 50 100 150 200 250 300 350
Time (Seconds)
Figure L.53: Mod AASHTO Density over time of 80% OMC (MOD-U) BSM-emulsion G2 Material
– 20kg Surcharge, Kango 637®
The refusal density profile of Figure L.53 also shows variability in the refusal
densities achieved, with densities ranging from 107% Mod AASHTO to
about 111% Mod AASHTO. The figure also indicates that the time to refusal
density is variable – between 210 seconds and 300 seconds depending on
the layer. As with the 80% OMC (MOD-U) 10kg surcharge the main reason
for the variation in refusal density results could be the grading of the
individual layers. Another possible reason for the variation is the cushioning
effect of subsequent layers. Layer 1 is compacted on the stiffest anvil (the
steel base plate) while each layer following layer 1 is compacted on the
scoured surface of the previous layer, the scoured surface creates a softer
anvil on which the next layer will be compacted, the cushioning effect may
also be a factor that influences the time to 100% Mod AASHTO compaction.
63

4.3.1.3 15kg Surcharge – Kango 637®
Time to 100% Mod AASHTO: BSM-emulsion G2 Material 80% OMC
(Mod-U) 15kg Surcharge, Kango 637 ®
Time (Sec)
80
70
60
50
40
30
20
10
0
Layer1 Layer2 Layer3 Layer4 Layer5
Layers
Figure L.54: Time to 100% Mod AASHTO compaction at 80% OMC (MOD-U) – 15kg Surcharge
BSM-emulsion G2 Material, Kango 637®
Figure L.54 indicates an interesting trend. The 15 kg surcharge at 80%
OMC (MOD-U) shows times to 100% Mod AASHTO generally fall in
between the 10kg and 20kg surcharge. The results, as with the 10kg and
20kg surcharge, are very consistent, only layer one has a notable
variation, taking almost 70 seconds to reach 100% Mod AASHTO.
Readings were taken at 180 seconds for the 15kg surcharge so as to
determine what density could be expected at an extended compaction
time from the time to 100% Mod AASHTO.
64

4.3.1.4 Comparison between Various BSM-emulsion
Experiments-Kango 637®
Time to 100% Mod Aashto AASHTO: compaction: Comparison Comparison between between 70%, 80% 70%, & 90% 80% OMC &
BSM-emulsion 90% OMC G2 Emulsion Material – Mix 10kg, - 10kg, 15kg 15kg & 20kg and 20kg Surcharge, Kango 637®
Time (Sec)
160
140
120
100
80
60
10kg Surcharge (70% OMC)
20kg Surcharge (70% OMC)
10kg Surcharge (80% OMC)
15kg Surcharge (80% OMC)
20kg Surcharge (80% OMC)
10kg Surcharge (90% OMC)
40
20
0
Layer1 Layer2 Layer3 Layer4 Layer5
Layers
Figure L.55: Comparison of Time to 100% Mod AASHTO Compaction For varying Moisture
Contents and Surcharges, BSM-emulsion G2 Material, Kango 637®
The trend described in the previous sub section (4.3.1.3) between the
10kg, 15kg and 20 surcharges at 80% OMC (MOD-U) is evident in Figure
L.55. The results of Mr. Hanekom’s experiments indicate a similar trend
for the 70% OMC (MOD-U) moisture content in the opposite direction.
The result is that the best time to 100% Mod AASHTO compaction that
will also allow for sufficient time for good particle orientation is at 80%
OMC (MOD-U) moisture content with a 10kg surcharge.
65

4.3.1.5 Bosch GSH 11E vs. Kango 637®
Kango 637® vs. Bosch GSH 11E® : Time to 100% Mod AASHTO
BSM-emulsion G2 Material 80% OMC (Mod-U) - 10kg Surcharge
70
60
50
Time to 100% Mod AASHTO:
Bosch GSH 11E®
Time to 100% Mod AASHTO:
Kango 637®
Time (sec)
40
30
20
10
0
1 2 3 4 5
Layer
Figure L.56: Comparison between Compaction times of Bosch GSH 11E® and Kango 637®
Figure L.56 provides a comparative view between the Bosch® and the
Kango® vibratory hammer with regard to BSM-emulsion. It is seen that the
Bosch hammer also provides consistent results with only layer 4 and 5
varying. The difference in compaction time between the Bosch® and
Kango® seems to remain consistent; a difference of 40 seconds is generally
noted with the Bosch® compacting much faster. This reduction in
compaction time by the Bosch GSH 11E® indicates that it exerts more
compacting energy than the Kango 637®.
66

4.3.2 BSM-foam
4.3.2.1 10kg Surcharge – Bosch GSH 11E
Compaction experiments carried out on the BSM-foam G2 material were
performed using the Bosch GSH 11E®. This was due, as was indicated earlier in
the report, to the fact that the Kango 637 ® had become permanently unavailable
due to technical difficulties. Only experiments performed on BSM-foam by Mr.
Rojean Hanekom were done using the Kango 637®, this was because at the time
of his experimentation the Kango 637® had not experienced the technical
difficulties which resulted in a replacement hammer.
70% OMC (MOD-U) - Kango 637®
Experiments performed on BSM-foam material at 70% OMC (MOD-U) were
performed by Rojean Hanekom (Rojean Hanekom, 2007).
Time To 100% Mod AASHTO Compaction: BSM-foam G2 Material
70% OMC (Mod-U) 10kg Surcharge, Kango 637®
70
60
Sample1
Sample2
Average
50
Time (Sec)
40
30
20
10
0
Layer 1 Layer 2 Layer 3 Layer 4 Layer 5
Layer
Figure L.57: Time to 100% Mod AASHTO compaction at 70% OMC (MOD-U) 10kg Surcharge,
BSM-foam G2 Material, Kango 637®
The work done by Mr. Hanekom is shown in Figure L.57 and L.58 Figure
L.57 indicates that there is considerable variability in the compaction of
BSM-foam material when using the vibratory hammer. Samples used for
67

experimentation show the extent to which compaction of the foam mixes
may vary. It is clear from this figure that the way in which the foam mixes
behave is considerably different from the BSM-emulsion. The result of the
70% OMC (MOD-U) foam mix shows that each individual layer requires a
separate compaction time. The middle layer appears to require the most
time to achieve 100% of Mod AASHTO density.
110.0
%Mod AASHTO vs. Time: BSM-foam G2 Material 70% OMC
(Mod-U) - 10kg Surcharge, Kango 637®
108.0
106.0
%Mod AASHTO
104.0
102.0
100.0
98.0
Layer 1
Layer 2
Layer 3
Layer 4
Layer 5
96.0
94.0
0 50 100 150 200 250 300
Time (Sec)
Figure L.58: Mod AASHTO Density over time of 70% OMC (MOD-U) 10kg Surcharge, BSM- foam
G2 Material, Kango 637®
The refusal density plot of Figure L.58 is again indicative of what was seen in the BSMemulsion.
The final refusal densities are variable, and vary from 105% Mod AASHTO to
108% Mod AASHTO. All these densities are achieved in the plateau stage of the figure,
consistently at 180 seconds. Layer 1 reaches the highest refusal density at 108% Mod
AASHTO; this is possibly due to the hard steel base against which it is compacted. Once
again the main factor affecting this variability may be the variability of the grading of the
individual layers.
80% OMC (MOD-U) – Bosch GSH 11E®
The 80% OMC (MOD-U) BSM-foam shows a similar trend to that of the 70%
OMC (MOD-U). There is much variability of the samples used for compaction,
but the nature of the curve when viewing the average result is again a clime
68

to the middle layers and then a tapering off of the final layer. In the 80%
OMC (MOD-U) BSM-foam Layer 4 requires the most amount of time to reach
100% Mod AASHTO. These results again indicate that when compacting foam
mixes with the vibratory hammer each individual layer requires its own
specified time.
Figure L.59: Time to 100% Mod AASHTO compaction at 80% OMC (MOD-U) 10kg Surcharge,
40
35
30
Time To 100% Mod AASHTO: BSM-foam G2 Material 80%OMC
(Mod-U), 10kg Surcharge, Bosch GSH 11E®
Average between 3 Samples
Sample 1
Sample 2
Sample 3
Time (Sec)
25
20
15
10
5
0
Layer1 Layer2 Layer3 Layer4 Layer5
Layers
BSM-foam Bosch GSH 11E®
69

118.0
%Mod AASHTO Vs. Time: BSM-foam G2 Material 80% OMC (Mod-U)
- 10kg Surcharge, Bosch GSH 11E®
116.0
114.0
%Mod AASHTO
112.0
110.0
108.0
106.0
104.0
Layer1
Layer2
Layer3
Layer4
Layer5
102.0
100.0
98.0
0 50 100 150 200 250 300 350
Time (Sec)
Figure L.60: Mod AASHTO Density over time of 80% OMC (MOD-U) 10kg Surcharge, BSMfoam
Bosch GSH 11E®
The refusal density plot of Figure L.60 shows again that the final refusal
density is variable. In this case refusal densities vary from 110% Mod
AASHTO to 116% Mod AASHTO. Layer 1 reaches the highest refusal density
at 116% Mod AASHTO; this is possibly due to the hard steel base against
which it is compacted. The result of this experiment shows how the
moisture content influences the densities. Refusal densities for 80% OMC
(MOD-U) are higher than for 70% OMC (MOD-U).
70

90% OMC (MOD-U) – Bosch GSH 11E®
35
30
25
Time to 100% Mod AASHTO: BSM-foam G2 Material 90% OMC
(Mod-U) -10kg surcharge, Bosch GSH 11E®
Average between 3 Samples
Sample 1
Sample 2
Sample 3
Time (Sec)
20
15
10
5
0
Layer1 Layer2 Layer3 Layer4 Layer5
Layers
Figure L.61: Time to 100% Mod AASHTO compaction at 90% OMC (MOD-U) 10kg Surcharge,
BSM-foam Bosch GSH 11E®
As with the 70% and 80% OMC (MOD-U) BSM-foam curves, the 90% OMC (MOD-
U) curves (Figure L.61) shows the same trend of climbing to the middle layers and
then tapering off at the final layer also there is clear variability in the compaction of
the various samples. The 90% OMC (MOD-U) requires the least amount of time to
reach 100% Mod AASHTO but varies vary little from the 80% OMC (MOD-U) time to
100% Mod AASHTO.
The refusal density plot of Figure L.62 shows a slightly different trend from the
previous two refusal density plots. Here Layer 1 again reaches the highest refusal
density (115.8% Mod AASHTO), but the remaining layers are all concentrated
around the same refusal density point of ± 113% Mod AASHTO. The result of this
experiment also shows that the refusal densities achieved are in the same bracket
as that of the 80% OMC (MOD-U) refusal density plot. Therefore it may be assumed
that there is vary little influence in the compaction for both time and density when
going from 80% to 90% OMC (MOD-U) moisture content.
71

118.0
% Mod AASHTO vs. Time: BSM-foam G2 Material 90% OMC (Mod-U)
- 10kg Surcharge, Bosch GSH 11E®
116.0
114.0
% Mod AASHTO
112.0
110.0
108.0
106.0
104.0
102.0
Layer1
Layer2
Layer3
Layer4
Layer5
100.0
98.0
96.0
0 50 100 150 200 250 300 350
Time (Sec)
Figure L.62: Mod AASHTO Density over time of 90% OMC (MOD-U) 10kg Surcharge, BSMfoam
Bosch GSH 11E®
72

4.3.2.2 15kg Surcharge – Bosch GSH 11E
40
35
30
Time to 100% Mod AASHTO: BSM-foam G2 Material 80% OMC
(Mod-U) -15kg surcharge, Bosch GSH 11E®
Average between 3 Samples
Sample 1
Sample 2
Sample 3
Time (Sec)
25
20
15
10
5
0
Layer1 Layer2 Layer3 Layer4 Layer5
Layers
Figure L.63: Time to 100% Mod AASHTO compaction at 80% OMC (MOD-U) 15kg Surcharge,
BSM-foam Bosch GSH 11E®
The results presented in Figure L.63 show a variation in the behaviour of
the material when compacting under a higher surcharge. Here the curve
climbs, but as apposed to peaking in the middle layer it drops off and the
climbs to the final layer. The characteristic that each layer requires an
individual time to reach 100% Mod AASHTO is still present in this
experiment.
73

120.0
%Mod AASHTO vs. Time: BSM-foam G2 Material 80% OMC (Mod-U)
- 15kg Surcharge, Bosch GSH 11E®
118.0
116.0
% Mod AASHTO
114.0
112.0
110.0
108.0
106.0
Layer1
Layer2
Layer3
Layer4
Layer5
104.0
102.0
100.0
0 50 100 150 200 250 300 350
Time (Sec)
Figure L.64: Mod AASHTO Density over time of 80% OMC (MOD-U) 15kg Surcharge, BSMfoam
Bosch GSH 11E®
The refusal density plot of Figure L.64 also shows that the final refusal
density is variable. Here the refusal densities vary from 112% of Mod
AASHTO density to 118% of Mod AASHTO density. Layer 1 reaches the
highest refusal density at 118% Mod AASHTO; this is possibly due to the
hard steel base against which it is compacted. From this experiment it may
be seen that the result of increasing the surcharge weight has little affect
on the refusal densities achieved, it is seen that after 300 seconds
compaction time densities are still increasing. For the BSM-foam the
compaction was stopped after 5 minutes (300sec) as it is believed that
compaction exceeding 5 minutes is not likely to take place in preparation of
samples. The bracket of refusal densities from this experiment is similar to
that of the 80% and 90% OMC (MOD-U) BSM-foam with the 10kg
surcharge.
74

4.3.3 Comparison between BSM-emulsion and BSM-foam G2
Material
Figures L.65 and L.66 provide a graphic view of the difference in
the response of BSM-emulsion and BSM-foam G2 material to
vibratory hammer compaction. The BSM-emulsion has a more
consistent time to 100% Mod AASHTO compaction where the BSMfoam
has a clear curve to it indicating that each layer requires a
different time to reach 100% Mod AASHTO compaction.
Time to 100% Mod AASHTO Compaction: BSM-emulsion G2
Material 80% OMC(Mod-U) 10kg Surcharge, Kango 637®
70.00
60.00
50.00
Time (S)
40.00
30.00
20.00
10.00
0.00
Layer1 Layer2 Layer3 Layer4 Layer5
Layers
Figure L.65: Time to 100% Mod AASHTO compaction at 80% OMC (MOD-U) – 10kg Surcharge
BSM-emulsion, Kango 637®
75

Time To 100% Mod AASHTO: BSM-foam G2 Material 80%OMC
(Mod-U), 10kg Surcharge, Bosch GSH 11E®
35.00
30.00
100% Mod AASHTO
Poly. (100% Mod AASHTO)
25.00
Time (Sec)
20.00
15.00
10.00
5.00
0.00
Layer1 Layer2 Layer3 Layer4 Layer5
Layer
Figure L.66: Time to 100% Mod AASHTO compaction at 80% OMC (MOD-U) 10kg Surcharge,
BSM-foam Bosch GSH 11E®
76

4.4 Time as a Fixed Unit with a Constant Moisture Content and Set
Dead Load: G2 Material
4.4.1 Vibratory Table: 50kg Dead Load
4.4.1.1 BSM-emulsion
The criteria used for the vibratory table was taken from the TMH1
Revised Addition (1990). The frequency for the table was set at
0.5Hz. Vibratory table compaction was done using a dead load of
50kg and the samples were compacted in 5 layers until 100%
Mod AASHTO density was achieved. The experiment was
conducted using two samples for an experiment and the average
time to obtain 100% Mod AASHTO Density of the three samples
was used to plot Figure L.67.
Figure L.67: Comparison of Vibratory Table to Kango 637® Vibratory Hammer 10kg
Time to 100% Mod AASHTO - Vibratory Table vs. Kango 637® with
10kg Surcharge: BSM-emulsion G2 Material 80% OMC (Mod-U)
70
60
Vibratory Table
Kango 637 ®
50
Time (Sec)
40
30
20
10
0
Layer1 Layer2 Layer3 Layer4 Layer5
Layer
Surcharge: 80% OMC (MOD-U) BSM-emulsion G2 Material
The results of Figure L.67 show a consistency in the time to
100% Mod AASHTO compaction. The vibratory table requires less
time to compact to the target density of 100% Mod AASHTO
when compared to the vibratory hammer, but the physical labour
involved in using the vibratory table was much more than for the
vibratory hammer. Typical problems that arose during compaction
77

were that the cross bars used to fasten the mould to the table
frequently came loose and the 50kg surcharge (deadweight
placed in the mould) was difficult to keep perpendicular to the
material surface, this resulted in layers not necessarily being level
after compaction but rather have a sloping effect. The end result
of this is that although the vibratory table requires less
compaction time per layer, the overall time to prepare a single
sample was much more for the vibratory table (± 45 – 60min)
than for the vibratory hammer (±30 – 35min) – almost double
the time.
78

4.4.1.2 BSM-foamed
The same procedure for the BSM-emulsion material (sub section
4.3.4.1) was applied to the BSM-foam material.
Time to 100% Mod AASHTO - Vibratory Table vs. Bosch GSH 11E®
10kg surcharge: BSM-foam G2 Material 80% OMC (Mod-U)
90
80
Vibratory Table
Bosch GSH 11E
Time (Sec)
70
60
50
40
30
20
10
0
Layer1 Layer2 Layer3 Layer4 Layer5
Layer
Figure L.68: Comparison of Vibratory Table to Bosch ® Vibratory Hammer 10kg Surcharge:
80% OMC (MOD-U) BSM-foam G2 Material
Upon comparing the vibratory table to the vibratory hammer for the BSMfoam
a similar trend is noted in the vibratory table compaction. The
compaction curve with time climbs and then tapers off at the final layer.
The difference being where in Layer 4 of the BSM-foam the peak occurs for
the vibratory hammer, this same peak occurs at Layer 2 of the vibratory
table. The time it takes for the vibratory table to achieve 100% mod
AASHTO is much higher than for the vibratory hammer, this is the exact
opposite what is seen in Figure L.67 of the BSM-emulsion where less time is
required by the vibratory table to achieve this target density of 100% Mod
AASHTO.
It must be remembered that although a difference between the results of
the BSM-foam and BSM-emulsion is evident, the BSM-foam was compacted
using the Bosch® hammer where the BSM-emulsion was compacted using
the Kango 637®. The results of the correlation experiment between the two
79

vibratory hammers (Figure L.56) indicates that the same trend will occur for
the BSM-emulsion as in the BSM-foam should the Bosch® hammer be used
to compact the BSM-emulsion.
Vibratory Table vs. Bosch GSH 11E® 10kg surcharge: BSM-foam G2
Material 80% OMC (Mod-U) - Final %Mod AASHTO Density
110.00
108.00
Vibratory Table
Bosch GSH 11E
% Mod AASHTO
106.00
104.00
102.00
100.00
98.00
96.00
Layer1 Layer2 Layer3 Layer4 Layer5
Layer
Figure L.69: %Mod AASHTO Compaction after 120sec Compaction Time relative to Bosch GSH
11E® 100% Mod AASHTO
Figure L.69 provides a Graphic view of the relationship of the
vibratory table, when compacting according to the full specification
of TMH1 (1990), compared to the result of compacting the BSMfoam
to 100% Mod AASHTO using the vibratory hammer. Not much
variation is noted in the density of the various layers –from Layer 1
to Layer3; a difference of about 1% is notable. Layer 4, however,
shows significant increase in the density, no explanation can be
given for this increase. Layer 5 shows that the material did not
compact any further than the 100% Mod AASHTO density mark.
As with the BSM-emulsion experiment, the vibratory table required
much more time (±40 – 45min) to prepare a single sample than
that of the Bosch ® vibratory hammer (±25 – 30min).
80

4.4.2 Temperature Variation: BSM-emulsion
From the results of experiments performed earlier the criterion for
these experiments were set up. Results showed that compaction
moisture content of 80% OMC (MOD-U) with a surcharge of 10kg
for the Bosch vibratory hammer, was the best criterion for
representative compaction and was therefore selected for the
performance of further experimentation. From the result of time to
100% Mod AASHTO compaction for this criterion and the correlation
experiment (Figure L.56), times were allocated to each layer as to
what would yield 100% Mod AASHTO compaction.
Figure L.70: % Mod AASHTO Compaction at Fixed Time Units: 80% OMC (MOD-U) 10kg
109.0
108.0
% Mod AASHTO At Fixed Time Units: Temperature Variation BSMemulsion
G2 Material 80% OMC (Mod-U) - Bosch GSH 11E ® 10kg
Surcharge
5 Deg C
15 Deg C
35 Deg C
107.0
%Mod AASHTO
106.0
105.0
104.0
103.0
102.0
101.0
100.0
Layer1 (10 sec) Layer2 (15 sec) Layer3 (15 sec) Layer4 (15 sec) Layer5 (15 sec)
Layer (Time sec)
Surcharge, BSM-foam Bosch GSH 11E®
The temperature of material is a factor which affects compaction;
Figure L.70 shows the effect of the temperature on the compaction
of the material. It is intuitive that at lower temperatures the
material would not compact as well as it would at higher
temperatures. This trend is not seen when comparing the results of
the 5 0 C to 35 0 C experiment. For the 5 0 C and 15 0 C experiments,
material was cooled in a fridge. The material was allowed to stand
over night and was then removed from the fridge and the
temperature measured. When the two bags marked off for the 5 0 C
81

experiment measured 5 0 C a single bag was taken to compact and
the remaining bags placed in the fridge; this kept the bags at their
target temperature as the temperature of the fridge measured
around 5 0 C. The bags marked off for the 15 0 C experiment were
allowed to stand out side of the fridge to warm up to 15 0 C (±2 0 C),
this was done was they had also reached 5 0 C. The trend from the
experiment shows a reversal in the expected trend with the
compatibility of the material becoming poorer as the temperature is
increased.
82

4.4.3 Temperature Variation: BSM-foam
As in sub section 4.4.1 the results of experiments performed earlier
were used to establish the criterion against which these
experiments for the BSM-foam would take place. Results showed
that a moisture content of 80% OMC (MOD-U) with a surcharge of
10kg was the best criterion and as such was used to perform these
experiments. From the result of time to 100% Mod AASHTO
compaction for this criterion times were allocated to each layer as
to what would yield 100% Mod AASHTO compaction.
%Mod AASHTO At Fixed Time Units: Temperature Variation BSM-foam
G2 Material 80% OMC (Mod-U) - Bosch GSH 11E® 10kg Surcharge
108.00
106.00
5 Deg C
15 Deg C
35 Deg C
% Mod AASHTO
104.00
102.00
100.00
98.00
96.00
94.00
L1 (10sec) L2 (28sec) L3 (30sec) L4 (34sec) L5 (30sec)
Layer (Time sec)
Figure L.71: %Mod AASHTO Compaction at Fixed Time Units: 80% OMC (MOD-U) 10kg
Surcharge, BSM-foam Bosch GSH 11E®
As in sub section 4.4.1 the trend of how material compacts at
different temperatures is seen in layers 1 and 2 and when
comparing the 5 0 C to 35 0 C experiment. Material for the 5 0 C and
15 0 C experiment were again prepared by cooling them in a fridge,
however the fridge temperature was not adjusted for the 15 0 C
material. Instead it was cooled to 5 0 C and allowed to stand in the
bags outside the fridge for roughly 30 – 40 minutes. A thermometer
and temperature probe was used to check the material
temperature. Once a bag reached the target temperature of 15 0 C ±
2 0 C, the bag was taken and the material compacted. Therefore no
83

excess water was acquired by the material as a result of melting ice
and no moisture was lost; moisture content readings show that the
moisture obtained for the 15 0 C experiment was around 74%OMC
(MOD-U). It cannot be explained why the compaction trend of
Layers 3, 4 and 5 indicates a drop off in the density achieved at
15 0 C compared with 5ºC.
84

4.5 Moisture Curves of the Vibratory Hammer
4.5.1 BSM-emulsion
The criterion of surcharge mass and compaction time per layer used
in sub section 4.4.1 was applied to this experiment. Here moisture
contents of 60, 70, 80, 90 and 100% OMC (MOD-U) were used.
Compaction was done using the Bosch GSH 11E® and the results
are provided in Figure L.72. The curves provided are the Mod
AASHTO curve for the BSM-emulsion and the moisture content
curve for the vibratory hammer compacted BSM-emulsion.
Moisture Curve Mod AASHTO vs. Vibratory Hammer - Bosch GSH 11E®
10 kg Surcharge: BSM-emulsion G2 Material
2300.00
Vibratory Hammer
2263.00
Mod AASHTO
2250.00
Poly. (Vibratory Hammer)
Dry Density (kg/m3)
2200.00
2190.00
2150.00
2100.00
2050.00
63.4% OMC
91.1% OMC
0 2
3.9%
4
5.6%
6 8 10 12
Moisture Content (%)
Figure L.72: Moisture Curve of Mod AASHTO vs. Vibratory Hammer: BSM-emulsion G2 Material
The result of the BSM-emulsion experiment shows that the OMC
(Vib Treated) point of the vibratory hammer is much higher (almost
30%OMC (MOD-U) higher) than for the Mod AASHTO OMC
(Treated) point. The vibratory hammer curves at 5.6% moisture
content where the Mod AASHTO curves at 3.9% moisture content,
this almost 2% more moisture for the vibratory hammer before it
curves. Analysing the densities, the vibratory hammer achieved a
density of 2263kg/m 3 ; this is much higher than the Mod AASHTO
peak density of 2190kg/m 3 . Another interesting point is that the
85

curve shows that the vibratory hammer is more sensitive to
moisture than the Mod AASHTO compaction. For the information
provided in the table below, the moisture contents ranging from
80% OMC to 100% OMC yield an average density of 102.86% Mod
AASHTO. Taking this average and referring to sub section 4.7.1 Site
Results: BSM-emulsion, the procedure used achieved the 95 th
percentile of the site compaction results of 102.71% Mod AASHTO
at these three moisture contents.
Table L.12: %Mod AASHTO Compaction Achieved Using the Vibratory
Hammer – BSM-emulsion G2 Material
Sample M.C %OMC Achieved Dry Density (kg/m 3 ) % Mod
AASHTO
3 4.96 80.65 2206.76 101.23
4 5.53 89.92 2257.19 103.78
5 6.2 100.81 2245.76 103.59
The average density for the last three samples is 102.86% Mod AASHTO
86

4.5.2 BSM-foam
The criterion of surcharge mass and compaction time per layer used
in sub section 4.4.2 was applied to this experiment. As in sub
section 4.5.1 moisture contents of 60, 70, 80, 90 and 100% OMC
(MOD-U) were used. Compaction was done using the Bosch GSH
11E® and the results are provided in Figure L.73. The curves
provided are the Mod AASHTO curve for the Foam mix material and
the moisture content curve for the vibratory hammer compacted
foam mix material.
Moisture Curve Mod AASHTO vs. Vibratory Hammer - Bosch GSH 11E®
10 kg Surcharge: BSM-foam G2 Material
2320.00
2298.00
Mod AASHTO
Vibratory Hammer
2270.00
Dry Density (kg/m3)
2220.00
2170.00
2138.00
2120.00
2070.00
2020.00
1970.00
71.5% OMC
78% OMC
2 3 4
4.4 4.8
5 6 7 8
Moisture Content (%)
Figure L.73: Moisture Curve of Mod AASHTO vs. Vibratory Hammer: BSM-foam G2 Material
Contrary to sub section 4.5.1 the OMC (Vib BSM-foam) point of the
vibratory hammer is lower than that of the Mod AASHTO OMC
(BSM-foam) point; the difference is not significant, only 7%OMC
(MOD-U) i.e. 0.4% moisture per mass of material. However when
the achieved densities are viewed a clear difference is seen. The
vibratory hammer achieved a density of 2298kg/m 3 ; this is much
higher than the Mod AASHTO peak density of 2138kg/m 3 . The
figure shows that the vibratory hammer achieves higher densities
than the Mod AASHTO compaction at the same moisture contents.
87

Figure L.40 shows, as in Figure L.39, that the vibratory hammer is
more sensitive to moisture than the Mod AASHTO compaction.
The maximum dry density achieved with the Bosch vibratory
hammer using the BSM-foam (2298kg/m 3 ) is slightly higher than
the maximum dry density achieved for the BSM-emulsion (2263
kg/m 3 : Figure L.72). The reason for this could be that the
compaction times per layer of the BSM-emulsion were generally
around 15 seconds per layer, where the BSM-foam experienced
longer compaction time, varying from 28 to 34 seconds depending
on the layer. The BSM-emulsion and BSM-foam show a similar level
of sensitivity to the moisture content. If the incline is taken as a
straight line then the gradients are 2% for the BSM-emulsion and
1.835% for the BSM-foam. From this the BSM-emulsion does
appear slightly more sensitive to moisture. Table L.10 indicates that
the longer compaction time for the BSM-foam does in fact have an
influence on the extent to which the samples were compacted. The
compaction densities achieved for the above curve exceed 100%
Mod AASHTO very comfortably; the density achieved near the peak
is 108.4% Mod AASHTO.
Table L.13: %Mod AASHTO Compaction Achieved Using the Vibratory
Hammer – BSM-foam G2 Material
Sample M.C %OMC Achieved Dry Density (kg/m 3 ) % Mod
AASHTO
1 3.32 53.97 2110.62 103.72
2 3.72 60.55 2215.71 107.14
3 4.09 66.52 2224.55 106.03
4 4.8 78.07 2262.49 105.97
5 5.53 89.9 2178.32 104.48
The average density achieved was 105.47% Mod AASHTO
As with the BSM-emulsion material, these results indicate that at
higher moisture contents higher compaction, in terms of %Mod
AASHTO, is achievable.
88

4.6 Correlation to Site Material and Site Achieved Compaction
In order to correlate the compaction experimentation of the G2 material to
compaction being achieved in the field material was acquired from a Cold in Place
Recycling (CIPR) Project along the N7. On this specific site a CIPR Project had
taken place in 2002/2003 where a lane was treated using BSM-foam. In 2007 a
CIPR Project took place again, but in this specific project BSM-emulsion was used
to treat the appropriate lanes. The field compaction results for the BSM-foam CIPR
were obtained from HL Theyse, CR-2003/23. The field compaction results and
relevant field information for the BSM-emulsion CIPR were respectively obtained
from Soil Lab and the engineers responsible for the project. The field results for
the respective mixes are provided in the statistical analysis section of this report.
Under this section (4.6) results marked N7 material are the results of the milled N7
material treated with either bitumen emulsion or foamed bitumen and then
compacted in the laboratory.
4.6.1 BSM-emulsion
The criteria of moisture content, surcharge mass and compaction
time per layer used in sub section 4.4.1 was applied to this
experiment. Compaction was done using the Bosch GSH 11E® and
the results are provided in Figure L.74 The figure indicates how the
N7 material correlates to the G2 material when using the same
criterion when both materials have been treated with bitumen
emulsion. From the statistical analysis results of site compacted
material, the 85 th percentile was chosen as the target density mark,
this in tern showed that the target dry density of the site
compacted material which would give its equivalent 100% Mod
AASHTO was 2138kg/m 3 , this is the target dry density of the N7
BSM-emulsion.
89

% Target Dry Density
110.0
108.0
106.0
104.0
102.0
100.0
98.0
96.0
Comparison Between G2 & N7 Compacted Material:
80% OMC (Mod-U), Bosch GSH 11E® 10kg Surcharge, BSM-emulsion
N7 Material
5 Deg C G2
15 Deg C G2
35 Deg C G2
94.0
92.0
90.0
Layer1 (10 sec) Layer2 (15 sec) Layer3 (15 sec) Layer4 (15 sec) Layer5 (15 sec)
Layers (Compaction Time in seconds)
Figure L.74: Comparison of G2 and N7 Material: BSM-emulsion
What becomes immediately clear is that the criterion for the clean
G2 material does not work for the N7 material. The N7 material
never reaches 100% of the achieved site density. The main reason
for this is the grading. From the grading curves presented in the
Methodology section it is clear that the grading between the G2
material and N7 material differ. Material properties also differ, this
because although the N7 material is of G2 quality, it contains RAP in
it as a result of the milling process; traces of cement may also be
present should the base have been treated with cement during the
initial construction of the road.
4.6.2 BSM-foam
The criterion of surcharge mass and compaction time per layer used
in sub section 4.4.2 was applied to this experiment. Compaction
was done using the Bosch GSH 11E® Hammer and the results are
provided in Figure L.75. The figure shows how the N7 material
correlates to the G2 material when using the same criterion when
both materials have been treated with foamed bitumen. A statistical
analysis was also performed on the results of site compacted
material taken from the CSIR report CR 2003/23 (Theyse, 2003);
the mean dry density was chosen as the target density to which the
N7 material would be compacted. This is due to the fact that on the
90

BSM-foam CIPR site compaction never reached 100% Mod AASHTO
compaction. The mean density was found to be 2177.33kg/m 3 .
Figure L.75: Comparison of G2 and N7 Material: BSM-foam
Comparison Between G2 & N7 Compacted Material:
80% OMC (Mod-U), Bosch GSH 11E® 10kg Surcharge, BSM-foam
108.0
Comparison between G2 & N7 compacted Material: Foamed Mix
108.00
106.00
104.00
% Target Dry Density
102.00
100.00
98.00
96.00
94.00
92.00
N7 Material
5 Deg C G2
15 Deg C G2
35 Deg C G2
90.00
88.00
Layers (Compaction Time in seconds)
L1 (10s) L2(28s) L3 (30s) L4 (34s) L5 (30s)
Layer (time sec)
As in sub section 4.6.1 it is immediately clear that the criterion for
the clean G2 material does not work for the N7 material. Once
again the N7 material never reaches 100% of the achieved site
density. The main reason for this is the reason discussed in sub
section 4.6.1 regarding the grading and material properties.
It is also clear that the mix type, BSM-foam or BSM-emulsion, is not
the cause of this behaviour of the N7 material. Densities achieved
for both mixes of the N7 material are similar e.g. layer 1 in both
cases is around 96% of the of the achieved site density.
91

4.7 Statistical Results
4.7.1 Site Results: BSM-emulsion
Statistical Calculations: N7 Site Information -
Emulsion Mix
%Mod Aashto
n Dry Density %Mod Aashto
1 2181 101.6
2 2227 103.6
3 2255 104.2
4 2255 104.7
5 2272 106.1
6 2287 106.2
7 2305 107.4
8 2323 107.5
9 2345 108.9
10 2345 109.6
11 2350 110.1
12 2367 110.7
Percentile
Mean = 106.72 % 75th : 104.61 %
Std Dev = 2.83 % 85th : 103.99 %
COV = 2.65 % 95th : 102.71 %
Outliers
n = 12
x0 = 101.6
|T0| = 1.81
T = 2.29
|T0|

The results of this statistical analysis show that the compaction
results on site have vary little variation; this is seen in COV value of
2.48%. It is interesting to note that on average, the dry density
achieved on site was in the order of 106% of Mod AASHTO density,
these are very high levels of compaction and the level of variation
was also very low, COV equalling 2.65%.
93

4.7.2 Site Results: BSM-foam
Foam Mix on N7
First Level analysis report:
HVS Testing of the Foamed
Bitumen-Treated crushed
Stone Base on N7/1 near
Cape Town Version: 1st draft
CSIR transportek
Pages 13, 14, 15
Fig 14 and Fig 15
Summary of the pavement and instrumentation detail detail of of the the test tst sections
Binder type 60/70
Binder Content 2.50 %
Field Dry Density (Fig 14)
Field Dry Density (Fig 15)
2137 kg/m3
2166 kg/m3
Appendix C: rehabilitation investigation
Distance (km) Field Dry Density Mod AASHTO Dry Density
17.7 2109 kg/m3 2260 kg/m3
16 2148 kg/m3
14 2146 kg/m3
11.99 2224 kg/m3
10 2218 kg/m3
7.8 2219 kg/m3
Mean = 2177.33 kg/m3
St Deviation = 49.15 kg/m3
C.O.V = 2.26 %
The results of this statistical analysis show, as in the analysis of the
BSM-emulsion CIPR, that the compaction results on site have very
little variation; The COV value for the BSM-foam is 2.26%. In
contrast to the field compaction of the BSM-emulsion, the BSMfoam
never obtained its 100% Mod AASHTO level of compaction;
on average a compaction density of 96.34% Mod AASHTO was
achieved.
Overall, the statistical results from field compaction indicate that the
extent to which field compaction varies is very low. The BSMemulsion
and BSM-foam CIPR results had COV’s of 2.48% and
94

2.26% respectively.
95

4.7.3 Laboratory BSM-emulsion G2 Material
10kg Surcharge
70% OMC
Sample1
Sample2
Time to 100%
Mod AASHTO
(sec)
Time to 100%
Mod AASHTO
(sec)
STD C.O.V
Layer
Mean Dev (%)
Layer1 30 15 22.50 10.61 47.14
Layer2 60 15 37.50 31.82 84.85
Layer3 45 60 52.50 10.61 20.20
Layer4 30 45 37.50 10.61 28.28
Layer5 30 30 30.00 0.00 0.00
Mean 39.00 33.00 36.00
STD Dev 13.42 19.56 16.49
C.O.V (%) 34.40 59.27 46.83
80% OMC
Sample1 Sample2 Sample3
Time to 100%
Mod AASHTO
(sec)
Time to 100%
Mod AASHTO
(sec)
Time to 100%
Mod AASHTO
(sec)
STD C.O.V
Layer
Mean Dev (%)
Layer1 35 55.37 50.28 46.88 10.60 22.61
Layer2 41 60.73 36 45.91 13.08 28.48
Layer3 51.1 55.82 29.86 45.59 13.83 30.33
Layer4 60 62.01 122 81.34 35.23 43.31
Layer5 25 58.41 75.37 52.93 25.63 48.42
Mean 42.42 58.47 62.70 54.53
STD Dev 13.65 2.93 37.49 18.02
C.O.V (%) 32.17 5.01 59.79 32.32
96

90% OMC
Sample1 Sample2 Sample3
Time to 100%
Mod AASHTO
(sec)
Time to 100%
Mod AASHTO
(sec)
Time to 100%
Mod AASHTO
(sec)
STD C.O.V
Layer
Mean Dev (%)
Layer1 12.52 11.65 10.93 11.70 0.80 6.80
Layer2 15.77 12.77 9.81 12.78 2.98 23.31
Layer3 9.47 13.29 9.6 10.79 2.17 20.11
Layer4 10.45 17.19 12.57 13.40 3.45 25.71
Layer5 11.72 14.9 13.07 13.23 1.60 12.06
Mean 11.99 13.96 11.20 12.38
STD Dev 2.42 2.15 1.58 2.05
C.O.V (%) 20.16 15.41 14.08 16.55
15kg Surcharge
80% OMC
Sample1 Sample2 Sample3
Time to 100%
Mod AASHTO
(sec)
Time to 100%
Mod AASHTO
(sec)
Time to 100%
Mod AASHTO
(sec)
STD C.O.V
Layer
Mean Dev (%)
Layer1 67.44 53 81.85 67.43 14.43 21.39
Layer2 28.68 23.47 40.37 30.84 8.65 28.06
Layer3 35.74 24.59 42.42 34.25 9.01 26.30
Layer4 25.89 26 39.57 30.49 7.87 25.80
Layer5 18.24 25.33 37.4 26.99 9.69 35.89
Mean 35.20 30.48 48.32 38.00
STD Dev 19.08 12.63 18.83 16.85
C.O.V (%) 54.21 41.42 38.97 44.87
97

20kg Surcharge
70% OMC
Sample1 Sample2 Sample3
Time to 100%
Mod AASHTO
(sec)
Time to 100%
Mod AASHTO
(sec)
Time to 100%
Mod AASHTO
(sec)
STD C.O.V
Layer
Mean Dev (%)
Layer1 130 210 110 150.00 52.92 35.28
Layer2 25 90 86 67.00 36.43 54.37
Layer3 25 46 165 78.67 75.50 95.98
Layer4 63 45 170 92.67 67.57 72.92
Layer5 32 60 145 79.00 58.85 74.49
Mean 55.00 90.20 135.20 93.47
STD Dev 44.77 69.39 36.23 50.13
C.O.V (%) 81.40 76.93 26.80 61.71
80% OMC
Sample1 Sample2 Sample3
Time to 100%
Mod AASHTO
(sec)
Time to 100%
Mod AASHTO
(sec)
Time to 100%
Mod AASHTO
(sec)
STD C.O.V
Layer
Mean Dev (%)
Layer1 17.3 18.87 20 18.72 1.36 7.24
Layer2 12.8 22 20.99 18.60 5.05 27.13
Layer3 16.32 28.03 12.28 18.88 8.18 43.34
Layer4 13 22.3 23.13 19.48 5.62 28.88
Layer5 24.91 18.37 16.44 19.91 4.44 22.30
Mean 16.87 21.91 18.57 19.12
STD Dev 4.92 3.85 4.27 4.35
C.O.V (%) 29.15 17.58 22.98 23.24
98

The statistical results of the laboratory experimentation provided on
the previous two pages show that for a 10kg surcharge the
moisture content of 90% OMC (Mod-U) required, on average, the
least amount of time to compact a layer to 100% Mod AASHTO,
12.38sec. The interesting result from this analysis is that on
average at 70% OMC (Mod-U) less time is required to compact a
layer to 100% Mod AASHTO density when compared to 80% OMC
(Mod-U); at 70% OMC (Mod-U) 36sec is the average time to
compact a layer to 100% Mod AASHTO where at 80% OMC (Mod-
U) it is 54.53sec, this is not what is expected as the moisture
content is increased.
It is also notable that for the 10kg surcharge the 90% OMC (Mod-
U) has a lower standard deviation (Std Dev) and coefficient of
variation (COV). The Std Dev is only 2 seconds and the COV is
16.55%. The COV may be regarded as high, but when compared to
the 70% and 80% OMC (Mod-U) of 46.83% and 32.32%
respectively, it is much lower.
The 15kg surcharge at 80% OMC (Mod-U) shows a consistency,
statisticaly, to the 10kg surcharge of 70% OMC (Mod-U). The mean
time to 100% Mod AASHTO differs by only 2 sec and the Std Dev
for 15kg surcharge is 16.85sec compared to the 16.49sec of the
10kg surcharge at 70% OMC (Mod-U). The COV’s are also very
similar, 44.87% for the 15kg surcharge and 46.83% for the 10kg
surcharge at 70% OMC (Mod-U). This result of the 15kg surcharge
indicates that similar behaviour of the material will occur when it is
compacted at 70% OMC (Mod-U) with a 10kg surcharge.
The trend differs when a 20kg surcharge is used. At 70% OMC
(Mod-U) the mean time to 100% Mod AASHTO more than doubles,
from 36sec at 10kg surcharge to 93.47 sec at 20kg. The Standard
Deviation and COV also increase dramatically from 16.49sec to
50.13sec and from 46.83% to 61.71% respectively. The resulting
effect is that as the surcharge load is increased, the material
becomes less responsive to compaction at 70% OMC (Mod-U). This
is due to the effect that at a higher surcharge load there is less
vibratory/impact affect and more of a general vibratory effect.
99

The trend for 80% OMC (Mod-U) with a 20kg surcharge has the
opposite effect to that of the 70% OMC (Mod-U) of the previous
paragraph. The mean time to 100% Mod AASHTO more than halves
and the Std Dev is almost one quarter of the Std Dev for the 10kg
surcharge. The COV however does not drop as dramatically as the
Standard Deviation and mean time to 100% Mod AASHTO; it does
however drop significantly, i.e. by 10%.
100

Temperature Variation: 10kg Surcharge 80% OMC (MOD-U)
5 Deg Celscius
Sample1
Sample2
STD
Dev
C.O.V
(%)
% Mod AASHTO % Mod AASHTO
Layer Achieved Achieved Mean
Layer1 (10s) 108.59 107.39 107.99 0.84 0.78
Layer2 (15s) 106.66 106.44 106.55 0.16 0.15
Layer3 (15s) 107.62 106.44 107.03 0.83 0.77
Layer4 (15s) 105.73 105.51 105.62 0.15 0.15
Layer5 (15s) 107.62 106.44 107.03 0.83 0.77
Mean 107.24 106.45 106.84
STD Dev 1.09 0.67 0.88
C.O.V (%) 1.01 0.63 0.82
15 Deg Celscius
Sample1
Sample2
STD
Dev
C.O.V
(%)
% Mod AASHTO % Mod AASHTO
Layer Achieved Achieved Mean
Layer1 (10s) 105.30 107.54 106.42 1.58 1.48
Layer2 (15s) 107.18 108.50 107.84 0.93 0.87
Layer3 (15s) 105.30 105.65 105.48 0.25 0.23
Layer4 (15s) 105.30 105.65 105.48 0.25 0.23
Layer5 (15s) 107.18 106.58 106.88 0.42 0.40
Mean 106.05 106.78 106.42
STD Dev 1.03 1.24 1.13
C.O.V (%) 0.97 1.16 1.07
35 Deg Celscius
Sample1
Sample2
STD
Dev
C.O.V
(%)
% Mod AASHTO % Mod AASHTO
Layer Achieved Achieved Mean
Layer1 (10s) 105.13 104.99 105.06 0.10 0.09
Layer2 (15s) 103.32 103.15 103.23 0.12 0.11
Layer3 (15s) 107.01 102.25 104.63 3.36 3.21
Layer4 (15s) 103.32 103.15 103.23 0.12 0.11
Layer5 (15s) 106.06 103.15 104.61 2.06 1.97
Mean 104.97 103.34 104.15
STD Dev 1.65 1.00 1.32
C.O.V (%) 1.57 0.97 1.27
101

The statistical results of the temperature variation experiment for
the BSM-emulsion show the reverse of the expected trend. The
trend is expected to show that as the material becomes warmer the
compaction will become better, but the opposite of this is reflected
from the results of this experiment. The reason for this is unclear.
The COV values across the samples show that the variability of the
results is very low, COV being in the order of 1 for all three
temperatures. When viewed independently, the 35 0 C experiment
produced the largest amount of variation (COV averaging 1.27%)
and the 5 0 C experiment showed the least amount of variation (COV
averaging 0.82%), but the difference in COV values between the
35 0 C and the 5 0 C experiment is 0.45%, this is extremely small and
may even be ignored. This shows that the temperature at which
samples are prepared does not influence the variability of the
samples, i.e. the variation in the level of compaction across samples
prepared at 5 0 C will not be more or less variable than the variation
in the level of compaction across samples prepared at 35 0 C in terms
of the COV values.
102

4.7.4 Laboratory BSM-foam G2 Material
10kg Surcharge
70% OMC
Sample1 Sample2 Sample3
Time to 100%
Mod AASHTO
(sec)
Time to 100%
Mod AASHTO
(sec)
Time to 100%
Mod AASHTO
(sec)
STD C.O.V
Layer
Mean Dev (%)
Layer1 150 150 180 160.00 17.32 10.83
Layer2 180 150 180 170.00 17.32 10.19
Layer3 180 180 210 190.00 17.32 9.12
Layer4 180 180 150 170.00 17.32 10.19
Layer5 180 180 150 170.00 17.32 10.19
Mean 174.00 168.00 174.00 172.00
STD Dev 13.42 16.43 25.10 18.32
C.O.V (%) 7.71 9.78 14.43 10.64
80% OMC
Sample1 Sample2 Sample3
Time to 100%
Mod AASHTO
(sec)
Time to 100%
Mod AASHTO
(sec)
Time to 100%
Mod AASHTO
(sec)
STD C.O.V
Layer
Mean Dev (%)
Layer1 5.44 6.02 5.81 5.76 0.29 5.10
Layer2 19.67 21.78 21.64 21.03 1.18 5.61
Layer3 37.92 22.73 15.26 25.30 11.55 45.63
Layer4 29.46 38 20.12 29.19 8.94 30.63
Layer5 14.26 37 31.61 27.62 11.88 43.02
Mean 21.35 25.11 18.89 21.78
STD Dev 12.71 13.12 9.42 11.75
C.O.V (%) 59.55 52.26 49.89 53.90
90% OMC
Sample1 Sample2 Sample3
Time to 100%
Mod AASHTO
(sec)
Time to 100%
Mod AASHTO
(sec)
Time to 100%
Mod AASHTO
(sec)
STD C.O.V
Layer
Mean Dev (%)
Layer1 5.62 8 6.25 6.62 1.23 18.62
Layer2 16.63 14.33 21.07 17.34 3.43 19.75
Layer3 14.71 17.77 29.93 20.80 8.05 38.70
Layer4 12.94 21.73 19.1 17.92 4.51 25.17
Layer5 14.15 18.83 18.93 17.30 2.73 15.78
Mean 12.81 16.13 19.06 16.00
STD Dev 4.23 5.26 8.46 5.98
C.O.V (%) 33.05 32.61 44.40 36.69
103

15kg Surcharge
80% OMC
Sample1 Sample2 Sample3
Layer
Time to
100% Mod
AASHTO
(sec)
Time to
100% Mod
AASHTO
(sec)
Time to
100% Mod
AASHTO
(sec) Mean
STD
Dev
C.O.V
(%)
Layer1 8.93 6.92 8.38 8.08 1.04 12.86
Layer2 31.65 21.67 25.45 26.26 5.04 19.19
Layer3 21.42 17.41 16.94 18.59 2.46 13.24
Layer4 27.27 26.57 19.92 24.59 4.06 16.50
Layer5 35.11 19.99 26.16 27.09 7.60 28.07
Mean 24.88 18.51 19.37 20.92
STD Dev 10.28 7.29 7.25 8.27
C.O.V (%) 41.32 39.39 37.41 39.37
The statistical results of the BSM-foam experiments show the
expected trend with regard to the moisture content. Under the 10kg
surcharge, the mean time it takes to compact a layer to 100% Mod
AASHTO density decreases as the moisture content increases. The
70% OMC (Mod-U) has a mean time of 172sec, at 80% OMC (Mod-
U) the mean time is 21.78sec and at 90% OMC (Mod-U) the mean
time is 16sec. From this it is evident that at 70% OMC (Mod-U)
compaction time to 100% Mod AASHTO is too high and therefore
inadequate to be used to produce samples. The mean time for the
80% and 90% OMC (Mod-U) moisture contents are much smaller
and differ by 5.78 sec. The 80% OMC (Mod-U) is the best option in
this case as the extra 5.78sec do not hamper production time and
will provide time to allow for adequate particle orientation during
compaction.
From the results it is also seen that as the moisture content
increases under the 10kg surcharge the Standard deviation also
reduces as des the mean time to 100% Mod AASHTO. As with the
BSM-emulsion experiments the 90% OMC (Mod-U) has the lowest
Standard deviation under the 10kg surcharge. Interestingly, where
the 70% OMC (Mod-U) of the BSM-emulsion under the 10g
surcharge has the highest COV, this is not the case with the BSMfoam.
Instead at 70% OMC (Mod-U) the BSM-foam has the lowest
COV with the 80% OMC (Mod-U) having the highest COV. This
indicates that although the 80% OMC (Mod-U) provides the most
104

adequate compaction time to 100% Mod AASHTO, it is also the
moisture content that provides the most variable results, where,
although the 70% OMC (Mod-U) takes the longest to compact to
100% Mod AASHTO, it also provides the results with the least
variation.
The 15kg surcharge at 80% OMC (Mod-U) has a mean time to
100% Mod AASHTO compaction similar to the 10kg surcharge at
80% OMC (Mod-U); less than 1 second difference between the two
surcharges. The Std dev and COV under this surcharge are however
much lower than the Std dev and COV under the 10kg surcharge.
This means less variability in the compaction results. The results
also indicate that the 15kg surcharge may be used as an alternative
surcharge mass to the 10kg surcharge at 80% OMC (Mod-U); this
may be done to reduce the variability of the compaction results.
105

Temperature Variation: 80% OMC (Mod-U) 10kg Surcharge
5 Deg Celscius
Sample1 Sample2 Sample3
% Mod AASHTO % Mod AASHTO % Mod AASHTO
STD C.O.V
Layer Achieved Achieved Achieved Mean Dev (%)
Layer1 (10s) 96.36 96.98 103.19 98.84 3.78 3.82
Layer2 (28s) 101.18 100.21 103.19 101.53 1.52 1.50
Layer3 (30s) 104.67 103.67 101.46 103.26 1.64 1.59
Layer4 (34s) 105.58 100.21 102.32 102.70 2.70 2.63
Layer5 (30s) 108.41 101.91 104.08 104.80 3.31 3.16
Mean 103.24 100.60 102.85 102.23
STD Dev 4.63 2.48 1.00 2.70
C.O.V (%) 4.49 2.46 0.97 2.64
15 Deg Celscius
Sample1 Sample2 Sample3
% Mod AASHTO % Mod AASHTO % Mod AASHTO
STD C.O.V
Layer Achieved Achieved Achieved Mean Dev (%)
Layer1 (10s) 100.96 105.67 103.32 3.33 3.22
Layer2 (28s) 105.39 102.96 104.18 1.72 1.65
Layer3 (30s) 104.47 102.09 103.28 1.69 1.63
Layer4 (34s) 101.82 102.09 101.95 0.19 0.19
Layer5 (30s) 102.69 104.75 103.72 1.46 1.41
Mean 103.07 103.51 103.29
STD Dev 1.84 1.62 1.73
C.O.V (%) 1.78 1.57 1.68
35 Deg Celscius
Sample1 Sample2 Sample3
% Mod AASHTO % Mod AASHTO % Mod AASHTO
STD C.O.V
Layer Achieved Achieved Achieved Mean Dev (%)
Layer1 (10s) 101.20 100.55 107.85 103.20 4.04 3.92
Layer2 (28s) 106.53 105.84 105.96 106.11 0.37 0.34
Layer3 (30s) 105.60 105.84 105.96 105.80 0.18 0.17
Layer4 (34s) 105.97 104.63 105.96 105.52 0.77 0.73
Layer5 (30s) 103.80 107.73 105.96 105.83 1.97 1.86
Mean 104.62 104.92 106.34 105.29
STD Dev 2.17 2.68 0.85 1.90
C.O.V (%) 2.07 2.56 0.80 1.81
106

The expected compactability trend between the 5 0 C and 35 0 C is
seen, as the 35 0 C experiment had a mean dry density of 105.29%
Mod AASHTO and the 5 0 C experiment had a mean dry density of
102.23% Mod AASHTO across the three samples of both
temperatures. The statistical results show that the influence of the
variation of the 15 0 C experiment from the expected trend was from
Sample 1. At 15 0 C Samples 2 and 3 both produced mean densities
of 103% Mod AASHTO, this is consistent with the expected trend,
but Sample 1 produced a mean layer density of 99.37% of Mod
AASHTO density, this is completely out of the expect compaction
trend therefore another sample should be prepared and compacted
so as to establish the trend.
Looking at the average COV of the three samples, excluding
sample1 of the 15 0 C experiment, for each temperature the
statistical results for the temperature variation experiments of the
BSM-foam indicate that at 15 0 C variability of the compaction results
are the lowest while the temperature that produced the most
variable results is at 5 0 C.
107

4.7.5 COV values of the vibratory hammer for BSM specimens
after determining the actual final mass and Dry Density
The statistical calculations provided previously were all based on
specimens in ideal circumstances, i.e. the calculated mass of
material placed into the mould is the mass of the specimen that is
produced. This is in reality not the case. Excessively wet samples
tended to loose mass as material seeped out between the mould
walls and the vibratory hammer foot piece (tamping foot).
Therefore the final actual mass and Dry Density of each BSM
specimen was noted, the Dry Density was then calculated. The
results for the temperature variation experiments were then
statistically analysed and the COV values for each temperature was
determined. The results are as follows:
Table L14: COV of vibratory hammer
COV* of BSM
Temperature
Emulsion Foam
5 0 C 0.13 0.24
15 0 C 0.39 0.42
35 0 C 0.84 2.34
*COV is of the vibratory hammer compaction
The results show, as the previous statistical analysis showed, that
as the temperature at which compaction takes place increases so to
does the variability of the mixes increase. Generally the COV values
are very low, i.e. they are comfortably below 1. The BSM-foam at
35 0 C however is well above 2, this jump is significant. The reason
for this is that of the three specimens compacted the third
specimen produced a final Dry Density almost 4% higher than the
first 2 specimens.
The results shown in Table L14 also show that the BSM-emulsion
produced specimens with less variability than did the BSM-foam. At
15 0 C however the variability between the two mixes is very similar.
108

4.7.6 Comparing the COV values of the vibratory hammer to the
COV value of the Mod AASHTO compaction method
The COV determined in Table L14, was then compared to the COV
of the Mod AASHTO compaction method. The COV of the Mod
AASHTO was determined by taking G5 material and compacting 4
specimens at the same moisture content. This was done so as to
determine the repeatability of the Mod AASHTO compaction
method. The G5 material was not treated with cold mix bitumen,
only water was added. The COV of the Mod AASHTO is shown
below:
• COV ModAASHTO = 0.45%
When this COV is compared to the vibratory hammer COV values, it
is found that generally the vibratory hammer COV is less than the
Mod AASHTO COV. This is however not the case when the vibratory
hammer was used to compact material warmer than room
temperature (25 0 C).
COV of the vibratory hammer: BSM-emulsion and BSM-foam
COV (%)
2.5
2
1.5
1.12
1
BSM-emulsion
BSM-foam
Poly. (BSM-foam)
Poly. (BSM-emulsion)
0.63
0.5
0
0 5 10 15 20 25 30 35 40
Temperature (Degrees Celcius)
Figure L.76: Projection of COV values of vibratory hammer at room temperature
Figure L.76 shows that the COV of the vibratory hammer at room
temperature is higher than the Mod AASHTO COV. The BMSemulsion
produces a COV of 0.63 which is close to the Mod
AASHTO COV of 0.45. The BSM-foam however produces a COV of
109

1.12. This is significantly higher than the Mod AASHTO COV
although it is still statistically low.
110

4.8 CT Scanning
CT scanning was used to determine the voids condition of the various compacted
samples. The sections shown in this report are of the middle scans of each of the
samples scanned.
Sample 1: S1
Section
150mm
Middle
scan
Middle
scan
Top:
S1B
Middle:
S1A
75mm
75mm
Emulsion BSM-emulsion:
Mix:
Vibratory Hammer Hammer
Figure L.77: CT scan S1B - Middle scan
Figure L.78: CT scan S1A - Middle scan
From Figure L.77 and figure L.78 it appears that the voids content in S1A are in fact higher
than in S1B.
111

S1B
85
E
75
65
Scan slice nummer (boorkern lengte in mm).
55
45
35
25
S
15
5
voids
Mortar
Stone
-5
0 20 40 60 80 100
Volume in %
Figure L.79: CT scan S1B – Graphic Plot, BSM-emulsion, vibratory hammer compaction
From Figure L.79 the voids condition and content may clearly be seen. The
analysis of the figure is taken from the point marked “S” to the point marked “E”
(this is done for each of the CT scan graphic plots). This is done as a result of the
anomalies present at the top and bottom of the specimen, the reason for these
anomalies is that the top and bottom sections of the sample are the sections that
112

were either cut or were the pieces that were the most frequently handled, this
may lead to the loss of aggregate in the sample at these points (this was seen to
happen on the surface of the section being cut). The figure for S1B shows that
the voids content is quite consistent, with only a temporary increase near the top.
The mortar (Emulsion) and stone configuration differ quite extensively. The figure
shows that as the mortar content increases so the stone content decreases and
the opposite is also true.
Figure L.81 shows a different picture; here the voids increase quite consistently till
the middle of the section and then decrease until the end. The reason for this
high increase at the centre of S1A is that the centre of this section falls extremely
close to the intersection of the last two layers of the original sample, therefore
this voids condition may in fact give an indication of the voids condition of the
scarified surface after compaction of the next layer, this reasons may also provide
an explanation for the bulge in S1B, Figure L.80 shows graphically where the
scarified surfaces sit relative to the section points of S1A and S1B (this diagram is
also applicable to the BSM-foam results). When the dimensions provided in the
diagram of Figure L.80 are projected onto the Figures L.79 and L.81, assuming
that the scans were performed from the top end of the sample moving
downwards, the bulge of S1B fits almost perfectly in place with the position of the
scarified surface. The bulge of S1A is slightly out, but this may be attributed to
some form of compaction variability, i.e. that the scarified surface of S1A was
possibly scarified to extensively, leaving a larger area for compaction.
113

The visual of the difference in voids at the centre of S1A and S1B may be seen in
Figures L.77 and L.78. S1A is the middle section of the specimen, so the fact that
the voids show an increase as apposed to a drastic decrease, it may be concluded
that no further compaction of sub layers takes place when compacting the top
layers (in this case S1B). The same would be true should the voids content remain
consistent throughout the sample length. When the mortar and stone content is
viewed, the same trend is seen as is seen in figure L.81.
S1B:
75mm
S1A:
75mm
60mm
15mm
45mm
30mm
60mm*
Scarified
area
60mm
60mm
Scarified
area
* The 60mm
sections are the
original layers of
the sample when it
was initially
compacted
Figure L.80: Graphical illustration of the scarified surface relative to the section points for
the CT scanned samples
114

S1A
85
75
E
65
Scan slice nummer (boorkern lengte in mm).
55
45
35
25
15
voids
Mortar
stone
S
5
0 20 40 60 80 100
-5
Volume in %
Figure L.81: CT scan S1A – Graphic Plot, BSM-emulsion, vibratory hammer compaction
115

Sample 2: S2
Section
150mm
Middle
scan
Piece Sent:
S2
±60mm
Emulsion BSM-emulsion: Mix: Mod Mod
Aashto
AASHTO Compaction Compaction
Figure L.82: CT scan S2 - Middle scan
The CT scan performed on the Mod AASHTO compacted sample (Figure L.83) showed a more
consistent voids content. There is a slight increase in the voids near to the bottom of the sample but
not really significant. The mortar and stone content show the trend again that as the mortar content
increases so to does the stone content decrease. The increase decrease rate is not extensively high,
so that there is a frequent cross over between the two plots, there is however only one point at
which the two plots intersect.
116

s2
85
75
Scan slice nummer (boorkern lengte in mm).
65
E
55
45
35
25
S
15
5
voids
Mortar
stone
0 20 40 60 80 100
-5
Volume in %
Figure L83: CT scan S2 – Graphic Plot, BSM-emulsion, Mod AASHTO compaction
117

Sample 3: S3
Section
150mm
Middle
scan
Middle
scan
Top: S3B
Middle: S3A
75mm
75mm
Foam BSM-foam:
Mix: Vibratory
Vibratory Hammer Hammer
Figure L.84: CT scan - S3B Middle scan, BSMfoam
Figure L.85: CT scan - S3A Middle scan, BSMfoam
The BSM-foam was compacted close to refusal density, while the BSM-emulsion
material was compacted to a target density of 100% Mod AASHTO. Figures L.84
and L.85 show that the top and middle section of the BSM-foam differ less
extensively in their voids content than does the BSM-emulsion.
118

S3B
85
E 75
65
Scan slice nummer (boorkern lengte in mm).
55
45
35
25
S
15
5
voids
Mortar
Stone
-5
0 20 40 60 80 100
Volume in %
Figure L.86: CT scan S3B – Graphic Plot, BSM-foam, vibratory hammer compaction
Figure L84 shows that the voids content of sample S3B varies very little, it is in fact very
consistent. In principle the trend of higher voids content at the points where scarification of
the layers took place should be applicable for the BSM-foam, yet this trend (as was seen in
119

the BSM-emulsion) is not found in sample S3B and S3A. This is due to the fact that the
BSM-foam was compacted for an excessively long period of time (3 minutes per layer) the
scarified layer would have had more time to be compacted and allow the particles to
orientate more thus reducing the voids at these points. The mortar and stone content of
the BSM-foam is also more consistent than in the BSM-emulsion. Figure L.86 shows the
same results as Figure L.87; the only difference is that the mortar and stone content of
Figure L.86 is less consistent near the top of the sample than in Figure L.86.
The voids content in both Figures L.86 and L.87 are consistently low, there is no excessive
decrease in the voids at any one point, and this indicates that no further compaction of the
subsequent layers takes place. This also indicates that no crushing of the aggregate takes
place; this is because at a refusal state the only time that further compaction may transpire
is when the aggregate begins to be crushed.
120

S3A
85
75
65
Scan slice nummer (boorkern lengte in mm).
55
45
35
25
15
voids
Mortar
Stone
5
-5
0 20 40 60 80 100
Volume in %
Figure L.87: CT scan S3A – Graphic Plot, BSM-foam, vibratory hammer compaction
121

4.9 Experimentation on Untreated G2 Material
Experimentation was carried out on untreated G2 material so as to determine
what the compaction trend is when only moisture is added to the material. This
could also provide a possible guide to develop an O.M.C. curve procedure for the
vibratory hammer.
2360.00
Moisture Curve: Untreated G2 Material - Bosch GSH 11E® 10kg
Surcharge
2340.00
2338.00
Dry Density (kg/m3)
2320.00
2300.00
2280.00
2260.00
2240.00
2220.00
5.9
2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0
Moisture Content (%)
Figure L.88: Dry Density vs. Moisture Content: Untreated G2 Material – Bosch GSH 11E ®
The compaction curve show in figure L.88 indicates that the vibratory hammer
compaction is quite sensitive to the moisture content. The climb to the peak at
first sight is quite steep. The gradient of the incline is 77.42; this is high as it is
very close to 100 (which in the case of this graph would be a vertical line). The
decline after the peak moisture content has a gradient of -25, the negative merely
indicates the direction of the line, where as the value of 25 shows that the
gradient is less steep than the initial incline. This indicates that after the peak
moisture content the compaction densities achieved are not as sensitive to the
moisture content as the initial compaction densities. A possible reason for this is
that as the moisture increases the effect of any pore pressure within the sample
would increase, as the moisture content is already high the effect of pore
pressures would become more prevalent, this would have an effect on
compaction, possibly making the sample less sensitive to the compaction. This is
because more of the compaction energy exerted by the hammer would be taken
up by the moisture, transferring less compaction energy to the material. In the
122

case of the incline the moisture is acting more as lubricant, aiding in particle
movement, rather than an energy absorber.
Moisture Curve: Untreated G2 Material - Mod AASHTO vs. Vibratory
Hammer Bosch GSH 11E® 10kg Surcharge
2350.00
2338.00
2300.00
Dry Density (kg/m3)
2260.00
2250.00
2200.00
2150.00
2100.00
Mod AASHTO
Vibratory Hammer
5.9 6.15
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0
Moisture Content (%)
Figure L.89: Dry Density vs. Moisture Content: Untreated G2 Material - Bosch ® Hammer
Compaction vs. Mod AASHTO
The above figure (Figure L.89) shows the comparison between the Mod AASHTO
compacted, untreated G2 material and the Vibratory hammer compacted
untreated G2 material. The first point that is evident is that the compaction
densities achieved by the vibratory hammer are significantly higher than the Mod
AASHTO densities for the same moisture content. Table L.15 shows the densities
achieved at the various moisture contents; the densities are expressed as a
percentage of the Mod AASHTO density. It is seen that compacting layer 1 for 10
seconds and the remaining four layers for 15 seconds each yields densities in
excess of 100% Mod AASHTO density.
123

The difference in the moisture content at the peak of the two curves, the OMC
value, is very small, only 0.25%. This indicates that for the procedure used that
the OMC of the Mod AASHTO and the OMC of the vibratory hammer are very
similar. The curve also shows that at lower moisture contents higher Dry Densities
may be achieved.
Table L.15: %Mod AASHTO Compaction Achieved Using the Vibratory
Hammer – Untreated G2 Material
Sample M.C %OMC Achieved Dry Density % Mod
(kg/m 3 )
AASHTO
1 4.24 68.98 2238.01 102.38
2 5.07 82.51 2269.32 101.67
3 5.73 93.10 2336.3 103.56
4 6.26 101.83 2329.61 103.08
5 7.4 120.27 2302.56 102.79
The average density achieved is 102.7% Mod AASHTO
124

4.10 Repeatability Experimentation: G5 Material
The purpose of the repeatability experimentation was to determine how well the
vibratory hammer compaction procedure developed using a G2 quality material
would perform using a different material; in this case a G5 granular material was
chosen.
4.10.1 Mod AASHTO Compaction of G5 Material
4.10.1.1 Untreated G5 Material
Moisture Curve: Untreated G5, Mod AASHTO
2240.00
2228.00
2220.00
OMC Mod AASHTO
2200.00
Dry Density (kg/m3)
2180.00
2160.00
2140.00
2120.00
2100.00
2080.00
2060.00
6.7
2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0
Moisture Content (%)
Figure L.90: Moisture Curve: Untreated G5 Material – Mod AASHTO
Figure L.90 shows the Mod AASHTO moisture curve used to
determine the OMC of the Untreated G5 material. The Max
Dry Density was found to be 2228 kg/m 3 with an OMC of
6.7%.
125

4.10.1.2 BSM-emulsion G5 Material
Moisture Curve: BSM-emulsion G5 Material, Mod AASHTO
2240.00
2220.00
2217.00
2200.00
OMC Mod AASHTO
Dry Density (kg/m3)
2180.00
2160.00
2140.00
2120.00
2100.00
2080.00
2060.00
6.8
3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0
Moisture Content (%)
Figure L.91: Moisture Curve: BSM-emulsion G5 Material – Mod AASHTO
Moisture Curve: Untreated G5 vs. BSM-emulsion G5 Material
Mod AASHTO
Dry Density (kg/m3)
2240.00
2228.00
2220.00
2217.00
2200.00
2180.00
2160.00
2140.00
2120.00
2100.00
∆OMC = 0.1%
2080.00
2060.00
Untreated G5
BSM-emulsion G5
6.7 6.8
2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0
Moisture (%)
Figure L.92: Moisture curve: Untreated G5 vs. BSM-emulsion G5 Material– Mod AASHTO
126

Figure L.91 shows the Mod AASHTO curve determined for
the BSM-emulsion G5 material; this curve was plotted
against the untreated G5 moisture curve and is shown in
Figure L.92. The results of comparing the two curves shows
that the OMC value of the BSM-emulsion was 0.1% higher
than for the untreated G5 material. The max Dry Density of
the untreated G5 material was however 11kg/m 3 higher
than for the BSM-emulsion G5 material. It is also seen from
the result of comparing the two curves that the gradient of
BSM-emulsion is steeper when compared to the untreated
material. The same point may be noted, with regard to the
gradient steepness of the two mixes, in the G2 material.
This indicates that the compaction of the BSM-emulsion is
more sensitive to moisture than the untreated material.
4.10.1.3 BSM-foam G5 Material
Moisture Curve: BSM-foam G5 Material, Mod AASHTO
2160.00
OMC Mod AASHTO
2150.00
2149.50
Dry Density (kg/m3)
2140.00
2130.00
2120.00
2110.00
2100.00
6.95
3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0
Moisture Content (%)
Figure L.93: Moisture Curve: BSM-foam G5 Material – Mod AASHTO
127

Moisture Curve: Untreated G5 vs. BSM-foam G5 Material
Mod AASHTO
Dry Density (kg/m3)
2240.00
2228.00
2220.00
2200.00
2180.00
2160.00
2149.50
2140.00
2120.00
2100.00
2080.00
∆ OMC = 0.25
Untreated G5
BSM-foam G5
2060.00
6.70 6.95
3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 12.00
Moisture (%)
Figure L.94: Moisture curve: Untreated G5 vs. BSM-foam G5 Material – Mod AASHTO
The same comparison, as to what was done for the BSMemulsion,
of the treated material to the untreated material
was done for the BSM-foam. The OMC of the BSM-foam for
the Mod AASHTO moisture curve was found to be 0.25%
higher than for the OMC of the untreated G5 Mod AASHTO
moisture curve (Figure L.94). The Maximum Dry Density of
the BSM-foam was however 78.5kg/m 3 less than the
untreated material, and when looking at the gradients, the
BSM-foam material has a slightly flatter incline compared to
the untreated material, but the decline of the moisture
curve is steeper in the case of the BSM-foam. This gradient
indicates that the BSM-foam is less sensitive to moisture
while moving up to the OMC point than the untreated
material, but after reaching OMC the decrease in the Dry
Density is more rapid and therefore more sensitive to the
moisture in the case of the BSM-foam.
128

4.10.2 Vibratory Hammer Compaction of G5 Material
4.10.2.1 Untreated G5 Material
Moisture Curve: Untreated G5 Bosch GSH 11E® 10kg Surcharge
2250.00
2240.00
2236.00
2230.00
Dry Density (kg/m3)
2220.00
2210.00
2200.00
2190.00
2180.00
2170.00
2160.00
2150.00
6.0
3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0
Moisture Content (%)
Figure L.95: Moisture curve: Untreated G5 Vibratory Hammer – Bosch GSH 11E®
2260.00
2240.00
2236.00
2228.00
2220.00
Moisture Curve Mod AASHTO vs. Vibratory Hammer Bosch GSH 11E®
10kg Surcharge: Untreated G5 Material
Dry Density (kg/m3)
2200.00
2180.00
2160.00
2140.00
2120.00
2100.00
2080.00
OMC Mod AASHTO
OMC Vibratory Hammer
2060.00
6.0 6.7
2.0 4.0 6.0 8.0 10.0 12.0
Moisture Content (%)
Figure L.96: Moisture Curve: Mod AASHTO Compaction vs. Vibratory Hammer Compaction –
Untreated G5 Material
129

Table L.16: %Mod AASHTO Compaction Achieved: Mod AASHTO vs.
Vibratory Hammer – Untreated G5 Material
Sample M.C %OMC Achieved Dry Density
(kg/m 3 )
% Mod
AASHTO
1 5.30 79.06 2161.00 100.05
2 5.91 88.26 2238.45 102.12
3 6.69 99.87 2234.07 100.27
4 8.09 120.69 2228.15 100.64
The average achieved Dry Density = 100.77% Mod AASHTO
The vibratory hammer compaction of the untreated G5
material was done using the same procedure developed for
the BSM-emulsion. This was decided based on the fact that
the emulsion is believed to act as a lubricant during the
compaction of BSM-emulsion, and in the case of untreated
material there is only lubrication being added to the
material in the form of water. Therefore layer one was
compacted for 10 seconds and the remaining 4 layer were
each compacted for 15 seconds.
The results of the experiment are presented graphically in
Figures L.94 and L.95. It is seen from these results that the
vibratory hammer compaction is more sensitive to moisture
on the incline as apposed to the Mod AASHTO compaction.
The decline shows a drop in moisture sensitivity for the
vibratory hammer, the Mod AASHTO compaction however
shows a higher level of sensitivity to the moisture content
after OMC than does the vibratory hammer. Across the
various moisture contents the Dry Densities achieved were
consistently around 100% Mod AASHTO (averaging
100.77%). On site the accepted level of compaction of a G5
material is 95 percent Mod AASHTO, considering that across
the various moisture contents an average Mod AASHTO Dry
Density of 100.77% is being achieved with the vibratory
hammer, it shows that the densities achieved with the
procedure are acceptable with respect to accepted levels of
site compaction for the G5 material. In the case of the G2
material the Dry Densities were on average around 103%
Mod AASHTO. The peak moisture content of the untreated
130

G5 material using the procedure was found to be 6% which
indicates that the OMC of the vibratory hammer may in fact
be lower than that of the Mod AASHTO compaction (which
in this case is 6.7%). The OMC results indicate that at a
lower OMC for a given material (in this case a G5 material)
the vibratory hammer compaction can produce Dry
Densities similar to or higher that the Mod AASHTO Dry
Densities.
The result of the achieved Dry Densities of the G5 material
indicates that the procedure is useable on lesser quality
granular materials such as G5 material.
131

4.10.2.2 BSM-emulsion G5 Material
2220.00
Moisture Curve: BSM-emulsion G5 Material, Bosch GSH 11E® 10kg
Surcharge
OMC Vibratory Hammer
2202.00
2200.00
Dry Density (kg/m3)
2180.00
2160.00
2140.00
2120.00
2100.00
7.0
4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0
Moisture Content (%)
Figure L.97: Moisture curve: BSM-emulsion G5 Material, Vibratory Hammer – Bosch GSH 11E®
2240.00
2220.00
2217.00
2202.00
2200.00
Moisture Curve Mod AASHTO vs. Vibratory Hammer Bosch GSH 11E®
10kg Surcharge: BSM-emulsion G5 Material
OMC Vibratory Hammer
OMC Mod AASHTO
Dry Density (kg/m3)
2180.00
2160.00
2140.00
2120.00
2100.00
2080.00
2060.00
6.8 6.95
4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0
Moisture Content (%)
Figure L.98: Moisture curve: BSM-emulsion G5 Material, Mod AASHTO vs. Vibratory Hammer –
Bosch GSH 11E®
132

Table L.17: %Mod AASHTO Compaction Achieved Using the Vibratory
Hammer – BSM-emulsion G5 Material
Sample M.C %OMC Achieved Dry Density (kg/m 3 ) % Mod
AASHTO
1 4.91 75.56 2106.52 101.47
2 5.53 85.06 2167.29 102.13
3 6.42 98.77 2186.11 99.82
4 7.01 107.83 2201.12 99.69
5 7.41 114.04 2197.18 101.07
Average Achieved Dry Density = 100.84% Mod AASHTO
G5 material was treated using Bitumen Emulsion and the
Dry Density vs. Moisture content curve was developed for
this mix using the vibratory hammer. For this compaction
the same procedure developed for the BSM-emulsion for
the G2 material was applied; figures L.96 and L.97 show
the results of the experiment.
The experiment shows that at the peak moisture content
the Dry Density achieved with the vibratory hammer is less
than the Mod AASHTO Dry Density. The peak moisture
content i.e. the OMC of both the vibratory hammer and of
the Mod AASHTO is very similar, only differing by 0.15%.
The criterion for the accepted level of compaction on site is
however still met, site compaction being accepted at 95%
Mod AASHTO and the vibratory hammer achieved a
compaction level of 99.82% Mod AASHTO at its OMC. It
should be noted that from Figure L.99 it is seen that there
is a moisture content bracket where the Dry Density
achieved using the vibratory hammer does not reach 100%
Mod AASHTO (it reached between 99.5% and 100%) but
outside of the bracket Dry Densities exceeding 100% Mod
AASHTO are achieved.
The BSM-emulsion experiment also showed that the G5
material was less sensitive to moisture overall when
compacting using the vibratory hammer as apposed to the
133

Mod AASHTO; this is shown by virtue that the vibratory
hammer curve is quite flat when compared to the Mod
AASHTO cure (figure L.97).
Comparing the result of the G5 BSM-emulsion to the G2
BSM-emulsion (sub section 4.5.1) a potential trend may be
noted, i.e. that as the material quality decreases (e.g. from
a G2 to a G5 material) so to does the extent with which
100% Mod AASHTO compaction is exceeded using the
vibratory hammer. This indicates that the material quality
influences the level of compaction.
The repeatability experiment for the BSM-emulsion shows
that the procedure is applicable to and can produce
acceptable result across material quality which varies as far
as from G2 to G5.
134

4.10.2.3 BSM-foam, G5 Material
Moisture Curve: BSM-foam G5 Material, Bosch GSH 11E® 10kg
Surcharge
2170.00
OMC Vibratory Hammer
2160.00
Dry Density (kg/m3)
2150.00
2140.00
2130.00
2120.00
2110.00
2100.00
6.85
3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0
Moisture Content (%)
Figure L.99: Moisture curve: BSM-foam Vibratory Hammer – Bosch GSH 11E®
Moisture Curve Mod AASHTO vs. Vibratory Hammer Bosch GSH 11E®
10kg Surcharge: BSM-foam G5 Material
2170.00
2160.00
OMC Vibratory Hammer
OMC Mod AASHTO
Dry Density (kg/m3)
2150.00
2149.50
2140.00
2130.00
2120.00
2110.00
2100.00
6.85 6.95
3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0
Moisture Content (%)
Figure L.100: Moisture curve: BSM-foam G5 Material, Mod AASHTO vs. Vibratory Hammer –
Bosch GSH 11E®
135

Table L.18: %Mod AASHTO Compaction Achieved Using the Vibratory
Hammer – BSM-foam Treated G5 Material
Sample M.C %OMC Achieved Dry Density (kg/m 3 ) % Mod
AASHTO
1 4.93 73.59 2133.19 102.66
2 6.26 93.40 2151.56 101.15
3 7.07 105.52 2158.96 100.46
4 7.94 118.54 2151.34 101.62
Average Achieved Dry Density = 101.47% Mod AASHTO
The final repeatability experiment was setting up the
moisture content vs. Dry Density curve for G5 material that
had been treated using Foamed bitumen. The results of the
experiment are presented in figures L.98 and L.99 as well
as in table L.15.
The experiment procedure used was similar to the
procedure developed for the G2 BSM-foam. Compaction
was done using the compaction times assigned to perform
the moisture content vs. Dry Density curve of the G2 BSMfoam,
these times are as follows:
Layer 1 = 10 seconds
Layer 2 = 25 seconds
Layer 3 = 25 seconds
Layer 4 = 34 seconds
Layer 5 = 25 seconds
The results of the BSM-foam showed, similar to the BSMemulsion
results, that the compaction density achieved
using the vibratory hammer is less sensitive to moisture
than what the Mod AASHTO compaction is; this is evident
by the lower incline and decline in the moisture content vs.
Dry Density curve of the vibratory hammer compaction.
The Dry Densities achieved averaged out at 101.47% Mod
AASHTO with all moisture contents exceeding the 100%
Mod AASHTO mark. This result is good in that it indicates
that the compaction procedure is capable of producing Dry
Densities well above the accepted level of site compaction;
136

95% Mod AASHTO for G5 material. The procedure may
therefore be used for lesser quality granular materials,
which have been treated using Foamed mix bitumen, to
produce samples so as to give an indication of the level of
compaction achievable on site.
In the Moisture vs. Dry Density curve of the BSM-foam G5
material, the sample treated for 80% OMC was ignored
when setting up the graph. When this ample was
compacted the point at which it fell on the graph was
outside of the other four (4) points. The grading curve in
Figure L.101 was constructed (dry grading) and the
possible explanation for this outlier is show in the en-circled
area. The 80% OMC sample and the 100% OMC sample
were both broken up after compaction and allowed to dry
out completely prior to doing the grading curve. The 100%
OMC was chosen as it lay on the plotted curve. Because the
remaining four samples all fell on the plotted curve it may
be assumed (considering that they are reconstituted
samples) that their grading is virtually the same. The encircled
section of figure L.101 shows a discrepancy in the
grading. When this was evaluated and the results converted
to mass of material it was found that the 80% OMC sample
had ±1100gm of material for the sieve fractions 13.2mm to
2mm more than the 100% OMC sample. Looking at the
fines, it was found that the 100% OMC sample had
±500gm for the sieve fractions 0.0425mm to 0.075mm
more than the 80% OMC sample.
137

Grading Curve: Foamed Mix G5 Material - 80% OMC vs. 100% OMC
120.0
100.0
80% OMC
100% OMC
80.0
% Passing
60.0
40.0
20.0
0.0
0.01 0.1 1 10 100
Sieve Size (mm)
Figure L.101: Grading curve: BSM-foam G5 Material – 80% OMC vs. 100% OMC
The difference in the material grading of the two samples
may be attributed to the segregation of the 80% OMC
sample during the weighing off of the various layers. For
the 100% OMC sample the moisture content and weighing
off of the various layers made use of almost all of the
material, after which there was in the order of 600gm
material left over where with the 80% OMC there was well
in excess of 1000gm of material left over after weighing off
of separate layers and moisture content samples being
taken, this large excess indicates that there is larger room
for segregation of the sample than for the 80% OMC
sample.
138

4.10.3 Grading of G5 material Before and After Compaction
90.0
80.0
Grading Curve of G5 Material: Grading Prior to Compaction vs. Grading
After Compaction
Grading Prior to Compaction
Grading After Compaction
70.0
60.0
% Passing
50.0
40.0
30.0
20.0
10.0
0.0
0.01 0.1 1 10 100
Sieve Size (mm)
Figure L.102: G5 Material Grading Curve – Before and After Compaction
The grading curve provided in figure L.102 was developed using the
dry grading method on untreated G5 material. The grading curves
of the material indicate that after compaction the overall grading of
the sample is different. The result does not appear to be as a result
of crushing, this is because there is not a distinctive increase in the
finer material. The reason may rather be due the segregation that
occurs during the weighing off process, or it could be due to the
loss of fine material through the seepage of material between the
walls of the mould and the foot piece of the vibratory hammer
when samples are compacted at higher moisture levels.
139

4.11 Zero Air Voids of the Curves
The Zero air voids curve was plotted against the moisture sensitivity curve
of the various mixes. This was done so as to establish were the compaction
curves lie with respect to the absolute maximum Dry Density, indicated by
the zero air voids curve.
To set up these curves the specific gravity (Gs) of both the G2 and G5 material
was determined in both the untreated and BSM states; the Gs of the bitumen
was taken as 1.00 and for both the BSM-emulsion and BSM-foam the binder
content was taken as 1.98%. The Gs values of the fine material were
determined using the TMH 1 method in the case of the G5 material. In the
case of the coarse material i.e. the large stones, the same method was used as
was used for the G2 material. A few of the largest stones (± 19mm) were
taken and the individual density of each one was determined, these were
averaged out and used as the representative density of the parent rock from
which the Gs value may be obtained; dividing this density with the density of
water in the case of the G2 material; for the G5 material the relative density
(RD) of the fine material was summed with the representative density of the
parent rock and divided by two, this gave the density from which the Gs value
was determined. The Gs values determined were as follows:
• Gs of the Untreated G2 = 2.873
• Gs of BSM G2 = 2.836
• Gs of the Untreated G5 = 2.788
• Gs of BSM G5 = 2.733
Figures L.103 to L.108 show the relationship of the moisture curve to the zero air
voids line.
140

4.11.1 Zero air voids vs. Moisture curve of the G2 material
Dry Density
(kg/m3)
Zero Air Voids Line vs. Moisture Curve of
Mod AASHTO
and Bosch GSH 11E®: Untreated G2
Material
2400.00
2350.00
2300.00
2250.00
2200.00
2150.00
2100.00
2050.00
Vibratory hammer
Mod AASHTO
Zero Air Voids
0.0 2.0 4.0 6.0 8.0 10.0 12.0
Moisture Content (%)
Figure L.103: Zero air voids line vs. Moisture curve – Untreated G2 Material
Dry Density
(kg/m3)
Zero Air Voids Line vs. Moisture Curve of
Mod AASHTO
and Bosch GSH 11E®: BSM-emulsion G2
Material
2350.00
2300.00
2250.00
2200.00
2150.00
2100.00
2050.00
2000.00
Vibratory hammer
Mod AASHTO
Zero Air Voids
0.0 2.0 4.0 6.0 8.0 10.0 12.0
Moisture Content (%)
Figure L.104: Zero air voids line vs. Moisture curve – BSM-emulsion G2 Material
141

Dry Density
(kg/m3)
2350.00
2300.00
2250.00
2200.00
2150.00
2100.00
2050.00
2000.00
Zero Air Voids Line vs. Moisture Curve of
Mod AASHTO
and Bosch GSH 11E®: BSM-foam G2
Material
Vibratory hammer
Mod AASHTO
Zero Air Voids
0.0 2.0 4.0 6.0 8.0 10.0 12.0
Moisture Content (%)
Figure L.105: Zero air voids line vs. Moisture curve - BSM-foam G2 Material
The zero air voids curve plotted for the G2 material show that the
untreated material finishes closer to the zero air voids line than dos
the BSMs. The expected result of higher compaction energy is seen
in the curves as the vibratory hammer curve shifts upward slightly
to the left, moving closer to the zero air voids line, with the
exception of the BSM-emulsion plot (figure L.104) which shifts to
the right however the movement is closer to the zero air voids line.
In view of these results, the vibratory hammer produces samples
with lower air voids than does the Mod AASHTO compaction
method.
142

4.11.2 Zero air voids vs. Moisture curve of the G5 material
Dry Density
(kg/m3)
Zero Air Voids Line vs. Moisture Curve of
Mod AASHTO
and Bosch GSH 11E®: Untreated G5
Material
2400.00
2350.00
2300.00
2250.00
2200.00
2150.00
2100.00
2050.00
2000.00
1950.00
Vibratory hammer
Mod AASHTO
Zero Air Voids
0.0 2.0 4.0 6.0 8.0 10.0 12.0
Moisture Content (%)
Figure L.106: Zero air voids line vs. Moisture curve – Untreated G5 Material
Dry Density
(kg/m3)
Zero Air Voids Line vs. Moisture Curve of
Mod AASHTO
and Bosch GSH 11E®: BSM-emulsion G5
Material
2350.00
2300.00
2250.00
2200.00
2150.00
2100.00
2050.00
Vibratory hammer
Mod AASHTO
Zero Air Voids
0.0 2.0 4.0 6.0 8.0 10.0 12.0
Moisture Content (%)
Figure L.107: Zero air voids line vs. Moisture curve – BSM-emulsion G5 Material
143

Dry Density
(kg/m3)
Zero Air Voids Line vs. Moisture Curve of
Mod AASHTO
and Bosch GSH 11E®: BSM-foam G5
Material
2350.00
2300.00
2250.00
2200.00
2150.00
2100.00
2050.00
Vibratory hammer
Mod AASHTO
Zero Air Voids
0.0 2.0 4.0 6.0 8.0 10.0 12.0
Moisture Content (%)
Figure L.108: Zero air voids line vs. Moisture curve – BSM-foam G5 Material
The results seen in the G2 material are evident in the G5 material
results. The G5 material showed more consistent results in terms of
how close to the zero air voids line the compacted samples finished.
The moisture curve of the untreated material did not finish
excessively closer to the zero air voids line than did the BSMs; on
the contrary when the position of the MDD of the vibratory hammer
of the G5 is viewed relative to the zero air voids line then it is seen
that the BSM-emulsion curve (figure L.107) is closer to the zero air
voids line. As with the G5 material the expected trend of the
moisture curve moving upward slightly to the left for higher
compaction energy is seen; however for the G5 material the BSMemulsion
conforms to this trend. In view of these curves the
vibratory again shows that it produces samples with lower air voids
than does the Mod AASHTO compaction method.
144

4.11.3 Zero air voids curve compared to CT scanning results
The comparison of the CT scanning results could only be done for
the BSM-emulsion of the G2 material as there was both a Mod
AASHTO compacted sample and a vibratory hammer compacted
sample of the material. The voids of Figure L.79 were used for the
vibratory hammer and an area on the sample where the voids were
at a minimum was chosen, therefore the voids from slice 20 to slice
54 were average out and used as this gave a comparable result,
(the scarification of the layer surface had not been identified prior
to the CT scanning, therefore the voids of Figure L.81 were not
used for this comparison). The voids of figure L.83 were used for
the Mod AASHTO compaction and the section of the sample where
the voids were at a minimum was chosen and the voids across that
section were averaged out; slice 8 to slice 62. The voids results are
as follows:
• CT scanning = 0.68% voids
• Mod AASHTO = 0.77% voids
These results show that the vibratory hammer produced lower air
voids content than did the Mod AASHTO, this confirms the results of
the zero air voids line, which showed that the vibratory hammer
produced samples with a lower air voids content.
145

4.12 Compaction Energy of the Compaction Methods
4.12.1 Bosch GSH 11E®
The compaction energy exerted by the Bosch® vibratory hammer over the
BSM-emulsion and BSM-foam samples as well as the untreated granular
material was determined after completion of the compaction experiments.
This energy was then compared to the compaction energy of the New
Zealand vibrating hammer compaction method and to the energy of the
ASTM vibrating hammer compaction method (see table L.Lit 3 and L.Lit 4 of
the literature study section of this appendix). The New Zealand and ASTM
compaction methods compact in two and three layers respectively to
produce samples of a height similar to the Mod AASHTO compaction
method. Therefore to compare the energies the new Zealand and ASTM
methods were evaluated over five layers; this allowed both methods to be
compared to the procedure developed by the University of Stellenbosch
(US).
The calculation of the energy of each hammer was done as follows.
n=5
• Total Energy = Σ ( Point Energy × (beats/min) × compaction
x=1
time of Layer x ÷ 60seconds)
• Average energy per Layer = Total Energy ÷ No. of Layers
The energy of the Bosch® hammer for the BSM-emulsion was determined
for two time scenarios:
1. Times assigned to the layers based on the experiment results.
2. Doubling the compaction times assigned to the layers based on the
experiment results.
The energies are assigned a lower limit and an upper limit. This is the range
in which the energies fall based on the minimum and maximum beats per
minutes of the specific vibratory hammer.
The point energy of the respective compaction methods are provided below
and the comparative results then follow.
• New Zealand = 7 Joules: 1500-3600 beats/min
146

• ASTM = 10 Joules (Bosch 11248 EVS ®): 1700-3300 beats/min
• US = 25 Joules (Bosch GSH 11E ®): 900-1890 beats/min
Compaction Energy over the BSM-emulsion samples: Times initially
assigned to each layer
Table L.19: Comparison of Total Energy of Vibratory Hammers BSMemulsion
Vibratory Hammer Lower Limit (kJ) Upper Limit (kJ)
New Zealand 157.5 378
ASTM: Bosch
11248EVS®
85 165
US: Bosch GSH 11E® 26.25 55.13
Table L.20: Comparison of Average Energy/Layer of Vibratory
Hammers BSM-emulsion
Vibratory Hammer Lower Limit (kJ) Upper Limit (kJ)
New Zealand 31.5 75.6
ASTM: Bosch
11248EVS®
17 33
US: Bosch GSH 11E® 5.25 11.025
The New Zealand and ASTM compaction methods make use of fixed
times per layer of 180 seconds and 60 seconds respectively. The
comparisons presented above assume that there compaction methods
are used as is to compact BSM samples. The untreated material at
Stellenbosch University was compacted using the procedure developed
for the BSM-emulsion samples. The compaction times for the US
procedure are as follows
• Layer 1 = 10seconds
• Layer 2 through to Layer 5 = 15 seconds each
Based on the given point energies the trend that is seen is that a higher
point energy yields a less total energy required to compact a sample. This
trend was noted in the literature study of this appendix and the compaction
procedure developed by the University of Stellenbosch confirms this point as
the Bosch GSH 11E® has the highest point energy (25 Joules) but uses the
least total energy and least average energy per layer to compact the
untreated and BSM-emulsion samples.
147

Compaction Energy over the BSM-emulsion samples: Times initially
assigned to each layer Doubled
Table L.21: Comparison of Total Energy – Compaction time per layer
doubled BSM-emulsion
Vibratory Hammer Lower Limit (kJ) Upper Limit (kJ)
New Zealand 157.5 378
ASTM: Bosch
11248EVS®
85 165
US: Bosch GSH 11E® 52.5 110.25
Table L.22: Comparison of Average Energy/Layer – Compaction
time per layer doubled BSM-emulsion
Vibratory Hammer Lower Limit (kJ) Upper Limit (kJ)
New Zealand 31.5 75.6
ASTM: Bosch
11248EVS® 17 33
US: Bosch GSH 11E® 10.5 22.05
The trend seen in the energies prior to doubling the compaction times
for the US procedure are still seen after doubling the compaction times.
It is expected that doubling the compaction times of the US procedure
will produce samples with a much higher final Dry Density which will
also be further above the upper limit of the accepted level of
compaction specified for site.
148

Compaction Energy over the BSM-foam samples: Times initially
assigned to each layer
Table L.23: Comparison of Total Energy of Vibratory Hammers BSMfoam
Vibratory Hammer Lower Limit (kJ) Upper Limit (kJ)
New Zealand 157.5 378
ASTM: Bosch
11248EVS®
85 165
US: Bosch GSH 11E® 44.63 93.71
Table L.24: Comparison of Average Energy/Layer of Vibratory
Hammers BSM-foam
Vibratory Hammer Lower Limit (kJ) Upper Limit (kJ)
New Zealand 31.5 75.6
ASTM: Bosch
11248EVS® 17 33
US: Bosch GSH 11E® 8.93 18.74
The BSM-foam samples show the same trend as the BSM-emulsion samples
in both the total energy required to compact a sample and the average
compaction energy per layer.
There was no need to double the compaction energies for the BSM-foam
samples as the samples produced from this procedure have final Dry
Densities that are well exceed the accepted level of compaction on site.
149

4.12.2 Mod AASHTO Compaction
The compaction energy of the Mod AASHTO compaction method was
determined from the formula:
• Compaction Energy (CE) = (No. of blows/layer) x (No. of layers) x
(weight of hammer) x (drop height of hammer)
In order to compare the overall energy of the Mod AASHTO method to the
vibratory hammer method the number of layer was taken as 5, the number
of blows per layer was taken as 55, the drop height was taken as 0.4572m
and the weight of the hammer was 4.536kg x g (gravitational acceleration
= 9.81m.s -2 ). The compaction energy is provided below:
• CE/Layer = 1.865 kJ
• CE Total = 5.595 kJ
This results, when compare to the Bosch GSH 11E® compaction energy, is
much lower than the vibratory hammer compaction energy. The Bosch®
hammer at the University of Stellenbosch exerted a CE/Layer of 11.025kJ
for the BSM-emulsion and 18.74kJ for the BSM-foam. The vibratory hammer
exerted energy per layer greater than the total energy of the Mod AASHTO
compaction method. The total energy of the vibratory hammer of 55.13KJ
and 93.71KJ for BSM-emulsion and BSM-foam respectively are vastly
greater that the total energy of 5.595KJ for the Mod AASHTO compaction
method.
150

4.13 Development of Material Tools
During the experiment process spot measurements were made of the layer
thickness just prior to compaction, i.e. directly after the material for the
respective layer had been added and the hammer had be lowered into
position, measurements were taken as to determine the starting thickness
of the layer. These measurements were taken by measuring the distance
from the lower end of the sleeve to either the zero line, in the case of layer
one, or to the final position of the previously compacted layer; the
measurements showed a starting height of around 92mm in the case of the
G5 material. These measurements were used to design a scoop. This was
done by determining the volume of the material prior to compaction and
then developing a scoop with either the same volume as the starting
volume of an individual layer or with a fraction of the starting layer of ½ or
1/3 of the volume. This was done so that as apposed to weighing of the
mass of material per layer a scoop could be used to place material into the
mould that would provide a similar mass as to what would have been
weighed off.
The result was a scoop with a volume of 1/3 the starting volume of a layer.
The dimensions are a diameter of 90mm and a height of 85mm. This design
allows for three scoops of aggregate to be added into the mould that will
provide a starting height of ±92mm.
151

4.14 Development of Specifications for Site Compaction
The vibratory hammer compaction method is to be used to provide specifications for
compaction levels on site. In order to develop the specifications the vibratory
hammer results were equated to the current specifications. This was done as
follows:
Material Type Current specification (%Mod AASHTO)
G2 100-102%
G5 95%
The average level of compaction achieved during the sensitivity analysis of the
vibratory hammer was determined for each of the individual BSMs. The standard
deviation for each BSM was also calculated. The Design Value i.e. the level of
compaction from which the specifications will be developed was calculated using the
following formula:
DesignValue = X − zS
Previously Ф was taken as 15% (Subsection 4.7.1) therefore z = 1.033
X is the average level of compaction of the vibratory hammer as a % of Mod
AASHTO
z is the dimensionless factor
S is the standard deviation
The results of this calculation were as follows:
Table L25: Design values for site compaction specifications of G2 material
BSM Average compaction level
(% Mod AASHTO)*
Standard
Deviation
Design Value
(85%
reliability)
Untreated 102.70 0.72 101.96
Emulsion 102.86 1.42 101.40
Foam 105.47 1.36 104.06
*After vibratory hammer compaction
152

Table L26: Design values for site compaction specifications of G5 material
BSM Average compaction level (%
Mod AASHTO)*
Standard
Deviation
Design Value
(85%
reliability)
Untreated 100.77 0.91 99.58
Emulsion 100.84 1.06 99.74
Foam 101.48 0.92 100.52
*After vibratory hammer compaction
The current accepted level of compaction is then divided by the Design Value to
obtain the level of compaction for the vibratory hammer. This level of compaction
for the vibratory hammer is equivalent to the current specifications. The
specifications determined in this report do need to be revisited as site compaction
data becomes available. The final specifications are as follows:
Table L27: Site compaction specifications for G2 material
BSM Site Specification (% vibratory
hammer compaction)
Untreated 98.6 - 100
Emulsion 98.6 - 100
Foam 96.1 - 98
The G5 material produced Design Values all in the order of 100% Mod AASHTO for
vibratory hammer compaction. Therefore the fraction of Mod AASHTO compaction
specified for the level of site compaction is to be used. An example of this is as
follows: for a G5 material the current specification for site compaction is 95% of
Mod AASHTO compaction. Therefore according to the previous statement the
accepted level of site compaction is as specified below.
Table L28: Site compaction specifications for G5 material
BSM Site Specification (% vibratory
hammer compaction)
Untreated 95
Emulsion 95
Foam 95
153

4.15 Outline of Proposed Protocol
From the results a potential compaction protocol may be identified. The protocol
is for the production of samples that will give an indication of what Dry Densities
may be obtained in the field i.e. using the vibratory hammer compaction as a
means to indicate what densities are possible on site by means of a sensitivity
analysis. A full, detailed description of the compaction protocol is provided in the
recommendations section of this appendix. A brief outline is described below.
Protocol: Sample production as an indication of achievable compaction
levels
For the compaction protocol the following information is needed:
1. Mod AASHTO moisture content curve
2. OMC (Mod-U)
3. Target Binder content if material is to be stabilized
4. Type of cold mix treatment i.e. bitumen emulsion mix or foamed bitumen
mix
The following tools and temperature information is also needed
5. Bosch GSH 11E® with a surcharge of 10kg
6. Compaction is done at room temperature
7. Drill with a drill bit longer than or equal to 300mm with a point marked off
10mm from the tip of the drill bit.
8. Extension piece for the mould so as to compact layer 5
9. Material scoop.
A Moisture curve for the untreated material ifs first developed, from this moisture
curve the OMC of the untreated material is known and this OMC is used to
develop a moisture curve for either the BSM-emulsion or the BSM-foam; which
ever stabilizing method is to be used. These graphs are the reference graphs for
the target Dry Densities.
Compaction using the vibratory hammer is done at various moisture contents
ranging from 2% moisture to 10 or 12% moisture; this depends on the type of
aggregate used.
154

The amount of moisture added to the material will depend on the type of cold mix
treatment. For the BSM-foam moisture is added to the samples till the sample is
reached which has a ratio of the water added to OMC of the untreated material of
0.8 (80%). At this point the remaining samples receive the same amount of water
as the sample with a ratio of water added to OMC of the untreated material of
0.8. The material is then treated using the foamed bitumen and the remaining
moisture required to achieve the targeted moisture content of each of the
samples is added. The BSM-emulsion samples must first have the moisture in the
emulsion calculated out of the targeted moisture content, the moisture is then
added followed by the bitumen emulsion.
The material is prepared; either BSM-emulsion or BSM-foam. The sample is then
compacted in 5 layers in a mould 300mm high with a diameter of 150mm.
Material is placed In the mould using a material scoop. Each individual layer
receives three scoops of material for the respective mix, this is placed in the
mould using a scoop with dimensions of 90mm diameter and height of 85mm and
the sample is compacted. Each layer is compacted for a set period of time until all
five layers have been compacted. Times for the respective layers are provided in
the recommendations section of the appendix.
After a layer has been compacted, that layer surface is then scarified using the
drill. Three scoops of material are then added using the material scoop and the
layer is compacted, this process continues until all five layers have been
compacted.
The sample is removed and the final height and final mass of the sample are
taken. A moisture content of the remaining material is also taken using the
standard oven drying method and the Dry Density is then determined. The Dry
Density for each of the samples is at their respective moisture contents is
determined. These are then used to plot the moisture curve of moisture Content
vs. Dry Density of the vibratory hammer. This curve is then compared to the Mod
AASHTO moisture curve set up for the respective BSM. The comparison is
expressed as a percentage of the Mod AASHTO Dry Density.
The results show that the Dry Densities achieved are above the recommended
levels of compaction for G2 and G5 materials. This shows that the vibratory
hammer may be used to specify the level of site compaction. Based on the results
of this research the level of compaction should be taken as 100% of the Maximum
Dry Density achieved using the vibratory hammer for the G2 material. In the case
of the G5 material the Maximum Dry Densities of the vibratory hammer
155

compaction was found to be extremely near to the Maximum Dry Densities of the
Mod AASHTO compaction, therefore the specification of 95% Mod AASHTO
compaction currently used to specify the level of compaction on site should be
applied to the vibratory hammer to specify the level of site compaction; i.e. site
compaction should be equal to or larger than 95% vibratory hammer compaction.
156

5 CONCLUSIONS
Based on the results of the experiments the following conclusions were drawn.
5.1 Influence of Time on Compaction with Varying Moisture Content of G2 Material
The experiments showed that as the moisture content increased the compaction time to 100%
Mod AASHTO compaction decreased. What was also seen is that once the moisture content was
at 90% OMC for the BSM-emulsion the compaction time reduced so far that it may be assumed
that sufficient time is not provided for the particles to become properly orientated within the
specimen during compaction at this moisture content and hence the representation of the
prepared sample to the compaction of the material on site may not be adequate.
At 80% OMC in both the BSM-foam and BSM-emulsion mixes the compaction time to reach
100% Mod AASHTO was sufficient to allow for proper orientation of the material particles, but
also short enough that the preparation of the sample does not necessarily become a very time
consuming procedure.
The influence of the surcharge weight is also evident in the results of the experiments. The
20kg load at 80% OMC showed a compaction time effect similar to that of the 10kg load at
90% OMC. The Kango 637® however suffered damages to the gear box at a 20kg surcharge
load and this damage is believed to be as a result of this load. Based on the short time the
20kg load gives and the fact that the Kango 637® suffered damages at this load the 10kg load
became the safest and best option. The 15kg surcharge showed compaction times in between
the 10kg and 20kg surcharge mass, it could therefore be considered as an option for the
surcharge weight used during compaction, however these results are all based on the Kango
637® results and when the correlation experiment between the Kango and Bosch hammer is
taken into account, the 10kg surcharge proves to be the best option.
In terms of the refusal density, it was generally seen that throughout the compaction of the
emulsion mixes the time required to achieve refusal density was between 2 and 5 minutes of
compaction time depending on the layer being compacted. The effect of moisture on this seems
minimal, as there are cases where at 80% OMC it took roughly 3.5 minutes to reach refusal
density and then at 90% OMC the layers all fall into a bracket of between 3.5 and 4 minutes.
The surcharges also had little effect, with the layers across the various samples, again taking
3.5, 4 minutes to reach refusal density. What is clear is that the achieved refusal densities are
influenced by the moisture content and surcharge load. The higher the moisture content, the
higher the bracket of the refusal density and the higher the surcharge load in combination with
the moisture content the higher the bracket in which the refusal density falls.
157

The foam mixes exhibited a similar trait to that of the emulsion mixes with regard to the refusal
density compaction. Once again as the moisture content increased so to does the bracket in
which the refusal density falls, and with the surcharge mass, the same trend was seen. A
difference was seen in the compaction of the foamed mix to refusal density, the density of this
mix continued to climb after 5 minutes of compaction. This results shows that the foamed mix
requires longer compaction time to reach refusal density than does the emulsion mix.
5.2 Comparison of Vibratory Hammer to Vibratory Table
The conclusion from this correlation experiment is that the vibratory hammer is a faster
procedure with less physical labour required (if no pulley system is available for the vibratory
table). Also the vibratory hammer gives more control and accuracy over both the target
densities and the final level of the surface of the sample (a more perpendicular surface face is
achieved with the vibratory hammer).
5.3 Correlation of N7 Material to G2 Material
Parameters used for the compaction of the clean G2 material i.e. time per layer etc. were not
applicable to the N7 G2 material. This is because, although the N7 material is a G2 quality, the
milling process of the recycling process changes the grading and quality of the material from its
original state when initially used in construction. The presence of RAP in the N7 material is
believed to have had an influence on the compaction of the material. More time is required to
reach the targeted site densities for the N7 material than for the clean bitumen treated G2
material to reach its 100% Mod AASHTO dry density. Simply because it takes 30 seconds to
compact the G2 material to its 100% Mod AASHTO does not mean it will take the same time to
compact the N7 material to its 100% Mod AASHTO density or the targeted site density. The
material properties, i.e. grading and the presence of other elements such as RAP or cement
from the initial construction influence the compaction time.
The correlation experiment shows an important conclusion. Time cannot be taken as a fixed
unit. Material properties influence the compaction time, therefore compaction using the
vibratory hammer to produce site related samples should be performed as a function of the
layer thickness and not as a function of time.
158

5.4 CT Scanning
From the CT scanning results the first conclusion that can be drawn is that the vibratory
hammer produces samples with very low voids contents. The extent to which the surface of
various layers is scarified has an influence on the level of voids at the intersection of two layers.
This indicates that when compacting, care must be taken that the surface of a compacted layer
is not scarified to extensively as this allows for higher voids contents in those areas, this is
because the compaction time assigned to compact a single layer (specifically for emulsion
mixes) is relatively short, this does not allow the loose material to necessarily arrange itself
properly even though the target Dry Densities are achieved. These points will inevitably be
week spots in the prepared sample.
The CT Scanning showed a consistency in the air voids across the two samples per mix that
were sent to the Netherlands. There is no sudden drop in the voids content in the lower layer of
the sample, this indicates that no subsequent compaction of the underlying layers takes place
nor does crushing of the material occur.
5.5 Repeatability Experimentation: G5 Material
The repeatability experiments showed that the compaction procedure developed using G2
material proved to be adequate for the compaction of lesser quality granular materials such as
G5 material. The G2 emulsion mix showed Dry Densities of around 103% Mod AASHTO and
using the same procedure on the emulsion treated G5 material Dry Densities of around 100%
Mod AASHTO were achieved; both these Dry Densities exceed the accepted level of site
compaction. Looking at the Untreated material the same results were seen as with the emulsion
mixes, the G2 material achieved Dry Densities of 103% Mod AASHTO and the G5 material
achieved Dry Densities of 101% Mod AASHTO.
5.6 Conclusions regarding variability
The variability results of Subsection 4.7 show that the vibratory hammer does have a low variability
in terms of specimen production. Therefore the vibratory hammer is capable of consistently
producing laboratory specimens. The results also show in Subsection 4.7.6 that the Mod AASHTO
compaction (at room temperature) is slightly less variable than the vibratory hammer compaction.
The difference is only slight. The largest difference came from the BSM-foam which showed a
difference of 0.67 from the Mod AASHTO COV at room temperature. The conclusion drawn from
such a small difference is that the variability of vibratory hammer is very similar to the Mod AASHTO
compaction.
159

5.7 Conclusions drawn from the Mod AASHTO curves
An interesting conclusion may be drawn when viewing the Mod AASHTO curves of the
untreated material against the BSMs. The maximum Dry Densities produced for the BSMs were
lower than the untreated material. This shows that the bitumen emulsion and foamed bitumen
influence the level of compaction that may be achieved.
5.8 Conclusions drawn from the Vibratory Hammer Moisture curves
The moisture curves of the vibratory hammer showed in general (with the exception of the
BSM-emulsion) that the graph moved slightly to the left of the Mod AASHTO moisture curve.
This shows that the OMC of the vibratory hammer is slightly lower than that of the Mod
AASHTO OMC; vibratory hammer OMC was typically around 90% of the Mod AASHTO OMC of a
specific mix. The curves also showed that in general higher Dry Densities are achieved using
the vibratory hammer at lower moisture contents.
The vibratory hammer moisture curves showed that in general the accepted level of
compaction, according to current standards, may be obtained at various moisture contents. The
bitumen emulsion and untreated materials produced compaction levels that are very close to
what would typically be accepted on site. The fact that the results are so close to accepted
compaction levels leaves room for variability (specifically in the case of the high quality G2
material which has an accepted level of compaction of 100 -102% Mod AASHTO (TRH 4) and
with the vibratory hammer levels of 102-103% Mod AASHTO were achieved). Therefore in
order to reduce this variability the compaction times may be increased by a factor of 2, this may
produce samples that are well in excess of the accepted levels of compaction, but the variability
may be reduced.
5.9 Overall Conclusions and Observations Drawn from the Experiments
A procedure using time to produce samples may be developed. This procedure should not be
used to compact a sample after a pavement has been constructed with the intention to
produced site representative samples. When the procedure using time to compact is compared
to the procedure followed by the USA (ASTM D 7382 – 07) and New Zealand it is evident that
the compaction time allocated from the experiment results is far less than what these two
countries have specified in their compaction procedures. The reason for this is the compaction
energy of the hammer. The compaction energy assigned by the USA is around 10 Joule, The
Bosch GSH 11E® used by Stellenbosch University has a point energy during compaction of 27
Joule, this is 2.7 times higher than specified by the USA, and will influence the compaction time
of the various layers.
160

The refusal densities of the various layers are variable. As was stated in the results, this is
believed to be as a result of the variability of the grading of the individual layers. Although the
final grading of a sample may be correct in relation to the grading curve of the material, the
material of the individual layers may in fact have variability in relation to the grading curve.
Much time is lost during the preparation of samples once compaction is completed. This is due
to having to unscrew the bolts of the split mould and then having to dismantle the entire mould
so as to remove the specimen and once removed the mould may at times be wiped clean,
depending on how dirty it has become during compaction, and then re-assembled for the
compaction of the next sample. This dismantling is necessary because the sample does have a
tendency to stick to the walls of the mould, therefore it cannot merely be slid out of the mould.
A not stick spray, this may be purchased at any supermarket, is used to treat the walls of the
mould prior to the compaction of a sample; this reduces the sticking of the sample to the mould
walls.
An important point that was also noted was that at high moisture contents, particularly 90%
and 100% OMC (Mod-U) the material in the mould began to stick to the bottom of the foot
piece. The effect of this was that upon raising the vibratory hammer, large amounts of material
from the compacted layer were removed from the mould. The compaction of the emulsion mix
for prolonged periods of time had the effect that emulsified bitumen began to ‘squeeze out’
between the walls of the mould and the foot piece. This ‘squeezing effect’ in the end results in
lower binder content of the compacted sample. Also small amounts of water were seen seeping
out at the base of the mould when compacting at high moisture contents (80%, 90% and
100% OMC (Mod-U)) for prolonged periods of time.
It was also noted that a times the interlocking between layers was not always adequate and the
sample tended to break in this case when being picked up. Therefore adequate scouring of the
surface of the compacted layer should be done, typically look to 10mm of scour. Also samples
should be picked up from the bottom end, this is because, as was seen in samples that were
too moist, the bottom end seemed at times as if it would fall off, and this is due to the size and
mass of the sample. The typical mass of a sample of 150mm × 300mm was in the order of
11kg and should the sample be picked up at the top end, the weight of the layers below each
interlocking face will exert a tensile force on the section and if the interlocking is not strong
enough the sample will break at those points.
During the compaction of the G5 material at higher moisture contents (90%OMC to 110%OMC)
it was noted that there was no seepage of water out from the bottom of the mould. Where in
the case of the G2 material there was clearly water seeping out. The G5 material when wet had
a very clayey appearance where the G2 did not. This clayey appearance may account for the
161

water retention of the G5 material, as clay materials do have a high level of water retention,
which may be the reason for the lack of seepage of water through the bottom of the mould.
162

6 RECOMMENDATIONS
Based on the results obtained from the experiments and the conclusions drawn, the following
recommendations may be made.
6.1 Reducing the loss of time
The dismantling and re-assembling of the mould takes up a significant part of the sample
preparation time. The removing and replacing of the bolts in particular consumes time.
Therefore the mechanism of loosening and fastening the split mould and fastening the mould
onto its steel base should be reviewed. As apposed to using nuts and bolts to fasten the mould,
a type of clasp should rather be used. i.e. instead of having to unfasten the bolts, the clasps
may merely be opened and the mould disassembled.
4
3
Closed Clasp
2
1
Opened Clasp
Figure L.109: Split mould set up with clasps
Figure L.109 shows the Mould (4) set up taken from the British Standards. A similar clasp
system to this is proposed to mount the mould to the base; these clasps are shown by number
2 on the figure and the base by number 1. The bolts used in the British Standards (number 3)
should be replaced by clasps similar to those shown on the right of the mould set up; a total of
four clasps would be need, two clasps on either side of the mould. It is also recommended that
as apposed to two clasps being mounted to the base, that two more clasps be mounted; a
clasp at the point marked 1 on the mould set up and a clasp directly across from it. This would
bring the total number of clasps used to fasten the mould to the base to four which would
provide enough support for the mould during compaction. It is believed that a clasp system
similar to the one proposed would greatly speed up the process of producing samples.
163

6.2 Reducing the loss of material
The extension piece used to extend the split mould height so as to accommodate the final Layer
of material just prior to compaction was taken from the moulds used for Mod AASHTO
compaction. This piece does not fit tightly around the split mould used for the vibratory
hammer compaction; instead there is a small surface space between the mould and the
extension’s inner circumference; this is where material gathers while compacting Layer 5. An
extension piece that fits exactly on the circumference of the split mould’s opening should be
made; this will therefore reduce the amount of material lost in Layer 5.
6.3 Vibratory Hammer Specifications
From the research done the following Vibratory Hammer Specifications are proposed:
Power rating:
Frequency:
Point Energy:
1500 Watt consumption
900 – 1890 beats/min (15 – 31.5Hz)
25 Joule
The vibratory Hammer should be mounted on two guide rods; one on either side of
the hammer.
The total mass of vibratory hammer, surcharge and mounting head should be 30kg
± 1.5kg.
The Bosch GSH 11E® vibratory hammer meets these specifications
164

6.4 Procedure to be followed for compaction using vibratory hammer
Based on the findings of this report the following compaction procedure is recommended when
using the vibratory hammer.
The compaction of a 150mm X 300mm sample of bitumen stabilized granular material will be
performed using a vibratory hammer, e.g. the Bosch GSH 11E®, with a surcharge of 10kg
mounted in a frame. Compaction of the material will take place with the aggregate at room
temperature i.e. 25ºC. The layout of the procedure is similar to the layouts provided in the TG 2
manual.
1 APPARATUS
1.1 A steel split mould 152mm in diameter and 300mm in height with an extension piece
and clasps to fix the mould to the base of the frame.
1.2 3 Circular papers with diameter of 152mm.
1.3 Non-stick spray e.g. non stick cooking spray purchased at any supermarket.
1.4 A Vibratory Hammer with the following Specifications (see note 4.3):
Power rating:
Frequency:
Point Energy:
1500 Watt consumption
900 – 1890 beats/min (15 – 31.5Hz)
25 Joule
The vibratory Hammer should be mounted on two guide rods; one on either side of
the hammer. A mounting head should be fitted to the vibratory hammer to allow a
surcharge of 10kg to be mounted to the vibratory hammer. There should be a pulley
system connecting the frame and mounting head. This allows for easy lifting and
lowering of the vibratory hammer.
The total mass of vibratory hammer, surcharge and mounting head should be 30kg
± 1.5kg.
1.5 A 150mm tamping foot
1.6 Material Scoop (90mm Ф x 85mm h)
1.7 Samples are compacted in 5 Layers.
1.8 Suitable marker e.g. permanent marker
1.9 Adjustable spanner to fasten and loosen surcharge load to the vibratory hammer.
1.10 Steel ruler of length >300mm
1.11 Chisel for tamping layers
1.12 Drill with drill bit of 300mm with a point marked off 10mm from the tip of the bit.
165

2 PROCEDURE
2.1 Preparation of the material
Preparing the sample of material for initial Moisture Curve
Determine the grading curve of the aggregate (TMH 1) and reconstitute the material to
produce samples that will be used to obtain the OMC of the natural (untreated) material
using the Mod AASHTO compaction method (TMH1: Method A7).
Develop a Mod AASHTO curve (dry density versus moisture content) for the appropriate
BSM using the OMC of the untreated material. This is done for various moisture contents
over a range of 60% of OMC (Mod-Untreated Material) to 110% of OMC (Mod-Untreated
Material) or till the appropriate moisture curve is obtained.
Preparing the sample of material for vibratory hammer compaction
From the grading curve reconstitute the material to the produce a sample of 14kg (see
note 4.1) of aggregate with a maximum particle size of 19mm. A total of 5 or 6 samples
of 14kg each are needed for a moisture curve.
The aggregate is prepared as follows:
1. For a moisture curve samples are to be compacted over a range of moisture
contents. Ranging from 2% moisture to 10 or 12% moisture; increasing in
increments of 2%.
2. Should cement or lime need to be added to a specific mix, add these first
to the sample:
3. For Untreated material i.e. no bitumen stabilization, add the fractions of
water to each of the individual sample masses.
4. For BSM-emulsion, the moisture content of the bitumen emulsion needs to be
calculated out of the physical mass of water that is added by hand. An
example of this calculation is as follows:
166

Assumptions made for the example:
A mix is to have a target moisture content of 6% moisture.
The bitumen emulsion content of the mix is 3%.
The bitumen emulsion is a 60/40 emulsion.
For a 60/40 emulsion, 40% of the emulsion is water. Therefore of the 3%
bitumen emulsion to be added to the mix, 40% of it is water. Therefore the
fraction of water being added to the mix from the emulsion is 40% x
3%÷100 = 1.2%
The fraction of water that is to be added to the mix in order to obtain a
moisture content of 6% is now 6%-1.2% = 4.8%. The physical mass of
water to be added to the mix in order to give a target moisture content of
6% when the bitumen emulsion is added is now 4.8% of the mass of the
sample.
First add the physical mass of water to the material and allow to stand for 40 –
60 minutes. After this time add the bitumen emulsion to the mix and also allow
to stand for 40-60 minutes to allow breaking of the bitumen emulsion.
5. For the BSM-foam add the fractions of moisture to the material checking
the ratio of the water relative to the OMC of the moisture curve developed for
the untreated material from the TMH 1. When the moisture to be added reaches
around 70-80% of the OMC of the untreated material, add this same moisture
to the remaining samples and calculate the amount of moisture to be added to
the sample to achieve the targeted moisture content. Prepare the BSM-foam in
accordance with the procedure outlined in the TG (2) manual. After the foaming
procedure add the remaining moisture to the samples mixed with moistures of
70-80% of OMC of the untreated material.
167

2.2 Compaction Procedure
2.2.1 Prepare the mould and vibratory hammer
Preparing the vibratory hammer
Fix the Mounting Head to the vibratory hammer and fit hammer onto the
guide rods. Place the 10kg surcharge load onto the mounting head and
fasten tightly – see separate drawing for Mounting Head, Subsection 5
(Kelfkens, 2008). Using the pulley system raise the vibratory hammer to the
maximum height it can be raised or to an adequate height that will allow
the operator to work beneath the vibratory hammer.
Preparing the Mould
Make sure the mould is clean and then spray the interior of the mould with
the non-stick spray. After a sample has been compacted and removed from
the mould, the mould should be cleaned by wiping of excess material from
the mould walls. This should be done prior to the compaction of the next
sample. Fix the mould to the base of the frame directly below the foot piece
of the vibratory hammer using the clasps. Place two of the circular paper
sheets at the base of the mould.
Lower the vibratory hammer into the mould, checking that the vibratory
hammer is perpendicular to the base of the mould i.e. the tamping foot is
flat on the base with no point of the foot slightly raised. Allow the vibratory
hammer to rest in the mould with no material present. Where the lower end
of sleeve of the mounting head rests on the guide rod mark that position
clearly on the vertical guide using the suitable marker. Raise the vibratory
hammer and measure up from the initial mark 300mm and mark this clearly
(non-erasable).
2.2.2 Compaction of the sample
Addition of material to mould
Material is placed in the mould 1 using the material scoop. Fill the scoop
with the prepared material and level off the scoop and place it in the mould.
Add three scoops of material to provide a starting layer thickness of 92mm.
168

Using the chisel, work the material around in order to evenly distribute it in
the mould; try to distribute the particles evenly as well i.e. not too much
fine material on top or to much coarse material on top, but rather a fair
distribution of each i.e. unsegregated. Make sure the material is as level as
possible before lowering the vibratory hammer till the foot piece comes to
rest on the material.
Compaction of individual layers
Samples are compacted according to set times for each layer. The
compaction times for the individual layers of a sample according to the type
of Bitumen Stabilized Material is provided in Table L.29:
Table L.29: Compaction times of individual layers for various BSMs
Compaction Time of Individual layers Number in Seconds
Mix type Layer 1 Layer 2 Layer 3 Layer 4 Layer 5*
Untreated 10 15 15 15 15
BSM-emulsion 10 15 15 15 15
BSM-foam 10 25 25 35 25
After the material of a layer has been compacted for the allocated time,
raise the vibratory hammer. Using the drill, scarify the entire surface area of
the top of the compacted layer to a depth of ± 10mm (see Note 4.2).
After the surface of a respective layer has been scarified, add the material
for the next layer and compact accordingly.
*After Layer four has been compacted and scarified, the extension piece
(collar) must first be fitted to the mould before adding the material for layer
5. After adding the material for layer 5 place a circular sheet of paper on
top of the material and then lower the vibratory hammer into position; the
paper helps prevent material of the final layer from sticking to the tamping
foot. Before raising the vibratory hammer the final height of the sample
must be measured, once this is done the vibratory hammer may be raised
and the sample removed.
Measuring the final height of the specimen
After Layer 5 has been compacted and prior to raising the vibratory hammer
take the steel rule and measure the distance from the zero line to the
169

ottom end of the sleeve. This distance is taken as the final height of the
specimen.
Removing and handling the compacted sample
Raise the vibratory hammer and remove the extension piece (collar).
Disassemble the mould entirely. Place a plastic bag over the sample and
remove it taking care to pick the sample up from the bottom end. Weigh
the sample after compaction to check the final mass of the sample.
Checking the moisture content of the sample
Take a small amount (750-950 gm) of BSM either just prior to during or
after compaction and using the standard oven drying method determine the
moisture content.
Determining the final Dry Density
From the moisture content determined, the final mass of the compacted
sample and the final height measured the final Dry Density of the sample
may be determined.
2.3 Moisture sensitivity curve
For the moisture sensitivity curve a total of 5 or 6 samples needs to be
compacted at various moisture contents. This is described in Subsection 2.1.
The curve is developed by plotting the final Dry Density of each of the
samples against their respective final moisture contents. The peak, point at
which the curve turns, is the OMC of the vibratory hammer and the
Maximum Dry Density.
Compaction specifications for site compaction
The MDD of the vibratory hammer may be used to specify site compaction levels.
The table below provides the levels of compaction:
Table L30: Specifications for the level of site compaction
Material Type/Quality Level of Site Compaction (% Vibratory Hammer)
Untreated BSM-emulsion BSM-foam
G1 and G2/G3 (High Quality) 98.6 - 100 98.6 - 100 96.1 – 98
170

For G4 to G6 materials the fraction of Mod AASHTO compaction specified for the
level of site compaction is to be used. An example of this is as follows: for a G5
material the current specification for site compaction is 95% of Mod AASHTO
compaction. Therefore according to the previous statement the accepted level of
site compaction is 95% of vibratory hammer compaction.
See Note 4.8
The Moisture curve of the vibratory hammer may be compared to the moisture
curve of the specific mix which was developed from the TMH 1: Method A7. The
OMC values are then compared and the lower OMC of the two curves is used to
specify the compaction moisture on site.
171

3 CALCULATIONS
3.1 Addition of lime or cement
Cement or lime content (C/L)
C/L (gm) = C/L (%) × 14000 ÷ 100
3.2 Addition of water for untreated material
Water (gm) = (target moisture content (%))/100× mass of sample (gm)
3.3 Addition of stabilizer and water to Bitumen Stabilized Material (BSM)
BSM-emulsion
Mass of bitumen emulsion
Emulsion mass (gm) = Emulsion content (%)/100 × dry mass of aggregate (gm)
Moisture contents
MC in BSM from emulsion (%) = (MC of emulsion (%))/100 × emulsion content (%)
Mass of water added to BSM = (∆MC) × mass of sample (gm))
100
∆MC = X (%) - MC in BSM from emulsion (%)
X = target moisture content of the mix
BSM-foam
The bitumen stabilizer is added according to the method provided in the TG 2 manual
for preparing BSM-foam.
Ratio of Moisture contents
MC mix : OMC = X/ (OMC (untreated material))
MC mix : OMC = Ratio of the targeted moisture content of the mix to the
OMC (untreated material)
X = target moisture content of the mix
172

For MC mix : OMC > 0.8 (80%)
Mf = X – (MC mix : OMC) previous mix × (OMC (untreated material))
MF = Moisture added after foaming
X = target moisture content of the mix
(MC mix : OMC) previous mix = Final sample with a ratio ≤ 0.8
3.4 Dry Density
Volume of the sample = π × 0.005625 × Fh
Fh = Final height of the sample
Dry Density = (Fm(kg)/(1+(MC(%)/100)) ÷ Volume of the sample
Fm = Final Mass of the sample
MC = Moisture Content
4 NOTES
4.1 For a final sample of 300mm high a sample mass of 14kg is recommended when
preparing the BSM.
4.2 Layers should not be scarified deeper than 10mm. The result of scarifying deeper
than10mm is that the layer being compacted does not bond adequately well to the
previous layer and hence there is an increase in voids at this interface.
4.3 Should the vibratory hammer not meet the specifications provided and where no
suitable alternative compaction hammers can be sourced, then a vibratory hammer
with a point energy of 25 Joule ± 2 Joule should be used. If the weight of the
hammer deviates from the specifications by more than 5%, then calibration tests
need to be made.
4.4 After a sample has been compacted and removed from the mould, the mould should
be cleaned by wiping off excess material from the mould walls. This should be done
prior to the compaction of the next sample.
173

4.5 When preparing the moisture curve, the material should be looked at carefully. It
must be noted at which moisture content the material becomes muddy. This is due
to when samples are prepared that are to be tested in the laboratory; material that
is muddy may not produce samples that are representative of the site compaction
(in terms of particle orientation). The samples should then be compacted at the
moisture content immediately below the moisture content at which the material
became muddy. Should samples be prepared in the laboratory for testing without
first preparing a moisture curve, then low levels of moisture should be added to the
sample and slowly increased until the operator is satisfied that the material is not
muddy but adequately wet.
4.6 When the sample prepared is to be used in the laboratory and the first four layers
have been compacted, add sufficient material to layer five so that that the final
height of the sample is 300mm or slightly higher. This is checked by viewing the
final position of the sleeve relative to the 300mm marked of point on the guide rod,
a tolerance of 2mm either side of 300mm is allowed. The sleeve may finish either
on the mark, slightly above the mark or blow the mark. For each of the finishing
positions a description of the procedure to be followed is given in a), b) and c).
a) If the sleeve finishes on the 300mm mark after compacting layer 5, the
sample is removed as previously described.
b) If the sleeve finishes above the 300mm mark after compacting layer 5 a
steel straight edge is used to cut of the piece of the sample extending out
of the mould. Material is then sieved through a 4.75mm sieve on top of the
sample. The vibratory hammer is the lowered and the sieved material is
compacted till the sleeve reached the 300mm mark. The sample is then
removed as previously described.
c) If the sleeve finishes below the 300mm mark after compacting layer 5, the
surface is scarified and three scoops of material are added. The added
material is then compacted for the same duration as layer five and checked
using a) and b) of note 4.6.
4.7 Should 7 ply shutter board not be obtainable then a wooden base with material
properties as close to those of the 7 ply shutter board should be used.
4.8 The compaction specifications for site compaction need to be revisited as
compaction data from sites become available.
174

5 DRAWINGS AND PHOTOGRAPHS
Positioning of
dead weight
Rubber
fitting
Sleeve
127
113
Sleeve
112
870
64
50
Figure L.110: Schematic drawing of the Mounting head for the Bosch GSH 11E®
Mounting
head
10kg
Surcharge
Vertical
guide rod
Rubber
Fitting
Sleeve
Figure L.111: Left view of Mounting head
Figure L.112: Front view of Mounting head
175

6 Summary of vibratory hammer compaction procedure
Outlined in this summary is the sequential procedure of the vibratory hammer.
Step 1
When material is obtained that is to be compacted in the laboratory, the first step is to perform
a grading on the material. Following the grading, samples for individual specimens are
reconstituted from the grading results.
Step 2
The first set of samples reconstituted is samples with a mass of 7kg that are used to perform a
moisture sensitivity analysis using Mod AASHTO compaction (TMH 1: Method A7). Two moisture
sensitivity analyses are performed. The first is an analysis on the untreated material, i.e.
material only having moisture added to it and not having under gone bitumen stabilization. The
OMC of the untreated material (OMC-U) is obtained from this the first analysis (Figure L.113).
MDD
Moisture Sensitivity curve -
Untreated Material: Mod AASHTO
Dry Density
(kg/m3)
OMC-U
Moisture Content (%)
Figure L.113: Moisture Cure:
Mod AASHTO- Untreated
The second analysis is a moisture sensitivity analysis on the BSM i.e. the material after it has
under gone bitumen stabilization. This analysis is performed using the OMC-U to determine the
moisture content of each sample. A fraction of the OMC-U is added to the material prior to
bitumen stabilization that will provide the target moisture content once the material has under
gone bitumen stabilization. The fractions of OMC-U start at 60% increasing in increments of 10
until 110%. From this analysis the OMC of the BSM material is obtained (Figure L.114).
176

Moisture Sensitivity curve - BSM:
Mod AASHTO
MDD
Dry Desnity
(kg/m3)
OMC-BSM
Moisture Content (%)
Figure L.114: Moisture Cure:
Mod AASHTO- BSM
Step 3
The second set of samples reconstituted is samples with a mass of 14kg. These are used for to
perform the moisture sensitivity analysis for the vibratory hammer. At least 5 samples should be
reconstituted.
Cement or lime is first added to the material; should the mix being prepared require these
stabilizers. Moisture is then added to the material in varying amounts across the samples. The
moisture added starts at 2% moisture for sample 1 increasing in increments of 2% until a
content of 10 or 12% for the final sample. The material is then allowed to stand for ± 60
minutes. The bitumen stabilizer (Emulsion or Foamed) is then added after the 60 minute time
period. The material is once again allowed to stand for ± 60 minutes to allow for the breaking
of the bitumen.
Step 4
The mould and vibratory hammer are prepared as outlined in 2.2.1 of the procedure.
Step 5
Each sample is compacted individually to produce a specimen for a specific sample. Specimens
are compacted in five layers. Material from the specific sample is placed into the mould using a
material scoop. Three scoops of material per layer are placed into the mould and each layer is
compacted for a set period of time (Table L.29).
Step 6
177

Prior to removing each specimen from the mould the final height of the specimen is measured.
The specimen is then removed and the final mass is measured. The remaining material from
the sample is used to perform a moisture content test for that specimen. The Dry Density of
each sample is then calculated.
Step 7
The moisture curve for the vibratory hammer is plotted. This is done by plotting the Dry Density
of each specimen against its own moisture content. All specimens are plotted on the same set
of axis. The Maximum Dry Density and OMC of the vibratory hammer are then read off the
curve (Figure L.115).
Moisture Sensitivity curve - BSM:
vibratory hammer
MDD
Dry Density
(kg/m3)
OMC-vib
Moisture Content (%)
Figure L.115: Moisture Cure:
vibratory hammer- BSM
Step 8
The Maximum Dry Density for the vibratory hammer is then used to specify the target level of
compaction for site (Subsection 2.3 of this procedure).
Step 9
The OMC of the vibratory hammer (OMC-vib) is then compared to the OMC of the Mod AASHTO
compaction for the specific mix. The lower OMC of the two is then selected for site compaction.
See the figures below.
Figures L.113 to L.115 show that the OMC of the vibratory hammer is typically lower than the
Mod AASHTO OMC. The BSM-emulsion however may provide a curve which has an OMC-vib
higher than the Mod AASHTO OMC. Therefore it becomes necessary to compare the moisture
curves of the vibratory hammer and Mod AASHTO compaction methods as the lower OMC is
used for site specification.
178

Step 10
Should specimens be prepared in the laboratory for testing purposes then specimens are
compacted at 100% of OMC-vib.
179

REFERENCES
Muthen, KM, 1998, Contract Report CR-98/077, CSIR
Hanekom, R, 2007, University of Stellenbosch, Thesis V08, Compaction of Cold Mixes using the Kango
Hammer Method
Prochaska, A.B., and Drnevich, V.P., (2005), “One-Point Vibrating Hammer Compaction Test for Granular
Soils,” Proceedings, GeoFrontiers Conference, ASCE, Austin, TX, January, 25 p.
BS EN 12697-32:2003, 2004, Bituminous mixtures-Test methods for hot mix asphalt Part 32: Laboratory
compaction of bituminous mixtures by vibratory compactor, University of Nottingham, Uncontrolled Copy, ©
British Standards
Thenoux, G, Jamet, A, Encina, C, 2004, A Study and Recommendations of a Mix Design Procedure Using
Gyratory Compactor for Foamed Asphalt Recycled Material, School of Engineering Universidad Católica de
Chile
Jönsson, M, Partl, MN, Flisch, A, 2002, Comparison of Different Compaction Methods Using X-ray Computer
Tomography, EMPA
Van de Ven, M, et al, , Gautrans, 1997, ITT Report 18.1-1997 Investigation into the Feasibility of Scaling
Granular Materials for use with the MMLS Trial Tests on G1, Waterbound and ETB, Institute for Transport
Technology
Theyse, HL, 2003, Confidential CSIR Contract Report CR-2003/23 First Level Analysis Report: HVS Testing of
the Foamed-treated crushed stone base on the N7/1 near Cape Town,
Theyse, HL, 2004, Restricted CSIR Contract Report CR-2004/38 The Compaction Potential of Foamed- and
Emulsified Bitumen Treated Material, Pretoria
ASTM Standard D7382, 2007, “Standard Test Methods for Determination of Maximum Dry Unit Weight and
Water Content Range for Effective Compaction of Granular Soils Using a Vibrating Hammer,” ASTM
International, West Conshohocken, PA, www.astm.org.
TNZ B/02, 2005, “Specifications for Construction of Unbound Granular Pavement Layers”, Transit New
Zealand
NZS 4402, 1986,”Test 4.3.1 New Zealand vibrating hammer compaction test”, New Zealand Standards,
Wellington,
180

Draft TRH 4, 1996, “Structural Design of Flexible Pavements for Interurban and Rural Roads”, Pretoria,
South Africa,
CSRA, 1987, Standard Specifications for Road and Bridge Works 1987, Committee of State Road Authorities,
Pretoria
Sabita, 1993, GEMS-The design and use of granular emulsion mixes, Sabita, Roggebaai, South Africa
MH 1, 1986, “STANDARD METHODS OF TESTING ROAD CONSTRUCTION MATERIAL”, National Institute for
Transport and Road Research, Pretoria, South Africa
Web Sites
Miller group & Associate authors, 2004, http://www.millergroup.ca/pavement/emulsion_mixes.html, July 22,
2008 (11:28 AM)
Patent Storm, http://www.patentstorm.us/patents/6451885-description.html, July 22, 2008 (11:29 AM)
181