A New Design Method for Permeable Pavements Surfaced with Pavers

by

John Knapton, Professor of Structural Engineering,Newcastle University

Ian D. Cook, Visiting Professor, Newcastle University

David Morrell, Marshalls Mono Ltd.

(Transcript of a paper published in the February 2002 edition of the Highway Engineer, the journal of The Institution of Highways & Transportation)

INTRODUCTION

Recent increases in levels of rainfall have led to The Environment Agency introducing guidance on the impact of development on flooding. Environment Agency's publication "Policy and practice for the protection of floodplains" (HMSO) states that: "Inappropriate development within floodplains should be resisted where such development would itself be at risk from flooding or may cause flooding elsewhere. To minimise any increased surface water run-off, new development must be carefully located and designed. Where appropriate, run-off source control measures which may also improve water quality should be incorporated into the development proposal."

Environment Agency's key engineering principles set out in their document are: "Development generally increases the amount of impermeable land in river catchments. This increases the amount and rate of surface water run-off which if unmanaged can increase river flows and the risk of flooding. The adverse effects of inappropriate development, however small, are cumulative and can lead to significant problems in the longer term."

The Environment Agency is empowered by government to advise planning authorities on development and flood risk matters. Government Circular 30/92 states: "The Government looks to local authorities to use their planning powers to guide developments away from areas that may be affected by flooding, and to restrict development that would increase the risk of flooding…"

Prior to the autumn 2000 floods, CIRIA Report C521, "Sustainable urban drainage systems" (1999) had already highlighted the potential drainage problems associated with unchecked urban development. It concludes: "Drainage methods that take account of quantity, quality and social issues are collectively referred to as Sustainable Urban Drainage Systems (SUDS). These systems are more sustainable than traditional drainage methods because they:

The role of permeable paving within SUDS can be appreciated from the CIRIA document's conclusion: "SUDS are made up of a series of structures built to receive surface water runoff working in conjunction with good management of the site. There are four general methods of control:

It is clear that all future developments need to address SUDS in order to gain planning approval. Planning authorities will be looking for evidence of innovative design in making their judgements. The inclusion of permeable paving within SUDS, possibly as one element in an overall sustainable environment design package, will greatly enhance the likelihood of a planning application succeeding. This Paper describes research undertaken into permeable flexibly bedded pavers and presents a design method based upon the results of that research.

 

BACKGROUND

The simplistic concept of allowing water to drain through the pavement and into the subgrade so eliminating entirely downstream drainage is unlikely to prove successful in the great majority of UK applications. This is because 96% or more of UK developments will be over clays which are not suited to accepting precipitation directly. In the UK, a permeable pavement is required to absorb 180litre/second/hectare. Whilst there is no difficulty in achieving this with pavement construction materials, most UK subgrades would be able to absorb only a small fraction of this. The remainder has to be retained in the pavement, either to gradually percolate into the subgrade or to be taken through a sub-surface drainage system. Such a system can be designed to constrict the flow and so act as a detention system, detention occurring in either the pavement or the drains (or in both). In view of the above, in addition to its conventional structural requirements, a permeable pavement has to be designed on the basis of the permeability of each of its courses and of the subgrade and it may also have to be designed on the basis of the volume of water which it can retain to attenuate downstream flow.

Allowing water to percolate into clay has the disadvantage that many clays in the UK lose much of their strength when wet. Before the pavement was constructed, the overlying vegetation growing in topsoil acted as a water recycling system. Rainwater entered the topsoil was absorbed by the roots of the vegetation and evaporated through transpiration: the water never reached the clay. With the removal of the topsoil and vegetation, the clay will then begin to absorb the previously recycled water and will weaken. This will cause structural difficulties within the pavement which will be irreversible.

 

FULL SCALE PERMEABLE PAVING TRIALS

To establish the relationship between paver joint width, jointing material characteristics and permeability, a series of preliminary tests was carried out on a single joint between two conventional pavers. Testing was undertaken using a box containing two 200mm x 100mm x 80mm thickness rectangular pavers installed in such a manner that water could be applied to the straight joint between them, collected and measured over time. Rainfall was simulated by a simple graduated vessel filled with one litre of clean water. The water was gradually poured over the joint between two pavers, maintaining a constant head of 3mm. Table 1 shows the results.

These results show that permeable pavers can have a standard joint of 1.7mm or a larger joint of 6mm. If a standard joint is specified, the jointing sand must have no particles finer than 212m m. If a wider joint is used, it must be filled with a sand with no particles finer than 2mm. The other combinations of joint width and sand grading failed to achieve the permeability rate of 180 litre/second/hectare when a 30% potential flow rate is applied.

The preliminary tests conclude that pavers are suitable as an infiltration medium. They conclude that jointing, bedding and underlying material particle size distribution is critical to the amount of water which can infiltrate a pavement. They show that infiltration well in excess of the UK requirement of 180litre/second/hectare can be achieved with conventional pavers and with pavers developed specifically with infiltration in mind.

Test No.

Joint Width (mm)

Minimum joint filling particle size (m mm)

Permeablilty

(litres/sec/hectare)

1

1.7

63

109

2

1.7

63

52

3

1.7

150

196

4

1.7

150

206

5

1.7

212

543

6

1.7

212

625

7

6

212

111

8

6

212

127

9

6

2000

1875

10

6

2000

2272

11

6

2000

2142

Table 1. Results of Newcastle University infiltration tests

A full scale test has been undertaken using Newcastle University Rolling Load Facility (NUROLF) on an area of pavement surfaced with permeable pavers comprising the following specification:

Permeable pavers

50mm Laying course material

350mm Crushed 20mm open graded gravel roadbase material

150mm DTp Type 1 granular sub-base material (local dolomitic limestone)

4% CBR subgrade material - boulder clay

The 9m long test site was divided into three sections - see Figures 1 to 4. In the first 3m length, the laying course and jointing material comprised a 4mm particle size brown washed natural gravel. In the remaining 6m length of the test site, the laying course and jointing material comprised Cloburn 6mm washed crushed red micro-granite available in the Central Lowlands of Scotland.

The central 3m of the test included a knitted geotextile fabric separating the laying course material from the roadbase material. The whole 9m x 2m test site was lined with a 1000 gauge waterproof polyethylene membrane as shown in Figure 1. This allowed the volume of water introduced into the pavement to be measured. The waterproofing enclosed the pavers, the laying course material and the open graded roadbase material shown in Figure 2 but not the underlying sub-base material. At each end of the 9m long trial area, a vertical drain was included in order to assess the level of water standing in the pavement and also to facilitate the removal of water from the pavement - see Figure 3.

The permeable pavers were installed to a 90 degree herringbone pattern as shown in Figure 4 so that the NUROLF vehicle ran parallel to and normal to the paver joints. In order to simulate the most adverse combination of traffic and climate, the test area was maintained in a saturated condition throughout the whole of the testing. This was achieved by a sprinkler system which was activated before and during all of the testing. Prior to commencing the trafficking, the sprinkler system was activated in order to fill all of the voids in the roadbase with water. It was established that the roadbase material could accept 32% by volume of water. The trial section was filled and emptied three times prior to a fourth filling which was maintained during the trafficking trials.

During the four filling phases, the level of the free water surface within the roadbase was observed and it was noted that a horizontal surface was maintained within the open graded roadbase material, even when all of the water was applied through a single point in the pavement. From this, it was concluded that the water was flowing freely through the roadbase material.

 

 

(FIGURE OMITTED TO SAVE SPACE ON THE SERVER)

Figure 1. NUROLF test site showing waterproof material prior to installation of crushed rock open graded roadbase.

 

 

(FIGURE OMITTED TO SAVE SPACE ON THE SERVER)

Figure 2. Open graded roadbase material installed in NUROLF.

 

 

(FIGURE OMITTED TO SAVE SPACE ON THE SERVER)

Figure 3. Vertical drains installed at each end of NUROLF allow the level of stored water to be measured.

 

 

(FIGURE OMITTED TO SAVE SPACE ON THE SERVER)

Figure 4. Permeable pavers under test at NUROLF

NUROLF applies a vertical wheel load of up to 5000 kgf through its offside wheel to the centre of the test site over a test length of 9m. It commences a cycle at one end of the site and accelerates linearly over half of the length of the test site so that the load wheel has attained a speed of 2.3 m/s at mid point. It then decelerates over the second half of the test site, becomes stationary and undertakes the second half of its cycle by repeating the above in reverse. It undertakes a complete cycle in 53 seconds and in so doing applies a horizontal force of 500 kgf always in the same direction, to the pavement.

This combination of a vertical load of 5000 kgf and a horizontal load of 500 kgf relates closely to the heaviest loading to which a permeable pavement is likely to be subjected. By comparison, the maximum non-steering axle load normally applied by a fully laden commercial vehicle is 9500 kgf, resulting in a wheel load of 4250 kgf.

BS 7533:1992 recommends that fully channelled loading should be equivalenced to normal wandering highway loading on a 3:1 ratio. By taking account of this factor and by applying the 3.75 power law relating axle load to pavement damage sustained, the number of standard 8000 kgf axles to which one pass of NUROLF equates is 6 i.e. each full cycle of NUROLF inflicts damage onto the pavement surface that would be inflicted by 12 standard 8000kg axles. When working continuously, NUROLF achieves 800 equivalent standard axles per hour. In this test, 32,000 equivalent standard axles (ESAs) were applied in 40 hours running. Initially, 16,000 ESAs were applied using a wheel load of 3000kg and the remaining 16,000 ESAs were applied using a wheel load of 5000kg.

The results are shown in Figures 5 and 6. In these Figures, the 4mm washed natural gravel laying course material is to the left whilst Cloburn 6mm washed crushed micro-granite is within the central and right hand zones. A knitted geotextile has been included beneath the Cloburn material in the central zone.

 

 

(FIGURE OMITTED TO SAVE SPACE ON THE SERVER)

Figure 5. Deformation of the test pavement in the wheel track following trafficking by 6000kg axles.

 

 

(FIGURE OMITTED TO SAVE SPACE ON THE SERVER)

Figure 6. Deformation of the test pavement in the wheel track following trafficking by 10,000kg axles.

The NUROLF results allow the conclusion to be drawn that when axle loads do not exceed 6000kg, permeable pavers will sustain regular channelised loading up to the levels which would be anticipated in even the most extreme situations in permeable paving. The testing was continued to 16,000 ESAs and rutting developed to a depth of 4mm in the Cloburn 6mm washed crushed micro-granite. Where the Cloburn material was separated from the underlying coarse graded gravel by a knitted geotextile, the rut depth was11mm. The zone installed over 4mm washed natural gravel deformed by 9mm. When the axle load was then increased to 10,000kg, the deformation increased by approximately 70% as shown in Figure 6 and remained at that level for a further 16,000 ESAs.

The above indicates that the deformation arises from an initial compaction of the laying course material and the roadbase material. The conventional gradual development of rutting as a result of fatigue does not occur in permeable pavements because in ensuring such pavements have structural stability, sufficient stiffness has been provided to ensure that fatigue is not a significant issue.

The NUROLF results suggest that when axle loads exceed 6000kg, initial deformation will be unacceptable and it will be necessary to introduce stabilisation to the open graded roadbase material. Cement content will depend upon aggregate grading but a figure of 180kg/m3 has been found to be satisfactory.

The development of significantly greater levels of rutting in the zone including the knitted geotextile had not been expected. Following the testing, an investigation revealed that the reason for this enhanced rutting value is the pressing of the roadbase material into the geotextile during loading. Effectively, during the construction phase, the geotextile spanned from high point to high point over the roadbase particles and the trafficking then stretched the geotextile, pressing it down into the depressions between the roadbase aggregate particles. It is recommended that the geotextile be omitted from pavements in order to reduce rutting.

ANALYSIS OF DESIGN RAINFALL EVENTS IN THE UK

BRE Digest 365 "Soakaway design" (1991) provides guidance on the assessment of the levels of rainfall likely to occur in the UK. For design purposes, it is assumed that hydrographs have a ‘block’ nature i.e. rain falls with constant intensity over a period, and that there is no attenuation of flow between the rainfall landing on the surface and the inflow to the infiltration system. Design rainstorm events are described in terms of intensity, duration and frequency.

Drainage systems are normally designed on the basis of a specific return period. The selection of design return period will depend upon the consequences of failure and a return period appropriate to the risk should be selected. In many cases a return period of 5 years is used as a basis for design. A longer return period may need to be selected depending upon the potential consequence of failure or as a requirement for adoption by an authority. At a particular location, for a specified return period, the rainfall depth varies throughout the country and so attention must be paid to the location of the permeable pavement. The Meteorological Office provides rainfall statistics. Alternatively the appropriate statistics may be calculated using published procedures.

The Institute of Hydrology has carried out an extensive analysis of rainfall statistics and has provided a method to determine the relationship between depth, duration and return period (Institute of Hydrology, 1975). This forms the basis for the method described here. The notation MT-D min is used to identify a storm. For example, an M5-10 min is the depth of rainfall of a 5-year return period storm event of 10 minutes duration.

It is conventionally assumed that the depth of rainfall occurring during a 60 minutes storm recurring every five years is 20mm throughout the UK. The depth of rainfall occurring every five years over storm durations other than 60 minutes is obtained as follows. The design rainfall depth for any given return period and storm duration can be found by multiplying 20mm by a factor Z1. Factor Z1 is read from Table 2 which requires a knowledge of "r", the ratio of 60-minute to 2-day rainfalls for a 5-years return period. Values of r are given in Table 3.

 

r

Rainfall duration (D)

Minutes

Hours

5

10

15

30

1

2

4

6

10

24

0.12

0.22

0.34

0.45

0.67

1.00

1.48

2.17

2.75

3.70

6.00

0.15

0.25

0.38

0.48

0.69

1.00

1.42

2.02

2.46

3.23

4.90

0.18

0.27

0.41

0.51

0.71

1.00

1.36

1.86

2.25

2.86

4.30

0.21

0.29

0.43

0.54

0.73

1.00

1.33

1.77

2.12

2.62

3.60

0.24

0.31

0.46

0.56

0.75

1.00

1.30

1.71

2.00

2.40

3.35

0.27

0.33

0.48

0.58

0.76

1.00

1.27

1.64

1.88

2.24

3.10

0.30

0.34

0.49

0.59

0.77

1.00

1.25

1.57

1.78

2.12

2.84

0.33

0.35

0.50

0.61

0.78

1.00

1.23

1.53

1.73

2.04

2.60

0.36

0.36

0.51

0.62

0.79

1.00

1.22

1.48

1.67

1.90

2.42

0.39

0.37

0.52

0.63

0.80

1.00

1.21

1.46

1.62

1.82

2.28

0.42

0.38

0.53

0.64

0.81

1.00

1.20

1.42

1.57

1.74

2.16

0.45

0.39

0.54

0.65

0.82

1.00

1.19

1.38

1.51

1.68

2.03

Table 2. Values of Z1 for rainfall duration D and ratio r

The procedure to calculate rainfall depth for a storm shorter or longer than 60 minutes is:

  1. From Table 3 determine the rainfall ratio r for the location of the permeable pavement (interpolating between contours). r is the ratio of the sixty minutes to two day rainfalls for a five years return period.
  2. Use r in Table 2 to determine Z1 for the calculation of the 5-year return period rainfall total, M5-D min, for different storm durations, D.
  3. Use the following formula to determine the depth of rainfall occurring for rainfall duration D:

M5-Dmin rainfall = M5-60min rainfall x Z1

The average rainfall intensity i is obtained by dividing the rainfall depth by the duration. Alternatively, a value of 180litre/second/hectare can be used as a matter of convenience.

City

r-value

Cambridge

0.45

London

0.45

Norwich

0.42

Birmingham

0.39

Bristol

0.39

Liverpool

0.39

Nottingham

0.39

Sheffield

0.39

Southampton

0.39

Belfast

0.33

Cardiff

0.33

Leeds

0.33

Manchester

0.33

Newcastle

0.33

Plymouth

0.33

Edinborough

0.27

Aberdeen

0.24

Glasgow

0.24

Table 3. Ratio of 60 minute to 2-day rainfalls of 5-year return period, r-values for some UK cities.

 

ASSESSMENT OF GROUND CONDITIONS

The specification of a permeable pavement structure depends upon the hydraulic and traffic loading characteristics and upon the properties of the subgrade, the ground directly beneath a pavement. Strength and permeability of the subgrade are interrelated - a wet subgrade is usually a weak subgrade.

The following tests are recommended on the soil samples, especially if the soil has clay content. These assist in evaluating the soil’s suitability for supporting traffic in a saturated condition while exfiltrating.

  1. Soil classification
  2. Moisture content in percent.
  3. Soaked CBR

For most UK soils, the maximum exfiltration available is 3.7 x 10-3 mm/sec (37 litre/second/hectare). This figure should be compared with the 5 years return period UK rainfall requirement of 180litre/second/hectare. This indicates that most UK pavements will be required to have a water detention capability.

Soils with tested permeability equal to or greater than 37litre/second/hectare usually will be gravel, sand, silty sand or other coarse grained materials. These are usually soils with no more than 10-12% passing the 75micron sieve. Silt and clay soils will have a lower permeability, and may not be suitable for full exfiltration from an open-graded roadbase.

Soils with permeability lower than 37litre/second/hectare can be used to infiltrate water as long as the soil remains stable while saturated, especially when loaded by vehicles. Pedestrian applications not subject to vehicular traffic can be built over soils with a lower permeability.

Table 4 shows the permeability of soils using the Unified Soil Classification System (USCS). It also shows typical ranges of Californian Bearing Ratio (CBR) values for these classifications.

USCS

Soil Classification

Coefficient of Permeability k, (m/s)

Relative permeability when compacted and saturated.

Shearing strength when compacted

Compressibility

Typical CBR range

GW-well graded gravels

10-5 to 10-3

Pervious

Excellent

Negligible

30-80

GP-poorly graded gravels

5x10-5 to 10-3

Very pervious

Good

Negligible

20-60

GM-silty gravels

10-8 to 10-4

Semi pervious to impervious

Good

Negligible

20-60

GC-clayey gravels

10-8 to 10-6

Impervious

Good to fair

Very low

20-40

SW-well graded sands

5x10-6 to 5x10-4

Pervious

Excellent

Negligible

10-40

SP-poorly graded sands

5x10-7 to 5x10-6

Pervious to impervious

Good

Very low

10-40

SM-silty sands

10-9 to 10-6

Semi pervious to impervious

Good

Low

10-40

SC-clayey sands

10-9 to 10-6

Impervious

Good to fair

Low

5-20

ML-inorganic silts of low plasticity

10-9 to 10-7

Impervious

Fair

Medium

2-15

CL-inorganic clays of low plasticity

10-9 to 10-8

Impervious

Fair

Medium

2-5

OL-organic silts of low plasticity

10-9 to 10-6

Impervious

Poor

Medium

2-5

MH-inorganic silts of high plasticity

10-10 to 10-9

Very impervious

Fair to poor

High

2-10

CH-inorganic clays of high plasticity

10-11 to 10-9

Very impervious

Poor

High

2-5

OH-organic clays of high plasticity

Not appropriate under permeable pavements

PT-peat, mulch, soils with high organic content

Not appropriate under permeable pavements

Table 4. Suitability of soils for infiltration of storm water and bearing capacity.

SPECIFICATION & STRUCTURAL DESIGN

The aggregate roadbase should have a porosity of at least 0.3 to allow void space for water storage. The structural strength of the materials should be adequate for the loads to which it will be subjected. The aggregate roadbase should be in accordance with either:

BS882:1992. "Specification for aggregates from natural sources for concrete". British Standards Institute, London

or:

BS1047:1983. "Specification for air-cooled blastfurnace slag aggregate for use in construction". British Standards Institute, London.

In the case of natural aggregate, the roadbase should comprise coarse graded crushed rock meeting the following requirements. The flakiness index, shell content and mechanical properties should be as set out in BS882 for coarse graded crushed rock. The 10% fines value should be 100kN or more. When tested in accordance with 7.2.1 of BS812: Section 103.1:1985, the amount of material passing the 75 micron sieve should not exceed 1%. In the case of blastfurnace slag, the material must be proven to be equal to or superior to the above in all respects.

Providing the above criteria are met, the roadbase material will have a porosity of at least 0.3 and a storage capacity in its voids (volume of voids/volume of roadbase) typically of 30% to 35%. A 30% void space means that the volume of the roadbase will need to be 3.33 times the volume of the water stored. The infiltration rate through 20mm graded crushed rock roadbase is over 70,000 litre/hectare/sec and this should be compared with the normally required value of 180 litre/hectare/sec.

To avoid the loss of laying course material into the roadbase, a laying course material which will not invade the surface of the roadbase should be used. The NUROLF trials indicated that Cloburn 6mm washed crushed micro-granite performed satisfactorily in this respect (available from Cloburn Quarry Company Ltd, Lanark, Scotland, ML11 8SR). The Cloburn material has the following properties which should be regarded as minimum acceptable values for alternative materials:

10% Fines Value 370kN (150kN or greater recommended)

Aggregate Crushing Value 14%

Aggregate Impact Value 10 (15 blows)

Plasticity Non-plastic

A 4mm washed natural gravel performed less well and should not be used in permeable paving. It can be presumed that material having similar geological and mechanical characteristics, particularly grading, will perform similarly. The 6mm Cloburn material (or similar) should be used for the jointing material. The NUROLF trials indicated that such material can be introduced into the joints using conventional paver installation technology.

DESIGN THICKNESS OF ROADBASE FOR STORM WATER STORAGE

The depth of rainfall occurring during a 60 minutes storm recurring every five years in the UK is taken to be 20mm. Table 2 gives values of Z1 which is the ratio of the depth of rainfall occurring in a given period divided by the depth of rainfall occurring in 60 minutes. It is recommended that permeable pavements be designed to store rainfall occurring during 24 hours, unless it can be proven that sufficient exfiltration can occur to ensure that the maximum storage required can be reduced to that required to store rainfall occurring in 6hr. The 6hr thicknesses should be used only when the subgrade has a Coefficient of Permeability (k) exceeding 10 -6 m/sec i.e. when the subgrade comprises sand or gravel and it is intended that the water entering the roadbase can exfiltrate into the subgrade. In some pavements, there may be sufficient surface or sub-surface drainage provided to allow the 6hr figures to be used.

Table 5 shows thicknesses of crushed rock roadbase required to store either 6hr or 24hr rainfall levels. Table 5 is derived using the figures from Table 2 and by assuming that 32% of the roadbase comprises void. Also, it is assumed that only the lower 60% of the voids in the roadbase should be saturated and that the upper 40% should comprise air. Note that the thickness of roadbase required depends upon the factor r, the ratio of a 60 minute storm rainfall depth to the 2-day maximum rainfall depth and this varies throughout the UK as shown in Table 3.

Ratio 24 hours rainfall to 60 minutes rainfall (r)

Roadbase thickness to accommodate 6hr rainfall (mm)

Roadbase thickness to accommodate 24 hr rainfall (mm)

0.12

275

600

0.15

250

500

0.18

225

425

0.21

225

350

0.24

200

325

0.27

200

300

0.30

175

275

0.33

175

250

0.36

175

250

0.39

175

225

0.42

150

200

0.45

150

200

Table 5. Thickness of permeable pavement roadbase required to ensure sufficient storage capacity. Thickness ensures upper 40% of roadbase remains unsaturated. Note that the thicknesses shown may need to be enhanced to ensure adequate structural performance.

STRUCTURAL DESIGN PHILOSOPHY

Permeable pavements contravene many of the traditionally accepted principles of pavement design. In particular, one of the objectives of a conventional pavement is to create an impermeable surface so that moisture ingress cannot weaken components of the pavement or the underlying subgrade. Many highway pavement specifications are predicated upon the requirement to keep the specified materials dry. The deliberate cascading of water through highway construction materials requires a radical approach to the selection of material thickness and properties. This impacts two areas of design. Firstly, an alternative approach is required for the assessment of loading. Secondly, material properties need to be selected taking into account the flow of water vertically downwards and the retention of water within the material. This means that the traditional structurally beneficial effects of fine materials will have to be foregone and an alternative methodology will be required to ensure stability, strength and durability.

Traditionally, highway pavement loading has been assessed in terms of the number of 8000kg Equivalent Standard Axles (ESA's) which a pavement will be required to withstand throughout its life. The loads applied to a pavement usually differ significantly from 8000kg but research has shown that axles of other load values can be equivalenced to standard ones. The Fourth Power Law is often used to equivalence a given axle load to a standard axle. In the case of permeable pavements, an alternative approach is required: one which assesses loading in terms of the maximum load which a pavement can be expected to withstand.

The reason for this alternative approach is that permeable pavements are designed on the basis of ultimate limit state analysis rather than serviceability limit state analysis. Conventionally, a pavement fails by becoming progressively unserviceable - by developing ruts progressively for example. A permeable pavement, on the other hand needs to be designed to ensure that it is stable. An underdesigned permeable pavement could fail catastrophically when a load was applied. This is because the materials used in the structure have less stability than those used in a conventional pavement. For this reason, the conventional ultimate limit state design approach is adopted. In this approach, firstly loads are predicted and are multiplied by a load safety factor which reflects the degree of accuracy of the prediction. Secondly, material strength is measured and is divided by a material safety factor which reflects the level of consistency which can be expected for that material.

The NUROLF results indicate that even when water is cascading through the pavement, the materials used in the test can withstand successive axle loads of 6000kg. This result is used as the basis for the derivation of the following design approach.

LOADING ASSESSMENT

Firstly, levels of traffic loading need to be assessed so that the pavement can be placed into one of four load categories as shown in Table 6:

Load Category

Maximum Axle Load Anticipated (kg)

Category 1 - Domestic (GVW = 2000kg)

1000

Category 2 - Light (GVW = 3500kg)

2000

Category 3 - Commercial (GVW = 7500kg)

5000

Category 4 - Heavy (GVW = 44,000kg)

11,000

Table 6. Classification of vehicles

 

Now take the load appropriate to the load category and multiply it by the Load Partial Safety Factor from Table 7.

Level of Certainty of Load

Load Partial Safety Factor

Certain

1.0

Well informed value

1.2*

Anecdotal information

1.5*

Table 7. Load Partial Safety Factors. (* For Category 4 vehicles, maximum Load Partial Safety Factor = 1.1)

Now proportion the pavement section from Table 8.

Factored Load (kg)

Course Thickness (mm)

Cement stabilised open graded crushed rock

Open graded crushed rock

1,400

-

150

1,600

-

150

2,000

-

175

2,800

-

200

3,200

-

250

4,000

-

300

6,000

-

350

8,000

150

150

10,000

200

150

12,100

300

150

Table 8. Pavement course roadbase design thicknesses. Note these need to be adjusted for ground conditions and for Material Partial Safety Factor.

If the subgrade CBR is greater than 5%, the above roadbase material can be installed directly above the subgrade. In poorer ground conditions, a conventional DTp Type 1 granular sub-base should be installed between the subgrade and the roadbase. In most design situations, an impermeable membrane should be provided between the roadbase and the sub-base. The thickness of the sub-base is shown in Table 9. When sub-base thickness exceeds 150mm, the additional thickness can be provided by capping material whose CBR should be 15% or more.

 

Subgrade CBR (%)

Thickness of DTp Type 1 sub-base material (mm)

>5

0

5

150

4

250

3

350

2

600

1

Subgrade improvement required

Table 9. DTp Type 1 sub-base thickness required. Note: when sub-base thickness exceeds 150mm, the additional thickness may be provided by capping material.

 

Finally, apply the Material Partial Safety Factor as follows. The stability of the open graded crushed rock material should be assessed according to Table 10 and the thickness of this course should be multiplied by the appropriate factor from Table 10.

Nature of open graded crushed rock

Material Partial Safety Factor

As stable as DTp Clause 803 material ("Type 1")

0.9

As stable as graded 20mm crushed rock to BS882

1.0

As stable as rounded 20mm graded gravel to BS882

1.3

Table 10. Open graded crushed rock thickness adjustment for Material Partial Safety Factor

DESIGN EXAMPLE

Consider a pavement on 3% CBR subgrade with a k-value of 10 -7 m/sec to be designed as a car park for vehicles a Gross Vehicle Weight of 4600kg. A 20mm open graded crushed rock is available. The roadbase is to be designed to retain rain falling throughout 24hr in Belfast.

Firstly, consider the thickness of the roadbase from a water storage standpoint. From Table 3, the r-value for Belfast is 0.33. From Table 5, the roadbase thickness required is 250mm from a hydrological standpoint.

Now consider the structural design. From Table 6, the loading falls into Category 3 (5000kg axle load). The load value is Well Informed rather than Certain or Anecdotal. Therefore from Table 7, use a load partial safety factor of 1.2. The design load is now 5000 x 1.2 = 6000kg. From Table 8, the design thickness of the open graded crushed rock roadbase is 350mm and there is no requirement for a cement stabilised open graded crushed rock roadbase.

The thickness of the open graded crushed rock roadbase is now adjusted by the Material Partial Safety Factor in Table 10. Assuming that the material is as stable as single sized crushed rock, select a Material Partial Safety Factor of 1.0 from Table 10. The thickness of the open graded crushed rock is:

350 x 1.0 = 350mm.

This is greater than the value of 250mm required for water storage so the design thickness is 350mm.

Because the CBR of the subgrade material is 3%, a DTp Type 1 granular sub-base material is required. From Table 9, the sub-base thickness is 350mm (the lower 200mm thickness can be a lower specification capping material).

Therefore, the full pavement section is:

Permeable Pavers

50mm Cloburn 6mm washed crushed red micro-granite

350mm Coarse graded crushed rock to BS882:1992

Waterproof membrane (e.g. 1000gauge Visqueen)

350mm DTp Type 1 sub-base material or 150mm Type 1 over 200mm capping

3% CBR subgrade