DCP pavement design and charts

DCP pavement design and charts


Introduction
This is a very simplified introduction to the use of the DCP. The DCP can be used at all levels from simple investigation to very sophisticated design. It can be integrated with FWD testing, linear elastic layer design (CIRCLY, ELSYM, LED). It’s a great tool for catalogue design (TRH 4, AustRoads). I know the South Africa and Australian design systems well, the British and American design systems reasonably well, and the DCP adds value to all of them. Great value. The DCP can be used on roads (of all types includes cemented layers - just drill through the cement), and runways (up to and included PCN 100 pavements for 747s). As well as carparks, container terminals, building foundations. About the only thing that it doesn't do well is open coconuts. I was on Christmas Island, and tried to open a freshly fallen coconut with the DCP - coconut stayed intact and I broke the DCP.

The following words are intentionally simple so they can be read by people who have English as second language. The DCP can be described in very difficult and complex language just as easily, so do not be fooled by the simple words.

The DCP results from the analysis described below can be used to divide the runway into uniform sections, get design subgrade strengths and thus feed into FAA design charts, or to estimate elastic moduli and feed into the various (APSDS, LED) computer elastic layer design programmes for runway thickness. They can also be used to help understand FWD and HWD test results.

Above all, I find the DCP is absolutely marvellous for sorting out the confusion that arises when all the various test results are put together. Usually nothing makes sense, and one test shows a section is about to fail, while another test shows this is the strongest section. The DCP is the "honest" tool to help sort out the mess.


DCP and pavement investigation
The DCP is a most useful tool for pavement investigation because it can 'look inside' a road, a runway or any other pavement. It shows field bearing strength which is the most important material property for roads.

Gradings, PI, and Atterberg limits are other tests used, but these are indirect or indicator tests and are less relevant than bearing strength. Because it is a quick, inexpensive pavement test, and because it relates very directly to the performance of a surfacing, DCP testing is often the main one suggested for use.



Effect of moisture on DCP and CBR results
Most road/runway materials become weaker if they become soaking wet, particularly clayey materials. In the field, the materials in the road/runway are usually drier than soaking wet. The DCP measures field CBR, and this is probably higher than a laboratory soaked CBR test. So a dry clayey gravel may have a good DCP-CBR (over 80), while its soaked CBR could be 45. The effect is most pronounced in the basecourse, because the lower layers are generally wetter than the basecourse.

Most of the time, this moisture effect does not matter much because the bitumen surfacing keeps the road/runway dry inside, even when it is raining.  But if water does get into the road/runway from underneath (if it is a wet area), or from on top (if the surfacing is cracked or potholed), it could matter.

Climate has an effect as well - in a wet climate (where the grass is green all year round, there are lots of lovely trees, rivers, ponds, etc), then the materials tend to be wetter. The DCP tests should be done in the wet season when the materials are at their weakest. The DCP tests should be compared to laboratory soaked CBR test results to see what the effect of soaking is. In a dry country, this is less important.

There are two three methods to estimate what the material will be like when it is soaked:
• take a sample, and do a laboratory soaked CBR test, or
• check the Plasticity Index (PI; from the Atterberg Limit tests). Low PI materials (PI < 6) will not weaken too much when wet, so field DCP is a good measure of soaked CBR. Higher PI materials will weaken when wet, so field DCP could be deceptively high.
• See the appendix to his document


DCP test frequency
DCP tests should be done at the following frequency:
• Upgrading a road/runway 2-4 DCP tests per kilometre, with the tests staggered as outer wheeltrack/inner wheeltrack one side, outer wheeltrack/inner wheeltrack other side, centreline, etc; at least 8 DCP tests per likely uniform section are needed to provide adequate data for the analysis.
• Failure investigation  2-4 DCP tests in the failed section, and the same in a nearby unfailed section.

It is useful to take at least 2 samples per kilometre for new roads, or 2 samples per section for failure investigations, to check laboratory soaked CBR, Atterbergs and insitu moisture content of each layer.  Also test the insitu basecourse density (optional for new roads, needed for failure investigations). Local soils laboratories can do all these soil tests.

In any event, never come back from the field without a heap of DCP tests been done. They are quick to do, and even half a day will give you 16 tests.


DCP test method
Assemble the DCP.  Place the DCP point on the surface, hold the DCP upright and start. Lift the hammer to the top of its travel, and release; do not throw it down, but let it drop. Record the depth of penetration every 5 blows. Continue to below 800mm below the surfacing.  Hitting a stone/rock will show as a horizontal line for 5-25 blows. When this happens during testing, just keep hammering with the DCP and usually it will break through.

Plot the depth versus blows directly on the sheet. Direct plotting guards against minor errors in reading.  Since the sheet only reads 50 blows across, transfer the depth at 50 blows across to the left hand side of the sheet and keep recording down


Drawing the DCP lines
Essentially, join up the dots on the DCP sheet to show continuous lines. It is helpful to draw this as only a few straight lines of "best-fit". The CBR at each depth can be read off the chart directly. Since most pavement layers are 150mm thick, it is usually possible to interpret the DCP line into actual pavement layers.


Use of DCP computer programmes
The use of DCP computer programmes is not recommended except by extra highly qualified and experienced pavement engineers. All current programmes necessarily have built-in assumptions, and these can be changed by experienced users. However if left unchanged, the programme will still operate and can make gross mistakes. The hand-drawn DCP charts are much safer to use. I only use hand drawn charts today, even though I have a number of the computer programmes available.


Pavement strength analysis

The DCP results can be used for determining uniform pavement sections, design subgrade strength, design strength of unbound (unstabilised) sub-bases and basecourses. It can even be used for quality control during construction of non-standard materials. The following steps give some illustration of how some designers use them.


STEP 1 DO DCP TESTING ALONG THE ROAD/RUNWAY
DCP testing is performed along the length of road/runway.  The frequency of tests should generally be in accordance with the standards here, but the visual inspection may indicate adjustments to the frequency.  If the road/runway is very uniform the frequency can be reduced, and if it is variable then it should be increased.  The basic frequency should be:

• test at the rate of 5 DCP tests per kilometre, with the tests staggered as centreline, outer wheeltrack one side, outer wheeltrack other side, centreline, etc;
• perform an additional test at every significant location picked up in the visual survey, such as particular failure areas;
• ensure that at least 8 DCP tests are performed per likely uniform section to provide adequate data for the statistical analysis.


STEP 2 DIVIDE ROAD/RUNWAY INTO UNIFORM SECTIONS
The results of the investigation, including the DCP testing and visual assessment, enable the length of road/runway to be divided in relatively uniform sections for the purposes of rehabilitation.

The minimum length of section should be 100 metres, and desirably 1000 metres. On long lengths of road with uniform conditions, it can be 10 000 metres.  Note that construction of sections shorter than 500 metres is awkward.  It may be that a low DCP result occurs in a spot which was identified in the visual survey as an isolated problem area; these are typical of an isolated drainage problem and consideration should be given to repair of these individually rather than taking them as representative of the section.

For runways with heavy aircraft, the design pavement thickness can be quite a lot, especially if the subgrade is weak. Very often on existing runways, the subgrade strength (and depth of existing pavement layers) is amazingly variable.  It is even more variable if you test outside the central 15 metres!  It is not unknown to have outer edges of runways with only half the strength of the centre of the runway - even though they have been built to the same thickness. The trafficking and compaction in the centre make it so much stronger there. So the DCP can be used to delineate the weakest areas. Then selective reconstruction can be used to fix those areas.

As an example, at Broome International Airport the 10/28 runway PCN was 28. Parts of the runway were even weaker than that and had failed. However by reconstructing 300m of the 2000m, the weakest areas and failures were replaced. The whole runway could be re-rated to PCN 35. Then an thin asphalt overlay could be used to bring the PCN up to 45.  The alternative was a thick asphalt overlay over the whole lot - the weak, failed and good areas of the runway. The cost savings amounted to 70% of the cost of the thick overlay.


STEP 3 CALCULATE THE DESIGN CBR VALUE FOR EACH LAYER IN EACH SECTION
The design CBR is found for each layer (such as subgrade, subbase or basecourse) in each section and calculated statistically to provide a safety margin against the variability of material within the section.  A normal distribution of data is assumed and the Student's T distribution at the 80% level is used:

First the design DCP is found:

Design DCP = average DCP  - .9 * (standard deviation)

Example   The DCP/CBR results of a basecourse in a section were as follows:
DCP/CBR:  125,  143,  120,  100,  145,  115,  140,  135
Mean (average) = 127,9       Standard deviation = 15,7
DCP = average DCP  - .9 * (standard deviation)
= 127,9 - .9 * 15,7   = 114
Note that the equation is using a one-tailed T-distribution for 8 samples and is reasonably robust for sample sizes from 5 to 30.

Then the design CBR is found by considering the relationship between CBR and DCP.  This is rather too complex to cover here. Instead, let me say what some designers do. In Australia, for drier areas, the DCP is taken as equal to the CBR for the subgrade and sometimes for the upper layers.  Easy, for dry areas.

In South Africa, they'll do a CBR test in the laboratory at soaked and at unsoaked conditions, and see how the strength varies with moisture. Then they'll consider how wet the inside of the pavement will get, before they pick the relationship between DCP and CBR. Often, a simple rule of thumb for areas in moderate climates is to reduce the DCP by 20% - so a design DCP of 80 becomes equivalent to CBR 60. The answer also depends on the assumptions of the design method being used, whether they are based on soaked or insitu strength, and how that has been handled in determining elastic moduli, etc. No easy answer there, sorry. But at least it keeps me in consulting work !



APPENDIX

Extract from:
THE PREDICTION OF MOISTURE CONTENT IN UNTREATED PAVEMENT LAYERS AND AN APPLICATION TO DESIGN IN SOUTHERN AFRICA
Division of Roads and Transport Technology, Bulletin 20, CSIR Research Report 644, Pretoria, 1988
Author: S. J. Emery

Table 22  Variation of CBR with moisture content

Material class    Soaked                  mc/OMCm ratio
(TRH 14)          CBR (%)         1,0           0,75        0,5
                                  Ratio of unsoaked to soaked CBR
G4                  80            1,34          1,88        2,5
G5                  45            1,77          2,57        3,6
G6                  25            2,35          3,56        5,4
G7                  15            3,00          4,71        7,6
G8                  10            3,65          5,88        9,9
G9                  7             4,33          7,16        12
G10                 3             6,50          11,41       22

Notes
1. Material class is South African material class - other countries can use the Soaked CBR column to identify their materials.
2. mc/OMCm ratio is the ratio of field moisture content (mc) to optimum moisture content (OMC at Mod AASHTO compaction). Thus a material, which has been prepared in the laboratory for an unsoaked CBR test, is at OMC because it has been compacted at OMC, and so the mc/OMCm ratio is 1.00.  A material in the field with a field m.c of 6% and an OMC of 8% has an mc/OMCm ratio of 0,75.
3. So a material with a soaked CBR of 45 can be expected to have an unsoaked CBR (at OMC) strength of 45 * 1,77 = 80. The same material, in the field, dried back to say 6% with an OMC of 8% which means a ratio mc/OMC of 0.75, would have a field CBR of 45 * 2,57 = 115.
4. These relationships are approximate and variable, and they also depend on particle size distribution, soil suction and the amount of clay in the material, none of which are explicitly addressed in the modelling. However they give an indication of the strength gain with drying out.