|4.6 SAND AS IMPORTANT NAMIBIAN ROAD BUILDING
4.6.1 COASTAL DEPOSITS FROM TERTIARY TO RECENT
The deposits discussed in this section are of a variety of ages und origins. They are grouped together, because they predominantly consist of deposits originated from the action of littoral processes. Surely the most striking feature of the Namibian coast is the Namib Desert which forms a long, narrow zone west of the Great Escarpment between the Olifants River in the Cape Province in South Africa und the Carunjamba River in south-western Angola. This desert consists of spectacular aeolian deposits in the several sand seas with the highest sand dunes in the world which in places are climbing to 300 m of height. Less prominent aeolian accumulations of sand can be found banked up against " inselbergs" und in local basins. In spite of the frequent storms accompanied by heavy dust clouds no loess deposits are known. The largest accumulation of sand covers some 34.000 km2 between the Lüderitz/ Koichab River area in the south und the Kuiseb River/ Swakopmund area in the north. Large parts of this central sand sea lie on a red aeolian rock known as the Tsondab Sandstone Formation which is dated to the Tertiary age und rests on a platform cut into the Cretaceous erosion surface .
The mostly westwards flowing Namibian rivers, which are all crossing or ending in the Namib Desert und which are flowing only in the short rainy season, have a Tertiary to Quaternary fill consisting of both coarse und fine fluviatile sediments und thick calcrete-cemented sands und conglomerates. Some gypcretes occur also in the western parts of the country. Several of these rivers have large fluvio-marine deltas which, in some places, contain layers of salt und gypsum-cemented sand.
The main components of the Namib dune sands consist of subrounded to sub-angular grains of quartz (70-90%) with some minor quantities of felspar (10-15%) and some heavy minerals . Surface sands are generally dry but become slightly moist with depth. The colour of the sand varies from yellowish-brown at the coast to red in the eastern sector of the Namib. The consistency is loose on the surface becoming medium dense with depth. Dune sands are normally not uniformly graded (bunch-graded) and fine to medium grained without any cohesion. They are finer than Kalahari sands but the -0,075 mm fines are absent due to the continuous wind action in the Namib Desert. See figure 1 at the end of this section which compares the gradings between a typical Namib sand with that of a Kalahari sand.
4.6.2 INTERIOR DEPOSITS FROM TERTIARY TO RECENT
The aeolian deposits of Namibia's interior all belong or have been derived from the uppermost formation of the Kalahari Group. The Kalahari Group comprises roughly the eastern third of Namibia and stretches from Owamboland/ Okavango/ Caprivi Strip in the north via Bushman/ Hereroland and Gobabis district in the central region to the eastern parts of the Mariental/ Keetmanshoop/ Karasburg districts.
In the northern sector, in Owamboland and Okavango, the Kalahari Group varies in thickness from 225 to 500 m . Deep borehole investigations indicate a succession which can be divided into a basal Beiseb Formation consisting of 15 to 30 m of reddish gritty or conglomeratic sandstone, more than 120 m of poorly consolidated reddish brown calcareous sandstone of the Olukonda Formation and the uppermost Andoni Formation consisting of greyish green sand or clayey sand with sandy facies towards the margins of the basin.
In the central sector of the Kalahari Group, in Bushmanland, Hereroland and the Gobabis district, three formations can be found. The lowermost formation of carbonate-cemented sandy conglomerate, grit and variably cemented sand is known as the Tsumkwe Formation. The overlying silcrete-cemented quartz sand, silicified limestone and calcareous sandstone belong to the Eiseb Formation, which is again capped by ferricrete and ferruginous sandstone of the Omatako Formation .
In the southern sector of this Group, in the Mariental, Keetmanshoop and Karasburg districts, well exposed formations have been grouped into a single stratigraphic formation, the Wei&rand Formation, where a basal conglomerate is overlain by grit and sandy limestone. The overlying sands of the Kalahari Group are currently stabilised except at the edges of rivers and omurambas where they are still mobile and a source of fresh sediments. These sands are mainly stabilised by vegetation, which in the north-eastern parts of Namibia is often thick bush or even semi-tropical forest.
The sedimentological characteristics of the Kalahari sands were investigated by different researchers . Four different types of sand can be differentiated, three can be encountered in Namibia:
The dominant particle sizes of the Kalahari sand vary in the different dune sections in relation to the variable wind velocities across the dune profile. The finer particles can be found more prominently in the interior of dunes where they penetrate into the voids between coarser particles and are protected from further wind transport. These fine sand particles, with iron oxide skins on grain surfaces and a small but significant clay component formed by in situ weathering of felspathic and ferromagnesian minerals, are responsible for the development of a collapsible fabric in some of the Kalahari sands.
Another property of the Kalahari Group is the low permeability of the Kalahari sands, and the low rainfall in this region is responsible that the groundwater aquifers are not recharged by rainfall where the sand cover exceeds about 15 m. The good aquifers found in the lower units of the Kalahari Group are recharged from other, distant sources as for instance by the artesian conditions which can be found in the area of Stampriet. The high concentration of soluble salts, mostly sulphates and chlorides, in these deep-seated aquifers, do render them in some cases unsuitable for use in reinforced concrete and even in the compaction of the selected road pavement layers, although salt was not controlled in subgrade layers and rendered no problems as long it was prevented to penetrate into the basecourse and the surfacing of the road as has been proved on main road 61 between Stampriet and Aranos. The same fact has been proved during the construction of trunk road 4/2 from Lüderitz to Haalenberg when seawater has been used for compaction purposes in subgrade layers. The soluble salts also didn't render any problems in the application for reinforced concrete as for instance during the building of bridge 412 over the Auob River at Stampriet.
One remarkable property of aeolian deposits lies in the fact that these deposits could possibly have a collapsible fabric . Such a collapsible fabric is normally associated with an open textured soil with individual grains being separated by a bridging material. In the Namibian parts of the Kalahari this bridging material often consists of calcium or gypsum, like, for instance, in the Gobabis area. Available data have been used in an attempt to obtain a relationship between the collapse potential index and a combination of easily determined properties such as moisture content, dry density and void ratio. To date no meaningful conclusions have emerged due to the small amount of available data and due to the complexity of the mechanism associated with the collapse phenomena of aeolian sands of the Kalahari Group.
The Kalahari sands have under a partially saturated condition a relatively high shear strength because of the effects of " apparent cohesion" imparted to the sand by pore water suction. Studies were carried out  during the construction of the road from Bulawayo to Victoria Falls in Zimbabwe over Kalahari sands which indicated that the composition properties of the aeolian sands in this area are dependent on the Plasticity Index 'PI' (LL - PL) and the percentage finer than 0,075 mm sieve. It was prevailed that when the sand is non-plastic a CBR of the order of 30 is obtained, but with an increase of the PI the strength of the sand was found to decrease to as low as a CBR of five.
Many problems have been encountered with pavements of roads and airfields in most areas of the Kalahari Group. Distress of flexible pavements were reported by several investigators . This distress is mainly caused by the densification of the aeolian sand subgrade which collapses under compaction by adding a road fill onto the subgrade and under the vibration of the road traffic due to the existence of collapsing bridging materials. This distress is associated with water penetration as well as with the static and vibratory traffic loads. Tests indicate  even under light traffic loads, collapse could occur to a depth of 700 mm in an aeolian soil subgrade. In the case of the presence of collapsible aeolian sands it is recommended that  the roadbed should be compacted to achieve 90% Modified AASHO for a depth of between 0 to 0,5 m and 85% Modified AASHO for a depth of 0,5 to 1,0 m. Several reports of the use of vibratory rollers are known . The results of these reports indicate that despite of using various techniques like the adding of water, ripping before compaction to break down the clay bonds, it is generally not possible to improve the in situ density below a depth of about 1 m. It is further reported that it is possible to achieve densities in excess of 95% Modified AASHO to a depth of 0,8 m and densities of 90 to 95% Modified AASHO between 0,5 and 1,0 m.
Experience gained by the Namibian Department of Transport proved, however, that sufficient bearing capacity without collapse of aeolian sand subgrades can be only achieved with compaction to 100% Modified AASHO. This density can be achieved by means of vibratory or even static rubber wheel compactors as well as the impact roller. The disadvantage of the impact roller is, however, that it needs a platform of gravel on top of the Kalahari sand before it can be used. This has been proved by trial runs of the impact roller on the Gobabis hospital complex during 1975 and during the construction of main road 61 between Stampriet and Aranos during 1976/78. The easiest way to achieve the required 100% Modified AASHO on collapsible sands is the use of the 50 t vibrating " super-compactor" which has been used during the construction of the airport Gobabis. 100% compaction has been achieved to a depth of 1 m and more. After more than 15 years of service of above two projects no distress due to collapse of Kalahari sands has been encountered.
The relatively high strength of Kalahari sands in comparison with Namib sands is due to their more uniform gradings and -0,075 mm fines fraction (up to 10% and sometimes even more) with the consequent higher CBRs. This part of the material is essential to attain, for example, stability in a gap graded asphalt mixture. See figure 1 at the end of this section which compares the gradings between a typical Namib sand with that of a Kalahari sand. Although the PI on the fraction passing the 0,425 mm sieve is very low (mostly non plastic), the PI obtained on the fraction of the Kalahari sand passing the 0,075 mm sieve can be up to 15. This clay fraction give some of the Kalahari sands a very good binding property.
Most parts of the areas covered by aeolian sands of the Kalahari Group experience a scarcity of other natural construction materials of high quality. The Department of Transport has carried out some studies and constructed some trial sections to determine suitable stabilisation methods to improve the properties of the aeolian soils sufficiently for them to be used as selected layers, subbase or even basecourse in the construction of paved roads. These studies indicated that the stabilisation of aeolian sands with bitumen or tar seemed to be the most promising procedure to improve the properties of Kalahari sands to basecourse quality materials. One example is the construction of the test section, the "holy mile" north of Andoni on trunk road 1/11 from Oshivelo to Ondangwa in Owamboland during 1965. Different stabilising agents have been used on this test section, like straight bitumen, various cut-back bitumens and coal tar. Due to the presence of pedocrete materials in most parts of the Namibian Kalahari it was always possible to use these materials before it was considered to stabilise Kalahari sands which was regarded to be uneconomical. The only exception is trunk road 8/6 from Kongola to Katima Mulilo ( Trans-Caprivi Highway) where a mixture of Kalahari sand and calcrete nodules has been used which was stabilised with bitumen. The subbase on this road has been constructed from sand which has been compacted to 97% Modified AASHO except one test section where the sand-subbase has been stabilised with 3% lime. The subgrade layers consisting of unstabilised Kalahari sand has been compacted to 100% Modified AASHO. In spite of the heavy army traffic until the middle of 1989 no distress, except some edge failures, has been experienced to date.
Particular attention must be also given to erosion protection and the correct design parameters for crossections of roads in Kalahari areas for both fills and excavations. The Namibian Department of Transport is using a side slope of 1:6 and critical sections are covered with a coarse gravel blanket to prevent erosion, as has been shown successfully on main road 61 from Stampriet to Aranos. Single size Kalahari sands have been even used as fine concrete aggregate as for instance for the construction of bridge 236 over the Nossob River on main road 41 between Leonardville and Zania because no river sands are obtainable in economic distances. Kalahari sand has been used as fine concrete aggregate for drainage structures on main road 75 between Tsumeb and Tsintsabis as well as for various smaller structures in the Tsumeb and Otjitue areas. In both cases this sand had to be blended with crusher dust to achieve the specified grading.
FIGURE 1 GRADINGS OF TYPICAL NAMIB AND KALAHARI SANDS
4.7 THE USE OF PEDOCRETES IN NAMIBIA
Pedocrete materials are common deposits in most areas of Namibia and they can be regarded as the natural road building material number one in terms of importance. Pedocrete is a convenient general term used in southern Africa for a group of materials known as calcretes (surface limestones), laterites and ferricretes (surface ironstones) as well as silcretes (surface quartzites). This term is appropriate, since these materials are not rocks, but owe their origin at least in part to pedogenesis (soil-forming processes). Thus pedocretes are composed of materials of two origins, namely the host material and the authigenic cement (authigenesis is a process of chemical reorganisation). With the developing pedocrete the authigenic component can increase and ultimately replace almost the whole host material. It is even possible that one pedocrete can replace the other one with resulting mixtures between them. Several horizons of both host material and cementing agent can be encountered in one pedocrete occurrence. Therefore it is difficult to date and correlate pedocretes within the stratigraphic system. Pedocretes are not sedimentary rocks, but materials which have been formed either as weathering residues or by cementation or replacement - sometimes both - of pre-existing soils by various authigenic minerals .
In terms of consistency these materials are of two types: indurated (e.g. hardpans, honeycombs, nodules) and non-indurated (soft or powdery forms). Pedocretes can be either cemented or replaced by a rather big variety of pre-existing materials, as for instance by carbonates with the resulting calcrete; dolomites with the resulting dolcrete; iron oxides with the resulting ferricrete, plinthite or laterite; silica with the resulting silcrete; manganese oxides with the existing manganocrete; phosphate with the resulting phoscrete; gypsum with the resulting gypcrete or magnesite with the resulting magnesicrete. Only in cases where more than 50% of the cementing or replacing materials are existing in the new deposit, they should be referred to with the term pedocretes. After the formation of pedocretes they will go through a weathering process like any other rock. The resulting boulders, cobbles and gravels may be transported as colluvium or alluvium. They can also be incorporated into another younger pedocrete.
The Namibian types of calcretes rarely exceed thicknesses of 1-2 m. In the case of Kalahari limestone, thicknesses of up to 30 m are reported but in excess of above mentioned 1-2 m, they are seldom homogeneous. Silcretes of 5 m thicknesses are reported to have been found in southern Namibia, but the younger Pleistocene pedocretes seldom exceed 2 m in thickness. Calcretes have mainly a pedogenic and not so much a groundwater origin. The main component to form a calcrete, the carbonate, is transported through the soil layers by rain-water. The carbonate itself can be originated from the surrounding soil or it can be transported in the form of dust or by rain-water. Ferruginous pedocretes, however, are formed by a process of accumulation. Ferricretes, silcretes, gypcretes and phoscretes all appear to be absolute accumulations which are necessary for induration . Typical constituents of some different pedocretes are summarised in table 19 :
TABLE 19 TYPICAL CHEMICAL COMPOSITIONS OF PEDOCRETES
Soils classifications for non-pedocrete materials are differentiating between discrete, hard, durable and solid particles which can be classified according to their grain sizes and Atterberg limits. These criteria cannot easily be applied to materials which are weathered, cemented or aggregated because gradings and Atterberg limits are too dependent on the test methods employed and on the methods of excavation and processing the material. Therefore it is advisable to add two extra symbols to the general soils classification in the case of pedocretes, i.e. to show the geological origin and the method of excavation . The sequence in which calcretes are forming has led to a simplified morphogenetic classification of calcretes which also can be applied to other pedocretes .
The different classes of pedocretes can be defined as follows :
The geotechnical properties of pedocretes are depending on mainly three factors: the texture of the host material, the stage of development to a pedocrete and the nature of the cementing or replacing mineral . For example, the properties of calcareous soils are closely akin to those of the host soil, whereas hardpan calcretes essentially behave as limestone rock. During pedocrete development, clay and silt become flocculated and cemented into larger silt to gravel-sized complexes of varying strength and porosity. These particles or aggregations may or may not break down during compaction of a road layer.
One of the most striking properties of some pedocretes is their ability to undergo a process of self-stabilisation. This self-hardening phenomena of, for instance, hardpan calcretes has been known for long periods in Namibia because these self-hardening materials have been extensively used as building blocks. Two of the many more examples are the old police station and the Lutheran church in Grootfontein which have been built from hand-sawn hardpan-calcrete self-hardening building blocks.
The reasons for the self-stabilising effect, as, for instance, one or other chemical or even organic constituent, are still not clarified as it is still not known which types of calcretes are more prone to self-stabilisation than others. The research is still ongoing, for instance, at the " Division of Roads and Transport Technology" (DRTT), the former ' NITRR' of the ' CSIR'. As far it can be ascertained it is one or other chemical reaction which is responsible for this self-stabilising process.
Test methods for potentially self-stabilising materials include the soaked CBR test, carried out after allowing the compacted material to dry in the mould or after moist curing, or following wetting and drying cycles . It is, however, not possible as yet to predict the grade of self-stabilisation of these pedocretes and determine its long-time effect on the bearing ratio of this building material. Thus the long time self-stabilising effect on a road layer must be currently regarded as an added and uncertain safety factor which cannot be used to establish the ultimate bearing limit of such a road layer. But, it must be categorically stated that the self-stabilising phenomena has been proved with the aid of CBR strength tests in the laboratory of the Namibian Department of Transport. The CBR tests have clearly indicated an increase in strength by time-dependent self-stabilisation of the calcretes but the problem still prevails to prove this phenomena in-situ on the road. An in-situ accurate test procedure in order to prove the self-stabilising effect on the increase of the strength of calcretes must still be found.
One of many proved examples for the self-stabilising effects of calcretes is the following: During the prospection for road materials for subbase and basecourse layers for main road 61 between Stampriet and Aranos numerous trial borrow pits have been opened by an obsolete bulldozer which opened the calcrete borrow pits without any effort. Six months later, during the starting phase of the actual construction of the road, it was encountered that some of the bulldozed materials in the borrow pits had become hard as concrete. During this period the natural moist of the calcretes had dried out and a hardening, self-stabilising effect had taken place.
In Namibia pedocretes, especially calcretes, have been used extensively in road construction. Calcretes are mainly used for wearing courses for unpaved gravel roads and for all layers of the road prism of a paved road, including basecourses and even surfacing chippings. In Namibia pedocrete surfacing chips have been, however, never used due to their excessive absorption of bituminous binder. In bituminous premix calcretes have been used successfully as crushed premix aggregate, as, for instance, on trunk road 1/10 between Tsumeb and Andoni. This hard calcrete premix aggregate behaved well for more than 15 years under extreme high traffic loads before a re-overlay became justified, this time with a premix aggregate of Abenab-dolomite.
Reports  that duripan pedocretes have been used in the Namibian south for all road layers up to subbase cannot be confirmed. It is also not on record that dolcretes from Tsumeb have ever been used. Silcretes which are mainly found in the Kalahari area are used predominantly for concrete aggregates because as material for road layers and wearing courses for gravel roads it is too hard and not sufficiently workable. Silcretes are mainly encountered in the Rundu and Gobabis areas but also in the vicinity of the Omatako Mountains between Okahandja and Otjiwarongo. The road building properties of silcretes are very similar to those of quartzitic materials. Gypcrete compacted with water nearly saturated with salt makes excellent unpaved " salt-gypsum roads" along the Namibian Atlantic coast. Ferricretes have been, during 1964, successfully used as basecourse and subbase layers for trunk road 1/7 between Okahandja and Otjiwarongo in the vicinity of Sukses.
As wearing course for gravel roads calcretes are getting automatically the first choice because it is superior to any known alternative material. Even if calcrete wearing course materials are not in coincidence with the materials specifications, it has been proved that a high success rate still exists in contrast with most other road building materials which would fail under the same conditions. In general it can be stated that pedocretes as road construction material behaved better than has been originally predicted and could have been expected on grounds of their gradings and Atterberg limits. Failures and ruttings are rare and have generally been confined to areas of poor drainage of the road prism. But, pedocretes can also cause problems in road construction which the designer and builder of roads should be aware of to avoid serviceability limit failures of this otherwise excellent and very economic road construction material. The problem-properties which can also be present in other road building materials, can be listed as follows :
Other potential problems include those of alkali-silica, carbonate and alumina reactions, which are, however, not experienced to date in Namibia, unusual reactions with lime and cement, heave, collapse and small to medium scale karst phenomena. Predicting methods of some of these problems have been determined as a very general guideline and are summarised in table 20 :
TABLE 20 RECOGNITION OF PROBLEM PEDOCRETES
It must be, however, stated that pedocretes are in many cases difficult to locate. Aerial reconnaissance and interpretation of aerial photography for the prospecting of calcretes can be used . For in-situ field prospecting vegetation indicators, soil colours, and in the case of calcretes a simple probing device can be used  which has been successfully proved. The oldest, thickest and best developed pedocretes occur as cappings to remnants of the African erosion surface  which can normally be readily located. The younger pedocretes under the Kalahari sands, which are also the smallest and most variable of all the pedocrete occurrences, are much more difficult to locate. Methods of how to use aerial photography and other locating techniques were developed . Generally it can be stated that pedocrete materials like calcretes are occurring in situations where stereoscopic interpretation of aerial photographs are revealing fairly flat grounds with a collinear drainage pattern and, as indication of a poor drainage, a mottled, grey tone as well as on terraces, associated with seasonal drainage features such as pans, fossil drainage lines and ephemeral drainage lines. The quality of the calcretes generally improves with increasing distance from the centre of the drainage feature.
In-situ tests in the field can use the colour of the covering soils which reveals a clear indicator for the occurrence of calcretes for any experienced road materials prospector. Other indicators for calcretes are vegetation indicators like the gabbabos ( Catophractes alexandri), the salie or bitterbos ( Pecheul-Loeschea-leubnitziae) and the vaalbos ( Tarchonanthus camphoratus). The calcrete probe is a simple but valuable device which enables the prospector to ascertain the presence or absence of calcretes to a depth of up to about 6 m within half a minute . But it must be kept in mind that a technique which has been proved to have been worked well in one area may not necessarily work well in another, even if it likes that this area is similar to the former one. Field calibration of the photographic image of the aerial photograph should always be followed up as quickly as possible by field checking procedures like soils checks, vegetation indicators and field probing. Difficulties can be especially expected in the zone of the Kalahari Group where the overburden thickness above the sparse calcrete occurrences can be substantial and prospect depths to up to 5 m can be experienced.
In the future pedocrete road materials will continue to be Namibia's natural road construction material number one. With the increasing provision of roads in so far underprivileged areas, pedocretes will play a special role in the establishing of labour-intensive low-volume roads. New research efforts will have to be used to prospect for and to use the right natural pedocrete materials at the right locations and especially to make an extra effort to investigate the self-stabilising properties of calcretes to make much more use of even sub-standard calcretes and to save millions to the Namibian taxpayer. As will be outlined in chapters 5 to 8 of this thesis, convenient and over-safe "first-world thinking" on the side of the road designer has to be replaced by the use of economically available natural materials.
A systematic approach to an appropriate system based on natural road building materials can only be based on cost and quality optimised roads models. Such cost and quality optimised models are a function of different road characteristics which will be described in the next chapters.
The tables in Section 4.7 will be translated into HTM Lnguage in the future.