May_1967 — Page 36

Fig. 11. Shed for mixing lime-soil material used in working course

fills. However, in areas amounting to 24,000 sq. yd out of a total of 523,000 sq. yd of Type 1 fill there was a noticeable heave under the proving roller and a working course of 8 in. of lime-stabilized soil was laid instead of the normal 4 in.

Lime-stabilized soil working course

Before work started, a series of CBR tests with the granite residual soils on the site were carried out using varying percentages of lime. On the basis of these tests it was decided to use 4% lime by weight with Type 1 or 2 soils for the working course.

Where the stabilized soil formed the base course, however, i.e. in the roads and shoulders, it was decided to use 6% lime. At the same time labora- tory samples were mixed and com- pacted to test the effect of ageing of the lime-stabilized soil. Lime was produced on the site.

At first the contractor spread loose Type 1 soil evenly on top of the pre- pared and sealed formation to a depth of about 6 in. to 7 in. and proceeded to pulverize this with the Bros Roto- mixer and the Howard Rotavators, but it was found impossible to achieve the specified fineness of 80% passing 3:16 in. sieve after as many as 64 passes.

At this stage the effect on the strength (CBR value) of lower degrees of pulverization was investigated in the laboratory.

Type 2 soil was then used in the same manner and was found to be much easier to pulverize. The fineness required was achieved much quickly and easily, but because of the frequency of rain in the afternoons little construction was possible.

more

It was then decided to set aside an area on the western edge of the flight strip between chainages 9,000 and 10,000 ft. as a centralized time-soil mixing area. All Type 1 soil was re- moved and shed 35 ft. wide, 600 ft.

60

long, with open sides and no internal obstructions, was built along the ex- The treme western side of this area. shed was filled in dry weather with dry Type 2 soil to a depth of 2 ft. 6 in. leaving

unobstructed an

clear height of 16 ft. for the machines to work under the shed is shown in Fig. 11. In wet weather the top 6 in. of fill in the shed was pulverized and mixed with lime. As soon

as the formation dried this layer of mixed material was shipped out by traxcava- tor and lorries and spread on the formation to fine levels by a grader.

Whenever weather permitted, pul- verizing and mixing took place in the open next to the shed, and the shed then provided a standby of five days lime-stabilized soil work in wet wea- ther. Nevertheless, soil from the shed was used for approximately 50% of the lime-stabilized work.

was

With experience it was found that the pulverization of the soil facilitated by spreading the lime after a single pass of the pulverizer. The lime and soil were mixed together in another pass and after a slight delay the pulverization was continued until 80% of the mix passed the 3/16 in. sieve.

Using this procedure the re- quired pulverization was fairly easily. achieved,

If the weather was continuously wet the mixed material was sometimes kept in the shed for a few days be- fore spreading and compacting. It was noted from tests that a delay of one week between mixing and com- pacting the mixture resulted in a loss of CBR of about 33%. The loss of strength with delayed compaction of lime-stabilized soils has been recorded by Webster and Sheary.2 Where there was a few days' delay an extra 2% lime was added and mixed in to compensate for the loss of strength of the original mix.

After spreading and grading, the lime-soil mixture was given initial

compaction by two or three passes of

a 22 ton three-wheel steel drum roller and ten passes of an 8 ton tandem steel drum roller. At this stage if the moisture content was at or below 23%, as was usually the case, and if rain was threatening, as was often the case, the surface was sealed with Class 1 B bitumen emul- sion at approximately 10 sq. yd/gal, which was sanded after breaking.

The main compaction was then continued. Initially a 24 ton self propelled pneumatic roller and the 18 ton towed steel drum roller were tried, but neither was able to reach the specified dry density of 95% B.S. heavy compaction. Finally a 21⁄2 ton towed vibrating roller was tried out and proved satisfactory. Seven passes of the vibrating roller after initial compaction were sufficient.

As described above, the degree of pulverization was continually checked and approved when more than 80% of the soil passed the B.S. 3/16 in. sieve. The rate of spread of lime was measured by placing a tray of unit area below the spreader and weighing the lime deposited on it. The uni- formity of mixing was controlled by the appearance and by occasional samples taken to the laboratory from which the lime content was measured by the E.D.T.A. method3. Density control of the compacted soil-lime mix was by sand replacement method and Speedy moisture content tester checked with oven-dried samples as in the earthwork compaction control.

In situ CBR tests were taken on the stabilized soil surface at seven days at the rate of one per 10,000 sq. ft. Other in situ CBR tests and uncon- fined compression tests on cored sam- ples were taken at different ages of the stabilized soil. A number of plate bearing tests were also carried out on the completed soil-lime slab. The re- sults of these tests were satisfactory and confirmed the choice of pavement thickness.

Cement-stabilized soil working course

In the pavement fills completed after September 1964 (mainly the taxiway and intermediate turn-offs) sand was used and a series of labora- tory tests was carried out stabilizing the Type A sand with varying cement contents.

In lieu of 4% lime stabilized soil as the working course below aircraft pavements the contractor was then allowed to use 5% cement with Type A sand at 9% moisture content, and in lieu of 6% lime-stabilized soil as the base course in the roads the con- tractor was allowed to use 6% cement with Type A sand.

The sand fills were placed and compacted up to formation level; the cement was mechanically spread on the formation and two passes of the Rotormixer were usually found suffi- cient to mix the cement and sand thoroughly. Immediately after mix-

Far East Architect & Builder May, 1967