Studying of the Gypsiferous Soil Suction Using Filter Paper Technique

Gypsiferous soils have specific feature under its suction studying. These soils contain salt solution that affects on the suction components at high water content levels. This paper investigates soil suction of three types of gypsiferous soils namely: AlMuhallabeia, Al-Jarin, and Al-Slmanii using filter paper technique. The investigated soils were located in Nineveh province (400 Km north Baghdad capital) and found to have a gypsum content of about 35, 23, and 6%, respectively. This study includes some factors (water content, gypsum content, and clay content) that affect the soil suction for gypsiforous soils. The effect of gypsum content on slope of the relationship between soil suction with pF unit and water content was examined. The results revealed that there is a linear relationship between soil suction and its water content for three studied soils. Both total and matric suction values increase as soil water content decrease. On the other hand, the difference between total and matric suction values increase with soil samples water content. The soil clay content has more effect on gypsiferous soil suction than its gypsum content.


Introduction
Gypsiferous rocks and its sediments of different origins are found through out various countries in the world especially in the arid and semi arid regions. It may be found near ground surface or at different depths depending on the climatic circumstances and the geological history of the area. Generally, there are four main forms in which gypsum deposit can be found [1]: i) Integrated beds through the soil layers. ii) Small lumps of gypsum within the soil layers. iii) Distributed small spots of gypsum within the soil. iv) Gypsum crystals form mainly in the surface horizon as a result of evaporation of ground waters.
Gypsiferous soil and soils with gypsum accumulation covers large parts of the world. Iraq is considering one of the countries that contains large area of gypsiferous soils. It is about 47792 km 2 which is equivalent to about 11.0 % of total country area [2]. Earlier many studies and researches were applied to obtain more knowledge of engineering properties for gypsiferous soil [3,4,5,6,]. On the other hands, others studies were done to understand behavior, classification, treatment, and land use potential of these soils [7,8,9,10,11,12].
gypsiferous soils in Iraq are mainly situated in the basin of Iraq. These soils are associated with Lower Fars formation of the Middle Miocene. The northern part of the upper AL-Jazierah region has a slightly gypsiferous soil which is formed over a gypsum bed rocks with gypsum content below 6 %. In the southern part of upper AL-Jazierah the gypsiferous soils are formed over gypsum and anhydrate rocks, and they are moderately to highly gypsiferous, with a gypsum content range from a few percentage to more than 60%. The lower AL-Jazierah is characterizing by gypsum desert where primary gypsum bed rock which is the main source of gypsum accumulation in other region [13].
The filter paper technique intensively used for measurements of soil suction because of its advantages over the other methods which are: low cost, simpler method for measuring soil suction both total and metric, can measure the entire range of soil suction, and the equilibrium period is acceptable. Soil suction measurements using filter paper technique has been studied by several researchers [14,15,16,17,18,19]. Al-Khashab [18] found good correlations between soil suction and different engineering properties of expansive soils selected from Mosul city. Al-Ashou et al [16] reported that the clayey soils a quire more suction value than the silty soils under the same moisture condition. Also, his results showed that the effect of density on the moisture-suction relationship is insignificant. This paper attempts to clarify some factors (water content, gypsum content, and clay content) that affect the soil suction value for gypsiferous soils.

Soil Suction
The theoretical concept of soil suction was developed in soil physics in the 20 th century. The soil-water-plant system was the main factor that leads to develop the soil suction theory. The practical studies of soil suction was appeared to explain the mechanical behaviors of unsaturated soils relative to engineering problems especially water flow in the zone of negative pore-water pressure (capillary flow) [20].
Soil suction has an important role that controls the behaviors of unsaturated soils through its effect on engineering properties of the soils. The variety of soil water content (i.e. degree of saturation) causing change in soil suction, "a relative humidity value less than 100% in a soil would indicate the presence of suction in the soil" [20]. It is commonly referred to as free energy state of soil water [21]. Total suction in soil can be divided mainly in to two parts express in equ. (1). The first one called matric suction: result from interplay of attraction and repulsion forces of clay particles and polar water molecules, together with surface tension force in water. The matric suction can be expressed as the difference between pore-air pressure (u a ) and pore-water pressure (u w ). The second part is osmotic suction: as solution potential due to presence of dissolved ions.

The soil
Three gypsiferous soils were selected from various regions near the Mosul city; Al-Muhallabeia (soil 1), Al-Jarin (soil 2), and Al-Slmanii (soil 3). Each of these soils contains various percentage of gypsum. Table (1) refers to some properties of the soils used in the study.
The three soils were oven dried at 60 ± 3C o for 48 hr and passed through a 4.75 mm sieve prior to determining its index properties and compaction characteristic.

Sample preparation
The required amount of soil and water were thoroughly mixed up to obtain homogenous mixture at the desired water contents. At present study the effect of dry unit weight on soil suction has not considered, because many researchers [16,22] pointed out that there is no effect of compaction energy on soil suction. Table (2) shows the molding water content and dry unit weight of the studied cases. The Mixtures were kept in the plastic container for 24 hr, and then the mixtures were compacted in a steel mold to produce a final dimension of the specimens with 63.5 mm in diameter and 25 mm in height.

29
A static compaction method under a rate of 1.0 mm/min was adopted to compact the soil samples using an electrical compressor to reach the required dry unit weight for each soil. The soil samples were kept under the static load for several minutes to avoid rebound, after that the soil samples transferred to a special container related to soil suction measurements using filter paper technique.

Soil suction measurements
Both total and matrix suction were determined using ASTM filter paper method (ASTM D 5298-03). For total suction measurement, the filter paper disks type of Scheicher and Schuell (S&S) No. 589 was fixed above the soil sample as shown in Figure (1).
The plastic ring is put on the soil sample to prevent contact between S&S filer paper and the soil samples. The matrix suction was measured by putting S&S filter paper sandwiched between larger size protective filter paper. Then these sandwiched filter papers are inserted in the soil sample in a very good contact manner, i.e. intimate contact between the filter paper and the soil. Hence, it should be pointed out that the protection filter paper is  Protective filter papers Figure (2), Calibration curve for filter paper. [15] used to prevent direct soil contact with the mid S&S filter paper. Then th i s sy stem was kept i n a (275) m l capacity container. The container was immediately closed with an airtight lid and was sealed with two wrappings sealing tape, then the specimen container was kept in a constant room temperature till the equilibrium state is reached (about two weeks).The water contents both of s oi l s a m p l e s a s w e l l a s th os e of filter papers were accurately determined.
The suction was calculated using the calibration suction-water content curves presented in Figure (2) [15]. Table (3) shows the measured values of total and matrix suction with its water content for soil (1), soil (2), and soil (3). It is obvious that both total and matrix suction varies with soil type and soil water content. The following paragraphs show the main factors that affect the suction values of gypsiferous soil.  Figure (3) shows the linear relationship between water content of the samples and both total and matrix suction for three studied soils. It is obvious that both total and matrix suction values increase as water content decrease. Both lines of total and matrix suction are closer as water content decreases. Table (3) presents total to matrix suction ratio for the studied cases. These ratios increase (i.e. the different between total and matrix suction values) with increasing water content of soil samples. This may indicates at higher soil water content

Al-Rafidain Engineering
Vol. 19 No.3 June 2011 31 the more effective part for soil suction value is the osmotic suction. Because of the existence of gypsum as dissolved salts in soil-pore water, this may increase soil suction value due to the increasing of osmotic suction. The concept of increasing osmotic suction due to presence of dissolved salts can be explained as: the water vapor pressure over a flat surface of solvent is less than the water vapor pressure over a flat surface of pure water. Thus, the relative humidity decrease with increasing dissolved salts in the pore water of the soil and in turn the total suction increases with the increasing of osmotic suction [20]. Figure (4) illustrates the relationships between total to matrix suction ratio and water content of the samples. The slope of these lines increases with increasing soil gypsum content. On the other hand, the highest ratios of total to matrix suction were for soil (1) (which have maximum gypsum content) followed by soil (2), and soil (3). Experimental data-total suction Experimental data-matric suction Bes t-f it line Figure (5-a) shows the relationship between matrix suction and water content. The matric suction of soil (2) and soil (3) are nearly equal to that of soil (1) when they are at relatively high water content. This is happen although the gypsum contents of soil (1) and soil (2) having relatively high values, also the clay contents of soil (2) and soil (3) is relatively having high values comparing with soil (1). This indicates the effect of clay content on soil suction is more pronounced than the effect of gypsum content. Figure (5-b) show the relation of total suction with water content, it seem that both the gypsum and the clay contents affect the total suction. Figure (4), Total to matric suction ratio for three soil studied.

Influence of gypsum content on slope of pF-value versus water content
Figure (6) shows another presentation for the relationships between soil suction with pF unit (pF = Log 10 │h│, where h is the suction in centimeters) and soil water content. The soil suction with pF unit was found for total and matrix suction as well as osmotic suction (osmotic suction represent the difference between total and matrix suction, equation (1)).  Water Content (%)

pF-Value (cm)
Experimental data soil 1 Experimental data soil 2 Experimental data soil 3 Best-fit line A best-fit line to the experiment data were made for all studied cases. The slopes of these linear relationships were tabulated in Table (4), the varying slope of pF-value versus soil water content relationship of total, matrix, and osmotic suction is due to varying gypsum content for the soils, see Table (4). The highest slope was for soil (1) comparing with soil (2) and soil (3) respectively A statistical analysis were adopted to the experimental results for obtaining the correlation between the slope of pF value versus soil water content and the gypsum content as shown in Table (5). The above formulas could be reformed as the curves shown in Figure (7). We can recognize that the three curves of total, matrix, and osmotic suction have the same trend. The effect of gypsum content on slope of pF value versus water content is higher for matrix suction than total and osmotic suction. Also, for a specific soil suction part, the slope of pF value versus water content has low value at soil gypsum content up to approximately 23 %, thereafter an increase in its value has been observed.

Conclusion
1. The three types of gypsiferous soils studied in this research have linear relationships between soil suction and soil water content (both total and matric suction. 2. Osmotic suction has more influence on the soil suction values for soil samples with higher water content as well as higher gypsum content. On the other hand, the soil clay contents more effect on gypsiferous soil suction than soil gypsum content. 3. The slope of pF-value versus soil water content depend on soil gypsum content as well as types of soil suction (total, matrix, and osmotic), The effect of gypsum content on slope of pF value versus water content is higher for matrix suction than total and osmotic suction. Finally, this slope were derived with respect to soil gypsum content by statistical analysis represented as coefficient of determination (0.9989, 0.9983, 0.9985) for total, matric, and osmotic suction respectively.