service life prediction of granite armourstone: a case study of...

Material Sciences and Technology 32

1. INTRODUCTION

South of Thailand is a part of Malaya Peninsular and located between the Gulf of Thailand In The South China Sea of the Pacific Ocean and the Andaman Sea (Fig. 1). Like many coastal countries in the world, Thailand is currently undertaking extensive shoreline developments, including several large reclamation projects, which involve the modifications of the environmental coast of tens of kilometres of new coastal defences. Also the consequences of earth’s changing climate, particularly, geohazard lessons from the 2004 Indian Ocean and 2009 South Pacific Tsunamis destroy catastrophic coastal area. Shoreline and coastal defence structures in Thailand are primarily rock armoured structures utilising locally-sourced rock won from the quarries (several in the Lower South of Thailand) at Songkhla, which outputs granite of variable quality. This paper describes a study conducted into the quality of this rock as a source of armourstone. The ultimate aim of this study is to quantify the reduction in armourstone mass which could be anticipated over the design life of a coastal structure, so that allowances could be made for this mass loss during the design stage. Mathematical models to estimate weathering losses in-service have been developed in the past. The first work was proposed by Lienhart and Stransky [2] proposed method of evaluation of rip-rap and armourstone. Latham [3], who presented a method for predicting armourstone mass loss in-service. Latham and Lu [4] suggested twelve factors quantitative indications for

Fig. 1 Location map of the Songkhla and adjacent coastal study area [1]. the classication of the blastability of a rock mass. Lienhart [5] presented a method of evaluating the quality of potential armourstone sources. These researches were combined into two separate degradation models by Latham et al. [6] and these models were presented in the second edition of the CIRIA Rock Manual [7]. Although using these methods is ‘by its nature inexact and burdened with difficult judgments’ [7] Both of these methods have been applied to Thung Wang armourstone in this paper.

ABSTRACT The main objective of this study has been to evaluate major factors involved in armourstone durability and long-term performance and deterioration of granite armourstone service life. The granite’s behavior in the field observed and measured granite at Thung Wang quarry, Songkhla province. To consider the combined effects of environmental stresses on armourstone, collected samples carried out several laboratory testings have been to evaluate the performance of stone subjected to both weathering and degradation. The tests result determined the quality and durability on mass density, water absorption, Schmidt impact index, compressive strength, point load strength, Los Angeles abrasion and MgSO4 soundness resistance. Long-term performance or deterioration of armourstone has been quantitatively monitored and a comparative study of armourstone quality designation (AQD) and Mocro-deval (MDE) methods analysed. The AQD method predicted service life of granite armourstone 2 and 4 tonnes a 70 and a 90 year and MDE method given a 135 and a 160 year, respectively.

. KEY WORDS: Granite / Armourstone / Durability / Songkhla coast / Laboratory testing

PPaappeerr IIDD 1155

Service Life Prediction of Granite Armourstone: A Case Study of Thung Wang Granite, Songkhla

P. Pantpong1, D. Tonnayopas2*, C. Aedpan3

1, 2, 3 Geotechnical and Innovative Construction Materials Research Unit (GICMRU) Department of Mining and Materials Engineering, Prince of Songkla University, Thailand, 90112

* e-mail: [email protected]


2. EXPERIMENTAL PROGRAME

2.1 Conventional degredation models

The Rock Manual describes two methods, the Micro-Deval (MDE) method, and the Armourstone Quality Designation (AQD) method. These two degradation models are greatly based on works carried out by Latham [3] and Lienhart [5,8], which were later redeveloped [9,10,11]. Both of these models calculate a parameter which represents the site aggressiveness (the Equivalent Wear Time Factor, X) and a parameter which represents the intrinsic durability of the rock (the Intrinsic Resistance to Mass Loss, ks) and use them to estimate mass loss over time with Equation 1.

XTk

XTk

MM ss exp95.030exp05.0

0

(1)

where M is nominal mass of armourstone at time T, M0 is initial mass of armourstone; kS = intrinsic resistance to mass loss; X = equivalent wear time factor; T = time since installation (years). The intrinsic resistance to mass loss kS is a property intrinsic to the rock material describing the resistance to weathering and the method of calculation is different for the two models. On the part of the Micro-Deval method it is solely related to the Micro-Deval Coefficient (MDE), whereas for the AQD method a number of indicators of rock quality are combined with a weighted average system. The AQD method therefore takes account of a large number of relevant factors when assessing ks and for this reason it may be preferred. However, the MDE method is probably better calibrated and the Rock Manual recommends that the results of both methods should be considered together The Equivalent Wear Time Factor, X, reflects the rock size, grading, shape and the conditions that the rock is subject to (wave conditions, climate, waterborne attrition, etc) and it is obtained by a similar method for both the MDE and the AQD methods. Nine parameters, designated X1 to X9, represent the various factors affecting weathering rates on site. The values of these parameters are obtained from look-up tables based on properties such as significant wave height, climate statistics, type of waterborne attrition agent, etc, and in the case of the MDE method, also properties of the rock, such as the block integrity and water absorption. Finally, the overall Equivalent Wear Time Factor is calculated as the product of each of these parameters via Equation 2.

987654321 XXXXXXXXXX (2) where X1 = rock size, X2 = rock grading, X3 = rock shape, X4 = wave energy, X5 = zone of structure, X6 = climatic weathering, X7 = waterborne attrition agents, X8 = concentration of wave attack, X9 = mobility of armuorstone

2.2 Micro-Deval coefficient

The Micro-Deval test is a test of a rock’s resistance to abrasion. The test procedure involves tumbling a sample of 10-14 mm aggregate (obtained by crushing representative armourstone) in a standard drum with a controlled quantity of water and steel balls for a standard period of time [12]. The result of the test, the MDE is the percentage by mass of material which passes a 1.6 mm sieve on completion of the tumbling process. The test is outlined definitively in BS EN1097-1 [13]. Since the Micro-Deval test result is essential to the application of the MDE method, and also has some (much more limited) influence on the AQD method, it was necessary to obtain an indicative Micro-Deval coefficient for Thung Wang rock. The Micro-Deval test is unfortunately not commonly undertaken on Thung Wang rock and it is thought that there are no testing laboratories capable of undertaking the test in the region. Another mill abrasion test (Los Angeles test) is commonly undertaken in Thailand, but studies in the past have found no significant correlation between the Los Angeles and the Micro-Deval test result [14]. However, quality assessment of large armourstone using an acoustic velocity analysis method was also performed [22] A certain number of samples were taken from the Thung Wang quarry and tested in a Geotechnical laboratory in the Department of Mining and Materials Engineering, PSU. The samples were taken from various parts of the quarry and of varying quality (as assessed visually). As a result just two Los Angeles abrasion tests were undertaken. Moreover, Shore hardness and impact value tests were applied to assess Thung Wang granite incorporation with Los Angeles test instead of Micro-Deval test. These values were tentatively applied to degradation modelling, but treated with considerable caution there were so few tests and because of the mixing of the samples.

2.3 Determination of the intrinsic resistance to mass loss

As described above, ks is obtained by a different method depending on whether the MDE or the AQD method is being used. For the AQD method, it is simply obtained AQD from the using equation 3 and also derived MDE from equation 4.

0.2032.0 AQDEks (3)

485.151012.4 MDEk s (4)

One of the great advantages of the AQD method is that properties which are not available from field evaluations and laboratory tests do not need to be included in the determination of the AQD value, and the weighted average AQD value can be based on whichever parameters are available. Of course, the more parameters that are available the more accurate will be the estimate of the AQD value, and therefore


the subsequent estimation of rock mass loss over time.

2.4 Meteorological climate weathering intensity

One area in which the CIRIA guidance appears to be not entirely suitable for estimating rock weathering rates in the South East Asia is in its recommendations to determine the parameter X6, which is representative of the aggressiveness of the meteorological climate. Lienhart [8] presented the recommendation is to use the Meteorological Climate Weathering Intensity (MCWI) index, and reproduced in Equation 4. This value is used (along with the water absorption, in the case of the MDE method) to look up values for X6. However, this parameter is based on The Royal Meteological secondary data only, and is considered to be suitable for the Lower Gulf of Thailand. In fact, since the MCWI index is calculated based on the mathematical product of a number of parameters, and one of these is the number of days when the temperature falls below freezing (which never occur in southern Thailand) the MCWI calculated in this way must be zero and the climate is treated as benign.

hfgcedbaMCWI )/()/()365/()/( (5) where a is mean (max) – mean (min) temperature range over several years, b is mean annual temperature, c is mean number of days max temp > freezing, d is mean number of days min temp ≤ freezing, e is extreme max and min temperature range over several years, f is mean number of days with precipitation > 0.25mm, g is annual precipitation in cm, h is total normal degree-days, base 18°C

3. RESULTS AND DISCUSSION

3.1 Geological setting

Thung Wang quarry, is Triassic biotite coarse grained granite to porphyritic granite [15]. It is noticeable that Thung Wang rock is of variable quality (Fig. 2a), with differences being visually evident in terms of the colour of the rock, and the presence of degree of alteration [16] and other quartz and feldspar phenocrysts in the rock fabric (Fig. 2b). Sometimes joints and faults are evident in the rocks at the quarry and sometimes these discontinuities are filled with rock fragments and wall rocks coated with chlorite or iron stain as shown in Fig. 2.

Fig. 2 a) Thung Wang quarry in Songkhla and b) Varied size and shape of fragmentations.

It has been investigated that joints are not necessarily detrimental, and may even be beneficial to the durability of armour stone, provided that they are small and evenly distributed. However, where these joints may form a failure plane or occupy a substantial part of the rock’s volume they have an obvious ability to reduce the integrity of the block (Fig. 3a). Moreover, Lindqvist et al. [17] reported influence of microstructure on functional properties and size of rock materials (Fig. 3b).

Fig. 3 a) Rocks containing joints and faults and b) size of blasting block distributions.

3.2 Index properties of the armourstone

A large number of shore protection works have been undertaken in this study, there is a reasonable amount of information available on rock quality based on tests undertaken to prove compliance with standard specifications of ISRM [18,19] and ASTM [20,21]. The test results for the granite is given in Tab 1 and Fig. 4. Furthermore, field investigation of rock mass were also assessed at the Thung Wang Quarry and are summarised in Tab. 2 Tab. 1 Index properties of the Thung Wang granite

Property Range (mean) CIRIA rating

Dry density (t/m3) 2.51-2.59(2.54)

Good

Water absorption (%) 0-0.93 (0.4) Excellent Compressive strength (MPa)

105-115 Good

Point load index (MPa)

2-6 (3.57) Good

Los Angeles (%) 39-40 Poor MgSO4 soundness (%)

0.03 Excellent

Impact value (%) 17.97-20 None Shore hardness 845 None

Tab. 2 CIRIA/CUR classification for Thung Wang Quarry

Criteria Description CIRIA Rating

Lithological classification

Biotite granite to pophyritic granite

Excellent

Weathering grade

I Fresh Marginal

Groundwater condition

Moist and completely dry

Marginal

a) b)

a) b)


Tab. 2 Continued Criteria Description CIRIA

Rating Production method

Conventional blasting

Marginal

Stone shape & weathering grade

10-15% of stones LT>3, 95% of stones Grade II

Marginal

Set aside Approx. 1 month Excellent

Fig. 4 a) Specimens for compression and point load tests and b) wear feature of granite after Los Angeles abrasion test. The Rock Manual indicates that density variation is a good indication of overall quality variation, and that the difference between the average and the 90% exceedance density should generally be not more than 50 kg/m3. Assessment of the 27 density tests which are available yields a 90% exceedance density of 2.58 t/m3, which is exactly 10 kg/m3

above the mean density, and so at the limit of the normal range suggested by CIRIA [7]. This suggests that the variability of rock produced at Thung Wang is low, which concurs with observations at the quarry and on site. It was noted in Latham’s paper [3] that salt crystallization in phaneritic igneous rocks can be accelerated by the climate in the South East Asia, and this concurs with observations of armourstone in service in Songkhla. Therefore, the original recommendations from Latham [3] were used to determine X6 instead of the MCWI index of Lienhart [4]. This decision has a considerable impact on the final estimates; X6, which would have been 1.5 (MDE) or 1.5 (AQD) if the MCWI method was used, is instead 0.2, and therefore the Equivalent Wear Time Factor would have been 4 or 5 times higher if the MCWI method proposed by CIRIA [7] had been adopted. For a typical Thung Wang rock, this could result in the final degradation estimate can double if the MCWI method were used instead of Latham’s method. Indeed, even if Latham’s method is applied the X6 parameter is somewhat sensitive, as it is 0.2 where the water absorption is above 2% and 0.5 when the water absorption is below 2%. Since the water absorption is estimated to be 0.45%, Thung Wang porphyritic granite is on the innerline, base upon a value of 0.5 has been adopted.

3.3 Determination of the intrinsic resistance to mass loss

In light of the uncertainty in the Micro-Deval results which have been obtained consideration needs to be given to the MDE value which is to be applied to obtain ks. The MDE values obtained were both 15% for the fresh and weathered samples mixture, 20% for the sample representative of rock appearing to be good quality and 35% for the sample of apparently excellent quality rock. If the range 15% to 35% is assumed to be representative of Thung Wang armourstone then this gives ks of 3.5210-3

to 4.6610-3. The method for determining the ks value when using the AQD method is somewhat more complicated, involving a weighted average of a number of factors. The same weights have been applied as recommended by the Rock Manual to obtain the AQD value, and this calculation is shown in Tab. 3. Tab. 3 Fractions of original mass ranges calculated for Thung Wang rock as armourstone in 50 years Initial rock size (M0) Fraction of original mass in 50y (M/M0)

Criterria CIRIA rating Rating value

Weighting

Weighted rating 4 3 2 1

Lithological classification 4 58 2.91

Regional in-situ stress 2 73 1.83

Weathering grade 2 73 1.83

Discontinuity analysis 2 95 2.38

Groundwater condition 2 73 1.83 Production method 2 95 2.38 Rock block quality 2 80 2.01 Set-asisde 4 73 3.67 Mass density

3.0 80 3.01 Water absorption UCS

3.5 88 3.87 Schmidt impact index PLS

1.5 88 1.66 Los Angeles abrasion

MgSO4 soundness 4 80 4.02

Sum 956 31.41 Note: Excellent = 4 Good = 3 n 12 12

Marginal = 2 Poor = 1 Mean 79.67 2.62

3.4 Determination of equivilent wear time factor

The estimation of the Equivalent Wear Time Factor in this study are summarised in Tab. 4.

b) a)


Tab. 4 Fractions of original mass ranges calculated

Criteria Description Rating value AQD MDE

X1

Considered typical primary armour in Songkhla of between: 0.5, 1.0, 2.0 and 4.0 tonnes.

0.63-0.79

0.63-

0.79

X2 Narrow graded armour is assumed, with (M85/M15)1/3 ≈ 1.2

1.64 1.64

X3 Assumed irregular based on inspection 1.5 1.5

X4

The 1 in 50 year Hs is estimated to be approximately 2.5 m for Songkhla’s north coasts. Based on such small waves a value of 4.0 is deemed reasonable.

2.6 2.6

X5 Assumed worst case (intertidal) 1.0 1.4

X6

Used recommendations from Latham [3] instead of CIRIA [7]. Assumed to be a hot and wet climate with water absorption < 2%

1.5 1.5

X7

Would generally be no waterborne attrition agents due to mild wave conditions, but the case of waterborne attrition by sand has also been considered (the seabed around Songkhla coast comprises sand and silt.

1.0 1.0

X8

Revetment slope angle is commonly 1 in 2.5 or gentler. Tidal range is generally less than 2 m throughout the Gulf coast of Thailand

0.3 0.3

X9

Assumed static design concept (generally the case for permanent structures in Songkhla), assuming Im50 = 2.56%

1.1 1.1

3.5 Comparison between degredation models

The value ranges obtained above for the Intrinsic Resistance to Mass Loss and the Equivalent Wear Time Factor were input to Equation 1 to obtain likely degradation envelopes shown in Fig. 5 for 2 and 4 tonne armours. The two methods show plausible estimates of degradation over time, although the MDE methods lowend estimates of mass loss do not seem likely. This is due the lower values of the estimated MDE indicating a rock of

0.300.350.400.450.500.550.600.650.700.750.800.850.900.951.00

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180

Years in Service Life

M/M

o (%

)

MDE, X = 2.79, ks = 0.003523AQD, X = 2.0, ks = 0.004662MDE, X = 3.5, ks = 0.003523AQD, X = 2.5, ks = 0.004662

Fig. 5. Predictly degradation envelopes based for 2 and 4 tonne armour on the AQD and MDE methods. ‘GOOD’ quality. The AQD method suggests the rock is of ‘MARGINAL’ quality. Regarding to the result of AQD method, concept designers accepted allow for design fraction original mass (M/M0) low until to 0.65, service life of a 70 year and a 90 year for 2.0 and 4.0 tonnes armour, respectively. Besides, MDE method, service life is longer nearly twice of AQD method result, it is given a 135 year and a 160 year for 2.0 and 4.0 tonnes amour (Fig. 5). In light of limited confidence in the Micro-Deval coefficient values which were obtained and the high sensitivity of the MDE method to this test value, it is considered that the AQD model provides the more reliable estimates in this case. In the viewpoint is re-enforced by engineering judgement, which would suggest that a ‘MARGINAL’ quality rating as suggested by the AQD method is a reasonable assessment of Thung Wang granite rock. The following analyses therefore consider the AQD method in isolation.

3.6 Rock size effect

Weathering rates of the rock size affects, as the removal of a ‘shell’ of given thickness has a proportionately greater effect on a smaller rock than a larger one. Degree of weathering is therefore greater in smaller rocks than in larger ones and this important factor must be considered when making allowance for degradation in design of Songkhla delineated area. Estimated degradation envelopes are illustrated in Figure 6 for four rock sizes which are typically used for primary armour in Songkhla; 0.5 tonne (X = 1.26), 1.0 tonne (X = 1.58), 2.0 tonnes (X = 2.0) and 4.0 tonnes (X = 2.5), using the AQD method. These degradation curves could provide indicative weathering allowances for the purposes of concept designs. In regard to the results of the AQD model, it is suggested that concept designers should allow for M/M0 of 0.62 for 0.5 tonne armour, 0.676 for 1.0 tonne armour, 0.726 for 2.0 tonne armour and 0.766 for 4 tonne armour, whole under for a 50 year design service life.


0.50

0.550.60

0.65

0.700.75

0.80

0.850.90

0.95

1.00

0 5 10 15 20 25 30 35 40 45 50

Years in Service Life

Frac

tion

orig

inal

mas

s (%

)

0.5t, X = 1.26, ks = 0.0046621t, X = 1.58, ks = 0.0046622t, X = 2.0, ks = 0.0046624t, X = 2.5, ks = 0.004662

Fig. 6 Degradation envelopes for different rock sizes based on the AQD method.

4. DISCUSSIONS AND CONCLUSIONS In the field observations by checking the produces of the granite in the Thung Wang quarry is lowly variable and the highland reveal that the granite showed marginal quality armourstone. The conditions that the rock is subject to are mild in terms of wave attack but the high salinities and temperatures in the region are likely to accelerate surface spalling and crumbling. When the AQD and MDE models were applied to the known parameters for Thung Wang rock, the MDE method suggests lower degradation rates than the AQD method. The results obtained on samples show significant variation in the resistance to wear. This is mainly due to the fact that the quality of the local rock varies within the same quarry [16]. This was confirmed by the density variation, described by CIRIA as a good indication of the overall quality variation. Because the MDE method relies uniquely on the MDE

coefficient for determining the intrinsic resistance to mass loss ks, the results obtained adopting this method are to be treated with caution, as it relies entirely on the specimens taken being truly representative of the output of the quarry. On the other hand, the AQD method takes into account a number of different parameters, which are weighted and averaged to give the AQD value used for the determination of ks. According to the limited confidence in the Micro-Deval values obtained ver estimation, it is considered that with the available data the AQD method provides a better general description of the rock resistance weathering in this case. With regard to the results of this model, it is proposed that concept designers should allow for M/M0 of 0.62 for 0.5 tonne armour, and 0.755 for 1 tonne armour for a 50 year design service life. Inside for M/M0 of 0.726 for 2.0 tonne armour, and 0.766 for 4 tonne armour also for a 50 year design service life (Fig. 6) The mentioned figure has been determined adopting rock proprieties and site conditions which are deemed generally representative of the local Triassic granite porphyry and for the Thailand’s southern Gulf east coast from Prachuap Khiri Khan

to Narathiwat. Designers should therefore carefully verify the applicability of these proposed degradation rates and refer to site specific investigation in order to correctly apply the degradation models on a case-to-case basis. For example design criteria (e.g. slope, grading envelope curves, stability number, frequency of storm, earthquake intensity, etc) could alter significantly the proposed degradation predictions. Likewise site conditions (e.g. tidal range, wave height etc) could differ throughout south Thailand, between the East (Andaman Sea) and the West (Gulf of Thailand) coasts. This could also lead to different predicted degradation rates. There are several number of factors concerning to the macro structure of the rock may not be realistically represented in the degradation models. Both models are primarily based on aggregate tests (i.e. the samples are produced by mean of crushing the rock down to the required size) and aggregates might not have the same resistance proprieties of larger rock used as armourstone. Especial, the presence of geological discontinuities (joint, fault, shear zone), alterations and mineral phenocrysts (quartz, feldspar) is noticeable in a number of rocks in service, whereas samples are generally clean from sand. The degradation in the mass of individual armourstones due to weathering poses questions about the implications on the different phenocrysts and phaneritic texture of the stones. It is expected that the reduced mass and changed shape of single rocks has a detrimental effect on the stability of the geo-materials as well as potential armourstone quality [23,24] and leads to an increased degree of mobility. The more the single rock elements become degraded, the more movement is allowed and the more the rock is prone to rock-to-rock abrasion and hence degradation or erosion. Thus it may be expected that weathering rates will increase over time, and the predictions (which rely on the assumption of degree of weathering) may not be accurate. This is especially a concern when the models predict high levels of weathering. One aspect where there seems to be a lack of guidance is in the determination of the X6 parameter for tropical monsoon regions like Thailand, where freezing temperatures never occur. The Rock Manual makes a general recommendation that the MCWI derived by Lienhart [8] should be used to determine X6, but in the absence of freezing days this renders meteorological weathering benign. However, it is considered that the extremely tropical climate (equator) accelerates alteration and crumbling of armourstone in Songkhla, and site inspections suggest that this may be a significant weathering mechanism. For the aims of this study X6 has been determined under the method described by Latham [3]. Therefore, there is rarely local data to support this and the refinement of the X6 parameter could be the subject of further study. It is important not only


to ensure that new benefit to obtain actual measurements of rock degradation on Thung Wang revetments against but also that the degradation models could be calibrated by methods like those described in Latham’s paper [3]. Such methods include block shape measurement; block surface profile monitoring and direct weight measurement, all have to work in tandem for a safe and economic design and to meet optimal standards which existing coastlines and infrastructure be retrofitted and rehabilitated. By other ways where confidence in these models could be greatly improved would be to undertake an expert geological assessment in rock mass and petrographic examination of the Thung Wang quarry carried out under thin-section method like distinguish type of granite of Tonnayopas et al. [25], and by further Micro-Deval testing.

ACKNOWLEDGEMENTS This research was carried out under mining engineering project of undergraduate programme which some part was financially supported by the Faculty of Engineering, Prince of Songkla University, Thailand.

REFERENCES [1] http://www.thefullwiki.org/South [last accessed

18.07.13].

[2] D.A.Lienhart, and T.E.Stransky (1981), Evaluation of the Potential Sources of Rip-Rap and Armourstone-Methods and Considerations. Bull Assoc Engng Geologists, 1981, Vol. 18, pp. 323-332.

[3] J.P.Latham (1991), Degradation Model for Rock Armour in Coastal Engineering, Quart J. Engng Geol, 1991, Vol. 24, pp. 101-118.

[4] J.-P.Latham, and P.Lu (1999) Development of an Assessment System for the Blastability of Rock Masses, Inter J. Rock Mech Mining Sci, 1999, Vol. 36, Issue 1, pp. 41-55.

[5] D.A.Lienhart (1998), Rock Engineering Rating System for Assessing the Suitability of Armourstone Sources, Engng Geol Special Public, 1998, Vol. 13, pp. 91-106.

[6] J.P.Latham, D.Lienhart, and S.Dupray (2006), Rock Quality, Durability and Service Life Prediction of Armourstone, Engng Geol, 2006, Vol. 87, Issue 1-2, pp. 122-140.

[7] CIRIA, CUR, CETMEF. (2007), The Rock Manual. The Use of Rock in Hydraulic Engineering, (2nd

Ed). C683, CIRIA, London

[8] D.A.Lienhart (2003), A Systems Approach to Evaluation of Rip-rap and Armor Stone Sources. Envi Engng Geosci, 2003, Vol. IX, No. 2, pp. 131-149

[9] J.P.Latham, J.V.Meulen, and S.Dupray (2006), Prediction of In-Situ Block Size Distributions with Reference to Armourstone for Breakwaters, Engng Geol, 2006, Vol. 86, Issue 1-2, pp. 18-36.

[10] J.-P.Latham, J.V.Meulen, and S.Dupray (2006), Prediction of Fragmentation and Yield Curves with Reference to Armourstone Production, Engng Geol, 2006, Vol. 87, Issues 1-2, pp. 60-74.

[11] J.-P.Latham, D.Lienhart, and S.Dupray (2006), Rock Quality, Durability and Service Life Prediction of Armourstone, Engng Geol, 2006, Vol. 87, Issues 1-2, pp. 122-140.

[12] J.Latham (1993), A Mill Abrasion Test for Wear Resistance of Armour Stone, STP1177 Rock for Erosion Control, McElroy and Lienhart, eds., pp.

[13] BS EN 1097-1. (2004), Tests for Mechanical and Physical Properties of Aggregates - Part 1: Determination of the Resistance to Wear (Micro-Deval). BSI, London

[14] P.R.Rangaraju, and J.Edlinski (2008), Compara-tive Evaluation of Micro-Deval Abrasion Test with other Toughness/Abrasion Resistance and Soundness Tests. J. Mater Civ Engng, 2008, Vol. 20, Issue 5, pp. 343-351.

[15] D.Tonnayopas (2012), Roadside Geology in Songkhla Province, Faculty of Engineering, Prince of Songkla University, 24 p. (in Thai)

[16] D.Tonnayopas, A.Sriya and C.Aidpan (2008), Weathering degree Effects on Geotechnical Properties of Porphyritic Granite in Songkhla. The 6th PSU Engng Conf., 8-9 May 2008, Songkhla, pp. 31-36. (in Thai)

[17] J.E.Lindqvist, U.Åkesson, and K.Malaga (2007), Microstructure and Functional Properties of Rock materials, Mater Charact, 2007, Vol. 58, pp. 1183–1188.

[18] ISRM (1981), Rock Characterization, Testing and Monitoring. International Society for Rock Mechanics, Suggested Methods, Pergamon Press, Oxford. Brown, E.T. (ed.)


[19] ISRM (1985), Suggested Method for Point Load Strength. Inter J. Rock Mech, Mineral Sci. Geotec Abst, 1985, Vol. 22, pp. 51-60.

[20] ASTM C535 (1989), Standard Test Method for Resistance to Degradation of Large-Size Coarse Aggregate by Abrasion and Impact in the Los Angeles Machine, American Society for Testing and Materials, pp. 285-287.

[21] ASTM C88 (1990), Standard Test Method for Soundness of Aggregates by the Use of Sodium Sulphate or Magnesium, American Society for Testing and Materials, pp. 37-41.

[22] M.S.J.Niese, F.C.A.A.Van Eijk, G.J.Laan, and P.N.W.Verhoef (1990), Quality Assessment of Large Armourstone Using an Acoustic Velocity Analysis Method, Int. Assoc Engng Geol Bull, 1990, No. 42, pp. 55–65.

[23] A.Özvan, I.Dinçer, and A.Acar (2011), Quality Assessment of Geo-Materials for Coastal Structures (Yumurtalık, Turkey). Mar Georesour Geotechnol, 2011, Vol. 29, No. 4, pp. 299-316.

[24] T.Topal, and O.Acir (2004), Quality Assessment of Armourstone for a Rouble Mound Breakwater (Sinop-Turkey). Envi Geol, 2004, Vol. 46, pp. 905-913.

[25] D.Tonnayopas, W.Kanchana, C.Aidpan, and S.Jantaramanee (2008), A New Approach for Assessment of Commercial Granite Slabs Quality. The 6th PSU Engineering Conference, 8-9 May 2008, Songkhla, pp. 593-598. (in Thai)

service life prediction of granite armourstone: a case study of...

Documents