Behaviour of Steel Fibre Reinforced Concrete Beam Under Cyclic Loading

IOSR Journal of Mechanical and Civil Engineering (IOSR-JMCE) e-ISSN: 2278-1684,p-ISSN: 2320-334X, Volume 6, Issue 3 (May. - Jun. 2013), PP 01-04 www.iosrjournals.org

Behaviour of Steel Fibre Reinforced Concrete Beam under Cyclic Loading Sreeja . M.D Department of Civil Engineering SRM University, Kattankulathur, Chennai.

Abstract: This paper describes the influence of steel fibre distribution on the ultimate strength of concrete beams. An experimental & analytical investigation of the behaviour of concrete beams reinforced with conventional steel bars and steel fibres under cyclic loading is presented. It is now well established that one of the important properties of steel fibre reinforced concrete (SFRC) is its superior resistance to cracking and crack propagation. As a result of this ability to arrest cracks, fibre composites possess increased extensibility and tensile strength, both at first crack and at ultimate load and the fibres are able to hold the matrix together even after extensive cracking. The net result of all these is to impart to the fibre composite pronounced post – cracking ductility which is unheard of in ordinary concrete. The transformation from a brittle to a ductile type of material would increase substantially the energy absorption characteristics of the fibre composite and its ability to withstand repeatedly applied, shock or impact loading. Tests on conventionally reinforced concrete beam specimens, containing steel fibres in different proportions, have been conducted to establish loaddeflection curves. It was observed that SFRC beams showed enhanced properties compared to that of RC beams with steel fibres. The experimental investigations are validated with the analytical studies carried out by finite element models using ANSYS.

Keywords: Steel fiber, concrete, properties, crack, ductility, technology. I.

Introduction

SFRC is a composite material made of cements, water, fine and coarse aggregate, and a dispersion of discontinuous, small fibres. These short discrete fibres are uniformly distributed and randomly oriented. They are mixed with concrete before pouring. The reinforced concrete structures are subjected to cyclic loads during dynamic loads such as earthquake shocks, traffic loads on the bridges, etc. It is well known that plain concrete is brittle and weak under flexural loads. To eliminate the disadvantages of plain concrete is added fibers into concrete mix. All admixtures meeting ASTM specifications for use in concrete are suitable for use in SFRC. Steel fiber-reinforced concrete (SFRC) has gained increased popularity in construction industries in recent years. Properties of Concrete Improved by Steel Fibres:  Flexural Strength: Flexural bending strength can be increased of up to 3 times more compared to conventional concrete.  Fatigue Resistance: Almost 1 1/2 times increase in fatigue strength.  Impact Resistance: Greater resistance to damage in case of a heavy impact.  Permeability: The material is less porous.  Abrasion Resistance: More effective composition against abrasion and spalling.  Shrinkage: Shrinkage cracks can be eliminated.  Corrosion: Corrosion may affect the material but it will be limited in certain areas. The different types of steel fibres which are available is shown in Fig 1.2

Fig 1.2 Types of steel fibres www.iosrjournals.org

1 | Page

Behaviour Of Steel Fibre Reinforced Concrete Beam Under Cyclic Loading 1.1 Objectives The objective of this study is to:  Investigate the mechanical properties like flexural strength, modulus of elasticity flexure beams and cylinders.  Investigate ductility requirement , moment rotation capacity, energy absorption capacity for SFRC beams under cyclic loading using hybrid steel fibers of different volume fractions  Improve the resistance of conventionally reinforced structural members to cracking, deflection and other serviceability conditions, by utilizing the inherent material properties of fiber concrete. 1.2 Need for the Research To improve:  Ductility.  Energy absorption.  Moment rotation capacity.  Structural strength.  Confinement characteristics. 1.3 Scope The scope of work is to study the behaviour of steel fibre reinforced beams under cyclic loading and comparing the results both analytically (using ANSYS- a general purpose finite element-modelling package) and experimentally. The parameters investigated are:  Midspan Deflection  Strain measurement  Location & type of crack occurrence 1.4 Research significance It is clearly seen from the literature that the behaviour of beam is affected by the properties of core concrete. Ductility is desirable in reinforced concrete frames under seismic loading. Ductility is generally achieved by providing closely spaced horizontal ties, but this causes difficulties in placing concrete in densely reinforced portions resulting in bad concreting. To avoid such closely spaced stirrups, confinement characteristics of core concrete has to be improved, which increases the ductility of the core concrete. In this respect, steel fibre reinforced concrete (SFRC) which posses ductility, toughness and tensile strength more than the plain conventional concrete can be considered to replace the plain concrete .It is expected that use of SFRC can eliminate partially or fully the ties ,thus avoiding the congestion of reinforcement. 1.5 Summary of review of literature  It is clearly seen from the literature that the behaviour of RC beam is affected by the properties of Steel fibre reinforced core concrete.  Many investigators have established that inclusion of high strength, high elastic modulus steel fibres of short length and small diameter improves the tensile strength and ductility of concrete significantly.  The combination of steel fibres and stirrups demonstrates a positive composite effect on the ultimate load, ductility and failure pattern of concrete beam.

II.

Materials and Methods

2.1 Materials: The materials used in the experimental work are tested for their properties and the details are furnished. Raw materials listed below were used for preparation of the specimens:  Ordinary Portland Cement (OPC)  Coarse aggregate with 10 mm maximum size  Fine aggregate  Steel Reinforcement: High Yield for main bar (12 mm,10 mm) and; Mild steel for shear reinforcement (8 mm)  Crimped and Hooked end steel fibres  Plywood with thickness 12 mm to prepared formwork.

www.iosrjournals.org

2 | Page

Behaviour Of Steel Fibre Reinforced Concrete Beam Under Cyclic Loading 2.1.1 Details of Fibres Fibre used – Hooked end, Crimped with following aspect ratios and diameters. The properties of fibres are given in Table 2.4

s.no

Fibre

1

Hook End -1 Hook End -2 Hook End -3 Crimped

2 3 4

Aspect ratio (l/d) 80

Length (mm)

Diameter (mm)

60

0.75

65

35

0.55

45

50

1.11

50

50

1

Table 2.4 Properties of Fibres

III.

Mix Design

Considering the capacity of beam testing machine, the grade of the concrete is M30. The mix design of the M20 grade concrete as per IS: 10262-1982. Water

cement

Fine aggregate

Coarse aggregate

214.24

465.7

506.52

1123.6

0.46

1

1.08

2.41

3.1 Design Of RC Beam The RC Beam is designed based on the load carrying capacity of the testing Equipment. The limiting load applied on the beam from the equipment is considered as 250kN.. The dimensions and reinforcement details for the beam are given below.The detailing of the beam is done as normal detailing and ductile detailing. L - 1800 mm B - 150 mm D - 225 mm 3.1.1 Fibre proportions in RC Beam s.no 1 2 3 4

Proportions Control beam without fiber & normal detailing as per IS456 Control beam without fiber & ductile detailing as per IS13920 Crimped fibre with AR-50+ Hooked End fibre with AR-80 Hooked end fibre with AR-80+ Hooked End fibre with AR-45

IV.

Experimental set up

All beams after casting are cured for 28 days prior to testing.Cyclic Loading is the application of incremental Push and Pull Load. The analysis is also known as Push-Pull analysis. The loads are applied incrementally as positive and negative loads.Hydraulic jack and load cells(10tonnes or 25 tonnes) are placed above and below the beam in order to attain the effect of cyclic loading. Each beam is tested for corresponding load increment and finally the4 behaviour of beam is studied. 4.1 Analytical investigation The study on the structural frame is basically a deflection controlled analysis; hence the analysis is carried out by the application of lateral force. The 3D model is analysed and the displacement results are observed from cyclic loading by applying each load increment both in positive direction and in negative direction. Figure showing the reinforcement details in ansys is shown below.

www.iosrjournals.org

3 | Page

Behaviour Of Steel Fibre Reinforced Concrete Beam Under Cyclic Loading

. Fig.4.1.1 Reinforcement details in ANSYS

V.

Schedule of work

Cylinders are casted by using the various fibre configurations in order to get the corresponding modulus of elasticity. The same results are compared analytically using ANSYS software.

VI.

Conclusion

The experimental values of the strain and deflection of steel fiber reinforced concrete beams are compared with that of the corresponding estimated values of the strain and deflection of reinforced concrete beams without fiber. The comparison reveals that the strength depends on the presence of fibre and it increases with decrease in the spacing of stirrups (increase in the percentage of web reinforcement). The ultimate strength of SFRC beams is analytically obtained. The analytical results were compared with the experimental results.

Acknowledgement I would like to thank Mrs K.GOMATHI, Assistant Professor, Civil Engineering Department, SRM University for guiding me in writing this paper.

References S. Pant Avinash1, R. Suresh Parekar (2010) “Steel fiber reinforced concrete beams under combined torsion-bending-shear” Journal of Civil Engineering (IEB) 38(1) 31-38 [2]. R.P. Dhakal and H.R. Song (2009) “Effect of bond on the behaviour of steel fibre reinforced concrete beams” ICI Journal, Vol. 1,No.4 [3]. Mukesh Shukla (2011)“Behaviour of Reinforced Concrete Beams with Steel Fibres under Flexural Loading” International Journal of Earth Sciences and Engineering ISSN 0974-5904, Volume 04, No 06 SPL, October, pp 843-846 [4]. Florian Finck(2010) “Acoustic Emission analysis of SFRC Beams under cyclic bending loads” Journal of Civil Engineering (IEB) 38(1) 21-28 [5]. Vengatachalapathy.V 1 , Ilangovan.R (2010 )“A Study on Steel Fibre Reinforced Concrete Deep Beams With and without Openings” International journal of civil and structural engineering, Volume 1, No. 3 [6]. Pant Avinash S, Parekar Suresh R(2009) “Steel Fiber Reinforced Concrete Beams Under Bending, Shear And Torsion Without Web Reinforcement” International Journal of Recent Trends in Engineering, Vol. 1, No. 6 [7]. I.J Sluys And R.De Borst (2010)“Computational modelling of impact test on steel fiber reinforced concrete beams” Journal of civil Engineering, Vol. 2, No.4 [8]. Hai H. Dinh and James K. Wight(2009) “Shear behavior of steel fiber-reinforced concrete beams without stirrup reinforcement” ISET Journal of Earthquake Technology, Technical Note, Vol. 44, No. 3-4 [9]. Hamid Pesaran Behbahani1, Behzad Nematollahi(2011) “Flexural behaviour of steel-fibre-added-rc (sfrc) Beams with C30 and C50 classes of concrete” Engineering structures ,Science Direct 37 [10]. Gustavo J. Parra-Montesano’s(2010) “Shear strength of beams with deformed steel fibres” International journal of applied engineering research, Volume 1, no1 [1].

www.iosrjournals.org

4 | Page

IOSR Journal of Mechanical and Civil Engineering (IOSR-JMCE) e-ISSN: 2278-1684,p-ISSN: 2320-334X, Volume 6, Issue 3 (May. - Jun. 2013), PP 05-14 www.iosrjournals.org

Influence of the speed in advance and the laser’s power on the zone affected thermically for steel C45 Sahbi Zantour 1, Meher Gasbar 2, Aleksandr Mikhaylov 3, Hassen Kharroubi 2 1

2

(Department electromechanical, Military Academy Fondek Jdid/ Tunisia) (Agricultural Machinery, Superior School of Engineers of Rural Equipment Medjez El Bab/ University Jendouba, Tunisia) 3 (The Machine-Building Technology Department, Donetsk National Technical University, Ukraine)

Abstract: The Laser cutting is a very important manufacturing technology. But this method has some disadvantages, among which we find the emergence of a Thermically Affected Zone ZAT can dramatically alter the characteristics of the processed material which affects its behaviour during its use. For this, we have tried in this article to study the effect of the forward speed and the laser power in this area (thickness, hardening). In this context, tests were made on steel C45 where we relied on the method of experiment plans to create a mathematical model Significant coefficients are obtained by carrying out a variance analysis ANOVA on the level of 5% of significance. We find that the speed in advance and the power of the laser have a great effect on the ZAT. Keywords: Cutting, Laser CO2, Heat Affected Zone.

I.

Introduction

The ZAT is obtained when the temperature of the sample of steel becomes higher or equal to the specific temperature of the first structure of transformed metal. Steel is transformed into austenite. During the cooling, austenite undergoes a phase shift of the solid state and is transformed into martensite. The sample preserves the same chemical composition that austenite, but it changes of crystalline structure quadratic. Martensite has a great tenacity for simple steels. The part of the steel which undergoes this austenitic transformation is called the ZAT (Thermically Affected Zone). The ZAT can be characterized by mechanical micro hardness testing and optically by an optical microscope. These measurements are used to estimate the width of the ZAT in the morphology of the optimal cut. Neila Jebbari al. [1] have studied the ZAT metal C45 with the assistance of CO2 gas. Fig. 1 shows some examples of grooves with different machining conditions. ZAT, characterized by whitish areas surrounding the grooves are clearly observable in the photos. The morphology of the ZAT is quite repetitive in general. It is possible to note a reduction in width when going towards the bottom of the groove. In Fig. 1 (a), irregularities on the edges of the groove can be observed. This can be attributed to the effect of feedback, which is created when the beam diameter is small impact. The effect of the laser power can be observed in Fig. 1 (c) (P = 2500 W), wherein the depth of the groove is twice as large as those of FIG. 1 (a) and (b) (P = 1500 W). When the cutting speed is high, the groove depth and width of the ZAT become weak due to the short time of laser-matter interaction (Fig. 1 (d)). Fig. 1 (b) is obtained in optimal machining parameters, where a great regularity in the form of the groove and of the TAZ can be noticed. ZZAT widths of ZAT were measured at the top of the groove and the average values are displayed on the experimental curves. Fig.2 shows the evolution of the experimental value of the width of the ZAT (ZZAT) depending on the diameter D of laser impact. (ZZAT) increases to a maximum at D = 0.17 mm, and thereafter, it decreases to a limit value. This result confirms the possibility of using the interaction time (increasing with D) and the laser power density (decreasing with D) to optimize machining.

Fig.1 Morphology of the ZAT www.iosrjournals.org

5 | Page

Influence of the speed in advance and the laser’s power on the zone affected thermically for steel C45 Fig. 3 (ZZAT) shows that changes linearly with the width of the ZAT as a linear function of

and at the same time, it is linear with V-1 (Fig. 4). In general, x V-1, it can be written as follows:

Fig. 5 shows that and depends on the diameter D and the impact of the beam. At the optimum value of the diameter D of the laser impact (D = 0.17 mm) and from the fitted curves of ZZAT (Fig. 5), it can be deduced that = 0.0012mm0.75 (kg s)-0,25 and = 0,027 mm.

Fig.2 Influence of the diameter of the laser beam at the impact on the width of the ZAT

Fig.3 Influence of the power of the laser across the width of the ZAT

Fig.4 Influence of cutting speed on the width of the ZAT

www.iosrjournals.org

6 | Page

Influence of the speed in advance and the laser’s power on the zone affected thermically for steel C45

Fig.5 Influence of and over the width of the ZAT In the same context, there is more research on the HAZ made by IA Choudhury and S. Shirley [2], these experimental trials have studied three different types of thermoplastics polypropylene (PP), polycarbonate (PC) and polymathy methacrylate (PMMA). Davim et al. [3] reported in their research that the HAZ PC was between 0.1 and 0.3mm and that of PMMA was between 0.05 and 0.25mm. Thus the results of this research [2] is reasonably agree with those of Davim et al. [3] The size of the ZAT reflects the quality of laser cutting of polymeric materials. The ZAT data were used to develop the model equations for the input parameters (cutting speed, laser power and air pressure). These equations are as follows:

These equations represent an overview of the extent of the HAZ and its dependence on the parameters of laser treatment. The analysis of variance for all these equations has proved sufficient, at intervals of 95%. Ratios lack of adjustment to the pure error in all cases was less than the F-statistic of 9.01. From these equations, it is clear that the ZAT is extended with the laser power and decreases with the cutting speed and air pressure. The laser power is the most important variable affecting the ZAT. More cutting speed, the lower the duration of combustion and therefore smaller ZAT. Other tests that highlight the influence of the feed rate made by J. Wang and WCK. Wong [4] concluded that if the feed rate and low ZAT will be of considerable thickness, so to get a good surface quality with minimal ZAT, it was necessary to increase the feed rate. This study is in agreement with the total research done with L. Shanjin and W. Yang [5], which has the same tests but changing the type of gas assistant using argon and compressed air. (Fig.11). After CO2 laser cutting, a thermally affected zone ZAT occurred is covered by a thin white layer [6], this area is under the influence directly by the laser power and Pu feed speed Vf. Experience: The experiments were made in the company LASER Industires in Tunis; the machine used is TRUMATIC L3030 that has a capacity of 5000W maximum switching. Was used as the material of steel C45 The chemical composition and mechanical properties are given in the tables (Tab 1 and Tab 2). The reason for choosing this type of material is its frequent use in general industry which gives great importance to the possibility of improving its cutting qualities. The choice of cutting conditions was based on experience and some preliminary tests. Assistant gas pressure is 16 bar, the laser frequency is 20 000 Hz with a distance de1.2 mm from the nozzle. We chose three different speeds (560 mm / min, 1400 mm / min and 2240 mm / min) with 4 different power levels (3 kW, 3.5 kW, 4 kW and 5 kW), where 12 samples. Tab 1: Chemical Composition of the C45 C 45

Fe % 98

S% 0.045

Mn % 0.5 0.8

C% 0.45  0.5

Si % 0.4

P% 0.045

Mo % 0.1

Tab 2: Physical properties of the C45 R (Mpa) Re (Mpa) A% C 45 730 410 17 Each sample was passed through a series of preparation before starting the analysis. Firstly, they were coated, and then did the polishing, then made chemical attack; it was used for the 3% Nital which consists of 3% Nitric Acid and 97% Ethanol. Each sample was visualized by optical microscope to operate the modification of the www.iosrjournals.org

7 | Page

Influence of the speed in advance and the laser’s power on the zone affected thermically for steel C45 ZAT metal and to measure it. They were then passed through the SEM (scanning electron microscope) in order to visualize the white layer. Then, it was microhardness tests to know the evolution of the hardness in the ZAT. After that, we measured the surface roughness with two different devices (3S Surface Scan for 2D profiles and MICROMESURE 2 for 3D profiles). II. Result and discussion : 2-1 Width of ZAT The principle of laser cutting is greatly heated surface of the material to be cut until it melts and the molten part will be removed with the assist gas. But on the other hand, it is absolutely necessary to find a game that is warm to a temperature that could change its metallurgical and mechanical characteristics; this part is called the heat affected zone (ZAT). Found below (Table 3) the results obtained under extreme conditions when measuring the width of the ZAT after an observation with an optical microscope (Fig. 6). Tab.3: Width of the ZAT as a function of cutting parameters Num 1 2 3 4

Vf (mm/min) 560 2240 560 2240

Pu (W) 3000 3000 5000 5000

ZAT (µm) 164,526 126,062 231,638 143,017

Fig. 6 Microscopique picture of the ZAT for a cuts out with surface Vf = 560 mm/min and Pu = 5000 W

III.

Modeling and Discussion

Pour cela on a utilisé la méthode statistique ANOVA avec niveau de confiance de 95% et ont a trouvé les résultats suivants (tab. 4). For better appreciate the effect of the feed rate Vf and the laser power Pu on the thickness of the ZAT, and Vf, Pu and the distance from the plane 0 (surface of the part) of the hardness of the HAZ, modeling was done using the method of experimental design [7]. For this we used the statistical method ANOVA with a confidence level of 95% and found the following results (Table 4). Tab.4: Table ANOVA for the measurement thickness of the ZAT of C45 Source of variance Vf

df 1

SS 8075,340995

MS 8075,341

Ftest 372,7911

>

Ftheoric 7,71

Pu

1

3533,602208

3533,60221

163,12568

>

7,71

Vf,Pu

1

1257,879052

1257,87905

58,068894

>

7,71

Error

4

TOTAL

7

86,6474 12953,4696

21,6618394

According to the ANOVA table (Table 5) we can see that the calculated Ftest feed speed Vf, the laser beam power Pu and their interaction are higher Ftheoric, where these three parameters affect the thickness of the HAZ and we can not ignore any parameters. The model calculation gives the following result:

www.iosrjournals.org

8 | Page

Influence of the speed in advance and the laser’s power on the zone affected thermically for steel C45 Tab.5: Comparaison between the exprimental resukt and the theorical resultsof the ZAT ZAT theorical Num of test Vf (mm/min) Pu (W) ZAT exprimental (µm) Error (µm) 1 560 3000 164,526 164,526 0,00% 2 560 3500 182,419 185,855 1,88% 3 560 4000 199,484 201,738 1,13% 4 560 5000 231,638 231,638 0,00% 5 1400 3000 137,973 139,245 0,92% 6 1400 3500 146,530 141,039 3,75% 7 1400 4000 154,370 150,917 2,24% 8 1400 5000 168,418 168,655 0,14% 9 2240 3000 126,062 126,062 0,00% 10 2240 3500 130,955 135,467 3,45% 11 2240 4000 135,347 141,751 4,73% 12 2240 5000 143,017 143,017 0,00% According to the model found above, it can be concluded that the ZAT and proportional to the laser beam power (Pu) (fig.7) and inversely proportional to the feed rate (Vf) (fig.8).

Fig.7: Effetc of the power of the laser beam Pu on the thickness of the ZAT

Fig .8: Effect of the speed in advance Vf on the thickness of the ZAT This conclusion can also be seen fromDiagramme the curve effects (Fig. 9), which we can conclude the iso curve response of the de surface de ZAT et Vf ; Pu thickness of the ZAT (Fig. 10).

240 210 ZA T

(µm)

180 150 120 4800 4200 Pu

(W)

3600 3000

2000

1600

1200 Vf

800

400

(mm/min)

Fig.9: Curve of the effects of the speed in advance and the power of the laser Pu on the thickness of the ZAT www.iosrjournals.org

9 | Page

Influence of the speed in advance and the laser’s power on the zone affected thermically for steel C45 Contour Plot of ZAT (µm) vs Vf (mm/min); Pu (W) 2200

ZA T (µm) < 140 – 160 – 180 – 200 – 220 > 220

140 160 180 200

2000

Vf (mm/min)

1800 1600 1400 1200 1000 800 600 3000

3500

4000 Pu (W)

4500

5000

Fig.10: Curve iso answer thickness of the ZAT I-2 Hard facing The Hardness is usually defined as resistance of the material resiliently stressed. It is considered one of the most important properties of metals, mainly because the hardness test is a very simple test to drive though included some complicated phenomena such as plastic multiaxial behavior. In this study we focused only on the hardness in the HAZ only. The microhardness measurements were performed with the measuring device: Testwell. Measurement type: Vickers hardness at a load of 50g. Hold time (Duel time) = 10.

Fig.11 : Hardness Test. The penetration is a straight pyramid with a square base and with an apex angle of 136 ° under a load F = 0.5 N. Measuring the diagonal "d" of the footprint. The test is performed at room temperature; the load is applied gradually and steadily until the selected force is reached. The holding time of the load is 10 seconds. The impression obtained has a shape of a pyramid with a square base (Fig.11). A light spot appears on an observation screen of square shape (base of the pyramid). It takes the average of the two diagonals of the impression: d, and calculates the Vickers hardness (HV50) in the form:

Found below (Table 6) the results obtained under extreme conditions during the hardness measurement made with the device of micro hardness. Tab.6: Measurements of hardness according to the parameters of cut and surface’s depth Num 1 2 3 4 5 6 7 8

Vf (mm/min) 560 2240 560 2240 560 2240 560 2240

Pu (W) 3000 3000 5000 5000 3000 3000 5000 5000

www.iosrjournals.org

D (µm) 10 10 10 10 510 510 510 510

Hardness (VH) 204 197 231 218 187 175 203 189

10 | Page

Influence of the speed in advance and the laser’s power on the zone affected thermically for steel C45

Fig.12: Print of the Micro hardness of the various zone Constituting the termically affected zone. Modeling and discussion: For better appreciate the effect of the feed rate Vf and the laser power Pu on the thickness of the ZAT, and Vf, Pu and the distance from the plane 0 (surface part) of the hardness of the ZAT, modeling was done using the method of experimental design [7]. For this we used the statistical method ANOVA with a confidence level of 95% and found the following results (Table 4). Tab.7: ANOVA Table (Hardness of C45 steel) Source of df SS MS Ftest Ftheorical variance Vf 1 1088,297858 1088,29786 2,0749306 5,32 Pu 1 3092,596658 3092,59666 5,8962934 5,32 D 1 4550,58 4550,58 8,6760602 5,32 Vf, Pu 1 24,808968 24,808968 0,0473004 5,32 Vf, D 1 10,829858 10,829858 0,020648 5,32 Pu, D 1 150,580658 150,580658 0,2870946 5,32 Vf, Pu, D 1 12,280968 12,280968 0,0234147 5,32 Error 8 4195,9875 524,498436 TOTAL 7 13125,96245 The ANOVA table and the Fisher test shows that a significant level of 5%. Fcalculated for Pu and D are higher Ftheorical so they are meaningful answer. But Fcalculated of Vf is less than Ftheorical but it is not negligible, so Vf is quite significant influence on the response. On all interactions, Fcalculated shows that they can be ignored and do not affect the response. The model calculation gives the following result:

Tab.8: Comparaison between exprimental results and theorical results of hardness

10 10 10 10 10

Theorical hardness(VH) 209,888 215,4493 220,3856 228,8884 200,6489

Exprimental hardness (VH) 204 214,5 223,5 231 205,2

3500

10

205,9653

210

2%

1400

4000

10

210,6843

220,65

5%

1400

5000

10

218,8128

225,45

3%

9

2240

3000

10

196,0687

196,5

0%

10

2240

3500

10

201,2638

204

1%

11

2240

4000

10

205,8751

211,11

3%

12

2240

5000

10

213,818

217,5

2%

Num of test

Vf (mm/min)

Pu (W)

D (µm)

1 2 3 4 5

560 560 560 560 1400

3000 3500 4000 5000 3000

6

1400

7 8

www.iosrjournals.org

Error 3% 0% 1% 1% 2%

11 | Page

Influence of the speed in advance and the laser’s power on the zone affected thermically for steel C45 1

560

3000

210

191,8503

197,25

3%

2

560

3500

210

196,9336

202,5

3%

3

560

4000

210

201,4457

207

3%

4

560

5000

210

209,2178

207,45

1%

5

1400

3000

210

183,4052

191,05

4%

6

1400

3500

210

188,2647

196,5

4%

7

1400

4000

210

192,5782

202,5

5%

8

1400

5000

210

200,0081

209,7

5%

9

2240

3000

210

179,2186

178,95

0%

10

2240

3500

210

183,9672

187,5

2%

11

2240

4000

210

188,1822

191,22

2%

12

2240

5000

210

195,4426

192

2%

1

560

3000

110

195,547

200,7

3%

2

560

3500

110

200,7282

211,5

5%

3

560

4000

110

205,3273

217,5

6%

4

560

5000

110

213,2491

226,5

6%

5

1400

3000

110

186,9391

192,22

2%

6

1400

3500

110

191,8923

202,5

6%

7

1400

4000

110

196,2889

208,95

6%

8

1400

5000

110

203,862

207,95

2%

9

2240

3000

110

182,6718

187,5

3%

10

2240

3500

110

187,512

195

4%

11

2240

4000

110

191,8082

191,65

0%

12

2240

5000

110

199,2084

202,95

2%

According to the model found above, it can be concluded that the HAZ and proportional to the laser beam power (Pu) and inversely proportional to the feed rate (Vf) and the depth from the surface (D) (Fig.13). This conclusion can be seen from curve effects (Fig. 14), where one can conclude Iso curve responses hardness (fig.15, fig.16, fig.17).

MB

Fig.13: Evolution of the hardness of the steel C45 according to the depth D

www.iosrjournals.org

12 | Page

Influence of the speed in advance and the laser’s power on the zone affected thermically for steel C45 Main Effects Plot for Dureté Data Means

Vf

Pu

210 205 200

Mean

195 190 560

2240

3000

5000

P 210 205 200 195 190 10

510

Fig.14: Effects of the speed in advance Vf and the power of the laser Pu and the depth D on hardness of C45 Contour Plot of Dureté (VH) vs Vf (mm/min); Pu (W) 2200

Dureté (VH) < 200 200 – 205 205 – 210 210 – 215 215 – 220 220 – 225 225 – 230 > 230

2000

Vf (mm/min)

1800 1600 1400 1200 1000 800 600 3000

3500

4000 Pu (W)

4500

5000

Fig.15: Iso response curve hardness according to the speed in advance Vf and the power of the laser beam Pu (D = 10µm) Contour Plot of Dureté (VH) vs Vf (mm/min); D (µm) 2200

Dureté (VH) < 175 – 180 – 185 – 190 – 195 – 200 > 200

2000

175 180 185 190 195

Vf (mm/min)

1800 1600 1400 1200 1000 800 600 100

200

300 D (µm)

400

500

Fig.16: Iso response curve hardness versus the depth from the surface D and the laser beam power Pu (Vf = 560 mm / min) Contour Plot of Dureté (VH) vs Pu (W); D (µm) 5000

Dureté (VH) < 190 190 – 200 200 – 210 210 – 220 220 – 230 > 230

Pu (W)

4500

4000

3500

3000

100

200

300 D (µm)

400

500

Fig.17: Iso response curve as a function of the hardness feedrate Vf and the depth from the surface D (Pu = 3000 W) www.iosrjournals.org

13 | Page

Influence of the speed in advance and the laser’s power on the zone affected thermically for steel C45 IV.

Conclusion:

After CO2 laser cutting of steel C45, there is a ZAT heat that appears undergoing metallurgical and physical changes important this area is under the influence directly: 

The speed in advance : - Is proportional to the thickness of the ZAT. - Is inversely proportional to the hardness of the ZAT.



The Laser’s power - - It is proportional to the thickness of the ZAT. - - It is proportional to the hardness of the ZAT.

Bibliographic reference: [1]

[2] [3] [4] [5] [6] [7]

Neila Jebbari, Mohamed Mondher Jebari, Faycal Saadallah, Annie Tarrats-Saugnac, Raouf Bennaceur, Jean Paul Longuemard (2007). "Thermal affected zone obtained in machining steel XC42 by high-power continuous CO2 laser". Optics and Laser Technology 40 (2008) 864-873. I.A. Choudhury, S. Shirley (2009). "Laser cutting of polymeric materials: An experimental investigation". Optics and Laser Technology 42 (2010) 503-508. Davim JP, Barricas N, Conceição M, Oliveira C. "Some experimental studies on CO 2 laser cutting quality of polymeric materials". Journal of Materials Processing Technology 2008; 198 (1-3):99-104. J. Wang, W.C.K. Wong. “CO2 laser cutting of metallic coated sheet steels”. Journal of Materials Processing Technology 95 (1999) 164-168. Lv. Shanjin, Wang Yang. “An investigation of pulsed laser cutting of titanium alloy sheet”. Optics and Laser in Engineering 44 (2006) 1067-1077. B. Tirumala Rao, Rakesh Kaul, Pragya Tiwari, A.K. Nath. “Inert gas cutting of titanium sheet with pilsed mode CO 2 laser”. Optics and Laser in Engineering 43 (2005) 1330-1348. Sivarao, T.J.S. Anand, Ammar, Shukor. “RSM Based Modeling for Surface Roughness Prediction in Laser Machining”. Inetrnational Journal of Engineering & Technology IJET-IJENS Vol: 10 No 04.

www.iosrjournals.org

14 | Page

IOSR Journal of Mechanical and Civil Engineering (IOSR-JMCE) e-ISSN: 2278-1684,p-ISSN: 2320-334X, Volume 6, Issue 3 (May. - Jun. 2013), PP 15-24 www.iosrjournals.org

Review on Design Optimization of Liquid Carrier Tanker for Reduction of Sloshing Effects 1

Sunil M Mahajan , 2Ashwin D Patil, 3V. N. Bartaria (Asst.Professor, SITRC Sandip Foundation ) (PG Student of SRES-COE, Kopergaon) (Professor & HOD, Mechanical Engineering, LNCT, Bhopal)

Abstract: This Paper Reviews Briefly The Current Research On Sloshing And Its Effect In Liquid Carrier Tanker. The Aim Of This Paper To Study The Basics Of Sloshing And Its Prevention (Mainly In Liquid Carrier Tanker) The Liquid Sloshing Is Free Surface Fluctuation Of Liquid When Its Container Is Excited By External Vibrations Such As Earthquakes. The Liquid Sloshing May Cause Various Engineering Problem, For Example Instability Of Ships In Aero Engineering And Ocean Engineering, Failures On Structural Systems Of The Liquid Container. The Tanker Used For The Transportation Of Liquid Over The Road-Ways Is An Integral Part Of The Carrier/Vehicle. The Tanker Is Expected To Withstand The Unbalanced Forces On Account Of Transit Over Uneven And Irregular Surfaces/Contours Of The Road As Also Due To Sudden Acceleration Or Deceleration (Due To Application Of Brakes). Keywords-Sloshing, Impact, Baffle, Simulation

I.

INTRODUCTION

Sloshing can be defined as dynamic load acting over a tank structure as a result of the motion of a fluid with free surface confined inside the tank. Liquid sloshing is a kind of wave motion inside a partially filled tank. The sloshing phenomenon is of great practical importance to the safety of the liquid transport. Under external excitations, with large amplitude or near the natural frequency of sloshing, the liquid inside a partially filled tank is in violent oscillations. In this paper we will see the background behind the sloshing phenomena to define the problem in proper manner. Also we will see the research work carried related to this problem, proposed methodology to go for solution and scope of work. The tanker used for the transportation of liquid over the road-ways is an integral part of the Carrier/ Vehicle. The tanker is expected to withstand the unbalanced forces on account of the transit over uneven and irregular surfaces/ contours of the road as also due to sudden acceleration or deceleration (due to application of brakes). As a result, `sloshing’ of the liquid is experienced within the tanker. Different aspects of analyses are necessary to design the tanker but sloshing analysis is also one of the prominent aspects for reducing its detrimental effects over structure of tanker. Sloshing can be the result of external forces due to acceleration/deceleration of the containment body. Of particular concern is the pressure distribution on the wall of the container reservoir and its local temporal peaks that can reach as in road tankers twice the rigid load value. In road tankers, the free liquid surface may experience large excursions for even very small motions of the container leading to stability problems. Analysis of the sloshing motion of a contained liquid is of great practical importance. Motion of a fluid can persist beyond application of a direct load to the container; the inertial load exerted by the fluid is time-dependent and can be greater than the load exerted by a solid of the same mass. This makes analysis of sloshing especially important for transportation and storage tanks. Due to its dynamic nature, sloshing can strongly affect performance and behavior of transportation vehicles, especially tankers filled with oil. In fact, a significant amount of research has gone into developing numerical models for predicting fluid behavior under various loads. Hence liquid sloshing is a practical problem with regard to the safety of transportation systems, such as oil tankers on highways, liquid tank cars on railroads, oceangoing vessels with liquid cargo, propellant tank used in satellites and other spacecraft vehicles, and several others

II.

LITERATURE REVIEW

Many researchers are studying the sloshing phenomena since 1950 for transportation of fluids as well as gases for better national communication. Here, in this chapter we reviewed some literature regarding sloshing and analyzing method.

www.iosrjournals.org

15 | Page

Review on Design Optimization of Liquid Carrier Tanker for Reduction of Sloshing Effects Donald Liu[1] has reported that the normal practice is to base the tanker's design on engineering analysis, and to use rule equations only as a check, for guidance and control. The potential benefit from utilizing engineering analysis is stress verification and rational distribution of steel, which results in steel weight optimization without sacrificing strength. Also it has been suggested the different types of engineering analysis required to design the tanker which are mentioned as below:  Ship Moment: The first step in the analysis consists of calculating the hull girder longitudinal vertical bending moment, for all anticipated loaded and ballast conditions.  Ship Motion: The next step in the analysis is to calculate the vertical, lateral and torsional dynamic components of the hull girder bending moments, as well as the motions, point accelerations and pressure distribution along the vessel's length.  3-D Global 3-D Finite Element Analysis (FEM): The data obtained from ship moment and ship motion above are used to carry out a three-dimensional global FEM analysis for the entire vessel, or for a portion of the hull girder. This analysis gives the overall structural response in the form of element stresses and displacements.  Sloshing Analysis: This analysis is used to determine the dynamic loads on the tank boundaries from the motion of the fluid within the tank due to ship movements in a seaway.  Thermal Stress Analysis: This analysis provides the distortions and stresses in the hull structure induced by non-linear temperature differentials in vessels carrying hot cargoes.  Fatigue and Fracture Mechanic Analysis: Based on the combined effect of loading, material properties and flaw characteristics, this analysis predict the service life of the structure, and determine the most effective inspection plan.  Vibration Analysis: This analysis is used to determine the extent of vibrations in the ship structure induced by the interaction of fluids, structure, machinery and propellers. Jean Ma and Mohammad Usman Jean [2] presented that, the sloshing phenomenon in partially filled fuel tanks is more pronounced when vehicle experience a sudden start or stop. Sloshing is undesired because it produces noise, high impact force on the tank walls and the challenge of low fuel handling. Today the solution for containing sloshing is to incorporate baffles inside the tank. The presence of baffles dissipates the energy that is induced by the fuel motions. Design of baffles is necessary step during the design of a fuel tank to meet required performance specification in service. After literature review it can be seen that, this problem is having FSI nature so the modeling and analyzing method should be susceptible to adopt such nature of problem. Here, in this chapter we will see different approaches to solve FSI problem along with explanation about FEA code useful for analysis. Approaches to solve FSI problem: This type of problem can be modeled in basic four approaches which are used for fluid structure interaction problem a. Lagrangian approach b. Euler approach Lagrangian approach Lagrangian formulation is usually used for describing a solid mechanics problem. The problem is described with a high number of mass particles, where the motion of every single particle is being observed in space and time. The problem is exactly defined when the motion of all the particles is known. The Lagrangian formulation is very simple and easy to use for one or only a few mass particles. However, the method becomes very complicated and complex for description of high number of mass particles. (Fig.1) In the Eulerian formulation the problem is being observed at one point in space which does not follow the motion of the single particle. In one time step t several mass particles may pass the observed point. Their motion is exactly determined in the moment of passing through that point. In the observed point the field variables are time dependent. (Fig.2)

www.iosrjournals.org

16 | Page

Review on Design Optimization of Liquid Carrier Tanker for Reduction of Sloshing Effects

Fig. 1 Langrangian Formulation

Fig. 2 Eulerian Formulation The basic difference between the Lagrangian and the Eulerian formulation is that at the Lagrangian formulation the magnitudes x, y and z are variable coordinates of a moving particle. At the Eulerian formulation those coordinates represent steady coordinates of the defined field point A Lagrangian mesh should be used to model structural parts of a problem. In the Lagrangian formulation one finite element represents the same part of the material throughout the course of the analysis. The fluid domain can be described with a material model which skips the calculation of deviatoric stresses. By defining a low bulk modulus for fluids such as water, the elastic shear forces become negligible, and by using a low yield stress, fast tran sition to plasticity can be achieved (e.g. by only considering the gravitation). Under high dynamic loading, the shear forces and any unreal introduced forces become negligible in comparison to the inertial forces of the fluid. (Fig.3) illustrates the solution process of a simple fluid problem using the Lagrangian formulation. It is presumed that the loading influences only the central node. The result of the loading is the shift of that node in a computational time step. If the influence of the loading does not stop or change, the node takes a new position in the next time step and the mesh deforms even more, since the mesh follows the material flow.

Fig. 3 Mesh Deformation in Langrangian Formulation

Eulerian Approach: It is also possible to apply the Eulerian formulation for fluid flow analyses, where the fluid flow through the fixed mesh in a space is observed. The material point moves from one finite element to another and the finite element mesh does not move or deform. Although the Eulerian mesh in appears not to move or deform during the analysis, it does actually change its position and form only within the single time step. The reason for this is the use of Lagrangian formulation in single time steps, which is much more advanced in. The Eulerian mesh is treated in a special way (Fig.3.4). To illustrate the use of an Eulerian mesh the same example is used as in the previous chapter. Because of the central node loading, the observed node changes its position during one computational time step (mesh deforms). After the time step the analysis stops and the following two approximations are performed: - Mesh smoothing: all the nodes of the Eulerian mesh that have been displaced due to loading are, moved to their original position; - Advection: the internal variables (stresses, flow fields, velocity field) for all the nodes that have been moved are recomputed (interpolated) so that they have the same spatial distribution as before the mesh smoothing. In this way the mesh smoothing does not affect the internal variable distribution. www.iosrjournals.org

17 | Page

Review on Design Optimization of Liquid Carrier Tanker for Reduction of Sloshing Effects The described procedure is being repeated for each time step of the analysis and provides the analyst with a non-movable and undeforms able Eulerian mesh.

Fig. 4 Mesh Deformation in Eulerian Formulation

Fig.5 Aligning Euler and Langrangian Elements

Fig.6 Tanker Mesh

Fig.7 Meshing of LPG in Unbaffled Tank

Fig.8 Meshing of LPG Tank with Full Enclosed Baffle at the

Fig.9 Meshing of LPG Tank with Two Modified Baffles

Fig.10 Eulerian and Langrangian Meshing of LPG in a Tank with Modified Baffles

www.iosrjournals.org

18 | Page

Review on Design Optimization of Liquid Carrier Tanker for Reduction of Sloshing Effects Simulation of sloshing: Simulation results are involving different iterations in that first iteration is Sloshing of LPG without baffle. After taking possible iterations graphs are plotted to show the pressure and velocity developed during respective condition of sloshing. Simulation results are involving different iterations in that first iteration is Sloshing of LPG without baffle which includes various stages of sloshing simulation at start which shown in Fig.11 and Fig.12 respectively. Similarly, disturbance in fluid is clearly shown in Fig.13 and Fig.14 Fig.15 to Fig.19 shows the gradually increasing of LPG sloshing amplitude with enclosed full baffle inserted in a tanker. Similarly, Fig.20 to Fig.24 shows the simulation of LPG sloshing with one modified baffle at the center of tank with various views. Then, simulation result of LPG sloshing with two modified baffles is given in Fig.25 to Fig.29 with various stages of sloshing phenomena.

Fig.11 Iteration 1 Sloshing of LPG without Baffle

Fig.12 Iteration 1 Sloshing of LPG without Baffle in Front View

Fig.13 Iteration 1 Sloshing of LPG without Baffle showing Side View

Fig.14 Iteration 1: Sloshing of LPG without Baffle showing Disturbance of Fluid

Fig.15 Iteration 2: Sloshing of LPG in a Tank with Enclosed Full Baffle

Fig.16 Iteration 2: Sloshing of LPG in a Tank with Enclosed Full Baffle at Initial Stage

www.iosrjournals.org

19 | Page

Review on Design Optimization of Liquid Carrier Tanker for Reduction of Sloshing Effects

Fig.17 Iteration 2: Sloshing of LPG in a tank with enclosed full baffle at final stage

Fig.18 Iteration 2: Sloshing of LPG in a tank with enclosed full baffle showing side view

Fig.19 Iteration 2: Sloshing of LPG in a tank with enclosed full baffle showing disturbance

Fig.20 Iteration 3: Sloshing of LPG in a tank with one modified baffle

Fig.21 Iteration 3: Sloshing of LPG in a tank with one modified baffle showing lift of fluid along baffle

Fig.22 Iteration 3: Sloshing of LPG in a tank with one modified baffle showing peak amplitude

Fig.23 Iteration 3: Sloshing of LPG in a tank with one modified baffle with side view

www.iosrjournals.org

20 | Page

Review on Design Optimization of Liquid Carrier Tanker for Reduction of Sloshing Effects

Fig.24 Iteration 3: Sloshing of LPG in a tank with one modified baffle showing disturbance of fluid in side view

Fig.25 Iteration 4: Sloshing of LPG in a tank with two modified baffle

Fig.26 Iteration 4: Sloshing of LPG in a tank with two modified baffle showing lift of fluid along baffle

Fig.27 Iteration 4: Sloshing of LPG in a tank with two modified baffle showing peak amplitude of fluid

Fig.28 Iteration 4: Sloshing of LPG in a tank with two modified baffle in side view

Fig.29 Iteration 4: Sloshing of LPG in a tank with two modified baffle showing disturbance of fluid

Graphs: After taking possible iterations graphs are plotted to show the pressure and velocity developed during respective condition of sloshing

www.iosrjournals.org

21 | Page

Review on Design Optimization of Liquid Carrier Tanker for Reduction of Sloshing Effects

Fig.30 Sloshing of LPG without baffle

Fig.31 Sloshing of LPG in a tank with one modified baffle

Fig.32Sloshing of LPG in a tank with two modified baffle

Fig.33 Sloshing of LPG without baffle

Fig.34 Sloshing of LPG in a tank with one modified baffle

Fig.35 Sloshing of LPG in a tank with two modified baffle

www.iosrjournals.org

22 | Page

Review on Design Optimization of Liquid Carrier Tanker for Reduction of Sloshing Effects

Fig.36 Sloshing of LPG in a tank with two modified baffle

Fig.37 Comparison of LPG sloshing without baffle and with enclosed full baffle

Fig.38 Comparison of LPG sloshing with one modified baffle and two modified baffles

Fig.39 Comparison of velocity variations of four iterations

Findings: From the simulation result of every iteration it can be found that, sloshing is reduced in considerable amount with the help of two modified baffles located at 2500 mm apart from each other. Also the maximum pressure generated in various iterations due to sloshing of fluid is as follows: Table.1 Maximum pressure generated in various cases Iteration No. 1 2 3 4

Case Name LPG sloshing without baffle LPG sloshing with enclosed full baffle LPG sloshing with one modified baffle LPG sloshing with two modified baffles

www.iosrjournals.org

Time in (sec)

Pressure in (N/m2)

0.00189

2.58 x 109

0.00200

1.17 x 109

0.0020

9.88 x 106

0.0050

2.87 x 106

23 | Page

Review on Design Optimization of Liquid Carrier Tanker for Reduction of Sloshing Effects Similarly the maximum velocity developed in various iterations because of sloshing of fluid is as follows: Table.2 Maximum velocity developed in various cases Iteration No. 1 2

3

4

Case Name LPG sloshing without baffle LPG sloshing with enclosed full baffle LPG sloshing with one modified baffle LPG sloshing with two modified baffles

Time in sec.

Velocity in m/s

0.049

62.13

0.045

53.48

0.0475

43.85

0.058

35.8

Therefore, from Table 1 and 2 we can say that, more time is required to generate maximum pressure and velocity in the case of two modified baffles compared to other cases. Also the maximum stress developed in modified baffle having 20 mm thickness is 2.20 x 108 Pa which is lower than the yield strength of baffle material (ASTM A576). Similarly, stress on wall of tanker is 2.05 x 108 Pa lower than yield strength of tanker material (AISI 1040).

III.

CONCLUSION

Analysis of cylindrical liquid carrier tanker is carried out using the finite element method. Studies of various methods in FEA are done and one particular method is selected to model fluid-structure interaction problem. These interaction problems are quite complex and they have been challenging as well. Using MSc-Dytran software defined problem has been modelled which uses Arbitrary Langrangian Eulerian method (ALE). Elements depicting the properties of the tanker and the liquid are selected and coupled. Water is selected first to see the nature of sloshing and to get the maximum pressure. Then iterations are taken for real liquid in the problem i.e. Liquefied Petroleum Gas (LPG). Analysis is done to obtain sloshing patterns, pressure and velocity parameters in different cases. From the results that obtained in analysis following conclusions can be drawn We can accept the challenge for transportation of liquid in partially filled tankers by using baffles in proper shape, numbers and location. In this problem sloshing of LPG is reduced in half filled cylindrical tanker at the speed of 40kmph by using two modified baffles. Also effect of sloshing over tanker and baffles are decreased with proper thickness. The pressure and velocity developed in two baffled condition is lower than unbaffled, one baffled and enclosed baffled condition. So it is recommended to use two modified baffles, 2500mm apart from each other with thickness 20 mm which can decrease the sloshing considerably and sustain the sloshing pressure

REFERENCES [1]. [2]. [3]. [4]. [5].

Donald Liu “Tanker Spills Prevention By Design” National academic Press, Publication Year1991 ISBN-10: 0-309-04377-8 ,pp.208 to 213 Jean Ma and Mohammad Usman Jean “Modeling of Fuel Sloshing Phenomenon Considering Solid-Fluid Interaction” 8TH International LSDYNA Users Conference Fluid/Structure Ranjit Babar and V.Katkar “Simulation of Fuel Tank Slosh Test-Coupled Eulerian-Langrangian Approach” Tata Technologies, TATA motors Ltd. Pimpri, Pune Kim Hyun-Soo and Lee Young-Shin “Optimization design technique for reduction of sloshing by evolutionary methods” Journal of Mechanical Science and Technology,Number1/January 2008,Vol.22 Jang-Ryong Shin and Kyungsik Choi and Sin-Young Kang “An Analytical Solution to Sloshing Natural Periods for a Prismatic Liquid Cargo Tank with Baffles” Proceeding of Sixteenth International Offshore and Polar Engineering Conference, California, USA, May 28June 2, 2006

.

www.iosrjournals.org

24 | Page

IOSR Journal of Mechanical and Civil Engineering (IOSR-JMCE) e-ISSN: 2278-1684,p-ISSN: 2320-334X, Volume 6, Issue 3 (May. - Jun. 2013), PP 25-36 www.iosrjournals.org

Determination of Mechanical Properties of Different Materials Using Inverse Finite Element Procedure for the Miniature Specimen Test Simulation Dr. V Chitti Babu, Mr. Bade Venkata Suresh, Sri P. Govinda Rao, Associate Professor,Department Of Mechanical Engineering Gmrit, Rajam, Srikakulam,Ap Pg Student, Department Of Mechanical Engineering Associate Professor,Department of Mechanical Engineering GMRIT, RAJAM, Srikakulam,AP

Abstract: Determination of mechanical properties of operating components is a key element in defining the strategies for maintenance and repair for extending component life. The assessment of the present health has to be done without disturbing the structural integrity and functional capabilities of the plant components. To achieve this, a miniature specimen can be extracted from an in-service component and its properties can be evaluated. This helps in avoiding the removal of large size samples from the in-service component for the evaluation of current material properties. The miniature specimen also has an advantage in finite element modelling with respect to computational time and memory space. An inverse finite element simulation has also been carried out on the miniature specimens of different varieties of steels. Inverse finite element simulated load-displacement curve has been compared with the experimental results taken from the literature. Properties of different varieties of steels are obtained when a perfect match is found between inverse FE and experimental results. Inverse finite element simulation procedure helps in finding the constitutive behaviour of the unknown material in combination with experimental output of miniature test. The approach seems to have potential to predict the mechanical properties of the inservice components. Keywords— Modelling, FEM, Miniature Specimen

I.

Introduction

Miniature specimen test techniques enable the characterization of mechanical properties using extremely small volume of the material. Though they have their origin in materials development programmes in industries. The miniature specimen test techniques find very promising applications in the field of remaining life assessment, failure analysis, properties of weldments, coatings etc. Determination of mechanical properties of operating components is a key element in defining the strategies for maintenance and repair for extending component life. The assessment of the present health has to be done without disturbing the structural integrity and functional capabilities of the plant components. To achieve this, a miniature specimen can be extracted from an in-service component and its properties can be evaluated. This helps in avoiding the removal of large size samples from the in-service component for the evaluation of current material properties. The miniature specimen also has an advantage in finite element modeling with respect to computational time and memory space.

II.

ANALYSIS OF MINIATURE SPECIMEN TEST

1.1 Materials: To study the effectiveness of the miniature specimen test simulation in getting the material properties from miniature specimens, five different materials have been selected. All these five materials show a broad range in their mechanical behaviour. A brief description of the selected materials is given below: ALUMINIUM ALLOY (AR66) This is one of the aluminium alloys. Zinc is the principal alloying element in this group. When it is combined with smaller percentages of magnesium and, in some cases copper, it results in heat-treatable alloys of very high strength. This alloy is used in aerospace applications, automobile industries, railways, shipping industry etc. and in many other engineering applications. DIE STEEL (D3) This is high carbon chromium steel with 2% Carbon, 12% Chromium, 0.3% Manganese and 0.3% Silicon. This type of steel has high dimensional stability with added wear resistance coupled with excellent edge holding qualities which are commonly applied in thread rolling dies, cold extrusion tools and dies, draw plates and dies, cutters, measuring tools, pressure casting moulds, blanking tools, reamer, finishing rolls for tyre mills and also for many applications in mechanical engineering. www.iosrjournals.org

25 | Page

Determination of Mechanical Properties of Different Materials Using Inverse Finite Element CHROMIUM HOT WORK STEEL (H11) This is a low alloy steel with Manganese (0.4%), Silicon (1%), Chromium (5%) and Molybdenum (1.1%). It has very good high temperature characteristics and excellent toughness combined with receptivity to heat checking, which is commonly used for extrusion dies, gripper dies, rams, drop forging dies, pressure casting moulds for light alloys, metal track pressure tools, high stressed internal boxes, little tube pressing mandrels for watercooling, die-casting tools etc. which are employed in many components and structures for various industries. LOW CARBON STEEL (LC) In this steel Carbon 0.19%, Manganese 0.44% and Silicon 0.14% are present. Other elements are Copper, Nickel, Chromium and Molybdenum etc. and their weight percentages are presented in Table. This steel is highly versatile and useful. It can be machined and worked into complex shapes, has low cost and good mechanical properties. It is widely used in automotive industries and in other areas where formability and stiffness are sought. MEDIUM CARBON STEEL (MC) In this steel carbon .35%, manganese 0.68% and silicon 0.15% are present. This type of steel has high wear resistance and can be used for components which are subjected to severe abrasion, wear or high surface loading. This steel is preferred for heat treated components having large sections and subjected to exacting requirements. Table 1.1 Uniaxial properties of the above materials: Different Mechanical Properties Sl. No .

Materia ls

Young’s Modulus (GPa)

Yield strength (MPa)

Tensile Strengt h (MPa)

Poisson ’s ratio ν

AR66 alloy

70.690

506.90

567.92

0.25

2.

D3 steel

201.00

483.00

745.71

0.30

3.

H11 steel

195.00

474.00

693.45

0.30

1.

III.

Finite Element Simulation Of Miniature Specimen Test

ABAQUS/CAE 6.12-2 Student Edition has been used for the finite element modeling of the given problem. The shell type finite element model has been created where in a circular dumb-bell shaped miniature specimen with two circular pins are taken to model the miniature specimen test simulation. 3.1 Geometric modeling: The test specimen has been modeled with 2-d four noded quad shell type element of plain stress type with thickness 0.5mm because of the shape of the body and convenience to mesh geometry. In order to represent the situation exactly similar to the experimental set up, A circular section of radius 5mm is taken to model. Two circular cut sections of radius 0.5mm at a distance of 3 mm from the center are modeled at the top and bottom of the specimen to grip the specimen. Two more cut circular sections of radius 2mm at a distance of 2.5mm from the center are modeled on left and right side of the specimen and two rectangular cut sections of 3mm width and 2mm height are modeled at right and left side of the specimen at distance of 2mm from the center to simulate the actual geometry of the specimen. Then the auto trim option is used to subtract the two circles of radius 2mm, another two circles of radius 0.5mm and rectangles from the larger circle of radius of 5mm to obtain the required model of circular dumb-bell shaped specimen.

www.iosrjournals.org

26 | Page

Determination of Mechanical Properties of Different Materials Using Inverse Finite Element TABLE 2.1 TRUE STRESSES AND TRUE STRAINS s. no 1. 2. 3. 4. 5. 6. 7. 8. 9. 10 11 . 12 .

AR66 alloy True True stress plastic 506.90 N/m2 0.0000 strain 512.32 0.0025 520.04 0.0035 526.58 0.0048 537.57 0.0084 552.50 0.0171 568.66 0.0299 591.39 0.0509 606.47 0.0667 612.65 0.0737 620.25 0.0826 626.75 0.0908

D3 steel True True stress plastic 484.4 N/m2 0.0000 strain 520.3 0.0056 9 544.7 0.0094 581.8 0.0153 616.4 0.0213 674.6 0.0335 712.9 0.0438 742.6 0.0541 759.9 0.0613 778.5 0.0705 801.5 0.0844 826.9 0.1041

H11 steel True True stress plastic 474.2 N/m2 0.0000 strain 518.8 0.0128 573.4 0.0173 633.3 0.0258 672.1 0.0352 697.6 0.0443 720.2 0.0555 734.6 0.0649 748.0 0.0761 758.5 0.0866 764.1 0.0932 772.1 0.1438

TRUE STRESS (σt): σ = P/Ai; Ai = AoLo/Li; σ = PLi/(AoLo); σt = P/Ao; Li = Lo+ΔL; σt = σ (1+ε) Li/Lo =1+ ΔL, TRUE STRAIN (εt): εt = dLi/Li; Li = Lo+ ΔL; εt = ln (1+ε); Ai = Instantaneous Cross Sectional Area; A0 = Original Cross Sectional Area; σ t = True Stress; σ = Nominal Stress; ΔL=Change in Length; 3.2 Material modeling: In the elastic analysis of simulation, the material properties of column were defined by elastic modulus and Poisson’s ratio. In the plastic analysis stage, material nonlinearity or plasticity was included in the ABAQUS by incorporating the yield stress and plastic strain values. In the finite element simulation of miniature dumb-bell specimen, the plastic properties are defined together with the isotropic hardening rule. It means that the yield surface size changes uniformly in all the directions such that the yield stress increases in all stress directions as plastic strain occurs. The curve of true stresses and true plastic strains were specified during the tensile test defines the basic flow characteristics of the material. The incremental plasticity model required the true stress-strain curve to the ultimate point of the stress-strain curve.

Fig: 2.2 Finite element model of the miniature specimen www.iosrjournals.org

27 | Page

Determination of Mechanical Properties of Different Materials Using Inverse Finite Element IV.

FINITE ELEMENT RESULTS AND DISCUSSIONS

4.1 AR66 ALLOY: The load-elongation solution is plotted below

FIG.3.1:Displacement of AR66 miniature specimen The stress distribution for AR66 is shown in below figure

FIG.3.1.1 Stress distribution of AR66 miniature specimen The experimental load-elongation curve is taken from the literature is shown in graph 2.1

Graph 3.1.2 The Load –elongation curve from finite element simulation is shown in graph 3.1.3 www.iosrjournals.org

28 | Page

Determination of Mechanical Properties of Different Materials Using Inverse Finite Element

Graph 3.1.3 Comparison of experimental results with finite element result

Graph 3.1.4 A comparison is made between experimental data and fem data. The maximum elongation of 0.23 mm at a load of 300N is observed by the finite element simulation. It may be observed that there is a fairly good agreement between the results of finite element simulation and the miniature test. 4.2 D3 STEEL: The load-elongation solution is plotted below.

FIG: 3.2 Displacement of D3 Steel miniature specimen www.iosrjournals.org

29 | Page

Determination of Mechanical Properties of Different Materials Using Inverse Finite Element The stress distribution of D3 Steel is shown in FIG.3.2.2.

FIG 3.2.1: Stress distribution of D3 Steel miniature specimen The experimental load-elongation curve is taken from the literature survey is shown in graph 3.2.2

Graph 3.2.2 The Load elongation curve from finite element simulation is shown in graph 3.2.3

Graph 3.2.3 Comparison of experimental results with finite element result

www.iosrjournals.org

30 | Page

Determination of Mechanical Properties of Different Materials Using Inverse Finite Element

Graph 3.2.4 A comparison is made between experimental data and fem data. The maximum elongation of 0.28 mm at a load of 382N is observed by the finite element simulation. It may be observed that there is a fairly good agreement between the results of finite element simulation and the miniature test. 4.3 H11 STEEL: The load-elongation solution is plotted in FIG 3.3.1.

FIG.3.3.1. Displacement of H11 Steel miniature specimen The stress distribution of H11 steel is shown in FIG 3.3.2

FIG.: 3.3.2. Stress distribution of D3 Steel miniature specimen The experimental load-elongation curve is taken from the literature survey is shown in fig: 3.3.3. www.iosrjournals.org

31 | Page

Determination of Mechanical Properties of Different Materials Using Inverse Finite Element

Fig 3.3.3 The Load elongation curve from finite element simulation is shown in graph 3.3.4

Graph 3.3.4 Comparison of experimental results with finite element result

Graph 3.3.5 A comparison is made between experimental data and fem data. The maximum elongation of 0.35 mm at a load of 380N is observed by the finite element simulation. It may be observed that there is a fairly good agreement between the results of finite element simulation and the miniature test.

www.iosrjournals.org

32 | Page

Determination of Mechanical Properties of Different Materials Using Inverse Finite Element V.

INVERSE FINITE ELEMENT SIMULATION

The finite element technique is used to simulate the structures having complex geometry or the structures having a set of complex initial conditions by supplying suitable input parameters. Then the desired output can be obtained for that model. This is the direct process. The inverse finite element process is the reverse of the direct process. Inverse problems could be described as problems where the answer is known, but not the question. In other words, the results or consequences are known, but not the cause. This unknown cause can be found out using the inverse finite element procedure. Inverse problem has attracted the attention of many researchers in diverse fields. Particularly it is very helpful in obtaining the solution of nonlinear problems where interest is focused upon the control action needed to achieve a particular form of output response. The load – elongation curves obtained from the miniature tension test on dumb-bell shaped specimens are used for the development of inverse finite element procedure to acquire the conventional tensile constitutive behaviour of the material. This is achieved by optimal matching of the experimental and the simulated load - elongation curve of miniature specimens. The proposed inverse finite element technique is used to compute the uniaxial tensile test behaviour of materials such as the Young’s modulus, yield stress and the post yield (flow) behaviour (true stress - true plastic strain curve). The inverse finite element procedure is carried out with the help of available student edition of finite element code ABAQUS/CAE 6.12-2. 5.1 INVERSE FINITE ELEMENT SIMULATION PROCEDURE The finite element simulation of miniature specimen test is capable of predicting the stress – strain behaviour and the load – elongation curve at any point of the test interval as described in the previous section. The results from the finite element simulations are closely matching with the experimental values. This is possible by supplying the required uniaxial tensile properties such as Young’s modulus, yield stress and post yield plastic flow behaviour of the material as input parameters in the simulation. We cannot find the output response of material for which tensile properties are not known because of lack of material to perform standard tests. To alleviate this situation an attempt has been made to develop an inverse finite element simulation procedure to determine the uniaxial tensile properties of the materials. Then by using these material properties the output response for the unknown material can be found out. The analytical procedures developed for the determination of constitutive behaviour of materials need interpretation of miniature test data in combination with the inverse finite element method. The procedures are essentially geared towards the prediction of conventionally measured tensile (through standard test) behaviour of the materials. The miniature test based experimental load – elongation curves are completely tried to match (up to the maximum load) with finite element simulated load - elongation curves to generate the true stressstrain curves for any unknown material. The elastic modulus is determined by matching the initial linear portion of the load – elongation curve from finite element simulation and the miniature specimen test. When both of these curves match exactly up to the yield point (start of non-linearity), the Young’s modulus of the material is obtained from inverse finite element method. Then, the post yield behaviour or flow behaviour in the form of true stress - equivalent plastic strain data are computed by the inverse finite element analysis. This way the developed inverse finite element procedure is helpful in computing the standard uniaxial tensile test behaviour of materials by using miniature specimen test data. The computed values of Young’s modulus and true stress - equivalent plastic strain data are compared with those of standard uniaxial tensile test data. In the direct finite element method, the simulation of miniature specimen test is carried out by giving the following input parameters: a) The Young’s modulus of the material. b) True stress - true plastic strain data at various locations from standard uniaxial stressstrain curve, starting from yield stress (corresponding to zero plastic strain) till the maximum stress point. The output from the finite element analysis is load – elongation curve, for the miniature specimen. In inverse finite element method the miniature test based experimental load - elongation curve is given as an input in a linear piecewise manner to the finite element software. The experimental load – elongation curve is used to compare with the curve produced from inverse finite element procedure. Figure (4.1) schematically summarizes the steps involved in the inverse finite element procedure developed for the determination of the uniaxial tensile stress - strain constitutive behaviour from miniature specimen load – elongation curve.

www.iosrjournals.org

33 | Page

Determination of Mechanical Properties of Different Materials Using Inverse Finite Element

FIG. 4.1 www.iosrjournals.org

34 | Page

Determination of Mechanical Properties of Different Materials Using Inverse Finite Element VI.

RESULTS:

6.1. Inverse finite element simulation for medium carbon steel

FIG. 5.1 6.2 INVERSE FINITE ELEMENT SIMULATION FOR PIPELINE STEEL

FIG. 5.2 6.3INVERSE FINITE ELEMENT SIMULATION FOR ALUMINIUM ALLOY

www.iosrjournals.org

35 | Page

Determination of Mechanical Properties of Different Materials Using Inverse Finite Element 6.4 RESULTS OBTAINED FROM INVERSE FINITE ELEMENT SIMULATION YOUNG’S YIELD TENSILE S.NO MATERIAL MODULUS STRENGTH STRENGTH (MPA) (MPA) (MPA) 1 MC STEEL 197 298 567 2 PIPELINE 200 490 597 STEEL 3 AL ALLOY 66 234 267

VII.

Conclusion

The finite element results are found to be in good agreement with the experimental results as reported in the literature. One can easily create the data base of load-elongation curves for different materials using finite element technique. Now for any unknown material the load-elongation curve can be obtained using the miniature specimen test. The experimental load-elongation curve can be matched with the finite element loadelongation data base and the tensile properties of an unknown material can be easily predicted. An inverse finite element simulation procedure has been applied successfully on MC steel and Aluminium alloy and Pipeline steel to find the Young’s modulus, yield strength and flow properties based on load-elongation curves obtained from the miniature test reported in the literature.

ACKNOWLEDGMENT The satisfaction that accompanies the successful completion of any task would be incomplete without introducing the people who made it possible and whose constant guidance and encouragement crowns all efforts with success. We express our sincere gratitude to the management of GMR Institute of Technology, Rajam. We are highly indebted to them.m for their guidance and encouragement to complete this work successfully.

References [1] [2]

[3] [4] [5] [6] [7] [8] [9]

G. Partheepan, D. K. Sehgal, and R. K. Pandey, “An inverse finite element algorithm to identify constitutive properties using dumbbell miniature specimen,” Modelling And Simulation in Materials Science and Engineering, vol. 14, pp 38 -50, 2006 Kullen, H. H. Smith, and D. J. Michel, “The shear punch measurement of the mechanical properties of selected unirradiated and irradiated alloys,” Journal of Nuclear Materials, vol. 58, pp. 57- 63, 1987 Jaeger, J.C.; “Moving sources of heat and the temperatures of sliding contacts”; Proceedings of Royal Society N.S.W., vol.76, pp.203-204, 1942 R. K. Pandey, S. Bhowmick, “Studies on miniature test technique for the assessment of residual life,” in Proc. of Trends in mechanical engineering Education and Research. 11th ISME conference, IIT Delhi, India, pp 542-547, 1999. M. P. Manahan, A. S. Argon, and O. K. Harling, “Miniaturized disk bend test techniques-development and application,” ASTM STP 888, pp. 17-49, 1981. M. P. Manahan, “A new post irradiation mechanical behavior test-the miniaturized disk bend test,” Nuclear Technology, vol. 63, pp. 295- 315, 1983. J. S. Cheon, and I. S. Kim, “Evaluation of thermal aging embrittlement in CF8 duplex stainless steel by small punch test,” Journal of Testing and Evaluation, vol. 24, no. 4, pp. 255–262, 1996. A. Husain, D. K. Sehgal, and R. K. Pandey, “An inverse finite element procedure for the determination of constitutive tensile behavior of materials using miniature specimen,” Computational Materials Science, vol. 31, pp. 84–92, 1994. G. Partheepan, D. K. Sehgal, and R. K. Pandey, “Design and usage of a simple miniature specimen test setup for the evaluation of mechanical properties,” International Journal of Microstructure and Material Properties, vol. 1, no. 1, pp 38-50, 2005. AnandKumar.M,(2000) determined the mechanical properties of different materials using miniature specimen technique and FEM, M.Tech. IIT, Delhi.

www.iosrjournals.org

36 | Page

IOSR Journal of Mechanical and Civil Engineering (IOSR-JMCE) e-ISSN: 2278-1684,p-ISSN: 2320-334X, Volume 6, Issue 3 (May. - Jun. 2013), PP 37-41 www.iosrjournals.org

Modelling of Permanent Deformation on Flexible Pavement Using Accelerated Pavement Testing Khan Shahbaz1, Tiwari, Devesh2, Nagabhushana M.N2, Jain P.K3 1 .Academy of Scientific and Innovative Research, CSIR-CRRI, India., 2 .Principal Scientist, Pavement Evaluation Division, CSIR- Central Road Research Institute, India. 2 .Principal Scientist and Nodal Officer APTF Program ,Flexible Pavement Division, CSIR-Central Road Research Institute, India 3.Chief Scientist and Head, Flexible Pavement Division, CSIR- Central Road Research Institute, India.

Abstract: The two major distresses encountered on flexible pavement under Indian conditions are fracture (cracking) and longitudinal permanent deformation (rutting) which affects the serviceability of pavement. Accelerated Pavement Testing Facility (APTF) is a tool which is a vital link for testing and measuring full-scale field behaviour of cracking and rutting of pavement at in-situ conditions. Recently, CSIR-Central Road Research Institute (CRRI) in India has procured a linear Heavy Vehicle Simulator (HVS) type of APTF which is presently being used for finding out the cracking and rutting behaviour of a flexible pavement consisting Dense Bituminous Concrete (DBC) as wearing course and Dense Bituminous Macadam (DBM) as binder course apart from the conventional granular layers above sub-grade. The present paper deals with the development of a statistical model and its approach for pavement rutting under numerous passes (bi-directional) for the layer specifications which are (i) 40 mm DBC (ii) 120 mm DBM (iii) 250 mm Wet Mix Macadam (WMM) and (iv) 300 mm Granular Sub-base (GSB) above the Subgrade, which is an Indian Specification widely used for 30 Million Standard Axles (MSA) at 5% CBR. The statistical model has been developed by observing / recording pavement surface profile using Laser Profilometer (off board) for every 5,000 passes upto 50,000 passes, thereafter at every 10,000 passes upto 175,000 passes and then at every 25,000 passes upto 275,000 passes. The details of methodology adopted, load applied, temperature and material properties have also been given in the paper. Keywords: Flexible pavement, Rutting, Accelerated Pavement Testing Facility,Modelling,Profilometer. I.

Introduction:

In Indian conditions, pavements are subjected to extremely varying seasons from hot to cold, therefore, possibilities of rutting and fatigue failures both can take place in flexible pavement. A rut is characterised by longitudinal surface depression within the wheel path and may have associated transverse displacement, thereby reducing serviceability and safety of a flexible pavement. Rutting can be the result of permanent reduction in volume (consolidation/traffic densification), permanent movement of the material at constant volume (plastic deformation/shear), or a combination of the both the two [1]. Bituminous concrete (BC) specification is a time, temperature and stress dependent material, which, when subjected to repeated loading exhibits visco-elastic or plastic responses. The visco-elastic properties are modelled by modulus of elasticity and Poisson‟s ratio since they do not contribute to permanent deformation. Plastic properties contribute to permanent deformation, which is cumulative under repeated loading. There are several factors that influence rutting. Variables such as vehicle speed, time and contact pressure are directly represented in the creep rate model, while temperature, asphalt/bitumen mixture characteristics and construction quality are represented as values of the constants. Shear resistance properties of materials, especially bituminous, need to be properly addressed for limiting the rutting. While pavement structural design practices mainly focuses on protecting the subgrade by excessive vertical strain, whereas, shear resistance of the bituminous layers has been left to mixture designers. Adequacy of shear resistance in bituminous mixture is commonly assessed with empirical methods, such as Marshall Stability. There is a conscious effort to shift towards the use of fundamental performance-related test methods. Two procedures have been used to limit rutting; one to limit the vertical compressive strain on top of the subgrade, the other to limit the total accumulated permanent deformation on the pavement surface which is based on the properties of each individual layer. In the process, the number of load repetitions (Nr ) to limit rutting is related to the vertical compressive strain (Ɛz ) on top of the subgrade (equation 1). Nr = f1 ×

1 f2

Ɛz

.............................(1)

Several agencies have defined different values for constants f1 and f2. As per the earlier research in India, the rutting model is defined from large number of data obtained from pavements subjected to rutting failure. Setting the allowable rut depth as 20 mm [2], rutting equation has been presented as equation 2. www.iosrjournals.org

37 | Page

Modelling of Permanent Deformation on Flexible Pavement Using Accelerated Pavement Testing Nr = 4.1656 × 10−8 ×

1 4.5337 Ɛz

.......................(2)

Nr = No of Load Repetition in Rutting Failure Ɛz= the compressive strain on the top of subgrade. To study the effect of rutting failure on flexible pavement, Accelerated Pavement Testing Facility (APTF) is the most appropriate tool and also herein used.

II.

Accelerated Pavement Testing for Rutting Evaluation

Accelerated Pavement Testing (APT) constitutes a vital link between the Pavement Engineering (P-E) tools for laboratory evaluation of pavement materials and to understand the full-scale in-situ field behaviour of these materials under a dynamic load within a pavement structure, as shown in figure 1. It supports most of the current pavement design methods and forms the basis for developing various theories about pavement behaviour and For this reason, APT has been regarded as a tool which provides pavement engineers the knowledge of understanding the behaviour of pavement materials and structures in a better way under varying traffic loading and known environmental conditions. APT aims to evaluate pavement sections under a range of loading and environmental conditions to improve the knowledge on the performance of the pavement layers and structure under a full range of field operational conditions. Development of rutting is caused by a combination of densification and shear-related deformation with an increasing number of load applications which may occur in any layer of a pavement structure [3]. Hence, the pavement structural designs, bituminous mix designs, and pavement construction aims at minimising permanent deformation. As a result, pavement structures are composed of number of layers possessing required performance properties. These layers attenuate load-induced stresses and limit subgrade stress and deflection. It must, therefore, be ensured that the flexural and shear resistance of the structural layers are good enough to sustain these high stress states. 1.1 APT Facility In the present case, evaluation of a flexible pavement has been done using linear Heavy Vehicle Simulator (HVS) type of APTF. The facility is in operation in Central Road Research Institute (CRRI) premises which is one of the premier laboratories of Council of Scientific and Industrial Research (CSIR) in India, as shown in Photo 1 and the salient features of HVS Table 1. Features of HVS Mark IV+ APT are presented in Table 1. Feature

Capability

Mobility

Towable over long distances; self-propelled over short distances

Load Application

Static & Dynamic Load Applications

Wheel Loads

Mean Load : 36 - 88 kN ; Peak Load : 30 - 105.6 kN

Test Wheel (s)

Single, Super-single, Dual (truck) or Aircraft (Airbus 4616)

Tyre Pressure

Typically 560-690 kPa on roads ; Upto 1450 kPa on airfields

Repetitions Per Day

Upto 32000 (Bi-directionally) ; Upto 16000 (Uni-directionally)

Trafficked Length

8m (test area length 6m)

Trafficked Width

Upto 1.5 m, depending on wheel configuration

Trafficking Pattern

Variable

Dimensions

Length = 22.8 m ; Width = 3.5 m ; Height = 4.0 m

Total Mass

47000 Kg

Power Source

Diesel (self contained) or electric shore power

Fig.1 Interrelationship Between P-E Tools [4]

Fig 2: HVS(APTF) Operating at CSIR-CRRI

www.iosrjournals.org

38 | Page

Modelling of Permanent Deformation on Flexible Pavement Using Accelerated Pavement Testing 1.2 Flexible Pavement Test Strip: A flexible pavement test section was designed (2) for a design traffic of 30 million standard axles (msa), with a subgrade CBR of 5%, layer of granular sub-base (GSB) materials of thickness 300 mm, wet mix macadam (WMM) as base layer of thickness 250 mm and Dense Bitumen Macadam (DBM) and Dense Bituminous Concrete (DBC) of thicknesses 120 mm and 40 mm respectively. The above mentioned layer construction of the pavement structure was accomplished incorporating the specifications of „Ministry of Road Transport &Highways‟ (4th Revision). The material properties are given in tables 2, 3 and 4. Table 2: Aggregate Gradations for DBC & DBM Mixture Sieve Size (mm)

DBC Cumulative % by weight of total aggregate passing, 100Values Specified 100 79 – 100 59 – 79 52 – 72 35 – 55 28 – 44 20 – 34 15 – 27 10 – 20 5 – 13 2–8

75 26.5 19 13.2 9.5 4.75 2.36 1.18 0.6 0.3 0.15 0.075

Gradation obtained 100.00 100 90.51 60.50 52.45 37.33 31.50 20.15 17.44 11.74 7.08 3.16

DBM Cumulative % by weight of total aggregate passing

Gradation obtained

100 90-100 71-95 56-80 38-54 28-42 7-21 2-8

100.00 100.00 96.08 67.92 49.26 29.96 9.14 2.06

Table 3 : Volumetric and Mechanical properties of DBC mix and DBM Binder type in mix

Optimum Binder Content (%)

Bulk Density(g/cc)

Marshall Stability (kN)

DBC DBM

5.1 5.0

2.280 2.301

10.22 10.15

Retained Marshall Stability % 83 79

Flow (mm)

Air voids (%)

VFB (%)

3.1 3.5

4.2 4.9

71.7 69.3

Table 4: Aggregate Gradations for WMM & GSB Mix Sieve Size (mm)

75.0 53.0 45.0 26.5 22.4 11.2 9.50 4.75 2.36 0.600 0.425 0.075 Max. Dry Density(MDD), g/cc Field Dry Density(FDD), g/cc CBR at FDD

Wet Mix Macadam(WMM) Cumulative % Gradation by weight of obtained total aggregate passing, Specified Values -100 100 95 - 100 100 -60-80 66 40-60 48 -25-40 29 15-30 19 8-22 20 -0-8 7

Granular Sub-base(GSB) Cumulative % Gradation by weight of total obtained aggregate passing 100 80-100 -55-90 -35-65 25-55 20-40 -10-25 3-10

2.26

78 58 49 37 21 8 2.19

2.23 71

100 91

2.16 58

www.iosrjournals.org

39 | Page

Modelling of Permanent Deformation on Flexible Pavement Using Accelerated Pavement Testing 1.3 Evaluation of Test Section using APT and Data Acquisition: A half axle load of 80 kN on dual wheels was applied on the test section having tyre pressure of 700 kPa. Single wheel path, bi-directional traffic was used for load application. The load was applied for 2,75,000 repetition for every 5,000 passes upto 50,000 passes, thereafter at every 10,000 passes upto 175,000 passes and then at every 25,000 passes upto 275,000 passes, adding upto 4.4 msa using AASHTO power law. The average speed was 10 kmph and test period was from Oct. 2011-Dec. 2012 where APTF was operated only during day time. Since, temperature and rutting magnitude along with mix properties are the most significant primary parameters for the study of asphalt failure in the form of rutting, therefore, temperature sensors were embedded at bituminous layers which were used for the measurement of pavement temperature. The details of bituminous mix properties are already detailed before in the tables, but environmental / temperature control measures were not able to be adopted in order to get data nearer to the true life situation. The continuous changes in profile resulting in rutting were also measured and recorded by portable laser profilometer. Prior to trafficking, cross-sections were marked out at 0.5-m intervals for 8 x 1 m (length x breadth) test section of the road.

III.

Data Analysis:

The temperature of pavement was taken at 50mm below the bituminous surface using K-type thermocouple: Table 5: MONTH Average Monthly Pavement Temperature, 0C OCTOBER 26.185 NOVEMBER 20.7375 DECEMBER 18.1775 As there was little variation in temperature in the evaluation period, so this was taken as constant parameter. 14

Rut progression

12 10 8 Rut Depth in mm

6 4 2 0 -2

0

0.5

1 1.5 2 Transverse width in m Fig3:Rut progression with number of passes

2.5

3

The rut progression was obtained using moving average of order nine for a smooth curve Correlation were attempted between dependent variable as rut depth and number of repetitions as number of standard repetitions keeping all other variables as constants (figure 3).

www.iosrjournals.org

40 | Page

Modelling of Permanent Deformation on Flexible Pavement Using Accelerated Pavement Testing RUT DEPTH VS NUMBER OF STANDARD LOAD PASSES PASSES IN TEN THOUSAND 0

5

10

15

20

25

30

RUT DEPTH (mm)

0 1 2 3 4 5 Fig 4: Best Fit Line of the Scatter Points IV. Conclusions: 1) Best fit line is obtained from trend line of second degree polynomial and was found to be quite realistic. 2) The best fit line equation for the scatter points of rut depth with number of passes was found out to be y = 0.005x2 + 0.2368x + 1.1988. 3) The value of T-test was found to be significant i.e. 7.3813 and also the coefficient of regression was also found out to be 95%. 4) The rut depth of about 1 mm is being visible even at the start of load repetitions is due to camber profile, since the measurement were done using laser sensors. 5) The best fit line is almost parallel to x-axis after about 200000 repetitions (3.2 msa)indicating a saturation level after 4 mm of rut depth. 6) The phenomenon of consolidation / settlement (rutting) in bituminous layers after few repetitions of standard load is quite evident in Figure 2. 7) After 4 mm of rut depth which is quite minimal, the effect of further number of load repetitions is not visible indicating the phenomenon of initial settlement level of DBC layers.

Acknowledgements: The authors are highly thankful to Dr.S.Gangopadhyay, Director, CSIR-CRRI for the encouragement, support and guidance in CSIRCRRI‟s APTF Program. The inputs of Mr. Abhishek Mittal, Senior Scientist, Flexible Pavement Division, CSIR-CRRI are duly acknowledged. Also, the authors express their thanks for all the member of CSIR-Central Road Research Institute who has helped in this project.

References: [1]. Dr. Roque.,Reynaldo, Dr.Birgisson Bjorn, Mr. Darku Daniel, Mr. Christos A.Drakos Evaluation of laboratory testing systems for asphalt mixture design and evaluation, Florida department of Transportation, www.dot.state.fl.us/researchcenter/Completed.../FDOT_BB888.pdf. ,2004 [2]. Guidelines for the Design of Flexible Pavement,Indian Road Congress -37:2001. [3]. Wang, H., Zhang, Q., and Tan, J. ”Investigation of Layer Contributions to Asphalt Pavement Rutting.” J. Mater. Civ. Eng. 21, SPECIAL ISSUE: China: Innovative Use of Materials for Highway Construction,2009 [4]. Hugo, F., B.F. McCullough, and B. Van der Walt, “Full- Scale Accelerated Pavement Testing for the Texas State Department of Highways and Public Transportation,” Transportation Research Record 1293, Transportation Research Board, National Research Council, Washington, D.C., 1991.

www.iosrjournals.org

41 | Page

IOSR Journal of Mechanical and Civil Engineering (IOSR-JMCE) e-ISSN: 2278-1684,p-ISSN: 2320-334X, Volume 6, Issue 3 (May. - Jun. 2013), PP 42-47 www.iosrjournals.org

Seismic Analysis for Safety of Dams 1

N.P. Gahlot, 2Dr. A.R. Gajbhiye

1

2

Sectional Engineer, Water Resources Department,Amaravati.. Professor & Head of Civil Engineering Department, Yeshvantrao Chavan College of Engineering, Nagpur.

Abstract: Geo-technical engineering as a subject has developed considerably in the past four decades. There has been remarkable development in the fields of design, research and construction of dam. India is capable of designing and constructing a dam that would withstand a seismic jolt. The country needs water and electricity to provide its people good living standards. Hydropower is the solution to the country's requirements, and this can be achieved by storing water in dams. In the past, earthquake effects may have been treated too lightly in dam design. Are such dams safe, and how have they fared in previous earthquakes, this Paper will be limited to the some of finding about one concrete types. What will happen to dams during severe earthquake shaking? It is obvious that at present engineers cannot answer this question with any certainty. But we are very much aware of the threat of disastrous losses of life and damage to property if dams should fail, and we are making great effort to increase our under standing of this complex topic. This Paper deals with the case study of totaladoh Dam Situated in Vidarbha Region of Maharashtra for Seismic Analysis by I.S.Code method (Simple Beam Analysis method). This also includes future scope of analyzing the same dam for Seismic safety by very accurate method i.e. finite element method.

Keywords: Earthquake, The finite element method, Indian Standard codes(I.S.Code), horizontal seismic coefficient (αh ),Hydrostatic pressure, Seismic analysis, I.

Introduction

As of now, there are about 23,000 large dams in the world. A large dam is defined by the International Congress on Large Dams (ICOLD) as one which has a height of 33 metres and above: Of these dams, only four, namely Koyna (India), Kremesta (Greece), Kar iba (South Africa) and Singfenkiang (China), have experienced earthquakes of a moderate magnitude of between 6.0 and 6.5 on the Richter scale within a few years of building. Although no earthquake-related failure of a concrete dam has occurred to date, no large concrete dam with a full reservoir has ever been subjected to really severe ground shaking. Such a possibility has many groups concerned, including the Division of Safety of Dams(DSD), a California state agency responsible for assuring the safety of California dams. The Division of Safety of Dams (DSD) has the power to order an updated seismic check of a dam if new information rises or if better analysis techniques are developed by researchers. In the early" 1970s, two events led the Division of Safety of Dams (DSD) to initiate a program to perform seismic checks on all major dams under its jurisdiction. The first event was the near collapse of Lower San Fernando Dam, a large earthen dam, during the 1971 earthquake; and the second was the development of the finite element method, a tool for computerized stress analysis. In this paper the methods of seismic Analysis of dam are discussed. A case study of Totaladoh Dam which is analyzed by simplified beam method. The Results obtained by this method will be compared by the results obtained by finite element method is the future scope of study.

II.

Methods for seismic analysis of Dam

Concrete gravity dam design was, and still is, based on two-dimensional idealizations (as illustrated in the figure No.1) because gravity dams, which are generally located in wide river valleys, are long and nearly uniform in cross section. Water loading from the reservoir behind the dam seeks to overturn or slide the dam downstream; and the dam's own weight resists this action. A proper choice of dam cross section provides stability. In addition, since concrete is weak in tension and since no steel reinforcing is employed, engineers equated. The presence of tensile stresses with failure.If their computations showed tensile stress at any point, they redesigned the cross section. Stress analysis was performed by treating the dam cross section as a beam of variable thickness cantilevering from the valley floor.

www.iosrjournals.org

42 | Page

Seismic Analysis For Safety Of Dams

α Pd αW

Ps

W a=αg Fig. No. 1 :Design of a concrete gravity dam (above) is based on a two-dimensional idealization. The design loads include the dam weight W, the static water force P, and the ground acceleration α (given as a fraction α of the gravitational acceleration g). The ground acceleration produces an inertia force α Won the dam and an additional water force α Pd (where Pd is the water force caused by a unit acceleration of the water into the reservoir).

Arch dams are built in narrow canyons, and they are true three-dimensional structures. They resist the water load by combining cantilever bending from the canyon floor with arch thrusting to the abutments. Usually their proportions are much thinner than those of concrete gravity dams. Tensile stress was again avoided in the design, but engineers found stress analysis much more complicated. An iterative relaxation method applied to independent arch and cantilever sections was developed, which produced many rooms full of engineers grinding out stress calculations. Which appeared to be a stiff structure, would move rigidly with the ground. Thus, if the ground accelerates at a fraction, designated α (alpha), of gravity, an inertial force of magnitude α times the dam weight is created and acts on the dam in the downstream direction. Moreover, additional water pressure is generated, proportional to the acceleration of the dam into the reservoir if water incompressibility is assumed. This feature was recognized in 1933, and it has been included in dam design ever since. Typical values of α were 0.05 to 0.15, and inclusion of earthquake effects stilI allowed the no-tension criterion to be observed in the design. Early design procedures were obviously great simplifications of reality. Dams are not really rigid; they are flexible structures that vibrate on their own when excited by ground motion. Stress analysis methods were approximate, and the maximum ground accelerations used were only fractions of what could occur. Vertical and cross-stream components of ground motion were neglected. But the pertinent question is, of course: How have dams designed by the methods described performed during past earthquakes? And the answer is: Fairly well,

III.

Seismic analysis of Dam by Finite element Method.

The finite element method transforms the governing differential equations (the equations of solid mechanics in the case of a dam) to a matrix equation that is solved on the computer. The structure to be analyzed is meshed into elements (see figure No.2), which are connected at nodal points. Associated with these nodes are displacement degrees of freedom, which become the unknowns of the matrix equation. Solution of the matrix equation yields the structure displacements, from which the stresses are easily computed. As long as the governing differential equations are linear, the finite element method produces remarkable solutions. Nonlinearities, however, are much more difficult to handle. An example of nonlinearity in dam behavior is the formation of cracks or opening of built-in joints due to the presence of tensile stresses. Even today, finite element techniques have not progressed to the point where this type of non linearity can be handled. A standard procedure using the finite element method was developed for computing the (linear) response of concrete dams. The sketch (Fig. No.2) illustrates this procedure. A finite element model is constructed of the dam and of a portion of the foundation region that extends out to an artificial boundary where earthquake motions are applied. The motion specified by the engineer is actually the free-field motion (that is, the motion that would occur at the dam-foundation interface if the dam were not present), and the engineer must www.iosrjournals.org

43 | Page

Seismic Analysis For Safety Of Dams back-calculate what motion to apply at the foundation boundary. Since the foundation boundary produces wave reflections that contaminate the computed dam response after a short time, the foundation mesh is usually assumed to be mass less. The alternative is to place the foundation boundary far away from the dam, but this results in a large, expensive to solve matrix equation. The water is included in the analysis by an added-mass approach. An appropriate volume of water is assigned to move with each horizontal, nodal degree of freedom at the upstream dam face. Treating the water in this manner neglects water compressibility (which can be important for deep reservoirs) and ignores the additional pressures generated by the vertical and cross-stream components of earthquake ground motion along the reservoir boundary. Obviously, the most pressing research need is for computational techniques that accurately model the cracking behavior, but little progress has been made to date. Recently, however, some headway has been reported on improved modeling of the dam foundation and of the water in the reservoir. The artificial foundation boundary can be replaced by mathematical transmitting boundaries, which reflect only a small fraction of an incident wave. For the water,We have developed finite element models (Fig. No.3) that includes water compressibility and the additional pressures generated. by vertical and cross-stream motions of the reservoir boundaries. Both of these effects have been shown to influence the earthquake response of concrete dams significantly.

Fig. No. 2 : Shows an illustration of how improved earthquake safety evaluations of dams have been made possible through use of finite use of finite element analysis. A common procedure employs a dam mesh, a mesh of finite region of foundation, an added masses on the upstream dam face to represent the water.

Fig. No. 3 : Finite element model of the water in the reservoir accurately represents its effect on the dam response to earthquake shaking, these model includes water compressibility and additional pressure generated by vertical and cross stream motions of reservoir boundaries

IV. Objectives of Study : Objectives of study is to estimate the stresses induced due to earthquake shock on the dam by I.S.Code method and compare these with the results which will be obtained for the same dam by the seismic analysis carried out by Finite element method.

V. Case Study - Analysis of Totaladoh Dam by I.S.Code method The Totaladoh dam is a masonry type Gravity Dam constructed in the year 1982-83 on the Pench river near village totaladoh, Nagpur district in Maharashtra(India).It is an interstate Hydro cum Irrigation project of Maharashtra and Madhyapradesh. The salient feature of Dam is as under. Controlling Levels : Length of Earthen dam – 2381 m. (I) Crest level of Spillway – 482.000 m. Length of Masonry dam – 680 m. (II) M.D.D.L. - 464.000 m. Total Height of Dam – 51.50 m. www.iosrjournals.org

44 | Page

Seismic Analysis For Safety Of Dams (III) (IV) (V) (VI) (VII)

F.R.L. T.B.L. M.W.L. Gross storage Live storage

- 490.000 m. - 495.500 m. - 493.000 m. - 1241 Mm3 - 1091 Mm3

Width at Top Width at Base Upstream slope Downstream slope Density of Masonry

- 6.70 m. - 43.90 m. - 0.10 : 1.00 - 0.83 : 1.00 - 23.3 KN/m3

Seismic Analysis Of Totaladoh Dam by I.S.Code method : For the study dam section is discredited into 13 segments. Wight of each segment is assumed concentrated at the centre of segment. considering the dam to be mathematically modeled as a cantilever fixes at base and free at top and assumed the dead weight of segments acting horizontally through the centre of respective segments.

A) Calculations of Stresses due to horizontal component of earthquake. As per I.S. 1893:1984 & 2002(Part I) Provision As the height of dam is less than 100 m. the analysis is to be done by the seismic coefficient method. The horizontal seismic coefficient αh can be calculated as Horizontal seismic coefficient αh = β I αo Where β – coefficient depending upon soil foundation system, for Dams β = 1.00 I – importance factor for dams I = 3.00 αo - as Totaladoh project lies in zone No. II ( latest map in fifth revision of I.S. code) hence αo=0.02 ∴ αh = 1.00 x 3.00 x 0.02 = 0.06 Weight of dam per metre length = 25374.87 KN. As per I.S.Code 1893 P.no.42 At the top of dam horizontal seismic coefficient shall be taken as 0.75 αh = 1.5x0.06 = 0.09 And reduces linearly zero at base. ∴ ℏ = {[6.7x51.5x(51.5/2)]+[1/2x4x51.5x(40/3)]+[1/2x33.2x40x(1/3x40)]} {(7.6x51.5) + (1/2 x4x40) + (1/2x33.20x40)} ∴ ℏ = 15.70 m. At section BB Base shear = VBB = 0.60W αh = 0.6 x 25374.87x0.06 = 913.50 KN. Base Moment = MBB = 0.9W ℏ αh = 0.9x 25374.87x15.70x0.06 = 21521 KNm. Stress at Base = ρBB =(±6 M/T2) = ± 6 x 21521.81/(43.90)2 = ±67.00 KN/m2 At section AA y/h = 11.5/51.5 = 0.223, C’V =0.20, C’m =0.09 Shear at AA= VAA = C’V *VBB = 0.20x 913.50 = 182.70 KN. Moment at AA= MAA = C’m MBB = 0.09 x 21521.81 = 1936.15 KNm. Stress at Base = ρAA =(±6 M/T2) = ± 6 x 1936.15/(6.70)2 = ±258.79 KN/m2 www.iosrjournals.org

45 | Page

Seismic Analysis For Safety Of Dams B) Calculations of Stresses due to vertical component of earthquake. Load point

Weight of segment

At section AA 1 546.39 2 624.44 3 624.44 Total 1795.27 Section at Base -BB 4 797.79 5 1144.50 6 7 8 9 10 11 12 13 Total

1491.20 1837.90 2184.61 2531.31 2878.02 3224.72 3571.42 3918.13 23579

αv= 0.75* αh

W* αv

Lever arm

Moment

Stress Calculations

0.045 0.041 0.038

24.59 26.19 24.01 74.79

3.35 3.35 3.35

82.37 87.74 80.43 250.53

X1 = 250.53/74.79 =3.35 m. ρmax = W* αv/B(1± 6e/B) = 74.19/6.70 (1+0) ρmax = ± 11.16 KN/m2

0.035 0.031

27.88 36.00

5.21 7.07

145.28 254.53

ρmin = 74.19/6.70 (1+0) ρmin = ∓ 11.16 KN/m2

0.028 0.024 0.021 0.017 0.014 0.010 0.007 0.003

41.70 44.97 45.81 44.24 40.24 33.81 24.97 13.69 353.31 428.09

8.93 10.79 12.65 14.51 16.37 18.23 20.09 21.95 20.16

372.34 485.19 579.54 641.87 658.67 616.40 501.65 300.59 4555.96

X2 =4555.96/428.09 =12.89 m. X = (74.79*3.55+353.31*12.89)/ (74.79+353.31) = 11.25 m. e = b/2 – X2 = 43.90/2-11.25 ∴ e = 10.70 m. ρmax = W* αv/B(1± 6e/B) =428.09/43.90(1+6*107/43.9) ρmax = ± 24.01 KN/m2 ρmin = W* αv/B(1± 6e/B) =428.09/43.90(1-6*107/43.9) ρmin = ∓ 4.50 KN/ m2

C) Calculations of stresses due to Hydro dynamic pressure of horizontal Earthquake acceleration y h At section AA 8.50 48.50

y/h

Cm

Cs

ρe

Vy

My

0.175

0.70

0.26

11.30

69.76

244.11

0.70

0.70

30.55

1075. 70

21486.51

Section at Base -BB 48.50 48.50 1.00

Stress Cs= Cm/2{y/h(2-y/h) +√ y/h(2y/h)} αh= 0.06 x1.50 = 0.09 stress are given by. ρAA = W/T ± 6M/T2 where W = 0 = 0 ± 6 x 244.11/(6.70) 2 ρAA = ± 32.62 KN/m2 ρBB = W/T ± 6M/T2 where W = 0 =0 ± 6 x 21486.50/(43.90) 2 ρBB = ± 66.89 KN/m2

D) Calculations of stresses due to Hydro dynamic pressure of Vertical Earthquake acceleration As per I.S. Code 1893-1984 (Clause 7.3.1.2 P.no.44) For Seismic coefficient method At top of dam it would be 0.75 times the value of αh and reduces linearly zero at base. Therefore in this case, as we are superimposing the stresses due to various conditions separately. The stresses due to vertical component are obtained by multiplying the stresses due to hydrostatic pressure by a factor ± αv i.e. ± 0.03 At section AA Stresses due to hydrostatic pressure = ± 27.84 KN/m2 (Calculated as in Static analysis) ∴ Stresses due to vertical component of earthquake= ± 27.84 x 0.03 = ± 0.83 KN/m2 At section BB(Base) Stresses due to hydrostatic pressure = ± 22.35 KN/m2 (Calculated as in Static analysis) ∴ Stresses due to vertical component of earthquake= ± 22.35 x 0.03 = ± 0.67 KN/m2

VI. Results : The results obtained of the seismic analysis of existing Totaladoh dam by I.S.Code method are expressed in the table below. Results showing Stresses at section AA & BB by I.S. Code method ( KN/m2) Loading Condition Dynamic condition

Section AA Upstream

Downstrean

Section BB Upstream

Downstrean

a) Horizontal Inertia b) Vertical Inertia Hydrodynamic Pressure

±258.79 ± 11.16

∓ 258.79 ∓ 11.16

±67.00 ±24.01

∓ 67.00 ∓ 4.50

www.iosrjournals.org

46 | Page

Seismic Analysis For Safety Of Dams c) Horizontal component d) Vertical component TOTAL

± 32.62 ± 0.83 ± 303.40

VII.

∓ 32.62 ∓ 0.83 ∓ 303.40

± 66.89 ± 0.67 ± 155.57

∓ 66.89 ∓ 0.67 ∓ 136.06

Conclusions :

As the design of this dam was done at 30 years ago that time the earthquake that may have been treated too lightly and used a very bold method for analysis.The finite element method is too complicated for analysing such structure due to contains of large calculation, but now a day such analysis is possible due to availability of various computer programs. The major work is the seismic analysis of dam by finite element method.

References: [1]. [2]. [3]. [4]. [5]. [6]. [7]. [8].

H.D. Sharma (1981) “ concrete Dams – Criteria for Design.” University of Roorkee. Metropolitian publication New delhi. Dr. D.S Saini (1983), “Dams and Earthquake”. Department of civil engineering, University of Roorkee. Bureau of Indian Standards Publication IS 1893-1984. Criteria for Earthquake Resistant design of structures. Engineering Seismology, Institution of Engineers ( India ),Kolkata publication journal, Vol. 83, Aug 2002. Dams and earthquake Safety. ( By John.F. Hall ) Dissertation Report on Seismic study of gravity dams for earthquake hazard mitigation and rehabilitation there of ( N.P.Gahlot ) Investigation of Liquefaction Failure in Earthen Dam during Bhuj Earthquake ( www.nicee.org ) Earthquake Tips ( C.V.R. MURTHY, Dept. of Civil Engg. IIT Kanpur )

www.iosrjournals.org

47 | Page

IOSR Journal of Mechanical and Civil Engineering (IOSR-JMCE) e-ISSN: 2278-1684,p-ISSN: 2320-334X, Volume 6, Issue 3 (May. - Jun. 2013), PP 48-55 www.iosrjournals.org

Analysis and study of vibrations in view of combustion gas forces for diesel engine Sushant S.Bhansali1, Dr. U.V.Kongre2, N.D.Shirgire3, P.R.Bodade4 1,2,3,4

(Department of Mechanical Engg./ J.D.I.E.T., Yavatmal (M.S.)/ Amravati University , India)

Abstract : The sound and vibratory behavior of the internal combustion engine is a highly complex one, consisting of many components that are subject to loads that vary greatly in magnitude and which operate at a wide range of speeds. Some origins of mechanically induced noise caused by various forces resulting from the combination of combustion and inertia forces which act on the moving parts of the engine to accelerate them across their running clearances and thus cause mechanical noise. The most of the investigations are done on the mechanical noise due to motion of mechanical system i.e. piston slap. The main purpose of this work is to analyze the vibration in diesel engine VCR (Variable Compression Ratio) cylinder liner considering combustion gas forces and cylinder liner temperature using finite element software ANSYS. Also Aluminium (HF 18) material is being tested in the software for this purpose. The output results were quite satisfactory to predict the behavior of deflection under different pressures. The combustion gas forces calculated for varying compression pressures. In this paper the results are presented for displacement and frequency which indicate the amplitude of vibration. By comparing the analytical results, the validity of the proposed analysis has been confirmed. Furthermore, this analysis is applied to evaluate the vibration of HF 18 material along with increase in thickness, and revealing the closer response according to the material and vibration. Keywords - VCR Diesel engine, Cylinder liner, Combustion gas force and Harmonic analysis. 1. Introduction Recently, a demand to the silence in the automobiles is increasing and the strengthening of the traffic noise regulations are carrying out. Thus the reduction of engine noise as the dominant source of vehicle noise is imperative. Some origins of mechanically induced noise caused by various forces (resulting from combination of combustion and inertia forces) which act on the moving parts of the engine to accelerate them across their running clearances and thus cause mechanical impacts. These impacts are found to be the main cause of predominant high frequency noise. Engine noise is divided roughly into two types. One is combustion noise, which originates from the combustion gas pressure in the cylinder, and other is mechanical noise, which arises due to the motion of the mechanical system i.e. piston slap impacts and other mechanical impacts in gears and bearings. Especially, the bearing noise imparted by impact and vibration caused by fluctuating forces due to various mechanisms during the engine cycle. Piston slap is a collision phenomenon between the piston and the cylinder liner when the lateral force acting on the reciprocating piston. As a result piston slap was identified and the instants of slap were determined. These were compared with the instants of slap calculated from a theoretical analysis of the dynamics of the moving parts of the engine [1]. Another investigation presented the various methods for estimating the pistons slap. This phenomenon of piston slap has been investigated by oscillographic and simulation technique to determine its relative magnitude, compared with the other noise sources of the engine. It has been found that this source of noise is important and its significance will become greater as other sources, such as combustion are reduced [2]. The finite element method (FEM) is employed to simulate the vibration response of the hull due to the excitations of diesel piston-slap and vertical inertia force of reciprocating masses. The numerical results show that, piston-slap imposed rolling moment on the diesel frame may cause a higher level vibration [3, 4, 5]. In an analytical model, predict the impact forces and vibratory response of engine block surface induced by the piston slap of an internal combustion engine. When slap occurs, the impact point between piston skirt and cylinder inner wall is modeled on a two-degree-of-freedom vibratory system. The equivalent parameters such as mass, spring constant and damping constant of piston and cylinder inner wall are estimated by using measured (driving) point mobility [6]. In this paper, modeling was done at the cross section of engine cylinder liner including 2-dimensional and 3-dimensional analysis. Beam and 3dimensional liner are consisting of thickness of cylinder and stroke length of cylinder. The results are considered for calculating the combustion forces for the certain ranges of compression pressure (between 45-75bar). Further, the effect of cylinder liner temperature was also analyzed.

2. Generation of engine noise vibration related to piston slap Looking towards the threat of vibration which is hazardous which may lead the total failure of the systems need to addressed. From few research efforts continuously focus their attention towards vibration or www.iosrjournals.org

48 | Page

Analysis and study of vibrations in view of combustion gas forces for diesel engine deflection analysis from different means which further causes the depth in analysis of vibration considering combustion gas force. In this work comparatively higher pressure zones were analyzed to understand the vibration behaviors of the engine. It imparts different material analysis with or without additional thickness to view the vibration with different aspects. Most of the researchers have found for the mechanical noise i.e. piston slap. The noise due to combustion gas force is the major source for the occurrence of noise in the engine cylinder. Very few efforts are devoted for combustion gas force analysis. Little attention was paid on noise and vibration due to combustion gas force. The work focuses on vibration issue and the main causes of mechanical noise are the “piston slap” phenomenon [3-5, 7]. The Problem is formulated as under study of the cross section of engine cylinder liner as 2-dimensional and 3-dimensional model with the help of ANSYS Version 10. Beam and 3-dimensional liner are consisting of thickness of cylinder and stroke length of cylinder. In this analysis, application of combustion gas forces and cylinder liner temperature on the piston cylinder components to investigate the vibration. The problem is described in a two dimensional beam and three dimensional model along with model meshing as shown in figure 2.1 and figure 2.2. Calculating combustion gas forces for the certain peal ranges of compression pressure (between 45-75 bar), considering cylinder liner temperature [8].

Fig. 2.1: 3-Dimensional Model Meshing      

Fig. 2.2: Combustion Gas Forces in 3-Dimensional Model after Meshing In order to simulate the model, prerequisites are as follows, Geometry: - Cross section of engine cylinder as a 2-D and 3-D model with the help of ANSYS. Boundary Conditions: - By fixing left side of model as all DOF zero and fixing the right side of model as UX as zero. Mesh: - Mapped (3-4 sided). Force Analysis: - Combustion gas forces applied at TDC position only. Parameters: - Analysis was done from nodal solution and graph. Analysis Method:- Harmonic Analysis

For the geometry analysis the work includes single cylinder, four stroke, and VCR (Variable Compression Ratio) Diesel engine. The specifications of engine are as shown in table 2.1. In dealing with this there are some basic assumptions that need to be addressed.

Fig 2.3: Figure showing single cylinder, four stroke, and VCR (Variable Compression Ratio) Diesel engine Table 2.1: Specifications of Diesel Engine [11] Features Specifications Model TV1 Make Kirloskar oil Engine Type Four stroke, Water cooled Diesel No. of cylinder One Combustion Principle Compression ignition Max speed 2000 rpm www.iosrjournals.org

49 | Page

Analysis and study of vibrations in view of combustion gas forces for diesel engine Min speed Min operating speed Crank Radius Connecting Rod length Cylinder diameter Thickness of cylinder Piston diameter Compression ratio Material of cylinder Stroke length Cut off ratio at 2000rpm

750 rpm 1200 rpm 39.5mm 234mm 87.5mm 10 mm 87.44mm 17.5:1 Grey cast iron 110mm 2.5

2.1 Assumptions i. The piston contacts the cylinder bore at the top and bottom of the piston skirt. ii. Piston skirt deformation and linear elasticity are considered negligible. iii. The skirt/liner oil film exhibits little effect on the transverse motion of the piston. iv. Take the range of Compression pressure as 45, 50, 55, 60, 65, 70, 75 bar. And from that pressure calculate the combustion gas force. v. The gas forces are assumed to act through the centre line of the piston so that the equations of motion apply no matter where the piston is in the cylinder. vi. Behaviour of gases is considered close to ideal gases. vii. Pressure developed after compression is calculated on the basis of assumption that compression process is following isentropic. 2.2 Calculation of Gas forces Calculation of gas forces were carried out to define the compression pressure during combustion in an engine. Following are the calculations performed for featured analysis of vibration due to combustion gas force. Force acting due to gas pressure on piston Force = P * п / 4 * dp2 [9] Where, P = Compression Pressure dp = Piston Diameter

3. Modelling and Simulation Simulation technique can never be complete replacement for an experimental testing. But they can provide a useful service in that the effect of parameter changes may readily be expressed without resource cutting metal, leaving experiment to confirmatory role. Experiment must also provide the evidence needed to validate the mathematical model in first instance. Since, vibration characteristic can be found with empirical relations it is also helpful to get the point load that impact on the side of the cylinder, data was taken from books, which were used to run the program. The studied system representing an I.C. combustion engine is modeled as a cylinder liner which is consisting of cylinder thickness and stroke length. Simulation with the finite element software ANSYS lends to the integration to the piston and cylinder liner. Finite element models were created based on the geometry of the system assembly. The modeling and simulation was carried out using the finite element software ANSYS [12]. For the problem use was made of a two dimensional beam and 3dimensional cylinder liner. A job was created with its attendant job name. I the pre-processor domain, modeling was picked and then area selected to create the job dimensions. In element type regions the pre-processor for selecting the element type for 2-D (plane 42) and for 3-D (Solid 45). The type of material and its properties were chosen in the material properties -material models for solid. An element size edge of 0.001 was used to mesh the model. This was done signifying the default attribute and the plane 42 and Solid 45 with the material model1. Plane 42 is selected to mesh a plane in 2-dimensional beam and solid 45 is selected to mesh a plane in 3-dimension model. Also apply the reference cylinder liner temperature 473K [9]. Material name Aluminum

Table 3.1 Properties of Material [10] Modulus of Poison’s ratio elasticity 22588(HF18) 90000 0.3

www.iosrjournals.org

Density 2700

50 | Page

Analysis and study of vibrations in view of combustion gas forces for diesel engine 3.1 Harmonic Analysis for 2-Dimensional Beam (Cylinder Liner) Harmonic analysis deals with the response of structure to harmonically time varying loads. It gives the ability to predict the sustained dynamic behavior of structure. While the steady state deals with point loads the harmonic analysis deals with varying loads with frequency. For actuating the harmonic analysis, the meshing of 2D beam were completed to process further analysis of the work. It includes entering into solution and to select proper harmonic analysis. Choosing Full solution method and selecting the Frontal solver. In this work the damping ratio 2% for vibration or deflection analysis. Apply boundary conditions, by selecting the define load, by selecting the displacement from structural, as fix the left side of beam (cylinder liner) as all DOF zero and fix the right side of beam as UX zero. Selecting the force / moment, apply the combustion gas force at TDC of cylinder liner and also apply gravity force. For the material HF 18 for 2-D

Fig3.1.1: Nodal solution for 45 bar compression pressure

Fig3.1.2: Graph for 45 bar pressure

Fig3.1.3: Nodal solution for 50 bar compression pressure

Fig3.1.4: Graph for 50bar pressure

Fig3.1.5: Nodal solution for 55 bar compression pressure

Fig3.1.6: Graph for 55 bar

www.iosrjournals.org

51 | Page

Analysis and study of vibrations in view of combustion gas forces for diesel engine

Fig3.1.7: Nodal solution for 60 bar compression pressure

Fig3.1.9: Nodal solution for 65 bar compression pressure

Fig3.1.8: Graph for 60bar pressure

Fig3.1.10: Graph 65 bar Pressure

Fig3.1.11: Nodal solution for 70 bar compression pressure

Fig3.1.13: Nodal solution for 75 bar compression pressure

Fig3.1.12: Graph for 70bar pressure

Fig8.3.14: Graph for 75bar pressure

3.2 Harmonic Analysis for 3-Dimensional Model (Cylinder Liner) For 3-dimensional model, drawing the two concentric circles, considering the thickness and diameter of cylinder and extrudes the same. After extruding, mesh the models. And start the harmonic analysis. In harmonic www.iosrjournals.org

52 | Page

Analysis and study of vibrations in view of combustion gas forces for diesel engine analysis, repeat the procedure as in 2-D up to boundary conditions. Instead of applying the combustion gas force, apply only compression pressure to 3D model. After that, repeat the same procedure as in 2D for getting nodal solutions. 3D model can deflect in two directions i.e. in Y and Z-direction. The output results of harmonic analysis are created using ANSYS on 3D model for all types compression pressures.

4. Results and Discussion This work describes the results obtained after the analysis which includes HF 18 material along with the directions. After applying gas forces to 2-dimensional and 3-dimensional cylinder liner, we get some results. In this chapter we discuss the node number at which we get the deflection due to each compression pressure. In following table, at node 22, the 2-dimensional beam deflects in Y direction due to combustion gas force. Table 4.1: Nodes at which 2-Dimensional beam Deflects in Y direction Compression pressure / Property 22588(HF 18) (Aluminum) 0.25161E-02 45bar 0.27965E-02 50bar 0.30762E-02 55bar 0.33558E-02 60bar 0.36355E-02 65bar 0.39151E-02 70bar 0.41948E-02 75bar

Fig.4.1: Compression Pressure vs. Deflection for 2D From above figure at 45 bar compression pressure the minimum value of deflection and the maximum deflection value at 75 bar pressure for the material aluminium (HF-18). The 3-dimensional model deflects in Y as well as in Z-direction due to combustion gas force. Firstly discuss the deflection in Y-direction. In following table, at node 158, the 3-dimensional beam deflects in Y direction due to combustion gas force. Table 4.2: Nodes at which 3-Dimensional Model Deflects in Y direction Compression pressure / Property 22588(HF 18) (Aluminum) 0.97098E-05 45bar 0.10789E-04 50bar 0.11867E-04 55bar 0.12946E-04 60bar 0.14025E-04 65bar 0.15104E-04 70bar 0.16183E-04 75bar

www.iosrjournals.org

53 | Page

Analysis and study of vibrations in view of combustion gas forces for diesel engine

Fig.4.2: Compression Pressure vs. Deflection in Y-direction for 3D From above figure it is seen that, at 45 bar pressure the minimum value of deflection and the maximum values of deflection at 75 bar pressure. In following table, the 3-dimensional model deflects in Z direction at node 109 due to combustion gas force. Table 4.3: Nodes at which 3-Dimensional Model Deflects in Z direction Compression pressure / Property 22588(HF 18)(Aluminum) 0.65291E-04 45bar 0.72546E-04 50bar 0.79800E-04 55bar 0.87055E-04 60bar 0.94309E-04 65bar 0.10156E-03 70bar 0.10882E-03 75bar

Fig.4.3: Compression pressure vs. Deflection for 3D 4.1 Reduction of Vibration In this work, second major part undertaken to reduce the vibration with design point of view. In addition to this it is proposed from theoretical findings from the analysis that if we add 1mm thickness of cylinder from outer periphery, we obtained better result for reduction in vibration. Thus to reduce the vibration in diesel engine due to combustion gas forces by optimization method through increasing the thickness of cylinder by 1mm found better and the same results were demonstrated by the analysis. Some of the results of 2-dimensional beam HF 18 are shown, vibration get reduced as compared to the conventional thickness of cylinder. Also same results were found for 3-dimensional model. ANSYS-10 version was utilized for post processing the results which are applied for 2-dimensional and 3-dimensional analysis.The results demonstrated by ANSYS were detailed discuss with reference to plots available for vibration reduction testing. In 2-dimensional beam, the maximum deflections for Aluminium at 75 bar (0.41948E-02). But by increasing thickness by 1mm, deflection which is less compared to old as shown in fig 4.4. Also in 3-dimensional model, the maximum deflections for www.iosrjournals.org

54 | Page

Analysis and study of vibrations in view of combustion gas forces for diesel engine Aluminium at 75 bar (0.16183E-04). But by increasing thickness by 1mm deflection which is less compared to old as shown in fig 4.5.

Fig 4.4: Reduction of vibration 2D

Fig 4.5: Reduction of vibration for 3D

5. Conclusion In this work vibration analysis was carried out for 2-D and 3-D model along with different direction for HF 18 material at different nodes. The analysis was done on HF 18 material for compression testing in the range of 45-75 bar. It is observed that during under compression pressure the vibration and deflection characteristics were small at lower range of comp. pressure. Further in order to reduce vibration level, thickness of outer periphery of cylinder increased by 1mm. Better results were obtained i.e. the level vibration is found to be less for HF 18 material. Therefore it is observed that we can predict and understand vibration levels in a better way during engine operating conditions.

References [1]. [2]. [3]. [4]. [5]. [6]. [7]. [8]. [9]. [10]. [11]. [12].

W.J.Griffiths and J.Skorecki [1964] “Some aspects of vibration of a single cylinder diesel engine”, Journal of sound and vibration vol.1 (345-364). S.D.Haddad and H.L.Pullen [1974], “Piston slap as source of noise and vibration in diesel engine”, Journal of sound and vibra tion 34(2), (249-260). L. Chabot PSA Peugeot Citroen, Oliver G.K. Yates Ricardo Group, “Noise and Vibration Optimization of a Gasoline Engine”, Fifth Ricardo Software International User Conference in Detroit [2000]. H. Zheng , G.R. Liu, J.S. Tao, K.Y. Lam, “FEM/BEM analysis of diesel piston-slap induced ship hull vibration and underwater noise”, ELSEVIER, Applied Acoustics 62 [2001] 341-358. S.H. Cho, S.T. Ahn And Y.H. Kim [2002], “A simple model to estimate the impact force induced by piston slap”, Journal of Soun d and vibration Vol. 255 (229-242). Takayushi Aoyama, Sigeo Suzuki, Atasushi Kawamto, Takashi Noda, Toshihiro Ozasa, Takeyushi Kato, Takashi Ito “Preventive Design and Analysis of Cavitation Noise on Diesel Engine”[2004], R and D Review of Toyata CRDL Vol. 40 No.1 36-42. Nobutaka TSUJIUCHI, Takayuki KOIZUMI, Shinya UEMURA Department of Engineering, Doshisha University, Japan[2006], “Modeling of Engine Block and Response Analysis of Piston Slap”. Anthony Deku, Subramanya Kompella, Department of Mechanical Engg., Blekinge, Sweedon, “Cavitation in Engine Cooling Fluid Due to Piston Cylinder Assembly Forces”, ISRN : BTH-AMT-EX-[2006]/D-13—SE. R.S.Khurmi and J.K.Gupta, “Machine Design”, S.Chand Publication[page no.[1137]. Dr. Sadhu Singh, “Machine Design”, Khanna Publication Table A.4, page no. 1126 and 1143. VCR Engine Test set up, 1-cylinder, 4-strike, Diesel (Computerised), Kirloskar Oil Engine make, Instruction manual, Apex Innovations. ANSYS Release 10 Documentation.

www.iosrjournals.org

55 | Page

IOSR Journal of Mechanical and Civil Engineering (IOSR-JMCE) e-ISSN: 2278-1684,p-ISSN: 2320-334X, Volume 6, Issue 3 (May. - Jun. 2013), PP 56-62 www.iosrjournals.org

Effect of Sodium Sulphate on the Index Properties and Compaction Behaviour of Neyveli Fly Ash- Shedi Soil Mixtures Dr.H.N.Ramesh1, Dr.H.S.Nanda2 and Dr. K.V. Manoj Krishna3 1

(Department of Civil Engineering, UVCE, / Bangalore University,JB Campus , Bangalore-56,India) 2 (Principal, Bangalore Technological Institute, / VT University, Bangalore-35,India) 3 (Department of Civil Engineering, Govt SKSJTI,/ VT University, K R Circle Bangalor-01, India)

Abstract: This paper presents the effect of abundantly available fly ash, on the index properties namely liquid and plastic limit and compaction characteristics of shedi soil. Shedi soil is a problematic soil that lies between top low level laterite and bottom high level laterites in the western coastal area of Karnataka, India. The effect of sodium salts on this shedi soil optimized with Neyveli Fly ash has also been studied. Considerable changes in the index properties and compaction characteristics were observed which are explained based on series of experimental results. Addition of Neyveli fly ash improved the workability of shedi soil considerably. The addition of sodium sulphate to the optimum combination of shedi soil-Neyveli fly ash mixture increases the shear strength of the mixture. The maximum dry density also found increased with the addition of sodium sulphate. Keywords - Compaction, Laterite, Maximum Dry Density, Neyveli Fly ash

I.

INTRODUCTION

Current energy policies worldwide consider thermal power generation as a major source of power for industrial development. Fly ash generated from the combustion of coal presently comprises of about million tones every year. Fly ash can be deposited in lagoons or in open dumps. The environmental impact of such disposal schemes in terms of land use, pollution of air and water is quite serious in addition to the enormous cost involved. Research work and studies have been carried out for utilizing fly ash for various purposes to achieve economy and disposal problem. Vast quantities of fly ash have been used in geotechnical engineering for construction of embankments, dams, as backfill behind retaining walls and for land reclamation fill (Faber et al., 1974 [1] ; Holm et al., 1983[2]; Torrey, 1978[3] and Li and Dutton, 1991[4]). The major advantages of using fly ash, as a fill material is its lower specific gravity and higher shear strength. Because of its pozzolanic property, the presence of free lime content and the coarser and inertness of its particles, it can be used for the stabilization of weak soils. Generally the plasticity, workability and strength properties of fine-grained soils are improved by the addition of small quantities of lime (Bell 1988a and 1988b[5]; Dal hunter 1988[6]). Lime-fly ash mixtures have proved to be very effective and economical for use in base and sub-base layers of pavement systems (Torrey, 1978[3]). Dispersive soils, which are highly susceptible to erosion on mixing with fly ash and curing for a sufficient period of time not only become resistant to erosion but also gain strength (Indraratna et al; 1991[7]). The mixtures of lime and fly ash are useful for stabilization of various Indian soils of classes CL, ML, SC and CH, Uppal and Dhawan (1968)[8].Lime-fly ash mixture improved the consistency, shrinkage and unconfined compressive strength of a wide variety of soils, Goecker et al. (1956) [9]. So far, no data are available on the effect of the fly ash content on the index properties and compaction characteristics of shedi soil. To enhance the strength further, sodium salts are often used with fly ash. (Ramesh et al. 1999) [10]. In the coastal area of Karnataka, India, a special type of problematic soil called as shedi soil is available from a depth of about 2 meters to 20 meters, which is underlying relatively good lateritic soil layer. This soil is having size distributive between JEDI (clay) and GODI (silt) soils but do not show the behaivour of the clay nor silt. These soils dissolve and flow like water when ever they come in contact with water which creates cavities and this leads to sliding of the top layers. Hence in this study an attempt is made to study the effect of Neyveli fly ash and sodium sulphate on index properties and compaction characteristics of shedi soil.

II.

EXPERIMENTAL WORK

2.1 Material used Shedi soil: Shedi soil was collected from shedi gudda near Mangalore, Karnataka State, India by open excavation from a depth of 2 m below natural ground level. The shedi soil was air-dried and pulverized in a ball mill after separating the pebbles. This pulverized soil which was passed through 425-micron sieve is used for the investigation. The physical properties of the collected soil sample are presented in Table 1.

www.iosrjournals.org

56 | Page

Effect of Sodium Sulphate on the Index Properties and Compaction Behaviour of Neyveli Fly AshTable 1 Physical Properties of Shedi Soil Parameter Colour

Value Light pink

Specific gravity

2.43

Grain size distribution: Gravel size fraction (%) 0.00 Sand size fraction (%) 0.00 Silt size fraction (%) 85.00 Clay size fraction (%) 15.00 Atterberg’s limits: Liquid limit (%) 27 .00 Plastic limit (%) 17.04 Plasticity index (%) 9.96 Compaction characteristics: Optimum moisture content (%) 13.70 Maximum dry density (kN/m3) 19.02 Coefficient of permeability (mm/sec) 3.652X10-6 Consolidation characteristics: Coefficient of consolidation (Cv mm2/sec) 70.8x10-3 2 Compression index (Cc mm /sec) 1.69x10-3 Shear strength parameters: Friction angle (Degree) 40o 2 Cohesion (kN/m ) 13.1 2 Unconfined compressive strength (kN/m ) 145.5 Neyveli fly ash: The fly ash was collected from Neyveli lignite corporation, Tamilnadu, India. It is pozzolanic fly ash belonging to the ASTM classification “A” Its physical properties are presented in Table 2.

Table 2 Physical Properties of Neyveli Fly Ash Colour

Dark grey

Specific gravity

2.67

Grain size distribution : Sand fraction (%) Silt size fraction (%) Clay size fraction (%) Atterberg’s limits: Liquid Limit (%) Plastic Limit (%) Compaction Characters : Optimum Moisture Content (%) Maximum Dry Density (kN/M3) Unconfined Compressive Strength at MDD (kN/M2)

Nil 87.00 31.00 39.30 NP 24.30 15.20 124.00

2.2 Test Procedure Soil was mixed with fly ash and all the tests were conducted as per BIS: 2720 guidelines, except for compaction test. Compaction test was conducted using both mini compaction test apparatus (Sridharan and Sivapullaiah,2005) [11] and Light compaction test as per BIS; 2720 Part VII (1980) [12].

III.

Results And Discussions

The index properties test like atterburg’s limits and compaction test were carried out for Shedi soil mixed with various percentage of Neyveli fly ash, shedi soil and optimum percentage of neyveli fly ash mixture treated with 1% sodium sulphate and the results are discussed. However, the Neyveli fly ash is found to be non-plastic material. www.iosrjournals.org

57 | Page

Effect of Sodium Sulphate on the Index Properties and Compaction Behaviour of Neyveli Fly Ash3.1 Effect of Neyveli Fly Ash on Liquid Limit of Shedi soil Addition of 10% Neyveli fly ash to shedi soil increases the liquid limit from 27% to 31.4% and remains almost constant up to 15days of curing and then reduces to 24.2% after 30days of curing. With further addition of Neyveli fly ash up to 90%, the liquid limit further increases with immediate testing as well as with curing up to 15days and then reduces marginally after 30days of curing. The variation of liquid limit of shedi soil with the addition of Neyveli fly ash at various curing periods is shown in Fig.1 (a) and Fig.1 (b). On addition of Neyveli fly ash, the liquid limit varies due to (i) the effect of free lime content in the Neyveli fly ash and (ii) the effect of coarser Neyveli fly ash, which acts as diluents, decreases the liquid limit. The increase in liquid limit of shedi soil shows that the effect of flocculation overrides the effect due to reduction in the diffuse double layer thickness and the effect due to dilution. With curing the reduction in liquid limit of shedi soil shows that the effect of dilution dominates the effect due to flocculation. 3.2 Effect of Sodium Sulphate on the Liquid limit of Neyveli Fly Ash-Shedi Soil mixtures With the addition of 1% Sodium Sulphate to Shedi soil, optimum Neyveli fly ash mixtures the liquid limit decreases on immediate testing. However, there is increase in liquid limit with curing up to 7days and remains almost constant with further curing. The variation of liquid limit of shedi soil optimum Neyveli fly ash mixtures treated with 1% Na2SO4 are shown in Table 3 and Fig.2. Table3 Liquid Limit of Shedi Soil -Optimum Neyveli Fly Ash(NFA) Mixture Treated with 1% Sodium Sulphate Liquid Limit (%) Curing Period 0 day 7 days 30 days 27 27 27 32.8 37.1 36.2

Mixture

Liquid Limit (%)

Shedi soil (SS) alone SS + 20% NFA SS + 20% NFA + 1% Na2SO4

27.7

36.9

36.2

Neyveli fly ash (NFA) alone

41.2

65.8

63.2

15 Days

30 Days

120 110 100 90 80 70 60 50 40 30 20

0 Day

0

20

7 Days

40

60

80

100

Percentage of NFA

Fig.1(a) Effect of Neyveli Fly Ash (NFA) on the liquid limit of Sheid soil

www.iosrjournals.org

58 | Page

Effect of Sodium Sulphate on the Index Properties and Compaction Behaviour of Neyveli Fly AshSS alone

NFA alone

SS + 10% NFA

SS + 20% NFA SS + 50% NFA

SS + 30% NFA SS + 60% NFA

SS + 40% NFA SS + 70% NFA

SS + 80% NFA

SS + 90% NFA

Liquid Limit (%)

120 110 100 90 80 70 60 50 40 30 20 0

5

10 15 20 Curing Time in Days

25

30

Fig.1(b) Variation of Liquid Limit of Shedi Soil Treated with Various Percentage of NFA

Liquid Limit (%)1

80 SS alone

70 60

NFA alone

50 SS + 20% NFA

40 30

SS + 20% NFA + 1% Na2SO4

20 0

5

10

15

20

25

30

Curing Time in Days Fig.2 Variation of Liquid Limit of Shedi Soil-NFA mixtures treated with 1% Sodium Sulphate 3.3 Effect of Neyveli Fly Ash on Shrinkage Limit of Shedi Soil Addition of Neyveli fly ash to shedi soil increases the shrinkage limit both with immediate mixing as well as with various curing period. The shrinkage limit increases with the increase in percentage of Neyveli fly ash. The variation is also shown in Fig. 3. The increase in shrinkage limit with the addition of Neyveli fly ash is mainly due to flocculation of clay particles by free lime present in the Neyveli fly ash and partly due to substitution of finer particles of shedi soil by coarser fly ash particles. Increase in shrinkage limit by addition of coarser fraction is relatively less where as flocculation of lime brings about substantial increase in shrinkage limit.

www.iosrjournals.org

59 | Page

Shrinkage Limit (%)

Effect of Sodium Sulphate on the Index Properties and Compaction Behaviour of Neyveli Fly Ash60 55 50 45 40 35 30 25 20 15 10 0

0 Day

7 Days

20

40

15 Days

60

30 Days

80

100

Percentage of NFA

Fig.3 Effect of NFA on Shrinkage Limit of Shedi Soil at Various Curing Periods 3.4 Effect of Sodium Sulphate on the Shrinkage Limit Behaviour of Neyveli Fly Ash-Shedi Soil Mixtures The optimum percentage of Neyveli fly ash for shedi soil is 20%. The optimum percentage of Neyveli fly ash mixed with shedi soil and 1% Na 2SO4 and the variation of shrinkage limit is studied up to 30days of curing period. The shrinkage limit of shedi soil + optimum Neyveli fly ash is 29.6% and increases initially to 34.3% with the addition of 1% Na2SO4. However the shrinkage limit decreases with curing period as shown in Table 4 and Fig. 4. While the presence of sulphate may not have any effect on capillary forces, may reduce the shear strength particularly at shrinkage limit water content. Thus it can be seen that, while the liquid limit of lime treated shedi soil has increased in the presence of sulphate, their shrinkage limit is reduced. This confirms that presence of sulphate does not enhance the strength of the soil.

Table 4 Shrinkage Limit of Shedi Soil-Optimum Neyveli Fly Ash Mixture Treated with 1%Sodium Sulphate Shrinkage Limit (%) Curing Period 0 day 7 days 30 days 16.8 16.8 16.8

A. Mixture Shedi soil (SS) alone Neyveli fly ash (NFA) alone SS + 20% NFA SS + 20% NFA + 1% Na2SO4

70

50.4 39.3

55.4 38.7

34.3

31.2

30.7

SS alone SS + 20% NFA

60

Shrinkage Limit (%)

34.7 29.6

NFA alone SS + 20% NFA + 1% Na2SO4

50 40 30 20 10 0

5

10

15

20

25

30

Curing Time in Days

Fig. 4 Variation of Shrinkage Limit of Shedi Soill – NFA Mixture Treated with 1% Sodium Sulphate www.iosrjournals.org

60 | Page

Effect of Sodium Sulphate on the Index Properties and Compaction Behaviour of Neyveli Fly Ash3.5 Effect of Neyveli Fly Ash on Compaction Characteristics of Shedi Soil As seen from the Table 5 the maximum dry density of shedi soil alone is 19 kN/m3 and the optimum moisture content is 13.6 %. Addition of 10% Neyveli fly ash reduces the maximum dry density and optimum moisture content. Further addition of Neyveli fly ash beyond 20% the maximum dry density still reduces and optimum moisture content increases. This is due to the fact that percentage of fine fly ash requires more water for interparticle lubrication and at the same time reduces the unit weight (Choudhary, 1994) [13]. The compaction curves for different percentages of Neyveli fly ash are shown in Fig. 5.

Table 5 Compaction Characteristics of Shedi Soil - Neyveli Fly Ash Mixture Maximum Dry Density (kN/m3)

Mixture Shedi soil(SS) alone SS+ 10% NFA SS + 20% NFA SS + 30% NFA SS + 40% NFA SS + 50% NFA SS + 60% NFA SS + 70% NFA SS + 80% NFA SS + 90% NFA Neyveli fly ash (NFA) alone

Dry Density (kN/m3)

20

19.00 18.50 18.30 17.70 17.30 16.80 16.50 16.20 16.00 15.30 15.24

Optimum Moisture Content (%) 13.60 13.30 11.70 15.20 15.90 18.80 17.70 21.60 23.60 23.70 24.00

SS alone

NFA Alone

SS + 10% NFA

SS + 20% NFA

SS + 30% NFA

SS + 40% NFA

SS + 50% NFA

SS + 60% NFA

SS + 70% NFA

SS + 80% NFA

SS + 90% NFA

18

16

14 5

10

15

20

25

30

Water Content in Percentage

Fig. 5 Density-Water Content Relationship of Shedi Soil Treated with Various Percentage of Neyveli Fly Ash 3.6 Effect of 1% Sodium Sulphate on Shedi soil-Neyveli Fly Ash mixtures Addition of 20% Neyveli fly ash has been chosen as optimum percentage to shedi soil. Table 6 shows the variation of maximum dry density and optimum moisture content for shedi soil alone, Neyveli fly ash alone, shedi soil + 20% Neyveli fly ash and shedi soil + 20% Neyveli fly ash + 1% Na 2SO4. Maximum dry density and optimum moisture content increases with the addition of 1% Na 2SO4 to shedi soil + 20% Neyveli fly ash mixture. This indicates the improvement in strength of soil for immediate testing with the addition of 1% Na2SO4. Fig. 6 clearly indicates the variation of the maximum dry density and optimum moisture content of shedi soil - optimum Neyveli fly ash mixture treated with of 1% Na2SO4.

www.iosrjournals.org

61 | Page

Effect of Sodium Sulphate on the Index Properties and Compaction Behaviour of Neyveli Fly AshTable 6 Compaction Characteristics of Shedi Soil+Optimum Neyveli Fly Ash Treated with 1% Sodium Sulphate Mixture Shedi soil (SS) alone Neyveli fly ash (NFA) alone SS + 20% NFA SS + 20% NFA + 1% Na2SO4

Dry

Optimum Moisture Content (%) 13.60 24.00 11.70 13.90

SS alone NFA alone SS + 20% NFA SS + 20% NFA + 1% Na2SO4

20

Dry Density (kN/m3)

Maximum Density (kN/m3) 19.00 15.24 18.30 18.60

18 16 14 12 0

5

10

15

20

25

30

Water Content in Percentage

Fig. 6 Density-Water Content Relationship of Shedi Soil+Optimum NFA+1% Sodium Sulphate

IV.

Conclusions

Based on the detailed experimental investigation and analysis of the results obtained the following conclusions have been drawn: 1.Addition of shedi soil with varying percentage of Neyveli fly ash, liquid limit increases up to 15 days of curing beyond which it reduces. 2.Addition of 1 % sodium sulphate to shedi soil-Neyveli fly ash mixture, liquid limit reduces. There is an increase in optimum moisture content and decrease in maximum dry density with the increasing percentage of Neyveli fly ash to shedi soil. 3. Maximum dry density and optimum moisture content increases with the addition of 1 % sodium sulphate to shedi soil-Neyveli fly ash mixtures.

REFERENCES [1]. [2]. [3]. [4]. [5]. [6]. [7]. [8]. [9]. [10]. [11]. [12]. [13].

J.H.Faber, and A.M.Digioia,(1974). Use of ash in embankment construction, Research Record No.593, Transport Research Board, Washington, D.C., U.S.A., pp. 13 - 19. G.Holm, R.Trank, and A.Ekstrom (1983). Improving lime column strength with gypsum. Proc. of VIII European Conference. on SM & FE, Finnish Geotechnical Soc. pp. 903 - 907. S.Torrey, (1978). Coal ash utlization - fly ash, bottom ash and slag, Noyes Data Corp.Park Ridge. NJ. K.S. Li, and C.Dutton,(1991). Geotechnical properties of pulverized fuel ash as a reclamation fill, 9th ARC, Theme 5. F.G. Bell (1988a). Stabilization and treatment of clay soils with lime part I - Basic principles, Ground Engineering, Vol. 21, pp. 10 15. Dal Hunter, (1988). Lime induced heave in sulphate bearing clay soil, Journal of Geotechnical Engineering Division, ASCE, Vol.114, No.2, pp.150 - 167. B. Indraratna, P.Nutalaya. and N. Kuganenthira(1991). Stabilization of dispersive soil by blending with fly ash, Journal of Engineering Geology, No.24, pp. 275 - 290. H.L.Uppal, and P.K. Dhawan, (1968). A resume on the use of fly ash in soil stabilization, Road Research Papers, No.95. W.L. Goecker (1956) . Stabilization of fine coarse grained soils with lime and fly ash mixtures, H .R. B. Bulletin 129, pp.63 H.N. Ramesh, Sivamohan and P..V. Sivapullaiah (1999). Improvement of strength of fly ash with lime and sodium salts, Ground Improvement, No.3, pp.163 - 167. A.Sridharan, and P.V.Sivapullaiah (2005). Mini compaction test apparatus for fine grained soils, ASTM Journal of Testing and Evaluation, Vol.28, No.3, pp. 240 - 246. BIS 2720 (Part VII) (1980), Determination of Light compaction, Bureau of Indian Standards, New Delhi. A.K. Choudhary, (1994). Influence of fly ash on the characteristics of expansive soil, Indian Geotechnical Journal, pp. 215 - 218.

www.iosrjournals.org

62 | Page

IOSR Journal of Mechanical and Civil Engineering (IOSR-JMCE) e-ISSN: 2278-1684,p-ISSN: 2320-334X, Volume 6, Issue 3 (May. - Jun. 2013), PP 63-70 www.iosrjournals.org

Experimental Investigation and Analysis of Extrusion of Lead from Round Section through Triangular Section Converging Dies: As Applied To Forward Metal Extrusion Debabrata Rath1, Dr. Sushanta Tripathy2 1

(Mechanical Engineering Department, CMJ University, India) 2 (School of Mechanical Engineering, KIIT University, India)

Abstract : An experimental investigation has been done on the changes of die angle, area reduction in dies, loading rate on the final extruded products, extrusion pressures of lead of circular cross sections of different length. The proposed method is successfully adapted to the extrusion of the equilateral triangular section from round billet through converging dies of different area reductions. Computation of extrusion pressure at various area reductions and finite element analysis of different parameters (stress, strain, velocity) both in dry and wet condition. Keywords - Converging dies, Extrusion of the equilateral triangular section, Extrusion Pressure

I.

Introduction

The extrusion process has been analyzed by various experimental, analytical and numerical methods. The majority of these techniques have been utilized for determining the extrusion pressure. For many metal working processes, exact solutions are not available and many attempts have been taken for approximate methods which could be adopted for estimating the loads necessary to cause plastic deformation. Now it is becoming essential to pay greater attention to the extrusion of section rod from round stock, as this operation offers the promise of an economic production route. Despite the advantage of converging dies, only a few theoretical approaches to the extrusion or drawing processes for 3D shapes have been published. Nagpal and Altan [1] introduced the concept of dual stream functions to express 3D flow in the die and analyzed the force of extrusion from round billet to elliptical bars. Basily and Sansome [2] made an upper-bound analysis of drawing of square sections from round billets by using triangular elements at entry and exit of the die. Yang and Lee [3] proposed kinematically admissible velocity fields for the extrusion of billets having generalized cross-sections, where the similarity in the profile of cross-section was assumed to be maintained throughout deformation. They analyzed the extrusion of polynomial billet with rectilinear and curvilinear sides. Johnson and Kudo [4] have proposed upper bound for plain strain axis-symmetry extrusion, for extrusion through smooth square dies. in this case materials was assumed to be rigid, perfectly plastic and work hardening effect being neglected. Hill et al. [5] proposed the first genuine attempt to develop a general method of analysis of three dimensional metal deformation problem choosing a class of velocity field that nearly satisfies the statistically requirements, by using the virtual work principle for the continuum. Prakash and Khan [6] made an upper-bound analysis of extrusion and drawing through dies of polygonal cross-sections with straight stream lines, where the similarity in shape was maintained. The upper-bound technique appears to be a useful tool for analyzing 3D metal forming problems when the objective of such an analysis is limited to prediction of the deformation load and study of metal flow during the process. P.K. Kar and N.S Das [7] modified this technique of discretizing the deformation zone into elementary rigid regions to solve problems with dissimilar billet and product sections. However, their formulation was also limited to problems with flat boundaries and as such; the analysis of extrusion from round billets is excluded from their formulation. However P.K. Kar and S.K. Sahoo [8] used the reformulated spatial elementary rigid region (SERR) technique for the analysis of round-to-square extrusion by approximating the circle into a polygon and successively increasing the number of sides of this approximating polygon until the extrusion pressure converged by using tapper dies. Narayanasamy et al. [9] proposed an analytical method for designing the streamlined extrusion have the cosine profile and an upper bound analysis is proposed for the extrusion of circular section from circular billets. Hosino and Gunasekara [10] made an upper bound solution for extrusion and drawing of square section from round billets through converging dies formed by an envelope of straight lines. Boer et al. [11] applied the upper bound approach to drawing of square rods from round stock, by employing a method of co-ordinate transformation. Knoerr et al. [12] have investigated the die fill, defect analysis and prevention, as well as for the prediction of part properties by using DEFORM FEM software. Kang et al. [13] have performed finite element simulation of hot extrusion for copper-clad aluminium rod to predict the distributions of temperature, effective www.iosrjournals.org

63 | Page

Experimental investigation and Analysis of Extrusion of Lead from Round section through Triangular stress, and effective strain rate and mean stress for various sheath thickness, die exit diameter and die temperature and validated with experiments. A.K. Rout and K.P.Maity [14] investigated the optimum pressure during extrusion of square sections from square billet by using curved dies. In addition to this the principle laid by Johnson and Mellor[15], the strain in the present case was calculated from the empirical relation.

II.

Experimental Investigation

Before applying the theoretically result to any practical situation, their adequacy needs experimental verification. The objective of present work is to compare theoretically predicted extrusion load with experimental value. Experiments are performed for TRIANGULAR section using converging die. Commercially available Lead was chosen as the working material for the experiments (extrusion). Different shape of same circular entry face and various reduction of triangular exit face were made (60%, 70%, and 90%).

2.1.Determination of Material Behavoiur A serious limitation of the tensile test even for cold working is that fracture occurs at a moderate strain; so that it is not possible to use this test for determination of yield stress, after very heavy deformation. The fracture is most easily avoided by adopting some compressive form of test. The average stress state during testing is similar to that in much bulk deformation process, without introducing the problems of necking or material orientation. Therefore, in compression test, a large amount of deformation can be achieved before fracture. By controlling the barreling of the specimen ends and the anvils with lubricants the strain can be varied under limits. 2.1.1. Experimental procedure of Compression Test Cylinders with a 50.20mm × 31.77 mm (H/D = 1.5 to 2) were used to obtain the stress-strain curve by a compression test using UTM (INSTRON 400 KN) at room temperature. The compression rate is maintained same as that adapted for the experiments. The specimen has oil grooves on both the ends to entrap lubricant during the compression test. The compression load is recorded at every 0.5 mm of punch travel. After compressing the specimen to about 10 mm it is taken out of the press, re-machined to cylindrical shape with original diameter, and tested in compression till the specimen is reduced to about 10mm. The stress-strain diagram is drawn and the curve is extrapolated beyond a natural strain 0.5. To simulate a rigid plastic material, the wavy portion is approximated by smooth line (Fig. 2.1). The average flow stress of the used lead is found to be 25460 KN/m2.

Fig 2.1.variation of true stress with true strain

2.2.Measurement of Friction Factor The local value of the friction cannot be easily determined. The coefficient of friction may actually vary through a working pass, as the lubrication deteriorates due to thinning of the film and extension of their surface. Experimental studies suggests, however that this is negligible for all well lubricated operations. There is at present no generally accepted method of measuring the value of the coefficient of friction for given surface and lubricant. Various factors can influence the result, chemical condition, lubricant film thickness; temperature, speed, environment and degree of deformation should match as closely as possible the actual conditions of the operation. The friction factor can be measured by the following methods. • Direct measurement of friction in metal working. • Coefficients obtained from correlation of theory. • Measurements depending upon shape change.

www.iosrjournals.org

64 | Page

Experimental investigation and Analysis of Extrusion of Lead from Round section through Triangular 2.2.1. Ring Compression Test If the coefficient of friction can be deduced from a change in shape, the yield stress will not enter the derivation, provided the material is homogeneous and there are no serious temperature gradients. Such methods are generally suitable for rapidly strain hardening materials. Ring compression test suggested by Kudo and Kungio and developed by Cockcroft utilizes axial compression of a ring between flat platens. When a flat, ring shaped specimen is upset in the axial direction, the resulting shape change depends only on the amount of compression in the thickness direction and the frictional conditions at the die ring interfaces. If the interfacial friction were zero, the ring would deform in the same manner as a solid disk, with each element flowing outward radially from the center. In case of small but finite interfacial friction, the outside diameter is smaller than in the zero friction case. If the friction exceeds a critical value, frictional resistance to outward flow becomes so high that some of the ring material flows inward to the center. Measurements of the inside diameters of compressed rings provide a particularly sensitive means of studying interfacial friction, because the inside diameter increases if the friction is low and decreases if the friction is higher.The ring thickness is usually expressed in relation to the inside and outside diameters. Under the condition of maximum friction, the largest usable specimen height is obtained with rings of dimensions in the ratio of 6:3:2 i.e, Outer Diameter: Inner Diameter: Height. The ring compression test can be used to measure the flow stress under high strain practical forming conditions. Thus, by measuring the ratio of internal, external diameters after axial compression of a ring of standard dimensions, it is possible to obtain a measure of the friction. Fig2.2 and 2.3 shows rings before and after compression.

Fig 2.2. rings before compression

Fig 2.3. rings after compression 2.2.2. Experimental procedure for Ring Test A ring compression test was carried out at dry condition and commercially available grease lubrication condition. The rings were compressed upto the 4 mm inner diameter, at each 0.5 mm of punch travel inner diameter and height was recorded. The friction factors were found to be 0.75 for dry condition and 0.38 for the lubricated condition by comparing our result with the calibration curve of Male and Cockcroft(Fig 2.4)

www.iosrjournals.org

65 | Page

Experimental investigation and Analysis of Extrusion of Lead from Round section through Triangular

Fig 2.4. theoretical calibration curve for standard ring 6:3:2 [21]

2.3.Design and Manufacturing of Extrusion Die

Fig 2.5. Converging die with 60% reduction

Fig 2.6. Converging die with 70% reduction

Fig 2.7. Converging die with 90% reduction www.iosrjournals.org

66 | Page

Experimental investigation and Analysis of Extrusion of Lead from Round section through Triangular

SL NO. 1 2 3

LENGTH SIDES(mm) 25.55 22.13 12.78

OF

TABLE 2.1. Dimensions of Different Dies AREA % OF AREA HEIGHT (mm2) REDUCTION DIE(mm) 282.67 60 40 212.06 70 40 70.72 90 40

OF

MATERIAL EN 8 EN 8 EN 8

2.3.1. Photographs of the extruded products are shown in Figs. 2.8, 2.9 & 2.10

Fig 2.8. 60% reduction

Fig 2.9. 70% reduction

III.

Fig 2.10 90% reduction

Result and discussion

Fig 3.1. Compressive Load vs Punch Travel graph for 60% reduction (m=0.38)

Fig 3.2. Compressive Load vs Punch Travel graph for 70% reduction(m=0.38)

Fig 3.3. Compressive Load vs Punch Travel graph for 90% reduction(m=0.38) www.iosrjournals.org

67 | Page

Experimental investigation and Analysis of Extrusion of Lead from Round section through Triangular

Fig 3.4 Compressive Load vs Punch Travel graph for 60% reduction(m=0.75)

Fig 3.5 Comparison of R 60% (m=0.38) with R 60% (m=0.75)

Section

Equilateral Triangle

Table3.1 Experimental Result tabulation for m=0.38 & m= 0.75 Punch Reduction ε Area Pav σ0 Condition load (%) (mm2) (N/mm2) (N/mm2) (N) 60 2.174 86000 706.86 121.665 25.460 Wet 70 2.606 95000 706.86 134.4 25.460 (m=0.38) 90 4.254 131000 706.86 185.33 25.460 Dry 60 2.174 102000 706.86 144.30 25.540 (m= 0.75)

Pav/ σ0 (Exp.) 4.78 5.16 7.28 5.65

Fig 3.6 Percentage Reduction vs Pav graph (m=0.38)

IV.

FEM Analysis of Extrusion of Triangular Section from Round Billet through Converging dies

FEM modeling has been carried out using 3D DEFORM software. A solid CAD model for the converging die profile is made using CATIA V5 software. The length of die is taken as 40mm. Similar solid CAD models are developed for three reductions. Young’s Modulus = 14 GPa; Thermal expansion = 2.9 e0.005 mm/mm/sec; Poisson’s ratio = 0.4; Thermal conductivity =34 N/sec/C; Heat capacity = 1.36 N/mm2/C; Emissivity = 0.3. In the modeling, the rigid plastic model for the material is assumed. The die is modeled as rigid body and stationery. The punch is also www.iosrjournals.org

68 | Page

Experimental investigation and Analysis of Extrusion of Lead from Round section through Triangular modeled as rigid body but moves with a velocity are equal to 0.03 mm/sec. The two frictional boundary conditions are taken. In one case, no lubrication is used. It is a condition of dry friction. From ring compression test it corresponds to a constant friction factor m=0.75. In another case, lubricant (commercial automobile grease) is used, it corresponds to a constant friction factor m=0.38.

4.1 Results and discussion From the present modeling, the effective strain, stress, strain rate, velocity and temperature are determined for different reductions as well as frictional boundary conditions. The equations of effective stress and strain are given as follows: Effective stress due to Von-Mises criterion

12 

1   2    2   3    1   3  2

2

2

   ,

(4.1)

,

where, σ1, σ2 and σ3 are the principal stresses; Effective strain rate

 is the flow stress in uniaxial compression (4.2)

.

.

.

where  1 ,  2 and  1 are the principal strain rates and σm = hydrostatic stress. It is observed that at entry plane the velocity is equal to the billet velocity and at the exit plane, the velocity is equal to the product velocity. The operation was carried out at room temperature. There is some rise in temperature in the deformation model. The maximum temperature is around 700C. The analysis has been carried out for triangular section.

Fig.4.1 variation of punch load with respect to punch travel for extrusion of triangular section (60% reduction)

Fig. 4.2 variation of punch load with respect to punch travel for extrusion of triangular section(70% reduction)

Fig. 4.3 variation of punch load with respect to punch travel for extrusion of Triangular section (90% reduction)

www.iosrjournals.org

69 | Page

Experimental investigation and Analysis of Extrusion of Lead from Round section through Triangular

Fig. 4.4 comparison of wet and dry extrusion pressure for extrusion of triangular sections It is indicated that the extrusion pressure increases with increase in reduction and friction factor. It is observed that the extrusion pressure is higher for dry friction.

V.

Conclusion

In this study, a die profile function have been developed for extrusion of triangular section from round billet using a mathematically converging die profile. Solid CAD models of converging die profiles have been developed for extrusion of triangular section. FEM modeling have been carried out by using DEFORM-3D software using tellurium lead as rigid plastic material for extrusion of triangular section from round billet. Various extrusion parameters have been investigated to determine their effects on extrusion product. It was found that extrusion pressure increases with increase in reduction and friction factor. Analysis of extrusion at other different reductions may be carried out to compare the results.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14]

[15] [16] [17] [18] [19] [20] [21]

V. Nagpal, T. Altan, Analysis of the three-dimensional metal flow in extrusion of shapes. With the use of dual stream function, in: Proceedings of the Third North American Metal Research Conference, Pittsburgh, PA, 1975, pp. 26–40. B.B. Basily, D.H. Sansome, Some theoretical considerations for the direct drawing of section rod from round bar, Int. J. Mech. Sci. 18 (1979) 201–209. D.Y. Yang, C.H. Lee, Analysis of three-dimensional extrusion of sections through curved dies by conformal transformations, Int. J. Mech. Sci. 20 (1978) 541–552. W. Johnson, H. Kudo, The Mechanics of Metal Extrusion.(1963) Hill, Analysis metal working process, mechanics of physical solids flow rule(1963) R. Prakash, O.H. Khan, An analysis of plastic flow through polygonal converging dies with generalized boundaries of the zone of plastic deformation, Int. J. Mach.Tool Des. Res. 19(1979)1-9 P.K. Kar, N.S. Das, Upper bound analysis of extrusion of I-section bars from Square, rectangular billets through square dies, Int. J. Mech. Sci. 39 (8) (1997) 925–934. S.K. Sahoo, P.K. Kar , K.C. Singh a. A numerical application of the upper-bound technique for round-to-hexagon extrusion through linearly converging dies, Int. J. Mat. Pro. Tec. 91 (1999) 105–110 Narayanasamy R, Ponalagusamy R, Venkatesan R, Srinivasan P (2006) An upper bound solution to extrusion of circular billet to circular shape through cosine die. Mater Des 27(5):411– 415. doi:10.1016/j.matdes.2004.11.026 J.S. Gunasekera, S. Hosino, Analysis of extrusion or drawing of polygonal sections through straightly conversing dies, J. Engng. Ind.Trans. ASME 104 (1982) 38–43. C.B. Boer, W.R. Schneider, B. Avitzur, An upper bound approach for the direct drawing of square section rod from round bar, in: Proceedings of the 20th International Machine Tool Design and Research Conference, Birmingham, 1980, pp. 149–155 Knoerr, M. Lee,T.Altan, application of the 2D finite element method to simulation of various forming process, International Journal of Materials Processing Technology,33(1-2), 1992,pp 31-55 Kang, B-Soo, Min-Kim,B. and Choi, J.C., perform design in extrusion by the FEM and its experimental confirmation, International Journal of Materials Processing Technology,41(2),1994,pp 237-248. A. K. Rout and K. P. Maity, Numerical and Experimental study on the three-dimensional extrusion of square section from square billet through a polynomial shaped curved die, published in an Int. J. of Advanced Manufacturing Technology, Vol-54, Issue 5, (2011), pp. 495-506. Johnson W, Experiments in the cold extrusion of rods of non-circular sections, Journals of Mechanics and Physics Solids,5,1957,pp.267-281. D.Y. Yang, M.U. Kim, C.H. Lee, A new approach for generalized three-dimensional extrusion of sections from round billets by conformal transformation, IUTAM Symp. Metal Forming Plasticity, Germany, 1979, pp. 204–221. Geoffrey W. Rowe, Introduction to the principle of metal working, Chapter-8 V. Gopinathan, Plasticity Theory and its Application in Metal Forming, Chapter 7 Dieter,G.E., Mechanical Metallurgy (Tata McGraw-Hill Publishing company ltd., New Delhi) DEFORM TM 3D Version 6.1(sp2),User’s manual, Scientific Forming Technologies Corporation,2545 Farmers Drive, Suite 200, Columbus, Ohio,2008,43235. Shabaik. A. and Kobayashi. S., Computer application to the viscoplasticity method, Transactions of the ASME, Journal of Engineering for Industry,89,1967, pp.339-352.

www.iosrjournals.org

70 | Page

IOSR Journal of Mechanical and Civil Engineering (IOSR-JMCE) e-ISSN: 2278-1684,p-ISSN: 2320-334X, Volume 6, Issue 3 (May. - Jun. 2013), PP 71-75 www.iosrjournals.org

Uniform particle distribution by a newer method in composite metal of Al/SiC Ajay kumar sahu , Payodhar padhi* Department of mechanical engineering ,Orissa engineering college, BPUT, India * Department of mechanical engineering , Konark institute of science& Technology, BPUT, India.

Abstract: Preparation of composites of metal with ceramic particle reinforced through the casting process is not uniform because of poor wet ability. The major difficulty is to get a uniform distribution of reinforcement especially in higher volume fractions. An innovative method of producing cast composites is tried in present study to overcome this problem we need homogeneity of matrix. The method involves multi axis rotation of liquid aluminum and silicon carbide particulates packed in a steel pipe inside a rotating drum. Up to 65 % volume of the metal (aluminum)is incorporated by SIC by this technique. Physical Properties like hardness, micro hardness, densities and microstructures have been studied. The distribution of particles as well the mechanical properties are better as compared to that of stir cast composites with similar volume fraction of silicon carbide reinforcement. The composite with 65-volume percentage of silicon carbide of particulates showed a Rockwell Hardness value of 67Rb.In few locations the microstructure showed a non-uniform distribution which can be neglected . There were segregation of silicon carbide particles at a particular location and the hardness obtained there was much higher. The particle distribution is a result of the combined influence of random mixing of particles and liquid aluminum and the solidification pattern obtained. Key word: Multi axis rotation, microstructure, MMC, Al- SIC matrix

I.

Introduction

High bulk and rigidity modulus, high mechanical strength, high wear resistance, low thermal and electrical conductivity and low sensitive to temperature variations of metal matrix composites (MMC) has drawn high attention for applications from high structural members to electronic packaging. General type of composites lacks behind the Particle reinforced MMCs since they show isotropic behavior. One of the most investigated MMC combination of Al/SiC particulates in a matrix form. This system has the potential for light and high structural components and low thermal expansion electronic packaging applications (Clyne, 1993 and Duralcan, 1996). Metal matrix composites can be processed using number of techniques. These techniques can be broadly classified as a) liquid phase techniques b) solid phase techniques c) semi solid techniques (Gupta et al., 1993). In present work an innovative method of producing cast composites has been attempted. The method involves multi axis random rotation (Mishra et al., 1993) of liquid aluminum and Sic particulates enclosed in a steel pipe which is kept inside a rotating spherical drum. The effect of vibration and mould rotation about one axis on the solidifying melt during casting process has been studied by many investigators (Campbell, 1981) with partial success. Centrifugal casting is known to enhance drastic segregation of heavy elements (Rohtgi, 1986) towards the periphery and concentrates lighter elements at the central portion. This study on the other hand is an innovative approach to mould rotation around more than one axis at random. The expected advantages of the application of this technique are to get homogeneity of distribution and improved mechanical properties (Sarath Chandra and Ramesh, 2011).

II.

Materials And Methods

Materials: Commercial Aluminum was used as matrix material for the present study. The chemical composition of Al is given in the Table 1. Silicon Carbide particulates were selected as the reinforcement. The composition of Sic particulates is in (wt %) in Table 2. Manufacturing process: Multi-axial random rotation technique (Mishra et al., 1993) was used to process the MMC in which a ball mill was used into which a steel cylindrical sample holder was dropped. The steps of preparation of the sample holder for multi-axial rotation are as follows. A steel tube of size 55mm internal diameter and 55mm length is served as the crucible for melting as well as dies for the casting. Commercially pure aluminum was melted in an electrical resistance pit furnace and cast into cylinders of 55mm matching with the internal diameter of the tube. From the height of the die and the volume fraction of silicon carbide in the composite the exact aluminum and silicon carbide needed was estimated. In the first attempt the www.iosrjournals.org

71 | Page

Uniform particle distribution by a newer method in composite metal of Al/SiC aluminum cylinder was kept at the bottom and rest of the tube was filled with Sic However the attempt was not successful in getting the particles uniformly distributed in the final composite. There was local segregation of particles as seen in micrograph Fig 1(a).In subsequent attempts the aluminum needed was divided into two equal halves and the silicon carbide was put in between the two pieces. Then the tube was sealed by a steel plate welded to the tube at top and bottom. Over the tube as above three metallic mild steel rings are suitably placed and welded giving a spherical out line to the assembly. It was done in order to get the random rotation along multi axis just like that in a ball mill. The closed tube as prepared above was heated to a temperature of 1000 C in a furnace to ensure complete melting of Al. Then these were charged into a ball mill with a cylindrical drum of length 600mm and 450mm diameter rotated by a three phase one hp motor at a speed of 50 rpm. The ball mill door was closed instantaneously, and the motor was switched on for the drum to rotate. The rotation was continued for 20 minutes to ensure that composite solidified completely inside the mold. Then the assembly was taken out and the MMC ingot is stripped of for further investigations. The cast cylinder so obtained is cut in the middle along the length and breadth for samples to obtained for study of mechanical properties of the composites

III.

Results And Discussions

Density Measurements: The densities of the aluminum matrix as well as Al/SiC samples were measured by Archimedes principle. The sample was collected at random. The results are given in Table 3. The density of the composite confirms to the expected trend i.e. more the volume fraction of particulates the higher is the density of the composite. This is mainly because of the higher density of the Sic particulates in comparison to Al metal. There is a variation in density from location to location. However this is not very significant showing that the particle distribution is uniform. Further the increase in density is not proportional to the increase the volume fraction of particles as seen in Fig 2. The increase in density of composite is a result of addition of Sic, which has a higher density. Density is also influenced by the presence porosity. The porosity level in cast composite is appreciable and it is reported to increase with increase in particle content. This is reflected in the density variation with increase in particle content. Hardness Measurements: Table no 4 gives the Rb hardness of composites. The hardness was measured in Rockwell hardness tester at a load of 100 kg with the suitable indenter as per B scale. Twelve samples of each volume fractions were tested for hardness. Fig3.clearly shows the upward trend of hardness increase with increased volume fraction of Sic particulates in the MMC. The variations of hardness show a trend very similar to the density variation clearly bringing out the influence of porosity in the samples. The hardness values obtained are higher as compared to the stir cast composites indicating that the porosity in the composites produced by the present method is lower. The hardness variation with location is also minimal showing that the particle distribution obtained is uniform. Microstructure: The SEM and optical micrographs are shown in Fig 1 with different volume fractions of SiC particulates. The micrographs show that there is uniformity of particle distribution as well as good interface between the particle and matrix except for the 10% volume fraction composite. Fig 1(a) shows the structure of composites obtained in the first set of experiments. It shows segregation of particles in a region .The hardness in this zone was also much higher. Micro hardness: The micro hardness values are shown in Table 5. The micro hardness was measured using LECO, DM –400 Micro hardness tester. Micro hardness measurements were made using a paramedical diamond indenter with a facing angle of 136 degree, 25gf indenting load and load dwell time of 15 sec on the matrix and at a specific distance ~ 15micro meter from the interface of the interface of the SiC particulates and the matrix. The hardness values vary in the region around SiC particulate matrix interface depending on distance from interface. However the nature variation does not show any clear trend. There is an increase in the micro hardness of the matrix with addition of the particles. The hardness increased from 28Hv to around 38Hv with addition of particles. However the volume fraction of particles does not have any significant influence of the matrix micro hardness. Increase in micro hardness of the matrix is possibly due to the higher dislocation density levels generated because of the presence of silicon carbide particle structure. There is large change in the micro hardness of the inter face with increase in the volume fraction of the particle. The hardness of the particles also varies with location.

Table 1, compositions of pure commercial aluminum element Wt%

Fe 2.96

Mg 0.43

Si 0.26

Al Balance

Table-2 ;-composition of sic Element Wt%

C 1.16

SiO2 0.65

Al 0.31 www.iosrjournals.org

Fe 0.02

SiC Balance 72 | Page

Uniform particle distribution by a newer method in composite metal of Al/SiC Fraction of Sic particle in( v) % 0 10 15 20 30 65

Table 3. Density Measurement (material ) A/SiC % Density in gm /cm2 Average Density in gm /Cm2 2.69 2.68 2.69 2.69 2.68 2.69 2.70 2.71 2.74 2.74 2.76 2.73 2.72 2.79 2.78 2.76 2.77 2.75 2.78 2.81 2.79 2.77 2.82 2.79 2.85 2.88 2.79 2.86 2.87 2.81 2.87 2.98 2.87 2.96 2.98 2.93

Table 4;-hardness measurements (Rockwell)(Material A/SiCp) Volume fraction 0 10 20 30 65

Load 25gf 25gf 25gf 25gf

Hardness (Rb) 18,2019,18,20,21,19,18,20,17,19,20 38,47,42,39,49,40,40,41,43,37,46,39 46,59,39,52,55,42,51,52,49,49,43,52 66,58,47,68,59,54,53,55,57,56,54,55 71,70,57,60,79,70,69,69,74,80,54,59

Table 5;-Micro hardness measurements Volume Matrix*(Hv) Interface**Hv) fraction 0 28 10 37 47 30 39 75 65 38 113

Average (Rb) 19.0 41.7 49.0 56.8 67,6

Particle (Hv)

2264 3079 2560

Fig 1-Micrograf illustrating the particle distribution in Al/SiC composites www.iosrjournals.org

73 | Page

Uniform particle distribution by a newer method in composite metal of Al/SiC (a)10% volume fraction (60x),(b)10% volume fraction ,(c) 20%volume fraction ,(d)30% volume fraction (g)65%volume fraction 2.95

Average density (gm/cm3)

2.90

2.85

2.80

2.75

2.70

2.65 0

10

20

30

40

50

60

70

% Vol fractions of SiCp

Fig 2: Variation in densities as a function of volume fraction

70

Average hardness (Rb)

60

50

40

30

20 0

10

20

30

40

50

60

70

% Volume fraction of SiCp Fig 3: Variation in hardness as a function of volume fraction

IV.

Conclusions

1: Multi axis rotation can be successfully used to get Aluminum matrix silicon carbide particle reinforced composites up to 65-volume percentage of particles. In most cases the particle distributions obtained is uniform. 2: There is a significant increase in hardness of the composites prepared by this technique. Rate of increase in hardness with particulate addition decrease possibilities due to increase in porosity level. 3: The micro hardness of the matrix increases with addition of particles. However increase in particles dose not influences this value significant. 4: The micro hardness at the interface increases significantly with increases in particulate addition

www.iosrjournals.org

74 | Page

Uniform particle distribution by a newer method in composite metal of Al/SiC References [1]. [2]. [3]. [4]. [5]. [6]. [7]. [8].

Padhi ,.P. PhD Thesis, 2006 C.M.Milliere and M.Surry. Mat Sc &Tech, 4, 41(1998) ―Duralcan Aluminium MMCs ― Duralcan(1994) J.Campbell, International Metals Reviews 26(2): 71-108 (1981) M.Gupta,T.S Srivastan,F.A Mohammad and E.J.Lavernia, J.Mat.Sc .28,2245 (1993) Misra .K. Asoka Metastable Structures (Principle, Design& Appl) Ind-US.w/s-Goa 111-114(1993) P.K.Rohtgi,R.Asthna and S.Das International Metals Review 31 (3): 115—139.(1986) T.W Clyne and P.J Withers,An introduction to MMCs, Cambridge University Press, Cambridge, U.K (1993).

www.iosrjournals.org

75 | Page

IOSR Journal of Mechanical and Civil Engineering (IOSR-JMCE) e-ISSN: 2278-1684,p-ISSN: 2320-334X, Volume 6, Issue 3 (May. - Jun. 2013), PP 76-81 www.iosrjournals.org

Comparative Assessment of Two Thermodynamic Cycles of an aero-derivative Marine Gas Turbine M. U. Bonet1, P. Pilidis2 1. Department of Mechanical Engineering, Nigerian Defence Academy, P.M.B. 2109, Kaduna, Nigeria 2. Buiding 52, School of Engineering, Cranfield University, MK43 0AL, Bedfordshire, United Kingdom Abstract: This paper explores the gas turbine potentials that are fully enhanced by the use of intercooling and thermal recuperation as an engineering option available in the design of gas turbines and offered for marine applications. It examines the off-design performance of two different cycle designs of a 25MW aero-derivative engine by modelling and simulating each of them to operate under conditions other than those of their design point. The simple cycle model consists of a single-spool dual shaft layout while the advanced model is represented by an intercooled-recuperated cycle that runs on a dual-spool and is driven through a three shaft configuration. In each case, the output shaft is coupled to a power turbine through which the propulsion power may be transmitted to the propeller of the vessel to operate in a virtual marine environment. An off-design performance simulation of both engines has been conducted in order to investigate and compare the effect of ambient temperature variation during their part-load operation and particularly when subjected to a variety of marine operating conditions. The study assesses the techno-economic impact of the complex design of the advanced cycle over its simple cycle counterpart and demonstrates its potential for improved operating cost through reduced fuel consumption as a significant step in the current drive for establishing the marine gas turbine engine as a viable alternative to traditional prime movers in the ship propulsion industry.

I.

Introduction

Producing great amounts of energy for its size and weight, the gas turbine has found increasing service in more than 40 years and in conjunction with growth in materials technology and increase in compressor pressure ratio, thermal efficiency has increased from 15% to 45%. It is a continuous flow non-reciprocating internal combustion engine whose initial concept was meant for aircraft propulsion so as to overcome the drawbacks of its reciprocating piston counterpart in the aerospace industry. When modified and adapted for land-based and off-shore applications, it becomes an aero-derivative gas turbine whose lightweight attracts it particularly to naval ship propulsion applications where weight and space are of primary importance [1] and over the last 40 years, Rolls-Royce (RR) and General Electric (GE), which are some of the world’s topmost original equipment manufacturers (OEMs) of gas turbines have since introduced a good number of them into the marine market, starting with the RR’s commercially successful version of the 3MW industrial Proteus engine followed by the Tyne (4MW), the Spey (12.5, 18 or 19MW) [2] and the Olympus models [3] which are all configured on the basis of the simple thermodynamic cycle. The implementation of the simple cycle technology by GE has shown tremendous success in its LM2500 series [4] [5] whose output power transmission is achieved through a power turbine and an output shaft unto which the desired load may be coupled. Another engineering option for advancing the performance of the gas turbine is that of the dual-spool, three-shaft MT30 Marine Trent MT30 [6] complex layout, composed of low and high pressure spools aerodynamically coupled to the free power turbine for the purpose of boosting the basic simple cycle thermal efficiency. In order to overcome the part-load performance challenges when used as a ship propulsion prime mover, the gas turbine can be installed and configured to operate in combination with other traditional ship propulsion prime movers such as the diesel engine either as ‘Combined Diesel and Gas turbine’ (CODAG) or as ‘combined diesel or gas turbine’ (CODOG) configuration. A big and small gas turbine may also be coupled together in a father and son relationship to either operate in a ‘combined gas turbine and gas turbine’ (COGAG) or ‘combined gas turbine or gas turbine’ (COGOG) in which boost power for top speed operations is to be provided by the bigger engine [8] while low speed fuel economy at low power settings is to be provided by the smaller one whenever necessary. Another combination ‘combined diesel Electric and Gas turbine (CODLAG)’ is also used in practice as an option that has been developed and adopted for the propulsion of high speed naval vessels in which a controllable pitch propeller is usually fitted so as to match the different operating conditions for each of the prime movers [9]. A general layout of both thermodynamic cycles is illustrated in Figure 1 showing the flow of the working fluid and how the desired output power is derived via a free power turbine. In practice, Rolls-Royce intercooled recuperated marine model WR21 [7] [8] whose SFC curve illustrated in Figure 2 is showing significant fuel saving capability when compared with other existing cycle options. Its cycle layout incorporates www.iosrjournals.org

76 | Page

Comparative Assessment of Two Thermodynamic Cycles of an aero-derivative Marine Gas Turbine intercooling and heat exchange for the purpose of improving part-load performance at low power demand operations. The intercooler, which represents an air-to-liquid heat exchanger, is useful during the compression process to enhance the specific power of the engine by cooling the airflow midway through the compression process [10].

Figure 1: The component layout and thermodynamic cycle configuration of the simple cycle [1] and the ICR model of the gas turbine [7] The added complexity results in a reduction of the energy required during compression which enhances a higher power output. The recuperator on the other hand transfers the waste heat from the exhaust gases to the engine airflow in order to reduce the fuel required during combustion thereby achieving better performance. Furthermore, the incorporation of an intercooler between the compression stages of the low and high pressure spools increases the specific power of the engine while the amount of work needed to drive the compressor is reduced and this increases the net power available. This advanced cycle is capable of achieving between 40% to 46% thermal efficiency which is considerably higher than the about 36% to 37% achievable in the simple cycle. This is only slightly less than the about 50% achieved by a combined cycle where a steam turbine is installed to use steam from boilers fired by waste heat from the gas turbine to generate additional power. With an estimated lifetime of up to 100,000 hrs the ICR heat exchanger may be of a compact plate and fin design [11] and in contrast to the combined cycle, it is inherently simpler, smaller and cheaper and does not require hours to warm up. It also has considerable waste heat left in the exhaust for possible onboard utilization along with a facility to generate more when the recuperator and intercooler are bypassed. Considered as the most advanced marine gas turbine currently available, it exemplifies the next generation of ship propulsion prime movers aimed at offering a combination of high power density, low fuel consumption and environmentally friendly solutions in the maritime industry. Another aero-derivative gas turbine with a regenerator is the General Electric LM2500R which also employs a recuperator (exhaust gas-to-compressed air heat exchanger) to recover high grade heat in the gas turbine exhaust by means of a counter-current flow compressor discharge air that, after heating, flows on to the combustor. In Figure 2, the specific fuel consumption (SFC) of some practical cycles are drawn to illustrate the successful design investments that have been made and are aimed at the improvement of gas turbine engine performance in which the ICR reveals the best results.

II.

Engine Performance Simulation Platform

The performance of any gas turbine can best be defined as the end product that any manufacturer can sell to its customers and the thread that sews all other gas turbine technologies together. It may be summarised as the thrust or shaft power that may be delivered for a given fuel flow, life, weight, emissions, engine diameter and cost which must be achieved while ensuring a stable and safe operation throughout the operating envelope under all steady and transient conditions. This investigation however limits itself to the running cost that relates to the consumption of fuel in the course of operating the selected models under state conditions. ‘Turbomatch’[12] which is a product of ongoing research efforts in industrial and aircraft gas turbines at the power and propulsion Department of Cranfield University in the UK [13], has the capability of carrying out design point and off design calculations. At its design point mode, it provides engine performance and size data as illustrated in Table 1, while at off-design it predicts the engine performance for varying engine throttle setting (rotational speed, combustor exit temperature, or fuel flow). Its working principle is based on a thermodynamic mass and energy balance from which the off design calculations may be conducted through an iterative generic method in conjunction with experimentally derived maps that are scaled to match the design point of the simulated engine. The choice of using the embedded maps is a result of the structure of the simulation program which comprises several pre-programmed modules that correspond to models of the www.iosrjournals.org

77 | Page

Comparative Assessment of Two Thermodynamic Cycles of an aero-derivative Marine Gas Turbine individual components, such as compressors, burners, turbines, mixers, nozzles, heat exchangers, splitters and the power turbines. Its modular structure enables the detailed design and simulation of any modern industrial or aero gas turbine engine. The simulation software has been validated against commercially sensitive data and further details can be found in [14], [15] and [16], while the working design point and off-design calculations are fully described in [12].

Figure 2: The specific fuel consumption (SFC) profiles of some existing Marine gas turbine engines [13]. Effectively, values of compressor pressure ratios, turbine entry temperatures and component efficiencies as well as pressure losses (regarded as the basic design parameters of any gas turbine) were specified as the main design point input variables. The procedure proceeded with the determination of the free stream static and total conditions at a given altitude using the International Standard Atmosphere (ISA) in order to enable the evaluation of the conditions at exit from the intake. Implementing the ‘TURBOMATCH’ program enabled the modelling and simulation of the selected gas turbine engines with the input data fashioned in manner that allows the building of an engine structure and the component parameters resulting in a compressor characteristic map. While taking many other important factors such as the expected component efficiencies, air bleeds variable fluid properties and pressure losses into account, detailed thermodynamic calculations over a wide range of pressure ratio and turbine entry temperature limits were conducted. The simple and advanced cycle models were then simulated at design and steady state off-design conditions by monitoring and recording every engine parameter and the output data was obtained in a spread sheet for further analysis. Predictable part-load performance in applications where considerable running at low power settings is of major importance and its effect on the specific fuel consumption (SFC) as well as the effect of high and low ambient temperatures and pressures on its maximum output are of high importance to the customer and as such, manufacturers must be prepared to guarantee the performance available at any specified condition hence the propulsion plant of any marine vessel has to be designed to allow for flexibility in operation over a wide range of power levels while still maintaining operating efficiency without incurring losses in fuel economy. The drawing up of a data file that consists of all necessary input and output parameters of the component parts when assumed to have been coupled together according to the cycle configuration of the model under investigation. The simulated off-design performance results have been interpreted and analysed through establishing component characteristics to enable the location of relevant corresponding operating points when the engine is running at a steady speed or in equilibrium for a series of non-dimensional speeds in the form of an equilibrium running diagram. A comparative performance analysis of the two engines was conducted and the result is demonstrated in the form of compressor characteristic maps for each of them as illustrated in Figure 4 for the simple cycle model and in Figure 3: for the advanced cycle model respectively. It will be observed that the ICR consists of a low pressure spool as well as that of a high pressure in order to facilitate the incorporation of the heat exchangers necessary for the cycle layout for it.

III.

Comparative Performance Results

Compressor performance has been evaluated and represented by plotting constant aerodynamic speed lines as a function of delivery pressure and corrected mass flow. The actual mass flow and speeds are corrected by a factor of

𝑇1 𝑃1

and

1 𝑇1

respectively through a non-dimensional method so as to reflect variations in inlet

pressure and temperature. This leads to the compressor performance being expressed through a variation of the delivery pressure P2, and temperature T2, the mass flow m, rotational speed N, and the corresponding inlet conditions in the form of the following non-dimensional groups: www.iosrjournals.org

78 | Page

Comparative Assessment of Two Thermodynamic Cycles of an aero-derivative Marine Gas Turbine 𝑓

𝑃2 𝑇2 𝑚 𝑇1

,

𝑃1 𝑇1

,

𝑃1

𝑁

,

𝑇1

= 0; where the suffixes 1 and 2 represent the compressor inlet and outlet pressures

and temperatures respectively, while the m and the N represent the mass flow and rotational speed respectively. Table 1: Design point parameters of the investigated gas turbine models GT Model/Design Parameter TET [K] Overall Pressure Ratio Intake Mass Flow [kg/s] Exhaust Mass Flow [kg/s] Exhaust Gas Temperature [K] Thermal Efficiency [%] SFC [g/KWh]

25MW ICR

25MW SC

1500 15.52 70 71.36 660 42.58 196.6

1505.5 18.75 72 73.54 804 37.78 221.65

The non-dimensional method of plotting the compressor characteristics implies that all operating conditions covered by values of every pair of

𝑚 𝑇1

and

𝑃1

𝑁 𝑇1

non-dimensional groups yield the relationship between the

pressure ratio, temperature ratio and isentropic efficiency. The surge line joins different speed lines where the compressor’s operation becomes unstable and the equilibrium running line shows the proximity of the operating zone to the compressor surge line such that when these two intersect each other, the engine will not be capable of being brought up to full power without the need for some remedial action before obtaining a stable operation. Note that a compressor in surge when its main flow passing through it reverses direction for a short interval of time during which the back (exit) pressure drops before the flow resumes its proper direction. The process is followed by a rise in back pressure that causes the low to reverse itself again and if allowed to persist, the unsteady process may result in irreparable damage to the engine. There exists also a condition known as ‘choke’ which indicates the maximum possible mass flow rate through a compressor at any operating speed. 9

Pressure Ratio

Pressure Ratio

4 3.5 3 2.5

2 1.5 1

10

30

50

70

90

Surge Line

7 6

5 4

3 2

1 10

NDMF Speed Lines

8

15

20

25

30

35

NDMF

ERL

Speed Lines

Surge Line

ERL

Figure 3: LPC and HPC performance characteristics of the ICR model An illustration of the LPC and the HPC performance characteristics for the ICR model is shown in Figure 3 while the characteristics of the single HPC of the simple cycle model is as demonstrated in Figure 4. The effect of varying ambient atmospheric conditions on engine fuel flow for each of the models was investigated to determine the engine performance response to variations in ambient temperature between (-30)oC and (+45)oC for the simple cycle and between (-25)oC and (+55)oC for the ICR. The variation of the turbine entry temperature (TET) at off-design was adopted at intervals of 50oC running from 950K to a maximum operating temperature of 1550K for the simple cycle and between 1000K and 1550K for the ICR model. This assessment reveals how the fuel flow (kg/s) of the two models differ with changes in TET such that at lower operating temperatures as low as 1200K, values of fuel flow for the ICR appear to be less affected by any significant change in ambient conditions.

www.iosrjournals.org

79 | Page

Comparative Assessment of Two Thermodynamic Cycles of an aero-derivative Marine Gas Turbine

Pressure Ratio

35 30 25 20

15 10 5 0

10

20

30

40

50

60

70

80

90

100

NDMF Non Dimensional Speed Lines

Surge Line

ERL

Figure 4: High Pressure Compressor characteristics of the simple cycle model Under the most adverse investigated scenarios with an ambient temperature of 55oC, the fuel flow still fell short of 0.6kg/s for the ICR model against a figure of over 1.3kg/s in the case of the simple cycle at a tropical ambient temperature of 45oC. The advanced cycle shows an off-design maximum value of 1.73 kg/s against that of the simple cycle which stands at 2.13kg/s when operating under a high ambient temperature of 55 oC and 45oC respectively.

Fuel Flow [kg/s]

Fuel Flow [kg/s]

2.5 2.0 1.5 1.0 0.5

0.0 950

1050

1150

1250

1350

1450

1550

2.00 1.80 1.60 1.40 1.20 1.00 0.80 0.60 0.40 0.20 0.00 1000

1100

1200

1300

1400

1500

TET [K]

TET [K] FF -25

FF - -30 FF - -15 FF - 0 FF - +15 FF - +30 FF - +45

FF -5

FF +15

FF +35

FF +55

Figure 5: Fuel flow variation for the engine models as affected by changes in ambient conditions

950 1050 1150 1250 1350 1450 1550

600

0.50 0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00

500 400 300 200

100 0 1000 1075 1150 1225 1300 1375 1450 1525

TET [K] SFC - -30 SFC - +30 Eff - 0

SFC - -15 SFC - +45 Eff - +15

Thermal efficiency

0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00

SFC [g/KWh]

1600 1400 1200 1000 800 600 400 200 0

Thermal efficiency

SFC [g/KWh]

A similar investigation of the thermal efficiency revealed a maximum performance of over 43% while generating a power output of over 28MW in contrast to a 37.24% at a power output of over 27MW. A reflection of this fuel flow pattern is further illustrated by the SFC and thermal efficiency curves in Figure 6 for comparing the impact of environmental conditions on the marine gas turbine which may depend on season and location of the vessel while navigating in the open sea at any time (season) of the year. It further reveals that the ICR is capable of a SFC of as low as 580kg/s under ambient conditions of 55 oC while the figure for the simple cycle goes as high as 1400kg/s at an operating temperature of 45 oC in the open sea. These results further demonstrate the significant part-load performance qualities of the advance cycle gas turbine engine for marine applications.

TET [K] SFC - 0 Eff - -30 Eff - +30

SFC - +15 Eff - -15 Eff - +45

SFC -25

SFC -5

SFC +15

SFC +35

SFC +55

Eff -25

Eff -5

Eff +15

Eff +35

Eff +55

Figure 6: Variation of SFC and thermal efficiency with TET for the simple cycle and the ICR respectively when operated under a variety of ambient temperatures www.iosrjournals.org

80 | Page

Comparative Assessment of Two Thermodynamic Cycles of an aero-derivative Marine Gas Turbine

SFC [g/KWh]

A further analysis is illustrated in Figure 7 where the SFC varies with changes in engine power output at the selected ambient temperatures of the study as earlier pointed out in Figure 2 above. It shows that the advanced cycle can be very economical at low power settings of as low as 8MW combining the gas turbine with either a smaller version or diesel engine for fuel saving purposes. 800 700 600 500 400 300 200 100 0 0

4

8

12

16

20

24

28

32

36

Engine Power [MW] SFC_SC@15

SFC_ICR@15

SFC_SC@0

SFC_ICR@0

Figure 7: Illustration of the impact of different ambient conditions on the SFC of the models

IV.

Conclusions

This paper assesses the behaviour of the marine gas turbine in response to varying ambient conditions and the potential performance benefits of cold weather operations over that of hot weather. It also explored the advantages of the advanced ICR cycle over the basic simple thermodynamic gas turbine cycle design revealing a higher thermal efficiency of over 43% contrary to the about 37% efficiency of the simple cycle option. In spite of its added complexity, it enhances an economic operating cost of any marine vessel installed with marine gas turbines as the propulsion prime mover. The study has also shown how the advanced cycle technology enhances the installation of a more compact propulsion plant that allows for more space to be made available for accommodation of more onboard cargo and avoids the installation of more complex layouts that may necessitate the combination of different types of prime movers such as CODOG or COGOG etc, so as to improve the partload fuel economy particularly during low speed operations. This has been demonstrated by a comparison of the SFC curves for both models, revealing a flat curve for the ICR at significantly lower engine output power of about 8MW for low speed operations much better than the simple cycle option.

References [1]

[2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16]

Hinks, A. R. (1999), "Aero-Derivative Gas Turbines in the Marine Environment", Gas Turbine Operation and Technology for Land, Sea and Air Propulsion and Power Systems, Vol. 34, 18-21 October, Ottawa, Research and Technology Organization (RTO), France, pp. 11-1. Woodyard, D. (2004), Pounder's Marine Diesel Engines, Elsevier Butterworth-Heinemann. Schamp, A., et al (1999), "Experience with Aero-Derivative Gas Turbines as Marine Propulsion Machinery", Gas Turbine Operation and Technology for Land, Sea and air Propulsion and Power Systems, Vol. 34, 18-21 October, Ottawa, RTO, France, pp. 9-1. GE- Marine, (2006), LM2500 Gas Turbine, General Electric, New York. GE-Energy (2011), www.ge-energy.com (LM2500 Gas Turbines for Offshore Applications - Fact Sheet) (accessed 05). Watson, B., (2006), LNG Shipping Operations, Rolls-Royce, UK. Shepherd, S. B. and Bowen, T. L. and Chiprich, J. M. "Design and Development of the WR-21 Intercooled Recuperated (ICR) Marine Gas Turbine", Journal of Engineering for Gas Turbines and Power, vol. 117, pp. 557-562. Parker, M. L. et al (1998), "Advances in a Gas Turbine System for Ship Propulsion", RTO AVT Symposium on "Gas Turbine Engine Combustion, Emissions and Alternative Fuels", Vol. MP-14, 12 - 16 October 1998, Lisbon, Portugal, NATO, US, pp. 2-1. Rowen, A. L. (2003), "Machinery Considerations", in Lamb, T. (ed.) Ship Design and Construction, SNAME, US, pp. 24:1-24:27. Groghan, D. A. (1992), "Gas turbines", in Harrington, R. L. (ed.) Marine Engineering, The Society of Naval Architects and Marine Engineers, USA, pp. 146-183. Wiens, B. (1996), "Turbo Expo Shows Off Latest in Gas Turbines", September, pp. 38. MacMillan, W. L. (1974), Development of a Modular Type Computer Program for the Calculation of GasTurbine Off Design Performance’, Doctoral dissertation, School of Mechanical Engineering, Cranfield Institute of Technology, Cranfield, UK Palmer, J. R, (1967), The ‘Turbocode Shceme for the Programming of the Thermodynamic Cycle Calculations on an Electronic Digital Computer, COA/AERO-198, College of Aeronautics, Cranfield Institute of Technology Palmer, J. R. and Pachidis, V. (2005), The Turbomatch Scheme; for Aero/Industrial Gas Turbine Engine Design Point/Off Design Performance Calculation, Manual, Cranfield University, UK Sirinoglou, A. (1992), Implementation of Variable Geometry for Gas Turbine Performance Simulation. Turbomatch Improvement, MSc Thesis, SME, Cranfield University, UK Van den Hout, F. (1991), Gas Turbine Performance Simulation Improvements to the Turbomatch Scheme, MSc Thesis, SME, Cranfield University, UK

www.iosrjournals.org

81 | Page

IOSR Journal of Mechanical and Civil Engineering (IOSR-JMCE) e-ISSN: 2278-1684,p-ISSN: 2320-334X, Volume 6, Issue 3 (May. - Jun. 2013), PP 82-86 www.iosrjournals.org

Effects of Ball Milling Conditions on Breakage Response of Baban Tsauni (Nigeria) Lead-Gold Ore. E.A.P. Egbe1,1* 1

Mechanial Engineering Department, Federal University of Technology, Minna, Nigeria.

Abstract: The effects of ball milling conditions on breakage response of Baban Tsauni (Nigeria) lead-gold ore was examined in this research work. The three factors considered were, the grinding media mass to ore sample mass ratio, the grinding time and the ball mill speed. The breakage response measured in terms of cumulative per cent mass passing the liberation size increased with increase in media mass to sample mass ratio and reached optimum at a ratio of 10:1. The response per unit time decreased as the grinding time was increased gradually from 2minutes to 10minutes, thereby indicating that long grinding circuits were poor in performance. The optimum ball milling speed of the ore was found to be 82.9% of the critical mill speed. This work established a rational approach in the choice of grinding speed. The effects of the three parameters on generation of fines were less pronounced. Key words: energy, fines, grindability, optimum, particle size.

I.

Introduction

Cumminution is an energy intensive process and a search for conditions that affect performance will lead to conservation of energy. The Bond’s Work Index is generally used as a measure of the resistance of a material to breakage [1], [2], [3], [4]. The standard Bond grindability test is carried out under specified conditions of mill volume, mill speed, feed particle size, a circulating load of 250% and a control sieve size that depend on the target particle size [2], [5]. The Bond’s work index is highly sensitive to the control sieve size and circulating load [6], [7], [8], [9], [10], [11], [12]. The foregoing indicates that the performance of a grinding mill is affected by the operating conditions in the mill. Excessive grinding of an ore increases energy costs and also leads to production of very fine untreatable slime particles which may be lost into the tailings and production of fines is affected by mill conditions. Thus grinding is a compromise between clean (high-grade) concentrates, operating costs, and losses of fine minerals [3].

1.1 Literature review The effect of grinding conditions on the response of an ore can be studied by varying one condition at a time and deducing information from the data [13]. Man [5] indicates that mill speed affects energy consumption and throughput. The ball size influence mill throughput, power consumption and progeny size [14], [15]. It is reported by [16] that each grain size has an optimum ball size. Energy wastage and drop in throughput are the consequence of deviation from optimum conditions [17], [18], [19]. The critical grinding speed (Vc) of a ball mill (in rpm) is given by [20] as:

Vc 

4 2.3 (D  d )

(1) where D is the mill diameter and d is the ball diameter. The critical speed denotes the speed to be avoided. It does not indicate the optimum speed. The standard Bond’s test is performed at a mill speed of 0.91Vc [5]. Sahoo and Roy [21] observed that ball mills usually operate at speed ranging from 65% to 75% of the critical speed. Adoption of different choice of mill speed by researchers is common {0.752Vc [2]; 0.8Vc [22]; 0.86Vc [23]and 0.85Vc [19]}. The foregoing suggests that further research work is required to determine the optimum ball milling speed for each ore. Overloading of the mill leads to accumulation of fines at the toe of the mill, which result in cushioning effect on the balls upon impact. Contrarily when the material load is low, excessive ball to ball contact retard the rate of breakage [20]. The optimum breakage response of an ore requires the maximum production of particles of the desired size while minimizing the production of untreatable fines. Thus the effects of milling conditions on response must be known in addition to the grindability of an ore.

www.iosrjournals.org

82 | Page

Effects of Ball Milling Conditions on Breakage Response of Baban Tsauni (Nigeria) Lead-Gold Ore. II.

Materials and experimental procedures

About 20kg of ore samples from Baban Tsauni lead-gold ore was used for this research work. An initial sample preparation aimed at ensuring a homogeneous feed for subsequent test runs was carried out. The crushed samples from the jaw crusher /roll crusher ( in that order) were subjected to the same grinding conditions of mill speed, ore mass to ball mass ratio, ball size and grinding time of 10minutes. The products of this batch milling operations were mixed thoroughly and passed through a Jones riffling sampler, until sets of 472g samples were obtained. The critical speed of the mill was determined by applying (1). The first response measured in all the experiments was the percentage of the product mass which passed the liberation size (355 microns) while the second response was the amount of fines generated in terms of per cent mass passing 50μm. The mill was charged with steel balls and the ore materials in layers, starting with the balls. Half a kilogram of the feed materials was used for each experiment. The mill was closed and placed on the rollers and the mill started. After grinding, the mill content was discharged unto a stainless steel bow and the balls were removed. A brush was used to clean the balls so that no material is lost. The content of each sieve fraction was weighed and recorded.

2.1 The effect of ball mass to sample mass ratio on breakage response The grinding speed was kept constant at 101rpm, the holding time at 10minutes, and the ball size at 25mm diameter. The ball mass to sample mass ratio of 2:1, 4:1, 6:1, 8:1, 10:1 and 12:1 were used. The product of each test run was subjected to sieving for 15minutes and the cumulative per cent mass passing 355μm and 50μm were determined.

2.2 The effect of holding time on breakage response The mill speed was maintained at 101rpm, the ball size at 25mm and the ball mass to sample mass ratio was kept constant at 10:1. The mill was run for 2min., 4min., 6min., 8min., and 10min. each. Each product was sieved for 15min. and the cumulative per cent mass passing 355μm and 50μm were determined.

2.3 The effect of milling speed on breakage response In this set of experiments the grinding time was kept constant at 10minutes, the ball mass to sample mass ratio at 10:1 and a grinding media size of 25mm was used. The critical speed of the mill, denoted Vc, was calculated by applying (1) and found to be 131rpm. Five milling speeds of 0.5Vc, 0.62Vc, 0.77Vc, 0.937Vc, and 1.14Vc were used. The product of each test was sieved for 15minutes and the cumulative per cent mass passing 355μm and 50μm were determined.

III. Results and discussion 3.1 The effect of ball mass to sample mass ratio on grinding response of the ore

Cumulative per cent mass passing (%).

The results of the effect of ball mass to sample mass ratio are presented in Fig. 1. The response increased gradually to a maximum of 97.474% at a ratio of 10:1. Production of fines was also highest at this ratio. The gradual increase in grinding output reflects increase in impact load and increase in surface area of media in contact with sample. The drop in output above the optimum was likely due to the increase in ball to ball collision which resulted in a decrease in effective impact energy available for grinding. A second reason was overriding effect of ball to ball contact above the increase in sample to ball contact. 120 100

% passing 355μm

80 60 % passing 50μm

40 20 0 0

5

10

15

Ball mass to ore mass ratio.

Figure 1: The effects of ball mass to sample mass ratio on ball milling output of Baban Tsauni ore. www.iosrjournals.org

83 | Page

Effects of Ball Milling Conditions on Breakage Response of Baban Tsauni (Nigeria) Lead-Gold Ore. Fig.1 shows that the rate of increase in the production of particles of desired size (upper curve) was higher than that of undesirable particles (lower curve). The generation of fines was almost non-sensitive to change in media mass to sample mass ratio at values below 8:1 as the curve in this region was almost horizontal.

3.2 The effects of holding time on ball milling output The results of the effects of holding time on ball milling output are presented in Fig. 2. As holding time was increased from two minutes to ten minutes the per cent mass passing the liberation size increased from 21.07% to 69% while production of fines increased from 5.58% to 21.9%.

Cumulative per cent mass passing.

80 70 60

% passing 355μm

50 2

y = -0.4939x + 11.839x - 0.4238 40 30 20 10

y = 2.0376x + 1.5066

0 0

2

4

6

8

1250μm % passing

10

Grinding tim e (m inutes).

Figure 2: The effect of holding time on ball milling output. To limit the production of fines to 10% for example the batch milling time must be limited to four minutes. The implication for closed circuit milling is that the classifier or the particle sizing system must be designed to remove materials finer than 355microns at the rate of 9.81 per cent of the feed load per minute. Fig.2 shows that the rate of production of fine was low when compared to the per cent passing the liberation size. At 8minutes for example, the rate of fine generation was 2.0376%/minute (slope of line) while that of particles passing liberation size (gradient of curve at 8minute) was 3.94%/minute. The upper curve in Fig. 2 show that the gradient decreased with time. This implied that the breakage per unit time decreased with increase in holding time. Fig. 3 was obtained by calculating rates of production of particles passing the liberation size. It show that breakage per unit time dropped from 9.86%/minute (at 2minutes) to merely 1.96%/minute at a holding time of 10minutes. The implication of this result is that long time grinding circuit is an energy wasting process.

Figure 3: The rate of production of particles passing liberation size. www.iosrjournals.org

84 | Page

Effects of Ball Milling Conditions on Breakage Response of Baban Tsauni (Nigeria) Lead-Gold Ore. 3.3 The effect of mill speed on breakage response The results on the effect of mill speed on grinding response are presented in Fig. 4. Mill speed was expressed in terms of critical speed (in line with common practice) by applying (1).

Per cent mass passing 355microns.

90 80 70 60 50 40 30

y = -273.71x 2 + 453.76x - 110.42

20 10 0 0

0.5

1

1.5

Speed to critical speed ratio

Figure 4: The effect of speed on ball milling output of Baban Tsauni ore. The graph showed that grinding output increased with grinding speed up to a maximum value and the relationship was expressed by (2).

y  273.71x 2  453.76 x  110.42

2 The optimum grinding speed was obtained by differentiating (2) and equating to zero. This gave a value of 0.829, which implied that the grinding output increased with speed to a maximum of 77.642% passing the liberation size at a speed of 0.829Vc and this pattern agreed with submissions in literatures [20], [24]. The output dropped to 51% at a mill speed of 1.14Vc. Grinding is not expected to take place at the critical speed nor above it. Literature [20] indicates that mills have to be rotated at speeds below the critical speed for grinding to occur. The expression for critical speed (1) was derived by considering the motion of one ball [20], [3]. In a typical ball mill the position vector of the various balls from the centre differ. Thus some of the balls lost their grinding effect at the calculated critical speed of the mill. This implied that the critical speed for the second layer of balls was higher than the theoretical value which was based on the diameter of the mill. This was why grinding output did not drop to zero above critical speed as expected. However the result still agreed with the general rule that optimum grinding takes place at some speed below the critical speed and was found to be higher than 0.7Vc which is commonly used [20], [3], [24]. The implication of this result is that the optimum milling speed of each material needs to be determined rather than just operating mills below their critical speeds.

IV.

Conclusion

The results presented in this work showed that ball milling conditions affect the breakage response considerably. Production of particles of desired size increased to an optimum value with increase in the ratio of media mass to sample mass. The grinding response decreased with increase in holding time and the quality measured in term of fines generated also decreased with increase in milling time. The milling output attained an optimum value at an optimum speed of 0.829Vc. Optimum ball milling response can only be attained with optimum grinding speed, optimum media mass to sample mass ratio and short grinding circuit.

References [1] [2] [3] [4] [5] [6]

Levin, J., Indicators of grindability and grinding efficiency. J. S. Afr. Inst. Min. Metall., 92 (10), 1992, 283-290, Deniz, V. Relationships Between Bond's Grindability (Gbg) and Breakage Parameters of Grinding Kinetic on Limestone. Proceedings of 18th International Mining Congress and Exhibition of Turkey-IMCET, 2003, 451-456. Wills, B.A. and Napier-Munn, T. Wills’ Mineral Processing Technology-An Introduction to the Practical Aspects of Ore Treatment and Mineral Recovery. Seventh Edition. (Amsterdam: Elsevier Science and Technology Books), 2006, P450. Doll, A., and Barratt, D. Grinding: Why So Many Tests? 43rd Annual Meeting of the, Canadian Mineral Processors, Ottawa, Ontario, Canada. January 18 to 20, 2011. Man, Y.T. Technical Note: Why is the Bond Ball Mill Grindability Test done the way it is done? The European Journal of Mineral Processing and Environmental Protection 2(1), 1303-0868, 2002, 34-39. Aksani, B., and Sönmez, B. Simulation of Bond Grindability Test by Using Cumulative Based Kinetic Model, Minerals Engineering, 13 (6), 2000, 673.

www.iosrjournals.org

85 | Page

Effects of Ball Milling Conditions on Breakage Response of Baban Tsauni (Nigeria) Lead-Gold Ore. [7] [8] [9] [10] [11] [12]

[13]

[14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24]

Deniz, V., and Ozdag, H., ―A New Approach to Bond Grindability and Work Index: Dynamic Elastic Parameters‖, Minerals Engineering, 16(3), 2003, 211-217. Ozkahraman, H.T..A Meaningful Expression Between Bond Work Index, Grindability Index and Friability Value, Minerals Engineering, 18(10), 2005, 1057-1059. Partyka, T., and Yan, D., Fine Grinding in a Horizontal Ball Mill, Minerals Engineering, 20(4), 2007, 320-326. Ścieszka S., .Grzegorzek, W. and Żołnierz, M. Laboratory methods for combined testing of abrasiveness, grindability and wear in coal processing systems, Journal of Scientific Problems of Machines Operation and Maintenance, 4 (164), 2010, 19-35. Muhammad, I.A., Abdul, G.P., and Abdul, H.M., Work index and grinding energy assessment of Dilband iron ore, Pakistan, Mehran University Research Journal of Engineering & Technology, 30(1), 2011, 29-34. Magdalinovic, N., Magdalinovic, S, Jovanovic, R. and Stanujkic, D., Is there a material constant which characterises mineral resources grindability? Annual of the University of Mining and Geology “St. Ivan Rilski”, 54(Part II), Mining and Mineral processing, 2011, 116-117 Egbe, E.A.P. Characterization and Beneficiation of Baban Tsauni, Gwagwalada Ore, Nigeria, Thesis Submitted in Partial Fulfilment for the Award of the Degree of Doctor of Philosophy, Department of Mechanical Engineering, School of Engineering and Engineering Technology, Federal University of Technology, Minna, Nigeria, 2012. Fuerstenau, D.W., Lutch, J.J. and De, A., The effect of ball size on the energy efficiency of hybrid high-pressure roll mill/ball mill grinding, Powder Technol. 105, 1999, 199–204. Kotake, N., Daibo, K., Yamamoto, T., Kanda, Y., , Experimental investigation on a grinding rate constant of solid materials by a ball mill—effect of ball diameter and feed size, Powder Technol. 143– 144, 2004 196– 203. Trumic, M., Magdalinovic, N., Trumic, G., , The model for optimal charge in the ball mill, Journal of Mining and Metallurgy A( 43), 2007,19–32. Katubilwa, F. M. and Moys, M.H., , Effect of ball size distribution on milling rate, Miner. Eng. 22, 2009, 1283–1288 Erdem, A.S. and Ergun, S.L. The effect of ball size on breakage rate parameter in a pilot scale ball mill, Miner. Eng. 22, 2009, 660– 664 Magdalinovic, N., Trumic, M., Trumic, M. and Andric, L. The optimal ball diameter in a mill, Journal of Physicochemical Problems of Mineral Processing, 48(2), 2012, 329-339. Gupta, A. and Yan, D.S. Mineral Processing Design and Operation- An Introduction. Amsterdam, Elsevier B.V., 2006, P693. Sahoo, A. and Roy, G. K. Correlations for the Grindability of the Ball Mill As a Measure of Its Performance, Asia Pac. J. of Chem. Engineering, 3 (2), 2008, 230-235. Sahyoun, C., Kingman, S.W. and Rowson, N.A., High powered microwave treatment of carbonate copper ore, The European Journal of Mineral Processing and Environmental Protection 4(3), 2004, 175-182. Ipek , H., Goktepe, F. Determination of grindability characteristics of zeolite, Physicochem. Probl. Miner. Process. 47, 2011, 183192 Damisa, E.O.A. Beneficiation Potentials of Nahuta Lead Deposit, Bauchi State, Nigeria, Thesis Submitted in Partial Fulfilment for the Award of the Degree of Doctor of Philosophy, Department of Metallurgical Engineering ,Faculty of Engineering, Ahmadu Bello University, Zaria, Nigeria, 2008.

www.iosrjournals.org

86 | Page

IOSR Journal of Mechanical and Civil Engineering (IOSR-JMCE) e-ISSN: 2278-1684,p-ISSN: 2320-334X, Volume 6, Issue 3 (May. - Jun. 2013), PP 87-94 www.iosrjournals.org

Effect of Lime Stabilisation on the Strength and Microstructure of Clay 1

Asma Muhmed and 2DariuszWanatowski

1

AsmaMuhmedis with the University of Nottingham, United Kingdom, on leave from Omar AlMukhtarUniversity, Libya 2 DariuszWanatowski is with the University ofNottingham, China

Abstract:Lime stabilization is one of the techniques that can be used for improving the engineering properties, particularly the strength, of soft clays. This paper aims to investigate the effect of hydrated lime on the strength and microstructure of lime treated clays. In order to illustrate such effect, a series of laboratory tests were conducted. Atterberg limits, compaction tests, unconfined compressive strength tests and scanning electron microscope (SEM) were carried out on kaolin clay mixed with 5% hydrated lime. The results indicated that the addition of lime resulted in a reduction in the plasticity of kaolin and an improvement in compaction properties. The unconfined compressive strength (UCS) of stabilized clay experienced an increase with lime addition. Two variables influencing the amount of strength developed were studied. These variables included curing time and water content. Curing time contributed to an increase in the UCS, from 183 kPa to 390 kPa, that is approximately twice of the strength of untreated kaolin. SEM analysis showed the presence of the cementious products in the kaolin clay resulted from lime-clay reaction. Key words:Lime, Stabilisation, Soft clay, Pozzolanic reaction, Mineralogy

I.

Introduction

Clay soils are normally associated with volumetric changes when subjected to changes in water content because of the seasonal water fluctuations. Furthermore, problems with clays, including low strength and high compressibility, can cause severe damage to civil engineering construction. Therefore, these soils must be treated before commencing the construction operation to achieve desired properties. Different methods are available to improve the engineering properties of problematic soils such as densification, chemical stabilisation, reinforcement and techniques of pore water pressure reduction. The chemical stabilisation of clays using lime is one of the commonest methods that can be used to upgrade the soils of poor properties to provide a workable platform to construction projects [1]. According to [2], the need for developing new materials for stabilisation started in the late 1940s. The chemical reactions between clay particles and lime can be categorised into two forms of improvement, short term reaction (modification) and long term reaction (stabilisation). In the first reaction the process of ion exchanges makes the clay minerals flocculate and agglomerate leading to a reduction in plasticity, swell and moisture content. The second reaction (pozzolanic reaction) accomplishes over a period of time creating cementing products that cause long-term strength gain [3]. This paper aims to investigate the strength and the microstructure of lime treated clay using unconfined compressive strength test and scanning electron microscope (SEM) analysis. For this purpose, Eedes and Grim's method was used to determine the required amount of hydrated lime to be mixed with commercial kaolin. Kaolin-lime mixtures were prepared with 5% lime and compacted at optimum moisture content (OMC), and at wet and dry of the OMC in order to investigate the influence of curing time required and water content needed for successful lime treatment.

II.

Experimental Program

In order to investiggate the effect of hydrated lime on the physical and engineering properties, both untreated and treated kaolin were subjected to similar laboratory tests. These tests include Atterberg limits test, compaction test, unconfined compressive strength test (UCS) and scanning electron microscope (SEM). The effect of curing time and water content on the UCS and microstructure of lime treated clay was also studied. The first step for stabilisation with lime in the present study was to determine the minimum amount of lime required for long-term improvement of strength. Test Materials The soil tested in this study is commercial kaolin clay. It was mixed with 5% lime (determined based on Initial Consumption of Lime). A fine ground hydrated lime (calcium hydroxide – Ca[OH]2) was used. Table 1 summarises the basic properties of untreated kaolin. www.iosrjournals.org

87 | Page

Effect Of Lime Stabilisation On The Strength And Microstructure Of Clay Table 1: Basic properties of untreated kaolin Value

Property

Specific Gravity Liquid Limit, % Plastic Limit, % Plasticity Index,% Maximum Dry Density, kg/m3 Optimum Moisture Content, % Unconfined Compressive Strength, kN/m2 pH

2.61 65.9 33.3 32.6 1422 29.9 183 5.43

Specimen Preparation The clay samples were firstly dried at 60oC before mixing. To prepare the lime-clay mixtures, the required amounts of dry kaolin were mixed with 5% of hydrated lime by weight of kaolin in the dry state until an even distribution of lime in mixture was obtained. The dry lime-clay mixtures were then mixed with distilled water. These mixtures were used for Attreberg limit and compaction tests. The moisture contents chosen for the present study corresponded to the optimum moisture content (OMC), wet side of OMC (WMC) and dry side of the OMC (DMC) of the treated and untreated kaolin determined from the compaction tests. These moisture contents were used for preparing the specimens for the unconfined compressive strength test.

pH

Test Results Initial Consumption of Lime (ICL) The pH test is the preferred method for identifying the lime content required to obtain a long-term pozzolanic reaction. In order to determine the minimum lime content required, the method developed by [4] was used. The plot of pH valuescorrected at 250C versus lime content are presented in Fig. 1. It can be seen that, the initial lime content needed for the lime treatment of clay is 2%. Therefore, this percentage of hydrated lime could be used. However, according to [5], for optimum improvement of the soil strength the optimum lime content for kaolin is in the range of 4% to 6%. Therefore, in the current study, 5% of hydrated lime by weight was chosen for thestabilization. 14 13 12 11 10 9 8 7 6 5 4 0

1

2

3

4

5

6

Lime content (%) Fig. 1: Determination of the initial consumption of lime

Atterberg Limit Test Attreberg limits were determined according to [6]. The liquid limit (LL), in the penetrometer method, of the soil is the moisture content at which an 80 g, 30o cone sinks exactly 20 mm into a cup of remoulded soil in a 5 seconds period. At this moisture content, the soil will be very soft. The plastic limit (PL) is the water content in which soil starts to crack when rolled between the fingers on glass plate to form a thread of 3 mm diameter. At this point, the soil has a stiff consistency. The difference between LL and PL is plasticity index.

www.iosrjournals.org

88 | Page

Effect Of Lime Stabilisation On The Strength And Microstructure Of Clay Proctor’s Compaction Test The optimum moisture content (OMC) and the maximum dry density (MDD) of the clay were obtained using the Standard Proctor compaction test carried out according to [6]. The test was performed by the compaction of the clay into a mould that is filled at a fixed moisture content in three, approximately, equal layers of clay. The compaction is achieved by a standard number of blows (27 blows) from a hand rammer of 2.5 kg dropping from a height of 300 mm. Unconfined Compression Strength (UCS) Test The UCS testing was performed according to [6]. For this test, samples of kaolin and lime treated kaolin were compacted at optimum moisture content (OMC) and at wet and dry side of OMC by hand in three layers in cylindrical mould of 38 mm in diameter and 76 mm in height. All specimens were taken out from the moulds and cured at 20oC for 7, 14 and 28 days. After reaching the specified curing time, the cylindrical specimens were subjected to a gradually increased axial compression load until failure. All the specimens were tested at an axial strain rate of 1% per minute. Scanning Electron Microscope In order to observe the influence of lime addition on the microstructure of the studied clay, Scanning Electron Microscope (SEM) analysis was used to investigate the microstructural development of the tested kaolin. After UCS testing, the tested samples were dried for SEM. Small specimens of both untreated and treated clay samples were prepared and the fractured surfaces of the specimens were coated with gold before scanning.

III.

Discussion Of Test Results

Influence of Lime on Atterberg Limits Fig. 2 shows Atterberg limits for 0% and 5% of lime. It can be observed from Fig. 2, there is a reduction in the Plasticity Index and an increase in both liquid and plastic limits with the addition of lime. The liquid limit increased by 20.6% when 5% of lime was added. The increase is more evident in the plastic limit which increased from 33.3% to 56.9%. According to [7] and [8], the reduction in the plasticity makes soils more friable and easily workable. Plasticity decreases because calcium ions from the lime reduce the volumetric changes. Furthermore, the aggregates, formed from the flocculation of the soil particles, behave more like silt rather than clay particles. Another study, conducted by [9], also offers evidence in support of the increase of the liquid limit. These investigations showed that the liquid limits of kaolinite rich soils increased when lime was added. However, liquid limit change is less important, for geotechnical engineers comparing with the plastic limit; this is because the maximum soil strength is attained when the soil is compacted at OMC that is close to the plastic limit as proved in this project. 86.5

5% lime 65.9

56.9

33.3

LL

0% lime

PL

29.6 32.6

PI

Fig. 2:Atterberg limits at 0% and 5% lime content

Influence of Lime on Compaction Characteristics Fig. 3 illustrates the moisture-density relationship of kaolin with 0% and 5% lime. As shown in the Fig. 3, the addition of lime tends to increase the optimum moisture content (OMC) and reduce the maximum dry density (MDD). For the same compaction effort, the optimum moisture content of the untreated clay changed from 29.9% to 33.3% when 5% lime was added to the clay.The maximum density decreased from 1422 kg/m3 to 1382 kg/m3. This change is considered as an indication of the improvement of the compaction characteristics of the lime stabilised clay. This is consistent with the study conducted by [10]. The reduction in the dry density www.iosrjournals.org

89 | Page

Effect Of Lime Stabilisation On The Strength And Microstructure Of Clay occurs because the agglomerated and flocculated particles of soil occupy larger spaces and the reason for increasing OMC is that, the lime requires more water for the pozzolanic reactions. [11] referred the decrease in density to the difference in the specific gravity of the clay and hydrated lime. 1500 0% lime

1450 Dry density, kg/m3

5% lime 1400 1350 1300 1250 45.0

40.0

35.0

30.0

25.0

20.0

Moisture content, % Fig. 3:Influence of lime on compaction characteristics of clay

Influence of Lime on Unconfined Compressive Strength (UCS) Fig.4 shows the stress-strain behaviour of untreated kaolin and kaolin treated with 5% lime and compacted at OMC. As can be seen, the untreated clay failed at the UCS of 183 kPa, whereas treated clay at different curing time failed at higher values of the UCS.

0% lime, 0 days 5% lime,0 days 5% lime, 7 days 5% lime, 14 days 5% lime, 28 days

400

Axial stress (kpa)

350 300 250 200 150 100 50 0

0

1

2

3

4

5

6

Strain (%) Fig. 4: Unconfined compression test curves

Figure 5 presents the relationship between the UCS and the curing time. As can be seen from Fig. 5, the limetreated soil exhibited a trend of increasing UCS with increasing curing time. For example, the UCS increased about twice of the original strength after 28 days curing. The summary of the UCS values is given in Table 2.

www.iosrjournals.org

90 | Page

Unconfined Compressive Stregth (kPa)

Effect Of Lime Stabilisation On The Strength And Microstructure Of Clay 450 430 410 390 370 350 330 310 290 270 250 230 210 190 170

Kaolin + 5% lime Kaolin

0

5

10 15 20 Curing Time (Days)

25

30

Fig.5: Influence of curing time on UCS of treated kaolin Table 2: Summary of the unconfined compressive strengths Description Kaolin Kaolin + 5% lime

Curing period (days) 0 0 7 14 28

UCS (kPa) 183 206.2 224.39 244.67 390.1

Fig. 6 presents the results of the UCS tests in terms of water content. The lime stabilised specimens at curing periods of 14 and 28 days achieved maximum strengths at optimum moisture content (OMC=33.3%) whereas for the wet side (WMC = 37.5%) and the dry side (DMC = 29.9%) of the OMC the UCS decreased for the same periods, even though they were still higher than the UCS of the untreated clay. The stabilised clay cured for 14 days experienced an increase from 189 kPa to 205.2 kPaonWMC, while at 28 days of curing the increase was from 252.25 kPa to 344.6 kPaonDMC. In other words, higher or lower than OMC water content decreases the UCS values compared to UCS at OMC. Similarly, [12] stated that the strength remains low with high water contents. The UCS decreases with the increasing water content because high water content prevents lime-clay contact; therefore, the bonding material created by the reaction between lime and clay particles is limited. This is on contrary to the priciples of the reaction required in lime stabilisation, where the chemical reaction needs more water to produce the pozzolanic materials.

390.1 268.73

344.6 252.25 189 205.2

OMC

14 DMC WMC

28

Fig. 6: Influence of water content on UCS of treated kaolin clay

Microstructure Analysis SEM analysis was carried out on the same samples tested in unconfined compression tests. Lime treatment changed significantly the soil fabric depending on curing time and water content. Fig. 7 shows the SEM-micrograph of untreated kaolin clay. As can be seen from the figure, hexagonal particles were observed and the clay displayed flaky texture which almost disappeared after 5% addition of hydrated lime. www.iosrjournals.org

91 | Page

Effect Of Lime Stabilisation On The Strength And Microstructure Of Clay The images of lime treated kaolin clay cured for 7, 14 and 28 days and compacted at moisture content of 33.3% (OMC) are shown in figures 8, 9 and 10 respectively. As observed from the Figures, the clay particles transformed from a flaky form into a flocculating structure (Fig.8). With increasing curing period, there were morphology changes at the edges of clay particles and sufficient cementing compounds due to pozzolanic reaction coated and joined the soil particles together making the voids in the soil less distinct (Fig. 10). Similar findings were reported by [13] who found that new compounds in the form of white lumps were observed for longer curing time as a result of adding 5% lime. It is expected that the new reaction products are fully formed after 28 days of curing.

Fig. 7: SEM (OMC=29.9%)

image

of

untreated

kaolin

clay

Fig. 9: SEM image of 5% lime treated kaolin clay (OMC=33.3%, 14 days)

Fig. 8: SEM image of 5% lime treated kaolin clay (OMC=33.3%, 7 days)

Fig. 10: SEM image of 5% lime treated kaolin clay (OMC=33.3%, 28 days)

In terms of water content, at the same compactive effort the clay particles became more oriented with increasing water content (Fig. 11). On wet side of OMC (37.5%), increasing water content led to a more homogeneous microstructure (Fig. 12). More parallel particles and less voids are observed in Fig. 13 after 14 curing days compared to the lime treated soil of OMC at the same period. It is clearly noticed that the SEM image of the treated sample compacted at WMC and cured for 28 days has more voids than that compacted at OMC and cured for the same time (Fig.14). Therefore, the reason for decreasing the unconfined compressive strength with increasing water content is related to these voids which are not filled with cementing products. The bonding of clay particles due to the pozzolanic reaction was limited by increasing water content. The microstructural analysis was in agreement with the results of strength tests. www.iosrjournals.org

92 | Page

Effect Of Lime Stabilisation On The Strength And Microstructure Of Clay

Fig. 11: SEM image of untreated kaolin clay (WMC=33.3%)

Fig. 12: SEM image of 5% lime treated kaolin clay (WMC=37.5%, 7 days)

Fig. 13: SEM image of 5% lime treated kaolin clay (WMC=37.5%, 14 days)

Fig. 14: SEM image of 5% lime treated kaolin clay (WMC=37.5%, 28 days)

IV.

Conclusion

Based on the investigations carried out in this study, the following conclusions can be drawn:  The addition of lime to the tested soil is found to increase the plastic and the liquid limits and decrease plasticity index. The plastic and liquid limits increased by 23.6% and 20.6% respectively, and the plasticity index decreased by 3%.  Lime treatment was able to increase the optimum moisture content from 29.9% to 33.3%. In addition, the maximum dry density reduced by 40 kg/m3with lime addition. Therefore, lime content of 5% improved the compaction properties of the clay.  The unconfined compressive strength of lime treated samples increased with increasing the curing period particularly after 28 days curing while it decreased with increasing water content. The added lime led to increase of around twice of the strength of the untreated clay. Contrary to curing time, water content, lower or higher than the OMC, decreased the strength of clay. These strength measurements correlate well with the laboratory data obtained by other researchers. www.iosrjournals.org

93 | Page

Effect Of Lime Stabilisation On The Strength And Microstructure Of Clay 



The behaviour and the fabric of treated soils shown by the SEM-micrographs indicated the formation of cementitous materials. These products resulted from the pozzolanic reactions contributed to the increase of the strength of stabilized clay. On the OMC, the microstructure of lime treated clay characterized by an assembly of clay materials whereas on the wet side of OMC, increasing water content led to a homogeneous microstructure. From the results obtained in this study, it can be concluded that the hydrated lime effectively improve the strength, plasticity and compaction properties of kaolin clay.

References [1] [2] [3] [4] [5] [6] [7]

[8] [9] [10] [11] [12] [13]

K. A. Gutschick(1967) Lime Stabilizes Poor Soils, the Aberdeen Group (Available at: http://www.concreteconstruction.net/concrete-articles/lime-stabilizes-poor-soils.aspx). T.Brandon, J.Brown, W.Daniels, T.DeFazio,G.Filz,J.Musselman, C. Forshaand,J. Mitchell (2009) 'Rapid Stabilisation/ Polymerization of Wet Clay soils; Literature Review', Air Force Research Laboratory, Blacksburg. D. N. Little (1999) Evaluation of Structural Properties of Lime stabilised soils and aggregates. The national lime association, Vol. 1. J.L. Eadesand R.E. Grim(1966) A Quick Test to Determine Lime Requirements of Lime Stabilisation, Highway Research Record 139, pp. 61-72. J.L. Eadesand R.E. Grim(1960) The Reaction of Hydrated Lime with Pure Clay Minerals in Soil Stabilization, Highway Research Board 262, pp. 51–63. British Standard 1377, “Methods of test for soils for civil engineering purposes,” British Standard Institution, London, 1990. A. Muzahimand L. Abdelmadjid(2008) Effect of Hydrated Lime on the Engineering Behaviour and the Microstructure of Highly Expansive Clay. In: The 12th International Conference of International Association for Computer Methods and Advances in Geomechanics, India, 1-6 October 2oo8. B. R. Phanikumar(2009) Effect of lime and fly ash on swell, consolidation and shear strength characteristics of expansive clays: a comparative study, Journal of Geomechanics and Geoengineering, 4(2), 175-181. F. G. Bell(1989) Lime Stabilization of Clay Soils, Bulletin of Engineering Geology and the Environment, 39(1), 67-74. T.S.Umesha, S.V. Dineshand P.V. Sivapullaiah(2009) Control of dispersivity of soil using lime and cement.International journal of geology.3(1), 8-15. D. T.Davidson, G. Nogueraand J. B. Sheeler (1960) Powder versus Slurry Application of Lime for Soil Stabilization. In: Papers on Soil 1959 Meetings, American Society for Testing Material, pp. 244-253. F. G. Bell (1993) Engineering Treatment of Soils, E & FN Spon, London. K. A. Kassim(2009) The Nanostructure Study on the Mechanism of Lime Stabilised Soil, Research Vot No: 78011, Department of Geotechnics and Transportation, UniversitiTeknologi Malaysia.

www.iosrjournals.org

94 | Page