Double Skinned Steel Tubular (DSST)

7/26/2019 Double Skinned Steel Tubular (DSST)

1/68

EXPERIMENTAL INVESTIGATION ON DOUBLE SKINNED STEEL

TUBULAR (DSST) COLUMNS SUBJECTED TO MONOTONIC LOADING

DISSERTATION

Submitted to

Visvesvaraya Technological University, Belgaum

In Partial Fulfillment of the Requirement for the Award of the Degree of

MASTER OF TECHNOLOGY

IN

STRUCTURAL ENGINEERING

By

DARSHAN.M.K

(USN: 1GC11CSE01)

Under the Guidance ofDr.N.S.KUMARProfessor & Director (R & D)

Dept of Civil Engineering, G.C.E,

Ramanagaram-571511

DEPARTMENT OF CIVIL ENGINEERING

GHOUSIA COLLEGE OF ENGINEERING

RAMANAGARAM-571511

2012-2013


2/68


3/68


4/68

ACKNOWLEDGEMENT

The satisfaction and euphoria that accompanies the successful completion of any task would be

incomplete without mentioning the people who made it possible.

I take this opportunity to convey my deep sense of gratitude to all those who have been kind

enough to offer their advice and provide assistance when needed which has lead to the successful

completion of the project.

I would like to thank sincerely, my project guide Dr. N.S.KUMAR, Professor & Director,

Department of Civil Engineering, and Ghousia College of Engineering Ramanagaram for his

valuable timely guidance, inspiration and continuous supervision during the entire course of this

project work, and for successful completion of the same on time.

I would like to express our deep sense of gratitude and indebtedness to Dr. MOHAMED

ILYAS ANJUM, Vice-principal, Prof. & Head, Department of Civil Engineering, Ghousia

College of Engineering Ramanagaram for his constant encouragement, guidance and inspiration

which enabled us to complete this project work.

I would like to thank our principal Dr. MOHAMED HANEEF, Ghousia College of

Engineering, Ramanagaram, for his support and inspiration.

I thank all the TEACHING STAFF, SUPPORTING STAFFwho have directly or indirectly

helped us in successful completion of our project work.

DARSHAN.M.K


5/68

ABSTRACT

Columns occupy a vital place in the Structural system. Weakness or failure of a Column

destabilizes the entire Structure. Strength and Ductility of Steel columns need to be ensured

through adequate strengthening, repair and rehabilitation techniques to maintain adequate

structural performance. Recently composite column are finding a lot of usage for seismic

resistant. In order to prevent shear failure of RC column resulting in storey collapse of buildings,

it is essential to make ductility of column larger. Recently most of the buildings utilize this

DSST concept as primary for lateral load resisting frames. The mortar used for encasing the steel

section not only enhances its strength and stiffness, but also protects it from fire damages.

In this Dissertation, Experimental work Analysis of cement mortar-steel double-skin

tubular member is carried out, with the emphasis being on its potential as key lateral and verticalload-resisting members in structures located in seismically active regions. In this new structural

member, the two constituent materials are optimally combined: the outer and inner tube is made

of steel, and the space in-between is filled with cement mortar. These members are highly useful

when they are used as columns. These members are monotonically loaded to their ultimate load

to study the behavior of DSST under increasing L/D ratio and keeping Thickness constant.

Hence, this member can be referred to Double-Skin Tubular Column (DSTC), In this

experimental programme, 57 samples have been tested.

Here, an attempt is made to study the strength of totally 57 specimens of the following

three models. As per IS: 2250-1981 (Reaffirmed 1990)-Third reprint, February 1993-Indian

Standard code of practice for preparation and use of masonry mortars, for masonry in

buildings subject to vibration of machinery, the grade of mortar shall not be less than MM 3

(Clause 7.1.4). Hence, in this dissertation work mortar grade of 1:3, 1:4 and 1:5 has been

selected as infill to fill the gap between outer and inner tubes. i.e. Double Skinned Tubes. Each

model is of three sets. One set consists of 1:3 ratio Mortar and remaining two consists of 1:4 and

1:5 ratio Mortar respectively which is filled in between the gaps of steel tubes of different

lengths and thickness. The lengths of the specimens used for study are 350mm, 450mm and

550mm of varying thickness 2.6mm, 3.2mm and 4.0mm.


6/68

Results have been analyzed using most recent soft tool Artificial Neural Networks

[ANN].The results obtained by experiment are validated using ANN model and the errors

corresponding to the obtained practical and analytical values are tabulated and concluded.


7/68

CONTENTS

CHAPTERS PAGE NO.

1. INTRODUCTION 1-3

1.1: General

1.2: Comparison of Different Types of Composite Columns

1.3: Ductility and Energy Dissipation Capacity

2. LITERATURE REVIEW 4-7

3. AIM AND SCOPE OF STUDY 8

4. EXPERIMENTAL PROGRAM 9-22

4.1: Preparation of Specimen

4.2: Experimental Study

4.3: Strain Gauge

4.4: Test Procedure by Using SCADA Software

4.5: Loading Scheme

5. TYPICAL RESULTS AND DISCUSSIONS 23-39

5.1: Result for Sample 1

5.2: Result for Sample 2

5.3: Tabulation of results of specimens.

5.4: Plots for Grade V/S Load

5.5: Plots for thickness v/s load

5.6: Plots for L/D V/S Load

5.7: Mathematical modeling

6. ANALYTICAL STUDIES USING ARTIFICIAL

NEURAL NETWORKS 40-51

6.1: Introduction

6.2: Artificial Neural Network

6.3. Work Flow

6.4. Prediction and Experimental Results


8/68

7. SUMMARY AND CONCLUSIONS 52

8. RECOMMENDATIONS FOR FURTHER STUDIES 53

9. REFERENCES 54

10. JOURNAL PUBLICATION 57


9/68

LIST OF SYMBOLS

AsSteel cross-sectional area

Am

Mortar cross-sectional area

DSST Double Skinned Steel Tube

D Diameter of circular steel tube

EmMortar modulus of elasticity

EsSteel modulus of elasticity

fyYield strength of steel

fcMortarcube strength

L Effective buckling length of column

PuUltimate axially compressive load

PuthePredicted ultimate axially compressive load

Axial Strain

Ponominal strength

A total Total cross-sectional area

Ductility


10/68

Experimental Investigation on Double Skinned Steel Tubular (DSST) Columns Subjected to Monotonic loading

Dept. of Civil Engg, GCE, Ramanagaram Page 1

CHAPTER-1

INTRODUCTION

1.1: GENERAL

DSST (Double Skinned Steel Tubes) Columns came into existence during early 1960.

Substantial research has been made to understand the behavior since then. The advantage of

using these DSST Columns have been found by Japanese first and employed in the

construction of multi-storied buildings effectively. Now, the analysis and Design of these

DSST Columns have found place even in Codes and Specifications. It has been envisaged to

study strength, stiffness and buckling characteristics by providing flutes to steel tube of

columns which enhances aesthesis of columns. Also, a fluted column enhances the strength

and also stiffness as the surface area of steel sheet and moment of inertia of the column

increases. The advantage of steel members having high tensile strength and ductility and

concrete members having better compressive strength have been better made use as a

composite member. Hence, it has been envisaged to check whether such a columns would act

as a slender column. Research has been in progress around the world on experimental and

analytical studies on double skinned Steel Tubular Columns for more than four decades.

Substantial contribution has been made since then in understanding the behavior of DSST

columns and to arrive at a design procedure. Quite few countries have incorporated the

design procedure in their respective codes also. Most of the researchers have considered the

contribution of geometric properties like shape, L/D ratio, t/D ratio, boundary conditions,

strength of materials and the loading conditions. It has been found that generally the failure

occurs by either local buckling or yield failure. It has been found that Euro code gives a better

design method which yields values nearer to experimental values.

Columns are considered as critical members in moment-resisting structural systems.

Their failure may lead to a partial or even a total collapse of the whole structure. Therefore, it

is important to improve the ductile deformation capacity and energy dissipation capacity of

columns so that the entire structure can endure severe ground motions and dissipate a

considerable amount of seismic energy. In recent years, double skinned steel tubes (DSSTs)

have become increasingly popular as columns in braced and unbraced frames, as they have

the advantages of ductile behavior as a result of confinement to concrete by the steel tube and

delayed local buckling of the steel tube due to the support from concrete, improved damping


11/68



behavior in comparison to traditional steel frames, ease for construction as the steel tube

serves also as the permanent form, and a high strength-to-weight ratio. Double Skinned steel

tube (DSST) columns combine the advantages of ductility, generally associated with steel

structures, with the stiffness of a concrete structural system. The advantages of the concrete-

filled steel tube column over other composite systems include: The steel tube provides

formwork for the concrete, the concrete prolongs local buckling of the steel tube wall, the

tube prohibits excessive concrete spalling, and composite columns add significant stiffness to

a frame compared to more traditional steel frame construction. While many advantages exist,

the use of DSSTs in building construction has been limited, in part, to a lack of construction

experience, a lack of understanding of the design provisions and the complexity of

connection detailing. Consequently, a joint was needed that could utilize the favorable

strength and stiffness characteristics of the concrete-filled tube column yet be constructible.

The inner void reduces the column weight without significantly affecting the bending rigidity

of the section and allows the easy passage of service ducts but in this experiment cement

mortar has been used instead of concrete due to very less gap between the two tubes.

1.2: Comparison of Different Types of Composite Columns

1.2.1: Comparison of the Steel-mortar DSST and steel-concrete DSST

a) A more ductile response of cement as it is well confined by the steel tube which

does not buckle. The steel tube is designed to have predominantly high strength with its axial

stiffness being nearly zero; by doing so, local buckling of the tube due to axial compressive

stresses, which is a common problem for steel tubes, is unlikely to happen.

(b) No need for fire protection of the outer tubes as the outer tube is required only as a form

during construction and as a confining device and additional shear reinforcement during

earthquakes. The steel tube with negligible axial stiffness contributes little to the load

carrying capacity of the hybrid member and is not expected to affect the structural resistance

during a fire. However, the outer steel tube of a steel-concrete DSST Columns takes

considerable axial loading, and when its structural resistance is lost during a fire, the

structural safety of the column is considerably compromised.

(c) No need for corrosion protection as the steel tube inside is well protected by the concrete

and some coatings for inside steel tube.


12/68



1.2.2: Comparison of the FRP-Steel-mortar DSST and FRP-concrete DSST

(a) Ability to support construction loading through the use of the inner steel tube. A

steel tube is superior to an FRP tube in taking construction loading, as the latter is more

susceptible to a buckling failure.

(b) Ease for connection to beam due to the presence of the inner steel tube, which enables

existing connection forms to be directly used.

(c) Savings in fire protection cost as the outer tube is required only as a form during

construction and as a confining device and additional shear reinforcement during

earthquakes.

(d) Better confinement of the concrete as a result of the increased rigidity of the inner tube.

1.3: Ductility and Energy Dissipation Capacity

Under seismic attacks, the ductility and energy dissipation capacity of a column are

the major concerns. Confinement to concrete is an effective means of improving the ductility

of a column in which concrete is a main material. It has been demonstrated by extensive

research that concrete confined by a steel tube outside can exhibit much better ductility

compared with unconfined concrete, either under monotonic loading or cyclic loading.

Extensive research on steel-confined concrete has shown that steel tube confinement to

concrete can also significantly enhance the strength and strain capacity of concrete, although

the stress-strain behavior of steel-confined concrete shown below.

Plot 1.1: stress-strain curves


13/68



CHAPTER-2

LITERATURE REVIEW

T. Yu, Y.L. Wong and J.G. Ten[1]

Six DSTC specimens with three different configurations were prepared and tested

under concentric compression. The results are summarized below. The columns all had an

outer diameter of 152.5mm, a height of 305mm, and the same steel tube inside. They were

provided with GFRP tubes of different thicknesses outside, which had fibers only in the hoop

direction. Tensile tests on steel coupons were conducted. It was found that the steel tube had

a yield stress of 352.7MPa, an ultimate tensile strength of 380.4MPa and a Young s modulus

of 207.28GPa. The FRP tubes were prepared by the wet lay-up process; the FRP used had a

nominal thickness of 0.17mm per ply, a tensile strength of 2300MPa and a Youngs modulus

of 76GPa based on this nominal thickness according to the manufacturers data. The elastic

modulus, compressive strength and strain at peak stress of the concrete averaged from three

concrete cylinder tests (152.5mm x 305 mm) are 30.2 MPa, 39.6 MPa and 0.002628

respectively.

During the test, all specimens exhibited a smooth load-displacement curve until

failure took place, when the outer GFRP ruptured and the load began to drop. The test results

shows that, Pco is equal to the unconfined concrete strength times the area of the annular

concrete section (=543.5 kN), while Ps is equal to the average ultimate load from three axial

compression tests on hollow steel tubes (=273.8 kN). Therefore, the ultimate load of the

hybrid column is 817.3 kN if the constituent parts do not interact and the confinement effect

of the GFRP tube is negligible.

Based on the results of this study, the following conclusions were drawn within the

scope of these tests:

1) This new hybrid structural member possesses good ductility and good energy dissipation

capacity. When subjected to concentric compression, the concrete sandwiched between the

two tubes may achieve significant enhancement in both strength and ductility overunconfined concrete. According to Teng et al. (2004), the concrete in a typical hybrid DSTC

may be confined as effectively as that in an FRP-confined solid concrete cylinder.

2) The new hybrid member shows good ductility under four-point bending, although

significant cracks will occur early in the loading process. Longitudinal fibers may be required

in the outer GFRP tube if the new hybrid member is to be used to resist bending only. In


14/68



addition, there may be a need to improve the bond between the concrete and the steel tube,

such as through the use of mechanical shear connectors to prevent possible premature slips as

observed in one of the beam tests presented in the paper.

3) Further tests, including eccentric compression tests, combined axial and cyclic lateral

loading tests and shaking table tests, should be carried out in the future to develop a more

complete understanding of the seismic performance of the new hybrid member and structural

systems based on this new member form.

Min-Lang Lin and Keh-Chyuan Tsai[2]

The purpose of this experimental study is to investigate the behavior of the double-

skinned concrete filled steel tubular (DSCFT) columns on the strength, stiffness and ductility

performance. The diameter-thickness (D/t) ratio and the hollowness ratio were chosen as

main parameters in designing the specimens. A total of 18 specimens were tested under

varied combinations of axial and flexural loads, and two specimens were tested under a

combination of constant axial load and cyclically increasing bending for comparison. Test

results concluded that the DSCFT columns can effectively provide strength and deformation

capacity even with a large D/t ratio.

Following conclusions were drawn from the above experiment,

1. Superposing the concrete and steel strength can predict the ultimate axial strength of

DSCFT Conservatively. It is illustrated that steel tube can improve the confinement of the

concrete, and the in-filled concrete can delay the occurrence of local buckling of the steel

tube with a large D/T ratio.

2. The DSCFT columns can have an optimal strength performance if the applied axial load is

less than 40% axial capacity.

3. Experimental results indicate that the behavior of DSCFT columns under cyclic loading is

as good as that under the monotonic loading.

Tao Yu, Yu-Bo Cao, Bing Zhang[3]

In total, eight identical hybrid DSTCs were tested, covering four loading schemes;

two specimens were prepared for each loading scheme. The specimens had an outer diameter

(i.e. the outer diameter of the annular concrete section) of 205.3 mm, an inner diameter (i.e.

the inner diameter of the annular concrete section and the outer diameter of the inner steel

tube) of 140.3 mm, and a height of 400 mm. The outer glass FRP (GFRP) tube had fibers in


15/68



the hoop direction only and was formed by a wet-layup process on hardened concrete. The

nominal thickness of the two-ply FRP tube was 0.34 mm (i.e. the nominal thickness was

taken to 0.17 mm per ply) while the thickness of the steel tube was 5.3 mm.

This paper has presented a series of cyclic axial compression tests on hybrid DSTCs. Hybrid

DSTCs have been shown to be very ductile under cyclic loading and their envelope axial

load-strain curves are almost the same as the corresponding monotonic axial stress-strain

curve. It has also been shown that repeated unloading/reloading cycles have a cumulative

effect on the permanent strain and the stress deterioration of the confined concrete in hybrid

DSTCs. Interfacial slips between the steel tube and the concrete may lead to noticeable

differences in the axial strain between them when the column is fully unloaded from an axial

strain level that significantly exceeds the yield strain of the steel tube.

Lin-Hai Han, Fei-Yu Liao, Zhong Tao[4]

The authors performed a series of tests on the CFDST columns subjected to static

loading, including 37 specimens under axial compression, 13 specimens under bending and

42 specimens under eccentric compression, respectively (Han et al., 2004; Tao et al., 2004;

Tao and Han, 2006; Tao and Yu, 2006). It was found that the behaviour of the CFDST

columns is generally similar to that of the conventional CFST columns. This is owing to the

fact that, generally, the section slenderness ratio of an inner steel tube is relatively small and

it can provide sufficient support to the sandwiched concrete. Otherwise, the premature local

buckling of inner steel tubes will have adverse effects on the load-carrying and deformation

capacities of CFDST columns.

This paper briefly summarizes some recent research outcomes of CFDST members

presented by the authors and their collaborators. From the experimental and numerical

results, it can be concluded that, when the hollow ratio () of a CFDST is within the normal

range of 0-0.5, the CFDST generally demonstrates a similar behaviour as that of a CFST,

whilst the fire resistance of the CFDST is superior to that of the latter. Apart from the

research results reported in this paper, ongoing numerical study is being carried out to

analyze the post-fire behavior of CFDST columns. Repair approach will be further

recommended. The authors also believe that there is immediate research need to put forward

suitable beam-to-column connections for CFDST columns, in which the load can be

transferred and shared by the three components simultaneously.


16/68



Zhang, B., Teng, J. G. & Yu, T[5]

The experimental program consisted of 6 hybrid DSTCs. All these specimens had a

circular section with a characteristic diameter D (the outer diameter of the annular concrete

section) of 300 mm and a void ratio of 0.73 (the ratio between the inner diameter and the

outer diameter of the annular concrete section). The inner steel tube had thickness ts of 6 mm

and an outer diameter Ds of 219 mm, leading to a Ds/ts ratio of 36.5. The outer GFRP tube

had an inner diameter of 300 mm and a thickness tfrp of 6 mm or 10 mm. The height was

1350 mm from the point of lateral loading to the top of the stiff RC column footing (4.5 times

of the column diameter).

This paper has presented the results of 6 large-scale hybrid DSTCs with HSC tested

under axial compression in combination with cyclic lateral loading. These test results suggest

that hybrid DSTCs can still show excellent ductility and seismic resistance even when high

strength concrete with a cylinder compressive strength of around 120 MPa is used.

L.Lam and J.G. Teng.[6]

In total, eight identical hybrid DSTCs were tested, covering four loading schemes;

two specimens were prepared for each loading scheme. The specimens had an outer diameter

(i.e. the outer diameter of the annular concrete section) of 205.3 mm, an inner diameter (i.e.

the inner diameter of the annular concrete section and the outer diameter of the inner steel

tube) of 140.3 mm, and a height of 400 mm. The outer glass FRP (GFRP) tube had fibers in

the hoop direction only and was formed by a wet-layup process on hardened concrete [2].

The nominal thickness of the two-ply FRP tube was 0.34 mm (i.e. the nominal thickness was

taken to 0.17 mm per ply) while the thickness of the steel tube was 5.3 mm.

This paper has presented a series of cyclic axial compression tests on hybrid DSTCs.

Hybrid DSTCs have been shown to be very ductile under cyclic loading and their envelope

axial load-strain curves are almost the same as the corresponding monotonic axial stress-

strain curve. It has also been shown that repeated unloading/reloading cycles have a

cumulative effect on the permanent strain and the stress deterioration of the confined concrete

in hybrid DSTCs. Interfacial slips between the steel tube and the concrete may lead to

noticeable differences in the axial strain between them when the column is fully unloaded

from an axial strain level that significantly exceeds the yield strain of the steel tube.


17/68



CHAPTER-3

AIM AND SCOPE OF STUDY

As the DSST columns are a new form of structural members, no existing studies have

dealt with their behavior and design. This reports research aimed at developing a good

understanding of the structural behavior of DSST and reliable Design methods for this new

form of hybrid columns. The report is mainly concerned With DSST Columns with two

concentrically placed circular tubes filled with mortar In between, so hereafter the term new

DSST Columnsor DSST Columnsis reserved for columns with a section unless otherwise

specified.

The stress-strain behavior of the confined concrete in this new form of hybrid

Structural members is the key to understanding their structural performance. To better

understand the behavior of concrete, it is important to understand how the concrete isconfined by the two tubes in these new columns and how the Inner void and the steel tube

affect the effectiveness of confinement.

Based on the above considerations, the research work presented in this thesis was

carried out with the following five specific objectives:

1. To obtain a good understanding of the Compressive behavior of DSST through

experimental work;

2. To clarify the confinement mechanism for the mortar in DSST, through Comparative tests

of different section forms;

3. To develop the mathematical models for previous Researchers contribution;

4. To develop ANN model using Mat Lab v7.12 (R2011a).


18/68



CHAPTER-4

EXPERIMENTAL PROGRAM.

4.1: Preparation of SpecimenFollowing are the major steps carried out to prepare the specimens-

4.1.1: Step1The Steel tubes of grade Fe-310 were cut into different lengths of 350mm, 450mm

and 550mm by using a cutting machine. The steel tubes mentioned above were of different

diameters and thickness of 21.3mm, 26.9mm, 33.7mm, 42.4mm and 2.6mm, 3.2mm, 4.0mm

respectively. The end faces of the specimen were properly machined to achieve exact

bearing.

Fig 4.1: Empty circular steel columns (before test)

4.1.2: Step2

Double skinned steel columns are achieved by selecting steel tubes of differentdiameter but of same thickness and lengths.

Fig 4.2: Double skinned empty circular steel columns (before test)


19/68



4.1.3:Step3As per IS: 2250-1981 (Reaffirmed 1990)-Third reprint, February 1993-Indian Standard code

of practice for preparation and use of masonry mortars, for masonry in buildings subject to vibration

of machinery, the grade of mortar shall not be less than MM 3 (Clause 7.1.4). The cement mortar of

mix ratios 1:3, 1:4, 1:5 is obtained and the corresponding compressive strength of the moulds

are shown below.

Ratio Wt of cement

(grams)

Wt of

sand(grams)

Wt of the CM

cube(kg)

Compressive strength of the

mould (KN)

1:3 200 600 0.701 7.0

1:4 200 800 0.705 6.2

1:5 200 1000 0.710 5.3

Fig 4.3: cement mortar cube Fig 4.4: compressive strength testing of cube

4.1.4: Step4

The mortar mix of above said ratio are filled in between the uniform gap of double

skinned steel columns and is well compacted to keep the steel tubes intact. The steel tube

placed inside remains hollow.

Fig 4.5: double skinned tube with mortar mix and curing of samples

The specimens prepared are placed for curing for the time duration of 7 days and are tested

for their compressive strength.


20/68



4.2: EXPERIMENTAL STUDY

4.2.1: Compression test

The compressive strength of the specimens under monotonic loading condition is

obtained using 200ton capacity monotonic loading machine.

4.2.2: Components of 200 ton loading machine

Fig 4.6: hydraulic compressive loading machine

Hydraulic press for testing load comprising Press frame; hydraulic cylinder

(dia320Xdia 250X250mm stroke). Hydraulic power pack 100 its with electric motor 5hp X

1440rpm,electrical control panel operating with PLC SCADA software,strain gauge SI -30 &

strain indicator.

4.2.3: 200 ton loading machine frame construction

The hydraulic press consists of press frame, mounting legs; hydraulic cylinder, spacers

12 nos to adjust the length of the specimen and load cell. The small amount of force can be

applied to the pump and used to compress very heavy objects. By working under Pascal s

principle the pressure in an enclosed liquid must be the same everywhere. The press frame is

2.2m height X 1m length X 1m width and the operation hydraulic cylinder is to move up and

down. Hydraulic cylinder (dia 320Xdia 250X250mm stroke) front flange mounting present in


21/68



button of the press which is stroking up/down. The height from end of the stroke to load cell

maintains 1m including spacers of (dia250X55) 12 nos. the spacers provided to fix the

specimen according to the length. Load cell which present in the top of the press frame

connected to electrical control panel which gives the reading of tonnage of hydraulic press

when the cylinder is in loaded condition.

4.2.4: 200 ton loading machine PACK

Pump flow: 6.2 lpm

Electric motor: 5HP X 1440rpm,3dia,415V & 50 Hz

The system employs with radial piston pump drawing oil from tank through suction

strainer. The pump is coupled with electric motor of 5HP. The proportional pressure relief

valve is used to set the system pressure to maintain the same pressure in the entire hydraulic

system. The solenoid operated direction valves employed to lift the hydraulic cylinder to up

and down. The corresponding voltages are given to the relief valve to get the appropriate

load. The voltage can be varied from (0-10v). air blast oil cooler will reduce the heat

generated in the oil tank and pressure filter will remove the dirts from the oil.

4.2.5: PLC Electrical Control Panel Using SCADA

Power supply: 415v

Phase: 3dia

Frequency: 50Hz

Input current: 5 amps

Electrical control panel is accommodated to lift up and down the cylinder using

hydraulic power pack. There are two separate operation Auto and manual present in the

control panel. Separate push buttons are provided for both the operations. When the specimen

is kept and pressure is applied for corresponding load the values can noticed in the control

panel. When the recipe is given the strain readings can be monitored in the indicator. Batch &

data wise reports can be generated using SCADA software and by the reading the stress v/s

strain, load v/s deflection, load v/s strain graph can be plotted.


22/68



4.3: Strain Gauge

The strain gauge with constant wire resistor is fixed on the specimen. The suitable

bonding is to be and the lead is connected to the strain indicator to absorb the reading. The

both full & half bridge circuit can be permitted from the above strain gauge. SYSCON make

strain accept signals from all type of strain gauges. Different type of gauge factors can be

connected right away with the instruments and measurements can be made without any

inaccuracy. Different types of strain values can be calculated while the specimen gets

compressed.

4.4: Test Procedure by Using SCADA Software

1.

Start SCADA in computer,

2. On desktop click on SCADA link, sure that machine should be in Auto Mode.

3. It will directly go to the run mode and the main process screen will open.

Fig 4.7: main menu of SCADA


23/68



4. Now click on Recipe Button, one Pop up will appear select any the desired recipe and

press ok.

Fig 4.8: recipe of SCADA

5.

By clicking OK button one recipe screen will open which contain different parameters to

be filled by user as per the requirement.

6. If the user wants to edit some parameters in the recipe, user can make changes, after that

just press EDIT Button. This recipe contains 50 segments; user can access these by

pressing Next Button.

7. If the user wants to create new recipe just press NEW button and fill the required

parameters and save it.


24/68



Fig 4.9: No. of segments in SCADA and loading condition

8. After filling the required parameters in recipe just press Load button, one popup will come

for conformation just press ok. Your recipe will be loaded to the PLC.

9. Just press exit to come on main process screen.

10.Fix the material in the machine. Now in SCADA just check Load (K-N) in Machine

Status. There will be some value just put the same value in Load Correction Factor.

11.Now press start button, one popup will appear. Fill the Batch No and Log interval and

Press Ok.


25/68



12. Batches will Start Running and the Process will be as per the recipe program, after the

particular time duration the process will be completed and one conformation popup will

appear.

Fig 4.10: starting of recipe and loading

13.For Report just click on Excel-Report select the desired Batch No in the popup and press

Batch Report button, the report will open in excel sheet format simply save it.


26/68



Fig 4.11: Result obtained form SCADA

14.

For Trend reports click on Trend button, and select the required trend i.e. Stress v/sMeasured Strain, Stress v/s Calculated Strain and Load v/s Deflection.

15.After selecting the required trend just click on plot and select the required Batch no and

press Ok, the trend will be plotted on the graph. Press Print Button to get the Print of

graph.


27/68



Fig 4.12: Screen shot of SCADA-Main Menu

16.To retrieve the previous data trend click on Trend button, and select Real History

trend.

17.Now in the Real History Trend Just click on Select Group and select the required

group from popup.

18.

Just Click on pause button and in the right hand side corner select the required

batch no and select the ENTER DATE/BATCH button. Press ok trend will appear for

the selected Batch No.

19.To save this trend click on save button and save the trend.


28/68



Fig 4.13: History Trend


29/68



4.5: Loading Scheme

Fig 4.14: axial loading

Here all the samples were monotonically loaded .monotonic compression involving

full reloading cycles have been conducted , where the reloading of each cycle was designed

to terminate at the loading displacement of the previous cycle) or after reaching the envelope

curve figure shows the loading scheme as shown below.

Fig 4.15: Setup for Monotonic loading condition


30/68



Fig 4.16: Picture taken during the test

Fig 4.17: Double skinned specimen at the end of test


31/68



Fig 4.18: No. of Double skinned specimen after test

Fig 4.18(a) Fig 4.18(b) Fig 4.18(c)

COMMENTS:Fig 4.18(a) shows failure at mid height. Primary buckling as occurred due to

axial compressive loading.

Fig 4.18(b) shows failure at 1/3th height from the bottom of the loading platform whereas

Fig 4.18(c) shows failure bulging and twisting at 1/4th

length from top and bottom supports.

In these specimens, local buckling of steel was delayed due double skinned and mortar infill.

As grade of mortar increase buckling occurred before yielding of steel.


32/68



CHAPTER-5

TYPICAL RESULTS AND DISCUSSION

5.1: Result for Sample 1 (1:3 ratio, 450mm length, 2.6mm thick)

Table 5.1: output obtained from the SCADA

Date TimeSegment

Number

Load (K-

N)Area

Stress

PV

Original

Length

(MM)

Measure

d length

(MM)

Changein

Length

(

Defelctio

n)

Measure

d Strain

Calculat

ed Strain

20/04/2013 15:52:27 1 14 892 0.02 450.00 74.33 524.00 1.77 116.44

20/04/2013 15:52:28 1 15 892 0.02 450.00 74.33 524.81 1.77 116.54

20/04/2013 15:52:29 1 15 892 0.02 450.00 74.33 524.81 1.77 116.54

20/04/2013 15:52:30 1 14 892 0.02 450.00 75.60 524.81 1.77 116.82

20/04/2013 15:52:31 1 14 892 0.02 450.00 75.60 526.10 1.77 116.82

20/04/2013 15:52:32 1 16 892 0.02 450.00 76.88 526.10 1.77 117.10

20/04/2013 15:52:33 1 16 892 0.02 450.00 76.88 527.35 1.77 117.10

20/04/2013 15:52:34 2 16 892 0.02 450.00 78.06 527.35 1.77 117.10

20/04/2013 15:52:35 2 25 892 0.03 450.00 78.06 528.50 1.77 117.37

20/04/2013 15:52:36 2 25 892 0.03 450.00 78.06 528.50 1.77 117.37

20/04/2013 15:52:37 2 38 892 0.04 450.00 79.17 528.50 1.77 117.61

20/04/2013 15:52:38 2 38 892 0.04 450.00 79.17 529.60 1.77 117.61

20/04/2013 15:52:39 2 48 892 0.04 450.00 80.27 529.60 1.77 117.86

20/04/2013 15:52:40 2 48 892 0.06 450.00 80.27 530.69 1.77 117.86

20/04/2013 15:52:41 3 48 892 0.06 450.00 81.39 530.69 1.77 117.86

20/04/2013 15:52:42 3 59 892 0.07 450.00 81.39 531.81 1.77 118.10

20/04/2013 15:52:43 3 59 892 0.07 450.00 81.39 531.81 1.77 118.10

20/04/2013 15:52:44 3 67 892 0.08 450.00 82.47 531.81 1.77 118.35

20/04/2013 15:52:45 4 67 892 0.08 450.00 82.47 532.88 1.77 118.35

20/04/2013 15:52:46 4 85 892 0.08 450.00 83.50 532.88 1.77 118.57

20/04/2013 15:52:47 4 85 892 0.10 450.00 83.50 533.78 1.77 118.57

20/04/2013 15:52:48 4 85 892 0.10 450.00 83.91 533.78 1.77 118.65

20/04/2013 15:52:49 4 98 892 0.11 450.00 83.91 533.93 1.77 118.65

20/04/2013 15:52:51 5 98 892 0.11 450.00 83.91 533.93 1.77 118.65

20/04/2013 15:52:52 5 99 892 0.11 450.00 83.97 533.99 1.77 118.66

20/04/2013 15:52:53 6 99 892 0.11 450.00 83.97 533.99 1.77 118.66

20/04/2013 15:52:54 6 100 892 0.11 450.00 84.02 533.99 1.77 118.6720/04/2013 15:52:55 6 100 892 0.11 450.00 84.02 534.05 1.77 118.67

20/04/2013 15:52:56 6 100 892 0.11 450.00 84.10 534.05 1.77 118.69

20/04/2013 15:52:57 6 102 892 0.11 450.00 84.10 534.13 1.77 118.69

20/04/2013 15:52:58 7 102 892 0.11 450.00 84.10 534.13 1.77 118.69

20/04/2013 15:52:59 7 105 892 0.12 450.00 84.26 534.34 1.77 118.73

20/04/2013 15:53:00 7 105 892 0.12 450.00 84.26 534.34 1.77 118.73

20/04/2013 15:53:01 7 107 892 0.12 450.00 84.44 534.34 1.77 118.77

20/04/2013 15:53:02 7 107 892 0.12 450.00 84.44 534.58 1.77 118.77

20/04/2013 15:53:03 8 107 892 0.12 450.00 84.86 534.58 1.77 118.87

20/04/2013 15:53:04 8 110 892 0.12 450.00 84.86 535.02 1.77 118.87

20/04/2013 15:53:05 9 110 892 0.12 450.00 84.86 535.02 1.77 118.87

20/04/2013 15:53:06 9 111 892 0.13 450.00 85.43 535.84 1.77 119.00

20/04/2013 15:53:07 9 111 892 0.13 450.00 85.43 535.84 1.77 119.00

20/04/2013 15:53:08 9 108 892 0.12 450.00 86.52 535.84 1.77 119.24

20/04/2013 15:53:09 9 108 892 0.12 450.00 86.52 536.95 1.77 119.24

20/04/2013 15:53:10 9 108 892 0.12 450.00 87.65 536.95 1.77 119.50

20/04/2013 15:53:11 9 96 892 0.11 450.00 87.65 538.10 1.77 119.50

20/04/2013 15:53:12 9 96 892 0.11 450.00 88.82 538.10 1.77 119.50

20/04/2013 15:53:13 9 83 892 0.09 450.00 88.82 539.29 1.77 119.76

20/04/2013 15:53:14 10 83 892 0.09 450.00 88.82 539.29 1.77 119.76

20/04/2013 15:53:15 10 66 892 0.07 450.00 90.05 539.29 1.77 120.03

20/04/2013 15:53:16 10 66 892 0.07 450.00 90.05 540.54 1.77 120.03


33/68



5.2: Result for Sample 2 (1:3 ratio, 550mm length, 2.6mm thick)


Date TimeSegment

Number

Load (K-

N)Area

Stress

PV

Original

Length

(MM)

Measure

d length

(MM)

Change

in

Length

(

Defelctio

n)

Measure

d Strain

Calculat

ed Strain

20/04/2013 16:03:25 1 5 892 0.01 550.00 74.24 624. 00 1.77 113. 51

20/04/2013 16:03:26 1 5 892 0.01 550.00 74.24 624. 74 1.77 113. 51

20/04/2013 16:03:27 1 5 892 0.01 550.00 74.24 624. 74 1.77 113. 51

20/04/2013 16:03:28 1 7 892 0.01 550.00 75.52 625. 99 1.77 113. 75

20/04/2013 16:03:29 1 7 892 0.01 550.00 75.52 625. 99 1.77 113. 75

20/04/2013 16:03:30 1 19 892 0.03 550.00 76.64 625. 99 1.77 113. 95

20/04/2013 16:03:31 1 19 892 0.03 550.00 76.64 627. 02 1.77 113. 95

20/04/2013 16:03:32 1 19 892 0.03 550.00 77.64 627. 02 1.77 114. 13

20/04/2013 16:03:33 1 50 892 0.06 550.00 77.64 628. 01 1.77 114. 13

20/04/2013 16:03:34 2 50 892 0.06 550.00 77.64 628. 01 1.77 114. 13

20/04/2013 16:03:35 2 73 892 0.08 550.00 78.23 628. 27 1.77 114. 23

20/04/2013 16:03:36 2 73 892 0.08 550.00 78.23 628. 27 1.77 114. 23

20/04/2013 16:03:37 2 75 892 0.08 550.00 78.29 628. 27 1.77 114. 24

20/04/2013 16:03:38 3 75 892 0.08 550.00 78.29 628. 33 1.77 114. 24

20/04/2013 16:03:39 3 82 892 0.08 550.00 78.74 628. 33 1.77 114. 33

20/04/2013 16:03:40 3 82 892 0.10 550.00 78.74 628. 93 1.77 114. 33

20/04/2013 16:03:41 3 82 892 0.10 550.00 79.12 628. 93 1.77 114. 33

20/04/2013 16:03:42 3 91 892 0.10 550.00 79.12 629. 21 1.77 114. 39

20/04/2013 16:03:43 4 91 892 0.10 550.00 79.12 629. 21 1.77 114. 39

20/04/2013 16:03:44 4 95 892 0.11 550.00 79.45 629. 21 1.77 114. 45

20/04/2013 16:03:46 4 95 892 0.11 550.00 79.45 629. 58 1.77 114. 45

20/04/2013 16:03:47 4 98 892 0.11 550.00 79.70 629. 58 1.77 114. 49

20/04/2013 16:03:48 4 98 892 0.11 550.00 79.70 629. 78 1.77 114. 49

20/04/2013 16:03:49 5 98 892 0.11 550.00 79.91 629. 78 1.77 114. 49

20/04/2013 16:03:50 5 100 892 0.11 550.00 79.91 629. 98 1.77 114. 53

20/04/2013 16:03:51 6 100 892 0.11 550.00 79.91 629. 98 1.77 114. 53

20/04/2013 16:03:52 6 102 892 0.11 550.00 80.12 629. 98 1.77 114. 57

20/04/2013 16:03:53 6 102 892 0.11 550.00 80.12 630. 26 1.77 114. 57

20/04/2013 16:03:54 6 103 892 0.11 550.00 80.54 630. 26 1.77 114. 65

20/04/2013 16:03:55 6 103 892 0.12 550.00 80.54 630. 74 1.77 114. 65

20/04/2013 16:03:56 7 103 892 0.12 550.00 81.23 630. 74 1.77 114. 65

20/04/2013 16:03:57 7 104 892 0.12 550.00 81.23 631. 58 1.77 114. 78

20/04/2013 16:03:58 7 104 892 0.12 550.00 81.23 631. 58 1.77 114. 78

20/04/2013 16:03:59 7 99 892 0.11 550.00 82.26 631. 58 1.77 114. 97

20/04/2013 16:04:00 8 99 892 0.11 550.00 82.26 632. 69 1.77 114. 97

20/04/2013 16:04:01 8 88 892 0.11 550.00 83.40 632. 69 1.77 115. 18

20/04/2013 16:04:02 8 88 892 0.10 550.00 83.40 633. 86 1.77 115. 18

20/04/2013 16:04:03 9 88 892 0.10 550.00 84.60 633. 86 1.77 115. 40

20/04/2013 16:04:04 9 73 892 0.08 550.00 84.60 635. 09 1.77 115. 40

20/04/2013 16:04:05 9 73 892 0.08 550.00 84.60 635. 09 1.77 115. 40

20/04/2013 16:04:06 9 61 892 0.07 550.00 85.86 635. 09 1.77 115. 63

20/04/2013 16:04:07 9 61 892 0.07 550.00 85.86 636. 36 1.77 115. 63

20/04/2013 16:04:08 9 48 892 0.07 550.00 87.16 636. 36 1.77 115. 87

20/04/2013 16:04:09 9 48 892 0.05 550.00 87.16 637. 65 1.77 115. 87

20/04/2013 16:04:10 9 48 892 0.05 550.00 88.47 637. 65 1.77 116. 10


34/68



Plot 5.1: load v/s deflection

Plot 5.2: load v/s deflection

COMMENTS:As deflection increased it can be observed from plot 5.1 and plot 5.2 it can be

observed load reached its peak value and suddenly decreased may be due to formation of

plastic hinges and internal crushing of the inner tube. Also sample1 give more strength when

compared with sample2 due its change in length.Similarly the results obtained from SCADA are tabulated below for their respective

ultimate load and deflection.

0

20

40

60

80

100

120

520.00 525.00 530.00 535.00 540.00 545.00

Load

(KN)

Deflection (mm)

LOAD v/s DEFLECTION

LOAD v/s

DEFLECTION

0

20

40

60

80

100

120

620.00 625.00 630.00 635.00 640.00

Load

(KN)

Deflection (mm)

LOAD v/s DEFLECTION

LOADv/s DEFLECTION


35/68



5.3: Tabulation of results of specimens.

grade of mortarsection properties (mm)

material

propertiesutimate axial

load

(experimental)

deflection

(mm)

D T L fc fy

1:3

33.7 2.6 350 7.0 310 123 424

42.4 2.6 350 7.0 310 133 425.23

42.4 2.6 350 7.0 310 139 57.74

33.7 2.6 450 7.0 310 111 535.84

42.2 2.6 450 7.0 310 124 531.86

42.4 2.6 450 7.0 310 135 530.74

33.7 2.6 550 7.0 310 104 631.58

42.4 2.6 550 7.0 310 119 624.01

42.4 2.6 550 7.0 310 120 292.8

33.7 3.2 350 7.0 310 165 498.2342.4 3.2 350 7.0 310 218 442.16

42.4 3.2 350 7.0 310 198 437.45

33.7 3.2 450 7.0 310 150 119.12

42.2 3.2 450 7.0 310 215 478.68

42.4 3.2 450 7.0 310 193 630.76

33.7 3.2 550 7.0 310 115 567.56

42.4 3.2 550 7.0 310 200 354.16

42.4 3.2 550 7.0 310 189 487.44

42.4 4.0 350 7.0 310 243 411.24

1:4

33.7 2.6 350 6.2 310 112 433.94

42.4 2.6 350 6.2 310 129 430.63

42.4 2.6 350 6.2 310 134 431.76

33.7 2.6 450 6.2 310 105 411.56

42.2 2.6 450 6.2 310 121 511.33

42.4 2.6 450 6.2 310 131 498.34

33.7 2.6 550 6.2 310 98 543.17

42.4 2.6 550 6.2 310 118 611.23

42.4 2.6 550 6.2 310 120 277.3

33.7 3.2 350 6.2 310 173 432.01

42.4 3.2 350 6.2 310 214 431.66

42.4 3.2 350 6.2 310 183 432.33

33.7 3.2 450 6.2 310 166 133.76

42.2 3.2 450 6.2 310 207 478.38

42.4 3.2 450 6.2 310 176 567.55

33.7 3.2 550 6.2 310 159 634.01

42.4 3.2 550 6.2 310 194 347.77

42.4 3.2 550 6.2 310 171 478.09


36/68



42.4 4.0 350 6.2 310 219 423.9

1:5

33.7 2.6 350 5.3 310 116 431.04

42.4 2.6 350 5.3 310 143 429.87

42.4 2.6 350 5.3 310 147 436.53

33.7 2.6 450 5.3 310 110 402.25

42.2 2.6 450 5.3 310 118 496.5442.4 2.6 450 5.3 310 126 477.13

33.7 2.6 550 5.3 310 94 511.04

42.4 2.6 550 5.3 310 113 578.36

42.4 2.6 550 5.3 310 117 435

33.7 3.2 350 5.3 310 158 434.27

42.4 3.2 350 5.3 310 217 430.06

42.4 3.2 350 5.3 310 177 248.45

33.7 3.2 450 5.3 310 142 109.38

42.2 3.2 450 5.3 310 191 456.76

42.4 3.2 450 5.3 310 164 523.23

33.7 3.2 550 5.3 310 123 498.34

42.4 3.2 550 5.3 310 183 375.85

42.4 3.2 550 5.3 310 157 411.44

42.4 4.0 350 5.3 310 215 411.65


5.4: Plots for Grade V/S Load

5.4.1: plot for 2.6mm thick, 350mm length

Table 5.4 Plot 5.3: 2.6mm thick, 350mm length

0

20

40

60

80

100

120

140

160

1:3 1:3 1:3 1:4 1:4 1:4 1:5 1:5 1:5

GRADE V/S LOAD

LOAD

GRADE

(X)

LOAD

(y)

1:3 123

1:3 133

1:3 139

1:4 112

1:4 129

1:4 134

1:5 116

1:5 143

1:5 147


37/68







0

20

40

60

80

100

120

140

160

1:3 1:3 1:3 1:4 1:4 1:4 1:5 1:5 1:5

GRADE V/S LOAD

LOAD

0

20

40

60

80

100

120

140

1:3 1:3 1:3 1:4 1:4 1:4 1:5 1:5 1:5

GRADE V/S LOAD

LOAD

GRADE

(X)

LOAD

(Y)

1:3 111

1:3 124

1:3 135

1:4 105

1:4 121

1:4 131

1:5 110

1:5 118

1:5 126

GRADE(x) LOAD(y)

1:3 104

1:3 119

1:3 120

1:4 98

1:4 118

1:4 120

1:5 94

1:5 113

1:5 117


38/68




Table 5.7 plot 5.6: 3.2mm thick, 350mm length



0

50

100

150

200

250

1:3 1:3 1:3 1:4 1:4 1:4 1:5 1:5 1:5

GRADE V/S LOAD

LOAD

0

50

100

150

200

250

1:3 1:3 1:3 1:4 1:4 1:4 1:5 1:5 1:5

GRADE V/S LOAD

LOAD

GRADE(X) LOAD(Y)

1:3 165

1:3 218

1:3 198

1:4 173

1:4 214

1:4 183

1:5 158

1:5 217

1:5 177

GRADE

(X)

LOAD

(Y)

1:3 150

1:3 215

1:3 193

1:4 166

1:4 207

1:4 176

1:5 142

1:5 191

1:5 164


39/68







0

50

100

150

200

250

1:3 1:3 1:3 1:4 1:4 1:4 1:5 1:5 1:5

GRADE V/S LOAD

LOAD

200

205

210

215

220

225

230

235

240

245

250

1:3 1:4 1:5

GRADE v/s LOAD

LOAD

GRADE(X) LOAD(Y)

1:3 115

1:3 200

1:3 189

1:4 159

1:4 194

1:4 171

1:5 123

1:5 183

1:5 157

GRADE

(X)

LOAD

(Y)

1:3 243

1:4 219

1:5 215


40/68



5.5: Plots for thickness v/s load

5.5.1: plot for thickness v/s load of 1:3 ratios

Table 5.11

Plot 5.10: thickness v/s load of 1:3 ratios


thick 350 450 550

2.6 112 105 98

2.6 129 121 118

2.6 134 131 120

3.2 173 166 159

3.2 214 207 194

3.2 183 176 171

Table 5.12

0

50

100

150

200

250

300

2.6 2.6 2.6 3.2 3.2 3.2 4.0 4.0 4.0

THICKNESS V/S LOAD

Series1

Series2

Series3

thick 350 450 550

2.6 123 111 1052.6 133 124 119

2.6 139 135 120

3.2 165 150 115

3.2 218 215 200

3.2 198 193 189

4.0 243

4.0 219

4.0 215


41/68





thick 350 450 550

2.6 116 110 94

2.6 143 118 113

2.6 147 126 117

3.2 158 142 123

3.2 217 191 183

3.2 177.0 164 157

Table 5.13


0

50

100

150

200

250

2.6 2.6 2.6 3.2 3.2 3.2

Series1

Series2

Series3

0

50

100

150

200

250

2.6 2.6 2.6 3.2 3.2 3.2

THICKNESS V/S LOAD

Series1

Series2

Series3


42/68



5.6: PLOTS FOR L/D V/S LOAD

5.6.1: Plot for 1:3 ratio, 2.6 mm thick

length L/D ratio

LOAD in

KN

350 10.38 123

1:3

ratio,

2.6

mm

thick

350 8.25 133

350 8.25 139

450 13.35 111

450 10.61 124

450 10.61 135

550 16.32 104

550 12.97 119

550 12.97 120

Table 5.14

Plot 5.13: 1:3 ratio, 2.6 mm thick


length L/D ratio LOAD in KN

350 10.38 112

1:4

ratio,2.6mm

thick

350 8.25 129

350 8.25 134

450 13.35 105

450 10.61 121

450 10.61 131

550 16.32 98

550 12.97 118

550 12.97 120

Table 5.15

0

20

40

60

80

100

120

140

160

10.38 8.25 8.25 13.35 10.61 10.61 16.32 12.97 12.97

Load

(KN)

L/D ratio

L/D v/s LOAD

L/D v/s


43/68






350 10.38 116

1:5

ratio ,

2.6 mm

thick

350 8.25 143

350 8.25 147

450 13.35 110

450 10.61 118

450 10.61 126

550 16.32 94

550 12.97 113

550 12.97 117

Table 5.16


0

20

40

60

80

100

120

140

160

Load

(KN)

L/D ratio

L/D v/s LOAD

L/D v/s LOAd

0

50

100

150

200

Load

(KN)

L/D ratio

L/D v/s LOAD

L/D v/s LOAd


44/68





350 10.38 165

1:3ratio,3.2 mm

thick

350 8.25 218

350 8.25 198

450 13.35 150450 10.61 215

450 10.61 193

550 16.32 115

550 12.97 200

550 12.97 189

Table 5.17



1:4ratio,

3.2mm

thick


350 10.38 173

350 8.25 214

350 8.25 183

450 13.35 166

450 10.61 207

450 10.61 176

550 16.32 159

550 12.97 194

550 12.97 171

Table 5.18

0

50

100

150

200

250

10.38 8.25 8.25 13.35 10.61 10.61 16.32 12.97 12.97

Load

(KN)

L/D ratio

L/D v/s LOAD

L/D v/s LOAd


45/68





1:5ratio,

3.2 mm

thick


350 10.38 158

350 8.25 217

350 8.25 177

450 13.35 142

450 10.61 191

450 10.61 164

550 16.32 123

550 12.97 183

550 12.97 157Table 5.19


0

50

100

150

200

250

10.38 8.25 8.25 13.35 10.61 10.61 16.32 12.97 12.97

Load

(KN

)

L/D ratio

L/D v/s LOAD

L/D v/s LOAd

0

50

100

150

200

250

10.38 8.25 8.25 13.35 10.61 10.61 16.32 12.97 12.97

Load

(KN)

L/D ratio

L/D v/s LOAD

L/D v/s LOAd


46/68



5.7: MATHEMATICAL MODELLING

5.7.1: To find the strength capacity of mortar filled steel Tubes (Inline with Research

paper-Min-Lang Lin and Keh-Chyuan Tsai[2]

)

The strength capacity Puof a specimen is defined as the peak value of the axial loads

observed in the axial load-strain curve. The corresponding strain is denoted as pe . The value

of Pois the nominal strength given by Euro code 4:

Po=Asf yt+Amf c (1)

Where As and Am are the cross-sectional areas of the steel and mortar section,

respectively. Thef ytandf care the yield strength of the steel tube and the actual compressive

strength of the mortar. Equation 1 differs from the AIJ specifications where a reduction factor

of 0.85 for the core mortar is not considered herein. It is observed that all values of Pu/ Po

observed lesser than 1.0 but not too significantly. Thus, it appears that Euro code 4 can

conservatively predict the ultimate axial strength of a DSST.

Specimen

(samples)

Pu

(kN)

Po(kN) Pu/Po

(kN)

p

(%)

Ecomp

(Mpa)

Ethe

(Mpa)

E comp/E

the

95

DS-

2.6mm,

450length

111 267.16

6

0.41 0.58 17745.87 200137.5

7

0.08 0.95

DS-

2.6mm,

450length

104 267.16

6

0.38 0.61 18927.49 200137.5

7

0.09 0.95

Table 5.20: Results of axial loading test


47/68



Figure shows the typical axial load v/s axial strain and axial ductility definition

5.7.2: Stiffness

The initial stiffness E compof a composite member is defined as the averaged initial slope of

an axial load-strain curve. It is calculated from its linear recurrence within the range of 0.05%

to 0.10% axial deformation, divided by the cross-sectional area total A of the composite

member. The theoretical stiffness, according to the theory of superposition can be expressed

as:


48/68



E the= (A sE s+ A c E c)/A total (2)

The values of Ecomp/Etheof all specimens are listed in Table. Apparently, the values of the

E comp/Ethe computed from Equation 2.seriously overestimates the stiffness of the specimen.

5.7.2: Axial Ductility

In this study, the axial ductility is defined as:

95= 95/y (3)

y=

75/ 0.75 (4)

Where 75and 95shown in above graph are the axial strains corresponding to the

75% and 95% of the peak axial load before and after the peak load was achieved,

respectively. Because the yielding point of a specimen is difficult to identify from the axial

load versus strain curve, the idealized yield strain yis extrapolated from 75 From Table.,

from above graph it is observed that all the specimens have similar performance in axial

ductility. The ductility value of the DSST specimen is slightly lesser than the another

specimens. That is, the strength degrading of the DSST specimen is slightly slower than the

another specimens, but not significantly. When a 0.03 axial strain is reached, all the

specimens can still retain more than 50% of its peak strength.


49/68



CHAPTER-6

ANALYTICAL STUDIES USING ARTIFICIAL NEURAL

NETWORKS

6.1: Introduction

Columns occupy a vital place in the structural system. Weakness or failure of a

column destabilizes the entire structure. Strength & ductility of steel columns need to be

ensured through adequate strengthening, repair & rehabilitation techniques to maintain

adequate structural performance. Recently, composite columns are finding a lot of usage for

seismic resistance. In order to prevent shear failure of RC column resulting in storey collapse

of buildings, it is essential to make ductility of column larger. Recently, most of the buildings

utilize this CFT concept as primary for lateral load resisting frames. The concrete used for

encasing the structural steel section not only enhance its strength and stiffness, but also protects

it from fire damages. Recycled aggregate concrete is used as an infill in order to achieve

economy.

One way of including specimen irregularities in the model is to use the results of the

available experiments to predict the behavior of composite tubes subjected to different

loading. ANN is a technique that uses existing experimental data to predict the behavior of

the same material under different testing conditions. Using this method, details regarding

bonding properties between fiber and matrix, strength variation of fibers and any

manufacturing induced imperfections are implicitly incorporated within the input

parameters fed to neural network.

In the current work, the prediction of the load-carrying capacities for axially-loaded

rectangular composite tubes is evaluated using ANN. To test the validity of using ANN in

determining the crushing behavior of these tubes, the study will compare the predictions

obtained to the experimental results using the neural network tool in MATLAB v7.12

(R2011a).


50/68



6.2: Artificial Neural Network

Figure shows the Neural Network

6.2.1: Introduction

ANN have emerged as a useful concept from the field of artificial intelligence, and

has been used successfully over the past decade in modeling engineering problems in general,

and specifically those relating to the mechanism behavior of fiber- reinforced composite

materials.

ANN generally consists of a number of layers: the layer where the patterns are applied

is called input layer. This layer could typically include the properties of the composite

material under consideration, its layup, the applied load, the tube aspect ratio etc. The layer

where the output is obtained is the output layer which could, for example, contain the

resulting deformation of this tube under the given loading conditions. In addition, there may

be one or more layers between the input and output layers called hidden layers, which are so

named because their outputs are not directly observable. The addition of hidden layers

enables the network to extract high-order statistics which are particularly valuable when the

size of the input is very large. Neurons in each layer are interconnected to preceding and

subsequent layer neurons with each interconnection having an associated weight.

A training algorithm is commonly used to iteratively minimize a cost function with

respect to the interconnection weights and neuron thresholds. The training process is

terminated either when the mean square error (MSE) between the observed data and the ANN

outcomes for all elements in the training set has reached a pre-specified threshold or after the

completion of a pre-specified number of learning epochs [1-4].


51/68



6.2.2 KolmogorovsTheorem

! Any continuous real-valued functions f (x1, x2, ..., xn) defined on [0, 1]n, , can be

represented in the form

f(x1, x2, ..., xn) =

where the gj's are properly chosen continuous functions of one variable, and the ij'sare continuous monotonically increasing functions independent of f.

Fig1: Block diagram of feed forward network

Given any function yxRImn

"# )(,: !! , where I is the closed unit interval [0,1],

can be implemented exactly by a three layer neural network with n input nodes, 2n+1 hidden

layer neurons and m output layer neurons, as represented in fig.1.

6.2.3 Multilayer Neural Network Architecture

6.2.3.1 Neuron Model

An elementary neuron with R inputs is shown below. Each input is weighted with an

appropriate w. The sum of the weighted inputs and the bias forms the input to the transfer

function f. Neurons can use any differentiable transfer function f to generate their output.


52/68

Experimental Investigation on Double

Dept. of Civil Engg, GCE, Ram

Fig

Multilayer networks reptansig is shown. Sigmoid outp

while linear output neurons ar

function purelin as shown in fig.

kinned Steel Tubular (DSST) Columns Subjected to Mo

nagaram

.2a

Fig.2b:Neuron model

esented in fig.2a, can use the an-sigmoid tra t neurons are often used for pattern recogniti

e used for function fitting problems. The li

2b.

notonic loading

Page 43

sfer functionon problems,

near transfer


53/68



6.2.4 Train the Network

Once the network weights and biases are initialized, the network is ready for training.

The multilayer feedforward network can be trained for function approximation (nonlinear

regression) or pattern recognition. The training process requires a set of examples of proper

network behaviornetwork inputs p and target outputs t.

The process of training a neural network involves tuning the values of the weights and

biases of the network to optimize network performance, as defined by the network

performance function net.performFcn. The default performance function for feedforward

networks is mean square error msethe average squared error between the network outputs a

and the target outputs t. It is defined as follows:

There are two different ways in which training can be implemented: incremental

mode and batch mode. In incremental mode, the gradient is computed and the weights are

updated after each input is applied to the network. In batch mode, all the inputs in the training

set are applied to the network before the weights are updated. This chapter describes batchmode training with the train command. Incremental training with the adapt command is

discussed in Incremental Training with adapt and in Adaptive Filters and Adaptive Training.

For most problems, when using the Neural Network Toolbox software, batch training is

significantly faster and produces smaller errors than incremental training.

For training multilayer feed forward networks, any standard numerical optimization

algorithm can be used to optimize the performance function, but there are a few key ones that

have shown excellent performance for neural network training.

These optimization methods use either the gradient of the network performance withrespect to the network weights, or the Jacobian of the network errors with respect to the

weights.

The gradient and the Jacobian are calculated using a technique called the back

propagation algorithm, which involves performing computations backward through the

network. The back propagation computation is derived using the chain rule of calculus.


54/68



6.2.5 Network Properties

The network type is feed forward back propagation. The training function is

levenberg-marquardt algorithm. The performance function is mean square error. The transfer

function is tan-sigmoidal and purelin.

6.3. Work Flow

The work flow for the general neural network design process has seven primary steps:

1. Collect data

2. Create the network

3.

Configure the network

4. Initialize the weights and biases

5. Train the network

6.

Validate the network (post-training analysis)

7.

Use the network

6.4. Prediction and Experimental Results

The Linear-Sigmoidal (linsig) and Tan-Sigmoidal (tansig) functions used to build the

model and train the network. The output is trained separately for both ultimate load and axial

shortening load. Also the best values of prediction are obtained for 11 layers.

The experimental results which are obtained are given as the desired outputs to the

feed forward backpropagation network . These results were used to predict the output values

and were in good agreement with the Kolmogorovs theorem. The output values and the

deviations are obtained were tested and validated from 3 hidden layers to 14 hidden layers.

Ultimate axial load prediction

0

50

100

150

200

250

300

1 5 9 13 17 21 25 29 33 37 41 45 49 53 57

PU

Pu prediction


55/68



The experimental values are obtained and verified for ultimate axial load . The

ultimate axial loads average deviations are tabulated in. The best result is obtained for 11

layers as per Kolmogorov principle and this is verified in the ultimate axial load deviation

histogram for all the layers .The performance is measured using mean square error (MSE).

The predicted values are tested, validated and plotted to obtain the best values on the

curve fit. The experimental inputs are tested from 3 hidden layers to 14 hidden layers and it is

verified that the deviations obtained for the 11 hidden layers gives the best result, also with

the best regression fit.

Axial Shortening prediction

-50

0

50

100

150

200

2501

2 3 4 56

78

910

11121314151617

18

192021

2223

242526272829303132

333435

3637

383940

4142434445464748

4950

5152

535455

5657

PU

Pu prediction

PU error

0

100

200

300

400

500

600

700

1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55

AS

AS prediction


56/68



The experimental values are obtained and verified for axial shortening load .The

values are tabulated in Table-3. The deviations are also tabulated to choose the the best

results . Again it can be seen that the results obtained for 11as the number of hidden layers as

per Kolmogorovs theorem and this is verified again with axial load shortening .The

deviation is also represented in the histogram.

The comparison of the experimental results and the predicted ultimate axial load for

11 hidden layers . The same procedure is repeated for axial shortening; The experimental data

are obtained after training the model to 1000 number of epochs and assigning the transfer

function as tansig with the given inputs and predicted values. The input is trained using

Lavenberg-Marquardt algorithm. The performance is measured using mean square error

(MSE).The predicted values are tested, validated and plotted to obtain the best values on the

curve fit. The experimental inputs are tested from 3 hidden layers to 14 hidden layers and it is

verified that the deviations obtained for the 11 hidden layers gives the best result, also with

the best regression fit.

6.5. Conclusion

The experimental behavior and corresponding ANN predictions of circular composite

tube subjected axial compressive load were presented and discussed. The ANN has been

shown to successfully predict the crushing behavior of wide range of circular tubes. The

predicted results obtained, are showed that the feed forward back propagation network with

11 hidden neurons consistently provided the best predictions of the experimental data. From

the current work it can be concluded that ANN techniques can be used to effectively predict

the response of ultimate axial load and axial shortening on composite tubes.

-600-400-200

0200400600800

12 3 4 5 6

78

9101112131415161718

1920

2122

232425262728293031323334

3536

3738

3940

414243444546474849

5051

525354

555657

AS

AS prediction

AS error


57/68



TABLE 7.1: Tabulated Experimental Values

Grade PU ASD t L fc fy

1:3

33.7 2.6 350 7 310 123 42442.4 2.6 350 7 310 133 425.23

42.4 2.6 350 7 310 139 57.74

33.7 2.6 450 7 310 111 535.84

42.2 2.6 450 7 310 124 531.86

42.4 2.6 450 7 310 135 530.74

33.7 2.6 550 7 310 104 631.58

42.4 2.6 550 7 310 119 624.01

42.4 2.6 550 7 310 120 292.8

33.7 3.2 350 7 310 165 498.23

42.4 3.2 350 7 310 218 442.16

42.4 3.2 350 7 310 198 437.45

33.7 3.2 450 7 310 150 119.12

42.2 3.2 450 7 310 215 478.68

42.4 3.2 450 7 310 193 630.76

33.7 3.2 550 7 310 115 567.56

42.4 3.2 550 7 310 200 354.16

42.4 3.2 550 7 310 189 487.44

42.4' 4 350 7 310 243 411.24

1:4

33.7 2.6 350 6.2 310 112 433.94

42.4 2.6 350 6.2 310 129 430.63

42.4 2.6 350 6.2 310 134 431.76

33.7 2.6 450 6.2 310 105 411.56

42.2 2.6 450 6.2 310 121 511.33

42.4 2.6 450 6.2 310 131 498.34

33.7 2.6 550 6.2 310 98 543.17

42.4 2.6 550 6.2 310 118 611.23

42.4 2.6 550 6.2 310 120 277.3

33.7 3.2 350 6.2 310 173 432.01

42.4 3.2 350 6.2 310 214 431.66

42.4 3.2 350 6.2 310 183 432.33

33.7 3.2 450 6.2 310 166 133.76

42.2 3.2 450 6.2 310 207 478.38

42.4 3.2 450 6.2 310 176 567.55

33.7 3.2 550 6.2 310 159 634.0142.4 3.2 550 6.2 310 194 347.77

42.4 3.2 550 6.2 310 171 478.09

42.4' 4 350 6.2 310 219 423.9

1:5

33.7 2.6 350 5.3 310 116 431.04

42.4 2.6 350 5.3 310 143 429.87

42.4 2.6 350 5.3 310 147 436.53

33.7 2.6 450 5.3 310 110 402.25


58/68



42.2 2.6 450 5.3 310 118 496.54

42.4 2.6 450 5.3 310 126 477.13

33.7 2.6 550 5.3 310 94 511.04

42.4 2.6 550 5.3 310 113 578.36

42.4 2.6 550 5.3 310 117 435

33.7 3.2 350 5.3 310 158 434.27

42.4 3.2 350 5.3 310 217 430.06

42.4 3.2 350 5.3 310 177 248.45

33.7 3.2 450 5.3 310 142 109.38

42.2 3.2 450 5.3 310 191 456.76

42.4 3.2 450 5.3 310 164 523.23

33.7 3.2 550 5.3 310 123 498.34

42.4 3.2 550 5.3 310 183 375.85

42.4 3.2 550 5.3 310 157 411.44

42.4 4 350 5.3 310 215 411.65

TABLE 7.2: Prediction OF Pu and Its Deviation

PU

Pu

prediction PU error

123 129.6084 -6.6084

133 161.1598 -28.1598

139 161.1598 -22.1598

111 112.0944 -1.0944

124 126.9117 -2.9117

135 129.1005 5.8995

104 105.6747 -1.6747119 125.7029 -6.7029

120 125.7029 -5.7029

165 171.9848 -6.9848

218 213.1588 4.8412

198 213.1588 -15.1588

150 159.7345 -9.7345

215 208.6071 6.3929

193 209.4455 -16.4455

115 133.1576 -18.1576

200 203.3088 -3.3088

189 203.3088 -14.3088

243 231.7198 11.2802

112 133.3728 -21.3728

129 141.5893 -12.5893

134 141.5893 -7.5893

105 113.5402 -8.5402

121 127.0825 -6.0825

131 127.0552 3.9448


59/68



98 100.4327 -2.4327

118 125.5173 -7.5173

120 125.5173 -5.5173

173 172.049 0.95103

214 187.8391 26.1609

183 187.8391 -4.8391

166 152.5117 13.4883207 188.3865 18.6135

176 185.2793 -9.2793

159 141.6543 17.3457

194 182.7252 11.2748

171 182.7252 -11.7252

219 183.9339 35.0661

116 109.1319 6.8681

143 146.7343 -3.7343

147 146.7343 0.26566

110 111.2282 -1.2282

118 129.8992 -11.8992

126 129.4207 -3.4207

94 105.062 -11.062

113 108.3131 4.6869

117 108.3131 8.6869

158 148.1193 9.8807

217 167.7527 49.2473

177 167.7527 9.2473

142 151.3536 -9.3536

191 167.4269 23.5731

164 166.5387 -2.5387

123 135.5481 -12.5481

183 151.5585 31.4415157 151.5585 5.4415

215 176.1332 38.8668

TABLE 7.3: Axial Shortening Predicted Values and Its Deviation

AS

AS

prediction AS error

424 633.9139 -209.9139

425.23 155.2998 269.9302

57.74 155.2998 -97.5598

535.84 524.6071 11.2329

531.86 465.4583 66.4017

530.74 466.3267 64.4133

631.58 545.6866 85.8934

624.01 431.9607 192.0493

292.8 431.9607 -139.1607

498.23 571.4514 -73.2214


60/68



442.16 539.2225 -97.0625

437.45 539.2225 -101.7725

119.12 201.335 -82.215

478.68 533.5634 -54.8834

630.76 555.3099 75.4501

567.56 59.0981 508.4619

354.16 587.304 -233.144487.44 587.304 -99.864

411.24 634.01 -222.77

433.94 450.607 -16.667

430.63 634.0098 -203.3798

431.76 634.0098 -202.2498

411.56 563.0175 -151.4575

511.33 634.01 -122.68

498.34 634.01 -135.67

543.17 634.01 -90.84

611.23 634.01 -22.78

277.3 634.01 -356.71

432.01 161.4997 270.5103

431.66 633.9684 -202.3084

432.33 633.9684 -201.6384

133.76 266.3004 -132.5404

478.38 634.01 -155.63

567.55 634.01 -66.46

634.01 634.01 4.09E-12

347.77 634.01 -286.24

478.09 634.01 -155.92

423.9 326.6426 97.2574

431.04 634.01 -202.97

429.87 146.5533 283.3167436.53 146.5533 289.9767

402.25 634.01 -231.76

496.54 634.01 -137.47

477.13 634.01 -156.88

511.04 634.01 -122.97

578.36 634.01 -55.65

435 634.01 -199.01

434.27 634.01 -199.74

430.06 416.4371 13.6229

248.45 416.4371 -167.9871

109.38 634.01 -524.63

456.76 634.01 -177.25523.23 634.01 -110.78

498.34 634.01 -135.67

375.85 634.01 -258.16

411.44 634.01 -222.57

411.65 389.8547 21.7953


61/68



CHAPTER-7

SUMMARY AND CONCLUSIONS

!

As the length of the DSST increases, the ultimate axial strength decreases.

! As the wall thickness of both inner and outer tube is greater, it can with resist more

axial load.

! Higher the total cross sectional area of DSST, better the ultimate axial strength.

! The infilled material cement mortar also acts as fire resistance.

! The steel tubular column gives good aesthetic appearance.

! linear behavior till yielding was observed al

Double Skinned Steel Tubular (DSST)

Documents

Transcript of Double Skinned Steel Tubular (DSST)