Computational Study of Structural Insulated Polystyrene Reinforced Concrete Sandwich Panel for Energy Efficient Building Construction PDF

Preview

Summary

Proceedings of ICTACEM 2014 International Conference on Theoretical, Applied, Computational and Experimental Mechanics December 29-31, 2014, IIT Kharagpur, India ICTACEM-2014/446 Computational Study of Structural Insulated Polystyrene Reinforced Concrete Sandwich Panel for Energy Efficient Building ...

Description

Proceedings of ICTACEM 2014 International Conference on Theoretical, Applied, Computational and Experimental Mechanics December 29-31, 2014, IIT Kharagpur, India ICTACEM-2014/446

Computational Study of Structural Insulated Polystyrene Reinforced Concrete Sandwich Panel for Energy Efficient Building Construction D Bandyopadhyay1, Anubhav Roy2 and Arka Dey2

1*

Professor, Dept. of Construction Engineering, Jadavpur University, Kolkata, India Undergraduate Student, Dept. of Construction Engineering, Jadavpur University, Kolkata, India

2

ABSTRACT Structural insulated polystyrene sandwiched between steel reinforced concrete, panels proved to be one of the energy efficient and cost effective approach in recent time. The dynamic behavior of the above mentioned composite panel under different support conditions are discussed in the proposed paper .Finite element analyses are conducted by using ABAQUS simulations to compute the modal parameters of the plate under different boundary conditions. Computing the modal parameters of the undamaged plate, various type, degree and locations of damages are in duced on the panel and parameters are evaluated. Changes in modal parameters due to different damage and support conditions are studied and discussed in the paper. These parametric modal studies seem to be quite important for the development of an inverse dynamic technique to detect the structural damage in a polystyrene sandwiched RCC composite panels. 1. INTRODUCTION & OBJECTIVE Structural insulated panel made of polystyrene foam sandwiched between reinforced concrete seems to be a sustainable approach of building construction. It has several advantages including better heat and acoustic impedance, lower self weight, higher strength to mass ratio, better resistance to earthquakes and lower costs and construction time. There are literatures available both numerical as well as experimental study of structural insulated panel in other countries. However, the computational study of the same panel is quite sparse in India. The proposed paper discusses the structural behavior of these panels for different boundary conditions. The composite plate is analyzed in finite element framework on ABAQUS platform to compute the modal parameters. Different type and degree of damages are 1*

Corresponding Author, E-mail: [email protected], Telephone: (91) 9433068776(M), Department of Construction Engineering, Jadavpur University, Kolkata, INDIA.

induced on the undamaged panel by computation modeling and modal parameters (mode shapes and natural frequencies) are evaluated by simulations. Changes of modal parameters due to induction of damage for various support situations are studied and discussed to explain the dynamic behavior of the mentioned composite panel. The main objective of the proposed paper lies in the fact of determining two aspects as mentioned below : i) the “active” damage factors those may affect the local or global shape functions as well as natural frequency of vibration for a particular boundary condition ; ii) the most “sensitive” mode numbers those are affected by induction of damage for a particular boundary condition. These parametric modal study seem to be quite important for the development of an inverse dynamic technique to detect the structural damage of these types of RCC sandwich composite structural panels of low cost energy efficient building construction.

2. DESCRIPTION The basis of the structural insulated panel construction system is based on a series of foa m polystyrene panels sandwiched between welded steel wire meshes, subsequently applying concrete spray plaster after panel installation. The aim of this type of construction is to provide a system of industrialized modular panels having most of the advantages of pre-cast construction technique including feasibility of rapid erection and assembling compared to the conventional systems, repetitive cost efficient quality fabrication to comply with structural and load -bearing functions, in addition to the thermal and sound insulations and a wide range of shape sand finishes that may be achieved during the building process. The membrane action of the walls, particularly in lateral load resisting considerations seems to be an alternative material of low cost relatively better earthquake resistant system than conventional brick wall of same thicknesses. The material of the panel element is a foam polystyrene core which should be non-toxic, self-extinguishing and chemically inert and sandwiched

between

reinforced

concrete

of

varying

thicknesses depending on requirement. The reinforcement is electro welded steel wire meshes made of galvanized drawn Fig 1: View of Sprayed Wall Panel

steel wires placed on both sides of the polystyrene sheet and

connected by means of spot welded shear connecter of the same material. In our study the diameter of the wire used in both the direction is 3.0 mm @ 100 mm c/c. The grade of concrete is M25 with adequate workability for proper spraying. 3. MODELLING DETAILS The proposed sample was modeled in finite element framework on ABAQUS 6.10-1 by assembly and hence the interaction of parts as of detailing given below : a) Core :The polystyrene core has been modeled by “3D-SOLID-Extrusion” modeling due to its all dimensional degrees of freedom. The core was 1m X 1m in dimensions with 60mm thickness. The „polystyrene‟ section was assigned to the part with material properties detailing as given in Table-1. b) Top and bottom face-sheets: The top and bottom concrete face-sheets has been modeled by “3D-SOLID-Extrusion” modeling again due to its all dimensional degrees of freedom. . The face-sheets were 1m X 1m in dimensions with 35mm thickness. The „concrete‟ section was assigned to the part with material properties detailing as given in Table 1.

Fig 2: Core and face-sheets

Fig 3: Reinforcing bars and shear connectors

Table 1 Property

Concrete

Steel

Polystyrene

Density (N/mm3)

2.5e-005

7.85 E-005

1.5e-006

E

25000

206000

3000

0.1

0.3

0.34

(N/mm2)

Poisson‟s Ratio

c) Reinforcing steel bars: The axial properties of the reinforcing steel bars are predominant for the flexural behavior of the concrete slab. Thus, the axial behavior of the reinforcing bars is considered for the numerical model as „3D-WIRE-Planar‟ part in the sandwich plate. The cross sectional area

and the length of the bars are 7.07 sq. mm. and 1m respectively. The „steel‟ section was assigned to the part with material properties detailing as given in Table 1. d) Shear connectors : The cross-chain shear connecters are assumed to be axial members and hence the shear connectors were modeled by “3D-WIRE-Planar” modeling. The cross sectional area and length of the bars are 7.07 sq mm and 1m respectively. The „steel‟ section was assigned to the part with material properties detailing as given in Table 1. The “INTERACTION” between the parts was done by applying internal constraints as follows: a) Applying general “surf-surf rough” contact between core and face-sheets allowing finite sliding between the two; b) Applying “TIE” constraint between shear connectors and reinforcing bars; c) “EMBEDDING” reinforcing steel bars in face-sheets; d) “EMBEDDING” shear connectors through face-sheets and core. 4. RESULTS AND DISCUSSIONS All the natural frequencies and corresponding mode shapes are simulated for induction of different degrees and locations of damages mentioned for different boundary conditions as shown below: 1)Plate fixed at one face 2) Plate Simply supported at two edges of bottom face-sheet 3) Plate Simply supported at all four edges of bottom face-sheet 4) Plate Fixed at all four faces of bottom face sheet Undamaged

Mode 1 2 3 4 5

Table 2: Natural frequency for undamaged plate One end fixed Simply supported at Simply two edges supported at four edges Natural freq. Natural freq. Natural freq. (hertz) (hertz) (hertz) 1.7326 3.5688 3.7989 5.0189 5.1889

2.4452 3.0858 4.5552 5.9253 6.4226

3.7976 4.9696 5.2901 7.4767 7.7275

All faces fixed Natural freq. (hertz) 5.1612 8.5483 8.55 11.14 12.607

The first five natural frequencies of the undamaged plate, for a different boundary conditions are computed and tabulated in Table 6. There, as the constraints are increased the natural frequency of vibration also gets increased accordingly as we know that as external/internal stiffness is increased, the natural frequency of vibration gets increased.

The simulated natural frequencies due to different damage inductions on face-sheets of the plate (DF) for particular boundary conditions are studied and shown in Fig.4. Table 4: Damage in face-sheets (DF) Both face-sheets damaged to same extent along a circular geometry radiating from center Top face-sheet damaged to greater extent than the bottom along a square geometry radiating from center Only top face-sheet damaged along a square geometry radiating from center Only bottom face-sheet damaged at three regions along circular geometries located near any three edges

DF1 DF2 DF3 DF4

Fixing one face 2nd Mode

Simply supporting at two edges

3rd Mode 5 Natural frequency

Natural Frequency

4

1st Mode

3 2 1

1st Mode

2nd Mode

UD

DF1 DF2 DF3 Damage status

4 3 2 1 0

0 UD

DF1

DF2 DF3 Damage status

DF4

1st Mode

2nd Mode

DF4

Fixing all faces

3rd Mode

10 Natural frequency

Natural Frequency

Simply supporting at four edges 6 5 4 3 2 1 0

3rd Mode

1st Mode

2nd Mode

3rd Mode

8 6 4 2 0

UD

DF1

DF2 DF3 Damage status

DF4

UD

DF1 DF2 DF3 Damage status

DF4

Fig.4: Change in Natural Frequencies due to Different Types of Laminated Damage

It is observed from the study that the natural frequency of the sandwich plate is affected by various degrees due to the induction of damage in face-sheets irrespective of boundary conditions. The change is different for different mode and can‟t be generalized. Now for DF1 lesser region of face-sheets are damaged compare to DF2 resulting a decrease of natural frequency from DF1 to DF2. But in DF3 scenario, the damage is only made on the bottom face-sheet which is again affecting a lesser face-sheet region than that of DF2 where both face-sheets are damaged to some extent. Now for DB4, when three circular regions are damaged in the bottom face-sheet near three

different edges of the plate, a greater part of face-sheet lacks stiffness contributing to reduction of global stiffness and resulted in a noticeable decrease in natural frequency at all the modes. Similarly, the mode shapes of the plate for „simply supporting at four bottom edges‟ boundary condition for UD, DF1, DF2, DF3, DF4, DF5 damage scenarios are visualized as shown in Fig.5 to Fig 10. .

Fig. 5: Second mode shape of UD for simply supporting at four bottom edges

Fig. 7: Second mode shape of DF2 for simply supporting at four bottom edges

Fig. 9: Second mode shape of DF4 for simply supporting at four bottom edges

Fig. 6: Second mode shape of DF1 for simply supporting at four bottom edges

Fig. 8: Second mode shape of DF3 for simply supporting at four bottom edges

Fig. 10: Third mode shape of DF4 for simply supporting at four bottom edges

It may be visualized from the above second mode shapes as shown in Fig. 5 to Fig 10, that the second mode shape function have a local deformation behavior on the plate in t he damage regions

for reducing the stiffness at both face-sheets (DF1 & DF2) locally, but the localized deformation is obviously of higher range. On the other hand when a local asymmetrical damage is induced on the plate (DF3), local as well as global deformation response is identified as shown in Fig.8. It might be perceived that for induction of damage quite symmetrically on the plate the modal parameters as well as the displacement vector of the plate is affected only locally in the regions of damage due to no major variations in stiffness distribution; on the other hand an abrupt change in stiffness distribution for induction of damage in any one of the face-sheets may lead to a global change in modal parameters. On inducing damages on three circular regions of the bottom face-sheet (DF5) the stiffness reduces drastically but local deformations are observed in each hole in each of the first four modes but with higher modes, the deformation gradually diverges to a global one. Hence any type or location of face-sheet damage might be considered as an “active” one to affect the modal parameters of the plate. Similarly, different damages were induced on the steel bars and shear connectors of the plate (DB) and the simulated natural frequencies due to different damage inductions for particular boundary conditions are studied as shown in Table 3. Table 3: Damage in shear connectors and bars (DB) DB1 4X4 and 3X3 intersections of damaged main bars are made at centre in top and bottom respectively DB2 A shear connector from last row is damaged DB3 A shear connector from third last row is damaged DB4 All shear connectors of a middle row are damaged DB5 All shear connectors of TWO middle rows from either sides are damaged DB6 All shear connectors of FOUR middle rows from either sides are damaged DB7 All shear connectors are damaged The natural frequencies of the plate for induction of DB are tabulated below in Table 4 Table 4: 1st, 2 nd and 3rd mode natural frequencies for induction of DB One face fixed 1st Mode

UD

DB1

DB2

DB3

DB4

DB5

DB6

DB7

1.7326

1.7318

1.7325

1.7325

1.7326

1.7325

1.731

1.2713

2nd Mode

3.5688

3.5678

3.5686

3.5687

3.5687

3.5684

3.5647

1.2739

3rd Mode

3.7989

3.7975

3.7974

3.7976

3.7984

3.7977

3.7957

1.2878

It is noted that the natural frequency of the plate reduces very slightly for induction of damages in bars or shear connectors until a severe damage is introduced as marked “DB6”, the change being greater for the case of DB1. The slight change may be attributed to the greater frictional resistance

between the laminates. It may be perceived that until DB7, in spite of affecting the stiffness of few shear connectors or bars, the remaining undamaged shear connectors or bars were quite capable to resist the plate against modal disturbance leading to a slight changes in natural frequencies but in case of DB8 when all shear connectors get damaged, the global stiffness and hence the natural frequency of the plate reduced drastically. Now the first mode shapes of the plate for „fixing at one face‟ boundary condition for UD, DB1, DB5, and DB8 damage scenarios are visualized and shown in Fig.11 to Fig 14.

Fig. 11: First mode shape of UD for fixing one face Fig. 12: First mode shape of DB1 for fixing one face

Fig. 13: First mode shape of DB5 for fixing one face

Fig. 14: First mode shape of DB8 for fixing one face

From the above mode shapes, the same slight change is manifested that of natural frequencies until all the shear connectors gets damaged, the undamaged shear connectors or bars are capable of maintaining the plate against any abrupt deformation changes leading the dynamic responses to remain unaltered. But for the induction of DB8 when all shear connectors are damaged, local deformation collapse between face-sheets and core are observed as from Fig. 14. Actually the flexural characteristics may not be influenced greatly by the through thickness transverse behavior. So the modal parameters of the plate doesn‟t seem to be much sensitive or “active” for above mentioned damage in shear connectors or main bars until all the shear connectors or bars get

affected for the assumed numerical model. Thus these modal parameters are insufficient and insensitive to detect minor damages in steel bars or shear connectors. The simulated natural frequencies due to different multiple damages induced to the plate (DM) as shown in Table 5 for particular boundary conditions are studied and shown in Fig.14. . Table 5: Multiple damages (DM) DM1 DF2+DB1 DM2 DF2+DB6 DM3 DF2+DB1+DB7

1st Mode

Fixing one face 2nd Mode

Simply supporting at two edges

3rd Mode

1st Mode

2nd Mode

3rd Mode

DM3

UD

DM1 DM2 Damage status

DM3

3rd Mode

1st Mode

5 Natural frequency

Natural frequency

4

4

3

3

2

2

1

1

0

0 UD

DM1 DM2 Damage status

Simply supporting two edges 2nd Mode

10

Fixing all faces

2nd Mode

3rd Mode

Natural frequency

Natural frequency

6

1st Mode

5

8

4

6

3

4

2

2

1

0

0 UD

DM1 DM2 Damage status

DM3

UD

DM1 DM2 Damage status

DM3

Fig.15: Change in Natural Frequencies due to Different Types of Laminated Multiple Damage

From the above plots it is quite evident that natural frequency reduces for DM1 where both the face-sheets as well as the main bars get damaged. The global stiffness of the plate is affected largely resulting to the decrease in the natural frequency. For DM2, as discussed earlier the resultant effect of DF2 and DB5 may be perceived to have a less effect on stiffness of the plate resulted in lesser reduction in natural frequency. By comparing DM2 and DM3 both shear connectors and main bars‟ stiffness damage, the global stiffness falls abruptly from DM2 to DM3 as we can perceive from the above plots.

The third mode shapes of the plate for „simply supported at two edges‟ boundary condition for UD, DM1, DM2, DM3 damage scenarios are visualized and compared.

Fig. 16: Third mode shape for UD simply supporting two edges

Fig. 17: Third mode shape for DM1 for simply supporting two edges

Fig 18: Third mode shape for DM2 for simply Fig.19: Third mode shape for DM3 for simply supporting two edges supporting two edges

From the above third mode shapes of above Fig. 16 to Fig. 19, all the multiple damages affect the global as well as local behaviors of the plate. For DM1, the top face-sheet as well as the main bars embedded in the top face-sheets is damaged resulting in change in distribution of global stiffness thereby exhibiting a global change in the mode shapes. The undamaged shear connectors in association with the inter-lamina friction for DM2 as discussed earlier are capable enough to hold the excess shear of the plate and hence the local as well as global deformations are observed in regions of local damage in the face-sheets. Now in case of DM3, s...