Investigation on Integrity of Existing Circular Reinforced Concrete Silo with Opening

Oh Chai Lian Department of Structure and Material Engineering

Universiti Teknologi MARA, UiTM Shah Alam, Selangor, Malaysia chailian@salam.uitm.edu.my

Gan Sim Keat Department of Structure Engineering Sepakat Setia Perunding Sdn. Bhd.,

Seri Kembangan, Selangor, Malaysia

Abstract – This paper presents an investigation on the integrity of an existing circular reinforced concrete cement silo that subjected to a new 600 mm x 600 mm opening. The opening is introduced at the cement silo side wall for airslide installation to improve the aeration system of the silo. Three-dimensional finite element silo model was generated based on the available construction drawings of the existing silo by using SAP 2000 software. The integrity of the silo was investigated through results from stresses distribution, hoop tension forces, shear forces and bending moments focusing on silo wall, wall opening and silo base slab. Assessment of the existing silo capacity is summarized and the results compared with the integrity checks show inadequacy of the silo base slab under the modification work. Besides, high vertical stress concentration at the localised wall opening is expected with adopted safety factor of 1.24.

Keywords – Integrity; Reinforced Concrete; Silo; Opening

I. INTRODUCTION The importance of storage structures grow with the needs of

mankind in storing their powdery materials such as cement and sugar or granular materials such as coal and wheat. Tall solid storage structures are generally referred as silos; either in metallic or concrete and usually are cylindrical shape. Understanding the behaviour of a silo wall under the static pressure induced by the stored material is important for their design. Extensive studies have been carried out by analytical, experimental and computational analysis to investigate the stress distributions of the silo wall due to dynamic forces such as loading or unloading, winds, seismic and thermal effects. However, the study on the performance of silo wall with opening is lacking. Nowadays, the modification works on the existing silos become popular due to the advance storage technology. The engineers started to investigate the integrity of the existing silos after the modification work, such as introducing an opening on the silo wall. This paper presents an investigation on the integrity of an existing circular reinforced concrete cement silo that subjected to a new 600 mm x 600 mm opening. The opening is introduced at the cement silo side wall for airslide installation to improve the aeration system of the silo.

Silo walls are mainly subjected to hoop tension, axial compression, meridianal and circumferential bending moments; and radial shear forces. The stored material in the silo tends to pressure the silo wall causing occurrence of the

hoop tension while wall self-weight together with the friction between the storage material and silo wall caused the axial compression in the wall. Early years, silo designers assumed lateral pressure induced by a stored material on a silo wall to vary hydrostatically without considering the interaction between the two materials; which gave conservative results. However the approach is still widely accepted amongst silo designers since the approach is adopted by most silo design codes [1]. Two classic silo theories, namely Janssen’s theory [2] and Reimberts’ theory [3] are preferred in United States of America and some part of Europe respectively in determining the static lateral pressures. Briassoulis [4] found that the Reimberts’ theory overestimates the lateral pressure compared to Janssen’s theory; however this is highly dependent on the silo geometry and the characteristic of the stored material.

There are several literatures available on the investigation of the silo strength and behaviour. Mohamed, Moore and Tarek [1] investigated the behaviour of ground-supported concrete silos that filled with saturated granular material by finite-element simulation. Gruyaert, Belie, Matthys, Nuffel and Sonck [5] investigated the structural stability of horizontal silo wall with straight and L-shaped wall panels experimentally. Prato and Godoy [6] studied on the bending stress of multi-bin reinforced concrete cylindrical silos. Martinez, Alfaro and Doblare [7] compared the induced pressure distribution in metallic silos with different standards. Vidal, Gallego, Guaita and Ayuga [8] performed finite element analysis on the filling of cylindrical steel silos with eccentric hopper. Elghazouli and Rotter [9] presented analytical investigation into failure in circular silo while Gurfinkel [10] studied the solution for strengthening impaired concrete silo.

In this paper, a numerical investigation into the capacity of the existing cement silo with the new opening is summarized. The attributes to the three-dimensional silo model such as the geometry modelling, material properties, static lateral pressure exerted by the stored material on the silo wall, boundary condition and analysis method by SAP 2000 software are generally discussed. Analysis results obtained such as stresses distribution, hoop tension forces, shear forces and bending moments on silo wall, wall opening and silo base slab are then compared with the assessment of the existing silo performance.

2010 International Conference on Science and Social Research (CSSR 2010), December 5 - 7, 2010, Kuala Lumpur, Malaysia

978-1-4244-8986-2/10/$26.00 ©2010 IEEE 298

II. METHODOLOGY An independent structural assessment was carried out to

check the adequacy of an existing cement silo with introduction a new 600 mm x 600 mm opening at the silo wall and installation of the steel cone frame on the silo base slab. The investigated reinforced concrete silo, denoted Silo no 3 is part of a twin circular reinforced concrete silo that were constructed in Perak, Malaysia, year 1984 by a turnkey contractor from Korea. The silo is approximately 53 m in height up to silo roof with a storage capacity of 10,000 tonnes. The internal diameter of the silo is 16 m with wall thickness of 350 mm. There is a reinforced concrete base slab with 1.5 m thick at level EL +95 300 mm that supported by 12 columns and separated from the silo wall proper. A new 600 mm x 600 mm opening at 8.107 m above level EL +83 200 mm will be made at the side wall of the existing silo for airslide installation to allow bulk loading as shown in Fig. 1. Steel framed cone will be installed on top of the base slab at EL +95 300 mm in order to improve the aeration system of the silo.

In this project, SAP 2000 computer software had been selected for the investigation due to its flexibility in geometric and analysis modelling. SAP 2000 features powerful graphical user interface and has latest numerical techniques and solution algorithms. A three-dimensional finite element model of the silo was performed using SAP 2000 based on available structural drawings and information obtained from the Client. The 53 m height circular cement silo was generally modelled by 500 mm x 500 mm square mesh and further refined to 200 mm x 200 mm square mesh near the opening.

The silo model was pinned supported at the base of the twelve columns. The supports were constrained in all translation Degree of Freedoms (DOFs) to avoid any moments induced to the piles foundation system underneath. Fig. 2 shows the typical three-dimensional meshed silo model.

Janssen’s theory was chosen to determine the lateral and vertical pressure induced by the stored bulk solids on the wall. Janssen’s silo theory suggested the horizontal pressure equation as

ρ = (γR / μ’) [1-e-μ’kY/R] (1) and vertical pressure as

q = ρ / k (2) where the k could be expressed as

k = (1-sin φ)/(1+sin φ) (3) in which γ is bulk density (kN/m3), R is hydraulic radius (m), e is eccentric discharge (m), μ’ is friction coefficient, k is Rankine’s coefficient of active earth pressure, φ is the repose angle (o) and Y is stored height (m) [11]. This theory accounts for the arching effect within the bulk solids and allows for transfer of part of their weight to the wall as axial compression [1]. Only gravity action by the stored material and silo self-weight were studied. The roof of the silo was excluded from the model; therefore the vertical uniformly distributed design loads from the roof were assigned to the silo wall. The impose pressures on the wall were gradually increasing with depth of silo. The loading on the silo base slab based on weight of fill material with consideration for wall friction was 439.25 kN/m2. The interaction between the silo wall and the

Figure 1. Location of the proposed new air slide for bulk loading (Source:SSP)

299

bulk solids contained was not considered in the modelling but the friction coefficient between the two materials was taken account in the Janssen’s Theory. The concrete was modeled with Young’s modulus, Ec = 25 GPa and Poisson’s ratio, vc = 0.30. The stored material parameter adopted were bulk density, γ = 16 kN/m3, friction coefficient, μ’ = 0.466, repose angle, φ = 20o and stored height, Y = 39.9 m.

A static linear analysis was performed to study the integrity of the silo components, mainly the silo wall, wall opening and silo base slab through stresses distribution, hoop tension forces, shear forces and bending moments. The finite element analysis results obtained were later compared with the existing silo capacity. The existing silo capacity such as tension capacity, shear capacity, moment capacity and compression capacity were calculated based on the material properties of the silo with referring to BS 8110: Part 1: 1987 [12]. Assessment of the existing silo capacity was summarized and the comparison with the SAP 2000 integrity checks results enable the determination on the adequacy of the silo components under such modification work.

III. RESULTS AND DISCUSSION The integrity of the existing circular reinforced concrete silo

after the proposal of new opening for airslide is examined here. The results will be focusing on three silo components, mainly the silo wall, wall opening and silo base slab.

A. Silo Wall It is found that the maximum hoop tension force with value

2896.73 kN/m occurs at silo wall at approximately level EL +97 400 mm is less than the respective existing horizontal tension capacity with value 3525 kN/m. The hoop tension forces are 20-60% lower than the existing horizontal tension capacity provided along the height of the silo wall. So, no apparent deficiencies are found. Fig. 3 shows the comparison

between distribution of hoop tension forces from the finite element analysis (FEA) results and existing horizontal tension capacity of the silo.

Besides, the vertical compression stress is approximately 14% less than the existing compressive strength of the wall. Furthermore, bending moments and transverse shear forces along the silo wall height are presented in Fig. 4(a) and (b). These results are compared with the existing bending capacity and shear capacity of the silo wall in Table 1. It is found that all bending moments and transverse shear forces along the silo walls only achieve approximately 12.5% of the existing silo wall capacity which lie on the conservative side.

TABLE I.

Comparison of shear forces and bending moments of the silo wall

Height (m)

Shear Forces (kN/m)

Existing Shear

Capacity (kN/m)

Bending Moments (kNm/m)

Existing Bending Capacity (kNm/m)

0.00 30.36 114.86 0.19 77.69 2.10 22.93 114.86 50.31 77.69 8.40 24.69 215.17 63.65 510.68

12.60 17.68 215.17 30.69 510.68 16.80 3.56 215.17 6.27 510.68 23.10 0.76 215.17 2.67 510.68 25.20 1.13 183.42 1.15 316.33 29.40 0.40 177.08 1.46 284.68 31.50 1.39 177.08 3.07 284.68 33.60 1.23 177.08 1.02 284.68 37.80 0.51 177.08 1.09 284.68 44.1 2.44 154.71 1.23 189.82 48.3 1.29 154.71 2.85 189.82

Figure 2. A typical finite element model

Figure 3. Comparison between distribution of silo hoop tension force

___ and silo hoop tension capacity ------

300

B. Wall Opening There is high concentration of vertical compressive stresses

at the corner of the proposed new 600 mm x 600 mm opening located at 8.107 m above level EL +83 200 mm. The vertical compression capacity of the existing 350 mm thick wall is approximately 23 N/mm2. The FEA results show the compressive stress near the opening varies from 23.5 N/mm2 to 29 N/mm2 as shown in Fig. 5. The total design load factor varies from 1.24 to 1.54 reveals that the existing wall is able to resist the vertical compression stresses created by the proposed new wall opening.

Furthermore, concentration of horizontal tension stresses of 5.7 N/mm2 above and below the edge of opening is noticeable. The tension stresses are further checked against the tension capacity provided by the silo wall. The horizontal tension

capacity provided by the existing horizontal steel bars of D29–150 c/c is approximately 10.07 N/mm2 which is capable to resist the horizontal tension stresses located above and below the opening.

C. Silo Base Slab The maximum sagging and hogging moment of the silo

base slab between columns are approximately +2400 kNm/m and -1200 kNm/m respectively as shown in Fig. 6(a). Sagging and hogging moment capacity of the existing silo base slab is +1189 kNm/m and -2379 kNm/m, where the sagging moment capacity has exceeded. By considering 30% moment re-distribution [12], the maximum sagging moment of the silo base slab has reduced from +2400 kNm/m to +1680 kNm/m and the maximum hogging moment has increased from -1200 kNm/m to -1920 kNm/m. Thus, the slab has a safety factor of 1.95 for the hogging moments and 1.11 for the sagging moments. It is observed that the sagging moment capacity barely exceeds the redistributed sagging moment marginally without adequate safety factor of 1.5.

Meanwhile FEA results show the maximum shear stress occurs at the distance of effective depth, d away from the column face with value of 0.17 N/mm2, and this is less than the existing slab shear capacity, 0.34 N/mm. The base slab therefore has sufficient shear capacity. Fig. 6(b) shows the shear stresses near the column at the base slab.

IV. CONCLUSIONS Based on the investigation, the following conclusions are highlighted: 1. The hoop tension forces results obtained from the three

dimensional finite elements modelling are 20-60% lower than the hoop tension capacity of the silo wall shows the adequacy of the existing silo wall.

Figure 4. Results obtained from the finite elements solutions: (a) wall transverse shear force, and (b) wall bending moment

Figure 5. Vertical compressive stresses at opening

301

2. Bending moments and transverse shear stresses along the silo walls are not exceeding the capacity provided by the silo walls and no apparent deficiencies are being found.

3. The high concentration of vertical compressive stresses at the localised corners of new 600 mm x 600 mm opening has exceeded the wall capacity. However, it is found that the silo walls are still performing with a total design load factor of 1.24. Opening at the silo wall is technically feasible based on the analysis results.

4. The slab finite element results show that the shear stresses at the slab is within the capacity. However, the sagging moment of the silo base slab has exceeded the slab capacity even with the 30% moment re-distribution consideration. A marginal safety factor of 1.11 for the sagging moment reveals the inadequacy of the slab. Strengthen for the slab is recommended to accommodate this modification work.

ACKNOWLEDGMENT This writing would not be possible without the information

and support from the following organization, Sepakat Setia Perunding Sdn. Bhd., Faculty of Civil Engineering and Research Management Institute of Universiti Teknologi MARA.

REFERENCES [1] AF. T. Mohamed, I. D. Moore, and AF. T. Tarek, “A numerical

investigation into behaviour of ground-supported concrete silos filled

with saturated solids,” International Journal of Solids and Structures, vol. 43, pp. 3723-3738, 2006.

[2] H. A. Janssen, “Getreidedruck in silozellen,” Z. Ver. Dtsch. Ing., pp. 1045-1049, 1895.

[3] M. L. Reimbert, and A. M. Reimbert, Silos-theory and practice, first ed., Trans Tech Publications: Clausthal-Zellerfeld, 1976,

[4] D. Briassoulis, “Finite-element analysis of a cylindrical silo shell under unsymmetrical pressure distributions,” Proceedings of Advances in

Civil and Structural Engineering. Civil-Comp Press, Edinburgh, 1998, pp. 31-39.

[5] E. Gruyaert, N. D. Belie, S. Matthys, A. V. Nuffel, and B. Snock, “Pressures and deformations of bunker silo walls,” Biosystem Engineering, vol. 97, pp. 61-74, 2007.

[6] C. A. Prato, and L. A. Godoy, “Bending of multi-bin rc cylindrical silos,” Journal of Structural Engineering, vol. 115, No. 12, pp. 3194-3200, 1989.

[7] M. A. Martinez, I. Alfaro, and M. Dolblare, “Simulation of axisymmetric discharging in metallic silos. Analysis of the induced pressure distribution and comparison with different standards,” Engineering Structures, vol. 24, pp. 1561-1574, 2002.

[8] P. Vidal, E. Gallego, M. Guaita, and F. Ayuga, “Finite Element Analysis under different boundary conditions of the filling of cylindrical steel silos having an eccentric hopper,” Journal of Constructional Steel Research, vol. 64, pp. 480-492, 2008.

[9] A. Y. Elghazouli and J. M. Rotter, “Long-term performance and assessment of circular reinforced concrete silos,”Construction and Building Materials, Vol. 10, No. 2, pp. 117-122, 1996.

[10] G. Gurfinkel, “Restoring an impaired concrete silo,” Journal of Performance of Constructed Facilities, Vol. 3, No. 2, pp. 87-99, 1989.

[11] S. S. Sargis and C. H. Ernest, Design and Construction of Silo and Bunkers, Van Nostrand Reinhold Company Inc., 1985.

[12] British Standards Institution, BS 8110: British Standard Codes of Practice for Design and Construction, London, 1987.

Figure 6(a) Sagging and hogging moment at silo base slab; (b) Shear stress at distance of effective depth, d away from column

302