You are on page 1of 10
ACI STRUCTURAL JOURNAI TECHNICAL PAPER Title no. 92-821 Designing Silo Walls for Flow Patterns* He & by John E. Sadler, F. Thomas Johnston, and Mostafa H. Mahmoud A procedure for design of concrete silo walls subjected to asymmetric flow patterns is presented. Differences beween concentric and asymmetric flow {or both the mas flow and the funnel flow cases are discussed. A method of ‘computing differential wall pressures resulting from flow channels is pre sented along with guidelines for selecting design flow channel geometry ‘and creaing a practical silo design for asymmetric flow. Design aids (hased on finite element modeling) for estimating circumferential moments ‘are included, Two silo design examples are presented; one for grain and ‘one for coal. Keywords: bending: bins: conerete construction; eccentric loads finite element method; granular materials; loads (Forces; reinforced concrete silos; stresses: structural analysis; walls, Silo codes! have traditionally recommended design pres- sures that apply a factor, usually called C, to an initial filling pressure (typically Janssen's®7) to cover the increased pres- sures during discharge. Most codes have limited their recom- ‘mendations to concentric flow pattems and have offered little guidance for asymmetric flow patterns, While Cy fac- tors are useful to cover the concentric case, a different ap- roach is necessary for the asymmetric case. This. paper discusses both the concentric and asymmetric cases but fo- uses on a method for direct design of walls for the asymmet- ric flow case.* FLOW PATTERNS Flow patterns that exist in silos are illustrated in Fig. 1. For Purposes of wall design, we can say mass flow exists when all material is in motion and funnel flow exists when only a Portion of the material is in motion. Funnel flow patterns where flow channels come near (Fig. 1(K)] or contact the wall asymmetrically [Fig. 1(G), (H), (I, and (J)] require spe- cial analysis. In cases where silos are unloaded from the top (Fig. 1(F)] or where flow channels are remote from the wall (Fig. 1(E) ‘and (L)], the wall pressures during unloading are not unlike the initial filling pressures, Initial filling pressure measure- ‘ments in full-size concrete silos*'? and metal or plexiglas ‘This approach is beng considered & potential aditon Wo revised version of ACI SI3°9. "Standard Pratie for Deng il Consrvtion of Conte Sits and Stacking Tes for Storing Granlar Matera" Comments ois pper ected aad published in SC Sacral Jour! wil be cnssdoted dung Commitee 31 de sion of he proposed evo, ACI Structural Journal / March-April 1995 models'*"? show localized peaks, valleys, and patches of higher pressure that deviate from Janssen's curve. These ap- pear to result from the settling material reacting to an imper- fectly shaped and irregularly surfaced cylinder, and from segregation during filling Such pressures peak slightly higher than Janssen's and for conventional slip or jump-formed concrete silos can be cov- ered by a C, factor of 1.35. The resulting extra hoop rein forcement, especially when placed in two faces, provides extra tension and moment capacity to cover the localized bending effects of patches of higher load. Alternatively, one may compute moments from a patch- type loading” which simulates the pressure peaks. Mass flow case During mass flow (Fig. 1(A)], the downward movement of the material is like the consolidation or settling movement during initial filling, but with higher velocity. In the mass flow regions of instrumented full-size’? and model!” si- los, the locations of pressure peaks did not change from their locations during filling, but their intensities generally in- creased. Some codes** recognize mass-flow pressure in- creases above the transition can be covered by a C, factor of 1.50 applied to Janssen’s pressures. Wall pressures at the top of the hopper (transition) are higher and must be computed using methods for converging flow channels. Walker's method" seems to predict verti- cal and horizontal pressures that are suitable for use in design of concrete silos. Other methods?"® predict higher spike pressures atthe transition, but examinations by the authors of wall distress, at least in concrete silos, indicate the stored ‘material is unable to impart such higher pressures. Funnel flow cases Where the flow channel is concentric and does not contact the wall (Fig, 1(B)], the C, factor of 1.35 can be used. Where the flow channel expands to the wall (Fig. 1(B), (C), and ACI Simetwral Journal V9, No.2, Mach Ape 1995, eco Sp. 19198 ad rived ut att lication pls, Coy fp 1998, Arnencan Conctet laste Alright esced.nludag te alee 1 copies anes permission s baned om he copyright popitrs:Perunen it usstn wil be published inthe January etme 1996 ACT Stracaral Journal received by Sep 11983, 219 ‘John E.Sadler, PACK, a concatantin Columbus, Oba anda menber of ACI Com ‘mince 313. Concrete Bins and Silos, He har more than 25 years of experince in ‘design and investigation of sles and has authored many arcs and technical bupers He ha beon the recipient of several anand, including the Delmar L. Bloem Distinguished Sesce Award for hs work past Charman of ACI Commitee 312. [ACI member E Thomas Johnston is President of Failiy Design nc, Combs, Ohio, a coming enincering frm specializing i the design of material handing land sorage facies, He te member of ACI Commitee 313, Concrete Bins and Sos and received his BS and MS degres fom Oho State Universi. He has nest ‘ated several hundred silo with low elaed problems and corrected man problems ‘bymodping Now pats. He has published several papers onl flares and prob sont. ACK member Mostafa H. Mahmoud ix Presiden of Slo Engineers & Consultan In, Columbus, Oka, He is Chairman of ACI Comminee 313, Concrete Bins and Sis, and received is BS from Caro University ond his MS and PRD degrees rom ‘Ohio State Univers He has more than 20 years of experienc in design and invest tein of silos and has published several ates om slo design and asymmetric flow (D)], the higher C, factor of 1.50 should be used at and above the transition. Ifthe asymmettie flow channel contacts the wall or has the possibility of tilting and contacting the wall [Fig. 1(G), (H), (D, and (), an analysis of pressure differentials and result- ing silo wall bending moments must be made. Ifthe channel is asymmetric but does not tilt to the wall [Fig. 1(K)}, an analysis for bending may still be necessary. Failure to recog- nize and provide for wall moments resulting from asymmet- ric flow channels, especially when they contact the wall, has Jed to numerous and serious problems!*! and has led many authors?27* to discuss solutions. DESIGN FOR ASYMMETRIC FLOW ‘The steps for designing the wall for a particular asymmet- ric flow channel are: 1) Determine hoop tensions due to ini- tial filling pressures; 2) Determine pressures in the flow channel; 3) Determine pressures in the static material; 4) De- termine circumferential bending moments in the wall that re- sult from the difference in horizontal pressure between the flowing and static material; and 5) Reinforce the wall to re- sist the circumferential tensions and moments. Selection of a flow channel configuration to be used for design is important. The recommended flow channel geom- etry will be shown in Fig. 3 and discussed later. Pressure in flowing material ‘The pressure within the flowing material after the flow channel has formed is different in the converging portion than in the constant diameter portion of the flow channel. ‘The converging portion of the flow channel below the transition (Fig. 2) may be treated as a hopper that has formed within the material, and vertical pressure q, can be approxi- mated using Walker's? mass flow formula for q, (Eq. (A-1) in the Appendix] with the following modifications. 1. Use 8 instead of © for the hopper angle. @ may be es- timated from Fig. 2, which shows the relationship between Sand 0 ;established by Giunta.” 2. Use ¢ instead of ¢’ for the hopper friction angle. If part of the surface is material and part of the surface is silo wall, use a weighted average for 6 220 ‘The portion of the flow channel with constant diameter 7 in Fig. 2 may be treated as a silo within the silo. The vertical pressure gycan be approximated using Janssen’s equation! ®? ‘with the following modifications 1. Use an R (R= A,/P)) appropriate for the flow-channel diameter rather than the silo diameter. 2. Since the material in the flow channel is flowing against the static material, use a wall friction coefficient of tan @. If the flow channel is partially exposed to the silo wall, use a weighted average of tan @ and tan 9”. ‘The horizontal pressure in the flowing material, where the channel is nonconverging, can be approximated by 9, = kay a ‘The ratio k in Eq. (1) may be taken as k ing @ Where the flow channel is converging and tilted with one side against the vertical silo wall, the authors estimate the horizontal flow pressure against the wall by substituting (@, +6) for o in Eq. (A-6) in the Appendix, Pressure in static material ‘The vertical pressure q, in the static material can be ap- proximated by summation of vertical forces at the level in question as follows. 1. Determine vertical pressure g; due to initial filling of the silo and the reduced vertical pressure gy in the flow chan- nel(s) after flow is started. 2. From the plan geometry defined in Fig. 3, determine Ay or if multiple flow channels, the summation of Ay ‘The vertical and horizontal pressure in the static material can be approximated by @) P= ka, @) where kis given by Eq. (2). Determine circumferential wall moments and tensions To obtain moments, apply the pressures py and p, as out- ‘ward radial forces on the silo wall as illustrated in Fig. 4(a) and use a method of analysis that recognizes the resistance the static material provides to the otherwise free deform: of the ring. To do so, a stored material modulus of elasticity Eycan be included in the analysis (ike the subgrade modulus in the classical “Beams on Elastic Foundations” problem). ‘Methods of analysis such as the ring formulas in Roark”? that do not recognize the subgrade modulus effect yield un- realistic deformations and overestimate the moments. For example, without support from the static material, the com- puted deflection of a 12-in, (305-mm) wall on a 70-ft-(21.3- m)-diameter silo with two 18-ft (5500-mm) flow channels, is ACI Structural Journal / March-April 1995 ass FLOW FUMEL FLOW ASYIMETRIC PATTERNS © FLOW IS NOT IN CONTACT WITH HALL Fig. 1—Flow patterns CRITICAL RATIO. q {FLOWING “4 ee) Q DEGREES TRANSITION: Fa Ba 3 g 0 20 30 40 50 60 70 ied vie DELTA (3) DEGREES RELATIONSHIP OF 8 TO 8 VERTICAL CHANNEL Fig, 2—Typical flow angles and channel shape per Giunta? ACI Structural Journal / March-April 1995 221 Yur. 0° 2h't tan Oy/cos Ot 4 $e FoR oy aa rt AUR fee . : ee al wac 4 i Aawite 7 bot ca oe ny = TAQH S089 : 1 \ Mo oa Ot aa BT 10 ws wok 11 47 «4 Fo a SION AID FOR LaF PF BETVEN LEVELS EB LEVEL 8 weno € = ten Oyish Oy 5, Dicos @ 60. No" Zien er fe \0 Fig. 3—Recommended design flow channel geometry 15 in, (381 mm), as shown in Fig. 5. Such large deformations are not supported by field measurements, ‘The authors use a linear elastic finite element model that approximates the horizontal interaction between silo wall and static material. A typical model for symmetric arrange- ‘ments of flow channels (two opposite channels) is construct- ced as illustrated in Fig. 4(b) and described in the following paragraphs. For arrangements of flow channels that are not symmetric, an analysis of the entire silo may be necessary. 1. A horizontal slice (usually of unit height) of the silo and its contents is modeled using three-dimensional finite ele- ments. Use of a slice model recognizes that the wall resists differential pressures primarily by horizontal spanning ac tion for most of the silo height. Near the roof and foundation, a different approach may be required to recognize radial re- straint. 222 2. The silo wall is modeled using three-dimensional shell elements with vertical and circumferential stiffnesses equal to E, and E,, respectively. The wall is assumed uncracked, but the effects of cracking can be investigated by using a re- duced wall modulus or thickness. 3, The static material is modeled using three-dimensional solid elements. The elements are assigned a modulus of elas- ticity , in the range of 50 to 100 ksf (20 to 40 MPa). Be- cause the wall is significantly stiffer than the material, Poisson's ratio is assumed to be A +k) where kis computed by Eq. (2). The analysis is not sensitive to the exact values chosen for Ey. 4, The material within the flow channel boundary is as- sumed to have no stiffness and is, therefore, modeled as a hole, ACI Structural Journal / March-April 1995 HIGHER ps PRESSURE stare patent. RADIAL DESIGN LOADS FOR WALL Fig. 4—Method of loading silo wall 5. Rather than apply the horizontal static pressure p, di- rectly to the walls, the vertical static pressure q, is applied to the upper face of the solid elements. This results in a more realistic distribution of p, on the wall. 6. The horizontal flow pressure p, is applied as a radially ‘outward pressure acting on the flow-channel boundary. 7. Because of small variations in vertical pressure from top to bottom of a silo slice of unit height, the base of the model is assumed restrained against vertical translation. This is equivalent to applying equal vertical pressures to the top and bottom. 8. Boundary conditions applicable because of symmetry are applied along the x- and y-axis. 9. The model is analyzed to determine the variation of ‘bending moments, hoop tensions, and deflections around the perimeter of the silo. Analysis shows, as indicated in Fig. 5, the maximum Mjp ‘occurs at the center of the flow channel, and the maximum ‘Mo, occurs outside the flow channel. Maximum deflection ‘occurs at the center of the flow channel, and hoop tensions decrease in the zone of the flow channel. However, itis rec- ‘ommended that the full F; tension rather than reduced ten- sion be used for design, Wall moments for single- and double-flow channels can be estimated using Fig. 6. The curves resulted from a series of finite element runs using the previous model. The param- eters D/h and L,/D were varied to compute a coefficient C) for the following Mrorar= C\(P,-P)D™ (3) ‘The moment Mjp is determined by multiplying Mrora. by the coefficient C; from the table of Fig. 6. The coefficient C, shows that for smaller L,/D ratios, Myp-is equal to or greater than Mop, in which case, more hoop reinforcement is re- uired on the inside than outside face. It is desirable, especially for slip-formed walls, to use the same reinforcement in both faces. Section 8.4 of ACI 318 ACI Structural Journal / March-April 1995 bs FINITE ELEMENT MODEL recognizes the possibility of formation of plastic hinges at points of maximum stress and allows redistribution of mo- ‘ment from the negative to the positive zone. The authors have found that plastic hinges do occur and that many times each face of the wall can be designed for a redistributed mo- ‘ment equal to the average of Myr and Mop. If the reinforce- ment is too heavy and the wall thickness must be increased, new moments can be computed from Fig. 6 using an updated Di ratio. Even though moments need not be considered below Lev- el A of Fig. 3, itis recommended thatthe extra reinforcement at Level A be decreased linearly to zero at the bottom of the pressure zone, ‘SELECTION OF DESIGN FLOW CHANNEL. CONFIGURATION Before the preceding analysis can be made, a “design” flow channel must be selected. Fig. 7 presents questions that raise the issues and help determine if bending from asym- metric flow is likely. The following also raises issues related to material properties, segregation, and geometry. As differ- ent potential flow channel sizes are considered, recognize that larger channels do not necessarily result in more mo- ‘ment than smaller ones (Fig. 6). Even though theoretical ‘considerations are important, final selection should be based ‘on actual flow channels observed in comparable silos or from evidence (wall abrasion, scour marks, cracking pat- terns, etc.) of actual flow channels having been active in comparable silos, if at all possible. ‘Segregation influences flow channel configuration Segregation is the primary determinant of whether a flow channel will tit. Although industry practice has been to as- sume center-filled silos with center outlets do not segregate enough to flow asymmetrically, such segregation does occur” [Fig. 1()]. Segregation occurs during filling when the trajectory off the conveyor or sloping spout causes the heavier particles to ‘ravel farther than the fines. Even if flow is baffled to stop the 223 Mop = 111 KFT Mig = 160 KFT (217 Kem (381m) ‘STATIC MATERIAL WITH T¥O 18° (5500 mm) FLOW CHANNELS AND 12° (305 mm) WALL CUNCRACKED) 0.7 IN (48 em) wat 11 FROM ‘STATIC MATERTAL Fig. 5—Comparison of wall moments and movements ‘trajectory, material rebounds and falls in a segregated pattern or the bafile wears through unexpectedly. The coarse material is typically less cohesive (weaker) than fines, so flow channels tend to tilt toward the coarser ‘material, which is usually against the wall. When contact is made with the wall, the low channel size often remains un- changed from the contact point to the top of the silo. Unless the material is uniform-sized (no fines), segrega- tion almost always occurs across the diameter regardless of ‘how the silo is filled. Therefore, any flow channel, especially if off-center, should be assumed to tlt toward the wall, Geometry influences flow channel configuration Fig. 1 illustrates how geometry influences flow channels in a variety of ways. Fig. 2 provides information to help size the channel. Flow channels develop by expanding from the ‘outlet size at an angle 6 , with the vertical. Depending upon the amount of cohesion (value of 8) and silo proportions, they may or may not reach a stable diameter. Investigations of flow channel development in laboratory- size model silos?" show a variety of results. For cohe- sive materials like coal, Giunta's work” seems to closely predict flow channel angles and sizes in full-size silos. Other studies may do better for noncohesive materials like grain. Fig. 3 defines the geometry of the recommended design flow channel for the tilted condition (9, > 0). The geometri- cal relationships of Fig. 3 result from the assumption that the flow channel tilts toward the wall and expands from its apex at arate equal to an included angle of 20 , a8 measured in a direction perpendicular to the plane of the paper. When the channel centerline hits the wall (Y= chord np) at Level B, the 224 channel must deviate away from the wall to continue to ex- pand to its maximum size at Level C. The channel then ex- tends the full height of the silo, over which and L,, remain constant, Ifthe P exceeds D [Fig. 1(H)], the analysis for asymmetric bending need only be done between Levels A and C using procedures described for the converging portion of the flow channel. Generally, the flow channel does not tilt more than 0 However, if the hopper surface adjacent to the outlet is suf- ficiently smooth and steep, and exceeds a= arctan (4tan9") o it is recommended the channel be assumed tilted to flow on the hopper surface adjacent to the outside of the silo. Then fi’ ‘equals the height of the actual hopper. Flow can also expand at a rate greater than the included angle of 20, in the direction perpendicular to the plane of the paper, if the hopper configuration permits such. Geometry dictates that as the number and size of flow channels around the perimeter of a multiple-outlet silo in- crease, the pressures in the static material and resulting mo- ments. increase significantly. However, when the flow channels increase to the point of merging, the moments de- crease, provided all feeders are active. If not, severe checker- board loading can occur. Itis, therefore, important to consid- cer realistic combinations of flow channels. Do not arbitrarily assume all feeders will be interlocked and operated together ACI Structural Journal / March-April 199 wo a +032] Sy oa vo orosiTe tow CHANNEL. 024] 022 +020} ove Cy og Lire none 4 7 heats a Fig. 6—Moment coefficients Sometimes, also, a hopper opening fails to become fully active because of a partially open gate, or a malfunctioning feeder or improperly proportioned feeder skirting. In such ses (Fig. 1(J)], flow immediately becomes asymmetric in the hopper, and an unexpected flow channel can form at the silo wall Variable material properties influence channel configuration Base the design flow channel on material properties repre- sentative of the material stored and the conditions of storage. ‘The material property that most dramatically affects the flow channel geometry is 8, which is a measure of cohesion. More cohesive materials have a higher 6. For high values of 8,0, approaches zero, and steep flow channels or ratholes occur. 8 varies with vertical pressure and approaches @ in the bottom of large silos where pressures are high. If segregation exists, 8 also varies and approaches @ where the coarse ma- terial is located, which is usually near the silo wall. Fig. 2 shows various granular materials and a range of 6 for each, ‘compared to Giunta’s curves ‘The values for 6 and 9’ also vary. Explore a range of rea- sonable assumptions for each ‘SERVICE LIFE CONSIDERATIONS Consider the following suggestions to improve the life and ‘maintenance of concrete silos. Owners and engineers have, ‘on many occasions, cooperated to design silos that lessen the severity of or even prevent asymmetric bending in their silos, 1. Strive to avoid configurations that have only off-center, outlets (Fig. 1(H), (I), (K)]. Try to incorporate a center outlet and use it as the primary means of emptying the silo. 2, Where outlets near the perimeter are necessary, try to lo- ‘cate them away from the wall to minimize flow channel con- ACI Structural Journal / March-April 1995 [ERS Jo [oar ne ram borer ee mr a faa ne sua we TET aon omar [vo fs we ro poe Fe foe vem [sp [OS We ie we we eco wer i seat ths ast te re pert — a Te a as Oe Sto inc Ars fe $s Toe owen [zen ha we) fers pe ie » here aw |_fiar oiee VL TeiJommet onc lea re ee a fone rox one Foe fsmetnc rw is ene jeri. fSunat "YeMETRE LONG CHE FG, 2X VL ACL RL. DG NE MATE ‘err LB one TY oF Fo TOE Fig. 7—Logic chart for determining if bending from asym- ‘metric flow is likely tact with the wall. This will reduce the forces on the wall and decrease the abrasion. 3. Consider limiting the use of outlets near the wall to those times when the level of filling is low. Consider more than the normal concrete cover over inside face reinforce- ment subject to heavy abrasion. 4. Avoid hopper configurations that can cause unpredict- able flow patterns or unexpected merging of adjacent flow channels. 5. Be cautious about using post-tensioned reinforcement for resisting bending from asymmetric flow. Compressive stresses can become critical under heavy bending plus axial load, 6. Reach an understanding with the owner regarding im- portant silo design and operating parameters for the silo, Record decisions on the drawings, such as: a, Stored material design parameters. », Design flow channel configuration. c. Basis for selection of hopper, such as flow studies, ete 4. Normal operating conditions. ¢. Emergency operating conditions, if one or more of the gates or feeders is inoperable or are partially open, causing hopper outlet to be only partially active. DESIGN EXAMPLES ‘Twosilo designs are presented in Fig. 8 and 9 as examples. The general procedure for an actual silo is as follows: 1, Establish values for 7.9.8.6. cones M's and k. 225 $20.5‘ Y=26" LEVEL ¢ GRAIN SILO (NO HOPPER) fy = 60 ksi YH [33_[er hy) fo 128 Vet) 26 [17 Ty (ft) [31.5 [18 Ty 70 —_I|1-05 [0.60 [0.31 [0.24 Agttt?) [404 [144 [37 [26 X {17.5° | 1 Hl |) PER ACI 313-91 CONCENTRIC FLOW fox 7 120° Py cet) [rz [as [2a [18 g 29.7 {30.1 [30.1 [30.1 ‘Gipsy |jere [ess [346 [271 (psf) [1153 [1460 [1582 | 1601 Gs (psf) [11520 [1672 [1651 [1656 Pr (psf) f]at2 [485 [262 [205 Pstpsf) [714 [ree [176 [178 Bs-P{___[302_|300 [513 [572 Cy 0135] .0258| 0167] .0127 C; (0.48 [0.60 [0.73 [0.74 Mg (kf¥) [1.75 [4.15 [5.53 [4-85 (eft [1.88 [2.78 [2.12 11.68 fa.13 [10.3 [11.2 [11.3 (0.30 [0.52 [0.65 [0.59 (0.31 [0.42 [0.39 [0.36 fo.61 [0.94 [1.04 [0.95 a erene p'=t.28’ Total Astin?) Fig, 8—Silo design example—Grain 2. From Fig. 2, determine 0,4, H’, and ¥. 3, Decide if and how much the channel will til. 4, From equations of Fig. 3, determine fy fh, and hy, 5, From equations of Fig. 3, for each depth Y, determine the value of ¥, Ly, Ly/D, Ap Pp and 6’. Use the design aid of Fig. 3 if depth ¥ falls between Levels A and B. 6. Determine gy using Janssen’s equation (if above Level ©) or Walker's Eq. (A-1) (in the Appendix) (if below Level ©. 7. Determine q for each depth ¥ using Janssen's equation. 8. Determine 4, Ps and p; from Eq. (1), (3), and (4). Use Walker's equation for py [Eq. (A-6)] if depth ¥ falls below Level C. 9, Determine pressure differential p, - py 10. Obtain moment coefficients from Fig, 7. Determine ‘moments, tension forces, and wall reinforcement using the standard methods of computing reinforcement for members subjected to both tension and moment. In the particular example of Fig. 8, the total A, required just below Level C drops because the flow pressure py in- 226 creases with respect to the static pressure p,, and the mo- ments decrease. It is recommended in such cases that the reinforcement envelope be constructed by extending a straight line from the point of maximum A, required above Level C to the point of maximum A, required below Level C. Generally, the extra reinforcement provided at the flow channel should be extended around the entire perimeter of the silo. The final design should never result in less rein- forcement than required under ACI 313 (or the governing ‘code) for the concentric flow case. SUMMARY AND CONCLUSIONS. ‘The guidelines and procedures presented are easy {0 use and applicable for the normal range of concrete silo sizes and flow channel configurations encountered. They are applica- ble to variety of stored materials and have resulted in prac- tical designs that have proven successful forthe authors over the last several years. ACI Structural Journal / March-April 1995 + 70" WITH 2-23” FLOW CHANNELS COAL SILO (WITH 60* STAINLESS STEEL PYRAMIDAL HOPPER = 60 Por nt = 157 Ft Yoas tt nh = 131 ft he t4 in ae Dm = 60 fy = 60 ksi WILL LIKELY FLOW ON HOPPER . YOR Tao [oo [130 [138 3 “UPER 313-91 TTT oO CONCENTRIC Gn [as _[a 5 [are . Loe Uy (FFT 25.4 [23-4] 25.4 | 22-0 2 Ty A_~[o.35]0.35.[0.33 [0.31] 7 Adi) [236 [238 |238 [208 Prcsiy [eo [60 [60 —[59.9 $133.4 [33.4 [33.4 [35.4 ‘a (pst) [aoa _[o79 ee? [eva ‘i (pst [e66 [2969 | 3738 | 3020 s(osF) [2013 [3264 | «rat [az0t Prtpsf) [336 [300 [362 [aay cost fas _|1338 | 1687 [1722 Bee fasr [376 [1335 [1278 Cy T0149 .0140). 01401-0133 po G; 11 [0.71 f0.71 [0.12 47.8 [65.2 [60.1 19.3[26.3 ]23.0 x 42-6 [52.2 [54 § 2.35 [3.17 [3.00 a 1.39 [1.82 [1.73 . 3.74 [4.99 [4.78 Fig. 9—Silo design example—Coal NOTATION sea of flow channel tension stel area usually per unit of height of wall cross-sectional area of silo critical ratio defined by Giunta29 (ig. 2) constant [Walker's Eq, (A) (sce Appendix) ‘oeticent For Morar coefficient for Mjp diameter (inside) of sila effective outlet diameter modulus of elasticity of bulk material modulus of elasticity for concrete hoop tension resulting from p cight of flow channel per Giunta” (Fig 2) length of design flow channel perimeter in contact with wall (rig.3) ‘moment causing tension on inside face of wall moment causing tension on outside face of wall sum of Mypand Mop usually per unit of height of wall perimeter of flow channel ratio of area to perimeter of horizontal cross section of stor age space (hydraulic ads) ‘dep from Surface of stored material o point in question diameter of flow channel per Giunta” (Fg, 2) tonstant in equation for hy (Fig. 3), effective depth of flexural member ACI Structural Journal / March-April 1995 Total Astin) af = disuances rom ace of wall toceterof enforcement nearest that face € = istance rom cemer of soto center of outlet (eccentricity of cout 0 sistance rom centr of ilo wo apex of excentrichoppe (Fig. > r rll hickness ‘y= height of hoper fom apes to top of hopper i depth below top of hopper opin in question ® hight from bopper apex to point wher tied low chanel cont wal Wy = height from hopper apex effective outlet of hopper (Fig. 2 and) X, eight fom hopper apex o point where Pfortited flow channel has expanded To equal the chord Iength ep (Fig. 3) k rata pt * onstant [Walker's a, (A-2) and (A-3)] (ee Appendix) n ina ng bor zona pressure eto stored eel horizontal pressure within flow channels) ™ esse normal hopper surface [Walker's Eg. (A6)] (see Rppeniny 7, —= hoon presure within static material around flow chan nels) 4 = Serial pressre at deh in converging channel [Walkers 4. (A-1) ee Append) a inl iin] cereal pressure de to sored materia a ‘erica pressure within low channel(s) % ‘erica pressure within static material around fw chan- nels) c= angle of hopper wih horizontal B = constant [Walkers Eq (8-5) (see Appendix) 227 eight per unit volume for stored material y 3, = effective angle of internal friction = coeffcient of friesion between stored material and wall or hopper surface = angle of interna friction . “angle of friction between material and hopper or wal surface 8) = angle of flow chanel wit vercal (Fig, and 3) @; = gpl oft of Now chanel axis with vera (i. 3,7-and ) p= angle of repose with horizontal REFERENCES 1, ACI Committee 313, "Standard Practice for Design and Construction ‘of Concrete Silos and Stacking Tubes for Storing Granular Materals—and “Commentary.” American Concrete Institute, Detroit, 1991, 2. DIN 1055, Design Loads for Buildings: Loads in Silo Bins, Part 6, Deutsches Institut fur Normung, Berlin, May 1987. 5. “Instructions fr Design of Silos for Granular Materials,” Soviet Code (CH-302-65, Gosstroy, Moscow, 1965 4, "Loads Due to Bulk Material,” ISO, Karlsruhe, July 1990. 5, Standards Association of Australia, “Loads on Bulk Solids Contain rs" Draft Australian Standard, SAA, Sydney. 1989 %. Janssen, H. A. "Versuche uber Geteidedruck in Silozellen.” VDI Zeitschrift, Dusseldort, V.39, Aug. 1885, pp. 1045-1049 7. Janssen, H. A., Zeitschrift des Vereines Deuscher Ingenicure, 1895, 39,(35), pp. 1045-1049 8, Sugita, M,, Flow and Pressures of Noncobesive Granular Materials, in Funnel-Flow Bins.” Publication No. 72-MH-20, American Society of ‘Mechanical Engineers, New York, 1972. 9, Nielson, J, and Kristiansen, N, O., “Pressure Measurements ona Silo in Karpalund” (in Danish), Nordic Group fr Silo Research, Report No.5, ‘Technical University of Denmark, Department of Structural Engineering, 1979, 10, Blight, G. E., and Midgley, D, “Pressure Measured in a 20M Diame- ter Coal Load-Out Bin," Journal of Powder and Bulk Solids Technology, V. 5, No.2, 1981, pp. 21-31. 11. Anderson, E. Y,, “Wall Pressures and Flow Patten in Fly Ash Silos Due to Various Aerstions,” Internal Joumal of Bulk Solids Storage in Silos, V.4, London, 1985, pp. 1-7. 12, Ooi J, Ys Pham, L.; and Rotter, M. M., “Systemat Features of Measured Pressures on Full-Scale Silo Wall Structures, V. 12, London, Apr. 199, pp. 7487. 13. Walker, D. M., “Approximate Theory for Pressure and Arching in Hopper,” Chemical Engineering Science, V. 21, 1966, pp- 975-997. Ts, Pieper, K., "Investigation of Silo Loads in Measuring Models.” Pub- lication No, 68-MH-30, American Society of Mechanical Engineer, New ‘York, 1968, 15, Deutsch, G. and Schmidt, L.C., “Pressures on cation No, 68 MH-24, American Society of Mechanical Engineers, New York, 1968 16, Jeike, A. W:; Johanson, J. R; and Carson, J. W., “Bin Loads, Parts 2,3yand 4," Publication No. 72-MHC, 2,3, American Society of Mechan- {cal Engineers, New York, 1972. 17. Clague, K., and Weight, H., “Pressures in Bunkers,” Iron and Steel Institute, Aug. 1973, pp- 336-246, 1, Jenike, Andrew, W.,"Denting of Cireular Bins with Eecentric Draw: points," Journal ofthe Structural Division, ASCE, V. 93, STI, Feb. 1967, pp. 2735. 19, Sadler, J. E., “Silo Problems,” Proceedings, International Confer- cence on Design of Silos for Suength and Flow, University of Lancaster, Sept. 1980. 20, Johnston, FT, and Hunt, F. A. “Solutions for Silo Asymmetric Flow Problems,” Proceedings, 2nd International Conference on Design of Silos for Strength and Flow, Sratford-upon-Avon, England, Nov. 1983, pp. 13. 21, Colin, H, and Peschl,V., *Non-Symmetrical Bin Flow Problems,” International Journal of Storing and Handling Bulk Materials, V.6, No.3, 1981, pp. 7995, 22, Salt, J, Ea etal, “Bins and Silos,” Handbook of Sirctural Con and Random Engineering 228 crete, Kong, FA; Evans, R. H.; Cohen, E; and Roll, F,eds., MeGraw- Hill Book Co., New Yerk, 1983, Chap. 29. pp. 1-33. 23, Wood, JG. M., “Analysis of Silo Structures Subject to Eccentric Discharge,” Proceedings, 2nd Imerational Conference on Design of Silos for Strength and Flow, Statford-on-Avon, England, Nov. 1983, pp. 132- 14 24, Hunt, FA. and Johaston, FT, “Solutions for Asymmetrical Flow Problems,” Coal Mining and Processing, Jan. 1984 25, Roiter, J. M., "Analysis of Steel Bins Subject to Eccentric Di charge,” 2nd International Conference on Bulk Material Storage, Handling land Transporation, Institution of Engineers, Wollongong, New Soulh ‘Wales, Australia uly 1986. 26, Blackler, M.., and Woods, J. G. M., Discussion on Technical Note 490, “Eccentric Discharge in Circular Silos.” by Hayal, H. M., Proceed- ings, Institue of Civil Engineers, Part 2, June 1987, pp. 475-480, 23. Blackler, M.J., and Wood, JG. M., “Toward the Design Limit of ‘Steel Silos” Proceeding, International Conference on Design and Con struction of Non-Conventional Structures, London, Dec. 1987. "28 Johnston, FT, “How to Design Large-Diameter Silos that La Powder and Bulk Engineering, May 1990, pp. 43-53. 29, Giunta, J. S, "Flow Patter of Granular Material in Flat-Bottom Bins” Publication No, 68-MH-1, American Society of Mechanical Engi neers, New York, 1968. 30, Roatk, RJ, Formulas for Stress and Strain, McGraw-Hill Book Co, New York 31. Bucklin, R.A Thompson, S. A; and Ross, I J, “Flow Patterns in ‘Model Grain Bins.” Transactions of ASAE, V. 34, No. 2, 1991, pp. $77 585. 32, Saperstein, L, W.,and Ozgen, N.C.,“Piping in the Flow of Granular Materials from Rectangular Bins," International Journal of Rack Mechan- les and Mining Science, Pergamon Press, V.8, 1971, pp. 1-16. "33. Carson JW, et a, “Predicting the Shape of Flow Channels in Fun ‘el Flow Bins and Silos,” ACI Committee 313 seminar, 1991 Spring Con- ‘vention, American Concrete Institute. APPENDIX ‘The following equations from Walker’ paper are repeated for conve rience, using annotations consistent with this paper. “The vertical pressure at depth hy below top of hopper is computed by ay fer gg computed by ane, a Force cnes, 2c, mootesstan 0 on For pane flow hope ny utmost 10 oo vo sndsn2 4) G sindcos2 (0 +B) ay ow 169+ aresin’ po¥fo’ arcana] os) Except fr the vertical end walls of plane flow hoppers, the pressure nor- ‘mal tothe hopper surface at a depth hy, below top of hopper is Computed by = Lb sindeos (28) Pe = [sind e052 (0+ BY a ACI Structural Journal / March-April 1995

You might also like