Problem Statement Methodology Finite Element Simplified Cross Section Model. Validation of Installation Requirements for

Transcription

1 Performance Evaluation of Buried Pipe Installation Preliminary Results Michele Barbato Assistant Professor Dept. of Civil and Environmental Engineering g Louisiana State University, Baton Rouge 2009 Louisiana Transportation Conference Baton Rouge, February 10, 2009

2 Outline Problem Statement Methodology Finite Element Simplified Cross Section Model Considered Modeling and Performance Parameters Performance SensitivityStudyStudy Validation of Installation Requirements for Realistic Implementation Situations Main results and preliminary recommendations Ongoing and Future Studies

3 Problem Statement Motivations: Buried pipe performance and reliability is not fully addressed by current specifications New competing materials (PVC, HDP) Objectives: Determine effects of geometric and mechanical parameters on soil-structure interaction developed in buried pipe installation Scope: Rational lbasis for revision i of current provisions i on buried pipe installation in Louisiana and USA.

4 Methodology The soil-structure interaction between pipe and soil is studied using the finite element method. Performance measures are defined from finite element response quantities. Parametric studies are performed on different geometries, materials, configurations, natural soil situations i Analysis is focused on performance sensitivity to design variables

5 Simplified Cross-Section Model P Parameters considered in this study: H p H b,pave H Pavement Pavement Base Cover Pipe material and geometry Bedding properties Fill properties and compaction Surrounding soil properties p Importance of facility D t W Backfill The load P is obtained from AASHTO design load, considering i an impact factor c = 1.5 B Bedding Other data are obtained from AASHTO material properties for pipe design and are provided by LTRC in order to represent typical soil conditions for Louisiana. B f

6 Cross-Section Properties of Pipes D t Critical approximations: Linear elastic materials Linear geometry Simplified boundary conditions Cross-section properties considered in this study: Pipe material: Concrete: E = 2,900 ksi Steel: E = 29,000 ksi PVC: E = 400ksi HDP: E = 140ksi Pipe diameter: D = 18, 36, 42, 48, 60 Pipe diameter/thickness ratio, D/t: Concrete: D/t = 10, 20, 30 (referred to as RC10, RC20, RC30) Steel: D/t = 50, 100, 200 (referred to as S50, S100, S200) PVC: D/t = 10, 20 (referred to as PVC10, PVC20) HDP: D/t = 10, 20 (referred to as HDP10, HDP20)

7 Design Parameters for Sensitivity Study The design parameters considered are: 1. Dip depth at the surface,, with design limit lim = 0.1in 2. Maximum stress at the pipe ring, max A third design parameter (ring deflection) is not considered herein, since itimposes constraints t which hare automatically ti satisfied when the requirements on are satisfied. Two different safety coefficients are introduced and used to evaluate design conditions: 1. SC = lim / 2. SC = lim / max IT IS FOUND THAT SC IS USUALLY SMALLER THAN SC Deformations control the pipe design and installation

8 Sensitivity Study (1) Surrounding soil stiffness (E soil ) 1.2 SC The natural soil stiffness is crucial Different minimum requirements are needed between 0.8 stiff and yielding soil 0.7 Stiff and yielding soils can be specified in terms of 0.6 initial 0 stiffness, choosing 4 5 a value 6 7 in the 8 interval ksi E soil (ksi) Calibration of this value requires more advanced nonlinear FE analysis studies, which will allow also to study the effects of soil strength Model properties: D = 60 ; H = 12 ; W = 18 ; B = 6 E fill = 30 ksi; E b_pave = 45 ksi Pavement type: Asphalt, h = 4, h base = 10 Sensitivity variable: soil stiffness E soil = ksi RC10 RC20 RC30 HDP10 HDP20 PVC10 PVC20 S50 S100 S200

9 Sensitivity Study (2) Trench width (W) 1.1 SC E fill /E soil = 6 RC10 RC20 RC30 HDP10 HDP20 PVC10 PVC20 S50 S100 S W (in) Model properties: D = 60 ; H = 12 ; B = 6 E fill = 30 ksi; E b_pave = 45 ksi; E soil = 5 ksi Pavement type: Asphalt, h = 4, h base = 10 Sensitivity variable: trench width W = in D W

10 Sensitivity Study (3) 0.95 Trench width (W) 0.9 SC E fill /E soil = The trench width has a moderate effect RC10 RC20 RC30 HDP10 HDP20 PVC10 PVC20 S50 S100 S on the displacement safety coefficient SC 0.6 The effect depends on the ratio E fill /E soil and becomes significant W (in) for E fill /E soil 10 Model properties: D = 60 ; H = 12 ; B = 6 E fill = 30 ksi; E b_pave = 45 ksi; E soil = 1 ksi Pavement For W > type: D/2 Asphalt, (i.e., h total = 4, width h base = 10 of the trench > D 2D), Sensitivity variable: trench width W = in this effect is always very small W

11 Sensitivity Study (4) 1.1 Cover height (H) SC d The cover height (for H D) has a significant positive effect on the displacement safety coefficient SC This effect increases for increasing flexibility of the pipe RC10 RC20 RC30 HDP10 HDP20 PVC10 PVC20 S50 S100 S Further study is required to evaluate H (in) the cover height effects for H D Model properties: p D = 60 ; W = 18 ; B = 6 E fill = 30 ksi; E b_pave = 45 ksi; E soil = 5 ksi Pavement type: Asphalt, h = 4, h base = 10 Sensitivity variable: to Cover model height soil H = arch effects in Nonlinear FE analysis is required H D

12 1.1 Sensitivity Study (5) Bedding thickness (B) Bedding thickness (B) E bedding /E soil = 1: uniform material 1.05 C SC RC10 RC20 RC30 HDP10 HDP20 PVC10 PVC20 S50 S100 S B (in) Model properties: D = 60 ; H = 12 ; W = 18 ; E fill = 30 ksi; E b_pave = 45 ksi; E soil = 5 ksi; E bedding = 5 ksi D Pavement type: Asphalt, h = 4, h base = 10 Sensitivity variable: Bedding depth B = in B

13 1.15 E bedding /E soil = 2 (6in) E bedding 1.1 /E soil = 6 (below 6in) Sensitivity Study (6) Bedding thickness (B) SC For uniform distribution of stresses, a layer of of loose bedding material is required (and sufficient) immediately below the pipe B (in) In presence of yielding soil, the bedding thickness can be increased and the additional material (at depth larger than 6 below the pipe) should be compacted at the same level of D the backfill material B Model properties: D = 60 ; H = 12 ; W = 18 ; E fill = 30 ksi; E b_pave = 45 ksi; E soil = 5 ksi; E bedding = 10 ksi Pavement type: Asphalt, h = 4, h base = 10 Sensitivity variable: Bedding depth B = in RC10 RC20 RC30 HDP10 HDP20 PVC10 PVC20 S50 S100 S200

14 Sensitivity Study (7) Backfill soil stiffness (E fill ) SC RC10 RC20 RC30 HDP10 HDP20 PVC10 PVC20 S50 S100 S The stiffness of the backfill material is very important E fill (ksi) Model properties: D = 60 ; H = 12 ; W = 18 ; B = 6 E b_pave = 45 ksi; E soil = 5 ksi Pavement type: Asphalt, h = 4, h base = 10 Sensitivity variable: Backfill soil stiffness E fill = ksi Adequate compaction and material quality must be ensured everywhere and required in the specifications

15 2.4 Sensitivity Study (8) Road pavement type Road Pavement Type 2.2 SC S RC10 RC20 RC30 HDP10 HDP20 PVC10 PVC20 S50 S100 S The road pavement Asphalt 4in type is a crucial Concrete factor 8in in the Asphalt 10in performance of buried pipes Model properties: D = 60 ; H = 12 ; W = 18 ; B = 6 E fill = 30 ksi; E b_pave = 45 ksi; E soil = 5 ksi Sensitivity variable: Road pavement type Asphalt, h = 4, h base = 10 Asphalt, h = 10, h base = 10 Concrete, h = 8, h base = 10 In presence of rigid pavement (concrete), any pipe configuration performs satisfactorily if the pavement is not damaged

16 1.7 Sensitivity Study (9) Diameter (D) Diameter (D) 1.6 SC S RC10 RC20 RC30 HDP10 HDP20 PVC10 PVC20 S50 S100 S The displacement at the road surface increases D (in) almost linearly for increasing pipe diameters Model properties: H = 12 ; W = 18 ; B = 6 Minimum requirements E fill = 30 ksi; E b_pave for = satisfactory 45 ksi; E soil = 5 ksi performance Pavement type: Asphalt, of buried h = 4, hpipes base = 10 need to be Sensitivity variable: Pipe diameter D = in functions of the pipe diameter

17 Validation of Installation Requirements The installation requirements are verified for all combinations of the following conditions: Road pavement type: Asphalt pavement with h = 4in; Asphalt pavement with h = 10in; Concrete pavement with h = 8in. Surrounding soil properties: Moderately stiff soil (E = 5ksi); Yielding soil (E = 1ksi); Pipe diameter: D = 18in D = 60in Pipe material: Concrete, Steel, PVC, HDP Importance of facility: Low ADT, Moderate ADT, High ADT

18 Current Performance (1) 1.6 Results for roads with Low ADT Asphalt pavement with h = 4in SC Performance is not satisfactory! t D = 18'' (MSS) D = 18'' (YS) D = 60'' (MSS) D = 60'' (YS) Moderately stiff soil (MSS): E soil = 5 ksi 0.6 Concre ete Ste eel Modeling assumptions: The road pavement type is asphalt with h = 4 and h base = 10. The pavement base material has stiffness E b_pave = 30 ksi. The backfill material has stiffness E fill = 30 ksi. The bedding material has stiffness E bedding = 10 ksi. PV VC HD DP Yielding soil (YS): E soil = 1 ksi

19 Current Performance (2) Results for roads with High ADT Concrete pavement with h = 8in SC Very high h safety factors! D = 18'' (MSS) D = 18'' (YS) D = 60'' (MSS) D = 60'' (YS) Moderately stiff soil (MSS): E soil = 5 ksi 2 Concre ete Ste eel Modeling assumptions: The road pavement type is concrete with h = 8 and h base = 10. The pavement base material has stiffness E b_pave = 45 ksi. The backfill material has stiffness E fill = 45 ksi. The bedding material has stiffness E bedding = 10 ksi. PV VC HD DP Yielding soil (YS): E soil = 1 ksi

20 Possible Improvements (1) H = D/4 = 15 Concr rete H = 12 Asphalt pavement with h = 4in H = D/2 = 30 teel H = 24 H = D/2 = 30 PVC H = 24 H = D = 60 HDP H = 48 8 As is Increased cover D=60 Yielding soil Recommendations: Flexible pavement 1. Increase minimum cover to Low ADT Concrete: H max(d/4, 12 ) Steel and PVC: H max(d/2, 24 ) Modeling assumptions: HDP: H max(d, 48 ) The road pavement type is asphalt with h = 4 and h base = 10. The pavement 2. Require base material high quality has stiffness and E b_pave = 45 ksi. The backfill material high compaction has stiffness of fill E material fill = 45 ksi. The bedding (Ematerial fill = 45 has ksi) stiffness E bedding = 10 ksi. S

21 Possible Improvements (2) Results for roads with Low ADT Concrete pavement with h = 8in In presence of rigid road pavement (concrete), 2.1 the safety coefficient are very high also for the 2 worst case scenario 1.9 D = The minimum dimensional requirements Yielding soil Flexible pavement 1.7 can be relaxed Low ADT 1.6 A more accurate study based on nonlinear FE 1.5 analysis is required to calibrate the relaxed requirements Concrete Steel Based on the current results, it is safe to keep the same requirements used for low ADT also for moderate and high ADT PVC HDP

22 Main Results 1. Performance in terms of maximum deflection (dip depth) at the surface is often the most demanding requirement for buried pipe installation 2. The performance of the soil-structure system is crucially dependent on the road pavement type. For rigid pavement (concrete), pipe deformation is generally very small 3. Natural soil stiffness, backfill stiffness, cover height are very important 4. Bedding thickness and trench width have smaller effects on deformation. These effects depend on the ratio between the bedding and fill material stiffness and the natural soil stiffness. 5. Pipe diameter, D, significantly affects surface displacement. For D 48in and flexible pavement (asphalt), cover height should be function of D

23 Preliminary Recommendations 1. For rigid pavement (concrete), minimum requirements can be relaxed. Continuity of the pavement is crucial 2. For D 48in and flexible pavement, the minimum requirement on minimum cover in presence of yielding soil should be function of the pipe diameter, e.g.: Concrete: H max(d/4, 12 ) Steel and PVC: H max(d/2, 24 ) HDP: H max(d, 48 ) 3. For uniform distribution of stresses, a layer of 6 of loose bedding material is required immediately below the pipe. Additional bedding thickness should be compacted as the backfill material

24 Ongoing Studies Different linear and non linear finite element models and programs are currently being used dto study: 1. effects of direct modeling of natural soil surrounding the pipe trench 2. finite element model mesh sensitivity 3. three-dimensional effects 4 effects of different modeling options for the 4. effects of different modeling options for the applied loads

25 Direct Modeling of Natural Soil Mesh Expansion for D = 18" (HDP) Def formation (in) Vertical (ft) Horizontal (ft) 0

26 Three-Dimensional Effects Stress in D18 RC Pipe for Varying Model Thickness 0 Vertical 0 Horizontal 0 Vertical Horizontal 8 Vertical 0 Horizontal 8 Vertical Horizontal Mo oment (k-in) Mesh Thickness (ft)

27 Suggested Model for Nonlinear Finite Element Analysis

28 Current and Future Research Nonlinear staged Finite Element analysis will be used for: 1. Calibration of the soil stiffness value to distinguish i between stiff and yielding soil 2. Study of effects of soil strength 3. Study of effects of material (cracking, fracture, plasticization) and geometric (buckling) nonlinearities 4. Study of effects of installation procedures 5. Calibration of other important parameters for cases in which the performance estimated from linear analysis is not satisfactory

29 Acknowledgements LA Department t of Transportation ti and Development Louisiana Transportation Research Center (project 08-6GT) LA DOTD PIPE Committee Dr. Zhongjie Doc Zhang, Pavement and Geotech Research hadministrator i t Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect those of the sponsors. Thank you! Questions?