Background Work Item 41: API 51 7 th Edition Ballot Item 6.5 Work Item 41 Flow Induced Vibration Guidance # Source Section Comment Proposed Change Volunteer 41 5/14/14 Email to M. Porter 5.5.1 I would like to propose to add "Flow Induced Vibration (FIV)" in the next API51 edition. In my opinion the new edition will include the basic phenomena of FIV for both of beam and shell mode vibrations of the pipe. EI guideline will be referred for the beam mode vibration and Chiyoda paper will be referred for the shell mode vibration. Recently I have a closed relation with Rob Swindell, key member of EI guideline, and he agreed with me to support to prepare the draft of this new part on FIV. Hisao Izuchi; Tom Bevilacqua; Georges Melhem; Eli Vatland Johansen and Jim Cowling to review Supporting presentation from API 51 Fall 015 meeting is given following the proposed changes.
Proposed Modifications to API 51 6 th Edition (New non-mandatory Annex F): Annex F (Informative) Flow Induced Vibration Pressure-relieving systems are usually sized at relatively high velocities and turbulence energies become enlarged after the tees, reducers, bends, valves, etc. due to vortex formations with pressure fluctuations. The pressure fluctuations become larger as the fluid velocities become higher and their frequency spectrum have broadband characteristic with peak at lower frequency region. Piping vibration may occur at relatively low frequencies due to these pressure fluctuations caused by turbulence relating to insufficient stiffness of the piping system and consequently this may have resulted in fatigue failure of the piping system. This phenomenon is called flow induced vibration. In particular the turbulence energy becomes extremely enlarged especially just after the expansion at laterals or reducers (enlargements) [56]. Common examples of the mitigation options to prevent piping fatigue failure due to flow induced vibration include but not limited to the following [56], [X1], [X] : a) Reducing the velocity by enlarging the pipe diameter. b) Adding piping supports and/or increasing wall thickness. References [56] Energy Institute, Guidelines for the avoidance of vibration induced fatigue in process pipework, Second Edition, 008, ISBN 978-0-8593-463-0 [X1] M. Nishiguchi, H. Izuchi, I. Hayashi & G. Minorikawa, Flow induced vibration of piping downstream of tee connection, Proceedings 10th International Conference on Flow-Induced Vibration, July, 01 [X] M. Nishiguchi, H. Izuchi, I. Hayashi & G. Minorikawa, Investigation of characteristic of flow induced vibration caused by turbulence relating to acoustically induced vibration, Proceedings ASME 014 Pressure Vessels & Piping Conference, July 014
Task Force on API51 API Fall Meeting 015 Flow Induced Vibration (FIV) Caused by Turbulence Hisao Izuchi Chiyoda Corporation Chiyoda Corporation 015, All Rights Reserved. Contents 1. Phenomena of FIV. Evaluation Method on FIV 3.Proposed Draft for API51 1
Increase of FIV Failure Possibility Increase of plant capacity Flow rates of pressure reliving system tend to increase Pipe diameters tend to increase Economic design of piping system Velocities of flare piping tend to increase Wall thicknesses of flare piping tend to decrease Exciting Force Increases + Piping Stiffness Decreases Relatively AIV (Acoustically Induced Vibration) FIV (Flow Induced Vibration caused by Turbulence) Instruction of FIV should be added into API51 Failure Caused by FIV High Velocity = High Turbulence = Large Exciting Force & Low Stiffness of Piping Sever Piping Vibration (Beam Mode) Fatigue Failure Piping failure example caused by FIV (presented by E. Zamejc, API Spring Meeting, 006) 3
Mechanism of FIV Norton and Karczub* explain as follows: (1) An intense non-propagating pressure field is generated in the immediate vicinity of the disturbances such as those produced by valves, bends, junctions, and other pipe fittings. () This fluctuating pressure field decays exponentially with distance from the disturbance, falling off to an essentially constant asymptotic state within a distance about ten diameters. This locally generated non-propagating pressure fluctuation could be source of the FIV. * M. P. Norton and D. G. D. G. Karczub, Fundamentals of Noise and Vibration Analysis for Engineers, nd Edition, Cambridge, 003 4 Pressure Fluctuation after 90deg Miter Bend p 10-1 M=0. X 10-10 -3 10-10 -3 10-4 10-5 10-10 -3 10-4 10-5 10-6 10-7 M=0.40 M=0.50 Large pressure fluctuation at low frequency region Undisturbed fully developed turbulent flow in a straight pipe at M=0.40 0.76 1.15 1.54 1.9.31 9.14 10.45 11.76 5.8 Undistributed fully-developed turbulent flow 10-1 10 0 10 1 10 p p U 0 a i /U 0 X x / d q 0 q 0 a i 1 U f 0 x : axis of pipe downstream of bend d : internal diameter of pipe f : density of fluid in pipe U 0 : center-line velocity at X=5.8 p : wall pressure fluctuation a i : internal pipe radius : radian frequency M : Mach Number M. K. Bull and M. P. Norton, On the Hydrodynamic and Acoustic Wall Pressure Fluctuations in Turbulent Pipe Flow due to 90deg Miter Bend, J. Sound and Vibration, pp561-586, Vol76-4, 1981 5
Pressure Fluctuation after 90deg Miter Bend Pressure fluctuation increases due to the effect of miter bend. Max. non-dimensional PSD of pressure fluctuation p becomes order of 10-1 This PSD fo fluctuating pressure field rapidly decays with distance at 10 diameter. Max. non-dimensional pressure fluctuation p becomes order of 10-3 at X = 11.76. This decreased fluctuating pressure field (Max p of 10-3 ) is still higher than that of undisturbed fully developed turbulent flow in a straight pipe (Max p of 10-5 ). Pressure fluctuations just after the disturbances such as valve, tee, miter bend etc. become the source of FIV (Broadband and random excitation with relatively low frequency) 6 Pressure Fluctuation after Expander, Miter Bend, etc. 90 deg. Miter bend and 1: expander would be sources FIV. (90 deg.) Also Combining tee with small branch area ratio to main line could be source of FIV similar to 1: expander. J. A. Mann, D. Eilers and A. C. Fagerlund, Predicting pipe internal sound field and pipe wall vibration using statistical energy approaches for AIV, Inter-Noise 01 7
Pressure Fluctuation at Combining Tee Results of CFD Analysis with HPC (High Performance Computer) Pressure Distribution at Centre Cross Section Periodic separation vortex structure moves from junction edge to bottom of the pipe and dissipates at downstream. 8 Pressure Fluctuation at Combining Tee Results of CFD Analysis with HPC (High Performance Computer) Pressure Distribution at Wall Wall Pressure [Pa] * Gray color shows Iso-surface of low pressure The separation vortex structure causes large pressure fluctuation at the pipe wall around the impingement point. 9
Pressure Fluctuation at Combining Tee Results of CFD Analysis with HPC (High Performance Computer) 1D Downstream D Downstream D Impingement Point 1D Downstream D Downstream 10 Screening Method in EI Guidelines Guidelines for the Avoidance of Vibration Induced Fatigue Failure in Process Pipework published by Energy Institute, 008 (EI Guidelines) offer a screening method of FIV failure risk for beam mode pipe vibration, however this method is only for screening purpose and not suitable for detailed design. In EI Guidelines, the following screening method (outline) is introduced: (1) Calculate LOF (Likelihood of Failure) V LOF FVF FV Fv: FIV Factor (Function of Piping Flexibility, D and D/t) FVF: Fluid Viscosity Factor for Gas There are four classes for piping flexibility, Stiff (14-16Hz), Medium Stiff (7Hz), Medium (4Hz) and Flexible (1Hz). Frequency in the parenthesis is typical fundamental natural frequency. LOF 1.0 1.0 LOF 0.5 () Main line shall be redesigned. Small core connection should be assessed. 11
Chiyoda FIV Study Chiyoda executed experimental study of FIV just after a combined tee. Backgrounds of study are (1) Large amount of flow in flare piping system () Economic piping design = Relatively thinner pipe wall thickness Increase of FIV risk Generally flare piping system is stiff for beam mode vibration corresponding to adequate pipe support span and large diameter. However, the shell mode vibration of the pipe would increase corresponding to relatively larger D/t similar to AIV. 1 Chiyoda FIV Study / Field Measurement Power Spectral Density of Circumferential Stress Piping Stress Strain 0.05 0.00 0.015 0.010 0.005 0.000 計測結果 66Hz (3rd mode) 0 50 100 150 00 Frequency [Hz] (a) Power Spectral Density of Piping Stress (corresponding to circumferential strain) Circumferential Stress (c) Time History of Piping Strain Axial Stress 0 1 3 4 5 6 7 8 9 10 Time Time[sec] 9' 0 11' 8' 1' 7' 13' 6' 5' 4' 14' 15' 16' 3' 17' ' 18' 1 0-1 - 19 (b) Distribution of Vibration Displacement (3rd Shell Mode at 66Hz) 18 3 17 4 5 6 7 8 9 10 11 1 13 14 15 16 Vibration stress of shell mode is higher than that of beam mode. H. Izuchi, Piping Integrity Design for Flare System on Acoustically Induced Vibration and Flow Induced Vibration, 0th World Petroleum Congress, 011 13
Chiyoda FIV Study / Experiment Facility 15 to 30 bara 45 and 90 deg. TP: Pressure Sensor TPA: Pressure Fluctuation Sensor M. Nishiguchi, H. Izuchi and G. Minorikawa, Investigation of Characteristic of Flow Induced Vibration Caused by Turbulence Relating to Acoustically Induced Vibration, ASME PVP014 14 Chiyoda FIV Study / Vibration Index From the fundamental random vibration characteristic the following vibration index can be introduced to express the magnitude of the vibration severity. F : Force m : Mass of Pipe f n : Fundamental Natural Frequency of Pipe g : Density of Fluid A 1 : Branch Flow Area v : Fluid Velocity at Branch p : Pressure Discontinuity at Branch (=0 at Subsonic Flow) r p : Density of Pipe D : Main Pipe Diameter : Wall thickness of Main Pipe t n 15
Chiyoda FIV Study / Experiment Results (90deg Tee) ] Proposed vibration index is quite effective to express the magnitude of vibration stress for wide range of piping specifications (branch area ratio and wall thickness) and process condition including choking condition at branch pipe. 16 Chiyoda FIV Study / Experiment Results (45/90deg Tee) Proposed vibration index is useful for both of 45 and 90 degree tees. ] The Vibration stress of the 90 degrees tee is approximately 1.3 times the vibration stress of the 45 degrees tee.. 17
Investigation Summary Failure risk of flare piping system due to FIV tends to increase corresponding to the increase of relieving flow rate. FIV occurs at just after the disturbances (valve, tee, miter bend etc.) which generate turbulence with broadband and random excitation characteristics with relatively low frequency. Both beam and shell mode vibrations could occur relating to lower stiffness of the piping system. Beam mode vibration would occur in case of insufficient support of the piping system. A screening method to evaluate FIV risk is described in EI Guidelines, however this method is still conservative for gas service after modification taking the dynamic viscosity effect into account. Vibration index introduced by Chiyoda is useful to evaluate the magnitude of vibration stress for shell mode vibration caused by FIV. 18 Draft for API 51 New Edition / Policy Design method would be difficult to be defined in API 51 though there are several publications on FIV. Caution of FIV should be described in API 51 because there would be piping failure risk caused by FIV. The proposed policy for API 51 description is to explain the following items: (1) Mechanics and characteristics of FIV () Key design points for FIV with referred publications. 19
Draft for API 51 New Edition (1/) 5.5.X Flow Induced Vibration Pressure-relieving systems are usually sized at relatively high velocities and turbulence energies become enlarged after the tees, reducers, bends, valves, etc. due to vortex formations with pressure fluctuations. The pressure fluctuations become larger as the fluid velocities become higher and their frequency spectrum have broadband characteristic with peak at lower frequency region. Piping vibration of beam mode may occur at relatively low frequency due to these pressure fluctuation caused by turbulence relating to insufficient stiffness of the piping system and consequently this may have resulted in fatigue failure of the piping system. This phenomenon is called flow induced vibration [56]. In particular the turbulence energy becomes extremely enlarged especially for the tail pipe ends just before the expansions at laterals or reducers (enlargements). In this case circumferential vibration in the pipe wall can occur with relatively low frequency in addition to the beam mode vibration relating to insufficient stiffness of the piping system for circumferential direction similar to high-frequency vibration caused by the noise source (see 5.5.1) [X1], [X].. 0 Draft for API 51 New Edition (/) Common examples of the mitigation options to prevent piping fatigue failure due to flow induced vibration include but are not limited to the following [56], [X1], [X]. a) Reducing the velocity by enlarging the pipe diameter b) Adding piping supports to mitigate the piping vibration of beam mode. c) Increasing wall thickness to mitigate the circumferential vibration in the pipe wall References [56] Energy Institute, Guidelines for the avoidance of vibration induced fatigue in process pipework, Second Edition, 008, ISBN 978-0- 8593-463-0 [X1] M. Nishiguchi, H. Izuchi, I. Hayashi & G. Minorikawa, Flow induced vibration of piping downstream of tee connection, Proceedings 10th International Conference on Flow-Induced Vibration, July, 01 [X] M. Nishiguchi, H. Izuchi, I. Hayashi & G. Minorikawa, Investigation of characteristic of flow induced vibration caused by turbulence relating to acoustically induced vibration, Proceedings ASME 014 Pressure Vessels & Piping Conference, July 014 1
Future Technical Development for AIV and FIV AIV and FIV are quite similar vibration phenomena caused by random pressure fluctuation in the pipe. The difference of AIV and FIV are the excitation sources. The sources of AIV and FIV are shock wave generated by restriction device and turbulence generated by disturbance such as tee, expander, respectively. The region of FIV is limited at the vicinity of the generating point. On the other hand the region of AIV is quite wide area at the downstream of the restriction devices. This difference in the region is introduced from the dissipation characteristics of AIV and FIV. Most of failure locations due to AIV is just downstream of combining tee as shown in Carucci and Mueller paper. FIV risk becomes also high at this combined tee downstream as shown in the next slide. This means the comprehensive evaluation method is desired to be developed taking both effects of AIV and FIV into accounts. This activity will be done in the AIV JIP organized by Energy Institute. C-M Data Review with FIV Evaluation 18 18 16 16 14 14 FIV Vibration Index (Vn) 1 10 8 6 4 Failure No Failure FIV Vibration Index (Vn) 1 10 8 6 4 Failure No Failure 0 0.0 0. 0.4 0.6 0.8 1.0 Pipe Diameter (m) 0 0 00,000 400,000 600,000 Flow Rate (kg/h) As shown above, FIV indexes for failure case are apparently higher than those of no failure case. This suggests that the role of FIV effect could not be ignored in AIV evaluation. 3
Thank you Chiyoda Corporation 015, All Rights Reserved. 4