Calculation on a valve housing flange

Transcription

1 Calculation on a valve housing flange numerical FEM analysis versus analytical calculation according to DIN EN Dipl.-Ing. Gerd Lannewehr Peter Thomsen Lannewehr + Thomsen GmbH & Co.KG Bremen, Germany info@flangevalid.com Slide 1

2 Overview of topics - Brief introduction of the lecturers - Consideration of the different flange calculations - AD 2000 data sheet B7/B8 - DIN EN / DIN EN VDI ASME Section VIII Div. 1 App. 2 and VIII Div. 2 App flangevalid / ANSYS - FEM analysis of a DN100 PN40 flanged joint with grooved metal gasket with graphite layers compared to the calculation according to DIN EN , revision 2011 and 2014 Slide 2

3 Slide 3

4 ISBN-13: ISBN-13: ISBN-13: Publisher: PP PUBLICO Publications, Slide 4

5 Drainage flat washers for horizontal flanges Safety blinds with vent Inventions Measurement flat washers Mechanically resistant insulating flanges that are permanently technically tight measuring with pre-tensioning force Slide 5

6 Standard Flange calculation methods and their application limits ( flangevalid) Mainline of force --- Gasket Auxil liary line of force --- Test standard Characteristic values Strength Tightness Proof German Clean Air Act (Technical Instructions on Air Quality Control) Process Engineering EN X EN X X X 1) CEN/TS ) X X X X 1) KTA AD 2000 rules X X DIN X X X 1) X MPA/VBG 3) X X X 1) Data sheet B 7 X X 4) EN Section 11 FEA Finite element analysis X X X X EN advanced analysis X X X X 1) Only in conjunction with a component test according to VDI 2440 and 2200 and a professional installation 2) Pre-standard since ) VBG (Verband der Großkrafwerksbetreiber) is the Association of Large Power Plant Operators 4) Only in conjunction with EN 1591 (supplement from the lecture by M. Schaaf, AMTEC, XVII Gasket Colloquium) Source: VDI 2290: , Table 1, supplemented and updated, as at August 2018 Source: VDI 2290: , Table 1, supplemented and updated Slide 6

7 AD 2000 data sheet B7/B8 (+B0) - pure strength test - V A IV III I II 2 VI 1 3 C flange bending moment from pressure/temperature 2 flange expansion by pressure/temperature 3 compressive force on system B8 1) 4 flange bending by screw pre-tensioning force B8 5 screw bending by flange face angle 6 lever for flange face angle B8 2) 7 screw pre-tensioning force B7 Installation B7 Test B7 Operation B7 A flange face thickness B8 B screw size B7 C sealing surface B7 3) 1) unrealistic, surface medium gasket diameter 2) unrealistic, always from gasket center to screw center, it should be on the outside edge of the gasket 3) unrealistic, independent of flange face angle Proof of tightness Force distribution even, unrealistic Variable, realistic Surface pressure: Gasket B7 1) Flare, plating or lining Flat washer (if equipped) Nut contact surface 1) averaged, without taking into account the lever Pipe forces B7 1) Torsion and/or bending moments Thermal expansion 1) unrealistic approach, connected shell is not taken into account B 6 Components/component strength: I flanges B8 II gasket B7 1) Setting behavior/relaxation III screws B7 2) 3) IV nuts B7 V flat washers (if equipped) VI plating or lining (if equipped) not recorded partially recorded recorded 1) Defined values, varying from the manufacturer specifications 2) Safety factor? Expansion screws = 1.5, Full-shank screws = 1.8 3) Special surcharge C5 for operation, only for full-shank screws Slide 7

8 DIN EN / DIN EN (does not calculate apparatus flanges, problems with metal gaskets) V IV I 2 1 C flange bending moment from pressure/temperature 2 flange expansion by pressure/temperature 3 compressive force on system 4 flange bending by screw pre-tensioning force 5 screw bending by flange face angle 6 lever for flange face angle 7 screw pre-tensioning force Installation Test Operation A flange face thickness B screw size C sealing surface Proof of tightness Force distribution even, unrealistic variable, realistic III II 3 A VI Surface pressure: Gasket Flare, plating or lining Flat washer (if equipped) Nut contact surface Pipe forces 1) Torsion and/or bending moments Thermal expansion 1) unrealistic approach, connected shell is not taken into account B not recorded partially recorded recorded 6 Components/component strength: I flanges II gasket Setting behavior/relaxation III screws IV nuts V flat washers (if equipped) VI plating or lining (if equipped) Slide 8

9 DIN EN (does not calculate apparatus or vessel-flanges) This is how the calculation was done: - no dished bottom - no separator plate - no flange supports for inlet and outlet - no pipe bundle plate - only one flange with recess - no gasket with center bar (would have 28% more sealing surface) - expanding sleeves as a replacement for screw lengths This is how it was actually built: fact or fiction? Slide 9

10 VDI (does not calculate flanges with gaskets in the main line of force) V IV I 2 1 C flange bending moment from pressure/temperature 2 flange expansion by pressure/temperature 3 compressive force on system 4 flange bending by screw pre-tensioning force 5 screw bending by flange face angle 6 lever for flange face angle 7 screw pre-tensioning force Installation Test Operation A flange face thickness B screw size C sealing surface Proof of tightness Force distribution even, unrealistic variable, realistic III II 3 A VI Surface pressure: Gasket Flare, plating or lining Flat washer (if equipped) Nut contact surface Pipe forces Torsion and/or bending moments Thermal expansion B not recorded partially recorded recorded 6 Components/component strength: I flanges II gasket Setting behavior/relaxation III screws IV nuts V flat washers (if equipped) VI plating or lining (if equipped) Slide 10

11 ASME Section VIII Div. 1 App. 2 and VIII Div. 2 App pure strength test - V IV I 2 1 C flange bending moment from pressure/temperature 2 flange expansion by pressure/temperature 3 compressive force on system 4 flange bending by screw pre-tensioning force 5 screw bending by flange face angle 6 lever for flange face angle 7 screw pre-tensioning force Installation Test Operation A flange face thickness B screw size C sealing surface Proof of tightness Force distribution even, unrealistic variable, realistic III II 3 A VI Surface pressure: Gasket Flare, plating or lining Flat washer (if equipped) Nut contact surface Pipe forces 1) Torsion and/or bending moments Thermal expansion 1) unrealistic approach, connected shell is not taken into account Components/component strength: B 6 I flanges II gasket Setting behavior/relaxation III screws IV nuts V flat washers (if equipped) not recorded partially recorded recorded VI plating or lining (if equipped) Slide 11

12 flangevalid / ANSYS IV V I 2 1 C flange bending moment from pressure/temperature 2 flange expansion by pressure/temperature 3 compressive force on system 4 flange bending by screw pre-tensioning force 5 screw bending by flange face angle 6 lever for flange face angle 7 screw pre-tensioning force Installation Test Operation A flange face thickness B screw size C sealing surface Proof of tightness Force distribution even, unrealistic Variable, realistic A III II VI 3 Surface pressure: Gasket Flare, plating or lining Flat washer (if equipped) Nut contact surface Pipe forces Torsion and/or bending moments Thermal expansions B 6 Components/component strength: I flanges II gasket Setting behavior/relaxation III screws IV nuts V flat washers (if equipped) VI plating or lining (if equipped) not recorded partially recorded recorded Slide 12

13 Comparison of the calculations Components and component strength Components and component strength AD 2000 B7/B8 EN VDI ASME Sec. VIII flangevalid ANSYS Flange Gaskets Setting behavior of the seals / relaxation Screws Nuts Flat washers (if equipped) Plating or lining (if equipped) not recorded partially recorded recorded Slide 13

14 Comparison of the calculations Surface pressures, additional forces and thermal expansion Surface pressures, additional forces and thermal expansion AD 2000 B7/B8 EN VDI ASME Sec. VIII flangevalid ANSYS Gasket Flare, plating or lining Flat washer (if equipped) Nut contact surface Pipe forces Torsion and/or bending moments Thermal expansion Proof of tightness Force distribution not recorded partially recorded recorded Slide 14

15 Comparison of the calculations Dimensions and requirements for the flange system Dimensions and requirements for the flange system AD 2000 B7/B8 EN VDI ASME Sec. VIII flangevalid ANSYS Flange face thickness Screw size Sealing surface Flange bending moment from pressure/temperature Flange expansion from pressure/temperature Compressive force on the system Flange bending by screw force Screw bending by flange face angle Lever for flange face angle Screw pre-tensioning force not recorded partially recorded recorded Slide 15

16 flangevalid / ANSYS Deformation under temperature and pressure Source: PVP Case study of temperature analysis In a 48" diameter heat exchanger flange The flange bending by the screws must be greater than the bending of the flange by the different thermal expansion; otherwise, the flange continues to tilt and the screws are relieved. If the flange is not sufficiently pre-tensioned, the interior pressure acts in a similar way. Slide 16

17 flangevalid / ANSYS Flanged joints with gaskets in the main line of force and the auxiliary line of force are too complex for an analytical calculation. Only by calculations based on the finite element method can the actual requirements for a flanged joint be calculated in any meaningful way. Slide 17

18 FEM analysis of a housing flanged joint on a ball valve housing DN 50 PN Goal of the FEM analysis With the FEM analysis, the gasket widths of PTFE gaskets used in actual practice are to be tested on a ball valve housing in comparison with EN Introduction In the design of the flange of the housing flanged joint according to DIN EN , Chapter 10, the procedure described below is followed. In so doing, a differentiation is made between direct load (gasket in the main line of force, Figure a) and no direct load (gasket in the auxiliary line of force, Figure b). Slide 18

19 2. Introduction (continued) Designs of flanged joints a) Direct load b) Not exposed to a direct load Extract from DIN EN , Chapter 10: The calculation of the flanges must be in accordance with or based on the specifications contained in DIN EN or DIN EN For flat gaskets with direct load, it is only permitted to use the characteristics of gaskets if they were derived from test standard DIN EN Slide 19

20 2. Introduction (continued) DIN EN , Chapter 10 refers solely to a strength test of the components used. To assess the sizes, here is some data of the housing flanged joint: The pre-tensioning force of a M12 screw at 70% utilization of the yield point is 23,310 N, i.e. in the case of 6 acting screws a pre-tensioning force of a total of 139,860 N is applied. The common metallic contact surface of the housing flanges is 4,515 mm²; the sealing surface amounts to 848 mm², i.e. the total contact surface of metallic contacts including gasket is 5,362 N/mm². According to the definition in Chapter 10, the average contact diameter of the metal surfaces is used to calculate the internal compressive force, i.e. d D = mm (120 mm+93 mm)/2. The design pressure or test pressure is applied to the resulting calculation surface. The internal compressive forces to be taken into account are F = 35,614 N from the design pressure of 40 bar (4 MPa) and F = 57,340 N from the test pressure of 64.4 bar (6.44 MPa). Slide 20

21 3. Load cases The installation pre-tension of the screw bolts is selected as load case, making use of the yield point of the screw material at 70%. Then the design pressure of 40 bar (4 MPa) as well as the test pressure of 64.4 bar (6,44 MPa) are applied. By the applied pre-tensioning force, the PTFE gasket is brought from a direct load (small protrusion of the gasket during installation) to an indirect load, i.e. both load conditions exist in accordance with the definition of DIN EN Slide 21

22 4. Materials The materials used are: (GX 5 CrNiMo ), DIN EN 10213, for the housing flanges A2-70 for the screws PTFE gasket A friction coefficient of 0.12 was taken as friction on the housing contact surfaces. 5. Geometry import In the FEM model, the metallic housing flange contact surfaces were taken into consideration as a friction-prone contact with a friction coefficient of The screw head contact surfaces to the flange surface were taken as composite contact since there are no additional thermal loads. The PTFE gasket was taken into account as a composite contact since the gasket is centered and guided. All the internal parts of the ball valve, e.g. the ball, the ball seals, the shaft with sealing, for the opening and closing of the ball were not taken into account in the analysis. Slide 22

23 5. Geometry import (continued) The flanged joint was made available as a STEP file, imported into the ANSYS Design Modeler and then imported into ANSYS Mechanical. Figure 1: Geometry import of the entire flanged joint in the ANSYS Design Modeler Slide 23

24 Figure 2: Networking the entire flanged joint Slide 24

25 Figure 3: Static, mechanical model of the entire flanged joint Slide 25

26 6. Calculation results Figure 4: Max. von Mises stress, load case: pre-tension on the housing counter flange; in the screw hole area, stresses of >108.7 N/mm² (MPa); maximum stress in accordance with DIN EN Slide 26

27 Figure 5: Max. von Mises stress, load case: pre-tension on the housing flange; in the screw hole area, stresses of >108.7 N/mm² (MPa); maximum stress in accordance with DIN EN Slide 27

28 Figure 6: Maximum surface pressure of the PTFE gasket, load case: pre-tension Slide 28

29 Figure 7: Contact status of the metallic sealing surfaces, load case: pre-tension Slide 29

30 Figure 8: Contact status of the gasket, load case: pre-tension Slide 30

31 Figure 9: Status of the screw head contact surfaces Slide 31

32 Figure 10: Permeation on the housing counter flange; load case: pre-tension in the screw hole area Slide 32

33 7. Summary The conventional analytical calculation in accordance with DIN EN : and varies significantly from the numerical calculation using the FEM. Table 1: Surface pressures according to FEM in comparison to a calculation according to DIN EN : Surface pressure Variance from FEM Rules / calculation MPa Factor % DIN EN : average FEM min / max min / max average Table 2: Surface pressures according to FEM in comparison to a calculation according to DIN EN : Surface pressure Variance from FEM Rules / calculation MPa Factor % DIN EN : average FEM min / max min / max average Slide 33

34 The values for the operating screw forces in accordance with DIN EN : vary 3.5-fold from the values of DIN EN : Table 3: Increase of the operating screw force (work load) according to the DIN EN calculation and according to FEM with the values from the DIN EN calculations Installation Operation Difference Rules / calculation N N % DIN EN : ,965 DIN EN : , FEM Values from DIN EN : FEM Values from DIN EN : ,280 88, ,652 98, Difference in % Slide 34

35 8. Evaluation The values for the increase of the screw forces during operation vary significantly from each other between DIN EN : and DIN EN : (Section 7, Table 3). The variance of the surface pressures on the grooved metal gasket during operation, with the same input parameters is 556% lower in the analytical calculation in accordance with DIN EN : than in the numerical FEM analysis. According to DIN EN : , it is still 204% (Section 7, Tables 1 and 2). Table 4: Difference of the calculated surface pressures in accordance with DIN EN Surface pressure Variance Rules / calculation MPa Factor % DIN EN : DIN EN : Slide 35

36 The differences between the analytical and numerical calculation methods especially the differences from the analytical calculations in accordance with DIN EN are so great (Table 4) that it is essential to conduct further examinations to verify a sensible application of DIN EN : or DIN EN : Slide 36

37 Thank you for your attention! Slide 37