Wind Energy Science Conference WESC 2017

Transcription

1 Wind Energy Science Conference WESC 2017 Fatigue of very large high-strength bolting assemblies in wind turbines Rasmus Eichstädt Prof. Peter Schaumann Lyngby, ForWind

2 Outline High-strength bolts in wind turbines Experimental fatigue assessment Test results on HV-bolt sets and M64 Validation of normative S-N curves Analytical fatigue assessment DOTI/alpha ventus Conclusions

3 Outline High-strength bolts in wind turbines Experimental fatigue assessment Test results on HV-bolt sets and M64 Validation of normative S-N curves Analytical fatigue assessment DOTI/alpha ventus Conclusions

4 Bolted ring-flange connections in wind turbines SENVION DOTI/alpha ventus M48 M64 M72 Up to 200 high-strength bolt assemblies (System HV) per connection High loads with large number of load cycles

5 Preloading and hot-dip galvanizing Preloading for limitation of fatigue loads R m s [N/mm²] Rupture Nominal preload F p,c * = 0.7 R p0.2 A sp R p,0.2 High mean stress affects the fatigue strength of bolts Hot-dip galvanizing for corrosion protection Lower fatigue strength as uncoated structural components Fatigue cracks initiated at shrinkage cracks in the zinc layer 0.2 A e [%] Rolled notch N = Limited validation of fatigue characteristics and normative S-N curves for large bolt diameters Source: Simonsen (2015)

6 Outline High-strength bolts in wind turbines Experimental fatigue assessment Test results on HV-bolt sets and M64 Validation of normative S-N curves Analytical fatigue assessment DOTI/alpha ventus Conclusions

7 Fatigue tests on large-size HV-bolt sets and M64 HV-bolt sets (10.9), rolled before head treatment S 1 load cycle Fatigue tests with constant stress amplitudes and high mean stress S a Sm = 0.7 R p0.2 = 630 N/mm 2 Test series for 3 boundary layer conditions: S m t Black bolts (B) M64 Normal temperature hot-dip galvanized (NT) High temperature hot-dip galvanized (HT) only NT B Influence of hot-dip galvanizing B NT HT Assessment of the size effect

8 Photo: Alexander Raba Photo: Matthias Döring Fatigue tests on large-size HV-bolt sets - tests High frequency pulsator Max: 1 MN M64 - tests Servo-hydraulic testing machine Max: 10 MN 1.1 m 2.1 m Mean load 515 kn Testing frequency ca. 50 Hz Over 100 specimens Mean load 1680 kn Testing frequency 2-4 Hz 18 specimens

9 Nominal stress amplitude S a [N/mm²] Fatigue tests on HV-bolt sets 150 Mean stress S m = 0.7 R p0.2 = 630 N/mm² HT-galvanized Black B HT NT NT-galvanized approx % 50% Survival probability Rupture Run-out E E E E+07 Load cycles N [-] Decrease of fatigue strength in accordance with guideline VDI 2230

10 Nominal stress amplitude S a [N/mm²] Fatigue tests on HV-bolt sets M64 Results in comparison to bolts Black bolts 200 Mean stress S m = 0.7 R p0.2 = 630 N/mm² Survival probabilities: Ps = 50 % 100 M64 Ps = 10 % Ps = 90 % 60 M Transition region HCF - Range to endurance limit E E E E+07 Load cycles N [-] Run-outs B B

11 Nominal stress amplitude S a [N/mm²] Fatigue tests on HV-bolt sets M64 Results in comparison to bolts NT-galvanized bolts 200 Mean stress S m = 0.7 R p0.2 = 630 N/mm² Survival probabilities: Ps = 50 % 100 M64 Ps = 10 % Ps = 90 % 60 M Run-outs NT NT Transition region HCF - Range to endurance limit E E E E+07 Load cycles N [-]

12 Nominal stress amplitude S a [N/mm²] Fatigue tests on large-size HV-bolt sets Comparison to S-N curves from Eurocode 3 Hot-dip galvanized bolts 200 M64 M48 (Marten) 100 NT HT NT NT 60 Run-outs 20 Rupture Rupture M64 Rupture M48 (Marten) 10 1,0E+04 1,0E+05 1,0E+06 1,0E+07 Load cycles N [-] Size reduction: M48 M64 EC3 FAT 50 0,25 30 Sc,red S : 0,96 S M48 : 0,89 S M64 : 0,83 S c c c c

13 Nominal stress range ΔS [N/mm²] Char. fatigue strength ΔS C (N = ) [N/mm²] Fatigue tests on large-size HV-bolt sets Evaluation of FAT-class and size-reduction acc. to Eurocode HV-bolts (tzn NT) Rupture Run-out Ps,50% Ps,95% Regression w/ fixed slope m = Size reduction of EC3 FAT EC3 FAT 50 (w/o reduction) 45 ΔS (N = ) = 51,9 N/mm² EC 3 FAT 50 ΔS c (FAT 50, ) = 47,8 N/mm² 20 1,0E+04 1,0E+05 1,0E+06 1,0E+07 Load cycles N [-] Test result M48 M64 72 Bolt diameter [mm]

14 Nominal stress range ΔS [N/mm²] Char. fatigue strength ΔS C (N = ) [N/mm²] Fatigue tests on large-size HV-bolt sets Evaluation of FAT-class and size-reduction acc. to Eurocode M48 HV-bolts (tzn NT) EC3 FAT 50 (w/o reduction) Rupture Run-out Ps,50% Ps,95% Regression w/ fixed slope m = Size reduction of EC3 FAT 50 w/o upper HCF test level ( ) ΔS (N = ) = 48,4 N/mm² EC 3 FAT 50 ΔS c (FAT 50, ) = 44,5 N/mm² 20 1,0E+04 1,0E+05 1,0E+06 1,0E+07 Load cycles N [-] Test result M48 M64 72 Bolt diameter [mm]

15 Nominal stress range ΔS [N/mm²] Char. fatigue strength ΔS C (N = ) [N/mm²] Fatigue tests on large-size HV-bolt sets Evaluation of FAT-class and size-reduction acc. to Eurocode 3 M64 HV-bolts (tzn NT) Rupture Ps,50% 55 Size reduction of EC3 FAT Ps,95% EC3 FAT 50 (w/o reduction) Regression w/ fixed slope m = 3 45 w/o upper HCF test level ( ) ΔS (N = ) = 48,0 N/mm² EC 3 FAT 50 ΔS c (FAT 50, ) = 41,4 N/mm² 20 1,0E+04 1,0E+05 1,0E+06 1,0E+07 Load cycles N [-] Test result M48 M64 72 Bolt diameter [mm]

16 Nominal stress amplitude S a [N/mm²] Fatigue tests on large-size HV-bolt sets Comparison to S-N curves from Eurocode 3 Black bolts 200 M B B 60 Run-outs 20 Rupture Rupture M ,0E+04 1,0E+05 1,0E+06 1,0E+07 Load cycles N [-] Size reduction: M64 M64 EC3 FAT 71 EC3 FAT 50 0,25 30 Sc,red S : 0,96 S M64 : 0,83 S c c c

17 Outline High-strength bolts in wind turbines Experimental fatigue assessment Test results on HV-bolt sets and M64 Validation of normative S-N curves Analytical fatigue assessment DOTI/alpha ventus Conclusions

18 P SWT [N/mm²] Local stress σ [N/mm²] Fatigue calculation with local concept (strain-life) FE -Model Local hysteresis Nominal loading e 1,max ,00 0,01 0,02 0,03 Local strain ε [-] Damage parameter S-N curve (P SWT ) Non-linear material implementation 32CrB4 Preloading: monotonic material law Cyclic loading: cyclic stabilized material law Base material fatigue strain-life curve PSWT ( sa sm ) ea E N(P SWT ) / N(S a, S m ) 100 1,0E+00 1,0E+02 1,0E+04 1,0E+06 1,0E+08 Load cycle number until initial crack Ni [-]

19 Nominal stress amplitude Sₐ [N/mm²] Local stress σ Fatigue calculation with local concept S m = 0.7 R p0.2 Test (Ps,50%) B Initial crack 80 Local strain ε 60 Initial crack with cyclic relaxation Analytical: 20 Initial crack (w/o relax.) Initial crack (with relax.) Rupture (with relax.) 1,0E+03 1,0E+04 1,0E+05 1,0E+06 1,0E+07 Load cycles N [-] Crack propagation until rupture (additional)

20 Nominal stress amplitude Sₐ [N/mm²] Fatigue calculation with local concept 200 S m = 0.7 R p0.2 Test (Ps,50%) NT Effect of hot-dip galvanizing is not covered in analytical approach! Analytical: 20 Initial crack (w/o relax.) Initial crack (with relax.) Rupture (with relax.) 1,0E+03 1,0E+04 1,0E+05 1,0E+06 1,0E+07 Load cycles N [-]

21 Outline High-strength bolts in wind turbines Experimental fatigue assessment Test results on HV-bolt sets and M64 Validation of normative S-N curves Analytical fatigue assessment DOTI/alpha ventus Conclusions

22 Conclusions Fatigue capacity of high-strength, large-size bolts significantly affected by hot-dip galvanizing EC 3 fatigue class FAT50 confirmed for bolts up to diameter M64 Size reduction necessary, better fatigue classification of uncoated bolts must be seen critically Analytical calculations with local concept show good approximation to experimental results for bolts w/o boundary layer influence

23 Thank you! Research partner: Funding: