ORNL is managed by UT-Battelle for the US Department of Energy SNS mercury target fatigue life prediction with Fe-Safe/Verity High Power Targetry Workshop 2016 Saulius Kaminskas

2 SNS Target Fatigue FATIGUE… Target design fatigue evaluation is NECESSITY 4 out of 13 SNS targets ended their lives prematurely because of …

3 SNS Target Fatigue Fe-Safe - structural fatigue analysis tool, developed by Safe Technology, part of Abaqus Extended Portfolio Center Baffle: = 153,462 cycles – Thermal cycle Fatigue Life Limit SNS Original Target Design – Thermal stress cyclic loading, S-N curve – “316 Stainless Steel” from Fe-Safe DB Using modern multiaxial strain based methods with conventional FEA results from Abaqus, ANSYS, or NASTRAN calculates fatigue life results at every FE node producing contour plots of fatigue lives and possible crack sites: Fe-Safe provides material database, automatic surface contact detection, parallel processing, integration in structural optimization tools, and etc.

4 SNS Target Fatigue Fe-Safe also calculates the probability for certain damaged area to survive requested fatigue life: Combines variability in the material fatigue strength and variability in the applied loading; A normal (Gaussian) distribution is applied to the variation in loading; A Weibull distribution is applied to the variation in material strength: Minimal Probability of survival = 75.2% for requested 12,000 cycles with 5% variation SNS Original Target Design – Thermal stress cyclic loading, S-N curve – “316 Stainless Steel” from Fe-Safe DB

5 SNS Target Fatigue Fe-Safe also calculates the Factor of Strength (FOS): The factor which, when applied to either the loading or stresses will produce the requested fatigue life number of cycles The ratio A/B is the factor of strength Indicates how much the stress could be increased (FOS>1) or decreased (FOS<1) to reach the Endurance Limit (the Goodman line) Center Baffle FOS = 0.57 for infinite material life (Endurance Limit) Baffle cracks at center line during operation! SNS Original Target Design – Thermal stress cyclic loading, S-N curve – “316 Stainless Steel” from Fe-Safe DB

6 SNS Target Fatigue There are 2 cyclic loads affecting Target Thermal Stress – Low cycle fatigue Up to 12,000 cycles per target life time Pulse Stress – High cycle fatigue Up to 1 billion cycles and more Thermal and Pulse cycle fatigue analyzed separately due to high ratio of cycle frequencies Thermal cycle fatigue static preloads: -Gravity -Mercury pressure Fe-Safe pulse cycle dynamic load definition: Fe-Safe thermal cycle load definition: Pulse cycle fatigue static preloads: -Gravity -Mercury pressure -Thermal stress (steady state)

7 SNS Target Fatigue Fe-Safe – Pulse cycle fatigue analysis for SNS FY 2016 target design FY16 – improved design: -Tapered wall thickness -Better joint weld geometry -Changed Hg flow patterns -Etc. Pulse stress cyclic loading (280 frames from 1 msec history), S-N curve – “316 Stainless Steel” from Fe-Safe DB Side Baffle: = 15.74*10 6 cycles – Pulse cycle Fatigue Life Limit Redesign is NEEDED Transition sectionFront section 1.54 MW pulse max stress: MPa at side baffle Stress history – max at 0.6 msec

8 SNS Target Fatigue Fe-Safe - Target pulse fatigue life at different locations with beam power up to 2.3 MW SNS Jet-Flow Target Design – Pulse stress cyclic loading, S-N curve – “316 Stainless Steel” from Fe-Safe DB 0.85 MW Current beam power range 2.30 MW

9 SNS Target Fatigue Verity - Fe-Safe add-on module developed by Battelle Memorial Institute Researchers, promoted by the Welding Research Council, recommended by ASME for fatigue analysis of welds Verity usually requires a separate FE mesh for weld material: Weld FY16 Target Design – FE front window welds Uses structural stress method based on nodal forces at the weld FE mesh insensitive Uses a single master S-N curve for: -different joint geometries -different loading modes -different plate thickness Validated against 3,500 welded joint tests Multiple master S-N curves are provided for various materials and probabilities of failure

10 SNS Target Fatigue Fe-Safe/Verity – Thermal cycle fatigue for SNS FY16 target welds Front-Transition joint welds: Thermal stress cyclic loading at 1.54 MW beam power S-N master curve – “Steel Weld (50%)” from Verity DB Minimal Fatigue Life: = 1.064*10 6 cycles Front-Transition welds Front Windows welds

11 SNS Target Fatigue Fe-Safe/Verity – Thermal cycle fatigue for SNS FY16 target welds Front Windows joint welds: Thermal stress cyclic loading at 1.54 MW beam power Fe-Safe S-N curve - “316 Stainless Steel” from Fe-Safe DB Verity S-N master curve – “Steel Weld (50%)” from Verity DB Minimal Fatigue Life – at Outer Window: = 2.472*10 6 cycles Life at Inner Window Upper joint Weld: = 8.933*10 6 cycles Life at Inner Window Lower joint Weld: = 2.618*10 6 cycles Thermal fatigue life time is well above 12,000 cycles limit for all welds

12 SNS Target Fatigue Fe-Safe/Verity – Pulse cycle weld fatigue analysis – FY16 target SNS targets survived significantly more beam cycles than estimated by Verity S-N curve There is little experimental data available for weld (very) high cycle fatigue supporting the bases of weld S-N curves Verity results compared to those of British Standard 7608 code also available in Fe-Safe for welds fatigue (Used most relaxed S-N curve B) Due to high strain rate with high cyclic loads the fatigue strength is expected to be noticeably higher than estimated 1,000 million 100 million BS7608 S-N curve B Verity: S-N master curve – “Steel Weld (50%)” Fe-Safe ASME 2013 Verity

13 SNS Target Fatigue “What if” exercise: Higher Hg pump speed  lower thermal stress  higher fatigue life At the same time: Higher Hg pump speed  higher Hg pressure  lower fatigue life -For simplified target FE testing model: -Using Fe-Safe and Isight from Abaqus Extended Portfolio -Created response surface -Indicates the point of optimal pump speed for longest thermal fatigue life The search for trade-offs among operation parameters is possible to maximize the vessel fatigue life

14 SNS Target Fatigue Summary  There are many uncertain factors in mercury target fatigue longevity Structural: -Material inclusions, porosities -Weld imperfections -Surface smoothness … Environmental: -Radiation damage -Cavitation erosion -Beam of-normal conditions …  Fe-Safe/Verity provide indications of confidence to survive to specified loading cycles (not exact lifetime to failure)  For uncertainties we need certain fatigue safety margins  Fe-Safe/Verity complement the target analysis tool set for new designs that aim for higher target reliability

15 SNS Target Fatigue Thanks for your attention Any questions?

16 SNS Target Fatigue Supporting slides

17 SNS Target Fatigue What is Fe-Safe ?

18 SNS Target Fatigue What is Verity?

19 SNS Target Fatigue VERITYTM WELD FATIGUE METHOD IN FE-SAFETM USING FEA Software Dr. Pingsha Dong Center for Welded Structures Research BATTELLE Columbus, OH

20 SNS Target Fatigue Fe-Safe/Verity fatigue analysis: FE model for Front Window welds

21 SNS Target Fatigue Fe-Safe/BS7608 weld fatigue analysis: The S-N curves are defined in terms of the nominal stress a small distance from the weld or weld toe; The welded joint analysis uses a principal stress calculation, with no mean stress correction; Fatigue strength is not sensitive to material tensile strength; Fatigue strength is not sensitive to changes in mean stresses in the applied loading; The weld geometry need not normally be modelled in detail (included in the S-N curve); Stresses at the weld toe may give unrealistically low fatigue life results. Assumptions:

22 SNS Target Fatigue Fe-Safe/BS7608 weld fatigue analysis Available S-N curves (analysis done with B, F, and W): Curve B is the closest to curve C used for fatigue with ASME code previously !

23 SNS Target Fatigue S-N curves:

24 SNS Target Fatigue SNS Target earlier fatigue evaluation methodology Using ASME S-N curves … … evaluated stresses at separate points from the table … … having chances to miss spots with critically low fatigue life …