Post on 04-Jan-2016
Felix Dietrich | AWLC14 | 14.05.2014 |
Stress simulation in the ILC positron target with ANSYS
Felix Dietrich (TH-Wildau), Sabine Riemann, Friedrich Staufenbiel (DESY)
Felix Dietrich | AWLC14 | 14.05.2014 | page 2
Outline
> Alternative source
> Goal of the studies
> Model
> Implementation into ANSYS
> Temperature calculation
> Results
> Summary
> Next steps
Felix Dietrich | AWLC14 | 14.05.2014 | page 3
Alternative source
> Suggested by T. Omori et al., see also NIMA672 (2012) 52
> Uses beam (“conventional” source) of 6GeV
> Target: Tungsten
> 300Hz scheme Stretch ILC bunch train (~1ms) to 63ms
lower peak heat load
Stacking of in the damping ring
> ILC time structure achieved by extraction scheme from damping ring
> Benchmark from SLC target: Peak Energy Deposition Density (PEDD) should not exceed 29J/g
> Electron beam parameters are selected accordingly
Felix Dietrich | AWLC14 | 14.05.2014 | page 4
Goal of the studies
> Beam parameters are given ( ~29J/g peak energy deposition @SLC)
> How is the dynamic stress evolving in the target with 300Hz scheme in the tungsten target
> Target lifetime depend on static and dynamic stress
• Energy deposition within fraction of a nanosecond instantaneous temperature rise
• Reaction of material within milliseconds
Stress dynamics
> Benchmark: SLC target But SLC number of positrons per second was about 50 times lower than at ILC
Felix Dietrich | AWLC14 | 14.05.2014 | page 5
Model
> Model 1– Modified SLC target Fe – ring & Ag – coat & W – core
Radius for beam line @ target = 10.75cm
> Model 2– cylinder with diameter corresponding to beam spot size
W – Layer & Ag – Layer
Beam spot radius =0.8cm
Radius of the whole model is r=2.2cm
> Parameters for both models Particle shower has been calculated with FLUKA
Average energy deposition in the target:
35kW without silver layer49kW with silver layer
Felix Dietrich | AWLC14 | 14.05.2014 | page 6
Implementation into ANSYS
> Mesh Mesh size is related to time steps
is given = velocity of sound =5180 m/s
Model 1 : 22124 elements
Model 2 : 9419 elements
Elements: Hexagonal elements
> Time structure 1st attempt: Every bunch is calculated. the length of one bunch is 250ps
2nd attempt: Linear rise of the temperature in one mini train
> Calculation & Results Model 1: first Calculations could be done. But there is still some work to do.
Model 2: works well, needs less computing time
Felix Dietrich | AWLC14 | 14.05.2014 | page 7
Temperature calculation
> With good mesh, the temperature change resulting from energy deposition by each bunch shows regular behavior with negligible numerical errors
Felix Dietrich | AWLC14 | 14.05.2014 | page 8
Temperature rise: 1st attempt
> The temperature increases per bunch by 1.6K
> The graph shows the temperature rise for the first triplet
0,0 0,2 0,4 0,6 0,8 1,0 1,2
50
100
150
200
250
tem
pera
ture
[°C
]
time[µs]
max. temperatur
Felix Dietrich | AWLC14 | 14.05.2014 | page 9
Results: Stress evolution – Model 2
> The Picture shows the equivalent stress (von-Mises) at diferent positions in the Material
> The Picture was taken at: 1.0205 µs
(t=5ns first bunch of mini train)
> Deformation is scaled by 730 for visualization; max. surface displacement is about 8µm
Felix Dietrich | AWLC14 | 14.05.2014 | page 10
Result: Short time – Model 2
0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0 1,10
50
100
150
200
250
300
350
4003rd mini Train
2nd mini Train
stre
ss[M
Pa]
time[µs]
stress max (von Mises)
stress z=1,4cm stress z=1,375cm stress z=1,35cm stress z=1,325cm stress z=1,3cm stress z=1,25cm stress z=1,2cm stress z=1,15cm stress z=1,1cm stress z=1,05cm stress z=0,9cm
1st mini Train
> Results are calculated for 1µs
> Red line shows max. stress at any given point at one time
> The other lines are calculated an the center of the disc at different depth (1.4cm equals the surface; 0.9cm equals a depth of 0.5cm)
Felix Dietrich | AWLC14 | 14.05.2014 | page 11
Result: Conclusion short time
> Max. temperature rise by one triplet is T=210°C within ~1ms
> Heat dispersion into the target body within microseconds is negligible temperature of the target body is 22°C
> Max. stress after one triplet 377MPa
> Every mini train adds a different stress load First mini train +225MPa
Second mini train +110MPa
Third mini train +37MPa
> Max. stress by one triplet is below fatigue stress limit 377MPa<520MPa
Felix Dietrich | AWLC14 | 14.05.2014 | page 12
0,0 0,5 1,0 1,5 2,0 2,5 3,00
50
100
150
200
250
300
350
400
4503rd mini Train
2nd mini Train1st mini Train
stre
ss[M
Pa]
time[µs]
stress max (von Mises)
stress z=1,4cm stress z=1,375cm stress z=1,35cm stress z=1,325cm stress z=1,3cm stress z=1,25cm stress z=1,2cm stress z=1,15cm stress z=1,1cm stress z=1,05cm stress z=0,9cm
Results: Long time
> First attempt 2 µs pause
> Same configuration like the first graph
> Results arepreliminary
Felix Dietrich | AWLC14 | 14.05.2014 | page 13
Temperature rise: 2nd attempt
> The temperature rises linearly over the whole mini train
> That saves computing time
0,0 0,2 0,4 0,6 0,8 1,0 1,2
50
100
150
200
250
tem
pera
ture
[°C
]
time[µs]
max. temperatur
Felix Dietrich | AWLC14 | 14.05.2014 | page 14
Results for 2nd attemept: Model 1
> Results are calculated for 3µs
> Max. stress is about 400MPa
> Results are preliminary
0,0 0,5 1,0 1,5 2,0 2,5 3,00
50
100
150
200
250
300
350
400
450
2nd mini train3rd mini train
stre
ss[M
Pa]
time[µs]
max_stress stress z=1.4cm stress z=1.375cm stress z=1.35cm stress z=1.325cm stress z=1.3cm stress z=1.25cm stress z=1.2cm stress z=1.15cm stress z=1.1cm stress z=1.05cm stress z=0.9cm
1st mini train
Felix Dietrich | AWLC14 | 14.05.2014 | page 15
Results for 2nd attempt: Model 2
> Results are calculated for 3µs
> Max. sress is about 420MPa
> Results arepreliminary
0,0 0,5 1,0 1,5 2,0 2,5 3,00
50
100
150
200
250
300
350
400
450
stre
ss[M
Pa]
time[µs]
max_stress stress z=1.4cm stress z=1.375cm stress z=1.35cm stress z=1.325cm stress z=1.3cm stress z=1.25cm stress z=1.2cm stress z=1.15cm stress z=1.1cm stress z=1.05cm stress z=0.9cm
3rd mini train2nd mini train
1st mini train
Felix Dietrich | AWLC14 | 14.05.2014 | page 16
Summary
> Maximum dynamic stress per triplet is about 377MPa (fatigue stress limit is 520MPa)
> longer time Stress rise up to 425MPa for both attempts
> Load cycles per year at one spot : cycles
> With target rotation the number of load cycles at the same beam spot volume can be kept below
Felix Dietrich | AWLC14 | 14.05.2014 | page 17
Next steps
> Simulations of a longer period: Few triplets
Include target rotation
Check whether interferences of stress waves are possible
> Degradation of material is expected due to irradiationevaluation of dynamic load has to be repeated with modified parameters
Felix Dietrich | AWLC14 | 14.05.2014 | page 18
Addition
0,0 0,2 0,4 0,6 0,8 1,00
100
200
300
400 stress(z=0,014m) stress(z=0,0135m) stress(z=0,013m) stress(z=0,0125m) stress(z=0,012m) stress(z=0,0115m) stress(z=0,011m)
stre
ss [M
Pa]
time [µs]
Felix Dietrich | AWLC14 | 14.05.2014 | page 19
Addition
Felix Dietrich | AWLC14 | 14.05.2014 | page 20
Addition
0,0 0,5 1,0 1,5 2,0 2,5 3,00
50
100
150
200
250
300
350
400
450
stre
ss[M
Pa]
time[µs]
stress_max longtime stress_max shorttime