A method of space radiation environment reliability prediction

Transcription

1 A method of space radiation environment reliability prediction Qunyong Wang, Dongmei Chen, Hua Bai, PhD Beijing San-talking Testing Engineering Academy Co., Ltd. (STEA)

2 Overview and outline Background and introduction Space radiation effects Basic model Prediction method Case study Summary and Conclusion

3 Background and introduction Space radiation environment is an important failure causing factor in space vehicles. Flux(cm 2 /s) Traditional reliability prediction method did not include the space radiation environment reliability.mil-hdbk-217, No? FIDES, No? It s the time to set up a method of space radiation environment reliability prediction. Dose(krad/Si) LET(MeVcm 2 /mg) GEO/MEO SEE LET Al(mm) GEO/MEO TID

4 Space radiation environment effects Main failure modes Single Event Effect (SEE) Total Ionizing Dose (TID) Displacement Damage (DD) Radiation sensitive devices CPU, FPGA, SRAM, DC-DC, CCD, etc.

5 Space radiation environment effects SEE instantiate effects,11 effects TID,DD accumulated effects,parameter drift or functional failure No. Effects Acronyms Features 1 Single Event Upset SEU 2 Single Event Transient SET 3 Multiple-cell upset MCU 4 Single event functional interrupt SEFI 5 Single Event Disturb SED 6 Single Event Hard Error SEHE 7 Single Eevent Latch-up SEL 8 Single Event Burnout SEB 9 Single Event Gate Rupture SEGR 10 Single Event Dielectric Rupture SEDR 11 Single Event Snapback SESB Non-destructive Destructive

6 Space radiation environment reliability The failure rate induced by space radiation environment is the sum of 3 independent failure rates of SEE, TID and DD. space Basic Model λ = λ + λ + λ TID DD SEE Failure physics and test support

7 Space radiation environment reliability Basic Model Embedded into the traditional reliability prediction models based on stress-physical failures, e.g. FIDES model λ = λ physical Π PM Π Pr ocess FIDES model physical NOspace Focus λ = λ + λ space Related to the processes of quality control, manufacturing and usage of component and equipment. Failure physics and test support

8 SEE reliability prediction method SEE is an instantiate effect, only happens during power on time of the sensitive devices. Power on working hours yearly λ SEE Annual _ time SEEtype = SEEtype i λseetype i i 8760 Calendar hours in a year The i th SEE failure effects in 11 types.

9 SEE reliability prediction method SEE failure rate prediction method under radiation stresses. Radiation stresses Device Intrinsic characteristics on radiation resistance λ = SEEtype i 1 2π f ( LET, θ, φ) σ SEEtype i ( LET, θ, φ) dφd (cosθ ) dlet 4-parameter-Weibull fit description σ SEEtype i ( LET ) = σ sat 1 e LET LET W th S

10 TID/DD reliability prediction method Sensitive devices TID/DD failure distribution lognormal Intrinsic radiation resistance failure distribution parameters (µ,σ) n 1 µ = ln( n i= 1 R TIDFAIL i ), n 1 σ = [ln( n 1 i= 1 R TIDFAIL i ) µ ] 2 1/ 2 RTIDFAIL i :the failed TID for the i th sample (unit: rad) Radiation stresses Device Intrinsic characteristics on radiation resistance S TID ( T) ln( R = 1 Φ spectid σ ( T)) µ

11 TID/DD reliability prediction method For simple use in engineering, define an equivalent constant failure rate upon the equal end life survival probability. λ TID 1 = T 1 = T ln ( S ( T )) TID ln( R ln 1 Φ spectid ( T )) µ σ Note:Equivalent only within the design life T. λ TID Instantiate failure rate End life TID Real λ TID Equivalent λ TID TID

12 Case study Assume satellite to be launched in 2015,GEO,T=12 Y. Mission space radiation stresses: TID 69.8krad, LET spectrum in figure 1. Key sensitive device: SRAM-FPGA,TID/SEU sensitive Ground test: 60 Co TID, LINAC Heavy ion SEU INTEGRAL LET SPECTRUM FPGA (DUT) BRAM CLB Control Data Select MAP FPGA (Monitoring) Control Data PROM Flux(cm 2 /s) LET(MeVcm 2 /mg) GEO LET Test Board JTAG FPGA Simulator Current monitor and protection Single Chip Upstream computer FLASH FPGA SEU monitoring system

13 λ TID prediction FPGA failed TID test data Case study No R TIDFAIL krad(si) λ TID µ=12.3 σ= ln(69800) 12.3 = ln 1 Φ = h 1

14 λ SEE prediction Case study BRAM(8192bits)SEE test data CLB ( bits)SEE test data Heavy ion LET MeVcm 2 /mg Flences cm -2 SEU # σ cm 2 /device Heavy ion LET MeVcm 2 /mg Flences cm -2 SEU # σ cm 2 /device C E E-5 Cl E E-4 Ti E E-4 Cu E E-4 C E E-2 Cl E E-1 Ti E E-1 Cu E E-1 σsat=3.8e-8cm 2 /bit LETth=0.86MeV cm 2 /mg S=1.35 W=11.42 λ SEE =0.04#/day-device σsat=2.2e-8cm 2 /bit LETth=0.86MeV cm 2 /mg S=0.99 W=6.06 λ SEE =50.3#/day-device

15 Summary Based on space radiation failure mechanism and statistical characteristics, a method of space radiation environment reliability prediction is setup upon failure physics, which expands the application domain of reliability prediction. To be applied on reliability design for future space vehicles.