Propellant estimation by Thermal Gauging Method (TGM) Dr Boris Yendler YSPM
Agenda • Introduction • How Thermal Gauging Method (TGM) can help satellite operator • How YSPM can help satellite manufacturer to make satellite TGM “friendly” • Basic of Thermal Gauging Method (TGM) • Requirements for using TGM • Comparison with book-keeping and PVT • Example of TGM estimation • Looking back – Past performance – Awards – Testimonial • Conclusion • References 2
How TGM will help Satellite Operator
Benefits to Operator • More accurate estimation of propellant remaining – TGM is more accurate than book-keeping and PVT at EOL • TGM is independent method – book-keeping (BK) and PVT methods are NOT independent (both use pressure transducer) • Increase confidence in accurate determination of EOL – use of independent methods increase reliability of estimation (BK and PVT methods are NOT independent) • TGM helps an Operator to make accurate business decision 4
How YSPM helps Satellite Manufacturer
Designing Satellite being TGM “friendly” YSPM will work with satellite manufacturer to make satellite TGM “friendly” on design stage. We will help to determine an optimal designs of: • Heater – Position on a tank – Shape – Ground control – Power • Temperature sensor – Position on a tank – Accuracy – Telemetry A/D and D/A conversion • Tank thermal connection: – To s/c environment, e.g. , optical properties of MLI, panels, etc – Between tanks (multi-tank system) • Allowable temperature rise 6
Thermal Gauging Method Basic
Basics Measure a propellant tank load using temperature rise • Temperature rise can be induced by: Tank heaters; Sun load; Equipment (e.g. IRU unit on BSS 601); etc. • Thermal Gauging Method (TGM) accuracy improves with load propellant load decrease because sensitivity of temperature rise to tank load is increasing when tank load drops • The method is capable of gauging: - individual tanks in multi-tank propulsion systems with no separation valve - Mono and bi propellant propulsion systems 8
TGM Phases Regardless of spacecraft type, Thermal Gauging method follows the same phases 1. Build integrated Thermal Model (Tank(s) and Spacecraft) 2. Prepare and Conduct in-flight test (tanks heating and cooling) 3. Calibrate integrated model per flight conditions 4. Find propellant load of each tank 5. Determine accuracy of the estimation 9
Requirements for estimation • Spacecraft design – to build Tank and Spacecraft Thermal Models • Tank temperature – typically propellant tanks have thermistors A mean of changing tank temperature – heater • (tank, bus unit, payload, etc), sun NOT MUCH 10
Comparison with other Propellant Gauging Methods
Methods of Gauging • Bookkeeping- calculate consumed propellant ( includes ∆ V, ranging, etc ) – Accuracy worse over time due to accumulation of error • Pressure, Volume, Temperature (PVT) - calculate remaining propellant based on Gas Law ( including variants like re-pressurization ) − Accuracy worse over time due to lost of sensitivity of He pressure to volume change in tanks with low propellant load • Thermal Methods - calculate remaining propellant based on temperature rise ( Including ESA TPGS, Comsat PGS, TGM, …) + Accuracy better over time 12
Bookkeeping vs. Thermal Gauging Method Tank Initial Load = 500 kg 450 kg consumed 50 kg remaining • Bookkeeping accuracy is calculated based on consumed fuel Assuming accuracy of 2% ; uncertainty – 450 kg x 2% = 9 kg • TGM accuracy is calculated based on remaining fuel Assuming accuracy – 12%; uncertainty – 50kg x 12% = 6 kg 13
PVT vs. TGM at BOL Assuming: propellant tank ≈500 liter; accuracy of PVT – 2%; TGM – 12% Beginning of Mission (BOL) PVT Thermal – gas volume 1 liter; using 1 – Propellant load 499 kg; liter of propellant doubles using 1 kg of propellant gas volume- pressure reduces mass by 0.5%; reduces 50% small change in slope of temperature rise – 2% accuracy of gas volume is 0.2 liter – 12% accuracy is 60 – ( ≈ 0.2 kg ) kg 14
PVT vs. TGM at EOL Thermal PVT – gas volume 480 liters; using – Propellant load 20 kg; using 1 liter of propellant 1 kg reduces mass by 5%; increases He volume by significant change in 0.2%- pressure reduces thermal response 0.2% – 12% accuracy is 2.4 – 2% accuracy of gas kg volume is 9.6 liters ( ≈ 9.6 kg ) 15
Comparison (example of generic spacecraft) High Accuracy Low End Beginning Book-keeping, PVT • – High accuracy at Beginning of Life (BOL) through Middle of Life (MOL) – Low accuracy at End of Life (EOL) • Thermal Gauging – High accuracy towards EOL 16
Example of TGM Extimation
Step 1a-Tank High Fidelity Model • 3-D propellant distribution in the tank using Surface Evolver • Grid for Finite Element Model (FEM) − high enough density to simulate temperature gradients • More then 20000 nodes • Detailed propellant and temperature distribution • Simulation run time (6 – 10 hours per run)
Tank High Fidelity Model-cont’d Tank Model Temperature Distribution (heaters are on domes)
Step 1b - Satellite Models West West East East BSS 601 (Ref.1) SpaceBus 2000 (Ref.2) StarDust (Ref.4) 20 EuroStar 2000 (Ref.3)
Step 2a- Test Procedure Operational Constrains • Avoid eclipse season (change of thermal condition) • No change in payload/Bus unit configuration (change of thermal condition) • No station-keeping maneuvers performed (change of propellant load, sloshing) • Enough time to cool-down for the tanks after turning heaters OFF • Tank temperature can not exceed qualification limit Get approval from Manufacturer before the test
Step 2b- in-flight test Heaters ON (Fig.4 from Ref.2 )
Step 3 - S/C Model Calibration • No ground calibration is required • Calibration is performed using current flight data • Calibration of satellite model to reflect current condition of the satellite
Step 4 -Propellant Estimation Lines – simulation results; Markers – Temperature Sensor reading Tank heaters were turned ON at t=0 Flight vs Simulation (Fig.5 from Ref.2)
Error Analysis – Step 5
Categories of Uncertainty Two categories of uncertainty • A least squares curve fit and associated uncertainty • Uncertainties of specific model parameters – Physical parameters – Temperature measurement – Numerical model
Error Analysis Starting Point • Satellite data: ( T i , t i ) • Simulation curves: T ( t , m , p 1 , p 2 , p 3 ,…, q 1 , q 2 , q 3 ,…) • Uncertainties for q parameters: σ qi 27
Least Squares Analysis Mismatch Function M [ ] ∑ = − 2 ( , , , ,..., , ,...) M T T t m p p q q i i 1 2 1 2 i 0.4 Load is determined by 0.35 Minimizing function M Mismatch Function 0.3 0.25 with respect to propellant 0.2 mass 0.15 0.1 0.05 0 0 1 2 3 4 5 6 7 8 9 10 11 12 Load [kg] (Fig. 4 from Ref.4) 28
Uncertainty • Assuming that the model is a good fit apart from statistical errors, 2 2 ∂ ∂ m m ∑ ∑ σ = σ + σ 2 2 2 ∂ ∂ m T q T q j i j i j • These can all be calculated. The variance of T i comes out of the least squares fit if we assume they are all equal. 29
TGM Accuracy of Estimation Bottom Line • Theoretical accuracy is determined by uncertainty analysis (Phase 5) • Theoretical uncertainty is conservative • Actual accuracy can be determined ONLY after tank(s) depletion • Existing flight data indicate that Actual accuracy of Thermal Gauging Method is about 12% - 15% of propellant remaining
Typical Schedule of TGM estimation • paper w ork SOW, NDA, Contract – 3 weeks • Model development – 2 weeks • In-flight test – 2 weeks – 2 weeks • Model Calibration • Propellant Estimation – 2 weeks • Uncertainty Analysis – 1 week • Final Report Total – 12 weeks 31
Typical Deliverables • One summary report with test procedure • One summary report with propellant estimation • One summary with accuracy of estimation • One final report 32
Looking Back • Past Performance • Awards • Testimonials
Past Performance - S/C Platforms • My experience includes more than 45 thermal gauging estimations during last 7 years including the following platforms: – Alcatel/TAS France SpaceBus 2000, 3000A – Astrium/EADS EuroStar 2000 – Boeing SS 376, 601 – LM A2100, Ax2100, series 3000, 5000,7000 – US Government – SS/Loral FS1300 – NASA (StarDust) 34
S/C Platforms – cont’ • Majority of spacecrafts have tank heaters and thermistors • Thermal gauging has being successfully used on spacecrafts not designed specially for the approach, like StarDust, SS/L FS1300, SpaceBus 2000, etc • Thermal gauging was even successfully used for BSS 601 which does not have tank heaters 35
Customers and Awards My customers include but not limited to : USA (Loral Skynet); US Government (USAF, NASA); Japan SkyPerfect (JSAT, SCC) ; Turkey ( Turksat); France (Thales); Canada (Telesat), Saudi Arabia (Arabsat); etc. COMSAT PGS group received 2006 US Air Force Chief of Staff Team Excellence Award 36
Recommend
More recommend
