CENTRADIANT LDR: COMPREHENSIVE RISK ANALYSIS
CONFIDENTIAL: DO NOT PUBLISH
Date: February 12, 2026
Author: Walter Kueffer
Organization: Centradiant Space Systems
Design Baseline: Disk LDR, per DESIGN-V2-SYNTHESIS.md
1. Executive Summary
The Centradiant spinning disk liquid droplet radiator has 0 CRITICAL risks and 0 HIGH risks. The risk profile:
- 61.7% five-year system reliability (series model, all subsystems included)
- 15 tracked risk items, none exceeding MEDIUM severity
- All major design risks resolved through architecture choices
- Only open item: spoke thermal bow FEA verification [LOW priority]
The architecture achieves 4.4 kg/kW(th) specific mass at 934 kg launch mass while maintaining robust margins (34.2 deg C junction temperature margin, 1,053x spoke stress margin, >99.99% droplet capture efficiency). The design is ready for preliminary design review (PDR).
2. Architecture Risk Mitigations
The disk LDR architecture addresses the challenges that have historically blocked liquid droplet radiator deployment.
2.1 Droplet Collection Without Electromagnetic Steering
| Attribute |
Detail |
| Historical challenge |
Prior LDR concepts relied on electromagnetic or electrostatic steering for droplet collection, adding complexity, parasitic power, and failure modes. None reached TRL 5+. |
| Centradiant approach |
Centrifugal force from a spinning disk (2 RPM, 0.045g at 10m rim) provides passive, zero-power droplet collection. Droplets are atomized at the hub and drift radially outward into the rim collector. |
| Result |
Zero parasitic collection power. Collection mechanism has no active components, no electromagnetic interference, and no sensitivity to droplet charge state. |
| Evidence |
analysis/V3_disk_design.py |
2.2 Thermal Interface Without Fluid Rotary Seal
| Attribute |
Detail |
| Historical challenge |
Spinning LDR concepts require heat transfer between rotating and non-rotating frames. Fluid seals at rotary interfaces are high-failure-rate components. |
| Centradiant approach |
Galinstan liquid-metal thermal joint (100mm R x 300mm L, 0.5mm gap). Working fluid (CB-DC705) never crosses the rotary interface. Heat transfers conductively through the liquid metal annulus with helical groove enhancement. |
| Thermal penalty |
delta-T = 2.9 deg C across the Galinstan gap (negligible vs 34.2 deg C total margin) |
| Joint reliability |
99.6% over 5 years |
| Evidence |
analysis/V3_rotary_joint.py |
2.3 Droplet Emissivity via CB-DC705
| Attribute |
Detail |
| Historical challenge |
Silicone oil droplets at 377 um diameter have low infrared emissivity (epsilon = 0.45 for bare DC-705) due to thin-film absorption limitations. Bulk emissivity (0.9) does not apply at small path lengths. |
| Centradiant approach |
CB-DC705: DC-705 + 100 ppm carbon black (20-50 nm particles). The carbon black provides broadband IR absorption, raising droplet emissivity to epsilon = 0.999 at 377 um path length. |
| T_j margin |
34.2 deg C below the 83 deg C GPU junction temperature limit |
| Nozzle compatibility |
CB particles are 20-50 nm; nozzle orifice is 200 um (ratio > 4000:1). In-line 100 um filters provide secondary protection. |
| Evidence |
analysis/V3_droplet_emissivity.py |
2.4 Minimal Mesh Area
| Attribute |
Detail |
| Historical challenge |
Enclosure-based LDR concepts require large mesh surfaces to capture omnidirectional sprays, adding mass and deployment complexity. |
| Centradiant approach |
The disk geometry constrains droplets to radial trajectories in the disk plane. Only the peripheral rim requires mesh, yielding 94 m^2 of capture area and 19 kg of Ti-6Al-4V dual-layer mesh. |
| Evidence |
analysis/V3_disk_design.py |
2.5 Power Budget Closure
| Attribute |
Detail |
| Risk addressed |
Pump power (585 W) is explicitly included in the electrical power budget. Solar array is sized to 145 m^2 to account for full electrical load: GPUs, pumps, avionics, and thermal control. |
| Evidence |
analysis/V3_hx_pche_sizing.py, analysis/V3_solar_reconciliation.py |
3. Active Risks
3.1 Droplet Capture at Mesh: MEDIUM [TRL 3]
| Attribute |
Detail |
| Risk |
Droplets impacting the rim mesh must be captured without splashing or rebounding. Weber number at impact We = 24, which is in the moderate-impact regime. |
| Mitigation |
Dual-layer Ti mesh with oleophilic coating; graded mesh (coarse outer + fine inner); micro-textured surface to promote wetting; pulsed spray option to reduce simultaneous impact loading |
| Advantage |
Centrifugal force continuously presses captured fluid against mesh (0.045g at rim), aiding drainage into gutter channels |
| Validation needed |
Ground test T1: droplet impact on mesh in vacuum with centrifugal analog |
| Evidence |
analysis/V3_disk_design.py (Weber number calculation) |
3.2 DC-705 Fluid Properties in Vacuum: LOW-MEDIUM [TRL 5-6]
| Attribute |
Detail |
| Risk |
DC-705 evaporative loss, UV degradation, and long-term property stability in the space environment |
| Heritage |
STS-77 IDGE experiment (1996) demonstrated DC-705 droplet behavior in microgravity |
| Evaporative loss |
~0.9 kg/yr (Langmuir equation, worst-case solar exposure) |
| Open question |
Long-term UV stability of CB-DC705 blend (carbon black may provide UV shielding benefit) |
| Validation needed |
Extended vacuum UV exposure testing of CB-DC705 |
3.3 GPU Radiation Tolerance (700km SSO): MEDIUM [TRL 4-5]
| Attribute |
Detail |
| Risk |
700km sun-synchronous orbit experiences elevated radiation flux. Total ionizing dose and single event effects must be within GPU tolerance. |
| Mitigation |
8mm Al shielding per V3_radiation_environment.py. ECC memory. Checkpoint/restart for GPU workloads. |
| Validation needed |
TID and SEE testing of target GPU (H100/B200) behind 8mm Al equivalent shielding |
| Evidence |
analysis/V3_radiation_environment.py |
3.4 Nozzle Clogging: LOW [TRL 8-9]
| Attribute |
Detail |
| Risk |
200 um laser-drilled SS nozzle orifices could be blocked by particulates, CB agglomeration, or DC-705 degradation products |
| Mitigations |
(1) 27,248 nozzles provide massive redundancy: losing 1% has negligible thermal impact. (2) Two in-line 100 um sintered 316L SS filters upstream. (3) CB particles are 20-50 nm, providing a 4000:1 ratio to orifice diameter. (4) 200 um orifice provides 4x margin over blockage threshold. (5) Closed-loop system limits external contamination. |
| Validation needed |
Ground test T4: laser-drilled nozzle lifetime test with CB-DC705 fluid |
| Evidence |
analysis/V3_nozzle_clogging.py |
3.5 PCHE Fouling: LOW [TRL 4]
| Attribute |
Detail |
| Risk |
Printed circuit heat exchanger (2000 x 3.0mm semicircular channels, 100cm length) could foul over time, increasing thermal resistance |
| Mitigations |
DC-705 is an exceptionally stable silicone oil with minimal degradation products. Closed-loop system with in-line filtration. CB particles at 20-50 nm are three orders of magnitude smaller than channel diameter. |
| Validation needed |
Ground test T5: long-duration PCHE flow test with CB-DC705 |
| Evidence |
analysis/V3_pche_fouling.py |
3.6 Deployment Latching: MEDIUM [TRL 3-4]
| Attribute |
Detail |
| Risk |
Six telescoping CFRP spokes must extend from 0.30m hub to 10m radius and lock via collet mechanisms in zero-g. Rim tape spring hinges must deploy and latch to form a continuous annulus. |
| Heritage |
Solar sail boom deployment (JAXA IKAROS, NASA NEA Scout); large mesh antenna deployables (Harris/L3); tape spring hinges (ESA Proba series) |
| Mitigations |
Motorized deployment with position telemetry at each stage; autonomous halt-and-retry logic; no irreversible pyrotechnic releases (all mechanisms retractable for re-attempt) |
| Validation needed |
Ground test T2: deployment mechanism in thermal vacuum chamber with gravity offload |
| Evidence |
analysis/V3_deployment.py |
3.7 Spoke Tension Under Rotation: LOW-MEDIUM [TRL 4]
| Attribute |
Detail |
| Risk |
Each spoke carries 112.5 N hub tension from centrifugal loading of the rim and spoke self-mass at 2 RPM. Over a 5-year mission, spokes experience approximately 5.26 million load cycles. |
| Safety margin |
1,053x CFRP tensile yield at the hub attachment point |
| Fatigue |
Published CFRP fatigue data shows > 10^8 cycle endurance at stress ratios far exceeding operating loads. The 1,053x margin puts operating stress well below the fatigue endurance limit. |
| Concern |
Creep at hub attachment fittings (metal-to-CFRP bonded joint); thermal cycling effects on adhesive |
| Mitigation |
Mechanical interlocking joint design (not adhesive-only); spoke tension telemetry for health monitoring |
| Evidence |
analysis/V3_spoke_dynamics.py |
3.8 Spoke Thermal Bow: LOW [ONLY OPEN ITEM]
| Attribute |
Detail |
| Risk |
Differential solar heating across the spoke cross-section (sun-facing side vs shadow side) induces a temperature gradient of approximately 4 deg C, causing the spoke to bow. |
| Analytical prediction |
0.8mm tip deflection over 9.7m spoke length (from V3_spoke_dynamics.py) |
| Impact |
Negligible effect on droplet capture geometry. The rim height is +/-0.75m (1.5m total), so 0.8mm bow is < 0.1% of the capture zone dimension. |
| Status |
This is the only remaining open analysis item. A 3D transient thermal-structural FEA is needed to confirm the analytical bending model. |
| Expected outcome |
FEA will confirm or slightly refine the 0.8mm estimate. CFRP's low CTE makes this inherently a small effect. |
| Fallback |
If bow is larger than expected, increase rim height (design margin available) or add passive thermal coatings to equalize spoke temperature. |
| Evidence |
analysis/V3_spoke_dynamics.py |
3.9 Galinstan Containment: LOW [TRL 4-5]
| Attribute |
Detail |
| Risk |
The Galinstan liquid-metal thermal joint (100mm R x 300mm L, 0.5mm gap) must retain Galinstan over 5 years without leakage. |
| Containment method |
Galinstan is retained by its own surface tension (high surface tension ~0.5 N/m) plus labyrinth seals at annulus ends. No dynamic O-ring seals required. |
| Material compatibility |
Galinstan corrodes aluminum. All wetted surfaces use stainless steel or nickel alloy interfaces. |
| Failure mode |
Graceful degradation: partial Galinstan loss increases thermal resistance (higher delta-T), reducing available cooling power. This is a performance degradation, not a catastrophic failure. |
| Validation needed |
Ground test T3: Galinstan long-term stability in representative geometry with thermal cycling |
| Evidence |
analysis/V3_galinstan_corrosion.py |
3.10 Gyroscopic Coupling: LOW [TRL 5-6]
| Attribute |
Detail |
| Risk |
The spinning disk acts as a gyroscope, coupling attitude maneuvers across axes. Slew commands produce cross-axis torques that must be managed by the ADCS. |
| Stability |
I_spin / I_transverse = 2.0. The disk is a major-axis spinner, which is inherently stable against wobble (energy dissipation damps nutation rather than amplifying it). |
| Angular momentum |
18.4 Nms at 2 RPM, modest compared to large spacecraft reaction wheels |
| ADCS sizing |
4 x 1.5 kg reaction wheels sized for worst-case gyroscopic torque during orbit-maintenance maneuvers |
| Mitigation |
Spin-axis pointing along orbit normal minimizes coupling with orbit-keeping burns. Precession-compensating feedforward in ADCS software. |
| Evidence |
analysis/V3_spoke_dynamics.py (gyroscopic coupling section) |
4. Risk Summary Matrix
| # |
Subsystem |
Risk Description |
Severity |
TRL |
Status |
Evidence Script |
| 1 |
Rotary Interface |
Thermal joint design eliminates fluid seal risk |
RESOLVED |
N/A |
Architecture feature |
V3_rotary_joint.py |
| 2 |
Thermal |
CB-DC705 achieves near-unity droplet emissivity |
RESOLVED |
5 |
Architecture feature |
V3_droplet_emissivity.py |
| 3 |
Structure |
Disk geometry minimizes mesh area to 94 m^2 |
RESOLVED |
5 |
Architecture feature |
V3_disk_design.py |
| 4 |
Power |
Pump power included, solar array sized |
RESOLVED |
5 |
Architecture feature |
V3_hx_pche_sizing.py |
| 5 |
LDR |
Droplet capture at mesh (We = 24) |
MEDIUM |
3 |
Needs ground test |
V3_disk_design.py |
| 6 |
Compute |
GPU radiation tolerance (700km SSO) |
MEDIUM |
4-5 |
Needs testing |
V3_radiation_environment.py |
| 7 |
Deployment |
Deployment latching in zero-g |
MEDIUM |
3-4 |
Needs ground test |
V3_deployment.py |
| 8 |
Fluid |
DC-705 vacuum properties |
LOW-MEDIUM |
5-6 |
STS-77 heritage |
--- |
| 9 |
LDR |
Nozzle clogging (200 um laser SS) |
LOW |
8-9 |
4x margin, massive redundancy |
V3_nozzle_clogging.py |
| 10 |
Structure |
Spoke tension fatigue |
LOW-MEDIUM |
4 |
1,053x margin |
V3_spoke_dynamics.py |
| 11 |
Structure |
Spoke thermal bow |
LOW |
4 |
Open |
V3_spoke_dynamics.py |
| 12 |
Rotary Interface |
Galinstan containment |
LOW |
4-5 |
Graceful degradation |
V3_galinstan_corrosion.py |
| 13 |
ADCS |
Gyroscopic coupling |
LOW |
5-6 |
Major-axis stable |
V3_spoke_dynamics.py |
| 14 |
Thermal |
PCHE fouling |
LOW |
4 |
Mitigated by filtration |
V3_pche_fouling.py |
| 15 |
Deployment |
Deployment complexity |
RESOLVED |
3-4 |
48 parts, heritage mechanisms |
V3_deployment.py |
Risk Count Summary
| Severity |
Count |
Notes |
| CRITICAL |
0 |
None |
| HIGH |
0 |
None |
| MEDIUM |
3 |
Droplet capture, GPU radiation, deployment latching |
| LOW-MEDIUM |
2 |
Fluid properties, spoke fatigue |
| LOW |
5 |
Spoke bow, Galinstan, gyroscopic, PCHE fouling, nozzle clogging |
| RESOLVED |
5 |
Architecture features addressing known LDR challenges |
No risk exceeds MEDIUM severity. All MEDIUM risks are addressable through standard ground testing.
5. Ground Test Requirements
The following ground tests address the remaining MEDIUM and LOW-MEDIUM risks. Full test specifications are defined in analysis/V3_ground_prototype.py (total budget: $225K, 6-month program).
| Test ID |
Description |
Target Risks |
Priority |
Estimated Cost |
| T1 |
Droplet capture: emissivity validation + mesh impact testing in vacuum with centrifugal analog |
3.1 Droplet capture, 2.3 Emissivity confirmation |
HIGH |
Included in prototype budget |
| T2 |
Deployment mechanism: spoke extension and rim deployment in thermal vacuum with gravity offload |
3.6 Deployment latching |
HIGH |
Included in prototype budget |
| T3 |
Galinstan joint: long-term thermal cycling and containment verification |
3.9 Galinstan containment |
MEDIUM |
Included in prototype budget |
| T4 |
Laser-drilled nozzle: lifetime test with CB-DC705 fluid (accelerated aging) |
3.4 Nozzle clogging |
LOW |
Included in prototype budget |
| T5 |
PCHE fouling: long-duration flow test with CB-DC705 at representative temperatures |
3.5 PCHE fouling |
LOW |
Included in prototype budget |
These tests advance the system from TRL 3-4 to TRL 5, supporting the ground prototype milestone and subsequent PDR.
6. Open Items
There is exactly one open analysis item:
Spoke Thermal Bow FEA Verification [LOW Priority]
| Attribute |
Detail |
| Current status |
Analytical model predicts 0.8mm tip deflection from 4 deg C sun/shadow gradient across 80mm OD CFRP spoke |
| What is needed |
3D transient thermal-structural FEA incorporating: (1) orbital thermal cycling, (2) CFRP orthotropic properties, (3) spoke-hub and spoke-rim boundary conditions, (4) radiation view factors |
| Expected outcome |
Confirmation of the 0.8mm order-of-magnitude estimate. CFRP's coefficient of thermal expansion (~1 um/m/K axial) inherently limits this effect. |
| Impact if worse than expected |
Rim height can be increased from +/-0.75m. Current rim captures droplets across a 1.5m band; even a 10x increase in bow (8mm) would be < 1% of capture zone. |
| Schedule impact |
None. This is a verification task, not a design driver. Can be completed in parallel with other PDR activities. |
| Priority justification |
LOW: the analytical model is conservative, the margin is large, and the fallback is straightforward. |
7. Conclusion
The Centradiant disk LDR architecture has a robust risk profile:
No CRITICAL or HIGH risks. The Galinstan thermal joint eliminates the need for a fluid rotary seal. CB-DC705 provides 34.2 deg C of junction temperature margin. These architecture choices resolve the challenges that have historically blocked LDR deployment.
Mass-efficient design. 934 kg launch mass with 94 m^2 of mesh (19 kg). The 4.4 kg/kW(th) specific mass is 6-8x lighter than conventional panel radiators, enabling Falcon 9 rideshare compatibility.
61.7% five-year system reliability (series model). Driven by high-heritage subsystems (5 of 9 subsystems at ISS or flight heritage levels), massive nozzle redundancy (27,248 laser-drilled SS nozzles with 4x clog margin), and graceful degradation modes.
Remaining risks are tractable. The three MEDIUM-severity risks (deployment latching, droplet capture, GPU radiation) are all addressable through standard ground testing and have clear mitigation paths. None requires a design change.
One open item. Spoke thermal bow FEA verification is the sole remaining analytical task. It is LOW priority with large design margins available.
The design is ready for preliminary design review (PDR). The $225K ground prototype program (V3_ground_prototype.py) addresses all MEDIUM-priority risks within a 6-month timeline, advancing the system to TRL 5.
Document generated from analysis suite. All quantitative claims traceable to scripts in analysis/.
Cross-reference: DESIGN-V2-SYNTHESIS.md (authoritative design document), V3_final_budget.py (mass/risk rollup), V3_executive_summary.py (program overview).