Critical Design Review (CDR)

GigaBIT M1 Primary Mirror

Optique Fullum · Atomaste ·

Agenda

  • Executive summary (decision + numbers)
  • Design evolution (PDR → CDR)
  • Selected design baseline (Option B)
  • Performance validation (WFE + margins)
  • Support system (whiffletree + lateral)
  • Reference frame & boundary condition strategy
  • Risks, schedule, next steps

Executive Summary

  • Status: design maturity sufficient to authorize blank procurement
  • Baseline for procurement: Option B (Conical V14)
  • Performance: WFE compliant with 36–71% margin
  • Mass: ~95 kg (allocation 103.5 kg)
  • Cost: $525k quote (0.4% above $523k target)
  • Schedule driver: 18‑week Schott lead time

Key Numbers (Option B)

  • WFE 40°–20°: 6.59 nm RMS (req ≤ 22 nm)
  • WFE 60°–20°: 14.79 nm RMS (req ≤ 22 nm)
  • Req interpretation: 22 nm RMS ≈ λ/25 @ 550 nm
  • Mass: ~95 kg (Δ = +8% margin to allocation)
  • Quote: $525,000 (received)
Section
Design Evolution (PDR → CDR)
Method improvements + scale of validation.

Transition to Atomizer Framework

  • Moved from HEEDS MDO → custom Atomizer framework
  • Full control of algorithms, tuning, and traceability
  • Automation: NX Nastran (SOL 101) + post-processing
  • Reproducible studies (version-controlled)

OPD‑Based Zernike Method (Upgrade)

  • Upgraded from Z‑displacement only → OPD‑based extraction
  • Captures lateral displacement (X,Y), not only axial (Z)
  • Directly supports the system WFE metric

Validation Scale

  • Total: 3,770+ FEA simulations
  • 14 design variables (support + ribs + blank geometry)
  • Algorithms: TPE, GNN, SAT v3, NSGA‑II, L‑BFGS
  • Convergence confirmed: multiple algorithms agree
Section
Cost Reduction Campaign
Manufacturing simplifications without sacrificing performance.

Cost Target & Outcome

  • Baseline quote: $625k
  • Budget target: $523k
  • Option B quote: $525k (0.4% above target)
  • Option C explored for manufacturability context (not procurement baseline)

Key Simplifications (Implemented in Option B)

  • Support interface taper simplified to 0° (cylindrical)
  • Center mini ribs removed (validated low sensitivity)
  • Guided by sensitivity + campaign data (not one-off changes)
Section
Selected Design Baseline
Option B (Conical V14) for procurement.

Option B (Baseline): Conical V14

  • Backface angle: 4.3° (optimized)
  • Center thickness: 85 mm
  • Mass: ~95 kg
  • Quote: $525k received

Figure placeholder: blank cross‑section export (CAD)

Why Option B (vs Flat‑Back Option C)

  • Option C has excellent polishing metric (MFG 90 ~28 nm) and slightly better WFE
  • Option B selected for maturity + procurement readiness (quote in hand, heritage)
  • Option C remains documented as a credible alternative for future trades
Section
Performance Validation
WFE compliance, dominant modes, uncertainty.

WFE Compliance (Operational)

  • Requirement: ≤ 22 nm RMS (relative to 20° reference)
  • 40°: 6.59 nm (≈71% margin)
  • 60°: 14.79 nm (≈36% margin)

WFE Breakdown (40° Case, Filtered RMS)

  • Dominant: Coma (J7–J8) ~60%
  • Astigmatism (J5–J6) ~33%
  • Higher order contribution is low

Uncertainty (RSS)

  • Combined WFE uncertainty estimate: ±1.2 nm
  • Worst-case 40°: ~7.5 nm (still far below 22 nm)
  • Support compliance target: reference frame deflection < 1 μm
Section
Support System
Architecture is defined; detailed design proceeds to FDR.

Axial Support: 54‑Point Whiffletree

  • Self‑equalizing hierarchical linkage (18 pads × 3 assemblies)
  • Load path: Mirror → pads → arms → 3 interface points → reference frame
  • RDOF joints provide thermal decoupling (insensitive in 100–200 Nmm range)

Figure placeholder: whiffletree schematic

Lateral Support (3‑Point)

  • 3 assemblies at 120°; determinate restraint
  • Includes spherical bearing (moment isolation) + radial slide (thermal) + Z stage (collimation)
  • At 20° elevation, each point carries ~35 kg lateral load (order of magnitude)

Figure placeholder: lateral support concept

Section
Reference Frame & Boundary Conditions
Procure now; develop frame in parallel.

Reference Frame (Scope Boundary)

  • Purpose: replicate fixed BCs used in optimization (stiff interface)
  • Status at CDR: conceptual; detailed design at FDR
  • Integration risk handled via ICD with StarSpec

Figure placeholder: reference frame concept topology

Why Ordering the Blank Now Is Safe

  • Diminishing returns: 3,770+ trials; improvements are flat
  • Separate problems: mirror optimization vs frame stiffness engineering
  • Parallel path: order blank now, design/validate frame while Schott fabricates
Section
Risks, Schedule, Next Steps
Residual risks are known and manageable.

Risk Snapshot (CDR)

  • Closed: design convergence (R1), mass allocation (R5), major methodology risks
  • Reduced: cost (R2), polishing difficulty (R6)
  • Open: Schott delivery delay (R3), interfaces (R7), material defect (R8)
  • New: reference frame stiffness (R4) — mitigated via parallel development

Path to FDR (High Level)

  • Schott PO: target Jan 20
  • Blank delivery: ~May 26 (18 weeks)
  • Reference frame concept review: Feb 28
  • Polishing start: Jun 2 → FDR: Sep

Decision Request

  • Approve CDR baseline: Option B (Conical V14)
  • Authorize Schott blank order (protect schedule)
  • Align on next steps: ICD + reference frame concept + support detailed design
Backup
Appendix
Detailed tables, parameters, and trades.

Backup: Option B Optimized Parameters (excerpt)

  • lateral_inner_angle 27.88°
  • lateral_outer_angle 13.19°
  • lateral_inner_pivot 7.41 mm
  • lateral_outer_pivot 11.53 mm
  • whiffle_min 58.90 mm
  • blank_backface_angle 4.30°

Backup: Requirement Matrix (excerpt)

  • REQ‑01 WFE 40°: 6.59 nm ≤ 22 nm ✓
  • REQ‑01 WFE 60°: 14.79 nm ≤ 22 nm ✓
  • REQ‑06 Mass: ~95 kg ≤ 103.5 kg ✓
  • REQ‑09 Thermal/STOP: pending (FDR)
Atomaste CDR · GigaBIT M1 Mirror