engineering-skills/additive-manufacturing/references/fatigue-design.md

Fatigue Design for Additive Manufacturing

Use this file whenever the component is subject to cyclic loads, vibrations, repeated impacts, or when the application is fatigue-critical (aerospace, biomedical, structural automotive, pressure vessel, moving mechanisms).

Static UTS strength alone is not sufficient to evaluate an AM part under cyclic loading.

---

1. When Fatigue Governs

Is the component subject to cyclic loads?
│
├── NO → static design with FS ≥ 2.0 on UTS. Done.
│
└── YES → how many cycles?
    ├── N < 10^3 cycles (extreme LCF)
    │   └── Use UTS with FS ≥ 1.5–2.0. Verify plastic deformation.
    ├── 10^3 < N < 10^4 cycles (LCF)
    │   └── Use σ_LCF ≈ 0.7–0.9 × UTS. FS ≥ 1.5.
    └── N > 10^4 cycles (HCF — the critical case)
        └── Use S-N data (section 3) + knockdown factors (section 2).
            The fatigue limit may be 30–60% of the UTS value.

Questions to ask the designer before proceeding:

• Expected total cycles (10^5, 10^6, 10^7)?
• Stress ratio R = σ_min / σ_max (typical: R = 0.1 pulsating, R = -1 fully reversed)?
• Will the fatigue-critical surface be machined or remain as-built?
• Is HIP planned?

---

2. Knockdown Factors for AM

2A — Kf from Surface Roughness (Ra)

The as-built roughness of AM parts is much higher than forged — surface valleys act as notches and drastically reduce fatigue life.

Ra (µm)	Typical condition	Kf (Ti-6Al-4V)	Kf (AM steels)	Kf (AlSi10Mg)
20–35	LPBF as-built (side)	1.8–2.5	1.6–2.2	1.5–2.0
8–20	LPBF as-built (top XY)	1.4–1.8	1.3–1.7	1.3–1.6
6–12	Bead blast post LPBF	1.3–1.6	1.2–1.5	1.2–1.4
3–6	Vibratory finishing	1.1–1.3	1.1–1.3	1.1–1.2
0.8–3	Electropolishing / SLA	1.05–1.15	1.0–1.1	1.0–1.1
0.4–0.8	CNC machining	1.0–1.05	1.0–1.05	1.0–1.05
< 0.4	Grinding / lapping	1.0	1.0	1.0

Effective fatigue limit = baseline S-N / Kf

Example: Ti-6Al-4V LPBF as-built side (Ra 20µm, Kf 2.2): effective fatigue limit = 260 MPa / 2.2 = 118 MPa — vs. 620 MPa for forged.

2B — Knockdown from Porosity

For every 0.1% of pore volume (measured by CT scan or Archimedes):

• Fatigue life reduction: −5 to −8% (LPBF metals, Ti and Al)
• Acceptance thresholds:
  • Porosity < 0.05%: acceptable for fatigue with HIP
  • Porosity 0.05–0.5%: acceptable only for non-fatigue-critical applications
  • Porosity > 0.5%: not suitable for cyclic loading — mandatory HIP required or reject

HIP closes spherical porosity (gas, keyhole) → reduces porosity from typical 0.3–0.5% to < 0.05%. HIP does NOT close planar LOF (lack of fusion) — these remain as cracks.

2C — Directional Anisotropy (Fatigue)

Process/Material	Fatigue Z/XY as-built	Fatigue Z/XY post-HIP
LPBF Ti-6Al-4V	0.60–0.75	0.92–0.98
LPBF AlSi10Mg	0.55–0.70	0.85–0.95
LPBF 316L	0.70–0.85	0.92–0.98
LPBF 17-4PH	0.70–0.80	n.d.
LPBF IN718	0.65–0.80	n.d.
SLS PA12	0.90–1.00	n/a
FDM PA12 (0°/90°)	0.35–0.55	n/a

Rule: orient the primary cyclic loading plane (σ_max) in the XY direction. In fatigue-critical zones, the load must be perpendicular to the layer lines, not parallel.

2D — Effect of HIP on Fatigue

HIP (Hot Isostatic Pressing) is the most effective treatment for improving fatigue life in AM metals:

• Closes spherical porosity → eliminates the primary internal crack initiator
• Reduces Z/XY anisotropy → from 0.6–0.75 to 0.92–0.98 for Ti
• Modifies microstructure → improves ductility but may reduce UTS if not followed by aging

Fatigue life improvement post-HIP (relative to as-built):

• Ti-6Al-4V LPBF: +50–100% in terms of cycles to failure
• AlSi10Mg LPBF: +30–60%

---

3. Baseline S-N Data (R = 0.1, HCF at 10^7 cycles)

These are representative ranges from the literature. Inter-lot variability ±15–20%. For critical applications: require testing on coupons from the same lot/build plate.

LPBF Metals

Material / Condition	Fatigue limit @ 10^7 cycles (MPa)	Notes
Ti-6Al-4V LPBF as-built	200–320	High scatter; Ra 15–25 µm side
Ti-6Al-4V LPBF HIP + machined	400–550	Close to forged
Ti-6Al-4V forged (reference)	620–700	Benchmark
AlSi10Mg LPBF as-built	90–130	Very sensitive to orientation
AlSi10Mg LPBF HIP + T6-equiv.	120–170	+30% vs. as-built
AlSi10Mg forged 6061-T6 (ref.)	95–110	AM comparable with HIP
316L LPBF as-built	180–220	Good relative to forged
316L LPBF HIP	220–260
316L forged (reference)	200–240	AM as-built nearly comparable
17-4PH LPBF H900	350–430	Only after mandatory aging
17-4PH forged H900 (ref.)	400–500
IN625 LPBF	280–380
IN718 LPBF (full HT)	350–450	Mandatory double aging
CoCr LPBF (biomedical)	500–600	Excellent for implants

AM Polymers

Material / Condition	Fatigue limit @ 10^6 cycles (MPa)	Notes
PA12 SLS	18–25	R = 0.1; sensitive to moisture
PA12 FDM (0°)	12–18	Layer bonding is the weak link
PETG FDM	10–16
ABS FDM	8–14	Highly anisotropic in Z
PEEK FDM	25–40	Only with heated chamber, correct orientation

Mean Stress Correction (Goodman)

For R ≠ 0.1, correct using the modified Goodman relation: σ_a / σ_fl + σ_m / UTS = 1

Where: σ_a = alternating amplitude, σ_m = mean stress, σ_fl = fatigue limit (from table).

---

4. Shot Peening and Surface Treatments

Shot peening benefit (quantitative)

Shot peening introduces compressive residual stresses at the surface that oppose the propagation of fatigue cracks.

Material	Fatigue limit improvement	Introduced σ_residual
Ti-6Al-4V AM	+20–40%	−400 to −700 MPa
AlSi10Mg AM	+15–25%	−200 to −400 MPa
316L / 17-4PH AM	+15–30%	−300 to −500 MPa
PA12 SLS	not applicable	—

Reference standard: AMS 2430 (aerospace); Almen intensity A8–A12 for Ti.

Deep Rolling (for cylindrical features: shafts, pins, fillets):

• σ_residual: −600 to −900 MPa (deeper than shot peening)
• Fatigue improvement: +30–50%
• Requires machine tool access; not applicable to complex geometries

Correct Sequence (DO NOT deviate)

AM Print
  ↓
Stress Relief (mandatory before removing from build plate for metals)
  ↓
HIP (if fatigue-critical)
  ↓
Specific Heat Treatment (aging 17-4PH, double aging IN718, etc.)
  ↓
Support Removal
  ↓
CNC Machining of critical surfaces (after HIP — HIP modifies dimensions ±0.05–0.2%)
  ↓
Shot Peening (ALWAYS AFTER all heat treatments)
  ↓
Final Inspection (CT scan + CMM)

CRITICAL: Do not perform shot peening before HIP — HIP relaxes the compressive stresses introduced by shot peening, negating its benefit.

---

5. DfAM Rules for Fatigue

1. Fillets in load paths: minimum radius ≥ 2× thickness of the adjacent wall. Sharp fillets (R < 0.5mm) → Kt 3–5 → guaranteed crack initiators.

2. Build orientation: primary cyclic load axis → XY plane. If not possible (complex geometry): specify HIP as mandatory.

3. As-built surfaces in critical zones: not acceptable for N > 10^6 cycles without treatment. Minimum: bead blast (Ra 6–12 µm, Kf still 1.3–1.6). Optimal: machining or shot peening.

4. Lattice in fatigue-critical applications:

   • Add solid skin ≥ 1.5mm on all surfaces exposed to cyclic loading.
   • The surface of an as-built lattice has Ra 30–80 µm → Kf 2–4 → drastically reduced life.
   • For fatigue-critical lattice: TPMS Gyroid + HIP + machined outer skin.

5. Holes in cyclic tension zones:

   • Kt of a circular hole in a plate = 3. In AM, the as-built hole edge has Ra 15–40 µm → effective Kf 4–6.
   • Solution: ream/mill the hole after printing (even with a manual reamer).

6. Support attachment marks (residual marks):

   • Support attachment points leave craters/bumps Ra 25–100 µm → critical initiators.
   • Do not leave support marks on fatigue-critical surfaces. Design orientation to avoid this.

---

6. Red Lines — Mandatory Stop

These scenarios require corrective action before proceeding. Ignoring them is not acceptable — communicate them explicitly to the designer.

Condition	Required action
Ra as-built > 6 µm + N > 10^6 cycles on metal AM	Mandatory: machining or shot peening. Do not proceed as-built.
CT scan porosity > 0.5% + fatigue-critical application	Mandatory: HIP. If unavailable, reject the part.
LOF defects detected by CT scan in load zone	Reject. LOF acts as a crack (Kf 3–10). HIP does not close LOF.
FDM polymer + N > 10^5 cycles in Z direction	Process not suitable. Switch to SLS PA12 or redesign orientation.
As-built lattice without skin + cyclic fatigue	Redesign: add skin ≥ 1.5mm or exclude lattice from critical zone.
Shot peening planned before HIP	Reverse the sequence. Shot peening must be the last thermal/mechanical step.

---

7. Fatigue Evaluation Plan (Output for the Designer)

When fatigue is identified as relevant, include in the final output:

## Fatigue Evaluation

**Regime:** HCF (N = __ cycles, R = __)

**Material/Process:** __ LPBF / condition: as-built / HIP / machined

**Baseline fatigue limit:** __ MPa (from S-N table, section 3)

**Applied knockdown factors:**
- Kf from Ra (__ µm, __ condition): __ → effective limit = __ MPa
- Porosity (CT scan planned? __): estimated knockdown __%
- Anisotropy (load in __ direction): Z/XY factor = __

**Estimated effective fatigue limit:** __ MPa

**Comparison with applied load (σ_max = __ MPa, σ_a = __ MPa):**
Fatigue FS = __ [target ≥ 1.5 for standard applications, ≥ 2.0 for safety-critical]

**Actions required to reach target FS:**
□ Mandatory HIP
□ Machining of critical surfaces (Ra target ≤ __ µm)
□ Shot peening (AMS 2430, Almen __)
□ CT scan post-HIP (acceptance: no LOF, porosity < 0.05%)
□ Fatigue coupons same build plate (required for qualification)

**Red lines:**
[List any stop conditions identified]