Thermal Shock vs Thermal Cycling
Quality & Inspection Standards
PRODUCTS CENTER
Get Custom Quote
PRODUCTS CENTER
Thermal Shock vs Thermal Cycling – Which Test Reveals Your PCB’s True Weakness?
A costly mis‑selection – thermal shock used where thermal cycling was needed
An automotive ECU supplier qualified a PCB using thermal shock (2.6.7.1, 100 cycles -40/+125°C, liquid‑to‑liquid). All samples passed. After 18 months in the field, solder joint cracks caused intermittent failures. Root cause: the actual under‑hood stress was slow thermal cycling (ramp rate 5°C/min), not rapid thermal shock. The solder joints had sufficient margin for shock but not for fatigue under slower, repeated cycling. The lesson: thermal shock finds brittle fractures and delamination; thermal cycling reveals solder joint fatigue. The supplier requalified using 1500 cycles of air‑to‑air thermal cycling (2.6.7) and detected the weakness.
Thermal Shock vs Thermal Cycling – Not Just a Speed Difference
Many engineers assume thermal shock is simply a faster version of thermal cycling. That is incorrect. The underlying physics of failure differ fundamentally.
| Parameter | Thermal Shock (2.6.7.1) | Thermal Cycling (2.6.7) |
|---|---|---|
| Medium | Liquid‑to‑liquid (typically two baths) | Air‑to‑air (single chamber with forced convection) |
| Ramp rate | >30°C/min (up to 60°C/min) | 5‑15°C/min (typical) |
| Dwell time | Short (5‑15 minutes, often just stabilization) | Longer (15‑60 minutes, to reach temperature equilibrium) |
| Primary failure mode | Brittle fracture (delamination, ceramic capacitor cracks, brittle IMC fractures) | Fatigue (solder joint cracks, copper fatigue, PTH barrel cracks) |
| Typical cycles to failure | 50‑200 cycles (high stress per cycle) | 500‑2000 cycles (lower stress per cycle, cumulative damage) |
| IPC‑6012 Class 3 requirement | Not required (space applications use liquid‑to‑liquid) | 500 cycles (air‑to‑air, -40/+85°C or -40/+125°C) |
| Acceleration model | Not easily modeled (complex fluid dynamics) | Coffin‑Manson (ΔT exponent 1.9, frequency exponent 0.33) |
Key insight: Thermal shock is a “go/no‑go” test for brittle materials and interfaces. Thermal cycling is a fatigue test for ductile structures (solder joints, copper). Using the wrong test can hide critical failure modes.
Failure Mechanisms – What Each Test Accelerates
The table below maps failure modes to the correct thermal stress test. Selecting the wrong test invalidates qualification.
| Failure Mode | Detected by Thermal Shock? | Detected by Thermal Cycling? | Why |
|---|---|---|---|
| PCB delamination (laminate) | ✔ High | Low (slow ramp may not create enough internal stress) | Rapid thermal expansion differential causes layer separation. |
| Ceramic capacitor cracks | ✔ High | Low (slow ramp allows stress relaxation) | Brittle ceramic fails under rapid ΔT. |
| Solder joint fatigue (BGA, QFN) | Low (fast cycles may not allow creep) | ✔ High | Cumulative plastic strain from repeated cycles. |
| PTH barrel crack (copper fatigue) | Low | ✔ High | Copper ductility requires many cycles to fatigue. |
| Brittle IMC fracture (ENIG black pad) | ✔ High | Moderate | Brittle IMC layer fails under shock, may survive slow cycling. |
| Underfill / epoxy cracking | ✔ High | Moderate | Fast temperature change stresses interfacial bonds. |
Test Methods – IPC‑TM‑650 2.6.7 and 2.6.7.1
Both methods are defined in IPC‑TM‑650, but they are not interchangeable. Understand the specific requirements of each.
- 2.6.7 (Thermal Cycling, Air‑to‑Air) – Samples are placed in a single chamber with forced air circulation. Temperature is ramped between extremes at a controlled rate (typically 10‑15°C/min). Dwell time must be sufficient for all test items to reach temperature (typically 15‑30 minutes per extreme). Used for IPC‑6012 qualification (Class 2/3: 500 cycles).
- 2.6.7.1 (Thermal Shock, Liquid‑to‑Liquid) – Samples are transferred between two liquid baths (e.g., fluorocarbon) at different temperatures. Transfer time ≤10 seconds. Used for military (MIL‑STD‑883) and space applications where extreme temperature gradients exist.
Note: Some specs refer to “air‑to‑air thermal shock” (e.g., using a dual‑chamber system with rapid transfer). This is closer to thermal shock than thermal cycling, but IPC distinguishes based on ramp rate (>30°C/min qualifies as shock).
Selection Decision Tree – Which Test Should You Run?
Start
│
├─ What is your dominant failure concern?
│ ├─ Solder joint fatigue, PTH barrel cracks, copper fatigue
│ │ └─ → Thermal cycling (2.6.7). Target 500‑2000 cycles.
│ ├─ Delamination, ceramic capacitor cracks, brittle IMC fractures
│ │ └─ → Thermal shock (2.6.7.1) or air‑to‑air rapid ramp.
│ └─ Both concerns (e.g., automotive under‑hood)
│ └─ → Run both tests or use combined profile (slow cycling then shock).
│
└─ What standard is mandatory?
├─ IPC‑6012 Class 2/3 → Thermal cycling (500 cycles, -40/+85°C or -40/+125°C)
├─ MIL‑STD‑883 / space → Thermal shock (liquid‑to‑liquid, 100 cycles)
└─ Customer internal spec → follow specified method
Acceleration Factors – Relating Test Cycles to Field Life
Only thermal cycling has a validated acceleration model (Coffin‑Manson). Thermal shock acceleration is difficult to quantify, making it unsuitable for life prediction.
- Thermal cycling – AF = (ΔT_test / ΔT_field)^1.9 × (f_field / f_test)^0.33. Example: ΔT_test = 165°C (-40/+125°C), ΔT_field = 40°C, f_test = 2 cycles/h, f_field = 0.042 cycles/h → AF ≈ 5.1. 500 test cycles ≈ 2550 field cycles.
- Thermal shock – No reliable acceleration model. Use as a pass/fail screen, not for life prediction.
Engineering rule: Never use thermal shock results to claim a specific field lifetime (e.g., “100 shock cycles = 10 years”). Use thermal cycling for life prediction.
Supplier Audit Checklist – Thermal Test Capability
- Does the supplier have both thermal cycling chambers (air‑to‑air) and thermal shock (liquid‑to‑liquid or dual chamber)?
- What is their maximum ramp rate for thermal cycling? (We require ≥10°C/min for effective fatigue testing.)
- Do they perform in‑situ resistance monitoring during thermal cycling (daisy‑chain)?
- Can they provide temperature uniformity data for their chambers (±3°C?)
- Do they have documented correlation between test cycles and field life for your application?
Related Engineering Resources
Frequently Asked Questions
A> No. Thermal shock does not accelerate the same failure mechanisms as thermal cycling. You may pass shock but fail in the field due to fatigue. Run the test that matches your field stress profile.
A> IPC‑6012 requires thermal cycling (air‑to‑air) per 2.6.7. For Class 3, 500 cycles -40/+85°C or -40/+125°C (depending on application). Thermal shock is not required by IPC‑6012.
A> MIL‑STD‑883, Method 1011: 100 cycles, liquid‑to‑liquid, -65°C to +150°C, transfer time ≤10 seconds, dwell 5 minutes each extreme.
A> Use the Coffin‑Manson model. For automotive under‑hood (ΔT_field ≈ 40°C, 1 cycle/day), 500‑1000 test cycles (-40/+125°C) typically represent 10 years. For less severe environments, fewer cycles may be sufficient.
A> No. Thermal shock is destructive and may pre‑damage samples, invalidating thermal cycling results. Use separate coupons for each test.
Reference standards: IPC‑TM‑650 2.6.7 (Thermal cycling), 2.6.7.1 (Thermal shock), IPC‑6012, MIL‑STD‑883 Method 1011, Coffin‑Manson acceleration model, JESD22‑A104.
Unsure whether to specify thermal shock or thermal cycling for your product?
We help define the right test based on your field environment, failure mechanisms, and industry standards.
Request a Thermal Test Consultation →Get Custom Quote
PRODUCTS CENTER
World-class PCB & PCBA manufacturer focused on high-frequency and high-reliability boards for defense, aerospace, automotive and industrial electronics.
You can unsubscribe at any time with a single click.
Copyright © 2025 UltroNiu Electronics Group All Rights reserved.