🧑🎓
What exactly is bond wire fatigue, and why does it happen in electronics?
🎓
Basically, it's the cracking and eventual failure of the tiny wires connecting a silicon chip to its package. It happens because the chip and the package expand and contract at different rates when the temperature changes—this is the CTE mismatch you see in the simulator. Each temperature cycle bends the wire a little, causing fatigue.
🧑🎓
Wait, really? So the "Temp. Swing" slider directly controls how much bending happens? How do we predict when it will fail?
🎓
Exactly. A larger ΔT means more bending strain. To predict failure, engineers use the Coffin-Manson law, an empirical model for low-cycle fatigue. In this simulator, it calculates the number of cycles to failure, $N_f$, based on the plastic strain amplitude in the wire from that bending. Try increasing the "CTE Mismatch" and watch the predicted life drop dramatically.
🧑🎓
That makes sense. So what's the practical effect of choosing a different "Wire Material" like Au vs. Al? Is it just about cost?
🎓
Not just cost! Gold (Au) is more ductile—it can handle more plastic deformation per cycle without cracking, which gives it a much longer fatigue life. Aluminum (Al) is stiffer and cheaper but fails sooner. Copper (Cu) is in between. Change the material in the simulator and see how the life changes, even with all other parameters like "Wire Diameter" and "Span" held constant.
The core model estimates the plastic shear strain amplitude in the bond wire caused by thermal expansion mismatch. The wire is treated as a beam fixed at both ends, forced to deflect as the distance between anchor points changes with temperature.
$$ \Delta \gamma_{pl}= \frac{\Delta \alpha \cdot \Delta T \cdot L}{d}$$
Where:
$\Delta \gamma_{pl}$ = Plastic shear strain amplitude
$\Delta \alpha$ = CTE mismatch between chip and package (ppm/°C)
$\Delta T$ = Temperature swing (°C)
$L$ = Wire span (length between bonds) (mm)
$d$ = Wire diameter (mm)
This shows why long, thin wires (high L/d ratio) are most vulnerable—they amplify the strain.
The strain amplitude is then plugged into the Coffin-Manson law to predict the number of cycles to failure.
$$ N_f = \frac{C}{(\Delta \gamma_{pl})^n} $$
Where:
$N_f$ = Cycles to failure
$C$ = Material ductility coefficient (e.g., much higher for Au than Al)
$n$ = Fatigue exponent (typically between 1.5 and 2.5 for metals)
This is a power-law relationship. A small increase in strain causes a large decrease in fatigue life, which is why controlling CTE mismatch and temperature swing is so critical.
Common Misconceptions and Points to Note
There are a few key points I want you to be especially mindful of when starting to use this tool. First, remember that "the calculation result is not an absolute lifetime." This tool is strictly for "observing trends" based on a simplified one-dimensional model. In reality, many more factors affect a wire's lifespan, such as its loop shape, interference with adjacent wires, and bonding strength. For example, even if the calculation shows a 100,000-cycle life, it's not uncommon for actual products to last only half that due to manufacturing variations or impurity effects. How you incorporate a safety margin becomes crucial in practical work.
Next, misconfiguring the "Temperature Amplitude ΔT" parameter is a common mistake. Please don't simply input something like "the operating temperature range is -40°C to 125°C, so ΔT=165°C." The actual temperature change the wire experiences is the sum of ambient temperature and self-heating. For instance, in a power device, even if the ambient is 85°C, the wire itself might momentarily reach 150°C due to Joule heating during current flow. In that case, ΔT would be 65°C (150-85). Identifying this "effective temperature amplitude" is the first step toward an accurate prediction.
Finally, be wary of blindly trusting material constants. The constants for gold, aluminum, and copper in the tool are representative values, but actual wire characteristics can change significantly with trace additive elements. For example, aluminum-silicon wire with 1% silicon added has higher strength and different fatigue properties compared to high-purity aluminum. After making comparisons with the tool, make it a habit to always consult the "datasheet for the specific material you are using" or "in-house measured data."
Related Engineering Fields
The concept behind this fatigue life calculation isn't just for bonding wires; it can be applied to reliability assessments for various "dissimilar material joints." The fundamental phenomenon is that stress arises where materials with different coefficients of thermal expansion (CTE) are joined and subjected to temperature changes.
For example, fatigue life evaluation of solder joints is a prime application. BGA (Ball Grid Array) solder balls also experience shear strain due to CTE mismatch between the substrate and the chip, leading to fatigue failure under thermal cycling. Variants of the Coffin-Manson law, like the "Engelmaier model," are used here as well. Furthermore, thermal components in automotive exhaust systems or aircraft engines face significant thermal fatigue issues at welds or bolted joints of dissimilar metals. Also, in MEMS (Micro-Electro-Mechanical Systems), thermal deformation and fatigue of microstructures are important research topics. Understanding the bonding wire calculation will help you grasp the fundamentals when reading papers or reports in these fields.
For Further Learning
If you want to delve deeper, start by understanding the difference between "low-cycle fatigue" and "high-cycle fatigue." This tool deals with low-cycle fatigue, where plastic deformation is dominant. In contrast, high-cycle fatigue, where elastic stresses are applied millions of times (e.g., from vibration), is evaluated using S-N curves (Wöhler curves). Next, I encourage you to think about the "meaning" rather than the "derivation" of the Coffin-Manson law that forms the tool's basis. The exponent $n$ in the formula $$ N_f = \frac{C}{(\Delta \varepsilon_p)^n}$$ represents the material's "ductility." Materials with a larger $n$ tend to show a more rapid decrease in life with increasing strain, indicating more brittle behavior.
As a next step directly relevant to practical work, I recommend learning about detailed simulation using the Finite Element Method (FEA). While this tool is a simplified 1D calculation, FEA allows you to compute detailed strain distributions, including the effects of the wire's 3D loop shape and substrate bending. In that process, the intuition you gain from this simple tool (like "how does changing the span affect life?") becomes a valuable "measuring stick" for quickly interpreting FEA results. A two-pronged approach—grasping the overall picture with the simple tool first, then using FEA to check local details as needed—is an efficient shortcut for design and verification.