Short To Ground Error Model

Description

The Short To Ground Error Model simulates a short-circuit fault from a component’s power input terminal to ground. This fault injection model can be attached to any powered component to study the effects of electrical failures on the power network. The fault triggers at a random time drawn from a Gaussian distribution, or can be manually activated. When triggered, the model injects a low-resistance path to ground into the power bus circuit.

---

Example Use Cases

• Fault Analysis: Study the impact of short-circuit faults on spacecraft power distribution systems.
• Protection System Validation: Verify that fuses, current limiters, or fault detection algorithms respond correctly to short-circuit events.
• Monte Carlo Reliability Studies: Use randomized trigger times to assess system behavior under stochastic failure scenarios.

---

Module Implementation

Fault Timing

The fault trigger time is sampled from a Gaussian distribution at the start of the simulation:

$$t_{\text{trigger}}=t_0+\max \left(\mathcal{N}(\mu ,{\sigma }^2),0\right)$$

where $t_0$ is the current simulation time, $\mu$ is the mean time to failure, and $\sigma$ is the standard deviation. The trigger time is clamped to non-negative values to prevent faults from occurring in the past.

Circuit Injection

When the fault activates, the model injects a circuit element between the component’s input node and ground. The behavior depends on the configured short resistance $R_{\text{short}}$:

• Resistive short ($R_{\text{short}}>0$): A resistor is added to the SPICE circuit, creating a current path to ground proportional to the node voltage.
• Ideal short ($R_{\text{short}}=0$): A 0 V voltage source is added, forcing the input node to ground potential.

Self-Solve Mode

When operating without a power bus (self-solve mode), the model approximates the fault effect by computing the current diverted through the short:

$$I_{\text{short}}=\frac{V_{\text{in}}}{R_{\text{short}}}$$

The forward current available to the component is reduced accordingly:

$$I_{\text{forward}}=\max \left(0,I_{\text{in}}-I_{\text{short}}\right)$$

For an ideal short ($R_{\text{short}}=0$), all current is diverted to ground and no forward current remains.

---

Assumptions/Limitations

• The fault is a one-shot event; once triggered, it remains active until explicitly reset.
• The short resistance is constant and does not vary with current or temperature.
• In self-solve mode, the approximation assumes a simple resistive divider and does not capture full circuit dynamics.
• The Gaussian trigger time can produce very early or very late faults depending on the chosen mean and standard deviation.
• The model does not simulate arc faults, intermittent contacts, or progressive degradation.