Cracks in ceramic capacitors, devices that store electric charge in electronic circuits, can cause damage to such disparate objects as medical implants and spacecraft. The cracks, which are often hidden initially, can start conducting electricity, depleting batteries, or shorting out electronics. Detecting these cracks in capacitors before they evolve into electrically conducting pathways and cause failures is something that researchers at the National Institute of Standards and Technology (NIST), along with collaborators from the University of Maryland, NASA Goddard Space Flight Center, and Colorado State University have spent years investigating.
Their research has demonstrated a new nondestructive approach to detecting these ceramic capacitor cracks proactively. In their study, the prototype method led to the rejection of more than 90 percent of sample capacitors with visible cracks.
Once further studies quantify and confirm detection levels, they say that the new technique may help prevent failures in medical devices, such as cardiac pacemakers and defibrillators and could avert electronics failures in satellites and other spacecraft. The method may also detect structural flaws in other types of materials, researchers say.
Because they are able to store a lot of energy in relation to their size, multilayer ceramic capacitors are widely used and have an annual market in the billions of dollars. But their failure rates, while low, have long been considered a problem in some applications. One NASA study notes that capacitors are the electronic component most likely to fail. Capacitors can crack during manufacturing, assembly, or use because ceramics are brittle and the devices are exposed to heat and mechanical stress. Industrial screening—such as automated visual inspection, X-rays, and acoustic microscopy—may not find subsurface cracks, especially near corners under capacitor endcaps, where stress can be highest. An FDA study of data for several million pacemakers and defibrillators implanted from 1990 to 2002 found that about 1 in 150 failed, about 25 percent of these failures were battery/capacitor abnormalities. The study also found that device malfunctions caused the death of 61 people.
How It Works
The new NIST crack detection method relies on acoustic measurements at frequencies much higher than humans can hear. Researchers briefly apply an electric field across the electrodes of a capacitor, exciting a vibration at a specific frequency. They then measure the decay over time of the signal. The data are analyzed to determine slight shifts in frequency versus the magnitude of the vibration. These frequency shifts are greater when cracks are present. The researchers say that this nonlinear approach—focusing on frequency shifts relative to signal strength rather than the frequency shifts alone—is especially useful because it is not affected by slight variations in size of the capacitors.
The ceramics in the NIST study are highly nonlinear, meaning the capacitors get less stiff and their resonant frequency drops when they vibrate more strongly. The new NIST method measures patterns in how this tone changes over time in relation to the strength of the vibrations.
Researchers measured 41 multilayer barium-titanate ceramic capacitors, each roughly 2 by 3 millimeters in size, before and after heating to high temperatures (189°C) and quenching in ice water. This thermal treatment generated surface-breaking cracks in 27 samples. The nonlinear acoustic results were strongly correlated with the presence of visible cracks: Measurements on 25 of the 27 visibly cracked capacitors yielded results that were outside the range of those for capacitors without cracks. (See Figure 1)
The study concluded that nonlinear acoustic measurements offer a promising approach for nondestructive detection of cracks in capacitors before electrical failure occurs, and that further work should be pursued to quantify the level of detection. NIST staff are continuing this research in collaboration with a capacitor manufacturer.
For more information, visit www.nist.gov/mml .