Electrolytic capacitor failure mode

Question

In a project we're working on, we had an electrolytic capacitor explode on us. It was a surface mount device, UCM1H470MCL1GS from Nichicon:

So far, we've seen our project run for many hundreds of hours without incident, so I'm fairly sure there is no fundamental error in the circuit causing these capacitors to be subject to over-voltage (they have a regulated 24v voltage across them).

However, our project does experience a lot of vibration and shock loading.

Question: Is it possible for vibration to cause a capacitor to fail short circuit, or in some other way that causes it to produce clouds of smoke?

The datasheet mentions the importance of careful handling, but doesn't seem to specify the failure modes in the case of mechanical abuse.

Answer 1

our project does experience a lot of vibration and shock loading.

The data sheet for the Panasonic FN series (for example) has this to say (look at #9): -

Note that I'm referring to the Panasonic data sheet because the picture in the question was originally a Panasonic FN capacitor: -

But anyway, the Panasonic list above isn't specific to Panasonic - it's a fairly standard thing for all suppliers. But, Panasonic, to their credit (and I do rate their electrolytic capacitors) are explicit about problems you may encounter.

So, you can assume that the list above pretty much applies to all electrolytic capacitors having a wet electrolyte.

Question: Is it possible for vibration to cause a capacitor to fail short circuit, or in some other way that causes it to produce clouds of smoke?

Yes it is.

Answer 2

As you might know, aluminium electrolytic capacitors are made from conductor foils separated by one or more dielectrics (e.g. paper).

Src: Research Gate

For a non-vibration-proof capacitor, excessive pressure or vibration may lead the separator to get mechanically damaged or move or another situation leading to a possible short between anode and cathode. Vibration-proof capacitors have extra supports inside to prevent any move of the foils and separators.

And a capacitor with shorted anode and cathode foils is happy to explode.

Answer 3

Here's one way to blow a surface mount capacitor: In our case, the housing of the PCB actually hugs the capacitor fairly close, keeping it roughly in position, even if it becomes detached from the board.

Vibrations caused the capacitor legs to fatigue and fail, but the capacitor was kept in place, wobbling and rotating with the vibrations.

After a while, it rotated a full 180º, and for a moment touched the PCB reversed. This caused it to explode.

There are burn marks on the housing which match marks on the capacitor which shows it was reversed polarised when it exploded.

Since it had already run for many hours, I think it's safe to say that it probably hadn't been soldered the wrong way round, but instead had rotated after the legs fatigued.