No, not that cat!
Catalytic converters - or "cats" - have a nasty habit of failing with age.
What are the symptoms associated with a cat that's gone bad?
Are there any tests that could definitively confirm that a cat has gone bad?
Cats can stop functioning in a couple of ways:
physical deformation due to high temperatures
According to this article, the catalyst melts at temperatures above 2100 °F (1200 °C).
In the event of a meltdown, the catalyst gets permanently damaged, at which point it doesn't scrub the nastiness out of the exhaust gases and serves as a major exhaust restriction.
inhibition of the chemical reaction, usually due to fouling of the catalyst inside
Unburnt fuel can accumulate in the cats over time, which inhibits the chemical reaction and increases the restrictiveness of the exhaust. The clogging leads to bigger problems related to engine inoperability.
A danger associated with built-up hydrocarbons is that if a spark finds its way into the cat, it will turn into a secondary combustion chamber, destroying the catalyst.
There are several ways to assess the health of a cat:
The catalytic reaction is exothermic in nature, so the exit side the cat should be hotter than the inlet side.
Paulster2's answer outlines a way to test for the presence of the chemical reaction by measuring temperature across the cat.
If exhaust side isn't hotter by 200 °F (93 °C), the reaction isn't taking place and the cat isn't doing its job.
Consult the lambdas
The big caveat with this approach is that it assumes the lambda sensors in question are healthy.
In narrowband setups, the voltage output of the pre-cat lambda sensor should flit between 0.1 V (lean) and 0.9 V (rich). Since the chemical reaction inside the cat consumes oxygen, the post-cat lambda should show a stable signal which corresponds to a rich condition.
The image below summarizes the expected results.
Check for restrictions
Clogging usually occurs due to the accumulation of unburnt hydrocarbons, which is often a sign of bigger problems with engine operability.
Clogged catalytic converters result in a very high engine backpressure, which results in noticeable loss of power and the engine's refusal to run at higher RPM's.
A very conclusive test shown by this video involves measuring pressure drop across the catalytic converter. It helps to have pre- and post-cat lambda sensors for this test:
unscrew the lambda sensors that sandwich the suspect cat
attach a fitting into the bung that allows for a pressure gauge or vacuum gauge to be plumbed in
In the video, Matt uses some pipe with a 22 mm thread cut into one end, but even a cork could do.
measure pressure drop. There should be minimal pressure drop across the cat. Any value greater than 2 psi would be a strong indication of clogging in the catalytic converter.
Since the post-cat exhaust pressure is almost atmospheric, it is possible to do this with just the pre-cat O2 sensor bung, with one caveat that restrictions may be caused by other downstream elements such as mufflers.
Use an OBD-II scan tool
The OBD-II protocol assigns P0420 and P0430 for Bank 1 and Bank 2 cat efficiencies respectively.
The exact logic by which the engine computer sets these codes will vary between manufacturers and models, but they will invariably involve assessing the signal coming from the pre-cat and post-cat lambda sensors, as outlined in 2.
Once again, this assumes fully-functional lambda probes, so a faulty lambda sensor may trigger a false-positive P0420 code.
Even before a cat plugs, you can test your cat using a laser thermometer. You are looking to compare the inlet and outlet temps on your cat. Do the following:
If the inlet temp is around the same temp or higher than the outlet, the cat is bad. A good running cat could have an outlet temp at least a hundred degrees F hotter than the inlet (should possibly be closer to 200). If you aren't seeing the temperature difference, it's most likely bad.
NOTE: Before you go believing the cat is bad due to an OBD-II check light (P0420), do this test to see if it's actually the cat or if it's the O2 sensor. The computer is fooled quite often by a bad sensor.
If an exhaust restriction is suspected a pressure gauge pre-catatlyst works just fine. My maximum cutoff pressure is 1.25 psi but most good systems will be well below that. The intake manifold vacuum is also a good clue. If it is not close to the idle reading at 3000 rpm suspect an exhaust restriction. In the case of a restriction the vacuum reading will be lower than normal. Also, the sound coming from the intake changes to a louder, deeper sound upon revving the engine.