We have been producing a variety of EGT probes over the last 20 years or so. At a guess, probably well over 200.000 so far.
On some installations the probes last "forever". on others not - even if similar installations.
There are a whole host of possible failure causes.
Yes, quality plays a role - up to a point. In principle you look for the quality of the thermocouple cable - the materials are dictated as far as the cores go. Electrical insulation - some probes I have seen use fiberglass, others a mineral fiber. The mineral fiber withstands higher temps - but is brittle and damages easily. The fiberglass can work - but only if the probes construction inside does not rely on it (probes can use ceramic sleeves inside which also makes them much stronger at high temps).
Typically 316 stainless is used - that's perhaps the best overall material. Wall thickness is again a compromise between strength and speed of reaction. Typically 0.4 to 0.5mm is the norm.
Most failures we have investigated are down to the cable itself - mechanical damage usually due to the cable flexing with engine movement. Just like bending a piece of fence wire repeatedly - the internal cores do not like that and break. Can also affect the insulation - that breaks and the cores short out or to the usual metal sheath of the cable. Most often the break is around the entry point to the probe itself (usually a spring is used to assist this area).
Then there are other issues. Fortunately more rare: Complete probe disintegration. When we get one of those - first thing we do is see if a magnet sticks to the probe. If it does - you have rust in the fuel. Rust particles instantaneously melt in the combustion flame (in particular with turbo changed engines), loosing the oxygen in the process so you have tiny liquid blobs of iron exiting the combustion chamber - usually by the time they get to the EGT probe they are ready to melt onto the surface. Layers of this get added over time (the probe diameter gets bigger). Trouble with this is iron has a very different thermal expansion compared to 316 stainless. The iron actually rips the probe to pieces over time. I have seen this a few times over the years - the probes totally disintegrate or if caught early enough you see them bent (sometimes at crazy angles) - this due to the iron coating not being even on all sides.
The "magnet check" shows quickly what you are dealing with - also on any of the iron you can scrape off the probe.
So that is one to watch out for - it will destroy any probe - no matter how much you paid for it.
Probes typically can go up to 1200 degrees C for short times - at this temperature they glow white. Very high flow velocities can conceivably bend the probes as they will be weak. I have not seen this happening with our probes though (but it's not impossible).
At normal combustion temperatures (non turbo, low revving engines) the probes are pretty much bullet proof - they barely even glow a dark red.
High revving engines (like Rotax's 912 range) can have much higher temps (the probes glow a nice satisfying red at up to 850 degrees C redline). Above that it gets orange and eventually white hot.
If you have a probe fail - investigate why it failed - first look at the cable very carefully (in particular where it enters the probe). If that all looks good - take a dremel with a cutting disk and cut the probe open lengthwise - the thermocouple itself should be close to the tip. Anything interesting to see there ? Of course if the probe is mechanically damaged then look elsewhere.
Rainier
CEO MGL Avionics