Are you trying to say one pole fails to open?
So i just got a breaker that failed today, and the 'B' phase was fused to its self on the line side of the breaker.
The breaker was mounted in a Nema 7/9 enclosure that has some significant corrosion.
Has any one had this happen?
and the 'B' phase was fused to its self on the line side of the breaker.

The damage in your pic looks like arcing, not overload (no discoloration). Could have been a faulty set of contacts or frequent tripping under a fault condition. Is it a bolted bus connection or stabs? If it was a long term problem, you should see more damage than just the breaker contacts. Check all of the bus connections closely....wouldn't want to ruin another breaker. Hard to say if it was a manufacturing defect or internal damage.
Is this breaker routinely used as a switch to turn a load on and off?Yeah funny thing about that is this is second time in 18 months this breaker was replaced.
In fact this breaker was replaced in mid-2016 1 month after a NETA test on the entire switchrack system. (Which checked out fine)
Last time the contact fell off the breaker line side.
Is this breaker routinely used as a switch to turn a load on and off?
So is it part of a listed motor starter assembly then?
That would be my guess. I think the OP said it has a fault capacity of "just" 17kA. Not really all that high for the duty.An HMCP has no thermal trips, so it only trips on very high magnetic current during a short circuit. If it is a bad enough fault, meaning the current exceeded the rating of the breaker, that sometimes welds contacts. It's not supposed to, but "stuff" happens. Now you need to look for the short circuit down stream of that breaker, it's going to be a bad one.
The assembly is not to code then. It sounds as though you have more than one S+S starter down stream of this HMCP then? If so, that would be a likely explaination of the failure. You can't do that.Unfortunately NO
The Starter /OL assembly is Spreckler and Schuh
with control circuity being various manufacturers in a Curlee enclosure
The assembly is not to code then. It sounds as though you have more than one S+S starter down stream of this HMCP then? If so, that would be a likely explaination of the failure. You can't do that.
Magnetic-Only circuit breakers like the HMCP can ONLY be used in listed and labeled combination starter assemblies where the breaker has been tested with a SPECIFIC overload relay.
...
Can you elaborate on your statement about HMCP and OL testing? My facility is older than most on this forum and starters have been pieced together with various parts through the years so every new starter I look at is a surprise.
(3) Instantaneous Trip Circuit Breaker. An instantaneous
trip circuit breaker shall be used only if adjustable
and if part of a listed combination motor controller having
coordinated motor overload and short-circuit and ground fault
protection in each conductor, and the setting is adjusted
to no more than the value specified in Table 430.52.
Informational Note No.1: Instantaneous trip circuit breakers
are also known as motor-circuit protectors (MCPs).
NEC Article 430.53.C. (emphasis added)
To your first point though, in a reversing starter the contactors should be both mechanically and electrically interlocked to avoid just that sort of issue. If the interlocking is only electrical, a burned aux. contact can lead to destructive events.
SO I dont have ground fault protection on this system. Does that not apply?
"Short circuit and ground fault protection" in this sense is just how the NEC describes a circuit breaker. It just means that the device senses a line to line fault as well as a line to ground fault.
