Electronic – NTC Inrush current limiter datasheet

Possibly; this page from Ametherm has a lot of description and guidance on how to select one. However, the cases they describe are mostly power supplies, where the inrush current is into a capacitor, and the NTC heats up and passes full current pretty quickly.

Given that you're building a triac control circuit, I suggest looking at "triac dimmer" circuits and using one of those instead. You can then control current either by "dead reckoning" (gradually increase power over a time period that gives the desired result) or by measuring current flow (eg Hall sensor or current transformer).

You also need to consider what happens if the heater fails short-circuit. The triac needs to survive until the fuse blows or circuit breaker trips.

It may be simplest to just overbuild the triac to handle the inrush current.

The inrush current limiter that you're using is under-specified and experiences too large of a temperature swing in your circuit. It fails due to excessive thermal cycling.

As expensive as it may sound, I've yet to find any inrush current solution more resilient than discrete resistors, and a contactor that bypasses them once the 24V control supply has started up. No need for any other delays: just a 24V switching supply, a contactor, and three resistors. The resistors would typically be in metal enclosures and mounted to a metal plate bolted to the DIN rail or the panel. You'd need to size them so that they can safely dissipate the energy they'll produce during the charging transient.