Back when I was building one, engine fires were all too common in Velocities. Detection becomes a real issue when the engine is behind you and the aileron and rudder controls pass through the compartment with no fire protection. Some of the thoughts I had:

 

  For sensing a fire, perhaps thermal fuses could be used. They're available in a range of temperatures- I have ones rated at 144dC up to 184dC. I imagine higher temps could be found easily. My thought is to run them in series to different points within the engine compartment. The loop would be connected to the winding of a relay, holding the contacts open. If one of the fuses blew the relay would close and set off the alarm.

 

  DC fire alarm components for hard wired systems might be another option- either temperature or rate-of-rise. It also might be possible to hard wire a battery-operated fire alarm as a sensor.

 

Bill,

I considered your approach when building my Lancair IV-P, and elected to go with two temperature sensors instead. Rationale was that I didn’t know how to “calibrate” the thermal fuses without testing that was beyond my capability…and probably interest. Basically I decided to install temperature sensors near each  cowling air exit rather than guess at what value to use for the thermal fuses. I hooked the temperature sensors to my MVP-50 engine monitoring system and recorded the actual temperatures seen by the two probes during initial test flying. After determining that the temps never went above 150F during flight or ground operation, I set the alarm “trigger” temp for 170F. My normal temperatures run between 120-135F in flight, meaning I will get a warning if the cowl exit temperature increased by 35-50 degrees from “normal.”  This is an arbitrary setting that has precluded any false alarm to date…but then again I’ve not yet had a real alarm either. I could probably generate an earlier warning by lowering the alarm temp, and may do so the next time I fiddle with the MVP-50 alarm settings. Of course, the objective is to generate an alarm for any “real” event, but to not generate false alarms. This took quite a bit of fiddling with most of the settings, but I am now very comfortable with what I have…and love the system.

 

Considering the consequences, I think a fire/overheat monitoring system in the engine compartment is a very good idea, no matter what system you use!

 

Bob

 

PS: The actual temperatures will be dependent on the sensor/fuse mounting location…so don’t use my figures; it’s easy to determine your own…