PPRuNe Forums - View Single Post - Jabiru engine failures
View Single Post
Old 5th Dec 2014, 22:29
  #263 (permalink)  
Jabawocky
 
Join Date: Jul 2007
Location: in the classroom of life
Age: 55
Posts: 6,864
Likes: 0
Received 1 Like on 1 Post
flightstick,

welcome and a very interesting post indeed.

I think if I may, a few subtle corrections and educational points, as you may have some assumptions based on correlation not causation.

he excess fuel drops the combustion temperatures, EGTs and cools the piston crowns themselves, keeping them within their safe thermal limits.
The extra fuel delays the peak pressure by slowing the speed of the flame front. This means the peak occurs further from TDC, in a larger volume and thus a lower peak pressure. It is the peak pressure that determines the CHT and piston temperature. There is a bunch more explanation, happy to do it via phone Skype or in class, but that might be a long way for you to travel.

these parts have thermal limits and high EGTs can decrease the life of these components. Rich mixtures decrease EGTs and exhaust valve temperatures.
You are on the right track, just correlation is confused with causation. The actual EGT is not the problem. In fact as funny as it sounds, the valve and seat temperatures are cooler at peak EGT than they are at 50dF ROP or as you approach best power (12.5-13.5 roughly).

It is not the temperature of the valve, it is the distortion of the head and valve components and subsequent valve guide wear that causes the valve to not seat nicely, and thus the "rattling into place" that stresses the valve and breaks the head off.

Here are a few graphs that contain much of the knowledge that has been passed down.





Just to finish off, if you can control your fuel injection on each individual cylinder to match the air intake (volumetric) and maintain an AFR that is consistent across all 4 or 6 cylinders, and do this in real time closed loop, you will not only have a wonderfully smooth Jab engine but one that will run Lean Of Peak at the top of the brown BSFC curve above, all day long and with superb reliability.

I would be interested to chat with you about it if you can. You can use a Lamda probe or EGT for feedback. But the Lamda probe would make the processing easier, but cost a lot more.

Cheers

David Brown
Jabawocky is offline