Forget calibration.  It is a waste of time.  All you will EVER get in the cockpit is relative temperature indications: when the exhaust gases are (probably) hotter or (probably) cooler.  That is ALL the EGT/TIT number tells you.  The absolute value of the numbers means not much.  The numbers are for comparison purposes only.

 

Measuring gas temperature in the exhaust pipe of a piston engine is very tough stuff if you want to really know the exact answer.  Reason?  There are many confusing effects, and the temperature and flow are constantly changing.  See attached picture for a bit more understanding if you want to wade into some of the details. 

 

The temperature displayed by the instrument is the temperature read by the thermocouple which is buried inside a probe which is sheathed with a stainless skin and filled with some kind of ceramic insulator.  These combine to yield a very slow response time, much slower than the cycle time of the engine (40 times per second at 2400 RPM).  So the thermocouple is time-averaging the temperature in the vicinity where the thermocouple is located.

 

I found that the EGT SYSTEM (thermocouple plus sheath plus surrounding thermal mass of exhaust pipe and cylinder head) take sabout three minutes to come to full equilibrium. 

 

Don't believe me? 

 

Do GAMI lean test.  Start rich, Lean down in steps waiting, say, one minute per step while you record data.  Then reverse, going from lean back richer.  Then plot the data.  The curve going leaner will not lie on top of the curve going richer.  Why? Because when you took the reading, it was still changing, slowly, but still changing.  It takes a lot of time and patience to get good data. 

 

But drawing the curve has value because it shows which cylinder goes lean first, second, and so forth.  But the absolute numbers are subject to huge errors as noted below. 

 

The instrument temperature reading is related to the gas temperature in the same way that a stopped clock is accurate twice a day.  That is, the EGT reading is only accurate when the GAS temperature is rising through the reading, or falling through the reading, which may occur once or more each cycle. 

 

Look at the figure.  The exhaust event lasts about 180 degrees out of a total 720 degrees required to complete the four stroke cycle.  So you get WHOOSH for about a quarter of the time, and then nothing except some ringing of the exhaust gas column back and forth as the pressure pulse oscillates up and down the exhaust pipe.  There is nothing like "steady state" (smooth and unchanging) occurring in the exhaust flow.

 

Similarly, there is nothing like steady state about exhaust gas temperature.  When the exhaust valve opens, the temperature of the escaping gases is highest, but falls as gas expansion occurs in the cylinder as the gas exits the cylinder. So the temperature starts out high, and ends up lower at the end of the exhaust pulse.

 

So both flow and temperature are constantly changing, but flow is nearly zilch about 3/4 of the time. 

 

Now the poor thermocouple is shielded from all this variability by the stainless steel sheath.  The sheath is heated by convection (passage of gases), and heated by thermal radiation from anything hotter than the sheath (portions of exhaust pipe, or exhaust valve perhaps?).  And the sheath loses heat to the surroundings via radiation going to any surface that is cooler than the sheath (some portion of exhaust manifold pipe probably, certainly the aluminum exhaust port in the head). 

 

And the sheath is also losing heat to the outside via conduction to the cooler exterior.

 

So the sheath is arriving at some time-averaged temperature that is probably close to the thermocouple temperature, but which is influenced by a heat balance of gain and loss between gas, conduction, and thermal radiation with the surroundings.  And the gas is active only 1/4 of the time, and the flow rate and temperature of the gas are changing during this short interval.

 

Getting the picture?  The thermocouple is related to the gas temperature, but is delivering a signal to the instrument that is averaging out all these variations in time and thermal balance heat flows. 

 

In short, we do not really know what the gas temperature is.  We only have a number that kinda sorta represents some sort of average of gas temperature and flow, installation geometry, and ambient environment under the cowl (cooling air blowing on exhaust pipes and thermocouples, radiation environment inside and outside of exhaust pipe, conduction losses and on and on).

 

The number in the cockpit is only valid for comparison with other numbers obtained during other power and mixture settings and flight conditions.

 

Some observations can follow from considering the figure and discussion above, such as:

For all these reasons, EGT spread normally means little unless one reading is WAY off the program in which case it does mean something (blocked injector, fouled spark plug, etc).  But the value of EGT/TIT is the TREND and THE COMPARISON to other operating conditions.

 

What about TIT limits, typically 1650F or sometimes 1750F?  The TIT is the best data we have, and if the turbo guys say we can expect trouble sooner or later if these temperatures are exceeded, it is based on experience and some safety factor using the best data available to the pilot.  Respect these limits.  Inaccurate data is better than no data in such circumstances.

 

But don't worry about absolute values or calibration.  They are meaningless values and exercises.  Watch the trend and comparison with other operating conditions.  Those tell the important tales we need to know and understand.

 

Fred Moreno