Today's Puzzle

[M20S Driver]

Here is today's puzzle:  CHT #6 fluctuates in cruise intermittently.  

This is the second time and both times was on a hot day.  There is hardly any change in EGT.  There is one EGT spike for EGT#6 at 1:38:42.  No sign of a major exhaust leak.  Baffles look fine (one year old), 

The flight was at 2430 RPM, WOT, 12500 feet, 50-70 ROP, 14.6 GPH fuel flow and take off was from KHND at 100 deg f, 130+ kts cruise climb at the lower part of the blue bar (EGT).

What could cause this?  

M20S_Driver

[Andy95W]

My first inclination is a bad probe or bad connection.  I suggest swapping it with the one next to it.

[M20S Driver]

3 minutes ago, N1395W said:

My first inclination is a bad probe or bad connection.  I suggest swapping it with the one next to it.

Good idea.  I was going down the probe path but when I saw the 30 deg spike on EGT #6, I thought there may be more to this issue.

[carusoam]

In real life, The thermal mass of the CHT doesn't allow for such fluctuations.  Chasing down the probe and the wiring seems to make the most sense.

Best regards,

-a-

update: consider searching on the words like saw tooth CHT pattern, and variations like that.  I think I may have seen something similar in the past...

[Vance Harral]

Swapping probes is always reasonable, but that sawtooth pattern looks too small in amplitude and and too regular in frequency to immediately blame on a bad probe or connection.  Politely disagree with carusoam that the thermal mass of the cylinder doesn't allow for such fluctuations in real life.  If I'm reading your graph correctly, the variation is only 20-30 degrees, spaced over 2-3 minutes.  That is quite possibly a "real" CHT change.

Recommend you sift through the monthly puzzlers on the Savvy Analysis site, where they show a graph each month and explain the root cause.  I just spent a little time looking there, as I think I've seen this sawtooth pattern before, but I was unable to find an exact match.

Edited September 15, 2016 by Vance Harral

[cnoe]

Here's another guess for you (based on past experience).

Look at ~2:25 and notice that 2, 4, & 6 all seem to move together while 1, 3, & 5 are more flat. I'd see if there's any way that half your ignition harness is in proximity to the CHT leads somewhere/somehow. Ignition can cause indication interference. Perhaps there's a broken cable-tie or clamp?

Or it could be something else entirely.

Sent from my iPad using Tapatalk

[ArtVandelay]

I had a problem with my CHT probe, after tightening the connection (under the fiberglass sleeve) and adjusting the wire so it wasn't practically touching the exhaust pipe it seem to fix the problem. Notice it stop recording and mark the probe as bad.

[M20S Driver]

3 hours ago, Vance Harral said:

Swapping probes is always reasonable, but that sawtooth pattern looks too small in amplitude and and too regular in frequency to immediately blame on a bad probe or connection.  Politely disagree with carusoam that the thermal mass of the cylinder doesn't allow for such fluctuations in real life.  If I'm reading your graph correctly, the variation is only 20-30 degrees, spaced over 2-3 minutes.  That is quite possibly a "real" CHT change.

Recommend you sift through the monthly puzzlers on the Savvy Analysis site, where they show a graph each month and explain the root cause.  I just spent a little time looking there, as I think I've seen this sawtooth pattern before, but I was unable to find an exact match.

The variations are 8 to 15 degrees.  It is very small and difficult to see in flight.  It is very visible in the chart format.  It seems to happen on very hot days!

I have seen this pattern before too but it was the EGT pattern not the CHT.  I will keep looking

[M20S Driver]

2 hours ago, cnoe said:

Here's another guess for you (based on past experience).

Look at ~2:25 and notice that 2, 4, & 6 all seem to move together while 1, 3, & 5 are more flat. I'd see if there's any way that half your ignition harness is in proximity to the CHT leads somewhere/somehow. Ignition can cause indication interference. Perhaps there's a broken cable-tie or clamp?

Or it could be something else entirely.

Sent from my iPad using Tapatalk

Thanks cnoe.   I will check the wires and clamps over the weekend.

Cylinders 1 and 6 behave differently than the rest.  Cylinder 1 seems to be slower in fluctuations and 2,3,4, and 5 look very close on the expanded chart when I move the data to Excel.  

Driver

[carusoam]

Expanding on the reasoning for CHT to be more stable than what it is showing....

1) The cylinder temperature is a result of a balance of heating and cooling.  In level cruise, these are usually in a good balance.

2) The heating is by burning fuel.  If the heating is stable, the EGT will be stable. (Generally speaking).  Compare the EGT for that cylinder.

3) The cooling is by airflow over the cylinders.  There are situations that block or leak cooling air that causes one cylinder to be hotter than the others.  The IO550 is known to have a warmer #5 cylinder because the alternator blocks some airflow.

4) Most 'real' CHT issues trend up and down over a much longer period of time, and do it in synch with EGT or changes in cooling.  Often they will do it in synch with a neighboring cylinder.  Blocked fuel passage in one injector, will go to the neighboring injector.

5) Thank you to Vance for pulling the details of temperature and time from the graph. (And being polite with the challenge)

6) Putting TJ's issue and CNOEs observations together.  Calls for an inspection of the wire locations...

7) The saw tooth pattern still seams familiar...

8) Also keep in mind the possibility of a thermocouple not making good contact in it's well.

9) the #6 cylinder is the front pilot side cylinder.  (Opposite numbering from Lyc)

 

This generates a great technical discussion.  I am no expert on this subject.  But, We have an MSer that is.  Paul at Savvy, I believe(?).

Savvy accepts downloads from JPIs etal... if you want to try that route...  The blue button at the top of the page that says...  Do you want us to analyze...

 

Best regards,

-a-

[Brian Scranton]

Have you removed your cowling recently? Perhaps a bit of the rear baffling near #6 is folded poorly resulting in airflow issues. Include that on your next preflight. 

 

[cnoe]

I've been a Savvy paid subscriber for 7 months now and am very pleased with what I get for my money. It's also nice to have Kortopates (Paul) on board here too.

Sent from my iPad using Tapatalk

[M20S Driver]

19 minutes ago, Brian Scranton said:

Have you removed your cowling recently? Perhaps a bit of the rear baffling near #6 is folded poorly resulting in airflow issues. Include that on your next preflight. 

 

Good question.  I found a 3 inch portion in the middle of the back baffle bent backward.  Why does it only effect cylinder #6?

[M20S Driver]

Looking closer at the EGT and CHT and comparing cylinders 3 and 6 shows some instability in EGT in E6.  I will add injector cleaning to the list

 

[carusoam]

3 & 6 are definitely going through the same cycles, but the jaggedness of the lines is suspect.

What is happening to the EGTs at the end of the flight?

 

I takes some pre-planning to get perfect flight data...  Start-up, taxi, mag check, T/O, climb, cruise, mixture adjustments, descent, throttle adjustments, approach, and landing, taxi and shutdown....  These steps are usually recognizable in the EGTs and CHTs.

there is wacky data at the end of the flight that is masking the usual steps.

Have you done a GAMI test to demonstrate the flow balance of your injectors?

Best regards,

-a-

[Brian Scranton]

The complexities of air flow are a mystery, but 3 inches will do wonders (that's what I told my wife before we got married)...fix that and go fly--you'll be surprised how much of a difference it makes. 

[M20S Driver]

3 hours ago, carusoam said:

Have you done a GAMI test to demonstrate the flow balance of your injectors?

Best regards,

-a-

GAMI index is less than .3 GPH.  

This behavior is intermittent and happens once in a while in cruise on a hot day. The flight in bound to KHND was normal and the return flight posted above had the CHT fluctuations.

I had a similar pattern flying from Palo Alto to Truckee and back. The pattern was there only on the return flight but not on the flight to Truckee.  We took off at near 8500 feet density altitude on a hot day and had to climb to 9500 to clear the mountains.  

[carusoam]

A quick swap of the CHT sensor with the one next to it will remove the sensor issue from the discussion.  While in there, check the well to see if there is any issues with it.  Looking for a broken spring, or something loose related to the mount.

I cannot connect why a hot day activates this challenge.  Temperature can soften the baffle seals...

PP/engineer thinking out loud...

Best regards,

-a-

 

[M20S Driver]

Found a small exhaust leak on the cylinder side of exhaust port.  This explains 80% of the problem-- temperature rise is due to exhaust leak.

My only explanation for the periodic nature of the temperature change is the 3" baffling that was folded back.-- the other 20%.

I hope to fly it over the weekend and report back

 

[carusoam]

That is interesting!  See if you can post a photo...

Best regards,

-a-

[M20S Driver]

The exhaust leak is a small one and only is detected at CHT of 345 deg f or more. TCM published a SB in 2015 changing the gasket type on their engines to correct the exhaust leak problem.

See attached.

SB14-8 (1).pdf

[M20S Driver]

Belated update:   re-torquing the exhaust nuts fixed the problem and this is a proven fix since it has been fine for the past three years

I did have the thicker gaskets.

[RLCarter]

34 minutes ago, M20S Driver said:

Belated update:   re-torquing the exhaust nuts fixed the problem and this is a proven fix since it has been fine for the past three years

I did have the thicker gaskets.

Nice follow up.....another thread comes to a close...lol

[Niko182]

On 9/14/2016 at 4:15 PM, M20S Driver said:

at the lower part of the blue bar (EGT).

if you're looking at the blue bar of the Moritz Gauges, Mine was always a decent chunk off of what the EDM700 said. I trusted the JPI a lot more.

[M20S Driver]

6 hours ago, Niko182 said:

if you're looking at the blue bar of the Moritz Gauges, Mine was always a decent chunk off of what the EDM700 said. I trusted the JPI a lot more.

They surely don't match on mine either.   EDM is more accurate and the probe location also makes a big difference.  My Moritz EGT probe is installed at the lower part of the mufflers but after the union of the three exhausts so the numbers are always 100 deg higher.