Bravo Exhaust Valve Failure - Root Cause

[alextstone]

In the midst of the COVID shutdown, Denise and I decided to fly down to Dauphin Island, AL for a short getaway last Thursday.  The flight down was uneventful.  We enjoyed the sun and each other's company along with one of our teenage sons.  Then, 19:30 minutes into the return flight at 4500 feet, I performed a leaning procedure to confirm I had the correct ROP setting.  As I moved past LOP through peak and toward the final ROP setting, I noticed that the engine was running just a bit rough.  I asked Denise if she felt it too and she said "yes, what's that"?  I then looked back at the engine monitor and saw the EGT for the #5 cylinder was high and climbing and the CHT was low and decreasing.  

I have to admit, I found it harder to commit to a diversion than it should have been.  Those thoughts of "how bad is it?, Can we make it to home base safely?, How on earth will I get the maintenance done at a different field?" all conspired to create a few moments of inertia.  However, I shook that off and I diverted to Trent Lott airport in Pascagoula MS.  The cylinder is not off and you can clearly see just how close we came to total valve failure.  This cylinder was installed in 2012 by the previous owner and it had about 525 hours in service, 200 of which were since I purchased the airplane.  In the photos below, you will see a borescope image of the valve at about 350 hours time in service.  I also show engine data during the failure sequence

Questions:

1.  What was the root cause?

2.  How often are borescope inspections necessary to identify these problems?

3.  What other early warning methods can one use?  BTW, the temps were normal for this cylinder up to point of failure and data was uploaded regularly to Savvy and no flags were made using their automated Exhaust valve failure analysis.

4.  What general "pearls of wisdom" come to mind with regard to this post?

Piece of the valve found lodged in the bottom plug

 

You can see where that piece of metal came from.  Oops.  

Rotator cap looks good

borescope of same valve at 324 hours (200 hours ago)

Engine data from entire flight

engine data for the two minute period around the time of failure

 

[Phil EF]

Bore scope is called for by TCM at every annual/100 hour inspection. Easy to do and best form of early warning. 

The final stage prior to coming apart is usually preceded with green coloration starting to show on the exhaust valve. 

[LANCECASPER]

Just now, Phil EF said:

Bore scope is called for by TCM at every annual/100 hour inspection. Easy to do and best form of early warning. 

The final stage prior to coming apart is usually preceded with green coloration starting to show on the exhaust valve. 

This isn't a TCM engine, it's a Lycoming TIO-540-AF1B.

[alextstone]

Just now, Phil EF said:

Bore scope is called for by TCM at every annual/100 hour inspection. Easy to do and best form of early warning. 

The final stage prior to coming apart is usually preceded with green coloration starting to show on the exhaust valve. 

Yep.  That last one WAS at the last annual.  I fly a lot.  I'm thinking now at every 4th oil change for me...every 100 hrs

 

 

[FoxMike]

I wondered how many hours were on those cylinders.  I am sitting here waiting for a FedEx delivery of new cylinders that I ordered the first of August.  I have about 2150 on mine.

[alextstone]

41 minutes ago, FoxMike said:

I wondered how many hours were on those cylinders.  I am sitting here waiting for a FedEx delivery of new cylinders that I ordered the first of August.  I have about 2150 on mine.

That's impressive!  This one will be the last of the cylinders that pre-dated my ownership...so from this point on, no excuses if they do not last...

[carusoam]

Great data and photo collection, Alex!

1) nice to see the pic of the valve to explain the vibration and funky EGT/CHT data...

2) Even Better to know that the missing part, got found...

3) Nothing beats pics from inside the cylinder... but, that takes scheduling and down time...

4) Often sticking valves have a similar ‘hot spot’ as depicted in your valve pic.

5) Sticking valves are the ones that stop rotating in place....

6) To find sticking valves as they happen... the EGT data often shows a remarkably odd detail to it...

7) The odd detail... the EGT graph has a saw tooth pattern to it for the offending cylinder... a jagged line with regular oscillations... as if they are normal...

8) If it started happening several flights ago... the data might be traced back to the point it started...

9) It would be nice if you could share the data file... for closer examination... copy the link from Savvy.

10) don’t forget to click the share button on the savvy site...

11) If Savvy didn’t see it, the chances of it being there are pretty remote.

12) One oddity is the extent of the failure... valves don’t shed parts that often from all the pics posted around here...

13) Pizza pics are normal valves, heat affected/sticking valves have the half moon coloration....

14) Valve rotation is a method for cooling the valve evenly, in this less than ideal heat transfer location...

15) Looks like something may have somehow stopped the valve from rotating... something to look for when the valve comes out...

16) Now that you know Valve#5 is struggling...

• take a look at its behavior just after start-up... just slightly irregular compared to the other five...
• take a look at the run-up... oddly, EGT5 is staying near the bottom and not really peaking as much as the others... different, but not standing out...
• Then the inflight mag check... She didn’t like going back ROP!  That is probably where the valve chipped the part off... and hot unburned fuel started escaping toward the exhaust...
• The TIT probably showed evidence of a slightly higher number as well...

17) Good call to end the flight... Too many unknowns to proceed. 

 

Thanks for sharing the Great details, please share the savvy data if able...

PP thoughts only, not a mechanic...

Best regards,

-a-

[LANCECASPER]

On 3/29/2020 at 11:54 AM, alextstone said:

That's impressive!  This one will be the last of the cylinders that pre-dated my ownership...so from this point on, no excuses if they do not last...

Sounds like the previous owner(s) flew it too hot too long. High CHT because of bad baffles, too high of a power setting, too high of a TIT comes back to bite later. If you haven't done anything with the exhaust or there weren't any recent logbook entries before you bought it, scrutinize it carefully. If they burned up the cylinders, they burned up the exhaust. A bad cylinder will cause you to do a precautionary landing and good job on that! You might not be so fortunate with a bad exhaust (turbo transition especially).

[alextstone]

33 minutes ago, carusoam said:

Thanks for sharing the Great details, please share the savvy data if able...

PP thoughts only, not a mechanic...

Best regards,

-a-

Thanks for the list of thoughts...here's the link to the data (actually to all flights)  In previous flights, the #5 cylinder CHT probe was the ring probe so it was reading 75-100 degrees cooler....

https://apps.savvyaviation.com/my-flights/19229/ac9a0959-c416-44a9-8897-42264fc5c3b9

[alextstone]

24 minutes ago, LANCECASPER said:

Sounds like the previous owner(s) flew it too hot too long. High CHT because of bad baffles, too high of a power setting, too high of a TIT comes back to bite later. If you haven't done anything with the exhaust or there weren't any recent logbook entries before you bought it, scrutinize it carefully. If they burned up the cylinders, they burned up the exhaust. A cylinder will cause to do a precautionary landing and good job on that. You might be so fortunate with a bad exhaust (turbo transition especially).

@LANCECASPER, I have overhauled the exhaust system and turbo 200 hours ago and the V-Band clamps are in good condition.  I've also changed (now) all six cylinders, overhauled the mechanical fuel pump and prop governor and replaced the starter.  Flexible hoses are in good shape.  GAMI's are tuned to .3GPH spread.  I hope this is the end of the consequences of deferred maintenance and infrequent flying.

[N201MKTurbo]

Why do valves burn?

There are a lot of theories and opinions about that.

The easy answer is that they stopped transferring heat from the valve to the valve seat. The hard question is why?

If the seat deteriorates from corrosion or some other high temperature chemistry, it can form a low spot where there is no sealing and also no heat transfer, but the valve rotates so the valve should survive. Unless the valve stops rotating, then it will burn at that spot.

If the valve deteriorates at some spot along its sealing surface from corrosion or some other high temperature chemistry, it will not transfer heat at that spot causing that spot to overheat which accelerates the deterioration. Even if the valve rotates, the deteriorated area will sill fail to transfer heat.

If the valve stops rotating, A bad spot on the valve can cause a bad spot on the seat and visa versa and they will both damage the other.

So, what can cause this to happen in the first place?

Some say it is worn valve guides that can cause uneven seating pressure.

I can imagine some debris sticking to the valve or seat keeping it from touching at some place.

A broken valve spring can reduce sealing pressure.

A defective lifter can cause valve lash to go away, preventing the valve from closing tightly.

The bottom line is that you will probably never know why it happened.

It is an airplane, $h!* happens!

BTW, you probably could have made it home. These things usually progress fairly slowly. That cylinder wasn't making much power, but probably wouldn't have hurt anything else. But then again, nobody ever died for diverting early. (No stories about people dying while diverting).

[alextstone]

18 minutes ago, N201MKTurbo said:

BTW, you probably could have made it home. These things usually progress fairly slowly. That cylinder wasn't making much power, but probably wouldn't have hurt anything else. But then again, nobody ever died for diverting early. (No stories about people dying while diverting).

Probably.  I thought of it.  I had my son and wife in the plane with me.  I might have tested that "probably" if I had been alone...

[N201MKTurbo]

FWIW,

When I borescope a cylinder, I like to do it with the valve open so I can see the sealing surface.

[N201MKTurbo]

So, I started surfing the web about valve failures and came up with some valuable advice that I hadn't considered before.

When borescopeing a cylinder, move the propeller and watch the valve as it seats. If it translates sideways at all when it seats, you have a warn valve guide.

I will start looking for that...

[carusoam]

I have only had one bad valve... I wouldn’t like another... 

Reviewing more of the flight’s data...

EGT5 is a pretty smooth line...

Compared to EGT2 has a more noisy/ more sawtooth looking graph...

Probably no chance that these two EGT sensors got swapped? 
 

More review of more flights would be desired to make any PP judgement to look in cylinder #2...

I was hoping to separate the TIT from the EGT data... but, the JPI puts the data on the same layer...

 

I like Rich’s take on the situation...

A valve that stops rotating isn’t usually catastrophic...

But, a valve that gets stuck while extending into the cylinder can cause a crash between the valve and piston... (My O360 experience...)

The heavy vibration, and lack of power that it creates can be mildly alarming... (more for a four cylinder engine)

I didn’t have an engine monitor at the time... 

PP thoughts only, not a mechanic...

Best regards,

-a-

[kmyfm20s]

If you change your own oil it’s really easy to pull the top plug and look at the valves while your letting it drain. I think Mike Bush said you would be able to identify that bad valve 75 hours before that happens, can’t remember exact numbers. Basically if you check the bore scope every oil change it shouldn’t catch you by surprise. Also if you look at previous EGT readings on that cylinder you should see a rhythmic little peak of the temp which was an indication it was burning.

You have motivated me to go change my oil today, thanks!

[carusoam]

To see an oddity That May be a sign of morning sickness...

Check the flights from 3/15 onwards... EGT5 demonstrates not following the curve of the other five cylinders... right after start-up... E5 is below all the others in sort of notable way... 

3/7 and Prior E5 was in the middle of the pack behaving as well as the others...

Without graphs being present in the cockpit it would be hard to notice... for anyone with less than perfect memory skills... 

And... no saw tooth.

Did you hear any valve sounds during the start up that were memorable? Tick, tick, tick, tick, TACK, tick.... repeating.....?

PP thoughts only, not a mechanic...

Best regards,

-a-

[N201MKTurbo]

While the Savvy folks offer a valuable service, it seems like this kind of analysis could be done in the engine analyzer itself. Modern embedded processors usually have a lot of extra processing power in terms of CPU time and memory, so all it would take is software and some way of displaying the information. Any system that allows you to download data could create a text file with the results of its analysis that you could download. 

[kmyfm20s]

11 minutes ago, N201MKTurbo said:

While the Savvy folks offer a valuable service, it seems like this kind of analysis could be done in the engine analyzer itself. Modern embedded processors usually have a lot of extra processing power in terms of CPU time and memory, so all it would take is software and some way of displaying the information. Any system that allows you to download data could create a text file with the results of its analysis that you could download. 

I always thought it would neat if the engine monitor displayed the last 15 minutes or so on the monitor. Similar to an EKG tape. I think it would be very powerful tool while in the air. How often do you hear a sound, feel a vibration or see a change on the monitor and wished you didn’t have to wait to download. 

[carusoam]

Speaking of things that stand out...

The TIT data in Alex’s data files... has a tendency to drop off line for a few seconds at a time...

Once per flight or so... not more than once...

Alex, let me know if you don’t see it... I can post a graph that shows it...

Best regards,

-a-

[alextstone]

1 hour ago, carusoam said:

Reviewing more of the flight’s data...

EGT5 is a pretty smooth line...

Compared to EGT2 has a more noisy/ more sawtooth looking graph...

Probably no chance that these two EGT sensors got swapped? 
 

Well, if it was swapped, then I just pulled the wrong cylinder and oddly, it had a valve failure too :-).  

[alextstone]

21 minutes ago, carusoam said:

Speaking of things that stand out...

The TIT data in Alex’s data files... has a tendency to drop off line for a few seconds at a time...

Once per flight or so... not more than once...

Alex, let me know if you don’t see it... I can post a graph that shows it...

Best regards,

-a-

I've seen it in the cockpit too.  Loose connection somewhere.  

[Phil EF]

10 hours ago, LANCECASPER said:

This isn't a TCM engine, it's a Lycoming TIO-540-AF1B.

Valves don’t know whose name is on the data plate. 

[kmyfm20s]

You stimulated me to go take some picture and change my oil:) I also posted a link to the article I was referring to.

 

 

[irishpilot]

I almost had the same thing happen. My #6 cylinder only had a few hundred hours, but was running CHT 35-50 degrees hotter than the others. It wasn't allowing me to run high altitude which was my indicator to get it scoped. Found the value discolored. It would not have made the next annual. Like yours, it showed no signs of distress during the annual.

I am a big fan the JPI units as I can see exactly what's going on in my engine. Worth the time and money investment.

Sent from my Pixel 3a using Tapatalk