Ovation IO-550: Cylinder 2 Suddenly Lean

[G-SLOT]

I wanted to share an issue I experienced during my last flight to see what others think. The engine has been running very well lately but, part through this flight, Cylinder 2 suddenly seemed to start running lean.

IO-550-G with GAMIjectors and electronic ignition. Normally smooth running LOP. First flight since an oil/filter change and, during the last annual (October 2020), the ignition harnesses, all spark plugs and the Cylinder 2 GAMIjector were replaced.

I'm thinking there might be something blocking the Cylinder 2 fuel injector or the fuel line itself. Magneto/electronic ignition OK and no change when I tried running on alternate air (though, oddly, and perhaps unrelated, the alternate air annunciator light did not illuminate when I pulled the alternate air handle).

Here is a Savvy link to today's flight (note that it's an older JPI700 and does not show the fuel flow - I'm thinking of a 900 or 930 but this is the best data have for now): Savvy 20210220

Referring to the elapsed times within this flight (all approximate):

• Time 00:13: Takeoff. WOT, 2,700RPM and mixture full rich. Circa 30-31 GPH. Reduction to 2500 RPM shortly after departure but WOT.

 

• Time 00:17: Level at 2,300' (under controlled airspace). 23"/2400RPM. Big mixture pull to around 12GPH and tweaked to about 13GPH which normally gives me 50 degrees LOP. All smooth and acting normally.

 

• Time 00:25: Mixture rich, 2500RPM and full throttle for climb to 10,000' (at 500fpm). Climbed ROP and leaned during the climb to maintain a broadly constant EGT.

 

• Time 00:41: Levelled at 10,000'. WOT/2400RPM. I wanted to assess my best power true airspeed (I normally fly LOP) so I was messing around with the mixture ROP and, when I then tried to find peak and LOP settings, I found that Cylinder 2 was peaking about 1.8GPH before the others (it peaked at about 14.5 GPH whereas the others peaked around 12.8 GPH - normal spread is around 0.5 GPH so this is quite unusual). Naturally, the engine was too rough to stay LOP so I stayed ROP for the rest of the flight, although I tried a couple of different times, including at a slightly lower MAP.

 

• Descent was at a reduced MAP and I enrichened the mixture on the way down as and when the engine became too lean and started running a little rough. Normal landing and shutdown.

I'll touch base with a local mechanic on Monday and think the first thing we should check is the fuel line to Cylinder 2 and the Cylinder 2 injector.

Does anyone have any other thoughts or can you see anything else on the chart that looks suspicious?

In case helpful, here is a link to the previous flight, right before the oil change, which was flown at about 2,500', 23 or so inches MAP / 2,400RPM and about 50 degrees LOP (approx. 13 GPH): Savvy Previous Flight

[PT20J]

When something goes amiss after maintenance, I always check the most recent things done. In this case, I would look especially hard at the #2 injector. You didn’t mention why the injector was replaced.

Skip

[G-SLOT]

The maintenance just before the flight (albeit a week before) was an oil and filter change. The injector was replaced at the annual in October (about 18 flying hours ago) because i was chasing smoother LOP operation and it was discovered it was the wrong type of injector (was type F while, according to GAMI, it should have been type E - replacement kindly supplied by GAMI). Running smoothly since (including LOP). The odd thing is that, during yesterday’s flight, it was running nicely LOP while I was at 2,300’ but, sometime between going ROP for the climb and settling into cruise at 10K, that cylinder became leaner, evidenced by it peaking way before the other cylinders. 
 

My best guess at the moment is that some small piece of debris became lodged in the fuel line or the cylinder 2 injector, but I’m very keen to hear other theories. 
 

That injector, and possibly the line supplying it, will definitely be the first thing I have checked. 

Edited February 21, 2021 by G-SLOT

[MarkD34M]

Must be something in the air today. I went out this am in my Ovation, I had the engine stumble when I reduced power at 1000 ft.  I climbed ROP to 7500 and as I went LOP, my number 5 cyl was running  leaner than normal.  I was thinking injector clog, air induction leak or plug? Interested what others have to say.

[PT20J]

8 hours ago, MarkD34M said:

Must be something in the air today. I went out this am in my Ovation, I had the engine stumble when I reduced power at 1000 ft.  I climbed ROP to 7500 and as I went LOP, my number 5 cyl was running  leaner than normal.  I was thinking injector clog, air induction leak or plug? Interested what others have to say.

It has to be a pretty big induction leak to affect cruise - usually induction leaks show up most at idle (I once had a R-985 on a Beaver that flew fine but was really rough at idle. Turns out the carburetor had been replaced and not torqued and was loose.)

One bad plug should increase combustion time and cause EGT to rise. Should be easy to check by switching mags. 

I assume you have GAMIjectors. I wonder if they are more sensitive to restriction or more likely to get restricted.

Always the easiest thing to do is clean and gap the plugs and clean the injectors in some Hoppes No. 9 and dee if that makes it go away. Not a bad idea to clean all the fuel screens also. 

Skip

Edit: Corrected comment that a bad plug should increase combustion time. 

[MarkD34M]

Thanks Skip - appreciate it.

[M016576]

4 hours ago, MarkD34M said:

Must be something in the air today. I went out this am in my Ovation, I had the engine stumble when I reduced power at 1000 ft.  I climbed ROP to 7500 and as I went LOP, my number 5 cyl was running  leaner than normal.  I was thinking injector clog, air induction leak or plug? Interested what others have to say.

Could also be a valve that’s warped or not sealing completely any more- that would show on a compression check

[1980Mooney]

6 hours ago, G-SLOT said:

The maintenance just before the flight (albeit a week before) was an oil and filter change. The injector was replaced at the annual in October (about 18 flying hours ago) because i was chasing smoother LOP operation and it was discovered it was the wrong type of injector (was type F while, according to GAMI, it should have been type E - replacement kindly supplied by GAMI). Running smoothly since (including LOP). The odd thing is that, during yesterday’s flight, it was running nicely LOP while I was at 2,300’ but, sometime between going ROP for the climb and settling into cruise at 10K, that cylinder became leaner, evidenced by it peaking way before the other cylinders. 
 

My best guess at the moment is that some small piece of debris became lodged in the fuel line or the cylinder 2 injector, but I’m very keen to hear other theories. 
 

That injector, and possibly the line supplying it, will definitely be the first thing I have checked. 

Have you pulled the cowling and actually looked at the #2 injector?  When the #2 injector was replaced the only place that the stainless steel pressure fuel line should’ve been disconnected was at the top of the injector.  If debris had entered the line at that point on top of the injector then it should’ve turned up immediately in the injector when the engine was run - not 18 hours later.  There would be no reason for the mechanic to disconnect the fuel line at the fuel distributor when installing an injector. If he did disconnect it at the fuel distributor because he thought the line was dirty then he should’ve cleaned all six lines since they’re all fed from the same common fuel distributor. 
 

I would look for blue stains around the fuel line fitting or on the fins of the cylinder. If that ball and socket fitting mating the injector to the the stainless steel fuel line wasn’t properly torqued it could’ve backed off from vibration during flight and be spraying fuel on the engine.
 

It happened to me. 

[carusoam]

Fantastic post G-slot!

All the data is present and easy to see...

Your discussion points right to what is going on... when...

 

Cylinder two is clearly mis-behaving...

As if something is blocking the fuel flow going to the injector...

How far up the line from the injector to the spider... we can’t tell from this set of data...

As far as Gami sensitivity... Gamis are all about the final injector diameter... in thousandth of inches...

Looks like you have something stuck, blocking the flow... see if it is stuck in the injector first...

Just because that is the logical place for things to get stuck... once they fit into the line...

See if you can catch, or photograph what ever caused the blockage...

PP thoughts only, not a mechanic...

Best regards,

-a-

[carusoam]

Something a bit odd about the previous flight data...

EGT1 and EGT2 are a bit colder than the other four...

And the data presentation has a bit of a skew going on... 

See the data in the screen pictured below...  note the double data points making EGT2 fall off the bottom of the list...

The order of the data 1-6 is not following the usual 1-6....

It might be an iPad issue...

I can invite Paul to have a look as well... see if there is something going on at Savvy... @kortopates

[1980Mooney]

9 minutes ago, carusoam said:

Fantastic post G-slot!

All the data is present and easy to see...

Your discussion points right to what is going on... when...

 

Cylinder two is clearly mis-behaving...

As if something is blocking the fuel flow going to the injector...

How far up the line from the injector to the spider... we can’t tell from this set of data...

As far as Gami sensitivity... Gamis are all about the final injector diameter... in thousandth of inches...

Looks like you have something stuck, blocking the flow... see if it is stuck in the injector first...

Just because that is the logical place for things to get stuck... once they fit into the line...

See if you can catch, or photograph what ever caused the blockage...

PP thoughts only, not a mechanic...

Best regards,

-a-

If an injector is blocked then he should see a reduction in total fuel flow as compared to the same rpm, MP before the condition.  If there is a leak at the injector the total fuel flow will be the same but the cylinder with the leak at the injector fitting will be lean. 

Edited February 22, 2021 by 1980Mooney

[carusoam]

2 minutes ago, 1980Mooney said:

If an injector is blocked then he should see a reduction in total fuel flow as compared to the same rpm, MP before the condition.  If there is a leak at the injector the total fuel flow will be the same but the cylinder with the leak will be lean. 

Not necessarily...

The fuel blocked from one injector tends to find the next easiest route out the neighboring injectors...

Measuring FF to the individual injectors would be sweeeet!

As would the air flow at any point...

Best regards,

-a-

[KLudwick]

I'd say you have a clogged injector, or a clog at the spider (less likely, but can still happen).  I just had this recently happen as well!

[1980Mooney]

7 minutes ago, carusoam said:

Not necessarily...

The fuel blocked from one injector tends to find the next easiest route out the neighboring injectors...

Best regards,

-a-

Wait a minute.  That is only if fuel pressure rises in the distributor.  At a given pressure in the fuel distributor the flow through the other five orifices will remain exactly the same regardless of whether #2 is open or obstructed. Bernoulli!

[carusoam]

Way to go Dr. Bernoulli...  

do you have any of the precision pressure instrumentation to see the difference?

block off an entire FI...

Sure... The pressure will try to rise...

Know that... The viscosity of 100LL is so low...

I don’t think... you will see a change in FP...  mostly because Ovations don’t have a FP gauge....

And no resulting change in FF... because that is set by relationship to air flow by the fuel injection system...

So...

I believe... the fuel continues to be delivered the same way to the spiders...

And.... the five spiders happily deliver the fuel instead of six...

Got a pic of how the fuel divider works?

This would be interesting...

Because it looks like the flow to both cylinder 1 & 2 may be affected... in the previous flights...

But only #2 seems to be suffering the most on the last flight...

 

PP thoughts only... I may have read about Bernoulli and his family before... and that guy who defined Newtonian flow...

All stuff I have read about around here... not any of my original research...

A pic of the fuel divider may hold the key... if the fuel injector shows up clean...

Best regards,

-a-

 

[1980Mooney]

1 hour ago, carusoam said:

Not necessarily...

The fuel blocked from one injector tends to find the next easiest route out the neighboring injectors...

Measuring FF to the individual injectors would be sweeeet!

As would the air flow at any point...

Best regards,

-a-

And if pressure does rise at at Fuel distribution Manifold, or spider, then as you say the flow will find its way to the other five cylinders. That means they’ll be running rich and one will be running lean. He will dial back the mixture to achieve optimum ROP or LOP on the remaining five and as a result you will see a reduced overall fuel flow rate which you can measure. 
 

if the overall flow rate has dropped as compared to before the condition then it is indeed an obstruction. If the flow rate is the same as before the condition then it’s a leak. 

Edited February 22, 2021 by 1980Mooney

[carusoam]

great point 1980!
 

The Gami spread is kind of the way this done...

Mostly because the other measurements are such tiny physical changes...

Some people are really skilled with knowing where their EGTs always are... under various engine operations...

And finding where peak usually occurs the same way...  (strong memory skills helps)

 

I would bet... the OP can do a Gami spread while on the ground, and a slow run-up for a second confirming piece of information...

To see if 1&2 are mis behaving the same way today...  or if it is just #2...

 

Having a pic of the flow diagram of the fuel divider may reveal if there is any relationship to 1&2 being so cold...

 

One of the things that adds to the challenge.... while looking at the previous day’s data... It is LOP.

And the 1 &2 are colder than the others...

I’m guessing that they are leaner... and thus colder...  (but should verify this, before going too far...)

The data is there, but, It requires a bunch of memory that some of us aren’t so lucky to have...   

 

PP thoughts only, not a mechanic...

Best regards,

-a-

[G-SLOT]

@carusoam, re EGT1 & 2 being below the others, this has pretty consistently been the case with my engine though I've always been advised to ignore the actual EGT values as they are very dependent on probe placement.

Plan is to pull the Cylinder 2 injector this week (before the next flight) and have it cleaned.

[carusoam]

Very true regarding raw EGT data...

Especially early on in the use of engine monitors...

By the time they were getting installed by the factory in the 90s... the consistency improved some...

Are you able to get a Gami spread test done, before sending it to see the mechanic?

Best regards,

-a-

[G-SLOT]

1 hour ago, carusoam said:

Very true regarding raw EGT data...

Especially early on in the use of engine monitors...

By the time they were getting installed by the factory in the 90s... the consistency improved some...

Are you able to get a Gami spread test done, before sending it to see the mechanic?

Best regards,

-a-

With the JPI700 not recording fuel flow, my GAMI spread test has to be done manually. Data I collected a while ago as well as experience running smoothly LOP pointed to the spread being within 0.5GPH but I have found that small (0.1GPH) variations in fuel flow create a small margin of error in this data. I really need a JPI900 or 930, which may be in my future plans (then again so is a second engine...).

I'm not planning to fly before the mechanic has the cowl off and cleans the #2 injector. If there is something blocking that injector, I don't want to risk an over temp on departure. The interesting thing is that, if something is blocking the injector, the blockage was created during the flight (at or after time 00:25 when I went ROP for the climb). That said, from my airborne investigations on Saturday, I determined the spread to be circa 2GPH (with cylinder 2 peaking first and the other five cylinders peaking around the same time about 1.8 GPH leaner).

[PT20J]

While the injector is soaking, I’d check and clean all the fuel screens and run some fuel through the fuel line to the injector into a clear container to see if any contaminants show up.

[G-SLOT]

4 minutes ago, PT20J said:

While the injector is soaking, I’d check and clean all the fuel screens and run some fuel through the fuel line to the injector into a clear container to see if any contaminants show up.

Good idea. The mechanic thought it would be difficult to see anything but it would be an excellent confirmation of the problem if something does come out. The next question would then be where it came from!

[carusoam]

Always good to attempt to capture whatever is blocking a fuel line...

Often it is a simple build up over time....

In this case something seems to be floating in the system, and found a place to rest in flight....

Similar to a plaque in an artery getting loose...   

M20Rs haven’t shown any of these that I’m aware of...

But, some IO360s around here have had some rubber fuel hoses related to the fuel injection system, shed internally... just like the plaque mentioned above...

Not a big deal, until it finds a smaller hole to block... 

Then it is time to replace a rubber hose...

There are a couple of screens in the system before the fuel injectors....

Unfortunately, I don’t remember the details enough to list...

We can invite @M20Doc to join the conversation... (IO550 fuel injector blockage question...)

Best regards,

-a-

 

[G-SLOT]

Update from earlier today: Cylinder 2 injector removed and cleaned (no visible sign of debris) and the fuel line from the distributor to that injector was also disconnected and had high pressure air put through it to clear out anything that might have been causing a blockage. Also, the area was inspected for any fuel staining that might have indicated a fuel leak and fortunately there was none. All put back together, fired up and flew for about 40 minutes. Delighted to say that smooth LOP operation has now been restored and it seems like some miniscule blockage of the GAMIjector was the likely culprit.

Looking back through the engine data from the flight in question, it does seem that whatever restricted the fuel flow to Cylinder 2 arose when I transitioned from LOP cruise to a ROP climb. Hopefully a one-off but thanks for all of the comments and suggestions. 

[carusoam]

Great follow-up G-S!

Without catching what ever was the issue...

Keep an eye out for it, just in case it shows again...

 

 

I had a blockage in a line once...took it to the mechanic... he said put more water through it next time... it’s dehydrated...

I followed his advice... the blockage resurfaced a few months later, much worse the second time...  he was the worst mechanic  in all of NJ...

Different kind of machine altogether... 

Best regards,

-a-