VAF Forums

VAF Forums (https://vansairforce.net/community/index.php)
-   RV Ongoing Maintenance Issues (https://vansairforce.net/community/forumdisplay.php?f=76)
-   -   Lost a cylinder on take-off (https://vansairforce.net/community/showthread.php?t=94516)

DakotaHawk 12-27-2012 11:51 AM

Here's the follow-up...

Last night I was able to follow up on a couple of the suggestions offered here.
  • I tested fuel tanks for water. No water found.
  • I checked all connections on Light Speed coils and plugs. No problems found.
  • I removed and cleaned the injector insert. Found some sooty colored black on the insert, but no blockage. All injectors looked good.
  • Checked intake gaskets and connector rubber for any intake air leaks. All tight.
  • Downloaded EMS data. Data shows a sharp spike of about ^300F on the #2 EGT about eight seconds after RPM went up to 2700rpm (take-off roll). About 2 seconds later, the #2 EGT dropped low (down to about 250F). Over the next 45 seconds, I reduced power to 1950 rpm, and then chopped power to land. When I chopped power, the #2 EGT picked back up and tracked with the other EGTs.


Ron Lee 12-27-2012 12:46 PM

Many of the colors run together so it is hard to ID some things. There are lines at the bottom that may be the four CHTs. Is the one that drops below the others CHT #2? If so, and this in not the norm for your engine, there may have been a combustion related issue with #2.

Mike S 12-27-2012 12:53 PM

That EGT line sure looks like an interruption to the fuel supply for that cyl from what I can see. Looks like a piece of crud got in the way, then let fuel through, then again partially blocked the fuel at the end.

Did you flow any fuel from the line when you had the injector orifice out??

BobTurner 12-27-2012 01:23 PM

I agree with Mike, it sure looks like a partially blocked fuel line or injector. Not that there aren't other explanations. It looks to begin just prior to takeoff, perhaps taxiing onto the runway, where all EGTs went up a bit but #2 is much higher, running lean. At full throttle mixture is so lean (?) that it quits firing, EGT and CHT drop toward ambient (CHT slowly). After retarding throttle #2 starts firing again but is still too lean, high EGT compared to others. Looks like you added throttle to taxi, #2 briefly quits again, but comes back but still too lean after you reduce throttle.

BobTurner 12-27-2012 01:49 PM

I would try the coke bottle test: remove all the injector line clamps (do not bend the lines), remove #2 and at least one other injector from the cylinder heads, put identical glass jars (coke bottles) under each injector, apply full throttle, full rich, turn on the fuel pump. Wait a minute or so, depending on the jar size, turn fuel pump off. Compare fuel levels in the jars. If #2 is obviously less than the others, you have a flow problem.

Tom Martin 12-27-2012 03:01 PM

A stuck valve could give the same readings. It happened, I believe, after a reduction of power. My experience with stuck valves has been on the ground, first start, and in the air after a power reduction.

N427EF 12-27-2012 09:48 PM

Hard to argue with experience Tom.
While you are correct in thinking that a stuck valve would result in the same kind of data read out, I believe a fuel problem is more likely.
Here is why:
DakotaHawk has been running mogas for some 500 hours by his own account and gumming up the valve stems with lead is somewhat unlikely, although possible. A stuck valve is almost exclusively the result of lead build up in the valve guide or on the valve stem.
Soooo, I think a plugged injector line even temporarily, has caused the rough running cylinder. I am also quite certain that this was not ignition related because of the almost ambient temperature reading on the EGT.
Having 2 ignition sources and one go out would elevate EGT's somewhat and having both go out at the same time on only one cylinder is highly unlikely.
My guess is an obstructed injector nozzle or line.
I don't think vapor lock or water contamination affects one cylinder only.
Let us know what you find, for educational purposes.

Lars 12-27-2012 10:29 PM

stuck valve
 
I'm gonna stick with my stuck valve assessment. While I am certain (would bet substantial money on) that the only other possible explanation is a clogged injector, all the evidence Scott has presented looks exactly what I saw as I chased my issue. Every time it happened was after a reduction in power from wide open throttle and max rpm. One giveaway was in the post describing sooty exhaust stack on the offending cylinder, but not on the adjacent one. Fire is continuing, but it's happening in the exhaust stack, blackening it, rather than inside the cylinder where it belongs.

What will help with troubleshooting is if it happens again. If it's the same cylinder, repeatedly, as it was in my case, then it's not the injector nozzle.

Mike S 12-27-2012 11:46 PM

Quote:

Originally Posted by Tom Martin (Post 728828)
A stuck valve could give the same readings. It happened, I believe, after a reduction of power.

Quote:

Originally Posted by Lars (Post 728919)
Every time it happened was after a reduction in power from wide open throttle and max rpm.

Look at the graph-------egt started up about 2 sec before throttle reduced, peak egt about a second before.

Not saying it isnt a valve, but if so, it was not after pulling power.

rv8ch 12-28-2012 03:53 AM

Sherlock Holmes
 
Impressive detective work going on here, gents. I don't have anything to add, except that this is really educational and entertaining. If anyone is not yet convinced that an engine monitor with logging is worth it's weight in avgas then they have not been paying attention!

Tom Martin 12-28-2012 04:41 AM

As a child conversing with my mentor neighbor about small gas engine problems told me that it was all very simple. "It takes three things to make and engine fire. Fuel, spark and compression.
Check fuel first, then spark then compression (mechanical items)
While it might be fuel, that is the easiest solution, the symptoms seem like valve to me. I have experienced partially blocked injectors but not a full blockage and so I do not have that to compare it with.
As far as the cylinder EGT peaking early and then falling off; Yes that could point to a partially blocked nozzle but we have no way of knowing if this cylinder normally peaks early, there is only one set of data here. Perhaps if we could go back a bit farther to a previous take off and look at that data we could see if things were similar.
If it were my plane, I would clean the injectors, flush the fuel system, fill with 100LL, change the spark plugs, and do the valve wobble check. These items are not expensive nor do they take a lot of time, why not?

N427EF 12-28-2012 07:32 AM

I too find the analyzing that is going on in this thread very enjoyable and most of all educational.
In finding the source of the problem, wether that be a stuck valve or a clogged injector, you have to go a little deeper.
If this was a stuck valve problem it can quite easily be duplicated since the stuck valve problems are the result of lead build up in the valve guide and or valve stem.
A clogged or partially clogged injector can be the result of dirt in the fuel
or other particle, inhibiting fuel delivery.
There is documented evidence that severely gummed up fuel lines sometimes
shed a flake of varnish and cause a clog in the injector nozzle.
This may be more llikely the result of stale mogas than avgas.
If this was the case then you will not be able to duplicate the problem until the next flake comes off the injector fuel line.
If you have done all the things that Tom describes above and the problem goes away, I would do one more thing, replace the injector lines.
Not expensive and then you can slice one open and see what it looks like on the inside.

DanH 12-28-2012 07:59 AM

Scott, can you full-scale the EIS timeline between 10:41 and 10:45?

If that's a Bendix FI, pull the servo inlet fitting and check the fine screen.

hydroguy2 12-28-2012 08:23 AM

Quote:

Originally Posted by Tom Martin (Post 728945)
?.....m
If it were my plane, I would clean the injectors, flush the fuel system, fill with 100LL, change the spark plugs, and do the valve wobble check. These items are not expensive nor do they take a lot of time, why not?

Me, too.

Coming home from OSH, I had a erratic EGT reading on #1. We were over Ewstern Montana at 12500 or so and just watched it....then got a stumble. I juggled the mixture a bit and started thinking of "Nearest airport". CHT was always rock solid on all cylinders. An hour later it happened again. I increased fuel flow thinking I was getting a lean condition on that cylinder. Fingers crossed the the remainder of the trip. Next day I did just what Tom described....inspected injectors, flow checked them too. All ok. Replaced the plugs(auto). Checked the wiring to the EGT connectors and then did a compression check. I found nothing....but have not had a stumble since.

Mine was different than yours, Scott, but I think the troubleshooting is the same. Do the cheap free stuff first. And even then sometimes it's tough to figure out what is going on.

Lars 12-28-2012 10:29 AM

Quote:

Originally Posted by Mike S (Post 728928)
Look at the graph-------egt started up about 2 sec before throttle reduced, peak egt about a second before.

Not saying it isnt a valve, but if so, it was not after pulling power.

True dat... why I should not post things after a 6 hour drive home, cocktail in hand...

Does speak to interrupted fuel supply, possibly.

Unfortunately, based on my experience, the only way to identify the cause may be to keep flying until it recurs. I was never able to duplicate the condition on the ground. The first couple of times it happened to me I was terrified- I flew straight back to the airport (I was always nearby, being in Phase 1 at the time) and landed. And the problem always "cured" itself before I landed. Eventually I realized that the engine wasn't on the verge of rapid inflight-disassembly and I was able to get some additional data.

Mike S 12-28-2012 10:40 AM

Quote:

Originally Posted by Lars (Post 729027)
a 6 hour drive home, cocktail in hand...

Wow, I would not have been able to nurse a cocktail that long:D

MauiLvrs 12-28-2012 12:08 PM

Quote:

Originally Posted by DanH (Post 728983)
Scott, can you full-scale the EIS timeline between 10:41 and 10:45?

If that's a Bendix FI, pull the servo inlet fitting and check the fine screen.

And add MP and fuel flow...

Scott Hersha 12-28-2012 03:44 PM

The definition of a cool head may have a different meaning for a pilot that flies upside-down a couple of feet from other aircraft close to the ground than it does for the rest of us (Kahuna). Trouble shooting in flight is called 'flight test' and should be done solo. Whenever anything goes wrong in flight - especially a rough engine in a single engine airplane - the number one measure of success is did you get the airplane on the ground safely without causing harm to people or property? You scored 100% in that regard, Scott...... Congratulations! The method you used to accomplish that goal is not important, the result is. You didn't further tarnish the name of experimental aviation in the eyes of the non-flying public or scare the young lady away. In fact you gave her a good glimpse of how possibly scary circumstances can be handled with success. When you get your problem fixed, I hope she goes up with you again.

DakotaHawk 12-28-2012 08:26 PM

Quote:

Originally Posted by MauiLvrs (Post 729048)
And add MP and fuel flow...



Ask and ye shall receive!

LifeofReiley 12-28-2012 08:29 PM

I'm thinking you had a stuck valve... short period IMHO. Fly the plane. :)

DakotaHawk 12-28-2012 08:49 PM

And here's the last chapter to this entertaining troubleshooting event...

This morning, my engine guru managed to pull himself away from his wife's honey-do list for some man cave time. We snuck away to my hangar to get to the bottom of this mystery.

We began by pulling the injector from all four cylinders. Using brake cleaner and Q-tips, this is what we got :eek:



Based on the stuff found in the injectors, we backed up to the inlet filter in the fuel servo, where we found more cruddies. The crud found in the inlet filter had the look of fine-hair fiberglass mat.

Based on the CRUD found in the fuel servo, we backed up to the fuel filter located upstream of the electric fuel pump. We found larger chunks of debris. The debris was smearable black material. The amount found in the fuel filter was not very much. It may have blocked about 1% of the fuel filter. The dismaying part is that the fuel filter had been cleaned during my last annual about 20 engine hours ago!

Based on the black, smearable material found, we postulated that it may be ProSeal that may be breaking off in the tanks and migrating through the filters. If you remember from a previous post in this thread, I run MoGas most of the time. (non-alcohol Premium grade) So I drained both fuel tanks, and allowed them to dry out for awhile. Using a q-tip taped to a long stick, I swabbed several ProSeal locations inside the tanks to see if there was any softness or failure in the ProSeal. I didn't get any smearing of the ProSeal.

Next, I hooked up a fuel pump to the fuel tank sump, added five gallons of AvGas, and recirculated the fuel (through two in-line see-through filters) back to the tanks for about an hour. The filters did not seem to get any dirtier, but I did not cut them open to examine the media up close. Finally, I filled the tanks with AvGas and went up for about twenty minutes of pattern work. Everything ran fine at this point, and I'm almost done with this chapter.

My last action item will be to open up the fuel spider, remove the diaphram, and clean out the orifices in the spider to make sure that there's nothing left in the system.

Although I've been flying for 480 hours on MoGas, I think that I'm going to have to stick to AvGas. I don't know if I got a bad batch of fuel that caused all of this trouble, or if this was building up and just finally reached a peak. (I think it's the bad batch of fuel - based on cleaning all screens and filters just a few hobbs hours ago.)

rocketbob 12-28-2012 09:24 PM

Scott, you're fuel filter is the problem, not the mogas. Its not fine enough to catch the crud that is making its way upstream.

digidocs 12-28-2012 10:29 PM

Scott,

Would you mind posting what fuel filter you are currently using?

Thanks,
David

DakotaHawk 12-29-2012 01:31 AM

Quote:

Originally Posted by digidocs (Post 729192)
Scott,

Would you mind posting what fuel filter you are currently using?

Thanks,
David

Van's Fuel Injection Installation Kit

I have the standard fuel injection system from Van's with the Airflow Performance fuel filter.


WAM120RV 12-29-2012 02:27 AM

Good job but.....
 
Hi Scott,

While all this stuff if bad I would not assume that it was the cause of your problem.......it may be, but for you loose a cylinder, the fuel feed to that cylinder would have to be temporarily cut off.

Compression, fuel and a spark are the three things you need, so a sticky valve could still be the thing that actually caused it.

I have often come across situations like this where you think you have found the culprit only for the problem to occur again, so take care for the next few flights.

Doug 12-29-2012 04:40 AM

The MoGas - do you transport it by some means to your airplane? If so I would look at these items first.

Walt 12-29-2012 05:23 AM

Quote:

Originally Posted by DakotaHawk (Post 729212)
Van's Fuel Injection Installation Kit

I have the standard fuel injection system from Van's with the Airflow Performance fuel filter.

The AFP filter is 125 micron so it's not an overly fine unit, I prefer to use a 74 micron unit.

BillL 12-29-2012 07:03 AM

Where to source?
 
Walt, where to you get that 74 micron filter and will it fit in the Airflow performance housing? Is the pressure drop much more to affect vapor formation? Thanks.

For reference:

Oil filters ~ 30 micron

Diesel prefilters ~ 15 micron

Diesel fuel filters ~ 5 micron

Walt 12-29-2012 07:20 AM

Quote:

Originally Posted by BillL (Post 729244)
Walt, where to you get that 74 micron filter and will it fit in the Airflow performance housing? Is the pressure drop much more to affect vapor formation? Thanks.

For reference:

Oil filters ~ 30 micron

Diesel prefilters ~ 15 micron

Diesel fuel filters ~ 5 micron

I use a FlowEzy 3000 unit, you may want to check with Don at AFP for a replacement element source on his unit.

erich weaver 12-29-2012 09:28 AM

See posts 28 and 29 in this thread for info from Don at AFP regarding the filter screen size:
http://www.vansairforce.com/communit...=Filter+screen

Erich

N427EF 12-29-2012 10:40 AM

Black gooo?
 
What is the black goo?
Has your proseal become soft?
Have you run mogas in both tanks?
How many years on mogas?

Glad you found the "apparent" problem but I would search a little further too.

DanH 12-29-2012 11:33 AM

Returning to the original issue.....I'll put my money on a partially blocked injector orifice.

This is the previous EIS graphic, stripped of everything except #2 EGT and fuel flow:



The cylinder runs fine (EGT is normal) below fuel flows of around 4 GPH. At 5 GPH it's going too lean, and at 6 or more it can't light. Put another way, the blockage is (was) limiting #2 fuel flow to about 1 GPH (1/4 of 4 GPH), which is fine at low air flow, but additional air (manifold pressure requiring more than 1GPH) just makes it too lean to run.

I asked Steve about checking the Bendix servo inlet screen because it incorporates an interesting detail. Fuel is delivered to the inner bore of the screen. The screen is held against the inlet fitting with a very light spring. If the trash load gets too high fuel pressure pushes the finger screen away from the inlet fitting during a period of high required fuel flow, allowing fuel to bypass the screen. That's good given the alternative (no or reduced fuel flow), but when it happens a clump of collected trash often escapes downstream into the servo, divider, and/or injector(s). That's how trash gets to an injector despite the upstream screens.

It's a small, very fine screen, easy to restrict, in particular if any kind of goo comes down the line. The Bendix literature says it should be cleaned every 50 hours. I'd bet most are not cleaned anywhere near that often. When you do clean one, remove it from the inlet fitting end, not the plug end, again so you don't spill collected trash into the screen bore.

Cleaning injectors: You really need a magnifier to examine injector bores, 10x minimum. Here's my long term favorite; I've even shot photos through it:

http://www.grainger.com/Grainger/BAU...agnifier-3G993

Scott, can you tell us a little more about what you found on the inner bore of the inlet screen? Fiberglass hair eh? Any of the black goo too?

MauiLvrs 12-29-2012 11:57 AM

Quote:

Originally Posted by Walt (Post 729225)
The AFP filter is 125 micron so it's not an overly fine unit, I prefer to use a 74 micron unit.

Keeping in mind that if the filter gets plugged ... all cylinders go to sleep :eek:
Quote:

Originally Posted by DanH (Post 729303)
Returning to the original issue.....I'll put my money on a partially blocked injector orifice.

Blockage somewhere ...

cdmiller 12-29-2012 01:16 PM

Check fuel distribution manifold
 
A blockage effecting a single cylinder can occur anywhere from the fuel distribution manifold to the injector. I would open and inspect the distribution manifold for foreign debris. I had this happen to a new IO-360M1 with similar symptoms to yours. In my case, the blockage would occasionally move around in the fuel distribution manifold, sometimes effecting a different cylinder and sometimes effecting none. Very frustrating to trouble check.

Dan Miller
RV-8

Lars 12-29-2012 01:20 PM

Quote:

Originally Posted by DanH (Post 729303)
Cleaning injectors: You really need a magnifier to examine injector bores, 10x minimum.

Agree in principle- I used my own 10x lighted eye loupe to examine mine, but I'm gonna guess that a particle big enough to reduce fuel flow by a factor of roughly 4 will be visible to the naked eye.

Also worth doing a flow test since it's so easy. 4 baby food jars work great- everyone has baby food jars in their hangar, right?

IowaRV9Dreamer 12-29-2012 02:45 PM

picture of bendix inlet screen?
 
Quote:

Originally Posted by DanH (Post 729303)
It's a small, very fine screen, easy to restrict, in particular if any kind of goo comes down the line. The Bendix literature says it should be cleaned every 50 hours. I'd bet most are not cleaned anywhere near that often. When you do clean one, remove it from the inlet fitting end, not the plug end, again so you don't spill collected trash into the screen bore.

Does anyone have a picture of this screen assembly, and how it comes out? I don't understand the part about "remove from inlet fitting end, not the plug end"

Thanks!

roee 12-29-2012 03:48 PM

Quote:

Originally Posted by IowaRV9Dreamer (Post 729358)
Does anyone have a picture of this screen assembly, and how it comes out? I don't understand the part about "remove from inlet fitting end, not the plug end"

Thanks!

The fuel inlet part of the throttle body is bored all the way through from left to right, and is threaded on both ends. You can choose which side you want the fuel inlet fitting to be on for your specific installation. So you thread the inlet fitting into the side you choose (and orient the inlet screen accordingly), and you thread a plug into the opposite side.

So, about removing the inlet screen for cleaning, as others said, you want to remove it by removing the inlet fitting, not the plug on the opposite side.

Documentation from Precision Airmotive:
http://precisionairmotive.com/Publications/25-020_a.pdf
See section 6.2 (pg 3), and drawings on page 8.

DakotaHawk 12-29-2012 03:57 PM

Quote:

Originally Posted by cdmiller (Post 729328)
A blockage effecting a single cylinder can occur anywhere from the fuel distribution manifold to the injector. I would open and inspect the distribution manifold for foreign debris. I had this happen to a new IO-360M1 with similar symptoms to yours. In my case, the blockage would occasionally move around in the fuel distribution manifold, sometimes effecting a different cylinder and sometimes effecting none. Very frustrating to trouble check.

Dan Miller
RV-8

Yep! Great suggestion. The spider (fuel distribution manifold) is off the engine and getting opened and cleaned today. I'll let you know what I find.

PS. The spider (according to my engine guru) is very, very, very susceptible to any impurities, and also to any mis-handling. So much so, that he asked that I deliver it to him un-opened. He'll open and clean it.

PPS. It's also expensive to replace if damaged!

cdmiller 12-29-2012 04:55 PM

Spider sensitive
 
Grumpy,

I didn't open and mess with my spider either. I sent it back to Bart at AeroSport Power and he found the debris and removed it. I think he said it was metal cutting(s) that could move around in the spider and cause partial blockage to a cylinder. No problems in hundreds of hours since.

Dan Miller

MCA 12-29-2012 06:12 PM

Here's something you might want to evaluate. I haven't used it yet, but looks interesting. https://www.savvyanalysis.com/home

The graphing stuff is free, and for a fee they will do some analysis. Mike Busch does an interesting presentation at OSH about reading engine data and how to interpret it.


All times are GMT -6. The time now is 06:55 AM.