Rocket Engine Stumble

[SpamPilot]

I've got an intermittent stumble/miss/hiccup on my Rocket 305.

On climbout, once or twice a minute you can hear a miss.  It lasts for about as long as you can snap your fingers.  EGT/CHT/Fuel Flow are all normal, and there's no question it's making rated power.

At 26"/2200 RPM cruise, leaned per the power chart, the stumble still happens, but less frequently.

Pulling back below 20" for slow cruise at 2200 RPM, it becomes much more noticeable.  Seems to happen a few times a minute.  Each stumble lasts about as long as you can count to four as fast as possible (call it 300-400 ms).  You can hear the prop speed slow down (a bit), then pick back up.  You can hear the slight RPM overshoot past 2200 RPM as the prop governor reacts to the sudden change in power.  EGT/CHT/Fuel Flow are all still normal.  Leaner or richer doesn't seem to make a difference.  The engine runs clean and strong between stumbles.  Hard to say if it feels like one cylinder is missing or they all cut out for a few revolutions, but I don't think it's all of them, because the sound doesn't drop to quiet during a stumble.    Out of the corner of my eye, maybe I see cylinder #1 EGT drop briefly and come back up, but also maybe I'm imaginging it.

On downwind to base, mixture full rich for landing, pulling the throttle back below 15" causes the engine to stumble so often you're not sure if it's going to keep running.  Adding a bit of throttle cleans it up a bit, but it still stumbles.

Back on the ground - no stumble.  Mag check at 1700 RPM is fine.  2000 RPM is smooth.  Any manifold pressure or RPM is smooth.  The only way I can get it to stumble is if I yank the throttle closed from 1700 RPM or so, in which case it coughs and sputters and then catches itself before it settles into idle.  600 RPM idle is a little rough, but anything over 700 RPM is clear as a bell.

Recently, a leaking fuel pressure regulator was replaced with an overhauled one.  The regulator was adjusted to ensure the correct fuel flow at max power.  The stumble did not happen (or was not noticed) before the overhauled regulator was put in, except that it did want a little bit of throttle on approach to keep running smoothly.  No other changes have been made to the fuel system, ignition system, or engine.

Any thoughts?

 

Edited December 16, 2018 by SpamPilot

[LANCECASPER]

It's fuel, air or spark. My first guess is spark. Magnetos, spark plugs or ignition harness.

     Have you done an in-flight mag check? That test is easy and almost free. Hopefully you have an engine monitor that you can change to one or two second sampling       and then download the results. How long since the mags were inspected?

      Next, if the spark plugs are Champions I'd test them for high resistance and start thinking about Tempest Fine Wire plugs. If the harness is old that might be a                  possibility.

      Maybe a clogged injector? The engine data should tell you that. 

[Cody Stallings]

Mine does that in the decent an on final.

i thought it was fuel related.

If I keep the mixture lean(within reason) it pretty well goes away, but if you push the mixture all the way in it gets really rough.

Im just learning to live with it.

[SpamPilot]

Thanks, guys.

I found Mike Busch's page on in-flight tests, including the ignition stress test, mixture distribution test, and induction leak test.

Mags were overhauled at last major 400 hours ago.  Tempest Iridium Fine Wire plugs with a new ignition harness were installed then, along with new GAMI injectors and a raft of other new or overhauled things you'd expect at major.

Cody, thanks, it's (somewhat) comforting to hear of another Rocket that may have the same issue.

Time to go make a data download cable for the EDM-700...

[M016576]

Worst case scenario would be a sticky valve

[SpamPilot]

Found this post here on MS, asking about possible fuel vaporization.  Jeff H's description fits my experience closely:

"the engine stumbled in the pattern and then quit on roll-out.  We were able to recreate the problem this weekend:
- 20" MP or less
- Intermittent stumbling."

I have had the engine quit on roll-out several times, though after my A&P and I replaced the leaking fuel pressure regulator and adjusted the idle speed from 550 RPM to 625 RPM, this has stopped happening.  (incidentally, the fuel pressure regulator, which is mounted to the firewall, had leaked so badly that I had blue fuel streaks coming out the cowl and down the left side of the airplane after a flight).  Now the engine "diesels" when I pull the mixture to idle cut-off, and doesn't stop easily unless I pull hard simultaneously on both the throttle and the mixture.  (it's not actually dieseling - it shuts off positively if I turn off the mags)

Seth reported that he had a similar issue, and suspected it's simply a problem with the way the mixture was set.  He talked about how it was a challenge to set the fuel controller so the idle mixture would be lean enough to avoid stumbling at low power while keeping the max power mixture rich enough to keep the EGTs and CHTs safely low.

Alas, it doesn't look like Jeff H reported on whether he ever found a problem or was simply able to live with the issue by leaning as Seth suggested.

 

 

[SpamPilot]

Thanks, M016576.  I'll have to give some consideration as to whether this might be sticking exhaust valves.  The engine runs rough immediately after start.  You can see in the EGTs where two or more cylinders aren't firing after start, and you can hear the engine get progressively smoother as they come in one by one.  I had assumed this was because it's 40F outside and I'm flooding those cylinders from all the priming and full rich mixture.  It can take 30 seconds for the last one to clear out.

Edited December 16, 2018 by SpamPilot

[flyboy0681]

Two years ago I lost all oil and at tear down the shop determined it to be a sticking exhaust valve, but that was for a Lycoming. Like you, all systems appeared normal during the flight and engine monitor data looked fine too when downloaded. What little oil was drained showed a lot of metal. You had better nip this one in the bud before it costs you an overhaul too.

[ArtVandelay]

Two years ago I lost all oil and at tear down the shop determined it to be a sticking exhaust valve, but that was for a Lycoming. Like you, all systems appeared normal during the flight and engine monitor data looked fine too when downloaded. What little oil was drained showed a lot of metal. You had better nip this one in the bud before it costs you an overhaul too.

Educate me how a stuck exhaust valve can cause loss of all oil?

[Guest]

1 hour ago, SpamPilot said:

Thanks, M016576.  I'll have to give some consideration as to whether this might be sticking exhaust valves.  The engine runs rough immediately after start.  You can see in the EGTs where two or more cylinders aren't firing after start, and you can hear the engine get progressively smoother as they come in one by one.  I had assumed this was because it's 40F outside and I'm flooding those cylinders from all the priming and full rich mixture.  It can take 30 seconds for the last one to clear out.

Sticking valves would be quite rare in a TCM engine.  I’ve never worked on a Rocket, but do take care of 8 NB series engines installed in Cessna products.

I from your data download you know which cylinders are cold on start up you should be able to pull the rocker covers and check the valves for sticking.

Assuming that the engine is similar enough to a Cessna installation, you should be able to make it run quite smoothly.  A set of accurate pressure gauges and a knowledgeable maintainer should be able to set the engine fuel system correctly.

Clarence

[SpamPilot]

This is a YouTube video of a Lycoming 4-cylinder reportedly with a stuck valve.  My roughness on start-up sounds nothing like that.

M20Doc, thanks, how would one " check the valves for sticking "?  I find lots of information on the web on disassembling the top end and dropping the valves to inspect them and ream the guides, but I haven't yet found something that describes a procedure to simply check for sticking (which, presumably, would be difficult to confirm with the springs in place).

[Guest]

You need to do something called the rope trick.  You need remove the rocker cover and the rocker and upper spark plug.  On the compression stroke fill the cylinder with a clean 3/8” rope, turn the prop to compress the rope between the valves and piston.  Then using a valve spring compressor remove the valve keepers and springs from the suspect valve.  Turn the engine backward to relieve the pressure and removed the rope, now grab the valve stem and move it in the guide.  It should be free in the guide.  If not you’ll need to ream the guide.

Clarence

[SpamPilot]

I think I've pretty much ruled out the possibility of a stuck valve, at least to my comfort level.  The primary evidence pointing towards a stuck valve is what could be described as "morning sickness", as I mentioned earlier.  But I now think it's really just a few flooded cylinders.  Today I tried a different starting technique.  Instead of starting with mixture full rich, I leaned the mixture to the point where I have it while taxiing.  I then primed normally and started the engine (overnight cold soak to 50F ambient).  Previously, this consistently resulted in a long crank and a rough running engine with a few cold cylinders until all cylinders cleared.  This time, it fired in four blades and was running on all cylinders immediately.  All EGTs and CHTs were coming up, no cylinders were cold.  Maybe one successful start isn't proof, so I'll continue with this technique and see if it consistently solves the rough starting problem.  Combined with the widely held belief that stuck valves are a Lycoming problem (sodium-filled valves that run with much hotter stems than the Continentals, thus accumulating deposits), and the fact that the engine only has 400 SMOH, I think I should be looking at a different branch of the fault tree.

I haven't flown again yet, but I did operate the engine at 1100 RPM for a while, with a run up to 1500 RPM, recording video of the EDM-700 (which is so old it doesn't have datalogging capability).  Leaned to slightly ROP, it sounds fine, with just a bit of variation well within the few percent of incomplete combustion cycles any engine is expected to have.  I can reproduce the rough running I encountered on final by pushing the mixture full rich, and I can completely clear it up with leaning.  I can bog the engine at idle by going full rich.  No cylinder is cold, no EGTs are fluctuating, and EGTs all respond correctly to both leaning and to running on one mag.  I think I'm just running excessively rich at the full rich setting.  Sounds like this is the same as Cody is seeing.  I think probably either the fuel pressure regulator or the fuel controller may need to be adjusted to fully dial things in.  Regardless, Mike Busch (who has lived with the TSIO-520 for decades) recommends lean operation for every phase of flight except takeoff and initial climb, so I think I need to pay closer attention to that before I conclude there's anything really wrong.

 

 

Edited December 17, 2018 by SpamPilot

[FloridaMan]

I just read your initial post again and I think I may have experienced something similar. At full rich mixture on takeoff and initial full power climb, I would experience what felt like misfiring. Leaning a bit to where the fuel flow numbers matched book make the misfire go away. The second, and I included that in a private message that I sent earlier, is that we discovered my mechanical tach read 60-100 RPM high. I believe that @Yooper Rocketman mentioned that he had seen the mechanical tachs in rockets read high before. The significance of this is that at an indicated 2200 RPM and low manifold pressure settings that you might just be bootstrapping. I observed the same behavior in my rocket as well, and after discovering that my tach read a tad high, I realized I was running less than 2200 RPM during the test. According to Mike Busch, this may be evidence of "Bootstrapping"

If you're flying a turbo and climbing up to the Flight Levels, you'll probably find that you can't use bottom-of-the-green RPM without "bootstrapping" (a condition where the wastegate is completely closed, the turbocharging system is operating unregulated, and significant MP excursions become evident in flight). The cure for such boostrapping is to increase RPM in small (50 RPM) increments until MP stabilizes. If you still have difficulty stabilizing MP at high altitudes, you may have an induction or exhaust leak or some other engine problem. The best way to diagnose this is to perform a "critical altitude check" as outlined in the service manual. Critical altitude for a turbocharged airplane is the maximum altitude at which the engines can develop full takeoff power. For my T310R, that's supposed to be 16,000'; for a T210 or 340 or 414, it's 20,000'. If you can't get full takeoff power at the airplane's rated critical altitude, then you have a problem that needs to be found and fixed.

https://www.avweb.com/news/airman/184350-1.html

[flyboy0681]

20 hours ago, ArtVandelay said:

Educate me how a stuck exhaust valve can cause loss of all oil?

Took off from the departure airport with 6.5 quarts of oil, landed with none and there was nothing on the belly. The shop who did the overhaul said we must have been trailing white smoke for quit a while.

[SpamPilot]

FWIW, I've had an oil control ring failure on my M20F cause the engine to lose four quarts in an hour.  Good thing I started with seven.  No smoke, made full power, only indications were a couple ticks lower EGT on that cylinder and slowly decresing oil pressure after about the 4 quart mark.  Not sure how a stuck exhaust valve would cause that much loss - there's just not that much opportunity for oil to get past the valve stem - but I guess stranger things have happened.

Antares, thanks again.  I did confirm by analysis of the audio from a video recording that the tach is reading accurately.

" At full rich mixture on takeoff and initial full power climb, I would experience what felt like misfiring. Leaning a bit to where the fuel flow numbers matched book make the misfire go away."

Yes, that's my experience.  And yes, by manually leaning to book, I can eliminate the misfire.  It's not very much misfire, to be sure, but a sensitive pilot would notice it.  Turns out my 38"/2650 RPM fuel flow was 38-40 GPH, not the 33 GPH the power table says, so it was running quite rich.  So my A&P has (re)adjusted the fuel pressure regulator (he turned it the wrong way the first time and increased rather than decreased the pressure).  Haven't had a chance to do a fully-stabilized climbout yet, but on a max power fast taxi we're seeing about 4 GPH less, so much closer to book.  Also the number of turns of the mixture vernier at 1000 RPM from full rich until you start to hear an RPM rise has decreased from 10 turns to 6 turns (not full turns, just what my wrist normally does).

Some additional data - before this most recent mixture adjustment, when takeoff fuel flow was 38-40 GPH, with leaning I could eliminate the misfire at 2300 RPM for any MAP above 18".  Runs like a sewing machine.  Below 18", and running 1300-1400F EGT, the engine would stumble once or twice a minute.  It was enough of a stumble to put me on high alert and make sure I had emergency landing options within gliding distance.  A passenger would definitely have asked "what's wrong with the engine?".  It would almost never register on the EDM-700, because it was over so quickly, but the one time it lasted long enough to show up (about half a second), I saw EGT for cylinders 1 and 2 drop to nearly zero.  No kidding, the EGT bars pretty much disappeared in less time than it takes to read "No kidding".  Then they came right back up and the engine was back to firing normally.  The other cylinders didn't appear to be affected.  Enrichening the mixture seemed to make the stumble more frequent, and leaning the mixture didn't seem to help, but I was afraid of leaning too far and shutting all the cylinders down since I had the control fairly far out already.  I kept the RPM and MAP up until I was lined up with the runway on final, and then I felt OK about bringing manifold pressure down towards idle for the landing, where it ran a bit rough but never actually threatened to quit.

Regrettably, I can't repro the issue on the ground, but then again I can't achieve the combination of 2300 RPM and 18" MAP on the ground.

My A&P and I still think it's a mixture issue.  Might be that the fuel controller can't adquately meter fuel between 18" and idle if the fuel pressure regulator is set too high.

 

Edited December 20, 2018 by SpamPilot

[LANCECASPER]

If I owned the Rocket and it was doing this, and especially since others are experiencing the same,  I'd call Darwin Conrad up at Rocket Engineering in Spokane. Even though they don't do Rocket conversions anymore he still knows the airplanes intimately.

[13scrappy]

I had something similar in my rocket earlier this year.  I had the fuel pump rebuilt and the problem has ceased.  It may not be the answer for you but I have no issues now with full power in all settings and it is running as good and as smooth as ever.  I noticed it mostly on takeoff as it would not reach max Manifold pressure and start to lose power and ran much smoother at 35-36 inches.

[SpamPilot]

Thanks everyone for their helpful thoughts and suggestions.

This is now pretty much resolved.  It was a mixture issue.

Two things contributed.  First, the fuel pressure regulator was found to be hooked up backwards - the fuel inlet and outlet hoses were swapped.  This regulator is described by some as a "limiter", meant to limit fuel pressure to some maximum value at high RPM, where the engine-driven pump might otherwise try to push too much fuel.  Hooking the fuel lines up backwards would likely prevent the internal piston/diaphragm from receiving enough fuel line pressure to push it open and limit the pressure as intended.  That probably explains the excessively high fuel flows and low EGTs at maximum MAP/maximum RPM, and the hiccups on takeoff, but I'm not sure it explains the stumbles around 2200 RPM and below 20".

Second, the controller was found to have been commanding excessively rich mixtures at all power settings.  This would explain stumbling across the entire operating range, as well as the very large RPM and EGT changes when moving the mixture control from full rich to stoichiometric.  Previously the idle mixture rise was about 100 RPM, which turns out to be well outside of specification.  SID97-3G says it should be 25-50 RPM.

Continental SID97-3G reads "Engine performance, service life and reliability will be compromised if the engine's fuel injection system is neglected".  I can't find any indication in the logs that the fuel injection system has received a tune-up since last the major was done by the previous owner.

After getting the pressure regulator hooked up correctly and leaning out the controller high- and low-end settings, the idle mixture rise is now only 25 RPM.  The max power fuel flow, with the mixture set to full rich, has come down from 40 GPH to 35 GPH.  I do sometimes get a short stumble on takeoff at maximum power with the mixture set to full rich, but by backing out the mixture four full turns, I can bring the fuel flow down to 33 GPH, per the Rocket performance table, and I get no stumbles.  I get no stumbles anymore at any cruise setting; indeed, I get no stumbles above 18" MAP.  I will start to get minor stumbles below 18", nothing severe, but if I back the mixture out another turn or two, I can clean those up.

Reducing MAP down to 10-12" on short final will cause more stumbling, again nothing serious, though at that point I have made the runway and I'm not going to mess with the mixture any further.  If I close the throttle fully in the flare and keep it closed, the engine may sputter and die before I am off the runway, so that problem has come back or wasn't fully resolved with the idle speed change.  This remaining stumbling suggests that the low-end mixture control is still set too rich, so I'll be tweaking that with the continued help of my A&P, and we'll see if we can adjust the upper-end mixture down a bit more, too.  My current operating procedure is to back the mixture out four full turns before takeoff (which still keeps TIT below the 1450F 100% power limit), then in cruise and descent lean the mixture for 1550F TIT (below the 1600F limit), and leave it there through landing, keeping RPM above 800 once I'm on the ground.  A go-around would require pushing the mixture forward to keep EGT down, which is increased workload, so I'm not 100% satisfied yet.  I also need to go do some simulated go-arounds to confirm that I get good transient response when it's needed (I'll be doing those at altitude or at a very long runway, thank you.)

A few technical details are worth pointing out.  Ground run-ups are limited in terms of what they can tell you about how your mixtures are set.  The reason I couldn't reproduce this problem on the ground, especially the nerve-wracking stumbling down around 15", is because you can't achieve a 15"/2200 RPM operating point with zero airspeed.  But 15"/2200 RPM is super-important to get right - that's around where you're going to be on downwind.  This and some other important operating points are not achievable on the ground, except maybe transiently if you run up the RPM and then close the throttle abruptly (which will give me a stumble, BTW).  Even the full power operating point that you experience on takeoff is not the same as what you get doing a static max power test.  Following SID97-3 may be your best bet, but that procedure doesn't test the full MAPxRPM operating range you will experience in flight.

BTW, a little math shows that the 33 GPH max power fuel flow setting from the Rocket power table is right around 12.5:1 air:fuel mass ratio, though in reality it's probably richer than that due to volumetric efficiency.  Cycle-to-cycle variations and fuel dispersion will temporarily create ratios even richer.  10:1 is the approximate upper flammability limit for gasoline in air; you'll have problems igniting the mixture as you go richer than that.  So if your fuel flows are much above book, you're risking misfires.

It's hard to get the fuel flow math right each cycle when you are stuck using mechanical/hydraulic/pneumatic computers (i.e. fuel pumps, controllers and pressure regulators).  That's why automobiles switched to electronic fuel injection a while ago.  They also have sensors that can measure air:fuel ratio and adjust it to account for changes in the engine over time.

 

Edited January 1, 2019 by SpamPilot

[SpamPilot]

Oh, and LANCECASPER, thanks, I did call Rocket regarding the fuel pressure regulator, and they said there was only one guy left there who knew anything about the Rocket 305.  I didn't write down his name, but he seemed to be someone who had worked on assembling them and was able to at least confirm the part number I was looking for to the best of his memory.  I notice that none of the various links around the Web to Rocket Engineering's Rocket 305 data pages are working.

Edited January 1, 2019 by SpamPilot

[aviatoreb]

2 hours ago, SpamPilot said:

Oh, and LANCECASPER, thanks, I did call Rocket regarding the fuel pressure regulator, and they said there was only one guy left there who knew anything about the Rocket 305.  I didn't write down his name, but he seemed to be someone who had worked on assembling them and was able to at least confirm the part number I was looking for to the best of his memory.  I notice that none of the various links around the Web to Rocket Engineering's Rocket 305 data pages are working.

That’s a major bummer.

[MB65E]

Nice Report SP! 

On some of the TSIO 550 Legacy’s, we needed to tie them down for run ups as the brakes would not hold. The port-a-port test set up I used annually to adjust fuel flows and pressures. It was a pain, but if the engine was set to the book values, it ran really well. Each year you could see changes and trends for the engine. The fuel pumps would ware and cause issues, but only due to the fact we were at the limits of the system at race settings. I’d recommend setting it up exactly as the Rocket manual allows with the test gauges. Idle fuel flows and pressures are really important. 

The Rocket is pretty much a dream certified airplane combination in my book!! Borderline Experimental! 

-Matt

[FloridaMan]

On 1/1/2019 at 6:43 PM, SpamPilot said:

Oh, and LANCECASPER, thanks, I did call Rocket regarding the fuel pressure regulator, and they said there was only one guy left there who knew anything about the Rocket 305.  I didn't write down his name, but he seemed to be someone who had worked on assembling them and was able to at least confirm the part number I was looking for to the best of his memory.  I notice that none of the various links around the Web to Rocket Engineering's Rocket 305 data pages are working.

http://web.archive.org/web/20070218094511/http://www.rocketengineering.com:80/rocket.html

[SpamPilot]

Following up on this issue for the record...

Further tuning has dialed things in pretty nicely.  The stumbles are completely gone.  If I pay attention, I can tell the engine gets a bit rough below 15" MAP at full rich, but it's minor and easily cleaned up with a small mixture adjustment.  This article was very helpful in understanding how the Continental Continuous Flow Fuel Injection System works for the TSIO-520, but there are a few practical issues it doesn't mention.  My A&P and I have found that the low- and high-end adjustments are very sensitive.  On the low end adjustment, 1/6 of a turn was enough to go from too rich at idle to borderline too lean.  On the high end adjustment, 1/6 of a turn is worth about 1 gallon per hour fuel flow at max power.  With the high end dialed in to 33-34 GPH at max power, I'm getting TIT of around 1375 F (below the 1450 F limit at 100% power), with per-cylinder EGTs down around 1325 F.  Reducing MAP and adjusting RPM down to the 88% Climb, 80% Max Continuous Power, or 76% Max Cruise operating points, EGT rises even if I leave the mixture at full rich, though the fuel flow and TIT stays within range.  This probably explains why Rocket drivers report cooler climbs at 100% power, though you do need to keep IAS high to provide adequate airflow to manage CHT.  This past weekend I climbed out at 100% power/150 MPH IAS/25C OAT/1000 FPM, and my hottest cylinder stabilized at 395F, with oil temp reaching a steady 220F and oil pressure dropping to 40 psi.  That's all still in the green, but a tad hot for my comfort level.  I reduced power to 76% and kept my speed up, but the engine didn't cool down.  It wasn't until I reduced power to 55% and climbed at 150 MPH IAS at 500 FPM that the temps came down and oil pressure increased.  In level flight at 2300 RPM, the CHT and oil pressure needles remain comfortably in the middle of their green ranges at any cruise power setting, leaned to the recommended limit.

With the low-end adjustment set to the borderline-too-lean point, I would need to blip the primer every few seconds to keep the engine running on startup until it warmed up and could hold idle on its own.  A 1/12 turn of the adjuster screw cleaned this up.  I may still need an occasional blip from the primer, but actually the TSIO-520 Operator's Manual says you might need to do this.

Possibly contributing to some of the roughness, after only 400 hours on the motor, I suddenly couldn't pass a mag check.  Turns out that when the engine was remounted after its last overhaul, the mechanic routed one of the plug wires between the #2 cylinder head and the rear baffle plate, where it doesn't belong.  Screwing the baffle plate into position squashed the wire against the head.  Ultimately the insulation failed and I was getting a spark at the baffle instead of the plug.  My A&P and I installed a new harness, routed the wires correctly, and that issue is now resolved.

Incidentally, loaded to near max gross, I am not getting the 1450 FPM climb rate listed on the performance chart.  At 100% power and climbing at 150 MPH (130 knots), I am seeing 1100 FPM +- 100 FPM.  Also, I'm about 10 knots slower than book.  Don't read that as a complaint, I know book numbers are optimistic and probably developed during a cold winter's day in a thermal over Spokane, WA.

I worked out a 2300 RPM economy cruise performance schedule for my aircraft, in case anyone wants to compare.  At 10500 feet, 10C OAT (about 20C above standard atmosphere), all speeds true airspeed, leaned to 50F rich of peak:

30" / 2300 RPM / 17.0 GPH: 208 MPH
28" / 2300 RPM / 15.5 GPH: 200 MPH
26" / 2300 RPM / 14.0 GPH: 193 MPH
24" / 2300 RPM / 12.5 GPH: 180 MPH

(2300 RPM to respect CSB09-11)