Flying an IO360 equipped aircraft with fuel injection, up a coastline after an uneventful over land flight. Weather is:
METAR 131820Z 26009KT CAVOK 16/08 Q1013=
1700' QNH, less than 1NM over the water flying parallel with the coast. 90kts, 2200rpm 6gph. Notice a reduction in power, increase throttle - no response. Full throttle gives no improvement. Excercise Alternate Air - SUDDEN LOSS OF ALL POWER! Alt Air to closed, regain 1900-2000rpm. Fuel pump on, change tanks, cycle mags no change. Declare a PAN and limp home. Power gradually improves but is by no means back to full by landing 10mins later. (still 6gph at full throttle)
Aircraft to be inspected by engineers tomorrow. Initial thoughts are icing to the fuel injection system, and if I'd left the alt air on (hindsight - great eh?) it would have cleared it as carb heat would. Doing some research this evening shows that Alt Air helps with impact ice only and that according to GASCO you can't get icing with fuel injection.
Any ideas? (ps still had fuel in both tanks on landing - before anyone asks!)
Loss of power is the first clue, but we're not given enough information. Did you adjust mixture or check your CHT/EGT at the time? The first thing that springs to mind is an induction leakage. If you experience an induction leakage and lose power (as you will), you'll have an imbalanced mixture to one or more cylinders. Opening alternate air with warmer airflow will increase these effects and produce a rough engine.
A plugged injector, partially or completely, may or may not produce roughness. Very often the only indication will be an increase in indicated fuel flow, where fuel flow is measured by differential pressure and not actual flow (depending on the system in use, of course)...what appears to be using more fuel is actually using less, and is often mistaken for something else entirely.
I've seen alternate air doors fail and block the induction before, or fail to seat (some use a magnetic latch which fails, others have mechanical control all the way through their range, and yet others may be opened in flight, but cannot be closed or secured until back on the ground...again, depends on the installation). I worked for a corporate department years ago in which an aircraft was lost (and all aboard killed) after one engine failed, and when alternate air was applied on the second, the door failed and blocked the induction...causing the loss of the second engine.
I've also seen ducting collapse which produced alternating and intermittant roughness or failure. This is more common in semi-rigid ducting such as SCAT hose, but I've seen it in injected and carbureted installations. Pull the cowl and look at the hose, it appears intact. Under low pressure or certain operating conditions, however, it can collapse.
The most common source of induction leakage in injected installations is the flexible couplings in the induction, typically hoseclamped onto the metal portions extending from the cylinders. As these age and are affected by heat and the clamps, they tend to develop leaks around the clamps. The net effect is uneven airflow to one or more cylinders, and depending on the fuel metering system in use, this can greatly affect the mixture to one or more cylinders. This may or may not show up as an obvious EGT problem, depending on whether you're using single point or multi-point EGT measuring (single point hides a lot of problems), but you an certainly experience a rough engine and cylinder damage, especially if you're already operating close to peak at high power settings.
Some talk has been made of icing. While icing sounds good, this isn't a carburetor, and doesn't ice like a carburetor, and doesn't clear like a carburetor. The original poster seems to be applying carburetor heat theory, believing that if he had induction ice in an injected engine, then using alternate air would melt the ice and cause a rough running engine. The problem is that while carburetor heat puts warmer induction air through the carburetor venture (where the ice is located) and thus melts it or prevents it's formation, that isn't true of the injected engine. In the injected engine, alterate air opens a door which merely provides an alternate air path.
I don't believe the original poster differentiated between the Lycoming IO-360 or the Continental IO-360. Each uses a different fuel system and induction; the induction is part of the fuel system and is used to determine the amount of air metered past the throttle body; therefore the two are tied together, but the diagnosis and troubleshooting are different.
A possibility for the loss of power in the first place was a problem with the alternate air door. Airworthiness Directives apply to certain installations regarding the deterioration, deformation, or failure of alternate air door asemblies. Cracks and distortions in the alternate air door are not uncommon, along with failures of the door, or hinge assembly, or actuation assembly (often a simple rigid cable). Parts of the door assembly can break free, or the entire door, and block the induction causing a partial or complete power loss. In some cases where a power loss occurs, a change in power setting may alleviate the problem, but often not.
A mechanical failure of the filter element in the primary induction path is also a possibility, especially if this partially occluded the throttle plate. Icing that blocks induction on an injected engine is typically at the filter, upstream of the throttle plate, rather than on the throttle plate, and it's for this reason that alternate air is provided...not to melt the ice, but to bypass it completely. If the filter becomes blocked, it can be by passed. If the filter fails, however, then alternate air may be of little use, or may contribute to the problem, depending on where and how the blockage occured.
When going to alternate air, you're going to need to make a mixture change. If you're already operating in a condition close to the operating limits (overly rich or overly lean), then you're in a position to cause the engine to quit or run rough until a mixture change is made.
The original poster stated that one can't get icing with fuel injection...this isn't true, but the icing isn't quite the same as what one gets with carburetion, and it's more tolerant.
Some induction systems also use more than one source for alternate air. Typically a source may be provided upstream of the throttle plate, but one is often provided down stream of the throttle plate, too; this is often an automatic emergency door which may or may not reset, and takes away some throttle control because of it's location. Sometimes referred to as a "suck door," it operates based on differential pressure between the induction and ambient, and is generally magnetically held closed. It's sucked in, to provide airflow; various installations are used in different aircraft and not all aircraft utilize them. A failure of this door to the open position can also cause some of the problems described.
Another thing to consider in your preflighting and subsequent use of alternate air is what's near the alternate air opening...which is generally located within the cowling. I've seen bird nests show up there which can then be sucked into the induction and cause an unrecoverable failure when alternate air is used, as well as mouse nests or other debris...check "under the bonnet or hood" when you can. Your preflight can't be too thorough.
I've seen bird nests removed three and four times in one day from the same engine...they can build them that fast; I've seen them built in hot engines. That airplane you flew in and parked an hour or two ago while you were having lunch may be a completely different airplane when you get back to it. Check it thoroughly.
The main thing is that you got back to the ground safely. Hopefully when it occured you immediately maneuvered to a place where you could execute an immediate emergency off-field landing, and kept yourself in a position to do so until you made it back to the airport. Remember that it's never a matter of if you'll have an engine failure. Only a question of when.