org

Some of you may remember my previous posts regarding use of a LS type engine in an aircraft, specifically a P51 replica. This requires the use of short individual stacks about 12 inches long. I realize this isn't the best performance wise, but it's what I need.

To me it appears there's no way to run O2 sensors on individual short stacks due to them being too close to the end of the pipe and outside air. Anyone have other thoughts? I'm not worried about sensing all cylinders, since I can pick the leanest and just assume the rest are OK for my purposes, so the accuracy of the sensor is the limiting factor.

Any thoughts? It looks like speed density for me unless there's a way to install the O2 sensors to a couple of the pipes.

Thanks.

jimmyblue

iTrader: (11)

You could tune with a test pipe perhaps and run open
loop, MAF or SD, with the short pipes normally.

66351coupe

carb it

org

Which way would the mixture go when the test pipe was removed, rich or lean? Also, how far from the port does the O2 sensor need to be and how far from the end, roughly?

I've thought about carbs, and there are several reasons the carb is attractive, but the difficulty in compensating for altitude changes makes FI more practical.

Thanks all for the replies.

org

Since my original post I've been searching the net and as usual I've found just about every opinion possible. Anyway, a possible solution is to use an offset for the bung to recess it slightly out of the main exhaust stream so I can place it fairly close to the port on two of the pipes. The tubes will be around 14" long plus a little and from what I've found, I can put the sensors about 10" from the port and still get useful signals during above idle operation. Since I'm not so concerned with the normal driveability issues found on a car, I think that's what I'll try. Since the sensors won't be in a close enclosure the recess should keep them from burning.

Worth a try anyway.

Dan Stewart

I'm with jimmy on this one. Use the test pipe to dial in your fueling, then lock it down in open loop and run speed density. This eliminates the possibility of stalling due to a faulty O2 or MAF sensor. Stalling at altitude would probably suck, so anything you can do to failsafe the engine from electrical problems, the better.

Dan Stewart

If you kept the O2 sensors in place, they would read a lean condition and this would cause the PCM to start dumping more fuel in. You would, in actuality be running rich. That is until the O2 sensors burned up from the skyrocketing exhaust temps from the extra O2 being pulled in by the reversions from the open portion of the exhaust pipe.

How far to mount the sensors? You need to calculate the venturi effect of the air flowing across the open portion of the exhaust based on your max airspeed. The faster your airspeed, the greater the venturi effect will be on the reversions and the farther back up the exhaust pipe the reversion will travel. I suggest looking in a copy of the Instrument Engineer's Handbook, Process Measurement and Analysis by Liptak. It will have all the info you need to calculate the distance to mount the sensor.

But, I highly suggest Jimmy's advice of open loop speed density for safety.

1bad lad

run a maf on it. the ecu only takes a baro reading when u first turn the key on. then fuel calculations are based off that. using a maf will be more accurate with the elevation changes i would think. put the o2 sensors as close the the head as possible. what kind of rpm r u lookin at spinnin it??? i wouldn't run it over 5500 for long periods of time. you'll wear out that engine so quick.

org

Actually the SD has always sounded best to me, but I was concerned with the initial settings. What you say sounds correct and I'll do some more thinking. The airspeed will be about 220 or so and the pipes are angled out of the cowling so the opening is about 3" outside the cowl, so there would be a fair amount of venture effect.

I have a 6.0 Vortec, but with prices dropping on LS3s and l92 crate engines they become more attractive. The GM ECM/harness setup for the LS3 looks good and easily installed. Before anyone says "Speartech" I already contacted him a year or so ago and he won't do anything that starts with "airplane."

Thanks for the thoughts. I've found this board to be invaluable.

SSAUTOCHROME

iTrader: (2)

How the F* are you going to get approval of putting that thing in the air again? Have you considered if you run Speed Density tuned and say the weather outside changes drastically... If the engine is not tuned for these changes (SD really can't keep up with them, hence a MAF) you engine could run like a$$ and possibly turn into a front end anchor!

org

SD will do fine if the fuel tables are correct. All the engine cares about is the manifold pressure (MAP) temperature of the air, and the rpm. If the tables are correct for those, it will run fine. Aircraft engines have been running for years simply by setting rpm, manifold pressure, and manually leaning. The ECM will take care of the leaning, the throttle takes care of MP, and the prop controls the rpm. My concern is for getting the initial setup right, not whether it will work. Homebuilt (called Experimental) aircraft with car engines number in the hundreds and have been flying for years. There are quite a few LSx powered airplanes, but only a few with short stacks. That tends to make quite a difference in setting it up.

The rpm range will be around 4800 for takeoff, 4000 for climb, and 3400 or so for cruise. LSx boat engines have been running a long time at similar power levels. The prop shaft reduction unit is a ratio of 2.43 to 1. you can see a similar setup at www.legendaryaircraft.com .

jimmyblue

iTrader: (11)

I guess I'd favor the MAF approach because this is a platform
that will see significant pressure and temperature variations.
The temperature is especially problematic, just see how much
IAT relocation seems to affect some behaviors. A MAF based
system pretty much doesn't care about anything, so long as
you calibrate it right. Which could be as simple as stock for
stock.

Because of the BARO = MAP@key-on deal, you might want
to make your fuel system a return type and manifold referenced
so you keep a constant pressure across the injectors regardless
of altitude. Otherwise the assumed BARO value will be off as
soon as you climb.

org

OK, let's see if my understanding is close to correct: If (on a stock engine, GMPP engine controller) I use a test pipe for each of the O2 sensors, plug everything in, and it should start. As long as the O2 sensor pipes are in place it would run normally. By taking the O2 pipes off, the PCM would revert to SD EXCEPT that the MAF would still input information on the actual mass of air entering the engine. I would get a warning light that the O2 sensors weren't operating and the PCM would manage the engine based on the (primary?) VE tables, but with reference from the MAF, which would provide more accurate information than without it.

On the other hand, same engine, same controller, initially startup with O2 pipes installed, then unplug the MAF. The PCM has referenced the MAF and O2 sensors and basically stores the information on the primary VE tables. After unplugging/removing the MAF and O2s it would revert to the secondary tables since there would be no input from either the MAF or O2 sensors? Is this basically correct?

In either case, the various warning lights would need to be shut off in the PCM.

I'm trying to figure out a way to run the engine in an unusual situation, I realize. I won't need to worry greatly about off idle response as long as the engine accellerates smoothly into the power producing band, where it will live for fairly long periods without idling. As long as I can count on it not loading up and quitting, and it answers the throttle smoothly, it doesn't have to have the response we expect on the road. I do need it to run well in the power band and smoothly get into the power band. It needs to compensate for barometric changes, temperature, load, and rpm. It needs to run at an efficient fuel mixture, if anything slightly on the rich side to avoid detonation, which will be hard to hear in this installation. I'm not sure how knock sensors will work with the additional vibration from a prop and gearbox, but I want to use them if they'll work.

All this would be easy if it was possible to use full time O2 sensors, but it appears that would be difficult.

Dan Stewart

Jimmy, have we verified that MAP is actually trimmed by key on BARO when in SD?

I know redhardsupra doubted this on the HPTuner forum, but I never read anything else after he stated he wanted to verify this. Aside from seeing this on Dodges, I haven't seen this in any of the GM pcm's yet (although, I don't have the opportunity to change my altitude by several thousand feet in one key on cycle), although I have seen MAP trimmed by the pcm mounted BARO sensor on the staged upgrades for the Cobalts.

If it stands true, then I have to agree and say keep the MAF in this case. If not, I still say go SD with it.

jimmyblue

iTrader: (11)

Dunno about trimming, but the BARO has to come from
somewhere and BARO-MAP (vacuum) is used to index the
injector flow table. It's my belief that BARO is a one-time
reading although it could be taken from any WOT, low to
mid RPM event.

Org: the test pipe is just a way to house an O2 sensor.
If you don't have the sensor there full time then you are
going to run open loop. Speed-density, MAF, alpha-N are
all different means of obtaining an air mass value and
can be used in open- or closed-loop systems.

org

Thanks, I understand about the test pipe. I'm wondering if the MAF has a role after the O2 sensors are removed, and what it is. What I've read before indicates that the removal of the MAF allows the use of MAP and temperature for the ve table to calculate density of the air entering the engine and by combining density with rpm (the speed of the air pump) the mass. Then the VE table tells the injection system how much fuel should be required for the given combination of MAP/temperature/rpm. The engine really doesn't care what the conditions are outside (except for the reduced exhaust backpressure at higher altitudes), only inside the manifold. Can the engine run in closed loop without O2 sensors and does the MAF actually do anything after startup without them?

Dan Stewart

Bingo!

I could have sworn that the whole "Massflow = (VE * MAP * RPM * Displacement) / IAT" formula would have been worthless in SD if this wasn't so.

But, I think tuning for SD and running a MAF sensor would probably work even better. You have the MAF making constant airflow corrections and you have a VE table that keep the engine from stalling if the MAF fails.

org

This being the case, the situation starts to look better. How far wrong would I be in thinking I could do the initial startup and some power runs (controllable props are great dynos) with the test pipes installed, then simply remove them and the O2 sensors, causing the engine to revert to basic SD but with the MAF still functioning? What would need to be reprogrammed, assuming I use a GMPP engine controller?

Thanks.