Compression ratios

Remember, you have the option to blend with 100LL if you want to pull the octane up. The problem with pump gas is the variability of the Reid Vapor Pressure (RVP). Winter fuel has a high RVP to aide in vaporization of the fuel in cold weather. (It also happens that what they add to winter fuel is also cheaper than the rest of the components. Bonus for the refineries.) This is really only a problem when you have winter fuel left in the tanks and you try to fly on a hot day. The fuel pump at the engine is attached to a skinny fuel line that runs to the tank. The low pressure on the suction of the pump, aggravated by the small fuel line, causes vapor lock. Turning on the electric fuel pump at the engine only makes this worse. Moving the electric fuel pump to as close to the fuel tank as you can get it will fix the issue.
For reference, take a look at the Peterson STC for Piper Cherokees. It costs several thousand dollars instead of several hundred. It includes two fuel pumps, switch(es), and wiring in addition to the typical fuel tank filler hole stickers.
 
GlennAB1 said:
It is probably just a difference in engineering theory or terminology. Some turbine manufacturers use the term "surge" and some use "stall" for the same phenomenon.
Click to expand...
I am neither an engineer nor an A&P but I thought it was the surge of air which caused the blades to stall. Thus the surge bleed valve opens to relieve the pressure.
 
Some manufacturers only use one or the other term in their maintenance manual, for example; off idle surge, and cruise power stall. There are numerous causes for the phenomenon, could be bleed valve, variable stator vane, fuel control, internal FOD, etc.
 
GlennAB1 said:
For entertainment purpose. How many of us are in piston engine research and development?
Click to expand...
Used to do engine controls / diagnostics / emissions (for more years than I care to think about) starting back when the job involved carburetor jets...

There is about a 15% chance that you drove to work in a car with my algorithms under the hood.
 
That's cool! I'll respect your opinions. But, being old school, and knowing that "detonation" is still a common term in numerous industries, I'm not going to go out and attempt to set everyone straight.
 
GlennAB1 said:
For entertainment purpose. How many of us are in piston engine research and development?
Click to expand...

Not anymore, but used to be on our engines.
 
Capt. Geoffrey Thorpe said:
Except, knock is not just a "temperature required for ignition" phenomena.

It's related very much to time / temperature / pressure history along with fuel properties. If you can get the flame front to the walls fast enough the pre-combustion radicals don't have time to form and you don't get auto-ignition even if the temperature is the same. And, different fuel types break down at different rates - long straight chains tend to go faster than more compact arrangements - that's why different fuels have different "octane" ratings even though the ignition temperatures and burn rates are about the same.
Click to expand...
I implied it will happen, you said why. It will happen just a matter of when. remember this thread started as a compression question. when the compression brings the temp above ignition temp. you will get ignition, wether the cylinder's is ready or not. spontaneous combustion will occur. When a diesel raises the temp above combustion temps they then inject the fuel at the pre-determined degree, and combustion occurs. but in a 4 cycle engine the fuel is already in the cylinder, so if the compression caused heating causes spontaneous combustion prior to ignition timing, detonation will be the result.
 
GlennAB1 said:
For entertainment purpose. How many of us are in piston engine research and development?
Click to expand...
I spent 2 years working at Curtiss Wright working my way thru trade school, doing just that.
 
GlennAB1 said:
That's cool! I'll respect your opinions. But, being old school, and knowing that "detonation" is still a common term in numerous industries, I'm not going to go out and attempt to set everyone straight.
Click to expand...

Google can only take you so far.
 
Dan Thomas said:
Knock sensors have been around quite a while. They protect the engine when low-octane fuels are used, though I sometimes wonder how much money is being saved by buying regular grade fuel and having retarded timing to prevent knock and end up with less power for a given gallon of fuel, which means less mileage.

I think knock sensors are also part of the emissions control stuff.
Click to expand...

Not how it works in most designs.

A failed knock sensor on most vehicles that run on regular fuel still means that it indicates knock when there is none, and the fuel economy falls off significantly because the computer won't advance the timing.

Bumped the fuel economy of a rather boring engine (flat 4 - Subaru) by 2 MPG by replacing the knock sensor that was firing all the time. The engine was set up (like most) to become less efficient and burn more fuel than necessary until the computer could once again reliably push the timing up and NOT see a knock until it pushes too far. If it even gets there.

They're usually used as a limiter on timing advance, and the engine is usually far less efficient without them working properly.
 
Actually, this is not how it works on most designs. It's the other way round. The engine never advances ignition beyond the values in the ignition maps, but it retards it based on knock sensor signal.
A bad knock sensor will mean the engine will constantly run on retarded ignition and depends on severity of the assumed knock, pulled boost and enriched mixture. Actually, an incorrectly tightened sensor will do the same thing. They are very sensitive. I wish more manufacturers would've copied Saabs older Trionics with ion current measurement for knock.

On Bosch systems starting from earliest EGAS Motronics (ME1.3->), the used ignition map is based on camshaft timing (if engine has VVT), this map is the highest value the engine will ever use, and should never activate knock prevention methods. Low loads it is very far from knock threshold, it's used to keep cat temperature high. High loads the Bosch calibration specifications call minimum 2 degrees from knock (Bosch threshold for knock is 15bar amplitude in the pressure curve during the knock window).
On Siemens (starting from EMS2000 and going all the way up to MSS6x/MSD8x ECUs), the engine either has two sets of ignition maps, or a delta map, based on fuel quality detection which it does constantly (this is a stored adaptative value). Denso apparently does this same, but I don't work with Denso so I'm not 100% on the algo.

So, important distinction - ECU never advances ignition by itself until it hits knock. The mapped value is never exceeded, the mapped value is the target unless there's an issue reaching it (such as a broken knock sensor).

denverpilot said:
Not how it works in most designs.

A failed knock sensor on most vehicles that run on regular fuel still means that it indicates knock when there is none, and the fuel economy falls off significantly because the computer won't advance the timing.

Bumped the fuel economy of a rather boring engine (flat 4 - Subaru) by 2 MPG by replacing the knock sensor that was firing all the time. The engine was set up (like most) to become less efficient and burn more fuel than necessary until the computer could once again reliably push the timing up and NOT see a knock until it pushes too far. If it even gets there.

They're usually used as a limiter on timing advance, and the engine is usually far less efficient without them working properly.
Click to expand...
 
mtuomi said:
Actually, this is not how it works on most designs. It's the other way round. The engine never advances ignition beyond the values in the ignition maps, but it retards it based on knock sensor signal.
A bad knock sensor will mean the engine will constantly run on retarded ignition and depends on severity of the assumed knock, pulled boost and enriched mixture. Actually, an incorrectly tightened sensor will do the same thing. They are very sensitive. I wish more manufacturers would've copied Saabs older Trionics with ion current measurement for knock.

On Bosch systems starting from earliest EGAS Motronics (ME1.3->), the used ignition map is based on camshaft timing (if engine has VVT), this map is the highest value the engine will ever use, and should never activate knock prevention methods. Low loads it is very far from knock threshold, it's used to keep cat temperature high. High loads the Bosch calibration specifications call minimum 2 degrees from knock (Bosch threshold for knock is 15bar amplitude in the pressure curve during the knock window).
On Siemens (starting from EMS2000 and going all the way up to MSS6x/MSD8x ECUs), the engine either has two sets of ignition maps, or a delta map, based on fuel quality detection which it does constantly (this is a stored adaptative value). Denso apparently does this same, but I don't work with Denso so I'm not 100% on the algo.

So, important distinction - ECU never advances ignition by itself until it hits knock. The mapped value is never exceeded, the mapped value is the target unless there's an issue reaching it (such as a broken knock sensor).
Click to expand...

Definitely not how this particular Japanese design is working in the real world, but then again, it's 16 years old, not fuel injected, and sans even so much as a CAT sensor.

You can watch the timing advance in real time on the OBD port and see that none of the above is what it's doing, either before or after the knock sensor fix.

And you could smell it at the tailpipe when the sensor was firing when it shouldn't have.
 
The old electronically controlled carbs/single point fuel injection systems are different, true. I don't know much about them, analogue knock control systems aren't my forte. But any modern design (by modern I mean MPI and ignition controlled by the ECU, with at least single lambda feedback loop for fuel control) works pretty much the way I described - mapped value which is then lowered if needed.

denverpilot said:
Definitely not how this particular Japanese design is working in the real world, but then again, it's 16 years old, not fuel injected, and sans even so much as a CAT sensor.

You can watch the timing advance in real time on the OBD port and see that none of the above is what it's doing, either before or after the knock sensor fix.

And you could smell it at the tailpipe when the sensor was firing when it shouldn't have.
Click to expand...
 
mtuomi said:
The old electronically controlled carbs/single point fuel injection systems are different, true. I don't know much about them, analogue knock control systems aren't my forte. But any modern design (by modern I mean MPI and ignition controlled by the ECU, with at least single lambda feedback loop for fuel control) works pretty much the way I described - mapped value which is then lowered if needed.
Click to expand...

Yeah this one's the ancient Subaru single port into a throttle body. I don't drive "them fancy cars" 'round here. Heh. Knock sensor croaked and you could about suffocate on the unburnt fuel vapors at cold idle. Haha.

The newest engine around here is a big fat stupid Ford Triton 5.8L on Karen's bling truck (Lincoln LT). Otherwise known as "Ford's Spark Plug **** Up They Mostly Got Away With". Well, other than that whole Class Action lawsuit thing.

Nothing newer than a Chevy small V8 or the 5.9L Cummins on "my" vehicles. (She drives the Subaru from time to time but rarely the Yukon and never the Ram 3500.)
 
You must log in or register to reply here.
Back
Top