Quote:
Originally Posted by ECONORAM
I had thought about altering my truck's air/fuel ratio, but your post makes me wonder if I would melt something down. With coated pistons and combustion chamber portions of cylinder heads (I am guessing both should be done), will this prevent this from occurring since the heat is reflected back into the combustion chamber?
|
I'm sorry, but as a long time lurker I thought I would chime in one this one, because lean and rich are sometimes used as relative terms ("too lean...") and that can be easily misunderstood in terms of what is good and bad for the engine.
A modern car spends much of the time running at stoichiometric ratio, or 'stoich'. This is nominally an AFR (air/fuel ratio) of about 14.7:1, but it varies with exact fuel mix. The vehicle doesn't spend all its time here, that's why OBD-II 'MPG' readings have to be fudged so much to match actual fuel usage, but it spends a lot of it.
Stoich is the optimum *thermal* reaction. It is the balance of air and fuel that releases the most heat. So, it also coincides with peak EGT. The only reason vehicles spend so much time running here is emissions. Stoich is a point where CO, HC, and NOx are all relatively low. If you go richer, CO and HC soar, if you go leaner, NOx soars. The peak exhaust gas temps are also needed for the catalytic converter to operate.
The car normally finds stoich using O2 sensors. Based on the name, you'd think they measure AFR, but they actually measure 'equivalency ratio', or how much O2 must be added or removed from a sample of the exhaust gas for stoich. Normally, we use the reciprocal of this and call it "lambda". So, stoich is 1.0. A smaller number is richer, a larger number is leaner. To get an approximate AFR, multiply lambda by the typical stoichiometric ratio for the fuel being used. So, with gasoline, that would be lambda x 14.7.
Stoich is good for emissions, but not for everything else. It is reasonable to think that peak thermal reaction is best for power, but the engine is not a steam engine. We are looking for the optimum combination of gas pressure inside the chamber, transferring optimum mechanical advantage. In other words, we aren't looking for the most heat from the fuel, but the highest mechanical advantage.
This is a point richer than stoich. It varies with engine design, but let's say .86 lambda, or roughly there about, on average. Although it is not a precise technical term, this is sometimes referred to as "Best Power". If we were looking at EGT temps, this would be roughly 125 to 150 degrees 'rich of peak' or 'enrichened until the temp is 125-150 degrees cooler than peak'.
Similarly, best economy is also not at stoich. That point is leaner than stoich, say 1.05 lambda, perhaps '25-75 degrees lean of peak' (lean until 25-75 cooler). The reasons are actually surprisingly complicated, but suffice to say that excess air helps insure that all the fuel is used and it is a point that still yields good mechanical advantage.
Sorry to be so long winded, but I wanted to make sure that my use of 'lean' and 'rich' that follows is clear.
There is a long standing MYTH that running an engine lean of stoich, or lean of peak EGT, is bad for it. It is clearly, and utterly a myth. Lindbergh disproved it in WW-II with P-38s, and helped turn the war in the Pacific. And we have seen it analyzed again, and again, and again, ever since. You are not going to detonate without heat, fuel, and pressure, and lean of peak, or lean of stoich, all three are at their lowest.
Likewise, you aren't going to burn anything up without heat, and CHT plummets lean of peak. The problems with running this lean are a) it is hard to keep the fuel distribution even to the cyls with such small amounts, so you can get rough engine operation and b) your emissions soar.
There is also a widespread believe that engines can't really run at Best Power because of heat. But thermal energy from the fuel goes basically three places, exhaust, cyl heads, and 'work'. At best power, the engine is doing a good job at 'work', so it is probably not surprising that both CHT and EGT are relatively low. The problems with running here guzzling fuel, and mechanical torques and stresses. Again, one of the issues is just how evenly you can distribute fuel.
The problem area for burning things up is just rich of peak, between stoichiometric and Best Power. This is where peak pressures and CHT's soar. In other words, this is when everything detonation needs is in plentiful supply.
So, when we are talking about the consequences of 'leaning out the engine', we need to be clear about which types of operations we are leaning. If we are leaning out 'closed loop' operation, or when the engine is running at 14.7:1, our primary concern is the environment. We can get so lean that we run rough, but you'll hear it.
It is possible to put in a wideband sensor and send a simulated narrow band O2 signal to an ECU and fool it into running leaner, but I think that this is counterproductive and a waste of time. You get into issues with long and short fuel trims, simulating cat efficiency, etc. and you ruin your cat while being bad for the environment. I think a more useful application is to use the wideband to 'tighten' the ECU's adherence to stoich. ECU's chase stoich, rich, lean, rich, lean. Following the narrow band sensors. The sensors have a little curve in their response, so the oscillations can get wide.
With a wideband (at least a fast one), you can simulate the speed of the response for the 'narrow band' simulation however you want. I've found that by adjusting the simulated angle of the narrow band, I can simulate a faster narrow band that the ECU will track. The swings are still there (they need to be for the cat to operate), but they hover closer to stoich. This can save a little fuel, but more importantly, it seems to lower my emissions when I do a test smog check. Cat efficiency is very dependent on peak temp. It plummets as we move off stoich. So, by fooling my car to run tighter to stoich, I can keep my cat operating at a higher efficiency. I could get it tighter if I could modify the control loop parameters in the ECU, but that is a whole different level of intervention.
If we are talking about leaning out things when we are 'open loop', or when the vehicle ECU is going rich because we are stomping on the gas to pass, etc., then going 'too lean' (being caught between stoich and best power) is a real concern. Although we are seeing more wideband technology in stock vehicles, most cars still just pour on a guess in excess fuel. And, to be safe, they generally error on the rich side.
From an eco point of view, I think that the answer is the same as in performance automotive. Tuning to as close to best power, closed loop with an add on wideband if you can, seems to make the most sense. Think about it, you are stomping on the gas for a reason. Tuning for max performance in these cases means that you will get the needed performance result down as quickly as possible, while using less fuel than a stock tune. Tuning any leaner than this is counter productive. You'll have to stomp harder and longer to get whatever you want done. It seems better to me to get the performance when you need it, and get back into closed loop operation as quickly as you can.
Sorry for the long post, but it is sort of 'my area'.
Regards,
-jjf