Saabscene Saab Forum - Saab Technical Information Resource banner
1 - 11 of 11 Posts

·
Registered
Joined
·
1,448 Posts
Discussion Starter · #1 ·
A little piece of turbo tuning explained.

This is based off information I have gathered from a few recently discovered sources I've read. It's really great stuff.

I'll start it off with EGT.

EGT is actually something very important with turbo engines in regards to tuning as well. It's very closely linked to total power.

People found early on that on a turbo engine, your best power output will be found by raising the boost until it knocks, pulling back timing, raising it some more to compensate for the timing until it knocks, pulling it back some more ... all the while waiting to make sure the EGT doesn't go higher than your parts can take.

That sounds contrary to how street engines are tuned generally, but for race engines finding peak power eventually comes down to a ballance between ignition advance and EGT. Too much ignition retard and your EGT goes sky high, too little and it knocks. Adding more agressive camshafts is similar to raising the boost.

So you find your best power at very high boost with lots of ignition retard. Turbocharging engineers refer to this as "fattening the curve" as you are increasing the length of the pressure time rather than increasing the peak pressure. Thus doing more work, even if not actually creating more pressure. Since you aren't creating more pressure or temperature close to TDC you aren't knocking, but you're still getting more horsepower. The limit is when your parts get too hot. Shiv from Vishnu knew this. On his Evo he runs plenty of boost/flow but with only 11 degrees of ignition advance. He was able to get somewhere around 440 wheel hp on 91 octane using this method. Though for obvious reasons I don't think he will be telling us what kind of ungodly EGT his car produces.

People found quickly that exhaust valves were one of the "weak links" and even the old turbo Corvairs had Nimonic (similar to Inconel) exhaust valves. My Saab uses Nimonic as well, and the SRT-4 uses Inconel.

Changing the exhaust valves, turbo housing, turbine, and exhaust manifold/header to a high temperature alloy can significantly increase the life of the engine when tuned in this manner, it can also increase the total peak power by allowing more agressive tuning.

For what it's worth on Saabs and other expensive European turbo cars the general tuning "limit" is around 1750-1850 degrees because of the expensive materials used. Some have seen 2000 without damage, though only for relatively short periods.

Ceramic coating the valves, pistons, combustion chambers, exhaust manifolds, turbines, and housings can increase the life of those parts in many cases. That would allow higher EGT than normal, and subsequently, more power. Though only after tuning.



Let's get some friendly discussion going.

Dubbya~
 

·
Registered
Joined
·
234 Posts
Very interesting, and nicely summarised.
In a turbo engine limited by knock, I'd always wondered whether it would be better to trade a little advance for a little more boost or vice versa.

I don't know about the later cars, but the very early 9000 turbo's have a vacuum advance/pressure retard unit which is supposed to retard the ignition under boost. I'm not convinced mine does, but that's another story

Looks like SAAB made the decision to go down the 'fat curve' route (or maybe just slightly plump ), and presumably had suitable exhaust valves to cope with the EGT's.

Cheers,
Alan
 

·
Registered
Joined
·
136 Posts
Good info. I have an EGT gauge and I find it to be my most important gauge. Unfortunately the gauge is after the turbo but even at that location it gets close to its maximum reading. The highest temps I have hit is 1500F or 815C. But the temps out of the engine are a few hundred degrees more. I was able to talk to Dr. Boost a while ago, he was one of the designers of the Saab Trionic system, and I asked him about EGTs. He said to keep it under 2000 F, 1100C, and I would be fine. Most casual driving is around 1200F and it takes full boost through three gears to his 1500F.
 

·
Registered
Joined
·
1,448 Posts
Discussion Starter · #4 ·
2000 degrees F might be a bit excessive for long periods. I think that's why many of the european tuners don't tune as agressively. If it were to run that hot on the Autobahn for long perdiods it would only be a minute or two before a valve melted.

Keep in mind the melting point of forged steel is generally just above 2200 degrees F, and as you get the valves close to that temperature they will weaken significantly. The maximum service temperature (intermittant service) is listed as about 1700F. Keeping below that should be safe for the valves, though the turbocharger is another story.

Short periods of above 1700 could be safe if they are sufficiently short.

For those with Viggens, Aeros, or MY2001+ 2.3L you'll have Nimonic valves rather than stainless. They list a melting point (solidus) of 2340 degres F, and do no list a maximum service temperature. Presumably, as none of the high heat superalloys list maximum service temps, it's fairly close to the melting point.

Another aspect of tuning that I've uncovered is a/f ratio. While richening it up does cool the gasses slightly to prevent knock. First some information on knock needs to be made more clear.

Firstly knock is not how most understand it. Firstly it's not the sudden auto-ignition of the remaining air/fuel mixture due to heat. That often accompanies knock, but that isn't the origin of the explosion that causes the pressure wave through the cyllinder.

Many SAE papers have been published saying otherwise, but it took NASA (at the time named NACA for National Advisory Committee for Aeronautics) to actually make a camera with sufficient frame rate to detect the true culprit.

Suffice to say one of the causes of knock is actually a spontaneous exothermic decomposition of CO2 ---> C0 that takes place due to the lack of oxygen present in the cyllinder and the presence of large amounts of hydrogen. High speed ultraviolet spectra of detonation processes indicate a large quantity of OH radicals immediately after the reaction. Indicating that possibly the reaction H + CO2 ---> CO + HO is spontaneous at combustion temperatures. This is partially re-affirmed by the tendancy to have excess CO in the exhaust emmissions when detonation has been present. Generally the reaction of 2CO + O2 --> 2CO2 occurs late in the combustion process, but since detonation always occurs on the trailing edge of the flame front there may be sufficient CO2 for the previous reaction to be significantly exothermic and forcefull.

This brings us to a/f ratio.

While richening up the mixture may cool combustion slightly, it also exacerbates the cause of detonation, which is essentially a lack of oxygen in later combustion processes.

Water injection has proved to help this slightly. It's generally considered more effective than richening up the miture, albeit requiring more effort to make reliable. This would seem to be because it does not exacerbate an already potentially damaging problem. But because the oxygen atoms in water are tightly bonded, it doesn't "help" the problem either.

I would theorize, that there is a possiblity that injecting a solution of water which has a small content of H2O2 (hydrogen peroxide) might provide enough cooling and oxygen to help stiffle the processes that lead to detonation. However it might be just as effective to run slightly leaner with a very good injection system of just water.

It would obviously take significant experimentation to find an ideal solution, and for most people richening the mixture works just fine. But this information may lead to som new tuning ideas. At the very least, injecting water instead of fuel saves the owner money as water is considerably cheaper, and less needs to be used as it cools 8 times better.

Food for thought.

Dubbya~
 

·
Registered
Joined
·
1,448 Posts
Discussion Starter · #5 ·
Bring in some more up to date information on Detonation Theory.

An automotive engineer and I are having a go at trying to properly explain the physics and chemistry behind it. Most SAE papers on it are either blatently incorrect, or misleading. Even some of the newer papers just don't fully adress it. I'm not entirely certain why, but we're trying to get to the bottom of it.

This is where we've gotten:

I had just looked up some reaction enthalpies for our suspect reaction as the culprit for detonation ...

When formed from basic constituents, CO has an enthalpy of -110.5 (kJ/mol), whereas H2O has an enthalpy of -241.8 (kJ/mol). That would leave around 130.3 kJ/mol of free energy.

Here's my theory:

Most of your gasloline is composed of hydrocarbons (Hydrogen and carbon), and in the early stages of combustion when there is a great deal of excess oxygen in the cyllinder both the hydrogen and the carbon get fairly equal shares of it. To be more specific, this even sharing would occur at the front of the flame wave where there is ample oxygen.

As the wave passes through a region thus depleting the oxygen and moves on the CO hasn't burned into CO2 yet (this reaction is known to take a while) and some of the H2 hasn't burnt into H2O yet (mixture is rich, and hydrocarbons are mostly hydrogen) ... so then suddenly the H2 molecules and/or individual hydrogen atoms suddenly claim oxygen atoms from the unburnt CO. This would release energy, but in order to do so quickly enough to detonate, it would need to be mostly H atoms, which would quickly form OH radicals (explaining the ultraviolet spectroscopic evidence of OH radicals), then form into H2O and create a great deal of energy. It would also explain why detonation appears to occur towards the end of the flame front, but why it also must be a place where combustion is not yet complete.

The excess CO found in the exhaust of a detonating cyllinder could simply be explained by incomplete combustion of the remaining mixture.

I think you might also explain the sensitivity to detonation as the relative rate at which the carbons vs hydrogens got ahold of the oxygen.

In chemicals like Methanol, Ethanol, and Toluene, the carbons will be the last to get any oxygen as they are completely surrounded by hydrogen in a physical sense. Since they carbons don't get as much oxygen prematurely, less oxygen is "stolen back" later on in the process, and voila: less sensitive to detonation.

The temperature difference may simply mean that at high temperatures there is a better chance that the carbon and hydrogen will ge equal shares DESPITE physical arrangement because the mollecules that are long (like octane) are easily broken, whereas short mollecules like Methanol and Ethanol are short and not easily broken. Same with Toluene. So short mollecules burn off the hydrogen quickly, then if there is sufficient air, burn the carbon too, which reduces the chances for detonation later on as hydrogen would steal the oxygen it eventually needs.

The physical shape would only make a small difference to the combustion rate, but in testing it only makes a small (relatively speaking) difference to detonation sensitivity. It seems to all be coming together nicely.

Dubbya~
 

·
Registered
Joined
·
1,219 Posts
ok..so what you really need is a short burst of air (or O2) just after combustion has started, to raise the O2 levels bihind the flame front... highly unlikely that it could be done...
 

·
Registered
Joined
·
1,448 Posts
Discussion Starter · #7 ·
Originally posted by Grentarc:
[qb]ok..so what you really need is a short burst of air (or O2) just after combustion has started, to raise the O2 levels bihind the flame front...   highly unlikely that it could be done...    
  [/qb][/b]
The next wave on engine technology will likely be direction injected gasoline engines. (Like Diesel is now) If it's done the way diesel is, you could controll when the fuel enters the cyllinder and force it to always burn at the proper rate. That would alleviate even the chance for detonation.

Another couple of possiblities have to do with water injection. Reducing how rich the mixture is, and running a water injection system instead has been shown to speed up the reaction of CO burning into CO2. Once burnt into CO2 it is no longer going to react with the hydrogen and detonate.

Dubbya~
 

·
Registered
Joined
·
1,448 Posts
Discussion Starter · #8 ·
Just thinking about that direct injection, and controlled puff of air combined.

You could program a direct injector to inject a puff of air just after the flame front leaves the injector. It might reduce the chances for detonation.

hmmm ...

Dubbya~
 

·
Registered
Joined
·
1,448 Posts
Discussion Starter · #10 ·
Audi is making a production direct injection gasoline engine. I have seen the Saab setup. Came across it a number of years ago. It would be pretty neat, but you know GM.


Dubbya~
 

·
Registered
Joined
·
1,219 Posts
just looking at a page describing saab's variable compression engine.. 1.6 Ltr supercharged (peak 40psi) 5cyl 225hp engine... has 14:1 - 8:1 compression ratio engine... that coupled with direct injection would be a nice setup. (love that 40psi!!)
 
1 - 11 of 11 Posts
Top