Compression Ratio vs BOOST - Knock Control
 

http://forums.pelicanparts.com/uploa...1205010265.jpg

I copied a version of the Porsche chart above on the ADVANTAGES OF TWIN PLUGGING thread. The more I studied it for my purposes the more I realized it needed it's own string. (Young guys, bare with me it's old school technical presentation)

I expanded the Porsche chart to higher boost pressure and higher fixed compression ratio. To this I added four starred data points for well know factory engines. There are two factory TURBO engines (middle/left) and two N. A. (lower right).

I added data points from three member's engines representing serious modded street TURBO motors and I added Porsche racing motors (upper/left). These points also show features, C.R., BOOST(or not), intercooled (IC), plugging - single(S/P) or twin (T/P)

Please take the time to understand the chart and the trends it shows. I think it goes a long way toward demystifying relationships and component options as they relate to staying within the confines of stable combustion on the air cooled 911 based motor, N.A or TURBO.

This chart does not predict HORSEPOWER.

The chart doesn't include the intercooler used, and the intercooler cfm airflow or outlet air temperature either... thats very important when figuring "Compression Ratio vs BOOST - Knock Control"

Compressing air quickly heats it up in a big way and that increases the chance of detonation.

So if you compress the air to around 14psi in the turbo and then cool it off good in the intercooler before compressing it very quickly some more in the cylinders, that helps alot for knock control or detonation.

Thanks great mapping idea. That indicates my Kremer car which boost to 1.4 bar must be built to the 6.5 of the early 934 935 as it does not detonate with 108 octane and has almost 10K miles on the original build. It has long been my suspicion that Porsche from the early baby 935 and onwards built low compression high boost motors as boost is adjustable for altitude, humidity longivity, etc and compression is fixed. My car was built in 84 and kremer had been king on tracks for many years by then. I do not know if the race cars were still 6.5 in 1984. Andial was competing closely with Kremer by then and my motor from them uses the 1974 factory and 84 Kremer low compression philosophy.

All true, getting into that level of detail would make the chart very busy but you do have a very good reading on the impact of adding a charge cooler when you compare the two factory TURBO motors. Apparently the factory's testing of their basic intercooler allowed a .75 C.R. increase.
In their paper they identified this addition upped the power and improved fuel economy 2-3%.
Additionally, if you look at the factory TURBO racers you see a two point C.R. increase with their large, highly efficient intercoolers plus twin plugging.

I placed the point using data from one of your postings. You know, assuming it isn't 6.5:1 and your C.R. is actually 7:1, it still places you right in line with Don and Les' cars. 12PLUS to 1 effective compression. That may be on the edge but you guys must be getting along with it.
Looks like it may be time for one of you three to pull out a couple more enabling technologies like water/methanol injection or knock sensors and set a higher point.
Well, Don you have a higher point out there at 13:1 but I think you told me you don't run that level.

I'm missing something; the numbers seem high to me. Please explain the math behind this chart. Surely a stock 930 does not run a corrected 12:1 C/R.
When I crunch the numbers 7.0 C/R @ 0.8bar = 10.9:1 effective C/R.

Manfred Kremers build spec sheet lists the HP @ 480 so that seems about right in line with the chart.

Yes, you are close enough for the point of the chart. Looking at the curved lines of effective C.R. the Factory 3.3L TURBO falls at 10.7:1 and the 3.0L. at 10:1

Craig 930 RS just PM'd to report he is running 7.5:1 and .9 Bar with twin plug. This drops right on the 12:1 effective C.R. and continues to support the chart for component selection and recommended boost level.

Thanks Craig.

No, I don't go above .8. I can't tell the difference between 500 and 525 hp driving around, but as your chart shows, it seems the engine knows. However, when I built my car, I built it for drivability and for me, that's torque and throttle response. The torque curves for my motor, whether its 525 or 500 hp, are nearly identical, so I see no need for 1 bar.

Hmm...I know of 7.5:1 C/R 3.6 twin plugged turbo converted motor getting dynoed to 1003 HP at 2.0 bar w/ race fuel...

This graph seems to be somewhat old....

Goran, there is no claim to HP on this chart however the only thing that is odd is the engine you just mentioned. Is the race fuel methanol? Please get more details, street?

While I find it an interesting chart, it is a little to theoretical. (sorry I got long winded)


One thing the chart does not account for dynamic compression, which will radically alter the results. Now I did a few numbers and here are how some things compare.

http://forums.pelicanparts.com/uploa...1205040775.jpg


Now you can see the differences in compression ratio, piston speed, and intake valve dynamics. All pistons speeds were calculated at 7000rpm except for the Baby 935 which was at 8000.

Now the 3.0 street engine will have a much higher dynamic compression ratio because of the camshaft timing than say the 935, but the 3.6 Turbo will probably have about the same dynamic compression with higher boost levels.

Now comparing boost between street motors and race motors is pretty much useless. Even comparing it between different generation street motors, say 3.0-3.6 is relatively useless. Because changes in camshaft timing, piston speed, compression ratio, intercooler design, intake design, exhaust manifolding turbo design, etc will all alter the measured restriction in the intake path (boost level). Pressure ratios would actually be a better way of comparing street engines, but even that is kinda irrelevant, except to show how improvements were made as Porsche learned more about turbo charging.

I suppose on a theoretical level you could perhaps calculate effective pressure, but it will be different in reality, because, for example the K27 will flow more airflow at the same boost levels than say a K26. So in the case of VoitureLTD he will have a different mean and maximum cylinder pressure than that of a stock 930. His turbo will pump more airflow into the cylinder than a stock turbo.

Also timing curve, piston speed, intake valve opening and closing points,cylinder head and piston design, rpm, etc will alter detonation resistance and cylinder pressure levels. This will also alter the burn speed (although somewhat controlled by fuel used) and combustion efficiency.

Maybe I wandered off topic. If so nevermind, I didn't mean to sidetrack the discussion.

Brett raised some really good points about how engine design can affect knock that skew straight octane vs compression ration/heat of compression. That I won't go into.
But, at full boost, turbocharging systems will be designed to flow enough air that excess must be vented out the wastegate (or bypass on newer engines). This means that no matter how much compressor "flow" available, the air density supplied to the cylinder is the same. 1.0 bar of charge results in the same amount of air no matter how much flow the turbocharger can supply.

Now, if your turbocharger went into choke at full throttle load, then the higher flowing unit will definitely provide more pressure. But, as I stated before, no manufacturer would design a turbocharger to hit its choke line.

A larger turbocharger combined with a correctly sized turbine wheel will likely hit a higher flow value at a lower turbine rpm (due to its compression curve), perhaps giving a higher boost pressure at a subsequent lower engine rpm. This is not guaranteed, but typically the K27 turbine and compressor wheels are designed for this. This means you may hit full boost pressure at a lower engine rpm, but if you have a 1 bar wastegate spring, the max pressure you will can get is still 1 bar.

Now, completely off topic but this reminded me of something someone said about turbochargers at high rpm. The statement was to increased boost pressure to account for the higher altitude. This is not true. Temperature being constant, the air density at seal level at 1 bar is equal to the density at whatever elevation. The only difference may be a slightly higher compressor discharge temp, due to higher delta P at max boost combined with lower cooling capacity from the intercooler, due to loss of air density all resulting in a higher intake temperature. At the same pressure, the higher temperature will result in a less dense mixture and lower O2 values.

I definitely wandered off topic. And I am pretty sure I had some really long, grammatically incorrect sentences :)

Oh, just to clarify, my comment was not against Brett's comments on dynamic pressures seen within an engine's combustion chamber. I believe those to be correct.

As I beleive he stated, it is really hard to compare engines that are not identical in all engine design components from intake to exhaust.

Wouldn't the airflow (mass) be lower at elevation though due to the lower atmospheric pressure feeding the turbo? Yes a turbo will generate the same level of boost at elevation as it will at sea level, it may take more rpm(turbo) for the turbo to reach that level. Your increased compressor outlet temps would come from the compressor working harder on less air, right?

All those opinions and infos are very interesting......does someone able to post a little bit more about the relation between those specs and hp results...?

Thanks,

I don't think there really is a relationship. Every engine will be different in what it wants as far as cylinder pressure rise and the resulting HP. I doubt you could lay out some hard ground rules as far as this boost and this compression ratio equal this HP.

Years ago I ask the question: "Would an engine with low compression and high boost make more power than an engine with high compression and low boost"? Well I got a million different answers and none of them really correct. It all depends on the application. I can take a modern pentroof four valve engine and run 11.5 with 10lbs and get excellent drivability, power and efficiency, but I would imagine that same engine running 9.0 would take more boost to make the same level of HP, thus adding more heat to the combustion cycle. It would be interesting to test it. Now with a 911 engine it will be a vastly different situation. The poor combustion chamber design and the aircooling/oilcooling require another approach. If I can get some time maybe I can build a motor with 10.0 and one with 7.0 and see what the differences would be. It would just be a matter of changing pistons. Leave everything else the same. Obviously it wouldn't be optimized for each condition, but it would be an interesting test.

One thing that would really affect the results of the test would be the usage of the engine. A street motor with limited track use, would be better with the higher compression, but the race motor would probably work better with lower compression since your turbo would remain spooled most of the time and you wouldn't have to worry about poor drivability outside of turbo.

Sorry I ramble.

I got it now, thanks.

One common data point you need to add to your chart:

Stock 3.3T 7.0:1 C/R @1.0 bar (using an intercooler)

VERY popular setup. My math says 11.6:1 effective C/R.

Brian:

This is the kind of consideration I would make of this data. It isn't a bunch of theoretical data about various engine configurations. It's data comes from the same basic engine from the maker who's evaluated it's likes and dislikes and has the real world experience you can accept for your use or reject.

It's easy to get "wrapped around the axel" in these discussions. Mainly because opening the population of data too broadly includes applications that defy the chart, then you end up discussing the exceptions rather than the rule, so to speak.

In response to your data point I might add I've always heard you can increase the stock 3.3L boost to 1 bar BUT only after upgrading the intercooler. Is this accepted wisdom correct, or is it conditional? Who really knows, in the end it's your choice.

My point is this allows you some reinforcement for the next component. You can visualize how far you maybe pushing the envelop. DonE made on observation using this that I came to also(but didn't tell Don). DonE gave me two boost levels and HP. The high data point is 525HP vs 500. IMO he rightfully assessed his engine didn't like the additional boost and the apparent risk is not worth an additional 25HP on an already very powerful street engine.

Yes, you are correct. Temperature rise will be based on differential pressure (work done to the air). And the turbo will need to spin a little faster, so your lag will be exaggerated. However, a higher pressure boost spring will not change the compressor curves. Max pressure will be higher, but the time to max will be the same.

At very high elevations, the compressor wheel design would need to be modified.

I thought the resulting effective C/R of the standard increase to 1.0bar on a 7.0:1 930 engine interesting because it falls nicely within what is considered a safe zone on your graph.
The operating environment for any vehicle can be very wide, from well below freezing to well above 100F at varying elevations. Having this graphed or mathematical information is necessary to make decisions of how far you can push your engine parameters given the environment it will operate.

I agree entirely, I would do it for street. I may be cautious on the very few extreme summer days here in IN. Regarding the factory TURBOS, there should be some leeway. These engines have mechanical timing and no knock sensors so there must have been some built in consideration for marginal octane and high ambient temperature.

Hello!

My comment was maybe a bit hasted, I wrote it after few beers. Sorry. ;)

Engine was a turbo-converted 964 N/A 3.6 with hot cams, twin Garrett GT35R's, twin huge intercoolers, cams hotter than GT2-evo and Rochester 1000cc-ish injectors. Pistons were 993tt stock pistons and it had H-beam steel rods. It was dynoed at 2 bar using standard (?) race fuel...I forgot which sort. On ordinary 98 octane RON fuel, boost could be upped to around 1.6 bar w/o knock. It had twin plugs as well (having N/A 3.6 heads)

I belive the graph is accurate but I suspect graph assumes 100% VE and low air temp. What I'm trying to say is that typical working installation is going to have different dynamic C/R at different revs, despite constant boost pressure.

As turbochargers go trough their different efficiency ratios, so does discharge temperature. The more heat in the air, the less oxygen is introduced into combustion chambers. Also, VE is going to vary wildly depending on cams, runner length etc.

Of course, Porsche aircooled heads aren't as good when it comes to knock as modern pent-roofed 4v designs with plug in the middle, but there are numbers of tricks one can use to prevent knock despite soaring dynamic C/R ratio. Ignition retard, twin-plugging, high octance juice and efficient intercoolers are some of them.

P.S. It's quite amazing what kind of boost you can run on modern engines. I have an old Audi UrS4 (93' model) that boosts 1.5 bar of boost on the top of it's 9.3:1 C/R.

Which brings me back my earlier inquiry on this thread and "930 timing specs for higher HP turbo" thread inquiry that at least in the beginning of Porsche turbo development low compression had its advantages especially under race conditions and for many years tuners such as Andial, Kremer and others, built to a known quantity standard. Now on to try to improve my motors using the applicable evolution, to adapt the benefits of more recent technology to update,tune what is there without changing every thing, as it my be applicable, to my 80s style motors. Lots of thoughtful information that may be good for me ,and other readers so far on both threads. I have been away from "hands on tuning and upgrading" for many years.I need a tune up before I proceed with these tasks.Thanks Tony@voitureltd

Goran, no problem on the comment. I often wish I wasn't geograhically displaced from ALL you guys so I could buy you all a few rounds.

Thanks for the information about this engine. Is this going to be a street engine?

We have talked of the importance of flowing, modern turbos and their efficiency (I like the name Garrett) and killer bar/plate charge air coolers (Garrett also).

Now assuming this engine has one other enabling technology, knock sensors, (only shown on the 3.6N.A. data point) how much have they entered into your moving this motor to very high TURBOcharged power levels?

Nope, no knock sensors. Just standard "deaf" MAP-based EFI and heaps and heaps of fuel and boost.

http://forums.pelicanparts.com/uploa...1150047914.jpg

It did have it's share of problems though, until everything was buttoned up:
http://forums.pelicanparts.com/uploa...1153688159.jpg

But amazingly, it never suffered from knock-induced damage. Head split after severe mapping-induced backfire. When you have injectors flowing that much juice, it just take one wrong cell in fueling table to hydro-lock the cylinder and do all sorts of nasty things.

I believe that best way to postpone knock in aircooled 911 engine is twin-plugging, efficient turbochargers and intercoolers, lot's of cooling fuel (=rich) and one thing that is often omitted: cool heads! Fan was severely overdriven on this engine, in order to keep the heads cool. So much that it actually munched trough several failed generators and belts until gen got moved and got it's own belt.

As far as I understand, alloy used in 3.6 heads will only take up to 240 deg. C before it goes soft (like plastic deformation) or ruin the things by pinging. If you push double or tripple HP trough heads while retaining same airflow, they will get hotter and ruin the day. Also 3.6 curved fan is quiet(er) but flows approx 70% of noisy straight-bladed turbo fan.

but yes, this is pretty academic discussion as nobody would run 2-bar on the street. ;)

But I've seen guys turbo-converting othervise stock 10:1 C/R Alfa Romeo V6 and running up to 1 bar of boost using E85 fuel! I also experimented myself with E85 and couldn't get darn corn-juice to ping no matter what I did :)

I'm speachless. That engine looks more like one of the hydrogen fueled main stage rocket motors on the space shuttle. Probably consumes fuel at the same rate...I love it!
Please help me understand, is this a dragrace type car? 1.6/2.0 Bar at what RPM?
Do you have any shots of the rear bodywork all buttoned up?
All of your comments on cooling are right on. At this power level there is so much waste heat energy in the cylinder/head system limits are reached in short order.
Still, detuned it's a very strong motor.
Thanks

Here you go:

http://forums.pelicanparts.com/porsche-911-technical-forum/254697-if2-thread.html?highlight=ItsFun2

Brian, thanks for the link. Lot's of reading, now I'm up to speed on project.

Once you've chosen your turbo, IC and twin plug, there are only two things you can manage to reduce detonation: timing and fuel. In my opinion, adding fuel to cool the charge to reduce detonation only works to a certain point. I've seen motors run at a very low AFR that successfully eliminates detonation only to wash the cylinders of oil and severely accelerate cylinder/ring wear (as well as dilute the engine's oil supply) and trash your valves. Being able to properly tune a modified, air-cooled, 30 year old turbo motor is kind of an art. For my motor, my lowest AFR is 11.8 and timing at .8 bar, 6000 rpm is 18 degrees.

Well just to stur the pot.........here goes. You all ready have a killer engine but you can cool the charge AND cool the flame with water/alcohol injection. That is a third option. W/A injection is more than intercooling. Wow, now I've asked for it.

Well of course. Water injection has been used from beggining of time on aircraft engine, to prevent pinging and improve mass flow. Unfortunately, I personally find it a bit cumbersome to carry extra system containing water/methanol. Also, there is a issue of redundancy...what if/when water runs out and pump stops? Do you pay for it by ruined pistons?

Personally, I find E85 solution much more elegant than gasoline/water/methanol combo. First of all, it takes roughly 40% more E85 than gasoline to reach same lambda. Second, E85 has octane ratio of 104, third, it's specific vapor energy is *four times* higher than for gasoline.

With other words, E85 will use 4x1.4=5.6 more heat from charge air to evaporate compared to gasoline (which keeps EGT's cooler), is very knock-resistent and negates the need of external water-injection system.

All-in-one solution and quasi enviromentaly-friendly too ;)

But E85 is close to 6.00$ US a gal to make. You can't buy it in many areas all over the country and most 930s get bad enough mileage as is. Actually the number is about 27% more E85 than regular gasoline, but you can make some huge power out of it.

Unfortunately E85 is a myth that keeps being perpetuated by the media. It is more expensive to produce than Gasoline and we don't have the land to actually supply it consistently.

I am not sure I understand the point of all this. You can probably get a 930 3.3 motor to easily break the 1000hp mark by a large margin. Its called good theory plus trial and error. Even today's OEM's work that way, but given advances in computing, the error part is reduced.
People running big block V8's could "theoretically" hit the same HP figures as top fuel cars. Is that a discussion worth having...

But, for all of these, you will pay in engine life.

But we can just go on naming more and more expensive and unrealistic (for the street) methods of making power.
very high octane fuel, high CR, optimized valve overlap and timing, valve sizes (heck - 4 valve per cyl, with water cooled heads), huge air to water intercooler, huge turbo, high boost pressure, blue printed engine, stripped car for weight, blah blah blah.

Hmm....I'm quite sure it takes roughly 40% more volumetric flow to get same lambda as gasoline. My WBO2 measurments also show this to be true.

Gasoline:
44 Energy content [MJ/kg]
735 Density [kg/m3]
305 Vapor energy [kJ/kg]

Ethanol
27 Energy content [MJ/kg]
785 Density [kg/m3]
840 Vapor energy [kJ/kg]

1L gasoline contains 735 [kg/m3] * 44 [MJ/kg] = 32340 MJ/m3 = 32.3 MJ/L
1L ethanol contains 785 [kg/m3] * 27 [MJ/kg] = 21195 MJ/m3 = 21.2 MJ/L

With other words, you need 52% more volume to obtain same amount of energy from pure ethanol ( 32.3/21.2 = 1.52 )

E85 usually contains 85% ethanol, 12.5% gasoline and 2.1% MTBE, so volumetric flow to obtain lambda 1.0 is usually around 1.4 of gasoline instead of 1.5 which would be the case with 100% ethanol.

So number is not 26% more but roughly 40% more.

I won't delve into enviromental merits of E85 as they are somewhat dubious. But as high-performance fuel for tuned turbocharged engines, it's quite handy. Fortunately, it's available at almost every pump overhere and cost's less per litre than gasoline. MJ per MJ, it costs roughly the same...

Regards,

Agree on the e85 comments and yes it is closer to 40% + to match gasoline.
The water met setup is easier here than finding and modifying a stock fuel system to accommodate e85....

Regarding the damaged engine , i would dis-agree that it was just a backfire that caused that damage , it looks like excessive EGT's , coupled with an engine misfired midbore ( lean condition) the resulting Cylinder pressures lifted and torched the cylinder head .

Well I guess we'll never know for sure but there were no signs of excessive EGT's. Pistons showed no pick-marks/melting typical for lean mixture (actually I believe piston survived but was binned beacuse of ovality). Head just exploded...clean break edges, no melting. It's somewhat atypical operating range for OEM 3.6 head though...

Excessive EGT's on that head don't appear the case. The head is soot riddled. I have heard that twin plugging our heads do weaken them somewhat. It does look like it started near or at the second plug threads. Was that a 930 head or a 3.2 head?

i would like to get your input. I am currently having the following work done on my '91 965 engine:
8:1 JE pistons
Clewett Engineering crank fire twin plug system
UTCIS didgital WUR
0.6 bar (large yellow) spring for TIAL wastegate
Blitz dual solenoid boost control

I have full a B&B exhaust system with a Zucz cold air intake on the car. What boost can I safely run on this setup and what else should I have done to the engine (sort of while your in there)?

I am hoping that the additional upgrades address the under the curve laziness while keeping power under boost. I am not considering EFI conversion at this time, but does the choice of parts (except for the UTCIS digital WUR) make sense with EFI?