Technical CIS operation & Cam Timing w/ CIS Discussion

[911sTarga]

smdubovsky

You wouldn't happen to have that in .pdf would ya?

I'd love a copy. I still think a systems approach to the CIS engine in whole is right answer. Compression and cams should be possible with a small tweak of the plenum volume, a shift in the cam centers, and some exhaust science. Agh...it is still just theory though.

I was hoping that someone would chime in with actually real world testing that they've done, especially since it appeared to be "common knowledge" that it is impossible to hot rod the CIS engine.

I'm not saying I'm right, I just want someone to prove me wrong before I go too far into this. If someone has the data, please step up!

Theory is just that, theory. One theory is no greater then the other, but hard data gives you numbers and I've learned that numbers just don't lie.

Jay

[Mysterytrain]

So then whats the feeling on boring out the throttle body?

[camgrinder]

Hopefully you go as far as flowtesting the heads and mapping the velocities. I am mostly interested in the low lift (1.5mm) velocities on both the intake and exhaust ports as a system.
If you do the wavelength analysis, what temperature will you
use for the intake and exhaust ports? It would be interesting to see how
the air cooled engine port temps compare to a typical water pumper.
You might have to build a stepped style header as well.
You can build a mule motor for the dyno and try a 964 style cam, put it in advanced 5 degrees and make a pull. Maybe retard it in one degree increments and make pulls until you had it 5 degrees retarded. This data might show certain trends you can take advantage of.

[jluetjen]

Wow, I looked away from this thread for a day or two and look where it's gone. Previous stuff that I've posted related to this has already been referenced, so just a few thoughts which I don't think have been covered yet.

1) Problem statement: The first post talked about more HP. HP is a function of air consumed that is generally a function of engine speed assuming engine size (2.7) is a constant. Basically, if you want more HP from a 2.7, you need to spec the engine's torque curve to be higher in the rev range. To get past 210 HP, I would expect that you'd need to be rev'ing the engine past the RS's rev range with it's 6300 RPM peak HP engine speed.
a) The peak HP engine speed is defined by the point where the torque drops off faster then the rev's increase. So it's really a function of the torque curve.
b) The 2.7RS had it's peak torque at 5100 RPM. So you are going to need to spec the engine to have it's peak torque at around 5500+ RPM. This will help you to figure out what cam you want.

2) How fast will your torque curve drop off? Personally, I'm of the belief that one of the things about CIS that limits HP is the airflow at peak RPM's. This should be easy to test by hooking a vacuum gauge up to a plenum and doing a couple of WOT pulls to redline. I suspect that past the peak torque point that a CIS system will start to generate a ~5% vacuum compared to ambiant. Unfortunately I don't have a CIS car in which to test this theory. Nobody else has stepped up to give it a try. Using my estimate for discussion, if CIS systems are pulling a 5% vacuum at peak RPM's, I would expect them to develop 5% less HP when compared to let's say an MFI system which has no such restriction. Or to put it differently, you may need to rev 5% higher then a similar non-CIS's motor to get the same HP, which implies even more radical cams. Note that you can flow the components of the system all that you want, unless you test the system as a whole you really won't know what's going on.

3) Driveability: The other place that I believe CIS sucks (so to speak) is in the driveability when using cams which are designed to hit the RPM's that we're talking about. This is the reversion issue that was discussed above. Once again, if you have overlap (as in "hot cams") at small throttle openings, when the frequency of the exhaust system results in a pressure pulse at the exhaust valve*, AND the intake port's pressure is below that of the exhausts, or even ambiant (due to the butterflies being closed and/or the plenum pressure being less then ambiant), the exhaust will contaminate the charge in the cylinder or in more drastic situations back up to the intake system. Eliminating overlap, or increasing the pressure of the intake system (ie: supercharging), or to a lesser degree removing restrictions in the intake design will mitigate the situation. Once you're up at the peak torque engine speed or above, I don't think that it's such a big deal since the velocity of the air mass in general overwhelms the affects of the pressure waves.

* Note: By my estimate, in a 911 engine's intake system with runners of .3175 meters, we're talking about a range of about 300 RPM between a pressure peak and an exhaust peak. The frequencies of the exhaust system will be a little different due to the different length. So at 4000 RPM your engine may be happy, at 4300 it might run like crap, and then clear out again at 4600. By tuning the lengths of the intakes and exhaust you can move the point on the curve where this occurs, but you really can't fix it by exhaust tuning. Stepped exhausts also apparently help.

My point? You're going to be juggling some conflicting system parameters in order to get where you are going.

PS: Another place to search might be some of the guys running CIS cars in the SCCA Production classes. In many cases the rules require the use of the stock injection system (with modifications allowed). Given the competitive nature of their cars they may not give away all of the secrets of the cam designs, but you might learn something. You might want to try posting the question on this SCCA Production Forum, but be forewarned -- the residents can get testy some times. Unfortunately the site was recently hacked and a couple of years of postings were lost.

[sammyg2]

I like my CIS. A.....lot.

I added 100 hp to an otherwise stock 3 liter and all I had to do to the CIS was ground the O2 sensor wire at 3 psi rising

[911sTarga]

Originally posted by jluetjen
1) Problem statement: The first post talked about more HP. HP is a function of air consumed that is generally a function of engine speed assuming engine size (2.7) is a constant. Basically, if you want more HP from a 2.7, you need to spec the engine's torque curve to be higher in the rev range. To get past 210 HP, I would expect that you'd need to be rev'ing the engine past the RS's rev range with it's 6300 RPM peak HP engine speed.
a) The peak HP engine speed is defined by the point where the torque drops off faster then the rev's increase. So it's really a function of the torque curve.
b) The 2.7RS had it's peak torque at 5100 RPM. So you are going to need to spec the engine to have it's peak torque at around 5500+ RPM. This will help you to figure out what cam you want.

The above is very true, but it is also quite vague and limited in truth because of that. I appreciate the input here, but above is a tad engine 101.

Like you said, horsepower is torque over time, period. SO what you said above is true, IF your torque numbers are the exact same as an RS engine. If so, you're just moving the curve later in the RPM range to extract more top end power at the extreme cost of RPM. There is a better way.....

You can (should) find ways to begin building torque sooner and (hopefully) max it to a higher level. This is done by keeping the curve flatter which will also extend its operating range. This is where matching the port volume/shape to the Valve head size (especially exhaust) to the intake runner length/volume to the plenum volume, and so much more become critical. It's a rubix cube of variables that involve port flow matched to theoretical CFM matched to the desired RPM range. It's extremely crucial to get them right or your power output goals will never be seen.

Again, the engine must be taken on with a systems approach, but we must first understand how each piece reacts to certain changes. The one true unknown and what we need to clearly understand is why a CIS induction system reacts the way it does with certain cams. Once we have that data, then we can make accurate decisions on engine modifications while retaining the CIS unit.

I don't want people to think that camshaft design is the only piece of the puzzle to make power with the CIS unit. That isn't the case at all. A discussion on how the CIS unit reacts to different cam profiles and what proof there is about the "myth" against hopping up a CIS engine is the underlining topic here.

Originally posted by jluetjen
2) How fast will your torque curve drop off? Personally, I'm of the belief that one of the things about CIS that limits HP is the airflow at peak RPM's. This should be easy to test by hooking a vacuum gauge up to a plenum and doing a couple of WOT pulls to redline. I suspect that past the peak torque point that a CIS system will start to generate a ~5% vacuum compared to ambiant. Unfortunately I don't have a CIS car in which to test this theory. Nobody else has stepped up to give it a try. Using my estimate for discussion, if CIS systems are pulling a 5% vacuum at peak RPM's, I would expect them to develop 5% less HP when compared to let's say an MFI system which has no such restriction. Or to put it differently, you may need to rev 5% higher then a similar non-CIS's motor to get the same HP, which implies even more radical cams. Note that you can flow the components of the system all that you want, unless you test the system as a whole you really won't know what's going on.

You are 100% correct, but you don't need to test an actual engine to prove it, let alone the CIS. It is pretty much common knowledge with any engine builder that you'll create vacuum in a shared plenum after you pass peak torque. Again, sort of engine performance 101. This is why a shared plenum intake will create more useable torque down low, but sacrifice top end power, when compared to an open plenum system tested over the same RPM range.

The benefit of the shared system is obviously torque over a longer period of time, so you change the volume of the plenum to match your desired RPM range. You lose a little off-idle drivability, but you'll get a fat torque curve deeper in the RPM band and increased HP output.

Guess what? You'll also remove some of the reversion problem with a larger plenum as well. Removing some reversion would allow the use of slightly larger cams. I see a trend here..... Anyway, I think we discussed plenum volume and how it works briefly in this thread, but I might be mistaken.

If you really want to test pressure drops and how the CIS reacts, put a manometer in the plenum and then one on both sides of the sensor plate. Take a drive and record the data at part throttle cruise and at WOT. Do this with a few different sets of cams and we'd have all the knowledge we need.

Keep me posted on your testing. (Just kidding around with you)

Originally posted by jluetjen
3) Driveability: The other place that I believe CIS sucks (so to speak) is in the driveability when using cams which are designed to hit the RPM's that we're talking about. This is the reversion issue that was discussed above. Once again, if you have overlap (as in "hot cams") at small throttle openings, when the frequency of the exhaust system results in a pressure pulse at the exhaust valve*, AND the intake port's pressure is below that of the exhausts, or even ambiant (due to the butterflies being closed and/or the plenum pressure being less then ambiant), the exhaust will contaminate the charge in the cylinder or in more drastic situations back up to the intake system. Eliminating overlap, or increasing the pressure of the intake system (ie: supercharging), or to a lesser degree removing restrictions in the intake design will mitigate the situation. Once you're up at the peak torque engine speed or above, I don't think that it's such a big deal since the velocity of the air mass in general overwhelms the affects of the pressure waves.

Please see above.

Originally posted by jluetjen
* Note: By my estimate, in a 911 engine's intake system with runners of .3175 meters, we're talking about a range of about 300 RPM between a pressure peak and an exhaust peak. The frequencies of the exhaust system will be a little different due to the different length. So at 4000 RPM your engine may be happy, at 4300 it might run like crap, and then clear out again at 4600. By tuning the lengths of the intakes and exhaust you can move the point on the curve where this occurs, but you really can't fix it by exhaust tuning. Stepped exhausts also apparently help.

My point? You're going to be juggling some conflicting system parameters in order to get where you are going.

You do understand that you can change this point with the camshaft profile and with the plenum volume too? Intake volume also changes the speed of the pulse as well. A "tuned" plenum with a stepped exhaust will all but completely tune that "300RPM flat spot" out (plus much more).


Before we go too far into thermal dynamics and how pressure should be measured inside the intake, let me show everyone my "flow bench". I don't have the typical flow bench found in your local machine shops. My system is used to test some very high profile clients designs (Ford MS, etc...), plus it can also test an air filter to ISO 5011 standards for restriction, efficiency, and capacity. The CFM flow rates can be changed, plus intake pulses can be introduced to simulate engine load and operation. Max flow is well over 2000CFM, so we have the room to test the CIS.

Here is a quick picture of the testing equipment.


Anyway, I hope to see the CIS unit installed on it real soon for some airflow *restriction* testing.

Enjoy the info guys!

Jay

[randywebb]

very nice

- what sort of data collection can you do? local pressures? local streamlines? local velocities (how measured)?

[Purkinjefibre]

Anyone out there lucky enough to have a Bosch K jet manual that they would like to share????

I am willing to scan in it and make a PDF to distribute amongst members on this board.

I'll do it on federal time (Cheers "W").

Brett

[mikensam]

I have a copy of C.Probst's Bosch fuel injection and engine management published by Bentley in 1989.Has 36 pages of theory on CIS,another 35 pages of troubleshooting and sevice,and 22 pages of performance tuning.Havent had the pleasure of purusing the Yellow covered K-jet manual that smdubovsky posted earlier,so i dont know how much theory it contains vs Probsts book.I can scan sections that supplment the bosch manual if there is interest.

Chris