bad cylinder distribution?

[slostar]

hey,
after a few ideas with my inlet manifold and water injection points. at this point its just running a simple single stage system. injecting about 500mm before the throttle body, with 100% distilled water. manly concerned about in cylinder cooling and det suppression. but doesn't seem to be doing alot to move it away from detonation. i have tuned the car up to the point, where it gets very slight det at one point in the rev range. (listening through a detset) then switching the water on, then adding more timing. although that's about where it stops, as i can only get about 1 degree more advance at the most. am i expecting too much, or is something not quite right? have tryed diffrent size jets m2, m3, m5, with same results. except with m5 it looses a good 5hp from injection starting point till rev limit

at the moment im pointing the finger towards the inlet manifold:

http://img.photobucket.com/albums/v178/slostar/starlet%20ep/Pic_000434.jpg

on the car:
http://img.photobucket.com/albums/v178/slostar/starlet%20ep/Pic_0018.jpg


the way the air/water is directed to the rear of the plenum, it seems the water could be completely bypassing the first cylinder. or even hitting the back to the plenum and falling out of suspense? witch would make sense, as i do get a bit of water hanging about. (all bad when the vac lines come from the bottom of the plenum, then water runs down to the ecu map sensor, when you somewhere you really don't want to break down. but that's another story :razz: )

i have thought about going to port injection. but only being 1.5L even four of the smallest jets may drown it?

any ideas or suggestions would be great.

kris

[sdminus]

Could you not inject the water just after the air filter. Im sure the intake tract would get hot when the car is running anyway. This would defo atomise the water better.

Scott

[slostar]

you mean before the turbo? i thought pre comp injection had little effect on in cylinder cooling? not to keen on pre turbo injection anyway

heres a better picture of the intercooler piping. intake off turbo is missing in this pic, air filter is mounted in behind bumper in the inner guard
http://img.photobucket.com/albums/v178/slostar/starlet%20ep/Pic_005923.jpg

injection point is about between the + and - terminals on the battery, just after the bend. you can see in the link in my first post

[sdminus]

I spose that is matter of opinion really Since the turbo is creating a lot of the heat.

How about moving it closer to the intercooler outlet. I think that the design of the inlet manifold is the main problem here TBH. What is your average inlet temp.

Scott

[JohnA]

hey,
...at the moment im pointing the finger towards the inlet manifold:...
I would also suspect this manifold when it comes to even distribution.
This also goes for air, as well as water mist.
Assuming that fuel distribution is even, then the cylinder that gets most air will be running leaner than the rest, won't it? :wink:

Ideally you'd get it on a flowbench to identify any major flaws and modify it accordingly.
Usually they tend to be tapered towards the most remote cylinder (looking from the throttle) but the internal shapes make a big difference as well.
It's hard to tell by simply eyeballing.

[sdminus]

You have uneathed a whole problem that prob affects 90% of water injection users out there. I have a similar prob im working on. The problem is the differance of the high presure air on the outside of the bend and the low pressure on the inside in my case.

So in your case the air will flow to the farthest cylinder best and the nearest the least. This is gonna leave you short in that cylinder. Even if you have a collector which then divides into 4 you would still suffer. The main problem as far as i can see is that the manifold is designed to flow air not WI as well. You would see the prob more if you were running a AIC down stream and a egt per cylinder. I know Lotus have flow bench equipment but private work is a strict no no........ i need to ask some of the guys when im there about this....I have a better pic using my car of what your problem is.

Scott

[sdminus]

ooopshttp://upload4.postimage.org/570272/01072006121.jpg (http://upload4.postimage.org/570272/photo_hosting.html)

[slostar]

sdminus,JohnA thanks for your comments.

at the time i built the manifold water injection wasnt in mind at all. i think the engine should have deceased long ago if there was a major cylinder distribution problem with airflow. i dont know alot about fluid dynamic's etc, so im probably way off here.. but surely a big step down from the throttle body to the plenum. would cause a low pressure point and along with the throttle aimed to the roof of the plemum, (probably gone a bit far with the angle) would cause more of a "filling" effect, rather than aiming the air down the runners. hence no angle on the plenum. from there letting the runners/bellmouths do all the work. that was the idea, but how well it works is another question. im sure with the correct angle on the roof of the plenum it would be an advantage. but the way i see it, is if the angle is wrong, say too steep would there not be more pressure over the latter runners? no angle seemed like the best option at the time.

heres a few more picures before the manifold was put together
http://img.photobucket.com/albums/v178/slostar/runners2.jpg
http://img.photobucket.com/albums/v178/slostar/Pic_0044.jpg

anyway:D

but as far as water goes, it seem it could be a bit of a lost cause. water being heavyer and keeping on its merry way to the end of the plenum and not just hanging about waiting to go in like the air.. the way the runners draw in from the side not down like most oem manifolds (on top of the other problems) surely cant be helping the cause either?

unsure on inlet temps, but are pretty low, out let side of intercooler never gets much past cold to touch inlet manifold doesnt get very warm either. i should test it i guess.


im unsure where to go next, a reliable port injection system is almost out of my price range. maybe a jet between bellmouths between cylinder 1&2 and another between 3&4 could work? probably still not so great..



sdminus: looks like you were lucky there, you were running meth wernt you? what are your plans now?

[simple]

This is AUDI set up:

http://www.audizine.com/gallery/data/500/10126intake-med.jpg

Cars known to have cylinder 2-3 to knock more than the rest. This is belived to be due to air distribution. This is not a problem for street car, but if you start pushing the limits of the motor you will be running in to the problem...

[JohnA]

... i think the engine should have deceased long ago if there was a major cylinder distribution problem with airflow. i dont know alot about fluid dynamic's etc, so im probably way off here..

If I were to build my own intake from scratch, I'd drill a hole on each runner so I can fit threaded things like WI nozzles (after the throttle of course). They would also be useful as pickup points for boost/vacuum gauges.
Imagine how multiple throttle-plates are synchronised :wink:
If you fit four cheap boost gauges next to each other (on a piece of MDF) and have them all connected to the intake, then a friend could be looking at them in the cabin while you are boosting on the road and tell you if they all move together. If you find runner # 1 having a few psi more than #4 then you know that you need to modify the design and try again.

This could only lead to a better design that makes more power and will be more reliable under stress.

Then you either blank off the holes or use them for water/meth nozzles or whatever.

[slostar]

interesting idea, will look into it.
setting it up on the floor with the first part of the intercooler pipe attached, including the water injection jet. then adding an air supply to the intercooler pipe. and doing a few tests, may be good enough to show the floors in the wi. wonder if a leaf blower would push enough air ? :D

simple: i seen that pic in your other post. does look pretty bad desnt it.. looking at it from the outside anyway

[simple]

Slostar : I have spent some time looking in possible solution to bad water distribution (for my case) the only reasonable and cheap is to fit one nozzle per intake. All other solutions simply introduce more variables which are difficult to account and controll :mad:

[hotrod]

I suspect you have a "wet flow" problem. Log style manfolds like that might move air quite well but when you put a liquid mist in the air flow things change rapidly.
It is very likely that you have one or more cylinders that are not getting their fair share of water.

You could play around with a flow bench and do some wet flow testing but it is a pain to set up as you have a lot of liquid in the air flow and most flow benches are set up as suction devices that pull the air through the motors.

You could try to find an UV dye that will mix with the water and run it for a bit then pull off the manifold and see if the dye points out the problem.

The more certain solution would be to configure that as a direct port injection after the runners break off of the plenum.

I suspect your getting the liquid mist centrifuged out of the air flow as it tries to make the turns into the runners and one or more runners is getting short changed on the water mist. That style of log manifold is notorious for having unequal fuel flow problems in carburated systems due to this sort of wet flow problem.

You might try reading your plugs for signs of detonation, that would clearly show if your only detonating on a single cylinder, and which one is having the problem.
Put new plugs in the engine, take it out and run it up until you detect that light detonation, then pull the plugs and carefully examine them. The cylinder that is detonating should show signs on the plug electrodes and center insulator of very small specs.

Larry

[JohnA]

Can I also add that air/mist distribution under boost is not necessarily as uniform as under vacuum.
You really need to test for both.

[simple]

Can I also add that air/mist distribution under boost is not necessarily as uniform as under vacuum.
You really need to test for both.
Will i be asking too much if I ask to explain why?

[slostar]

only just realized this forum was working again.. thanks for your comments

i ended up doing a quick test with a garden leaf blower. first connected to the section of the intercooler pipe with the jet. with no manifold connected. it had a nice mist coming out the end of the intercooler pipe. now with the manifold connected, there was very little mist. be it with all the runners open, or 3 blocked etc. from memory some were slightly better than others, but not alot. so the mist was basically puddling in the plenum, then dribbling out the runners. sure its alot different to being on the car, but was enough for me to bin the single jet setup and try something else.

so its now running port injection. unfortunately i haven't had time to test it, against the original setup.
had a drag meet coming up, and managed to get a hold of some better fuel. witch made it easy to tune quite a bit more power out of it. but soon as ive ran out of the good fuel, will do some more testing.


http://img.photobucket.com/albums/v178/slostar/starlet%20ep/Pic_0008inlet.jpg

[JohnA]

Can I also add that air/mist distribution under boost is not necessarily as uniform as under vacuum.
You really need to test for both.
Will i be asking too much if I ask to explain why?

It's a matter of scale.
Vacuum is milder, usually it creates air movement due to pressure differences of a few psi, doesn't it?
...while boost can lead to pressure differences of dozens of psi, so airflow will tend to be far more turbulent. Minor eddies (appearing under vacuum tests) could turn into serious airflow obstacles under boost.

[keithmac]

slostar, I`m in the same boat as you with the log style manifold, going to try port water injection myself as well, hopefully have some good results..