Injecting this manifold?
Turbo Mini Forums

Ok, here it goes, after reading in books and checking threads on this forum about the work Paul S, Rod S and Graham T and ofcourse a lot of info by Jean I have found some courage to come before the knowledgable ones.. :)

So I want to inject this manifold.

When considering injection first I thought I should just make it a wet manifold.
But looking at the possibilities I should at least try to do it the ´proper´ way.

The problem with my manifold is that it was not initially designed to be injected.
Result is that I can´t put the injectors to close to the ports.
Then I thought about using pencil pattern injector shooting across the port.


Using the pencil pattern injectors I was hoping to bridge the distance and get more accurate fueling per port.
I was hoping pencil injectors would also reduce wall wetting and combined with injectors being timed after or in final phase of valve overlap reduce the charge robbing effect.

Am I correct that Rod S is also trying to work with timing to take care of the charge robbing effect? (I didn’t quite get the “skewed VE table” bit.)

The two main problems / questions I foresee are:
1.) Heatsoak; I was planning on using some slim EV6 injectors or anything alike. However the injectors still remain close to the exhaust manifold, about 15mm.
Now I can make some heatshielding but will be enough?
2.) Precision of injectors timing; how precise injectors would I need to get injection right if I want to do it this way? I was thinking about using 4x 400cc injectors (or something close).

My engine spec:
1293cc
9:1 CR
Mg metro cam
1:1.3 rockers
10psi boost
M45 supercharger
21352 pistons 8.4cc dish
Ported head (MSE6 with) 28cc chambers
Megasquirt + 2 widebands

Now, the mini is purely a hobby car, I enjoy working on it very much (I am a mechanical engineer) but also looking forward to driving it…
Not looking for mega performance, just the extra little kick in the butt… :)
So basically, is the idea I have worth a go or am I getting myself in a lot of trouble and should I better stick to the wet manifold plan and see if I can stick 2 injectors in the vertical bit before the manifold splits to the ports?

Your thoughts are much appreciated!

Edit typo..

Edited by Yo-Han on 8th Nov, 2010.

Dazed and Confused....

With 4 injectors set like that you wouldn't be using the siamese code but regular sequential injection. That also means that you need to have the injectors large enough at about 25% duty cycle for your power target.

Timing for the injection won't be that much more critical than it would be with the 2 pulse mode of the siamese code but the issue is that you'll have to use the trim tables to adjust the fueling between the inner and outer cylinders which means you'll need to copy the data from one pair of cylinders to the other pair. Also, you won't be able to use staged injection since you'll already be using the 4 injector drivers but that may not be an issue.

Jean

http://www.jbperf.com/

Although your drawing makes it look as if they are each aimed at their own specific inlet valve - hence I think Jean's suggestion of using regular sequential injection - another approach, with 4 injectors mounted like that, would be simply to wire them in parallel on each runner and use the siamese code as if there was only one larger injector on each runner.

This is what Paul has on his 998 at the moment although his 4 are in different positions without any angle towards individual valves.

Again, you would not be able to run staged injection if they were wired in pairs but the opposing angles would theoretically cancel out and you wouldn't need pencil type injectors as the two would would be mixing into one pulse.

Graham's initial work on staged injection shows that the angle of the injectors (if used independantly) does cause a further imbalance between inner and outer cylinders and Paul's staged injection design uses 3 per runner to cancel out any effects of angle.

If wired in parallel, 400cc each is going to limit power - Paul has 480cc paired together for approx 120HP and my initial setup is 1000cc individual injectors on each runner for an intended similar output.

Schrödinger's cat - so which one am I ???

My only comment, in addition to what has already been said, is that due to the turbulent flow within the port, it is highley unlikely that angling the injector towards a particular valve will have any benefit.

Once the fuel is in the port, it will travel in the same direction as the air flow, ie along the port centreline, then turn into whichever valve the air is going.

Having said that, the injector positioning will be fine and you have sufficient flexibilty within the code to get it to work, one way or another.

Are you sure that the manifold is going to clear the bulkhead?

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

http://www.jbperf.com/

I think that if the angle of the injectors is equal and opposite, then wiring them in parallel and treating them as a single injector will give no more issues than normal siamese mode.

However, it does prevent staging being used unless a third one can be squeezed in.

Using "normal" sequential is an interesting idea but I think would only work if you could actually direct an injector at a specfic valve with absolute reliability and have really large sizes so as to keep the duty well under 25%.

But it's all experimentation at the moment so worth considering.

Schrödinger's cat - so which one am I ???

Thanks for the input guys, much appreciated!

The main reason for going this route was because I can't really get the injectors any closer to the ports. I thought I could 'reduce' a bit of this distance by using pencil injectors.

Earlier I thought about putting 'ordinairy' injectors in the manifold but as they will be about 5cm away from the head this would pretty much make it a wet manifold and would not be able to use the Siamese code.

Summarising the input leads me to believe I would make it pretty hard on myself continueing this route.
I am going to see if there is a possibility to redesign the manifold, something like the manifold Paul S has on his 998 but than upside down.... This way creating space for two injectors on top of the runners much closer to the head.

But I really need to see how much space there is for this.
Your thoughts are more then welcome..
Thanks again for you thought!

(@ Paul S, yes, manifold clears the bulkhead.)

Dazed and Confused....

Unfortunately I can't find enough time to make a proper manifold for using the siamese code.
So I have decided to settle for the less perfect wet manifold injection for now and try to make a proper manifold later.

I used an injector calculator which says I need just over 400cc/min to get at 125 hp (0.55 BFSC, 0.9 Duty cycle)

Now, I am planning to fit 2 MPI injectors into the 'vertical' runner before the T split.

Does this make sence (apart from going to wet manifold)?

Edited by Yo-Han on 19th Jan, 2011.

Dazed and Confused....