site with fuel injection for straight eight .

[39CENT]

Ive been trying to take all this in, but my feeble mind can,t keep up. What about using a throttle body injection? won,t the fuel be fairly evenly fogged into the manifold?


ohh found somethin in google kinda answers my ??? so I need to get into this further, get edgycated.

 

[pmuller9]

Ive been trying to take all this in, but my feeble mind can,t keep up. What about using a throttle body injection? won,t the fuel be fairly evenly fogged into the manifold?

Yes, throttle body injection would be a good option. You get a nice homogenous mixture as with a carburetor only with electronic control. The problem is not fuel separation. The problem is the cylinders that share the same intake port do not get the same amount of air during the intake stroke making it difficult if a person wants to use an EFI unit which uses a fuel table based on all cylinders having the same Volumetric Efficiency.
It is a challenge that cannot be ignored. Fun stuff!

 

[Jyrki]

Valve timing

This thread finally got interesting!
Paul, I spent 20 minutes yesterday writing a reply with my new laptop, only to find out the system threw me out just before submitting my reply, and I lost it !

Anyway, you do have valid points there!
Yes, the #2 piston is already 71 degs past TDC while #1 piston is 71 degs past BDC, #1 valve closing and time to inject to #2 cylinder. You are right, at low speeds, there is probably reversion from #1 cylinder during the compression stroke (0 to 71 degs after BDC). The same holds true for #2 cylinder. So, if we begin injection to #2 at 71 degs past TDC, we only have 109 degs before BDC (and the start of possible reversion)

You are right, we might be able to start the injection sooner than 71 degs ATDC at low speeds. But the things change at higher rpm's or when you add a supercharger - air will move in as long as the valve stays open. I don't think you can change the injection start point in respect to RPM or boost, with any of the aftermarket EFI systems? I'm counting on the huge injectors to give the dose while there's still ample air movement to #2 cylinder. The 200 lb injectors are an overkill, just to provide a margin of safety in this respect.

Advancing the cam is a great idea, and should be done regardless of EFI, and especially with EFI! It would provide more squirt window before compression stroke (and airflow reversion).

Scattering the cam timing events between cylinders is a more sophisticated idea that I never thought about!

The stock cam timing looks rather radical on paper, especially the late intake closing. If I were to have the cam reground now, I wouldn't add "advertized" intake duration, but rather the "area under the curve", that is , more lift and faster ramps. I actually had my cam regound to stock specs some 10 years ago, with just additional lift. I wish we had this forum back then! I'm going to mock up the EFI using the stock cam, and build a "wild" engine later.

Why is the stock EXH timing shorter than INT timing? Who knows? Take any SB Chevy dual pattern cam, it's the opposite, even for naturally aspirated engines, not to mention supercharged.

 

[Jyrki]

Throttle Body Injection

39 Cent,
the TBI (Throttle Body Injection) is a viable option. For instance, the Hart/Collison team uses a system together with a roots supercharger, which could be described as TBI.
The reasons why I didn't choose TBI are:

1. I want a symmetrical, "dry" manifold design which requires port EFI
2. The TBI is not that much different from carbs, although it's so much easier to tune.
3. The TBI doesn't work too well with a blow-thru forced induction (turbo or centrifugal supercharger)
4. I get a "kick" out of solving problems. Going with TBI would be too easy for me. I want to go the hard way. That's why I chose not to swap a small-block Chevy in the first place :bgrin:

 

[pmuller9]

Jyrki, you probably already figured out that using Microsoft Word (or some other) to write a long reply, then copy and paste to the Site solves the timeout problem. I don’t feel so rushed plus I get to use spell check.

Do you know what the cam lobe durations are at .050” lobe lift? This probably gives more realistic duration numbers to work with.

After reading both our replies, I realized something important (a nice way of saying I was being stupid). At low rpm the VE numbers in your fuel table will be small meaning the injectors will be on for a shorter period of time even at WOT. Also the centrifugal blower doesn’t build max boost till later so injection time will still be a lot less than max.
At higher RPM there probably is enough airflow (Ram effect) so as the intake valve for #1 closes the difference between the airflow in the port and the air being taken in by #1 has to transfer over to #2. This transition occurs before #1 intake valve completely closes. So injecting earlier may be OK.
Ideally (meaning if you have the money) on a Dyno, all 8 EGT,s and the Air fuel ratio gets recorded as you move the injector timing around. Obviously the goal would be to get 1,3,6,8,operating as close as possible to 2,4,5,7. The budget method would be to monitor just 1 and 2 with a laptop on the passenger seat while going down the road.

One of the cars I work with uses the Bigstuff3 EFI unit. It will handle up to 3 injectors per port (Total of 24). Presently we are using two 160 lb injectors per port with the crossover point set at 25% duty cycle.

Which Procharger are you planning on using?
What is your horsepower goal?

 

[Jyrki]

Paul, once again, good points!
I don't have the .050" lift numbers, but I recall Alleycat had checked some of the cams and to confirm the degrees at .050" lobe lift were "modest" at best!

The volume squirted by injectors increases as load increases. The squirted volume doesn't increase with RPM's, but the injectors just squirt more often. The volume is increased by lengthening the injector open time. When you step on it at idle or low speeds, you do need a lot of fuel. But, in high RPM's there's not that much time for the injector to be open, while there's ample time at low rpm's. At 1000 rpm you have 5 times more time to squirt compared to 5000 rpm.

I'm going to use a P-1SC Procharger at 15 psi to get something around 400 HP. My intention is to utilize two wideband sensors to compare mixtures between cylinders, and I'm going to weld bungs on each primary header tube for EGT thermocouples.

Tell me more about your application!

 

[Jyrki]

Paul,
I forgot to mention you'e right with the Volumetric Efficiency (VE) being lower at low rpm's, and peaking at max torque rpm. Now that I read my reply, you get a feeling I'm trying to educate you which is not the case. My comment about the volume not increasing with rpm but load, was to simplify the basic case. You obviously know a lot more than I do, but I wanted to bring the "spectators" to the same "table". There are over 1000 views to this thread already, but for some reason people are shy to participate in this discussion.

So please shed some light on your project, even though it wouldn't be a STR-8 !

Tell me about the Bigstuff EFI. The ViPEC EFI is not capable of driving staged injectors in full sequential mode, but only in batch-fire mode. The Bigstuff is apparently able to both and more?

 

[39CENT]

Ok I now I got it, when you said basically you wanted to solve the problem. In that case why not bore out the ports as much as you can, and put in separators, would that work or have you already answered that?? gggg Ohh and I guess you would need 8 injectors then, and of course a blower cam.

 

[39CENT]

I have read the Leo Santucci book on building the siamese port 63 and up chevy inline 6,s, and in it there is where one racer developed a special cam timing for these ports. Maybe that would be an answer?? Lotta info there on porting.

 

[pmuller9]

Ok I now I got it, when you said basically you wanted to solve the problem. In that case why not bore out the ports as much as you can, and put in separators, would that work or have you already answered that?? gggg Ohh and I guess you would need 8 injectors then, and of course a blower cam.

There is very little room for porting without running into water and the head bolts get in the way. The stock head requires a "Y" or "V" port reconstruction. Someone is already working on this using several possible approaches.
Thanks for your input and reference to Leo Santucci's book. I'm adding that to my Christmas list
Paul

 

[pmuller9]

Jyrki….Thanks for your replies and great comments.
You said it correctly about this being an open forum for all to read and participate in.
The more participation, the more years of experience and info can be shared. As for me, the more I learn, the more I can see that I don’t know, so school is always in session for me. Keep the detailed comments coming. The following contains work you have already done so please bear with me as I continue with more detailed comments
.
It looks like the difficult area is the response time of a large injector. Can it deliver a small enough amount at idle and also deliver the full amount within .005 seconds when at 6000 rpm?

There are three types of injector nozzles for this application. Pintle, Disc and Ball. The Pintle is the oldest and most common using a tapered needle and seat but is also the slowest due to the mass of the armature and needle. The Disc is the lightest and fastest.

There are two types of drivers. Voltage controlled (Saturated) and Current controlled (Peak-Hold). The Peak-Hold initially drives the injector coil with a 4 amp pulse to build the magnetic field quickly then drops the holding current to 1 amp so the magnetic field is weaker requiring less time to collapse. Since this driver is current regulated the injector impedance can be low to decrease the time constant allowing coil current to rise and fall faster.

At 400 HP target, multiplied by a BSFC (Brake Specific Fuel Consumption) of 0.6 lbs of fuel per hour for each horsepower (for a supercharged engines) divided by 4 injectors and again divided by .5 duty cycle, gives an injector size of 120 lbs/hr. For big margin or overkill, injector size can be 160 lbs/hr or more for this application.
Ideal injector would now be a low impedance 160 lb/hr (Or more) disc type with a total response time of around 1 ms. The drivers in the EFI unit must also be Peak-Hold.

Looks like with two wideband sensors and EGT thermocouples you will not only be able to fine tune but collect some good data. Great to see someone spending the time and money to do research and development.

Will comment on the Bigstuff3 and the car in the next post.

 

[Jyrki]

Great to see someone spending the time and money to do research and development.

Paul, that's why my progress is so painfully slow! I've always been short of money, but now I can't find the time either!

I've chosen the Injector Dynamics ID200 injectors because so many people have told me they are the best, and the only ones linear enough to be used in street applications, that is, to be able to controlled precisely during idle and low loads.

See www.injectordynamics.com

This is a quote from T1 Racing who sells these:

The new Injector Dynamics ID2000 flows 2200cc/min at 3 bar (43.5psi) and is high impedance. That means no injector resistor or injector driver is needed. With the capability of running up to 130psi and flowing 3700cc/min, the needs for running multiple injectors per cylinder are gone on all but the most demanding methanol applications.
With the most linear flow in it’s class, great atomization and full injector dead time information provided, there really is no competition for the ID2000. Check out this video to see how well they run in a worst case scenario, a bone stock 120hp car!
http://www.youtube.com/watch?v=RX1xIvOWcNk

 

[pmuller9]

Jyrki, Thanks for the link to Injector Dynamics. I learned a lot. Your right about being a great injector and fitting your application. Beautiful linear graphs in the tech forum.

Again you probably have looked at this but for the benefit others, more detailed observations.

Looking at the Fuel Flow vs. Pulse Width graphs you will notice that at 16 volts all injectors needed 1 ms to begin flow and around a 9.5 ms pulse width to reach full flow. At 80% duty cycle for normal application that equates to 10,000 rpm. In your case the window is 5 ms at 6000 rpm. The graph specific to the ID2000 shows a 1250cc flow rate at a 6 ms pulse width which is close to your calculated requirements. Since it takes 1 ms to begin flow, if you make that your lead time, then there is 5 ms of spray time starting at 71 degrees ATDC.
Closing time on a high impedance injector can also be up to 1 ms so actual spray time can be beyond the end of the injector pulse. This is based on a fuel pressure of 43.5 psi. Raising the base pressure will give the needed margin.

Am I looking at this correctly? Please give additional reply.

What will you be using for a fuel pump and fuel regulator?

I know I promised to comment on my projects in the next post, but I got excited about these injectors.

Paul

 

[pmuller9]

Here is the short story.
My primary project is a 1949 Ford set up for Drag Racing in the Pacific Street Car Association, Outlaw 10.5 tire class. Since we run twin 88 mm turbos, the car and driver must weigh 3200 lbs. Top qualifiers typically run 6.60 quarter miles which for us means about 2700 horsepower. The present engine is a 522 cu. in. GM based Hemi running on gasoline. We only have one complete pass down the track with this new engine. 7.049 @ 198.55 mph with only 20 lbs of boost. We run an MSD Promag 44 triggered by the Bigstuff3 EFI unit. The Bigstuff3 will run up to 3 injectors per cylinder for 8 cylinders in sequential mode. The injectors for each cylinder can also be staged. The rev limiter for staging uses an algorithm that shares cylinder misfire evenly providing even cylinder temps at launch and a smoother running engine while being limited. Depending on the sample rate you choose it will record every parameter for minutes. Torque management is an option where you simply pull up a recorded run on a graph showing the driveshaft speed of what you might consider an ideal run and apply your graph points just above as a limit for tire speed. For further details see http://bigstuff3.com .

The present Buick project is a 263 S8 with an aluminum symmetrical 16 port head, 8 velocity stacks with EFI for a 1933 Buick 2 door Victoria and possibly a 1952 Buick Woody.

 

[39CENT]

yes the GMC's have a Y type port and run well. Also the small series 1935 and earlier Buicks, have Y ports and stock combustion chamber is milled out. Just see what the old 6 cyl inliners do. I,m just throwing some stuff out for discussion, if its not correct, well we can just discuss and correct wrong opinions. As for fuel injection, well thats 'whats happening' today, and a whole new world to me. I,ll be listening.

 

[Jyrki]

Injectors

Paul, thanks for taking it a step further - I never got that far :bgrin:
You are right. Five milliseconds is my window at 6000 rpm (although I doubt I'll ever rev that far up). According to your notion, the flow of of ID2000 is barely adequate at boost. Yes, I'd probably raise the fuel pressure, as it even improves fuel atomization. I need to consider advancing the injection start point by 20 degrees, as #2 piston speed (= airflow) is high at the time #1 valve is closing. So even if the valves are both open, #2 is the winner. Anyway, adding 20 degrees to 180 degrees would only yield a 5.5 ms window at 6000 rpm, the same effect as revving only to 5500 rpm.

I recall the pump and regulator are Aeromotive stuff. I haven't touched them since I bought them almost a year ago!

Wow, mind numbing HP numbers...
The best system for me would have probably been to use staged injectors, one in the port and another under the throttle body. And you're saying that the Bigstuff could have handled this? Now I'm stuck with the ViPEC - sucks!

Oh yes, there was another thread about the billet head project! You should give us an update on that original thread!

 

[pmuller9]

Jyrki, You are right about piston speed to airflow relationship. Advancing injection by 20 degrees and also advancing the cam by 6 to 8 degrees will definitely get injection started well before the velocity peaks. When you are done with this project you should compile all of these notes along with all of your findings and write an article (Or a book) entitled “EFI for the Buick Straight Eight”.

Also consider that since you are adding a supercharger there is no need to rev beyond 5000 rpm for 400 hp. At 5200 rpm 400 hp = 400 ft-lb’s of torque.

Max hp for a P-1SC ProCharger is 825. Max Impeller rpm is 62,000 with a 4.10 internal step-up. Given a 6,000 max engine rpm a 2.5 overdrive is possible. As you already know centrifugal supercharger output increases exponentially with impeller rpm. Gear the supercharger to bring in 15 lbs of manifold pressure by 2,000 rpm and use a bypass valve to blow off the extra pressure in order to regulate the 15 lbs up to 5,000 rpm. The torque curve should then be relatively flat between 2,000 and 5,000 rpm. (With more valve lift)
With porting and valve lift close to .500, 400 ft-lbs of torque should be attainable throughout that range.

Keeping engine rpm under 5,000 rpm and increased fuel pressure should give margin using a single injector.

If I where to use a second injector, Injector #1 would supply regular pump gas and injector #2 would supply high octane fuel where you would switch from injector #1 to injector #2 after 5 lbs of manifold pressure. You can do this and keep the ViPEC.

The Buick valve stem diameter is .372”. Consider bushing the stock valve guides and use a smaller stemmed valve for better flow. SBC is .343 for example.

Another consideration: Fill in the combustion chamber using brass brazing rod till the chamber is ½ inch deep. Use an end mill to bring center depth to .575”. Finish the chamber edge with a radiused cutter. Chamber volume should be about 85 cc giving 8.5 compression ratio with a flat top piston. You can sink the spark plug deeper into the head bringing the tip closer to the center of the chamber. Install valve seats compatible with unleaded fuel. Install longer but thinner valves to make up the difference in chamber depth. Attached picture shows similar aluminum 263 chamber with SBC valves.

What compression ratio are you planning on using?

What octane gas would you be using?

 

[Jyrki]

I now have setup that gives a 2.25:1 overdrive. Using the maximum overdrive and bleeding off excess boost at lower rpm's might be worth considering :thumbsup:

In my current engine I'm using BB Chevy EXH valves as intake valves, as they were the same stem diameter and length as 320 Buick. SBC valves aren't long enough for the stock 320 head. Then again, I'm sure they make thin-stem BBC valves, too.

Actually, I have been thinking about using bronze inserts to fill the 320 "open" chamber and make it "closed", then using flattop pistons. That would provide some "squish" to increase combustion speed.

We currently have two grades, 95E5 and 98E5, both of which contain 5% alcohol. In the beginning of 2011, the 95E5 will be replaced by 95E10 containing 10% alcohol. These are RON octane, and 98 RON roughly equals 93 pump octane.

 

[pmuller9]

Jyrki, Your solution to Siamese port EFI for a non modified head is the best I’ve seen.
I have learned a lot in our exchange.

With 2.25:1 overdrive, will manifold pressure need to be bled off at higher rpm to keep it at 15 lbs?

In my original chamber design the valves were kept deep and there was a squish area making the chamber into a heart shape. I took it to a shop that does NASCAR heads for suggestions. They said: Because the valves are so deep in the chamber, the extra shrouding of the squish section really restricts airflow around the head of the valves. Also because the chamber diameter is small there is not enough flat squish area to be effective. They’ve had best results with a shallow open chamber with a shallow, round dish in the piston. The result is the chamber shown in post #37.
The other part of this is getting the spark plug closer to the center of the chamber.

What is the compression ratio going to be?

What are you going to use for an intercooler?

I will tell you more about our 263 build-up including the billet head next reply

Paul

 

[Jyrki]

Paul,
I have no experience with the charger yet, so everything's just guessing. Procharger won't let out any compressor charts, unfortunately.

We should listen to what the NASCAR guys are saying. I need to reconsider the chamber design. On the other hand, this is not a high-compression NASCAR engine. But I see what they mean by shrouding the valves.

I'm aiming at 9:1 compression. Procarger says it's doable with 15 psi intercooled design. And, the ambient temps in Finland are so much lower...

I'm thinking of using a cheap Chinese intercooler, the biggest I can fit in.