Intake Manifold project-turned-nightmare

[Xnke]

I've got this posted up on HybridZ too, but figured you guys might like a look at the job. I am currently running a hybrid intake made from an L28 fuel-injection intake lopped off about 1.25" from the head, then welded to the twin SU intake after the flange was cut off. The SU's were replaced with a pair of Mitsubishi Eclipse throttles, so it's a twin 50mm throttle system.

 

This works fine, from about 3500RPM and up. Below that, I think I'm seeing bad reversion effects and charge robbing on some cylinders, as the car has to run extremely rich at idle and will foul plugs on four cylinders, while the other two remain quite fine. I've swapped injectors around a few times, and the plug fouling doesn't follow the injector...Note that the engine will only foul plugs at idle, once you're driving it keeps them clean. Leaning out the idle results in poor running and lean snaps. All this goes away at higher RPM.

 

So, to eliminate charge robbing and to improve the overall cylinder-to-cylinder distribution in preparation for 8-10lbs of boost, I've started a new manifold project. Here's the outline:

 

The manifold is a fabricated aluminum part. I have the ability and facilities for aluminum castings up to 20lbs on site; and *may* work with that eventually, but not for now.

 

The runner ID is 38mm, and was chosen based on where I want peak torque to fall. I don't buy into the "manifold length is the torque determining feature" because the data shows otherwise. Manifold length should be chosen to complement runner ID in my opinion. For an L28, 38mm ID runners will provide a torque peak at 5300RPM, and will have a port velocity of 200ft/second to 520t/s from 2800RPM to 7000RPM. This is within the ideal range, although on the bottom end we're starting to see the velocity drop off. To complement this, a runner length of 15.5" was chosen, as this is close to the 3rd wave length for 5300RPM, and has usable nodes from 3800 to 7400RPM, coinciding nicely with the goals of the project.

 

The absence of charge robbing will be pushed for by making this an ITB intake, using throttles from a FWD VG30e Maxima, as they are plentiful in the parts yards, even if they are a little oversize. I'd prefer 40mm throttle plates, but these are 45mm. They'll do. At 10 bucks each, I'm not going to shell out nearly 400$ per each from Jenvey for 40mm units.

 

Injector ports are currently being fitted, and it's a right pain in the rear to get the TIG torch down in the tight spots. It's currently giving me fits, but hopefully a gas lens setup will help with this...it's on order.

 

We'll start here, with the plenum:

 

 

Yep, those are fire extinguishers...

 

 

 

Then we press a table leg into service, wrapping it with aluminum to form a secondary plenum assembly from 2.4mm sheet:

 

 

Got them welded up, the main plenum pulled pretty good in the middle. Fix it later.

 

 

Fitting up the throttle plate/velocity stack plate:

 

 

Laying out the cylinder head flange:

 

 

The runner tubing is fit into hole-sawn holes, cut with both a 38mm and a 41mm hole saw in a drill press. The ports are cut in at a 15* angle, and if I had done my math right, they would line up with the bottom of the OEM port. They don't...I'll have to lift the manifold about a mm.

 

 

 

Well, let's have a go at it, shall we?

 

 

 

 

Yep, welding moves. A lot. 2mm!

 

 

Before straightening up, this is the "concept" shot.

 

 

Now, that's done. Let's get started on some "velocity stacks"....I hesitate to call them that, because they don't have as much radius as I wanted. Not enough room between the throttle bores. But, they are tapered in, and have a 1.5" radius cut into them. I really don't like the thin edge there, but they're already in and I can't change it now.

 

 

 

Plenum assembly!

 

 

I'll put up the rest of the main runner assembly in the next post.

[Xnke]

Got all the bits straightened and machined flat, here's the mockup fitting to make sure I haven't totally goofed up:

 

 

 

Going to have to fit up a throttle shaft extension here to correct for this...

 

 

More photos of ugly welds this evening, I'm having a rough time of getting the injector bungs welded into place.

[garvice]

I really enjoy threads like these. Looking forward to end result.

[Sirpent]

Welcome from across the pond Bud,

 

Peter Mc will be along sooner or later and see this thread, and apart from a number of other guys on this forum he will probably give you the best opinion on what your up to and aspiring to.

 

Cheers

 

John

[Xnke]

Good. I'd like a professional (or even semi-professional) builder's thoughts on it; our pros back at Hybrid have gone into retirement/hiding for now...they're hard to come by, anyway.

 

These injector bungs look like shivere. Two of them are going to need touchup machine work, which bites. Maybe I'll get some heat in the shop tonight and see about welding up the rest.

 

Here's a little extra about the project:

 

[KatoKid]

I like where this is going!

[Six_Shooter]

I like where this is going!

 

Me too!

 

I love hand fabbed parts, especially intakes and headers.

 

I'll have to look into the throttle bodies you mention for another 6 cylinder project I am going to build.

 

What was originally attached to the inlet of the throttle bodies?

 

Keep the pics coming.

[d3c0y]

What sort of engine management are you going to run on it?

[Xnke]

The original inlet of the throttles has a 90* turndown on it, that goes to the OEM nissan airbox. I thought about trying to figure a way to use it but it was going to be a lot of trouble for less than optimal usage. I left them at the junkyard in the passenger seats of the Maximas I pulled the throttles from.

 

The engine management is handled by an MS1extra controller, and an EDIS ignition controller. I am currently running this on the car now, and about 95* of all the bugs are worked out.

 

The new installation will let me fix all the remaining bugs, which are mostly wiring related. The wiring harness will be split into a cabin harness and an engine harness, as to make the wiring much much less convoluted. Also, the harness won't have to be wrangled out of the car each time the engine gets removed.

 

Every time I go out to the shop to do some welding, it's so cold my breath fogs the mask and I can't see what I'm doing...

[jamo240]

That's tough Xnke...it's 100 degrees here today...no risk of mask fogging!

 

Certainly an interesting project to take on...will be interesting to see what the dyno figures reveal about your design.

 

Cheers

 

 

Jamo

[peter mc]

Hi Xnke

mate have you tryed to tune a itb style boosted engine before , it can be a real bastard , after tuning many gtr engines i have found many ecu have not got what it takes , ,,

 

love the manifold good luck with it all ,

pmc

[Xnke]

Yep, done it before. At the specific power outputs we're looking at here, it shouldn't be an issue. That said, I am fully willing to accept ANY advice from someone who has worked with the GTR boosted system at the level you have. A GTR though, is a whole 'nuther beast. You're looking at 600HP minimum, I'm looking at 300HP maximum....the max this particular supercharger (eaton M62 from a series 1 3800) can push is about 270RWHP. I would have to move to an M90 to get more.

 

As it is, I'm running a short-runner, large-runner-diameter, large-throttle-plate intake that requires a LARGE throttle cam (we're looking at a 5:1 ratio of pedal-to-throttle motion at the early part of the cam) to even be remotely drivable. Tip-in is awful. Idle vacuum is 55kpa best case, and the engine cruises at 75kpa. WOT is reached at 82% throttle position. Don't shift up under 4000 unless you really want to baby it. (makes in-town driving a pain.)

 

4000RPM and up and this intake really shines....below that and it's a terrible compromise and tunes very, very similar to ITB's. ITB's with charge robbing. Worked perfectly fine for a track car with a head the flowed a little better than mine does; but on my daily no, it's causing me headaches.

 

If I'm going to drive with all the issues inherent in ITB's, then I might as well pick up the one benefit to them. Also, MS now has a new load system specific to ITB systems....I plan to give it a shot. If the new load sensing mode can't handle boost, then there is always Hybrid Alpha-N mode, which I have used before and it works very well.

 

Got any tips on preventing standoff or reversion? I'm worried that the straight-shot runners might cause problems with that. Thinking about porting the head to 39mm or 40mm (careful, careful!) to get an anti-reversion step, as suggested by Racer Brown in the article reproduced on Datsport. Head is a P90, and I've done a little port work before. I still need to practice some on my exhaust ports, but I think I've got the intake down. Flowed an E88 head casting to 221cfm at .540" lift, and only lost a very few cfm under .300". The P90 port looks like it'll take less work to get there.

[peter mc]

Hi xnke

i think reversion wont be a problem with that length inlet  ,, the cam plays such a big part as does the exhaust manifold  , last year i engine dyno a engine that was spraying fuel 3 feet out from the carbs and we changed the exhaust runner length 2in and it stoped , its not in the inlet port from all the testing we did last year , all my heads have 39mm + ports and never have steps in them , things to note short runner manifolds with 300d+ cams  and long primry pips will have reversion , you need at least 13in from valve to trumpet ,

 

as for the gtr even stock engines that make 280hp are very hard to tune with some ecu , the map signal is so week and the tps can be is 5 places with boost  and vacuum , at part throttle there a real thing to tune ,

[Xnke]

Good to know. 15.5" from the back of the valve to the runner opening, so we're a little longer than the 13" you mention. Header is this one, built in my garage. Total length to final merge is 34", primary length is 20". All tubes are within 1/8". Header tubing is 1.625", 2.5" collectors, and is mated to a 2.5" exhaust. Header was intended to be mated with a 3" exhaust, however I haven't managed the packaging constraints yet.

 

 

Camshaft is a Isky custom profile; 0.480" gross valve lift, 280* seat-to-seat duration, 232* duration at 0.050" lift. Lobe separation angle 113*, intake open 25* BTDC, closes 75* ABDC, exhaust opens 71* BBDC, closes 29*ATDC, Overlap 54*. Camshaft has not been delivered yet, but has been ordered and a core delivered to be ground. I will be calling tomarrow to see where it is. Hopefully it'll be better for the application than the current cam, which is a similar lobe profile cut on a 107.5* lobe seperation angle. Current cam has shit-for-all vacuum at idle, and was generally a poor choice for a street application.

 

Compression ratio is *supposed* to be 9.7:1, however I am reading conflicting reports from my construction logs. Compression may be lower than desired, however that can be easily, if expensively, fixed. Current engine is 79mm stroke, 88mm bore, 133mm rod, 9.32cc dish in piston, and moderately ported N42 head, shaved 0.030". Calculations showed that this *should* have put chamber volume at 42cc, however I need to measure the head thickness again and be extra certain. If I'm lucky, then the compression will come out at about 9.0:1, and we'll see what we will see.

 

Target boost is 8PSI, intercooled. Compressor will be 65% adiabatic effeciency from 2000RPM til 6500RPM, dropping to 60% or so at redline. Volumetric efficiency of the compressor will be 88% or higher over the same range. Hoping to get at least 80% volumetric efficiency over this rev range in the engine; if I manage that then I'll put the power goal to the ground. (assuming the transmission will handle it!)

 

As far as the tuning goes, the Hybrid Alpha-N mode I mentioned uses both MAP and TPS to gauge load, using the same math the Ferrari F40 used. I've used it prior to this, and it's pretty slick, just takes more time to tune properly. Initial tuning will be done on some carefully chosen roads that have (measured) grades of 3%, 7%, 10%, and a 12% grade, each about a mile long, in a straight line. It's pretty easy to setup passes on each grade at incremental throttle position, then move to the next grade, so tuning is slow, safe, and steady. I think I can get at least 85% of the tuning done this way, before shelling out the money for dyno time. It's a 4 hour drive, plus fuel costs, plus 185$/hr to use the dyno, plus lunch for the dyno operator, so I like to make that time count.

[Xnke]

Finished welding in those *$%&#@^ injector bungs today, those little bastards are a pain. I would have been better off boring them through 10mm and forgetting the lower o-ring seat until the welding was complete, as it is I'll have to rig up a long-pilot counterbore to re-machine the o-ring seats. Had MAJOR problems with arc wander in one spot on every bung. Might see if it can be set up in a seat-and-guide machine...that might be the hot ticket.

 

Regarding the tuning, I DO have the option to run a hot-wire MAF sensor, there are several out there that are compatable with MS. Would it be beneficial to go with at MAF setup for load sensing, over TPS/MAP combo? The TPS/MAP combo can be rigged to run several ways...TPS as primary table, with MAP modifying, or MAP primary table with TPS modifying, or the mode that I've used before, which is used on the Ferrarri F40 system...TPS is primary, and MAP is multiplied in. In this mode, there is only one table to tune, the TPS table. MAP is linearly applied based on the current MAP value. This is the only mode that I actually *KNOW* how to use, and the only one I've been able to convert to run the spark map from the MAP value and not TPS. (this last bit seems important to me, since TPS-only won't be an accurate load metric under boost)

 

Hoping to get the interior of the manifold cleaned up and finish-ported this week, and figure out how to accurately re-machine the injector bungs in-situ...I don't have a milling machine and I'm not sure I can rig the lathe to mill a part this large. Probably going to rig up a custom piloted counterbore from some drill rod and call it a day.

[Six_Shooter]

Finished welding in those *$%&#@^ injector bungs today, those little bastards are a pain. I would have been better off boring them through 10mm and forgetting the lower o-ring seat until the welding was complete, as it is I'll have to rig up a long-pilot counterbore to re-machine the o-ring seats. Had MAJOR problems with arc wander in one spot on every bung. Might see if it can be set up in a seat-and-guide machine...that might be the hot ticket.

 

That's what I had to do with a Crower mechanical injection manifold for a SBC. The first attempt made the bungs useless, the second attempt, the bungs were drilled with 5/16" or 3/8" holes initially, and those holes were used to line up the bungs with the fuel rail, using some fabricated inserts. Then the bungs were drilled out after the fact to fit the injectors.

[Xnke]

So I got the counterbore made this evening, hardened and sharpened up. It's an old-style D-bit reamer, with a pilot on it. I use these things all the time in my gunsmithing stuff.

 

 

Got the injector bungs counterbored and test-fitted the lower injector seals.

 

 

More on the TPS sensor this weekend, when I get it fitted up.

[Xnke]

Only thing that has me hung up now is the PCV system. I *think* I've got a good way to work it figured out, but not positive yet.

 

Instead of having the PCV valve on the manifold, it will be on the catch can. The catch can will have full engine vacuum on it all the time, be mesh-filled and have an oil separator in place. The PCV valve will still be fully functional; but the can will act as a vacuum resivour as well as stopping most of the crud that would go into the manifold.

[Xnke]

This weekend I'll make the fuel rail mounts, and will get started with the fuel rail itself, I hope. All the material is here, but i've got a hung caliper on the datto and may have to split funding to rebuild the front suspension...You know, the while-am-at-it's will get me...

[Xnke]

Been busy with some other things; so it's a big update.

 

Got the throttles opening up together!

 

 

Fuel rail stanchions made...chunky things!

 

 

Rail mounted, but not quite finished...Still needs a regulator.

 

 

Feed line fitted up, and the throttle cable installed.

 

 

Fuel pressure regulator housing machined and welded onto the rail, with the regulator installed.

 

 

Just pop the circlip out...

 

 

Pull the regulator out of the housing...

 

 

BAM! ready to drop the new replacement in, OR quickly change fuel pressure in preset steps...2.5, 3.0, 3.5, 3.8, 4.0, 5.0 BAR regulators are available from Bosch...Or your local VW/Audi dealer!

 

Starting to get the fuel lines run here:

 

 

Compressor bypass vacuum control lines wrap around the cylinder head, like so...

 

 

 

And, short some vacuum lines, wiring, and IAT sensor, this is how it should all look:

 

 

 

Of course, we'll need more stopping power, since we're adding more going power.