Fabbing sheetmetal intakes?

[DocsMachine]

Newly delurked guy with a question:

Does anyone have some good info on the whys and wherefores of fabbing sheetmetal intakes? I'm a decent machinist and I'm learning/relearning the TIG under some professional guidance, but I'd like to find out some more info, preferably from someone who has already done it.

The application is an EFI manifold for a daily driver, so things like plenum volumes and runner sizes aren't quite as critical as they'd be for a pro-stockers' tunnel ram, but I'm still sweating things like building in proper clearance for gaskets, should I bother with a heat crossover, and will I have corrosion problems from the coolant? (Iron block, copper radiator.)

As well as things like how thick should I make the gasket faces? 1/2"? Will 1/4" flex too much to seal well? Is 3/4" too thick? (I know it'll be a pain to weld...) Is injector angle critical? Should they spray right at the valve, or is "pretty much anywhere into the runner, heading in that general direction" okay?

Thanks for any and all replies.
Doc.

[GPster]

Welcome, One of the things you'll lean here is that with the number of people here someone has usually tried that but it might help if you are more specific. What kind of engine? Gpster

[DocsMachine]

Well, I thought it was more a fabrication question than an application question, but on rereading I note I didn't even mention it was a V-8, so yeah, a few details couldn't hurt.

I have a '66 Toronado (that I've been working on for considerably longer than Jay Leno has been doing his, but I'm still gonna be called a "copycat" ) with a '68 455 that I want to convert to EFI. The problem is the car's very low hoodline- there's almost no room for a bolt-on TBI injector like a Pro-Jection, even the 2-barrel unit, since the stock intake is literally "sunk" in the middle.

Here's a pic- the hood insulation actually compresses a touch on the air cleaner hold-down knob. I have just under 3-1/2" from the peak if the valve covers to the hood.

Nobody makes a port-injector system for the Olds (not off-the-shelf anyway) and even if they did, it wouldn't fit under the hood.

What I plan to do is fabricate an intake that looks something like a wider, lower version of the Chevy LT-1 intake. (Pic stolen at random from Google image search.)

Mine will be a bit lower and almost 4" wider, simply due to the width of the 455. But the layout will be about the same- central plenum/doghouse, front-mounted twin-nostril LT-1/TPI type throttle body, runner-mounted injectors. I'll be using the MegaSquirt DIY controller to run it in closed-loop standalone, batchfire mode.

Due to the location of the throttlebody, I plan in moving the thermostat down and to the driver's side, in what will probably be a custom-made housing. I plan on using an aftermarket 52mm to 58mm throttlebody, and I found a couple of sources for both weld-in injector cups and extruded fuel rail. I can make both from scratch, but the prefab stuff is cheap enough to buy.

As far as the fabrication goes, I have in mind, at the moment, to make the cylinder head faces from 3/4" 6061 flatbar simply for the thickness and rigidity. It'll be a bear to weld, but considering the bar will be seriously reduced (port passages and bolt holes, etc) by welding time might make it a bit easier.

I wondered if 1/2" were thick enough to seal properly- it'd be easier to weld, definitely, but also more likely to warp either during welding or later in use. Some of the photos of pro-stock type intakes I've seen online appear to have 1/2" or even 3/8" faces, but then again, they're racers, regularly rebuilt and serviced, while mine will be a long-term daily driver.

The runners will be 1/8" wall extruded 1-1/2" by 2-1/2" rectangular tubing, which internally almost perfectly matches the dimensions of the stock ports. The base valley cover, most of the plenum and other parts will be 3/16" sheet 5052, with a few 'billet' 6061 pieces, like the TB face, threaded water temp bungs and other bits.

I'm worried about making sure I allow for gasket clearance, and/or leaving enough room that I can have the faces surfaced after all the welding's done to take care of any warpage. I'm worried about placing the injectors at the proper angle, and wondering how much volume I should give the plenum. I know it's "dry" air, but should I minimize volume? Will it matter? Will a smaller plenum give me a better vacuum signal?  (Power brakes, trans modulator, etc.)

For that matter, where can I pull "ported" vacuum for the distributor advance? Just thought of that one, lemme write that down...

So actually I guess it's a two part question; manifold theory and design, and actual fabrication techniques.

Doc.

[enjenjo]

I don't think you would have a welding problem with 3/4" material, by the time you are welding most of the vertical cross sections will be less than 1/2" It will be tough to control warping of the head faces, so I would allow about .125 to be milled off to the finished size.

On the injector positioning, it is my understanding for several reasons you want the shot directly to the backside of the valve if possible. By starting with 3/4" material, it might be possible to machine the bungs right into the plates.

On the ported vacuum for the distibutor, why not go full electronic advance on the ignition? You will have most of the sensors needed to control it in place already. Seems to me it would be fairly easy to adapt a Caddy Northstar ignition system.

[GPster]

Just some ideas to get the discussion going. I personally think that 3/8" plate is enough but there might be some suggestions I would make to the process. Find yourself a dummy block and some dummy heads. Maybe if you do all your fabricating on a solid work surface things will stay where you put them. You might also put the heads on the block with 2 sets of gaskets or something else to use as a spacer. When you get the manifold made you can have it surface milled like they do with intake manifolds after they mill heads to get everything back together. If you start bigger than maybe you can mill the manifold to standard and straight. One other suggestion. If you're going to computors with the fuel injection maybe you can look for some kind crank fired ignition. Maybe if you remove one more piece off the top of the motor you'll have more room under the hood to play in. Now I have to finish my dinner before CHAT TIME. GPster

[DocsMachine]

I hadn't through about milling the injector pockets into the plates themselves... Interesting. Of course, I haven't picked up the injectors yet and  haven't even moved into the mocking-up-with-cardboard stage, but that's a great idea. Would you happen to know the correct dimensions for a GM injector pocket?

The Olds has a pretty wide runner spacing, wider than an SB Chevy, so the wall between runners will be fairly big- thankfully, I'm not sure how I'd weld 'em if they were any closer.

On the ignition, you have to realize I'm not an electonics whiz. I can get the batteries into a flashlight the right way two tries out of three if I'm given enough time to practice beforehand. As far as factory ECUs, I'd have little or no documentation, and no laptop support, besides needing additional sensors like knock and crank-position.

The two things I like about the M'Squirt setup is that it's well documented (you can even compile and flash your own code if you want) and has a built-in laptop interface for tuning on the fly and recording.

They're working on an add-on upgrade to run electronic ignitions (like the 7-pin HEI, or Ford's distributorless) but as I understand it, that system isn't ready for release yet. I'll probably upgrade to that when it's available (they're also working on a wideband O2 sensor interface) but for the moment I'd like to keep it relatively simple. Just fabricating the manifold will be farily involved undertaking.

Doc.

[unklian]

Any of the race manifolds I've seen,
have runners that taper in cross section.
These runners are made in 2 pieces,and hammer formed,
then trimmed and welded.

For the street,it's probably not critical.

Longer runners will improve low speed torque.

Optimum plenum volume is probably related to displacemnt.

[DocsMachine]

GPster- I have a partly-dismantled junkyard block that I plan on doing the fabricating on. It's a rebuildable core, but filthy- Won't have any problems dropping shavings or aluminum dust into it, it needs tearing down and rebuilding anyway.

If I were to build the thing to the block and heads directly- as in no gaskets- that would leave me some room to have the faces milled back to straight, wouldn't you think? Stock Olds intake gaskets are just pressed sheetmetal with a valley pan though... I wonder if I can get composite gaskets or a Fel-Pro with the rubber lines? The sheetmetal might only be ten or fifteen thou.

Doc.

[enjenjo]

I hadn't through about milling the injector pockets into the plates themselves... Interesting. Of course, I haven't picked up the injectors yet and  haven't even moved into the mocking-up-with-cardboard stage, but that's a great idea. Would you happen to know the correct dimensions for a GM injector pocket?

I don't know the dimensions, but you can rent a cutter from Kinsler injection.

The Olds has a pretty wide runner spacing, wider than an SB Chevy, so the wall between runners will be fairly big- thankfully, I'm not sure how I'd weld 'em if they were any closer.

On the ignition, you have to realize I'm not an electonics whiz. I can get the batteries into a flashlight the right way two tries out of three if I'm given enough time to practice beforehand. As far as factory ECUs, I'd have little or no documentation, and no laptop support, besides needing additional sensors like knock and crank-position.

The knock sensor is no problem, GM has at least a couple, one that is 1/4" pipe thread, and the other is 3/8" 16 thread. there are others that can be used. For the crank trigger you can use one from MSD or the like. There are at least a couple companies that have stand alone CCCI ignitions available.

The two things I like about the M'Squirt setup is that it's well documented (you can even compile and flash your own code if you want) and has a built-in laptop interface for tuning on the fly and recording.

They're working on an add-on upgrade to run electronic ignitions (like the 7-pin HEI, or Ford's distributorless) but as I understand it, that system isn't ready for release yet. I'll probably upgrade to that when it's available (they're also working on a wideband O2 sensor interface) but for the moment I'd like to keep it relatively simple. Just fabricating the manifold will be farily involved undertaking.

From what I have discovered while working on EFI, it's not as complicated as it first seems, particularly batch fire. Just research each system as you are doing it, and pretty soon it becomes clear.

[PeterR]

On the injector positioning, it is my understanding for several reasons you want the shot directly to the backside of the valve if possible

On sequential fire systems this seems logical because every injector releases the fuel into the moving air stream. However I have often wondered if this is true for batch fire where every cylinder will have different fuel timing and some will even have the fuel fired into dead air.  Under these circumstances it would seem sensible to have the injector further away from the head to promote mixing.

There may be some other variable I am missing, so Choco et al can you enlighten me.

[Sean]

The Olds has a pretty wide runner spacing, wider than an SB Chevy, so the wall between runners will be fairly big- thankfully, I'm not sure how I'd weld 'em if they were any closer.

Tig Torch for tight spots... http://www.esabna.com/ESAB/showdetl.cfm?&DID=8&User_ID=523851&st=3668&st2=71225618&st3=48199375&Product_ID=1652&CATID=9 used with a Button style cap, this should get into some pretty tight spots.


They make them even smaller than this, with a longer reach. We have a head thats on a 6" long flexible neck for really tight spots. Its kind of a pain to use because the Tungstens for it are only about 1/2" long and the Cup is very delicate and easily broken. There have been times we couldn't have done without it though. The one we have would not run anywhere near the amperage needed for heavy aluminum however.

[sirstude]

As long as you are looking at the MegaSqirt stuff, take a look at the MegaJolt site also for your ignition.  There seem to be a couple of options there (and I have not looked for probably a year) one useing all their stuff, and another adapting some direct fire stuff from one of the Ford V8s.  

Doug

[Bruce Dorsi]

On the injector positioning, it is my understanding for several reasons you want the shot directly to the backside of the valve if possible

On sequential fire systems this seems logical because every injector releases the fuel into the moving air stream. However I have often wondered if this is true for batch fire where every cylinder will have different fuel timing and some will even have the fuel fired into dead air.  Under these circumstances it would seem sensible to have the injector further away from the head to promote mixing.

There may be some other variable I am missing, so Choco et al can you enlighten me.

================================
I have read somewhere that aiming the injector discharge at the back of the valve uses heat in/from the valve to promote vaporization of the fuel.

While a batch-fired system may not be as precise as a sequential system, fuel introduced in a metered amount, near the valve, should be more exact than mixture/distribution variations from a carb on a manifold.

[DocsMachine]

I found a reference earlier today that mentioned developing an EFI system, and having to work at keeping the fuel from puddling on a port wall due to the spray angle.

Interesting point on the spray cooling the valve though... that's an interesting bit of data.

I found a reference on the injector ports as well- the end cup dimensions are .530 to .535" ID. Doesn't give a depth, but from the diagram it looks like it could be anywhere from .250" to .500". The fuel rail pockets are the same. So it seems I can just use a 13.5mm endmill, which is easy enough.

Stude- I saw the M'Jolt on the 'Squirt site, but hadn't looked through it yet. At the moment I wanted to keep things simple- I already have an HEI on it, and that works fine. I just hadn't thought about the vac advance until just now. If necessary, I'll convert to the electronic advance HEI, but for the moment, I need to concentrate on the injection itself.

Sean- Interesting torch. One of my Richard Finch TIG books has a pic of one smaller still, though it might be hard-pressed to weld 3/4".

I still need to get the materials in and mock the thing up in realtime before I even know what sort of room I have. If it works the way I have in mind, I may be able to weld both ends of the runners from the inside, no external welds.

Doc.

[river1]

Sean- Interesting torch. One of my Richard Finch TIG books has a pic of one smaller still, though it might be hard-pressed to weld 3/4". Doc.

if you're needing to weld thick stuff in a tight space you might look at water cooled torches. they are pretty small for the amperages they can handle.

later jim