322 Cylinder Head Flow Specs

[Beemon]

Has anyone flow tested a stock 322 cylinder head and have the numbers lying around? This semester at the University, we're working with fluid dynamics in Solid Works and I'm planning on creating a EFI intake manifold for the final project. Any help will be greatly appreciated!

[old-tank]

42 minutes ago, Beemon said:

I'm planning on creating a EFI intake manifold for the final project.

I don't know the answer to your question, but will look forward to the project.  I had a short glance at a street rod with a nailhead and the injectors and fuel rails were underneath and almost invisible. ..and very clean!

[Beemon]

I'm not sure how injectors would work under the manifold, form over function? I would want them top dead on the intake runner since the cylinder head port angle is so steep from the flange. Having them under the intake runners would most likely disrupt flow, rather than on top where they go with the flow. I'm looking to adopt the 5.3L LS type intake with the "hurricane" style swirl effect, and a 90MM throttle body. They are 324ci, so should be fairly similar. Looking forward to the project as well. The first example the professor used for introducing fluid dynamics was a basic venturi. Been there, done that...

[avgwarhawk]

 

[60FlatTop]

Pick up this copy of HRM and plagiarize. The engine at his right elbow looks like a shot.

 

Give your professor the stutters. Design the individual runners with an equalizer pipe. Feed the equalizer with a throttle body up to about 1600-1800 RPM. Then switch to port injection for higher revs.

There are a couple of tricks you can do in overdrive with low RPM if you put some mileage gears in. Once you work up the basics and steal a few other ideas it will get crystal clear.

[Beemon]

45 minutes ago, avgwarhawk said:

 

 

Too many variables to get right, much prefer one butterfly with everything up front. To get the individual ports to sync, you would need some type of MAP sensor, IAC sensor below each throttle plate and at least two TPM sensors to accurately feed all 8 cylinders.. then you would need to assume the butterflys all seal correct, and adjust both bank's idle position separately, fab up a wonkey throttle rod... with the front mount throttle body, it's one of each sensor feeding an ECU, then I could run a cable from the current throttle position to the throttle body. I know Hillborn is old school, but it's already been done to death. If I'm going to turn this into a final project, it's going to be a one of a kind. Maybe stamp WEIS on the top in the center, with 322 above or below that, or just omit the 322 altogether and have people ask "what the hell is that?"

[jackofalltrades70]

Weis supermarkets will love the marketing......cleanup in aisle 3.

[NTX5467]

The earlier single-port-ganged injection, as pictured, was considered the most powerful intake system in existence.  Everything straight into the port, using the length of the pipes to tune the rpm band.  Shorter = higher rpm, longer = lower rpm.  Lets call that earlier "single port" system "Single Port Throttle Body FI".

 

The "jets" in the housings put the fuel into the air stream at a right angle, which meant that dispersion would happen, no matter what.  In later years, "injector targeting" became more important with the injector aimed at the intake valve's head, physically.  BUT . . . there can be problems with that, if done strictly geometrically!!!  Reason is due to the flow dynamics of the port itself AND any affects of the boundary air flow with respect to the central-region air flow (the main flow itself).

 

As for the "beer barrel" intake manifold, GM LS-series, Chrysler LA Magnum, or whatever, it's a more compact version of the old Chrysler Ram Induction Manifold from 1960.  Just packaged smaller and in a "dry flow" environment of port fuel injection.

 

There ARE formulas for determining the "tune length" of the runners which go to the cylinder head, not specifically the total length.  In the "long ram" earlier versions, the intake runners were siamesed, with the divider between the head ports in the runners being what determined "long ram" or "short ram".  In order to raise the torque peak "bump", the manifolds were cut apart, that divider shortened, and everything welded back together of the resultant "short ram" manifold.  The total runner length was important as the longer runners were considered to reach a point of there being too much flow restriction in them if they were too long.  "Friction" in the air flow, I believe is how they termed it, back then.  In later times, it might be termed the boundary layer of the flow rumbling or getting in the way of the main flow (usually center of the port/runner).

 

In the '50s and '60s, much port work was done "by eye", as flow benches weren't generally used.  Sure, development engineers could tap various points on each port, but much of it was done by the way it looked.  This led to the round port Pontiac RamAirr III (I believe) round ports, plus the Ford Boss 302 round ports.  As massive as the RA III ports were supposed to be in flow, when somebody finally got around to flow-benching them in the 1980s, it turned out that they didn't flow any better than the RA II heads of normal port shape.  Until flow benches became somewhat common in the 1980s, when "flow dynamics" and "swirl" were really seen and researched, if it looked good, it had to work good.  To be sure, the round ports were usually seen as the "latest thing" in serious competitive racing, so putting them on a USA street motor was the "baddest" thing you could do to an engine for high-rpm power.  

 

It will be interesting to see what the flow numbers are on this historic Buick engine!  BUT, don't be surprised if they are not terribly impressive.  Which, in combination with the exhaust manifolding of those times (related to the width of the engine bay), might explain why a 322cid V-8 can do very nicely with an OEM 450cfm 4bbl carb.

 

The Edelbrock Tarantula series of intake manifolds were the first ones to seek to optimize mixture flow to each cylinder, to be more even when compared to the existing dual-plae intakes.  That was in 1969.  The 20% (maybe 15%) variation in air flow to the individual ports probably was not that much of an issue as the intake ports didn't all flow the same. In the case of big block Chevy motors, they were known to have "good ports" and "bad ports" on the intake side.  Conventional wisdom of the time meant you spent more time on the bad ports to get them to flow more, in porting time and reshaping.  When aftermarket cylinder heads finally came out with all good ports, it was discovered that they didn't tune the same as what people were used to, OR expected.  Reason?  The "bad ports" tuned to a lower rpm, as the "good ports" tuned to a higher rpm.  The broad torque curves suddenly became shorter and peakier, thus trickier to tune for max power.  And they ran slower in the quarter mile, usually, than the allegedly mediocre factory heads.

 

The LS-series cylinder heads are some of the best flowing stock heads in production, with various tweaks and changes over the years.  Cylinder head flow dynamics, combustion chamber dynamics, exhaust system flows, and what happens in the intake manifold make these engines run well and make the power they do.  In many earlier engine designs, the cylinder head ports were probably a limiting factor in power ultimate power production with the intake manifold and camshaft specs contributing to the desired broad power/torque curves of the 1950s-70s V-8 engines.

 

In the 1950s, there were MANY theories of combustion chamber dynamics that were being tried, but had limited results due to the fact that an understanding of what was actually going on didn't happen until flow benches were used by the NASCAR and drag race engine builders.  "Stratified charge" would equate to the later "swirl" or "active air orientation in a gasoline-fueled engine, but also applicable to diesels.

 

There probably is a CATIA program you can use to determine the  optimal runner lengths for various flows and power results.  From that, you can then determine how to best package it all.  Plus work at injector location too.

 

Best of luck and keep us posted on your progress!

 

NTX5467

[Beemon]

The program we're using, SolidWorks, is a cad specific program that is used to first develop the part, then bench test the part given a specific material (density) and it's characteristics. You can simulate loads, fluids and actual assembled systems to see if the materials will fatigue or, in terms of fluid dynamics, flow. By imputing certain lengths as variables, these flow dynamics can be changed on the fly until a desired result is found. If laminar flow techniques are used, such as an LS intake for example, then surface friction can be reduced to almost nothing.

 

As I understand it, wall friction in cast intake manifolds is usually where torque comes in? Where broader torque ranges are accomplished by disrupting the flow just enough until a certain RPM is reached where wall friction forces become negligible?

 

As we get further into the unit, I would like to have some numbers to play with. I might just end up asking my sister to drop one of my spare heads off at a flow shop near home to get some data. I proposed this idea to a group on Facebook, but they're more interested in the 401/425 engine... no love for the 322. Those numbers would be a baseline charge, hopefully within +/- 10% per intake port, to determine what the exit velocity of the intake runner needs to be to achieve max flow through each port. With a type of polymer material, laminar flow can supposedly be achieved and restriction plates could be added to disrupt flow at lower RPMs, kind of like the checker pattern found in Edelbrock intake manifolds.

[NTX5467]

I believe you're mis-reading some of these concepts.  Laminar flow is always there, to some extent.  Can't get rid of it, only minimize OR use it to your advantage.

 

In the early days of porting, "polishing" was also involved.  The orientation was that a smooth surface leads to more flow.  Then carb tuning went flaky, as I read it.  It seems that the existing laminar flow did slide easier on the smooth port surface, BUT it interfered with the main flow, so actual flow was discovered to be less than more.  They discovered that a slightly rough texture worked best, producing more power, as it kept droplets of the fuel mixture better dispersed into the main mixture flow.  So, the best performance was by deleting the "polish" from "Port and Polish".  With a "dry flow" situation, as with Direct Injection fuel systems, surface finish might not be quite as critical as with a "wet flow" situation.  But in the case of the cylinder head portion of the port, any fuel added to the air flow prior to the combustion chamber, would probably be considered a "wet flow" system, at least partially.

 

When the "wall friction" was being discussed, in relation to the Chrysler Ram Induction system, this was in the earlier 1960s.  A few years later, "other types of tuned intake manifolds were developed without the "long ram" design.  It was not long thereafter when exhaust tuning was considered as a complement to the intake tuning . . . as time progressed.  The "wall friction" was what led to the length limit of the tuned intake manifold runners, which stretched OVER the opposite side of the motor, with the carb plenum over the exhaust manifold.  That whole system was worth about another 30lbs/ft or torque in the 3000rpm range, which was a lot of "high tech scientific" horsepower back then.  Power peak was still about the same rpm as the camshaft did not change.  When the shorter and narrower 2x4 intakes of 1962 emerged in the Chrysler race inventory, that intake set-up could use bigger cams and carburetors for better drag strip performance.

 

The CATIA program, from the 1990s, was used to design the 2nd gen Chrysler V-6s of the late 1990s.  The Chrysler 300M-generation 3.5L and the smaller 2.7L V-6s.  The engines were done completely on disc, tested, dyno tested, all by the program before any engine was ever put together for the assembly line.  I suspect there are more advanced programs by now.

 

The "checks" in the plenum of intake manifolds, even troughs or ribs, probably have several purposes.  One, if exhaust heat is present, is to give the area a little more surface area with which ensure the mixture is heated for best cold ambient performance.  It can also help keep the larger fuel droplets get back into the airstream, especially at low vacuum/high flow times.  I somewhat doubt there is much laminar flow in this area of the intake as it's where the mixture changes directions AND is pushed/pulled by pressure pulses related to valve timing events.  There would be a great possibility for laminar flow in the runners of the manifold and the intake ports of the cylinder head.  Where the mixture is flowing more ONE direction (although reversion can exist during valve overlap times).

 

SO . . . if there's a shop with a flow bench, you can do some experimenting.  Have them do a flow with the "unmolested" stock ports.  One with an NHRA-legal port job and a polished texture on the port.  Then use a rougher grit to simulate the finer texture of modern aluminum castings.  Then get readings at .050 increments of valve lift, up to about .450" lift.  There's a standardized pressure drop to obtain the air flow figures.

 

Rather than deal with a range of values for air flow and such, why not do the figures (in the program) for a 322 V-8 at 3000rpm, probably figuring about 70% volumetric efficiency?

 

Enjoy!

NTX5467

[Beemon]

I wouldn't be looking for range of flow, I'd be looking for the maximum flow so the intake runner can flow the max, if not more, air than the heads. Once I have the max flow through the intake, I can simulate air flow through the intake in the program. As I understand it with the Nailhead family, the angle between the intake and head is almost two 90 degree angles as the mixture makes the S-turn into the cylinder bore. The second limitation, of course, is the intake valve and then the cylinder head runner. After doing a bit more reading, since the setup would be a "dry flow" design, the wall of the intake manifold is regardless as the mixture starts at the injector. What would be really interesting, would be rifling the intake manifold to create a vortex into the cylinder head for added atomization. I'm sure this has been done before... my "in-depth" knowledge of automobiles internal working is limited to my Buick, and a 2002 Jeep Liberty. Always a learning experience, of course. Tomorrow I'm going to head on down to the library to retrieve as many journals on the subject as possible, if available.

[avgwarhawk]

15 hours ago, Beemon said:

 

Too many variables to get right, much prefer one butterfly with everything up front. To get the individual ports to sync, you would need some type of MAP sensor, IAC sensor below each throttle plate and at least two TPM sensors to accurately feed all 8 cylinders.. then you would need to assume the butterflys all seal correct, and adjust both bank's idle position separately, fab up a wonkey throttle rod... with the front mount throttle body, it's one of each sensor feeding an ECU, then I could run a cable from the current throttle position to the throttle body. I know Hillborn is old school, but it's already been done to death. If I'm going to turn this into a final project, it's going to be a one of a kind. Maybe stamp WEIS on the top in the center, with 322 above or below that, or just omit the 322 altogether and have people ask "what the hell is that?"

 

 

 I do not see a need to have multiple sensors for this set up.  But that is neither here nor there as you want to do something different.  Perhaps individual coils and injectors for each cylinder as seen today. It would take some engineering.  Crank trigger and all sensors hooked to the correct ECM.   Would certainly be different and a lot of modifications to the existing intake!   

 

Maybe just a throttle body with two injectors as seen on some 80's vehicles.    

[60FlatTop]

Take a look at the operation of 1976 Cadillac Bendix fuel injection. It is a classic example of complicating a process by monitoring and reacting to too many variables. Really study it. It would be a great learning experience.

Then compare the simplicity of the evolution that minimizes the process to a few relatively simple components. Don't forget the stoichiometeric cruise control mode, it is neat.

Take into account that a carburetor is only a "pot" of fuel on top of the engine with a few holes that fuel gets sucked through, not a very scientific or predictable process, but worked in a wide range of technologies for over 100 years. It ain't rocket science,  they designed the Bendix system.

 

The flow information you ask for is pretty easy to calculate with minimal information. Send me a PM with your email and I will send you a reverse engineering guideline I did for flow and capacity of steam heat exchangers when specs are lost. I think you can apply the thought process.

Bernie

[dei]

Enjoying this thread guys.

With having my son soon finishing his Masters in Mechanical Engineering and able to stay at home, have had more than a few conversations with him about some of the things he has been dealing with.

I've only had High School Auto Class and a lot of old school Back Yard experience but enjoy learning about how things are evolving.

 

Ben, just don't put out too much info down the road or someone might take your ideas, run with it and claim it as their own... 

(A revolutionary idea might just be worth patent rights?)

[60FlatTop]

Patent rights are a myth started by the conniver Ben Franklin when he wanted the Potash family's fertilizer recipe (Patent #1) to feed a new nation.

 

If you have a good idea just make it and don't tell anyone the details. And don't enter a partnership when you have skills the other partner doesn't. Once, due to a management change, a large engineering company demanded I sign a non-disclosure agreement for a process I designed and had to teach them to do. I refuse to work with them.

 

Patents are expensive. I had two for refrigerant recovery and reclaimation. There are more application and maintenance fees than Robin had Merry Men in Sherwood Forest. In ten years I had about $40,000 tied up in them. And a major tool provider to GM uses my non-condensible gas discriminator on their machine. The legal burden for proof of patent infringement required I purchase one of their machines as evidence. I walked away rather than spend another $15,000 on their machine and more fees.

 

I highly recommend you avoid any "sharing" of profitable ideas no matter what thin veil of protection you perceive.

 

Most importantly, remember, it you come up with ideas and the other guys don't, you won't just have one. But they had none and had to steal yours. That is important. Don't hold all your ideas so closely that no one knows your capability. Put a couple out as seed. Just be careful with the really good ones.

 

Mediocrity will look like brilliance to many of the duds (not dudes) in the engineering community. Seed with mediocre ideas and watch their amazement.

Bernie

[Beemon]

6 hours ago, avgwarhawk said:

 

 

 I do not see a need to have multiple sensors for this set up.  But that is neither here nor there as you want to do something different.  Perhaps individual coils and injectors for each cylinder as seen today. It would take some engineering.  Crank trigger and all sensors hooked to the correct ECM.   Would certainly be different and a lot of modifications to the existing intake!   

 

Maybe just a throttle body with two injectors as seen on some 80's vehicles.    

 

With a Hilborn injector, you need to assume all the throttle plates are set up the same, and assume that air flow through the stacks are all the same. Because it's mechanically injected from the high pressure fuel block, you must also assume each cylinder is getting the same amount of fuel. Depending on if/where the MAP and IAC sensor is, it will measure from only one cylinder to create a fuel charge for that cylinder, on all 8 cylinders, which could be inaccurate. There's just too many assumptions.

 

The whole purpose of the project is to not re-use and modify an intake, or buy another intake, but to make a brand new intake from scratch. I've been studying ram air theory, and all I can really say thus far is I'm glad I got my cam degreed. Because the intake manifold will be dry-flow, port injection, the intake runners can be tailored to allow pulsations and simulate forced induction at a specific peak curve dictated by the cam (and of course the flow of the heads). Air comes in through the throttle plate, leaves the plenum and swirls through the runners and meets the injector and goes into the valve. Valve closes and the rush of air pulsates back through the runner, rebounds off the plenum and goes back down in a rush back to the intake valve with another rush of fuel. Cam timing is important, to find where the peak curve is, and dictates how long the runners will be. At peak, you would want the pulse wave to return to the intake manifold exactly when it opens again. This can all be simulated in the program. Finding a peak RPM for power would be pretty simple I think, depending on driving habits and where, say 35-40MPH is, maybe around 2400 RPM? Lastly, crank position sensor would be in the distributor.

 

Bernie, will like to see those equations. Thanks for offering, I'll be sending a PM. I'm reviewing my Thermodynamics book, too, to see what I can gain from heat pumps and the otto cycle.

[NTX5467]

You want "swirl" in the intake runners?  Several add-on carb spacers have tried to accomplish that, key word "tried".  The main place you want "swirl" is when the air enters the combustion chamber.  The whole stratified charge issue is designed around that and the related layers of mixture density.  I believe the Mitsu 3-valve system (1980s) is supposed to have some relation to that, too.

 

There was one model year of 5.0L and 5.4L Ford V-8 that intentionally shrouded the intake valve (ala Ford Y-block) in order to induce swirl in the mixture as it came into the chamber,  but it was ONLY for that one model year.  The shrouding was on the outside edge of the chamber.  The classic Chevy 4" bore cylinder head normally used either 1.72" or 1.94" intake valves.  2.02" intake valves can also fit in that chamber, but the edge of the valve is right at the edge of the chamber, resulting in not much additional air flow from the larger valve.  The "fix" was to put a recess in the side of the head's chamber casting so the valve was not shrouded.  This is how you tell a factory 2.02" cylinder head OR one that's been converted.  Just be careful of any head with 2.02" intake valves and no such machine cut in the edge of the chamber.

 

Trying to put swirl in the intake runners might make for a soggy throttle response.  Too much "spring" action in the air column.  That air would be tired by the time it stopped going in circles by the time it gave an unenthusiastic "boom".  Nor would it use that extra length to boost low end torque, as the physical port length relates to that, not how far the air in the port might travel.

 

The Hilborn system was very crude and only really worked well on drag race engines.  Something of a "controlled flood" fuel supply.  The pressure pump had a "pill" that controlled the return fuel, and the fuel amount that went to the injectors, as a result.  The things which made it "imprecise", as mentioned, didn't really matter in the real world of a drag race car or boat.  You read the spark plugs after each run, making adjustments on the fly, by changing the pill to get the best spark plug insulator color.

 

Check out the "High and Mighty" Plymouth drag race car built and raced by The Ramchargers (a group of Chrysler engineers).  Quite an interesting story and vehicle!  Probably some information at Allpar.com.  There's probably some information on the Chrysler Ram Induction intake manifolds on the Chrysler 300 Letter Cars on that website, too.  The tuned length formulas are pretty straightforward, as I recall.

 

Enjoy!

NTX5467

[EmTee]

https://itstillruns.com/tuned-port-injection-5398218.html

[NTX5467]

I believe the FI systems on the 3300 V-6 Buick was also called "tuned port".  Didn't look like the Chevy system, but more like the earlier Rochester FI system on '50s-'60s Corvettes.

 

The '80s Chevy TPI system also has some aftermarket versions that had more air flow capacity for higher power levels  Even chrome!  The later "beer barrel" style intakes (circular) is a different method of packaging.

 

Thanks for that link!

 

NTX5467

[Guest]

Might b a resource