Don Terrill’s Speed-Talk

I am hoping someone can give me some guidance on modeling a chevy 292 inline six on Engine Analyzer Plus. My sticking point is how to deal with the siamesed intake ports. The ports are siamesed in both the head and intake manifold. Presently I am assuming I use the full area of the siamesed port for each cylinder since the firing order is far apart for the shared ports. Any help would be appreciated.

For information:

The firing order for the 292 is 153624. One and two share a port, three and four share a port, and five and six share a port.

Questions Regarding Siamesed Intake Ports (9 Port Head)
See thread below 2nd from bottom of page

ProPower:

I had seen the thread started by Enigma57. I am hoping he will see this thread. Not sure where he ended up for a model, one approach mentioned in the thread was to treat the siamesed port as a large plenum until the ports diverge in the head. This would leave relatively short port lengths.

Thanks, Cris

The pairs do not have the same crankshaft rotational interval for their intake strokes.
The 1-2 is 240-480-240, the 3-4 is 360-360, and the 5-6 is 480-240-480.
This affects jetting: the even interval required richer jetting (per Jaguar specs)

Panic:

The most problematic for the model is when the 1/2 and 5/6 cylinders are 240 degrees apart. I am trying to model the stock performance first and the cam has only 188 degrees of duration with a 111 LSA. My hopes are, though there is some overlap with 240 degrees of separation, the effect is minimal with the small cam. The other cases of 360 and 480 degrees of separation have very little or no overlap.

I have been playing with the model and the best match I am getting to GM measured data for torque and horsepower comes when I model the siamese ports as (large) plenums. The intake port length is very small for this case.

Here's the GM plot and the EA calculated profile of HP and torque.

Cris wrote: ↑Sat Nov 21, 2020 5:24 pm ProPower:

I had seen the thread started by Enigma57. I am hoping he will see this thread. Not sure where he ended up for a model, one approach mentioned in the thread was to treat the siamesed port as a large plenum until the ports diverge in the head. This would leave relatively short port lengths.

Thanks, Cris

Hi, Cris! Just saw your post here. As Panic and ProPower noted, there is a small amount of overlap (shared draw) on the paired siamesed runners at each end of the head (cylinders 1&2 and 5&6) but the center 2 cylinders (3&4) have none. This being due to firing order and port layout.

Apologies to Shakespeare (Henry III) "A horse, a horse! My kingdom for a horse!" But when you start looking at these old 9-port heads with their siamesed head port ports and working up an intake (I refer to my 292 as the 'tractor engine', it actually was in an old combine on a farm when I found it)...... The first thing that comes to mind is (to paraphrase) "A 12 port head, a 12 port head! My kingdom for a 12 port head!"

Actually, Sissel has a very nicely done 12-port aluminum head for these engines, but it would bust the budget for my build. And the older Duggan 12-port heads are nearly impossible to find. So I have determined to rebuild one of Sissel's older brazed lump ported siamesed heads I found used.

My original intent with my 292 was to remove the headbolt boss from the center of each of the paired (siamesed) runners in the head and when working up my intake manifold, include (weld in) a hand fitted center divider extending into the siamesed head port that would functionally eliminate the siamesed runners so I could build an Isolated Runner intake for my Weber carbs. That would allow me to optimize port diameter/velocity and set runner length to take advantage of 2nd order harmonics.

Unfortunately though, there is insufficient material in the OEM iron head casting to isolate the runners with a center divider and then open each runner to provide sufficient cross section to spin the 292 inline 6 to even 5,500 RPMs redline without grinding through and into the water jacket.

Same for the late '80s OEM 12-port Brazilian heads that had individual intake and exhaust ports. Flow characteristics were worse than a reworked siamesed port US head and the Brazilian 12-port castings were thin and prone to cracking.

I even looked (in passing) at building one of the 230 or 250 Pontiac OHC engines and stroking it with a reworked 292 crank. That was a non-starter because the long stroke 292 engines are RPM limited and have harmonics issues above 5,000 RPMs. The quick reving OHC Pontiac design would not flow any better than a reworked lump ported Chevy head and the thing it does best - spin up quickly and rev higher - would be defeated by stroking a 250 by 9/16", creating a terrible rod length / stroke ratio and putting more sideloading on cylinder walls due to having to run very short rods in a block that is around 1-3/4" shorter than the 292 block.

So all in all, given my limited budget for this build and the intended use of the engine...... I have settled on running a reworked lump ported OEM siamesed runner 9-port iron head.

Regarding your question about the siamesed runners...... They can be made to flow more than you will need to run a 292 inline 6 to survivable redline RPMs. But forget optimizing port velocity. No way to dial it in precisely as with a true Isolated runner head and intake combo. Just get the flow your engine requires and don't exceed what you need in port cross section to do that and you will be OK.

Yes, the siamesed portions of head ports and intake manifold port runners act as plenum volume. And due to firing order, I suspect that the unused portion of the 2nd paired port (the divided part) acts functionally as plenum volume, as well.

The following is excerpted from pages 102 - 103 of "Chevrolet Inline Six-Cylinder Power Manual' (2nd edition) by Leo Santucci......

"Flow Breakthrough

Headrick's breakthrough was understanding more fully than anyone before him the real nature of airflow in the Siamesed ports of the Chevy-six head...... What Jim realized was that each port did not draw 50% of the volume of that port, but rather, because the firing sequence never fired two adjacent cylinders, one cylinder was actually using the other Siamesed part of the port as a plenum, thus creating, in effect, a much larger port flow than one would expect.

What Jim did and what was so unique for its time, was to change the timing events of the camshaft to take advantage of this 'larger port'. In effect, each cam lobe was ground on a different centerline so as you went from #1 cylinder to #6 cylinder some lobes were advanced, some were retarded. Jim shared with me that they interbred two different cam concepts to create enhanced airflow. To my knowledge, this design was never made generally available.

There were two elements that complemented the cam design......

Flow Tips

First, Jim was adamant about three points:

1. If you go smaller than 64cc combustion chamber, you lose airflow.

2. Maximum size valves are:
Intake 1.970"
Exhaust 1.625"
The bore size imposes these limits.

3. Velocity in the ports is critical."

Secondly, in order to support this airflow, more carburetion was needed as the power band moved into the higher RPM ranges (8,000 - 10,000 RPM)......

No further info is given as to the specifics of Jim Headrick's cam design, but someone a lot smarter than me could probably take the numbers Panic posted plus the flow numbers of their head and figure it out.

My engine will go into a road car that will do occasional towing of a boat trailer up and down long, steep grades in the Texas hill country. That's why I decided to build up a 292 inline 6 for my '57 Chevy sedan. I need the cubes and the added torque so settled on the 292 over the 230 and 250 primarily for this reason.

The 292 has harmonics issues and tends to sling flywheels if you spin it over 5,000 RPMs, so I decided to build and cam my engine for the most average power from off idle to a self imposed redline of 5,000 - 5,200 RPMs rather than higher peak HP and TQ. For this application, I want a broad powerband with nearly a flat torque curve. So that is what I designed my solid lifter cam to do. Its ground on 108 degree LSA, but depending on what you will use the engine for, that could be tightened up a bit.

Interestingly, Rhoads designed a SBC hydraulic flat tapped cam to be used with their lifters some while back and duration @ 0.050" was very close to what Mike used on his boat engine. But Rhoads ground theirs on 104 degree LSA to boost throttle response and power off idle and due to the way their lifters would affect (add) duration and lift as they pumped up at higher revs (above 3,500 RPMs).

Hope this gives you some ideas,

Harry

W/r/t "each cam lobe was ground on a different centerline"
Vizard describes doing this to a BMC Mini L4 way back, not sure who was first.

EngMod4T might be able to model it.
It has far more options for developing a manifold than other 1D simulation programs.

SchmidtMotorWorks wrote: ↑Mon Nov 30, 2020 3:54 pm EngMod4T might be able to model it.
It has far more options for developing a manifold than other 1D simulation programs.

It can. Speedtalk Sir Yun documented how to do this here and on his blog: https://aseriesmodifications.wordpress. ... /engmod4t/ He disappeared to work on his PhD I believe.

One thing is you'll need quite a bit of computing power.

One of the best A-Series builders in the US, said the biggest problem with the Siamese intake ports is you lose the inlet ram effect to some degree. He has A-Series engines making 127hp/L with the stock 5 port design and head castings. Mike Jones is doing cams for him and they are not scatter patterns.

Harry,
Clay Smith was mentored on camshafts by a fellow named Pete Bertrand; he ground flathead cams on different lobe centers to compensate for the different valve angles on the left and right banks. When he was killed at Bonneville Smith ended up with the Bertrand masters. I don't think Smith continued with this practice in later years but have none of his cams to check. Pete up at D&L Automotive (Speed-Talk Vendor) has a ton of f/h history stored in his brain maybe he will chime in; I don't think Smith ever practiced the concept on 6 cylinder cams. I ran some Crowers in 235's (45-50 years ago) and they were on the same lobe centers cylinder to cylinder. I still work with the 250-292 engine platform quite a bit and nearly drove myself crazy mapping cylinder sequences and lobe centers. I realized when I was chasing that rabbit that since my operating rpm range was less than that of Headricks the benefits probably didn't offset the effort. When Vizard used the practice he was class racing in a division where 3-4 horsepower would make a difference; that's not relevant to most. Working on seat angles and ports is more in the wheelhouse for most of us; perhaps a scatter pattern of valve sizes, seat angles and throat percentages would achieve more in lower rpm applications?

Clay Smith would not have has access to any of the resources required to make an informed design with regard to gas dynamics.
At best, he used intuitive reasoning to guess.
With the advantage of modern resources, we know that the system is more complex than human intuition can imagine.

SchmidtMotorWorks wrote: ↑Tue Dec 01, 2020 10:27 am Clay Smith would not have has access to any of the resources required to make an informed design with regard to gas dynamics.
At best, he used intuitive reasoning to guess.
With the advantage of modern resources, we know that the system is more complex than human intuition can imagine.

This.

And the gas dynamics process going on in an A-Series (or any other siamese port engine) is even more complex. What actually does make power, or at least makes things run more evenly which you need at higher outputs is different ignition timing for the cylinders that share an intake port vs the ones that don't. On the A-Series, it's enough the middle two cylinders (shared exhaust port) need more timing. I know Aaron Kelly, eventually just went to different temp plugs to get away from a more complex ignition setup, at the advice of Darin Morgan. I believe when he went to EFI on his SCCA GT engine, that issue went away.

Totally agree with the two posts above; staggering heat ranges on spark plugs in the 9 port engines is common practice for racing. I sometimes do it on stock builds for ag. applications IF the customer understands the reasoning behind it; sometimes it is not worth the effort to explain. Some of the early manifolds connected the p.c.v to the 3/4 runner with a small diameter steel tube in the runner directed towards the head to combat the inherent richness from the centrally mounted carb.

Guys, thanks for stepping in. Enigma57, I was hoping you would see this post. I appreciate the feedback.

Maybe I should sketch out my goals for this 292. I presently have a 1963 C10 with a 230. I love the way this engine behaves in that it will not complain if loaded at 1000 RPM. But, the 230 is a little short on oomph. I found a 292 on Craigslist that is one of the later GM replacement engines (Hecho en Mexico). It was in a running truck and I was able to hear it run and perform a compression test. The engine looks to be in excellent condition.

My plan was to tear this motor down completely, but it is in such good shape, I am rethinking how to proceed. I plan on doing a leakdown test in the next week as a final check.

I built the EA model to see if there is any low lying fruit to improve the engine without tearing it down. I now have some faith in the model, and surprise, surprise, the only parameter that picks up performance without sacrificing the low RPM torque of the stock motor is higher compression. The stock 292 has 7.65:1 compression. Bumping it to 9:1 gives a very nice boost across the RPM range.

I'm not the guy to try the more sophisticated modeling suggested. I am hoping my $200 EA Plus can steer me in the right direction to make these basic decisions.

The other noteworthy change I will add to the 292 is a Holley Sniper two barrel EFI.

Regards, Cris