Primary pipe length formula

[Kevin M]

Adam,

This stuff is tremendous. You could do a book. What’s needed is not a Masters level engineering text, as you mentioned those already exist, but a simple practical book giving exactly the sort of information you are giving here.

No one could do a book like this using a Dyno, it would take way too long and cost too much. You have the great advantage of being able to run a large number of variations in a very short time compared to actually rebuilding an engine for each trial. (By the way, if you change your mind, I have a friend who is an editor with 20 some years experience. I’m sure she could put you in touch with a ghostwriter, etc. Another option would be to put you in touch with Scott Parkhurst, technical editor for Popular Hot Rodding, they do books on this sort of thing and he might have an interest.)

As for the Intake expose, you certainly have my vote for it.

Kevin

 

[Steve Toner]

Adam,

This is really good stuff you're posting. Consider putting it all together into a document for the Articles Page . And yes, by all means do an induction expose as well!

 

[Adam C.]

OK, back for more. Now we are going to look at the induction side. I thought about starting a new thread for induction, but we might as well keep it all together, I've already thoroughly hijacked this thread with my ramblings anyway.

So we will start just as we did for the exhaust side. The engine is our 5.0L numerical test mule. The exhaust is the original 36"long primary, 36" long secondary 180 degree system we started with in the exhaust study. Now we are going to only manipulate things that are related to the intake side.

The explorer EFI manifold on our mule has runners that are 19.5" long, from plenum to valve stem. This is the entire primary runner, including the upper and lower manifold, and head port. Although the runner diameter varies somewhat along it's length, it is nearly constant at 1.68" in the upper manifold where we will be playing.

First is primary runner length. Because I am all about making horsepower, we are going to make the primaries progressively shorter in 2" increments.

Not too suprisingly, as we shorten the runners, the botom end suffers while top end power increases. There is a limit in how much we can increase the top end. We are seeing diminishing returns at around 15.5 to 13.5". Going any shorter will only kill the low end, with no real benifit up top. We can see that if we wanted to go higher in the rev range, we would need to rethink our cam, heads, and exhaust system.

 

[Adam C.]

So now on to primary runner diameter. Here we will only manipulate the diameter of the upper manifold, which is approximately 10" of our primaries for the 19.5" baseline case.

Here we vary the cross-sectional area of the runner by 25% from the baseline of 1.68". The primary runner diameter has an huge effect on high RPM performance.

Because I love horsepower, the 1.88" case is looking pretty good. Maybe we'll use that in the future.

 

[Adam C.]

So here is one that I know everyone is interested in, the old single 4 barrel vs. webber.

For this comparison I chose two dimensions that were looking pretty good for horsepower in our earlier experiments. Our manifolds both have primary lengths of 15.5" and diameters of 1.88" in the top 5.5" of the manifold, that then taper down somewhat to the cylinder head port diameter. The primaries for both manifolds are identical. We will only change the position of the venturies and throttles.

For the 4 barrel carb we have a single plane mainifold, for the webbers we have primaries that just stick out into atmosphere.

With all else being equal, there is hardly any difference in power. This may be suprising to some, but this is because in the real world, all else is often not equal.

The webber carb leads a hard life. Because there is a carb for each runner, the venturies see violently oscilating flows due to the intake valve opening and closing. At low RPM there is even a significant amount of backflow, where exhaust gas is flowing out the intake valve. This makes it very difficult to control the A/F ratio. The velocities in the primary runner are also very high. Because the throttle and venturi are in the primary, the flow losses caused by them are HIGHER than if they all were connected to a plenum.

The 4 barrel carb on an engine with 5 or more cylinders leads a much more civilized life. It sees a nearly steady state flow, especailly at high RPM. Because the cylinders breath pretty much one at a time, all four barrels are providing air to each cylinder. Because the cross-sectional area of the four barrels is greater than that of a single webber, the flow velocity and therefore flow losses due to the venturies and throttles are lower.

I love the look of webbers, especially on a GT, but they really don't make much sense to me. /ubbthreads/images/graemlins/confused.gif

 

[Adam C.]

So now for what is the most exciting topic to me, cylinder head flow. We already showed that you really can't have too much exhaust valve flow, but is this true for the intake valve?

For this test we are using the single plane manifold in the previous example because it offered more power than the EFI manifold with longer runners. Now we are going to vary the intake valve flow up and down by 25%. This just so happens to equal 180 CFM on the low end (GT40 iron heads) and 300 CFM on the high (TFSR).

On the low end, higher flow means less torque. This is because we have a cam with a good amount of overlap and it is letting some of the fresh air in the cylinder reverse and go back out the intake before the intake valve closes. We could probably make the same power at low RPM with all three heads by adjusting the cam, but again the top end would suffer.

On the top end, more flow is better, period. There is this myth that you can have too much flow for small engines, or those with low compression, or low RPM, etc. This is simply not true. The reason these engines often perform poorly is because the PORT DIAMETER has been significantly increased, loosing flow velocity and shifting the natural frequency of the induction system too high. At low RPM these engines suck because they have a cam with an intake valve closing event that is probably too late.

This is the difference between recognizing patterns, and examining things systematically. Fortunately in engine simulation you can change exactly one thing at a time and see the effect of that one thing, even if it may not be possible to implement in real life.

So with the TFSR's (or AFR 205's for Ron since he is so smitten by them /ubbthreads/images/graemlins/grin.gif) our little 302 is rockin at 7000 RPM. We're talking nearly 110% vol eff! Just for comparison, our original 5.0L with a complete aftermarket mustang exhaust system managed about 325 HP at the flywheel. Our new configuration with the single plane manifold, TFSR's, and GT type exhaust and no mufflers is now sporting about 465 HP. This is with with a very streetable 10.3:1 compression and a relatively mild hydraulic roller cam (wich would probably cause valve float above 6000 RPM in real life).

There is not much left now but the cam, compression, and everybody's favorite, displacement. Do we press on?

 

[Kevin M]

Go Adam Go!!

One question. Does the venturi in the IR - Weber scenario behave the same as a straight section of port of the same diameter?

Which is to say, if we decide that a 2.30 sq in cross sectional area is right for our intake system. Do we want the venturi to be 2.30 sq in (43.5 mm) as well? Or, does flow behave differently through a venturi causing us to want the venturi to be somewhat smaller than the runner. (And if so, how do we calculate how much smaller?)

Kevin

 

[Adam C.]

Kevin,

From a tuning standpoint you want the open crossectional area of the webbers to be equal to the ideal crossectional area. This means that you want to subtract the crossectional area of anything that is in the runner, like the throttle blade and shaft at full open, the boosters (is that what their called?) etc. This would mean that the bore of the webber would actually have to be slightly larger than the ideal runner diameter. The venturi has no real effect in tuning except that it is an area change and a flow loss.

Now, with that being said, there may be reasons to have smaller diameters for carb operation, I'm not sure.

Hang on, the final installment is coming.

Adam

 

[Adam C.]

And now for the final installment of our engine tuning trillogy. The items that we have left to look at are the camshaft, compression ratio, and displacement. The camshaft has several parameters, leading to an infinite number of possible combinations. Here we look at the major parameters one at a time.

First up is cam lift. Most magazines say that it is pointless to run a cam that has more lift than that of the cylinder head at max flow.

Here we vary the cam lift by 10% up and down from our good old FMS X303 cam. The engine is still our 5.0L described earlier, with the single plane manifold and TFSR's.

It is important to note that while our heads in this simulation flow as well as the TFSR's, they still have the small 1.94" intake valve of the GT40X. This means that the max flow is still realized at about 0.5", where the real TFSR's attain max flow at about 0.65" I think. This means that to get the results that we are showing here, you need to run a cam with more lift than the X303.

For our comparison however, the lift of the X303 is about .050" higher than the max flow of our physically impossibly ported GT40X's, so it will still work for comparison.

In this case, the magazines seem to be pretty accurate. There is almost no difference in vol eff with the lift varying a whole 0.1". It should be noted that all three of the lifts tested were at or above the head max flow lift. Going below this will hurt power.

 

[Adam C.]

Next up is Cam phasing. Until now the cam has always been installed "straight up". Now we are going to advance and retard the cam 4 degrees to see the effect.

Advancing the cam shifts the VE curve down in the RPM range, while retarding shifts it up. The shift is relatively small, but not negligible.

If you are considering shifting the cam further than 4 degrees because you want to move the VE curve further, you need to pick another cam.

 

[Adam C.]

Now we look at Lobe Separation Angle. This, combined with the duration, determines the valve overlap duration.

The x303 is ground with a 112 LSA, which is typical of most mild SBF cams. Here we narrow and widen it to 108 and 116, respectively.

Typically, narrowing the LSA increases power on the top end, while decreasing on the low. While this did occur here to some degree, the lines actually cross a couple of times. I think this indicates that we just about have this thing tweaked in.

 

[Adam C.]

Last up for the cam is duration. The x303 has a duration of 224 deg at 0.050". We will now decrease it to 216, and increase it to 232.

Going to the larger cam devistates the low end, with no real benifit up top. The smaller cam however, has improved bottom end, with not much lost up top. This indicates that our cam is probably just a bit on the large side for our little 302 at 7000 RPM and under. Suprising? The results might have been different if the heads didn't flow so well.

 

[Adam C.]

So with the cam finished, we turn our attention to the cylinder block. There aren't a whole lot of things that we can fiddle with here.

First we will look at compression ratio. I often hear that in order to run a high flowing cylinder head you need to have high compression (I even had a twisted wedge rep tell me this once). Where does this stuff come from?

Here we move the compression ratio up and down a whole 2 points. On the low end we have what a supercharged or turbo engine would run, on the high end we're talking 110+ octane.

Remember, we are looking at the effect of compression ratio on TUNING, not power. We are seeing if by changing the compression ratio, we are changing the amount of air the engine is able to take in.

There is little difference in VE by changing the CR, in fact, the low CR case is better over most of the range. Again, this is with our huge flowing heads (300 CFM).

Note that the difference in horsepower due to compression ratio is due primarily to Thermodynamic Efficiency, and not to airflow. Here we are only looking at airflow.

 

[Adam C.]

And so this brings us to our last topic, displacement.

In reality the displacement should be the first thing that is set, and then all other parameters determined. Suppose that for this case we have built our little 302, spent a ton of cash on ported heads, a custom cam, and custom length manifold and headers, but we aren't happy with the insane amount of power we are making.

So we pull out the crank and start stroking the engine (well past what a 302 block can hold).

First we are looking at VE for engines with displacement of 302, 327, 351, 377, and 402 CI. If there is some flow potential left in our cylinder heads or if our resonance tuning is set a little too high in the RPM range for our 302, one of these engines should exploit it.

In looking at VE, we are measuring how well the engine is being filled. Because we are increasing the displacement, the same VE at a given RPM in a larger engine would mean that we are actually taking in more air.

By increasing the displacement, we are shifting the resonance tuning further down in the RPM range. Notice that the peak VE is also lowering in magnitude. So are the bigger engines making more power? To know that for cedrtain we need to look at horesepower.

 

[Paul Bearman]

Fantastic!!!
Many thanks Christain, this is far more info than I could have ever expected. I may not fully understand it all (yet) but this has given me a great insight into what is required to produce a well balanced unit.
Thanks also to everyone else who has contributed.

 

[Adam C.]

So here is the plot of horsepower for these engines.

It is important to point out that this is the indicated power from the engine, neglecting friction due to rotating components and sliding pistons. The amount of friction will also increase with displacement at a given RPM.


The power is increasing throughout the RPM range with increasing displacement. It is beginning to become constant at about 575HP at 7000 RPM for the 402 CI engine. This is indicating the limit of our heads, induction, and exhaust. If we wanted more peak power than this, we would have to go back to the drawing board.

If we continued to increase displacement with this combination, the peak power would not go up, but the lower RPM range would continue to increase. Adding displacement will not necessarily increase max power, but it will almost always increase the area under the curve.

 

[Adam C.]

And so that brings us to the end of our little crash course on resonance tuning.

The aim here was not to give the majic formula (I keep that for myself /ubbthreads/images/graemlins/grin.gif), but to show what is important and the directional trends due to these parameters.

Our example engine was 302 CI and operating between 1500 and 7000 RPM. Due to the number of parameters, there are multiple solutions that would get similar results for this engine, but they won't be radically different.

I hope this helps those who are interested. This forum has tought me a tremendous amount about GT's and I though I would give back by contributing something I have some experience in.

Thanks all,

Adam

 

[Mark Worthington]

How about a round of applause for Adam? And don't forget to tip your waiters, ladies and gentlemen. /ubbthreads/images/graemlins/smile.gif

Seriously, Adam, that was some great stuff. I still need to review it in detail to be able to really digest it.

Thanks for sharing.

 

[Kevin M]

Adam,

Wonderful stuff.

I suspect that this thread will become a major part of the future tech articles section.

Thanks Again,

Kevin

 

[Lee]

Thanks very much Adam. In nice little bite size chunks as well. Like Mark, it will take time to digest, but a great insight into the black art of tuning. Thank you.