By Dimitri N. Elgin
The goal in rebuilding an engine is to return its performance and reliability to what it once had new. The goal in “building” an engine is to increase its power within the capabilities of that engine without unduly ruining its other performance factors — drivability, mileage, reliability, and perhaps smog-law compliance.
Building a performance engine is not just a matter of tossing “speed parts” like a big cam into it, nor is building high performance or racing engines anywhere near as simple as some people imagine. Change one thing and that’s a different engine. Change many things and you have entered another profession, as an engine developer.
The package of professional skills you acquired as a rebuilder still apply to building high performance engines, whether for yourself or for customers. But a few advanced skills must also be acquired (or rented). So must further, up-to-date insights, for making effective decisions long before the first cutting edge touches iron. Then knowledge is power.
The topic of engine performance is enormous, and enormously complicated. So one article can lead only so far. We try the possible. The May 2004 issue of Shoptalk carried my article entitled “Camshafts” that covered the basics of how engines operate across six working cycles, how cam designs affect those cycles, and some recommendations. This article looks for a moderate but real increase in power to a production engine that must run well at reasonable rpm on pump gas.
Find the Balanced Package
Before changing anything on an engine, closely examine it in detail. See how each aspect of the engine balances against the others. Maximum usable rpms are limited by resistance to gas flow through the engine, and maximum piston speed is limited by stresses from the inertia of moving parts. Bring the weaker features of the engine up to the performance level of its strong points.
Focus your attention close to the action, in the combustion chamber and ports. The better the burn and freer the flow through those, the more power develops. The better the balance among features in the combustion chamber and ports, and the better the cam(s) choreograph the activity there, the more power develops. But the further away you wander from the action, the less improvement comes from your efforts. Chroming the cat-back exhaust system looks pretty but accomplishes nothing.
In the big 2-valve iron V8s most of us are most familiar with, the short block is generally a pretty strong assembly, usually capable of handing quite a bit more power than it now puts out in a regular passenger car. With the exception of its pistons and static compression ratio, build much the same short block as you would for 100,000-mile service life. Power potential lies upstairs, in the heads, the intake and exhaust systems, and especially in the right performance cam to direct the action.
Coming in the other door are many modern, perhaps unfamiliar, alloy 4-valve engines. Most of their heads already flow air extremely well. Slightly different camshafts can release further potential from these heads and make more power than the production lower end can handle in one piece. So here your first task becomes structural. Replace aluminum threads with steel. Fit a permanent “girdle” around free-standing tops of cylinder bores to encourage them to sit still for the honing head, and later for faster-moving pistons. Then fine-tune the top end.
Figure out what the engine want to do. Give it the cam(s) that make that happen.
Somewhat surprisingly, the connecting rod affects intake flow. More specifically, the ratio of the center-to-center length of the connecting rod to the stroke of the engine — termed the rod/stroke ratio or just rod ratio — has a significant effect on Volumetric Efficiency. More surprise: the effect of rod ratio differs for 2-valve vs. 4-valve engines.
Airflow in a normally-aspirated engine is driven into the cylinder only by the pressure difference between the 14.7 psi of the atmosphere and whatever less pressure is inside the cylinder at that instant. The greatest difference in pressure occurs shortly after the piston is moving downward at its fastest. Piston velocity peaks when the rod and the crank throw are at right angles to each other.
The exact number of degrees ATDC for maximum piston velocity can be found in any trigonometry table. The Tangent of the angle ATDC is twice the rod ratio for that engine. Then add 2-3° for time for that news to reach the intake valve at the speed of sound and affect airflow there. The sum should come between 70° and 80° ATDC. The shorter the rod ratio, the earlier that piston velocity peaks.
Airflow in 2-valve heads begins slowly. So airflow through these heads responds to a long rod ratio, close to 2:1, for maximum draw after 75° ATDC.
By comparison, a 4-valve engine flows a lot more air at the lower and mid lifts through its smaller valves and ports. This allows its rod ratio to be smaller without hurting power, more like 1.55:1. Check the rod ratio in a Honda. Airflow demand in the 4-valve engine occurs closer to 70° ATDC.
Another flow difference between 2-valve and 4-valve heads is the ratio of exhaust flow to intake flow. Exhaust valves and ports are always made smaller that the intakes, because exhaust flow gets pushed first by high cylinder pressure, then by the piston on the way up. In 2-valve engines the exhaust port flows between 60% and 80% of the intake. Exhaust flow in 4-valve engines is very high, somewhere in the 80% to 90% region. Later we’ll see how these factors affect cam selection.
It is important to realize that the engine sees three different compression ratios. One is the static ratio which we are all familiar with: clearance volume + swept volume, divided by the clearance volume. A number like 9:1 is a common static compression ratio.
The second is the effective compression ratio, which the engine sees when the intake valve closes against the valve seat. A number like 7:1 is common. This is determined by the interactions of the static compression ratio, the rod ratio, and cam timing for closing the intake valve. (Wrist-pin offset has an additional but minor effect.)
Third is the dynamic compression ratio which is when the engine is in the peak power range and the volumetric efficiency is above 100% then the cylinder pressure-compression when the intake valve closes is at its highest, example above 8:1.
Building an engine for more performance often means raising the static compression ratio close to
10:1, but keeping the effective compression ratio not much over 7:1. Anything lower gives up power. Anything much higher will not run at low speed with WOT on pump gas without detonating and destroying itself.
Before selecting a cam, long before changing anything, take benchmarks to determine what a production head already provides, feature by feature. Trust no published specifications. Measure everything. What diameter and width are the valve seats? What angle? How did the factory finish the top and throat cuts? Is there a good radius behind the seat to direct airflow through the open valve curtain? Does the backside radius of the valve complement that?
Which way does the port aim mixture into the cylinder? 2-valve heads direct flow around the circumference of the cylinder in a motion termed swirl. 4-valve heads send flow in head-over-heels tumble down the bore. Too little or too much of either motion stalls flow. Is the mixture aimed dead into the cylinder, around its circumference in a swirl, or splat up against a cylinder wall? Hello BBC.
Is airflow shrouded after it passes the valve by running into a side of the combustion chamber? Was the chamber cast with some intentional shrouding? Check a SBC Vortec chamber. GM did something clever there for guiding flow past the valves. Is the spark plug well out of the way of the incoming air/fuel charge, or does it look as if it will be soaked silly before it attempts to fire? Will that affect how plugs should later be indexed?
Look deeper down the ports. Do the machined cuts blend smoothly into “as cast” surfaces? Do lumps or ridges protrude where casting cores once almost aligned? Feel inside. How sharp is the turn to the critical short-side radius? Did casting cores leave a sharp edge there to turbulate and stall flow?
What is the overall shape of the port, both lengthwise and in cross-section? Does the port narrow gradually along its length to accelerate airflow, or change size suddenly? Does it address the port from a high angle and turn smoothly into the bowl area, or does it send airflow along a long flat trip across the head then demand a sharp drop at the valve? (Hi there, Jaguar XK.)
Is the cross-section square with dead corners of zero flow, or perfectly round so flow spins this way and that but not in? Is the exhaust port just a big hole in the head? (Hey, Mopar B. Meet Ford Cleveland.) Where does the cross-sectional area of the port become smallest, and how small is that? A new skill at casting with latex can bring that shape outside for more direct and insightful analysis.
What does the finish of the chamber look like? It does not have to be polished (although doing that does not hurt), but surfaces should be reasonably smooth with no sharp edges. More power makes more heat. Heat seeks peaks. Gently bevel sharp edges, or the engine will pre-ignite off its own internal glow plugs.
Go with the Flow Bench
So far we have examined heads much as any avid gearhead would on his garage workbench. We now know what they look like. But how do they flow? Take them to a flow bench.
A flow bench is the measurement tool to analyze airflow through the ports. A bench can now be used in three modes. First, to measure the airflow rate through each port as a function of valve lift. Then to probe inside ports to analyze the details of airflow. Finally – and this is very new – attach a wet-flow adapter to the bench and actually see the dyed flow leaving the port and entering the cylinder. All three procedures add up to help you decide how to address these heads.
Air flows through intake ports into an “adapter” exactly the same diameter as the cylinder and two bore lengths long. Air flows from the chamber through exhaust ports into a stub stack the same diameter as the header pipe and at least eight inches long. Part of the fun of operating a flow bench is getting to fabricate and inventory an interesting assortment of adapters.
Take airflow measurements in cubic feet per minute across the entire range of valve lift, at every increment of .050”. Continually adjust the flow bench so it keeps working at the same “depression”, the working pressure difference through the port, to keep your measurements meaningful. For later comparison with other heads or your modifications to these, always work with the same depression.
The SAE recommends a depression of 28” of water (about 1 psi). Some experienced engine developers use less, maybe 16”. A few use more. Careful. Depressions of much more than 40” produce falsely optimistic flow numbers. Truly bad ports can be made to flow great numbers at unreasonably high vacuum. But later they’ll die on the engine, and that customer will be very unhappy about your services. As with a dyno, never rig the system just to make big numbers. It’s pointless. Accuracy is an operator skill, painstakingly acquired. And findings vary from bench to bench, even among the same model, by as much as 10%. Conversion factors for readings at different depressions are not reliable. So don’t bother “racing” somebody else’s flow numbers. Believe what you see.
Measure every port. If one port flows less than the rest, that cylinder will need different spark timing from the others, and total spark advance will be compromised for the weak cylinder rather that set for best power from the good ones. Low to mid lift is very important on the exhaust valve. Mid to high lift is more important to the intake. But measurements of intake flow at lift as low as .050” is also important. That gets flow going and also accepts the final pulse of mixture arriving by inertia before the valve closes at high rpm.
Findings should be tabulated for later reference, but for analysis graph all the flow numbers as a function of lift. Trends jump out at you from graphs. At what lift does the rise in flow level out on the intake and exhaust sides? What is the ratio between intake and exhaust flows? The ratio should remain quite steady across the range of lift. If not, there is an opportunity to find power. I prefer to port most heads to achieve exhaust flow 75-80% of the intake. Exhaust flow above 90% may make power at the drags, but a lot of intake charge goes sideways out the exhaust valve. Fuel economy suffers and so does torque.
Experiment. Go a bit beyond your expected maximum lift, to find out what happens there. Try a valve with a different shape to its backside radius. See how much a clean back-cut improves flow. If you have a surplus (say, cracked) head to play with, try different seat angles, widths, and multiple top and throat cuts. Notice how a 30° seat flows great at low lift but dies at high lift. Radius the top edge of the exhaust valve margin, and compare flow numbers with a square margin. If the customer demands bigger valves, try just one first to see where this program is headed.
What the bare head tells you is a baseline. Now attach the intake manifold and carburetor (or FI intake and throttle body). Everything changes. I have seen a loss of 10-60 CFM after the intake system was installed, on cylinder heads properly ported with valve job completed. After the intake was installed I have also seen great variations between flow numbers among different intake ports, as much as 20 CFM. Most CFM numbers quoted are without the intake manifold installed. This is mistaking the mission. It is flow through the entire system that the engine sees. Corrections here pay off big.
Next, probe the flow with a Pitot wand. Find pressure differences, flow velocities, dead spaces, and intriguing mysteries. Does the smallest cross-sectional area of the port restrict the air flow, creating turbulence, or is the port so large there that very little velocity develops down the port? Is the intake port efficient all over its area, or is the floor and one corner dead? When you find a dead space, fill that with clay and see what happens next. If it’s good, epoxy the solution in permanently. Check along the floor of the port on the short-side radius. Does flow follow the curve, or is it breaking off?
The very latest bench equipment can now view wet flow, dynamically and in detail. This is an extremely promising process. Airflow is one thing, but a running engine flows air “wet”, mixed with gasoline in a ratio around 14:1. Wet flow more closely approximates what happens inside engines. And for the first time you can watch dyed mixture flow into the cylinder – or better – videotape and freeze-frame it across the range of valve lift.
Wet flow is a powerful tool for analyzing mixture motion into cylinders. Testing has shown that the flow through each intake tract must be individually matched as a partner to its port so that each cylinder receives the same amount of mixture and close to the same degree of mixture motion. If some cylinders are weak from poor motion, spark timing among cylinders could vary as much as 6°. Timing the engine for the poor cylinder loses power. Timing set for the better cylinders burns the piston in the weak cylinder.
Wet benches map the flow vortexes inside the chamber. Even the big boys are learning from this. Chrysler’s new head for Pro Stock couldn’t win. A wet bench discovered a huge vortex soaking the plug. Chrysler reshaped the chamber. Dart, the respected maker of aftermarket high-performance heads, now wet-flows their designs.
Engine Cycles and Cam Timing
Once you have the particular performance characteristics of the heads carefully mapped, use that information to select the one cam that makes the engine make power like it wants to. Compare your test results against what the engine experiences while it goes through all six cycles.
Engine cycles are determined by the direction the piston is traveling and the timing of the openings and closings of the valves — collectively termed valve-timing events. Timing these events becomes complicated because a lot of compromise must be made in order to balance out all of engine operating cycles. The May 2004 article outlined the six engine cycles of a four-stroke engine. This time we’ll walk through each cycle in turn, now from the viewpoint of determining how changing every valve-timing event affects other cycles, and how balance builds power.
· The Power Cycle: TDC to Exhaust Valve Opening
By the time the crankshaft reaches 90° ATDC, cylinder pressure has dropped greatly and most of the power that can be recovered from it already has been. So opening the exhaust valve well before BDC loses less power from the power cycle than it later gains across the following cycles. The lower the rod ratio, the faster cylinder pressure drops.
· The Blowdown Cycle: Exhaust Valve Opening to BDC
The blowdown cycle relieves excess (but unrecoverable) cylinder pressure and begins clearing exhaust gases off the energy of their own pressure. Otherwise the piston would have to push all the exhaust gases out of the cylinder on the next up-stroke, lowering horsepower from a pumping loss.
The timing for Exhaust Opening is the least important of the four valve events. It can be anywhere between 50° and 90° BBDC, so its timing is easily adjusted to match the performance characteristics of that engine.
With higher compression ratios the burn rate is faster, so the exhaust valve can be opened earlier, which aids in the cylinder blow-down. With lower compression ratio (static 8:1 or lower) you want to delay the exhaust opening as late as possible in order to utilize the last usable bit of pressure that is on top of the piston. But that hurts the top end horsepower, because the blow-down period is no longer as effective.
· The Exhaust Cycle: BDC to Intake Valve Opening
The piston reaches maximum velocity at about the same number of degrees BTDC as it did ATDC on the way down, or a degree or so sooner with offset wrist pins. The exhaust valve must be open sufficiently by this time so that spent gases in a hurry meet little resistance against being pushed out.
How far the valve must be open is known from flow-bench data. The proper cam meets that need from a combination of timing, total lift, and its rate of lift (its “velocity”).
· The Scavenge Cycle: Intake Valve Opening to Exhaust Valve Closed
The scavenge cycle occurs during the overlap period, when intake and exhaust valves are both open at the same time. The intake valve is just opening. The exhaust is closing but not yet seated. Overlap is what the cam and valves are doing, dictated by the combination of total cam duration and the locations of lobe centers. Scavenging is what the engine is doing with that.
A good number of engine processes (and a few unsolved mysteries) are going on now simultaneously. The most important are (1) scavenging the last of the exhaust gases as much as possible from the clearance volume, where the piston cannot reach to push them out, and (2) initiating intake flow into the cylinder without wasting very much of it out the open exhaust valve.
Overlap duration increases as total duration increases, and it also increases as the lobe center decreases. Increasing the time for Overlap makes more time for scavenging at high rpms. Residual exhaust gases kill power twice over: they displace their volume in incoming charge, and later during combustion they absorb heat that should have gone into making power. At 5000 rpm an engine with a high-performance cam carrying 55 degrees of overlap must complete the entire scavenge cycle in less than two thousandths of a second.
In standard engines, valves are open together for only 15-30 degrees of overlap. In a race engine operating between 5000 and 7000 rpm, the overlap period is more like 60-100 degrees. The penalty for so much overlap in a street engine is very poor running at lower rpms, when a lot of the intake charge has time to sidetrack directly out the open exhaust valve. Mileage goes South. Heads overheat from fuel burning in exhaust ports. The engine runs hot. The exhaust system gets fueled like a blowtorch. The tailpipe turns white. Catalytic converters fry. The buyer blames the cam grinder.
Timing Exhaust Closing must be balanced against flow through the intake port. If the intake port flows poorly from being too small (or too large) then later Exhaust Closing might help to initiate intake flow. I consider this only as a last resort for kick-starting a lazy intake port. It always carries some charge out the exhaust valve, wasting fuel and all that.
Make the overlap period as short as will complete the job of scavenging. Factor in the effects from the combustion chamber size and shape (including the shape of the piston top) and shrouding near valves. Balance power goals with other requirements for the intended usage, such as idle quality, low-speed throttle response, fuel economy, and smog test compliance.
· The Intake Cycle: Exhaust Valve Closed to Intake Valve Closed
I consider Intake Valve Opening the second most important valve timing event, because that does two important jobs. (1) It initiates the Scavenge Cycle and (2) it begins lifting the intake valve out of the way of the incoming charge. The air/fuel mixture began entering the cylinder during the Scavenge Cycle, builds to a maximum, tapers off, then packs in a final gulp.
The intake valve is in a race with that pressure differential at maximum piston velocity that drives intake flow. The valve always loses this race, because max draw happens between 70° to 80° ATDC, yet the intake valve does not open fully until it reaches centerline, down around 105° to 115° ATDC.
When you can’t win, do your best. Get the valve out of the way as far as possible by giving it a fast rate of lift, a “high velocity”. Much the same could be accomplished by more valve lift, but then the nose of the cam gets pointy and real stiff springs are needed for closing the valve – a combination not favorable to very long service life.
The Intake Closed point – when the valve seals on the seat – is the most important valve-timing event. This event governs both the engine’s rpm range and its effective compression ratio. Closing the intake valve later optimizes intake flow for high rpm and allows inertia to pack in its last gasp of air. The drawback to that is back-flow at low rpm. But closing the valve earlier shuts down rpm. Pick your operating range.
· The Compression Cycle
The piston compresses the air/fuel mixture to a high enough pressure and temperature for it to be ignited efficiently by the spark. The effective compression ratio must be high enough to compress and pre-heat the air/fuel mixture for a fast, complete burn.
But too much heat and pressure kick off the whole charge at once in the destructive explosion of Detonation. When pistons taken from a blown engine show ring lands melted as if by a cutting torch, that was by Detonation. (If a hole has been blasted through the center of the piston crown, that came from a hot spot in the chamber pre-igniting the mixture.)
Tweaking for idiosyncrasies, nitrous, supercharging, turbocharging
I Prepare the cylinder head and fit a good, small-diameter exhaust system. To increase the rpm band, increase the compression ratio. I do not advocate extra high lift, long duration, or very high compression in a street car. I use velocity. I have never seen a normally aspirated engine make more power by lifting beyond the flow capacity of the head.
Within limits, experienced cam makers can juggle timing events to customize valve action to the special requirements of that particular engine. Careful here. Not all valve timing is equally important. Exhaust Opening may be re-timed with little impact elsewhere. But Intake Close is tied closely to the static compression ratio, and cannot be re-timed very far without upsetting the dynamic compression ratio, cylinder pressure, resistance to Detonation, burn rate, rpm range, and just about everything else that makes power. Intake Opening is slightly less important, and Exhaust Close less than that. So juggle needs of the engine against the importance of timing events.
I find it amusing to see people treat the 4-valve engine like 2-valve engines when they select cam durations and lifts. Timing and lift for the 4-valve engine must be made different, due a 4-valve’s quicker air flow velocity as well as its high ratio of exhaust to intake flow. Even with only moderate total cam duration, a cylinder under a 4-valve head sees enough air flow by 75° ATDC. For example, a high-performance 2-valve engine for the street would need 270°-280° of total duration, but a 4-valve engine would require only 250°-260° total duration for equivalent performance. Any more duration closes the intake valves so late that the engine would become very peaky, hardly suitable for street driving. 280°-290° for the 4-valve engine would be the equivalent to 310°-320° on the 2-valve engine.
In a 4-valve engine the intake and exhaust cams can use the same duration until the intake cam gets into the 270°-280° duration range. In some production 4-valve engines, its good exhaust flow can over-scavenge cylinders. Less duration on the exhaust would help that.
Timing the Exhaust Opening event should be re-examined whenever an engine begins using a supercharger, turbocharger, or nitrous oxide. Be careful with a 2-valve supercharged engine. The extreme pressure still in the cylinder can bend the valves and pushrod if the exhaust valve tries to open too early against it. Turbocharging requires wider lobe centers to narrow the overlap period. Nitrous responds to slightly wider lobe centers and more duration so the exhaust valve opens earlier to relieve the higher cylinder pressures nitrous generates. Pick you power goal. Balance the package for it.
I hope that this information as well as the May 2004 Shoptalk article will help you better understand this very complex internal combustion engine. Happy tuning!
For a glossary on camshaft terminology, visit Dema’s website: www.elgincams.com.