Cylinder Heads, Part I

Lined up to resemble infantrymen...        read full caption Lined up to resemble infantrymen marching into battle, we gathered castings representing the range of 1967-and-up production heads.

Swirl readings are taken from...        read full caption Swirl readings are taken from Quadrant Scientific’s sensitive swirlmeter. Readout is in in-oz of torque. A swirlmeter is a must-have for serious cylinder head development.

All the airflow evaluation...        read full caption All the airflow evaluation was done on a Quadrant Scientific computer-integrated Flowlab. Repeatability was astonishing. It sure beats trying to read numbers off a tube. All testing was done at 28 inches of H20 pressure drop. Values at 275 cfm ran 9 percent lower than those of the most prevalent flowbench in many back-to-back tests.

Here is the compact closed...        read full caption Here is the compact closed chamber of the 915 casting. The flat area opposite the spark plug (A) allows for the squish and quench effects.

Back-cutting the valve 30...        read full caption Back-cutting the valve 30 degrees (right) relieves the valve of its flow impeding ridge (left). The result--a significant gain in low-lift flow (up to 0.350 inch) at all levels of modification, and virtually free horsepower.

The old standby--the equivocal...        read full caption The old standby--the equivocal Mopar Performance porting templates. The first step is to cut the templates along the black lines.

Jumping into the porting fray...        read full caption Jumping into the porting fray doesn&8217t require a lot of gear. Evaluating the results is another matter, though. Carbide cutters make quick work of removing cast iron, while cheap, high-speed steel is useless. Stones and porting rolls are available at Standard Abrasives, and the drive tool can be air or electric. Proper safety gear and lighting are musts.

The template kit leaves no...        read full caption The template kit leaves no guesswork in aligning the templates. Just mark them according to the directions. Note the location of the casting number on the combustion chamber side below the exhaust ports.

This is the stock port form...        read full caption This is the stock port form in the bowl area, under the seat of a 346 casting. Template porting removes a lot of material. The large relieved area (A) in the open-chamber design is a detriment to squish.

The idea is to remove just...        read full caption The idea is to remove just enough metal to allow the template to drop into position. The templates register off the valve guide for alignment. The templates were a tad large for our standard 2.08-inch valve heads, and were adjusted slightly to fit without cutting into the seat. Ruin the seat or the approach to it and, at best, the head can be salvaged only by installing a larger valve.

Once an area has been ground...        read full caption Once an area has been ground sufficiently to fit the template, the adjacent area is attacked. Blend evenly from each checking point.

In critical areas in which...        read full caption In critical areas in which we grind the approach to the seat, it pays to paint the wall with a black felt marker or machinist&8217s blue to boldly depict the grinding progress (arrow).

A fine grinding stone can...        read full caption A fine grinding stone can be used to blend the rough surface left by the carbide cutter. Further polishing can be done with the sanding roll. Although the effect on airflow is insignificant, the rougher surface helps promote the boundary-layer turbulence desirable for keeping the fuel in suspension.

If factory big-block Mopar heads cause confusion, it's no wonder. Other than tidbits of information released more than 20 years ago by Chrysler's Direct Connection factory tech bulletins, there has been almost no comprehensive information published on the selection and modification of these commonly used heads. Although several professional porting establishments will gladly provide modified factory heads, the techniques used and the resultant flow and performance gains are, as expected, closely guarded. As a result, the enthusiast must resort to dropping the dough and choosing level X, Y, or Z in a Pro-ported head. Alternatively, those who seek to improve their big-block's performance can order a beginner's porting kit, haul their factory castings to the basement, and make cast-iron dust. Unfortunately for our would-be home porter, without access to a flow bench to evaluate progress, or spe-cific information on what works, the effort is a hit-or-miss proposition at best. At worst, the heads can show little or no gain, or flow less.

Factory Castings

Since the 1967-78 heads are the most readily available, and the most commonly used in street performance applications, we will limit our discussion to these heads. The year 1967 was significant for Mopar big-block heads with the introduction of the 915 casting, identified by the last three digits of the casting number. The 915 was a performance-minded redesign of the big-block head, featuring revised ports while retaining the closed-chamber configuration (typical factory chamber volume of 78.5 cc) of the earlier 1964-66 big-block heads. The 915 casting was found only on the 440ci big-blocks, as the 1967 383 engine retained the 516 castings of the previous year. In standard form, the 915 was fitted with 2.08-inch intake valves, and the same small 1.60-inch exhaust valves found in the older 516 heads. The year 1967 also was when the 440 Magnum engine was introduced. In step with the performance requirements of the Magnum, the engine’s exhaust valve size was increased to 1.74 inches, a size used successfully in some earlier high-performance packages.

Over the years, confusion and mythology arose around the 915 casting, and it mistakenly became known as the 1967 440 Magnum head. This same head was also found on all 440s in 1967--from your old man’s 440 New Yorker to the boss’s 440 Imperial. The only difference was the smaller exhaust valve in the non-Magnum engines. A throat cut and the bigger exhaust valve evened the score. For 1968, the big-block cylinder head was modified again. The new 906 head was identical in port configuration to the previous year’s 915. However, the 906 was cast with open combustion chambers (typical factory volume 88 cc). Also, the smaller 1.60-inch exhaust valve, which had been fitted in standard engines, was discontinued. The 1.74-inch exhaust valve now was used in all B/RB heads. Despite the popular misconception that the 906 head was exclusive to the 440 Magnum and the new-for-1968 383 Magnum, it was actually used across the board, with the same valve sizes in all big-block engines for 1968. From the 383 two-barrel to the 440 Magnum, the heads were unchanged. Some of this confusion can be attributed to Chrysler’s advertising of the day, in which the 383 Magnum was heralded as "coming with the 440 Magnum’s free-breathing cylinder heads." What the mavens on Madison Avenue weren’t telling was that the standard 440, the 383 2V, and all passenger car big-blocks had the same 906 heads with the same valves--although the Hi-Po Magnums came with stiffer valve springs fitted. With the exception of some industrial, heavy truck, and motorhome application castings, this one-head-fits-all philosophy was carried through the end of big-block production in 1978.

While the 906 and its open-chamber configuration presented a small flow improvement over the previous year's 915, it was a step in the wrong direction in terms of performance. Racers in the know have long coveted the 915's closed-chamber heads, not for better flow, or even the higher compression offered by the closed chamber's smaller volume. The racers value the heads because of the superior combustion characteristics associated with the closed chambers. The 906, and all later heads that shared the same open-chamber configuration, did have an advantage in one area--lower emissions. The less intense burn and less effective scavenging resulted in reduced levels of nitrogen oxides (NOx). The 906 casting was retained as the B/RB head for 1969 to 1970 with no notable changes.

In 1971, a new casting was introduced--the 346, again used across the board on all big-blocks, B, and RB (although some early 1971s were reputed to have slipped through with the 906s). Although the 915 of 1967, and 906 of 1968 to 1970, had the same ports, the new 346 was a significant departure in terms of intake port configuration. Exhaust ports were essentially unchanged. Factory information on the 346 casting, as published in the old Direct Connection racing manuals, and carried over into today’s Mopar Performance engine book, was limited to one line: "In 1972 (and late 1971), the new emissions head with the flatter intake port was introduced on all ’B’ engines, and has a casting number of 3462345." Other than printing the wrong year and casting number, it was clear that this was a smog head to be avoided. No supporting information was provided, but in the minds of Mopar enthu-siasts, the written words have long been accepted as gospel. Interestingly, these same smog heads were the recommended heads for Grand National racing when the Wedge returned to NASCAR, and in modified form were used quite a bit by Richard Petty in his domination of the circuit in the early 1970s. The 346 head was retained as the standard big-block head through 1973.

From 1974 to the end of big-block production in 1978, several different castings were used--the 902 in 1974, the 975 in 1975, and the better-known 452 from 1976 to 1978. All these heads were minor variations on the 346 casting from 1971, with differences associated with durability details such as hardened valve seats, improved crack resistance, and an enlarged guide boss in the valve seal area. The intake port remained unchanged from the 1971 346 casting. The exhaust port was never significantly changed and remained true to the form introduced back in 1967 on the 915 heads. To sum up, two intake port configurations were used beginning in 1967 on big-block heads. The first was the 915 closed-chamber head and its successor, the 906 open chamber. The second intake port type was introduced in 1971 on the 346 open-chamber head. This port had a substantially lowered roof and greatly reduced short side, with a "Humber Hump" configuration on the cylinder wall-side floor. This port was carried through on the 902, 975, and the last of the production big-block heads--the 452, all of which were open-chamber designs. On the exhaust side, the 915, 906, 346, 902, 975, and 452 all featured the same basic port. As a result of the need to run unleaded fuel, retrofitting to the late 452 heads, with their induction-hardened valve seats, became a popular swap on earlier engines. Interestingly, the 452 gained a reputation as a desirable performance head casting; through experience, Mopar enthusiasts discovered the performance was on par to that of the esteemed 906 castings of the musclecar era. This regard did not return to the earlier 346 heads, even though the port configuration was virtually identical.

Stock Intake Port Flow

In an effort to provide useful information about the selection of production cylinder heads, we compiled examples of popular production cylinder castings that cover the period from the major redesign in 1967 to the end of big-block production in 1978. For our evaluation, we brought our heads to the flow lab of the renowned cylinder head flow researcher David Vizard. David, an aftermarket performance consultant, univer-sity lecturer on performance engine theory, and the author of scores of comprehensive books on performance engine modifications, allowed us access to his flow lab during the seven months we compiled the information in this series of articles. His book, How to Build & Modify Chev-rolet Small-Block V-8 Cylinder Heads, should be required reading for anyone who contemplates modifying cylinder heads for improved performance. Although, as the title suggests, the focus is on the enemy's small-block, Vizard's theory of cylinder head modification and port flow dynamics is among the most comprehensive and universally appli-cable work available. As advised by ancient philosophers, "Know thy enemy, in battle," and in racing.

The intake ports of the 915, 906, 346, and 452 castings in dead stock form, with OE valves, were flowed on Vizard’s Quadrant Scientific flowbench. The results are shown in Chart 1. The 902 and 975 were omitted because the intake port configurations were essentially the same as that of the 452/346 heads. Although the 346 and 452 also share the same port design, both were tested individually since the former is saddled with a reputation as a dog, while the latter is held in high regard. As expected, the 346 and 452 heads in stock form exhibited virtually iden-tical flow. Similarly, the 915 and 906, again sharing the same intake port, flowed in virtual lockstep until 0.350-inch lift. The 906 showed a flow advantage between 0.400-inch and 0.550-inch lift, probably due to the combustion chamber differences. By 0.600-inch lift, the 906 and 915 flow curves converge, and the 906’s advantage disappears. The bottom line--from 0 to 0.350-inch lift, all 1967-and-up castings flowed virtually the same way. From 0.400- to 0.550-inch lift, the average flow difference from best to worst was about 5 percent. Again, at higher lifts there was no considerable difference. If analyzed in terms of area below the flow curve, the difference between the castings narrows to 2.9 percent, best to worst.

Making comparisons in terms of the absolute flow number on our stock Mopar big-block heads versus other common domestic V-8 engines is found in Chart 2. It becomes clear that the larger displacement Chrysler engines are somewhat starved for intake airflow. Chrysler never envisioned the production big-block head as the last word in performance. These heads were required to meet the need for bread-and-butter passenger car applications, while providing adequate airflow for street performance use in relatively mildly spec'd engines. Airflow was enough to meet the compromises required of a head to be used in a broad range of production applications. For racing applications, the specialized Max Wedge, large-port, big-block head was developed in the early 1960s to fill the void. When the competition heated up, instead of extensively redeveloping the B/RB Wedge head theme, Chrysler shattered the opposition by releasing the Hemi engine in 1964.

Stock Exhaust Port Flow

Since the exhaust port remained virtually unchanged over the years examined in our study, we would expect the flow results to reflect near-equal flow for the various castings tested. The one complication was that our 915 casting was originally the standard version fitted with the 1.60 exhaust valve. Although the seat on our 915 was re-ground for a 1.74-inch exhaust valve, the throat diameter entering the bowl was not cut open to the larger specs of the other heads. As a result, the flow was dramatically reduced throughout the lift range. Since the port is basically the same as that of the other heads, a machine operation to enlarge the throat to the other heads' 1.74-inch valve specifications would probably bring the flow in line with the others. The exhaust ports on the other heads were untouched factory stock, with OE 1.74-inch valves. Flow figures for the production exhaust ports are given in Chart 3.

Again, we find no dramatic difference in flow between the factory castings, except in the case of our 915 with its small 1.60-inch valve throat size. At lifts above 0.450 inch, our 452 casting began to show a flow disadvantage when compared to the 906/346 castings. When analyzing the stock exhaust ports, the short side exit off the valve seat is quite shallow, leaving a dead air space in the hollow area as the flow velocity increases at higher lifts. The minor difference in higher lift flow could easily be attributed to production tolerances in the casting core placement, resulting in the poor form in this area. Again, comparing the stock exhaust flow to common performance V-8 engines of the era, the Chrysler exhaust port split the difference between the efficient big-block Chevy port, and the terribly restrictive, small-valve 385 series (429/460) Ford big-block port. This is displayed in Chart 4.

Port Flow vs. Velocity

It has been established that an engine's ability to fill the cylinders is ultimately the limiting factor that allows it to produce maximum power. In normally aspirated form, the way the cylinders are filled (barring the effects of ram or shock wave tuning) is the pressure differential created by the piston moving down the bore. The filling efficiency is measured in terms of the volumetric efficiency--the ratio between the mechanical volume available within the cylinder, and the volume of mixture the cylinder actually ingests. If the cylinder is completely filled, the volumetric efficiency is 100 percent. The timing events of the valves opening and closing (a matter of cam profile) is the primary factor in determining the rpm range in which the engine will achieve peak volumetric efficiency. The airflow capacity, and the velocity of the flow, must work in concert with the valve timing in order to produce a performance engine with desirable performance characteristics. Although the flow requirements of an engine increase as rpms increase, the amount of time to fill the cylinders decreases. Once the engine's airflow demand outstrips the ability of the heads to deliver increased airflow, volumetric efficiency tails off, as does peak power.

If achieving big flow numbers was the last word in producing cylinder heads that perform well, port design would not represent a significant engineering challenge. Utilizing sewer pipe-sized ports would be the quick solution. Unfortunately, as the port volume rises, velo-city drops. A huge port between the carb and valve will create a proportionally large volume of air. Since the motion of air through the port is dictated by the increasing volume in the cylinder--as the piston descends in the bore on the induction stroke--an increase in the size of the port means the air will be lazy in response to what's happening in the cylinder.

Furthermore, since the air column is the only means by which the cylinder communi-cates with the carburetor, a large, lazy volume of air in the port is slow to provide a signal at the carb's venturi, which actually dampens the signal. The net result is a lazy throttle response--remember Ford's Boss 302? Implicit with higher port velocity is the notion that the air column is moving faster. Basic physics tells us that a body in motion (in this case, the air column) will build momentum and will want to keep moving in the same direction, even as the force acting on it is removed. In the case of the air within the port, the force acting on the air column is the pressure differential in the cylinder as the piston moves down. When the piston slows as it reaches the bottom of the cylinder at the end of the induction stroke, good port velocity is helpful to continue filling the cylinder, due to the air column's inertia (especially if there is good low-lift flow, as the valve is now closing more quickly).

In the exhaust port, many of the same considerations apply. While here, we are not drawing air in to fill the cylinder. Maintaining good velocity helps to scavenge the cylinder of spent gasses to promote good cylinder filling of fresh air/fuel charge. Consider an oversized exhaust port with poor flow velocity. In this situation we have a large, slow-moving tub of burnt mixture hanging around just outside the valve. As the piston reaches the top of the exhaust stroke, lack of inertia in our lazy, oversized port can allow the spent gasses to stop, and actually reverse direction during the overlap phase of the valve timing. This creates dilution of the fresh mix. This effect is progressively worse at lower rpm, where flow rates and velocity are at their lowest. Increased overlap with high-performance cams magnify the problem. The result--an engine that is slow to get on the cam as rpms pick up velocities in the port, and poor low-speed operation.

The bottom line: Both the intake and exhaust ports must balance flow velocity and flow capacity to achieve flexible performance over a broad range of rpms--a necessity in a street performance engine. We should maximize the efficient use of the port to achieve the required flow, instead of simply making bigger ports.

Swirl

Anyone with more than passing interest in performance engines has probably heard the term "swirl" thrown around in recent years. Swirl is considered a recent development in cylinder head design. OEMs have placed a greater emphasis on swirl because of some desirable effects reaped from achieving intake ports with good swirl characteristics. The first question is, "What is swirl?" As you might expect, swirl is obtained when the air/fuel mixture enters the cylinder in a definite rotational pattern (Helical pattern down the bore). The next question would have to be, "So what?" Research has found that in typical internal combustion engines, there is a lot of variation in the timing of peak cylinder pressure from one combustion cycle to the next. The key to power production is to manage the peak cylinder pressure in relation to the piston position during combustion. Most of an engine's power production occurs in the early stages of combustion, while combustion pressures are the highest and the piston can take maximum mechanical advantage of the expanding gasses. Unfortunately, the variation in the pressure curve from one burn cycle to the next is the most erratic during the critical early stages of combustion. This is not good, if maximum efficiency and power are the goal.

Swirl has been found to reduce the combustion pressure variation, due--in theory--to a more consistent gas mix in the vicinity of the plug when firing. This promotes more efficient combustion. It was also discovered that the action of the swirling gasses during ignition helps speed flame propagation, providing a higher resistance to detonation, and a quicker pressure rise at the critical early stages of the combustion process. Less erratic combustion boils down to the poten-tial to produce greater torque, and the quicker burn promoted by swirl increases the detonation limit. This allows for an increase in compression for even more power. Chrysler first took advantage of swirl design in the 1985 318 two-barrel, and true to that theory, increased the compression ratio. OK, swirl can be a good thing, but my 915s were cast 20 years before all the hubbub about swirl. Whether it was by design or coincidence, the production Mopar big-block heads exhibit excellent swirl characteristics. In fact, the swirl numbers for our heads in stock form compare favorably with many of the latest high-swirl head designs. Creating high swirl numbers alone is not difficult to achieve (shrouding half the valve is one quick way to do it). However, providing high-flow efficiency with adequate swirl is another matter. In the case of production B/RB heads, the trick is not to kill the swirl when modifying these heads for improved airflow.

Squish and Quench

Along with the rage in swirl-promoting designs, and after a long romance with open-chamber designs, OEs have returned to embrace the closed-chamber configuration. Again, the reasons stem from greater combustion efficiencies. A faster burn and a more turbulent mixture are obtained when the compression height of the piston is set to take the maximum advantage of the flat quench face of the closed-chamber head. In other words, if the area between the flat face of the head and the piston top is sufficiently close, as the piston reaches TDC on the compression stroke, the mixture that occupied the area over the piston is forced at high speed into the direction of the advancing flame front. The net effects are faster burn and greater turbulence at the time of combustion--advantages for efficient combustion. This is the squish effect. A second benefit is the mechanical resistance to detonation is improved when a closed-chamber head is used with a close piston-to-head clearance. Detonation begins at an area other than the main flame front. With a flat-top piston, the area opposite the spark plug in the quench area is prime detonation territory. If this area is sufficiently tight, the ability for it to detonate reduces since the combustion at the desired flame front occurs quicker (squish effect), and lessens the time for heat rise in the mixture at the far end of the chamber. Furthermore, the thin section at the squish area exposes only the small volume of mixture to an equally high surface area, diminishing the heat rise in this part of the mixture.

This explanation is where the term quench originated. Soaking this heat from the area most likely to detonate means the tendency toward detonation is reduced, and the compression tolerance is increased. Of the 1967-and-up B/RB heads, only the 1967 915s had closed chambers with the improved ports (the 516 heads of earlier B/RBs and the 1967 B-engines were also closed chambers). If we throw those 915s on a low-compression 400 B-engine, with the stock-type low-compression pistons in the bore, 0.100-inch at TDC, we pick the compression up, but the quench and squish effects are lost. That 0.100-inch (plus head gasket thickness) minimum clearance between the piston and head is prime detonation territory. The same result applies to the 906, 346, 902, 975, and 452 heads, all of which come through with a 0.100-inch recess open chamber. Filling the quench area (welding or metal spraying), or using special quench-dome pistons, has been the solution for engine builders who seek the quench and squish advantages with these open-chamber heads. The former method is definitely the better alternative.

Modifications: Back-Cutting the Intake Valves

Before we considered any grinding of the port, the first modification we looked at was the valve itself. The stock intake valve has a lip backing the valve seat face, which acts as a ski jump to the airflow on an engine where the airflow predominates across the back of the valve. The first step was to remove this lip with a 30-degree back-cut to the valve. This simple modification was applied to a 915 and 452 casting, representing both intake port types in the production heads. As seen in Chart 5, the result was a dramatic increase in low-lift flow--0.350 inch. Gains of as much as 18 percent on the 452s, and 15 percent on the 915s, were achieved at the lower portions of the lift curve. Even if porting is not considered, a back-cut valve represents free and easy horsepower.

Modifications: Back-cutting the Exhaust Valves

Similar flow gains can be achieved by back-cutting the exhaust valve, as seen in Chart 6. However, the lip on the exhaust side does serve to impede flow reversion--we've already discussed that.

Modifications: MP Porting Templates

When modifying the production head castings, one of the first tools for most Mopar enthusiasts is a set of Mopar Performance P4120437 porting templates. Developed many years ago, these templates were designed as a guide to porting the old Stage IV Direct Connection performance castings available about 20 years ago. The templates also work with production castings, and since the demise of the Stage IV, this is where the templates are usually employed. The porting templates provide a clear-cut guide to opening the bowl area to about an inch under the valve. Because with most heads, the greatest flow improvements are made for the least effort in this area, we began our evaluation using the templates as a guide.

The concept was to template-port each of the OE castings, and see how the various castings respond to this standardized modification. The back-cut valves were just as effective in the template-ported heads, and were installed for the tests shown here. Chart 7 tabulates the results. The 346/452 ports showed a clear advantage at higher lifts in comparison to the earlier castings, while the earlier ports maintained an advantage in the midrange lift. All the heads showed improved flow throughout the curve, but the late castings were more responsive to this level of modification, with an average improvement over the lift curve of roughly 12 percent. The average flow improvement on the early-style port was slightly better than 8 percent. On the exhaust side, with the same template-porting modifications and back-cut exhaust valves, flow for the various castings were equivalent in Chart 8. This is not a surprise, though, considering the port's similarity in all heads tested. We followed the same theme with an average gain over the curve of 7.5 percent.

Although we gained a decent flow improvement, we were a little disappointed with the results. The feeling was that there was a great deal more flow to be achieved from the factory castings. We had done the standard no-brain modifications. It was time to hunt for some serious flow. The template modifications shown here are a solid base from which to start, but don't put your die grinder away just yet, and don't draw too many conclusions on the relative worth of the castings. We've left some serious flow on the table, and next month we will discuss what's involved in serving it.