In Part I of this series, we looked at a variety of production big-block cylinder head castings from 1967 through the end of big-block production in 1978. We've discussed production history and design changes, identified the stock port types, and compared the stock flow. We've also looked into some of the theory and characteristics of cylinder head design, including the balance of port flow and velo-city, quench and squish effects, and swirl. We've considered valve modifications and their effect on airflow. Finally, we've compared the flow of the various ports when the stock valves were modified, and the port bowls opened with the Mopar Performance porting templates.
In this story, we will examine more serious intake port modifications and the gain in airflow provided by such mods. All the modifications in the tests shown were performed by the author of this article, and cylinder head flow expert David Vizard, at Vizard's flow research facility. Cylinder head porting is not for everyone. Don't be misled into believing that a novice with a couple of spare hours and a few grinding stones can magically transform 30-year-old production passenger-car cylinder heads into the last word in race heads. Many pros, who have been at it for more than 30 years, still admit there is more to gain. That is not to say porting can't be done at home--just realize that porting takes a great deal of time, effort, skill, knowledge, and patience. The results of a well-ported set of heads can be truly phenomenal. Professionally prepared heads that utilize the modifications shown here are available from Jerry Goodale at Specialized Motor Service. Jerry has spent years working on Chrysler cylinder heads, and has also supplied ported cylinder heads for the pro-race motors of Walker Evans.
The do-it-yourself porter, beyond having a good assortment of grinders and cutters, must know how to use them with a deft sculptor's touch. Although we will forewarn you of the major danger spots, overzealous grinding will result in the destruction of a perfectly good stock casting in the search for flow. Michelangelo's masterpieces have immense value, even though he started with just a hunk of worthless rock. A stock head is the porter's worthless rock--the value lies in the know-ledge and labor that transforms the head into a performance masterpiece. The truly serious home efforts will benefit immensely from access to a flow bench to evaluate the results. We will review a wide range of modifications that are proven to work, and also check out a few commonly performed pitfalls that don't. The intangiblity is in the execution. By carefully following the concepts shown, the production Chrysler big-block heads can deliver outstanding performance.
Terms
Before we get into a discussion of the modifications to 440 heads, we ought to take a minute to define the various parts of the port, as identified in the text.
Short-Side Turn: This is the turn from the port's floor to the valve seat below, so called because the turn is at a short radius.
Long Side: The path along the roof and down to the seat.
Port Wall: The port walls are the vertical sides of the port. Since paired ports are mirror images of each other, "right" and "left" won't adequately identify the sides. The port walls are identified as they relate to the cylinder wall.
Inside Wall: This is the port wall which leads to the center, or inside, of the cylinder (or chamber). On the 440 heads, it is the pushrod side, or the curved wall.
Outside Wall: This is the port wall which leads to the outside, or cylinder-wall side of the cylinder (or chamber). The outside wall is the straight wall and is the center, or siamesed wall, of the pair of ports.
The Ports
As we pointed out in Part I, Chrysler produced six different production cylinder head castings, from the major redesign in 1967 until the end of big-block production in 1978. Essentially, the intake port configurations can be distilled down to two basic ports: the early intake port, found in the 1967 915 open-chamber casting and the 1968-70 906 closed-chamber casting; and flatter intake ports found on later heads (these are all open-chamber).
The Early Intake Port
One of the first common modifications to the intake port is replacing the factory 2.08-inch valves with the readily available 2.14-inch valves. The familiar advice is to have the seats enlarged for the larger valves and blend the seat's bottom cut into the port's bowl below. These suggestions represent typical performance effort. We put a set of 915s--prepared in just this way by a local machine shop (Photo 1)--to the test. The flow figures are shown in Table 1, Column 3. The flow was abysmal, barely edging out a totally stock 915 (Column 1). If we compare the template-ported 915 fitted with a back-cut stock 2.08-inch valve from last month's Part One (Column 2), we will notice that the big valve head was crushed through the midrange, and caught up at the top of the lift curve.
The lesson here is with the early ports, getting improvements in airflow is not as simple as installing a larger valve. As we will see, in the early port, the 2.14-inch valve does not provide the flow advantage we might expect, without complementary modifications to utilize the valve's greater flow potential. How would the 2.14-inch valve operate in a template-ported early head? The templates prescribe more extensive bowl grinding than what was done on the 915 head, resulting in a larger throat leading to the seat. We installed the larger valve in a template-ported 906, using the usual three-angle 30/45/70-degree seat, and a de-shrouding cut radially out to the chamber wall. The results can be seen in Table 2, Column 3. The 2.14-inch valve worked much better with the bowl work, soundly beating a stock head. Compared to last month's template-ported 906 with a back-cut stock valve, the 2.14-inch was weaker at low lifts, but showed an advantage from 0.350-inch lift-and-up. Overall, the improvements with this basic setup were far more modest than normally thought.
Flow Separation
Studying the flow numbers on our 906 casting (Table 2), one can see that as the flow increases, it reaches a peak and then tails off, even as the valve is opened further. Why the drop-off at higher lifts? It's called flow separation--look down the floor of the early intake port (Photo 2). The floor comes in at the manifold face at quite a low angle to the valve, with the air building speed as the port tapers down the straight portion of the runner. Toward the end of the runner, the floor forms a ski jump, reaching a ramp that angles upward toward the roof, then the bottom drops out at the shortside turn. At that turn, the floor abruptly changes profile from the rectangular-port form to the round profile off the seat--with no transition. The early port's form makes it difficult for the airstream to make the turn-in toward the valve. As the flow (and hence, velocity) reaches a critical point (at 0.400-inch lift in our last test), it is simply deflected off the floor and blows the turn completely.
It's not the greater valve opening in itself that causes a port to stall--at greater valve lifts, we move more air through the ports. If a problem exists with the port dynamics that will cause the flow to stall, it will happen at a critical point in the lift curve, when a critical flow rate is achieved. This flow separation is a major problem with the early port, especially as port flow increases. This is a dilemma even in a stock port. Overcoming this deficiency is the major battle in getting the early intake port to perform.
Shortside Mods
It is ironic that the old Direct Connection tuning manuals and the Mopar Performance Engines book have always advised to avoid modifying the shortside. Now the contro-versy--this is perhaps the most critical area for getting the early-port heads to work! Two things can improve the situation, in terms of getting the air to flow around the corner with the early port. First, the floor profile can be modified to provide an improved flow path. Second, like a race car braking before entering a corner, the airstream can be slowed down as it enters the bowl and begins to turn in toward the valve. Slowing down the air for a given level of airflow is accomplished by increasing the volume of the port in the area where the air needs to slow down.
At Vizard's facility, we achieved the best results by utilizing a combination of the two modifications. First, the shortside form was reprofiled to drop down at 90 degrees nearly right off the seat. With the larger valves, a considerable amount of material overhangs in the flow path to the valve seat. In extreme examples, such as the 915 casting tested in Table 1, Column 3, this would require the flow to make a U-turn to get around the shortside. The transition around the shortside to the floor was blended to provide as large and as gentle a radius as possible. The abrupt transition from the rectangular port to the round seat was modified by putting a slight convex shape into the shortside--from the round valve seat to the port floor. A word of caution about dropping the center of the shortside floor to get the rolled convex radius--there is water under the flat area of the floor, where it approaches the shortside turn (Photos 3 and 4). So, the amount that this area can be reprofiled is limited! At the place where the template-ported, big-valve head shut down at 0.450-inch lift, with the modified shortside, the port picked up almost 20 cfm. That was a huge gain at a quite usable part of the lift curve (Table 2, Column 4). The lower-lift flow was substantially improved as well, and gained between 10 and 15 cfm throughout the midrange. The port was now able to handle 235-plus cfm before reaching stall velocity, and this flow rate was reached at 0.450-inch lift. This was a tremendous improvement, but more help was necessary.
Increase Volume
The next modification involved 1) modifying the profile of the guide boss, 2) rolling the boss' sides around the back of the guide, and 3) cutting back the web between the guide and the pocket (Photo 5). The theory is to increase the volume in the tight bowl area and allow the air to slow down--this helps it use more of the valve. The guide boss is a large chunk of solid cast iron that takes up space. By cutting it back and rolling the transition from the roof into the pocket, the volume here is increased, and the form is vastly improved for airflow. The results of this modification included significant gains from as low as 0.150 inch, and huge gains at the higher end of the curve (Table 2, Column 5). Stack on another 9 cfm at 0.450 inch over our last test, and 21 cfm at 0.500 inch! These gains are quite usable, even in street applications, with top-end flow now reaching real performance levels. At this point, the port stalling problem was greatly diminished. Our modified head now had better flow at only 0.350-inch lift than the stock head achieved at 0.650 inch, with a 25-percent flow improvement at 0.500 inch, and we had untouched, raw cast iron covering 90 percent of the port!
Take Care!
Cutting back the web offers great flow gains, but it is precision work. The web itself is solid from the guide back. However, water dwells where the web meets the pocket and roof. The cut must be made tight to the back of the guide, removing the minimum amount of material to roll it into the pocket and roof channels (Photo 6). The final form will change from the original large block to a rolled-up wedge, only slightly lower at its peak. Blending the roof channels into the pocket also is tricky with the early intake port because of a hollow spot in the pocket, in the path of the join. If we dig this area out to make the transition perfectly smooth, we will get dangerously close to water in the areas on either side of where the guide-boss web was cut! Grind judiciously, and compromise the form slightly to be safe.
Port Myths
Note that up until now, the restriction at the port where it is kinked in for pushrod clearance has not been touched. This is where the novice head porter usually starts to grind, since it seems as if it's a major point of restriction. But the truth is, this modification doesn't have any effect until the rest of the port is able to handle serious volumes of air. At more than 255 cfm on Vizard's bench, we were well past Max Wedge flow levels. Opening the pushrod restriction on a less developed port can, in fact, cost flow. Our next modifications began with opening the port at the pushrod restriction by nearly 0.100 inch, thus working to get a smooth, blended transition on the inside wall. On the outside wall, there is a hump for the rocker oil feed, which can be blended down considerably (don't strike oil by cutting through). The rest of the outside wall's profile was made as straight--leading to the pocket--as practical. With the port walls finally done, the transition from the walls to the floor, and into the shortside turn, were blended for a smooth profile with a minimum of material removed. The flow (Table 2, Column 6) again was improved from down to 0.150-inch lift, with the entire curve fattened up nicely, up almost 8 cfm at the high end. The airflow improvement in the lower end of the curve probably was from the blending of the corners into the shortside turn. The top-end improvement was more likely a result of reducing the pushrod restriction and straightening the outside wall.
The final modifications to this head were in the roof area. The roof channels were widened by cutting the guide boss' sides into a more square profile. Further widening was done at the outside wall only, and was achieved by cutting back the outside wall adjacent to the roof channel. The lead-in of the guide boss on the inlet side was also reduced substantially in profile and streamlined. The last modification involved a transition--the port runners into the pocket along both port walls were mildly blended, for a smooth profile. All this work had us knocking on 270 cfm at very high lifts, which--on Vizard's Quadrant Scientific bench--was cooking. Results down low were mixed (Table 2, Column 7). All in all, our modified early port was now moving copious quantities of air throughout the lift curve, and actually surpassing many pure race heads in the lower-lift ranges. The downside is that getting the early port to work requires a concentrated effort, and subtle changes in porting form dramatically influence airflow. This port is sensitive.
Take a Seat
The importance of a well-designed valve seat cannot be overlooked in the bid for improved flow. Racers go through great pains to optimize seat design and gain a competitive edge, especially in classes where port modifications are restricted. At low lifts, the valve seat form and the valve curtain area (valve circumference x lift) are the primary determinants of flow. A poor form can adversely affect flow throughout the curve. With a given valve diameter, the seat form must be optimized to get the most from the package. Traditionally, a high-performance Wedge, three-angle valve job is performed with the stock 45-degree seat angle, with a 30-degree top cut and a 70-degree bottom cut. While the production seat angle on Chrysler engines is 45 degrees, a properly done 30-degree seat often can pay big dividends, in terms of low-lift flow. Because of the geometry involved at low lifts, the 30-degree seat will open more area to airflow for the same amount of valve lift--this can provide an advantage. Switching from a 45- to a 30-degree seat can be done only when the valve size is increased.
With the Chrysler big-block production heads, we can take this opportunity when substituting the common 2.14-inch valve for the stock 2.08-inch valve. We have consistently found flow improvements on the order of 25 to 35 percent at up to 0.050-inch lift, and 10 to 20 percent at 0.100 inch lift, by using the 30-degree seat. Of course, the angle on the valve face will require a change to a matching 30 degrees, as well as a mandatory back-cut at a shallower angle to reduce the face width (the actual angle will depend on what works with your valve profile). Cutting the 30-degree seat is a precision operation best performed with a Serdi-type seat machine, which uses a cutting tool rather than stones. The configuration we used is a 10-degree top cut, followed by the 30-degree seat 0.050-inch wide, with a 0.100-inch radius coming off the seat 15 degrees from tangent, and a tangent join to a steep 75-degree throat. Say that three times fast!
The Chrysler big-block intake ports, with their characteristic flat-approach angle to the valve, are especially responsive to this modification, and consistently show dramatic increases in low-lift flow. For lifts up to 0.250-inch on modified heads with the 30-degree seat, the ports obtained phenomenal flow levels. Remember, the valve may kiss a high-lift number and those big flow figures once at max lift, but it will visit each of those lesser points on the curve a second time on the way back down. Typical improvements in low-lift flow over a conventional three-angle seat are shown in Table 3.
Chamber Mods
The act of polishing the chamber existed for as long as people have modified cylinder heads, and it's still a good idea today. The smooth chamber surface will markedly reduce carbon buildup in a street engine, remove pre-ignition-inducing hot spots, and sap less power from the burn in terms of heat loss. The raised ring--where the as-cast chamber surface meets the seat's top cut--should be blended smooth. An old valve of slightly smaller diameter than the final valve size should be dropped in, to protect the seat area during polishing. A once-over with a fine grinding stone, followed by progressively finer sanding rolls, gets the job done in short order.
The effects of valve shrouding can have a serious impact on both intake and exhaust flow, especially as the valve size is increased. With an inline-valve engine such as the Chry-sler big-block, the valves open straight down, keeping the outside edge of the valve in proximity to the chamber wall. With a Hemi or canted-valve engine, the valve moves away from the cylinder and chamber walls as it opens, thus minimizing the shrouding effect. When installing larger valves in our inline-valve big-blocks--at a minimum--the chamber wall adjacent to the valve should be plunge-cut at a fixed radius from the seat, and out as far as the gasket line will safely allow. We performed this modification as a matter of course in all tests, with larger valves presented here except where noted. We gained as much as an additional 8 cfm at high lifts by hand-blending the plunge cut into the chamber wall, up to the gasket line. The cut made on closed-chamber heads is slightly different than the simple blend done on the open chamber, as shown in the photos.
The Late (Smog?) Intake Port
As we saw last month, the later flat intake port introduced with the 346 casting in 1971 flowed virtually the same as the early port up to 0.350-inch lift, with the early port gaining a small advantage from 0.400 to 0.550-inch lift. Above this, they all flowed practically the same. This would indicate a small edge in performance by the early intake port, with stone stock heads and stock lift levels. The late port responded much better to our basic template porting and the backcut stock valve. With these modifications, the late-port had nearly a 20 cfm advantage over the early port at lifts above 0.500 inch, and was noticeably better above 0.450 inch. In this form, the early port seemed to respond slightly better in the midrange, gaining a small edge up to about 0.350-inch lift. With a street cam and template-type pocket cleanup only, it seems like a wash when running the stock-size valves. The early port showed a slightly fatter curve in the midrange, while the top-end flow advantage of the late port was just beginning to come into play with typical street-type valve lifts.
The attention-getter for us was that the late port didn't have the same tendency to stall at higher lifts. This means more air could pass through before battling the port dynamics that cause the early port to stall. In the early-port testing, we started right off the bat by installing the larger valve. Since the late port responded so well with just the template pocket cleanup, we decided to modify it while retaining the back-cut, stock 2.08-inch valve, to see what it would do. The first modification was to blend in only the back of the guide boss, cut back the web, and blend it into the bowl and roof. With the flatter roof, and without the hollow spot in the bowl of the early port, this task was much easier to perform, and required a lot less dangerous metal removal to get a nice form. The results of the first modification are shown in Table 4, Column 3. The entire curve was fattened up considerably, with a remarkable peak gain of 11.6 cfm at 0.300-inch lift. Next, we moved to the shortside radius--what little of it is there. The shortside of the late port comes in quite flat, with the odd "Huber Hump" on the outside corner. The shortside was given as large and smooth of a radius as possible, removing a minimal amount of material and fully removing the sharp ridge from the production throat machining. The results are in Table 4, Column 4. High-lift flow was increased dramatically, gaining 16 cfm at peak flow. The curve was beefed up noticeably below 0.250 inch, with not much change in between.
At this point, the lead-in of the guide boss was still untouched, so it was narrowed and blended into the roof. The guide boss on the late port is much smaller and extends far less down the port in comparison to the early port, resulting in a much simpler task. While we were at it, the transition from the roof to the walls was given a tickle adjacent to the guide boss. The flow from Table 4, Column 5 again showed a strong improvement from as low as 0.100-inch lift all the way to the top. Top-end airflow broke more than 250, which on our bench means it's working, while the mid-range number of 192 at 0.300-inch lift was outstanding. At this point, this mildly reworked 346 "dog" head with a stock 2.08-inch valve and the entire runner left untouched would kill many ported early heads with a 2.14 valve.
Since we were looking for more airflow, we decided to go for the kill and open up the pushrod restriction at the manifold side by 0.100 inch. As shown in Column 6, nothing was there. In fact, the whole curve went on a diet. The runner was not the major restriction we would imagine at this flow level. We attempted to redeem the effort, so we returned to the runner and blended the outside wall, where it bulges, to the oil feed hole, lightly blended the "Huber Hump" in the shortside, blended out the lump from the valve cover bolt boss, and never moved closer than 1 cfm from where we were in Column 5.
30-Degree Seat
It was clearly time to move on and try the late port with the bigger 2.14-inch valve. The seat was cut with the 30-degree seat angle discussed above. Why throw away 25-percent better low-lift flow? With no other changes except for blending the bottom cut of the new seat, flow figures in Table 4, Column 7 were obtained. The 2.14 valve and the 30-degree seat bumped low-lift flow to nearly 30 percent more than our last test at 0.025-inch lift, and 22 percent at 0.050 inch. The already excellent midrange flow reached astoundingly fat numbers, with strong double-digit gains all the way up to 500-inch lift. The top-end flow showed moderate improvement. In this form, this is a killer head for street or race applications, with moderate cam lifts. From our testing, we know that when the 2.14-inch valve is installed in an early port with only the template bowl cleanup, the resulting flow is disappointing. It's not until the port is substantially reworked--to allow for greater flow without stalling-- that the early port can take useful advantage of the bigger valve. Since the late port has shown less of a ten-dency to drop off at higher lift, we wondered how it would respond to the 2.14-inch intake with only a template-type bowl cleanup.
We took a fresh template-ported late-port head (a 452 casting this time), and cut it with our favored 30-degree seat. With just a two minute blending of the new 75-degree throat cut into the bowl, and around the shortside, the flow figures in Table 4, Column 8 were obtained. Flow improvement was very strong throughout the lift curve, compared to the same type of template-bowl porting with the smaller 2.08-inch valve. The 2.14-inch valves definitely work in the late port, even with only minor bowl porting. In a mildly modified form, with the 2.14 valves, the late port is clearly superior to a similarly modified early port.
Intake Port Type: What's the Call?
It was crystal-clear that getting the late-type intake port to work is much easier than the early port. The early port can be made to flow bountiful quantities of air, but this is a tricky, dangerous job. How many guys are running around kidding themselves with 915s or 906s that pack big valves and a pocket blend, such as the ones we tested in the third columns of Tables 1 and 2? We would guess too many. For the home porter--or even some pros--without access to a flow bench, to get a modified early port to work as well as a modified late port is unlikely. The early port can be thrown off by 0.010-inch of metal in the wrong place, and requires serious metal carving to reach the flow gains outlined here. The late-port modifications are more intuitive once you know where to cut, and require much less heavy-handedness with the grinder to gain meaningful results. The intake port result? For the average home porter, that old geezer Cordoba's 400 could have better heads than that Road Runner's 440+6.
Both of our intake ports showed gains in total area under the flow curve of close to 25-percent-over their respective stock ports. That's more than three times the flow gain achieved with last month's mods on the early port, and more than twice the improvement on the later design. Both ports were significantly fatter from way down low, all the way up to maximum lift. Pick any point along the lift curve, and the total area improvement will remain relatively constant, making these modified ports extremely flexible in a wide range of applications. Is there more flow in the stock heads? Of course--our work with the late intake port ended without even looking at the effects of chamber mods, and experimentation with even bigger valves seemed promising. How much is this worth in the real world? We'll put the heads on the dyno in an upcoming issue to find out. In the next issue, hang on as we discuss the exhaust port mods and more surprises on the intake side.
