AFR Eliminator 195 Street Heads
Designed for emission legal street use on small block Chevy’s - GFN tests to see just how good they are.
Good things rarely happen over night and so it was with the AFR Eliminator series of heads. They were out a while before I got to even look at them. Here things are about to be set straight.
History
The development of the 195 Street Eliminator heads from AFR started back in early 2004. The goal was to develop a head that would produce very positive results for the most popular displacement of small block Chevy’s built. That is 350’s and 383’s. The target was as much flow as possible without an overlay large port and a format that could be rapidly CNC machined and therefore contain costs. The question we pose here is how well were those goals met?
Starting Point.
Being continually busy and caught up in the bustle of exciting projects often causes me too loose any realistic perception of passing time. The case in mind here is the precise point at which I started to take an interest in the AFR 195 cc intake runner Eliminator heads. I know the ‘how’ but just not the ‘when’. As for the ‘how’ this happened while visiting a big league NASCAR engine builder buddy of mine who, as it happened, had just built a street 383 using a set of the 195 Street Eliminator heads. When I asked how they performed his answer was pretty emphatic – ‘--- made about 30 hp more than I expected and 10 more than I hoped for’. At the time I thought this was an interesting answer from a guy who spent his working life building high output Chevy engines and about 20% of that time dynoing them. But, to an extent, caution stepped in here. I reasoned that this guy had way more experience building $80,000 race engines than $8,000 street engines. What might look good to him may only be what is expected to someone who builds street motors almost all the time. Still, I felt such a creditably expert opinion should not be dismissed too lightly and decided I should keep this experience in mind.
Some time later (I can’t say just how much – it’s that time lapse problem again) I am over at T& L Engines in Stanfield NC (20 miles due east of Charlotte) when the boss, Lloyd McCleary, asks if I can get hold of a set of AFR 195’s really fast as he a) has a customer who wants them on his 408 small block Chevy, and b) this is an opportunity to see how well they may work. The problem is this all has to happen in about 4 days. At this point I well appreciate that this is an opportunity for me to see what goes with these heads so I call Rick Sperling, the boss at AFR to see if he has any in stock ready to go and if so can he next day air them to us. The answer is yes and yes so, though somewhat of a rush we looked ready to go.
The 408 Build
The short block for this AFR Eliminator 195 head test was a nice but not exotic parts combo. First you need to understand that getting 408 inches from a 350 block entails more than just dropping in a long stroke crank. Our story here is about the heads but just for the record T&L builds these stroker’s by first selecting a block with a sonic tester. These blocks are then cut to clear a modified stroker rod and then pressure tested. Only half the blocks make it past the sonic test and only 2/3rds of the remaining blocks pass the pressure test after being cut for rod clearance. So getting 408 inches entails scrapping a block or two – but if you want 408 inches that’s the price to be paid.
For this particular application the crank and rods T&L uses are from K1 Technologies. These are nice looking pieces at a cost effective price. For all that stroke length to clear even in this specially cut block some mods are required to the K1 rods.
The crank and rods were mated with Mahle pistons to complete the internal rotating assembly. For what it’s worth those of you currently planning an engine should seriously look at these pistons as they have much going for them – especially in a stroker motor. What makes them a good choice, among other things, is that they have smaller cross section area rings which allow them to seal up with less bore friction. The longer the stroke the greater the loss due to friction so this piston design tends to counter this negative aspect of a longer stroke.
Light weight, skinny rings, anti-friction skirt coatings and an overall stout construction make these Mahle pistons a good choice for a high performance stroker application.
Before we actually get to discuss the AFR heads it looks to be worthwhile to just point out the merits of testing on a 408. Let’s face it; at the end of the day power is all about cylinder heads being able to satisfy the air demand imposed by the bottom end. This is dictated by displacement times rpm. Our 408 will create a big air demand without having to turn as many rpm as would a 350. Putting some numbers to this we see that if the 350 targeted say 8000 rpm the a 408 would only have to turn 6960 rpm. As it happens our test engine peaked at a little over 6100 rpm. That’s the equivalent of 7100 rpm for a 350. The advantage of the cubes, as apposed to rpm, is that the valve train is not so fussy as there is more time to get the job of opening and closing the valves done and that in turn means less spring force. The cam by the way, for this build was a Comp Xtreme hydraulic roller speccing out at 230/236 at 0.050 tappet lift.
Another important aspect toward supplying big inches with air is high valve lift. Here a set of Comps 1.65/1 and 1.5/1 on the intake and exhaust respectively were used. The added duration on the exhaust compensates for the lesser lift of the 1.5/1 rocker on the exhaust. Also the heads we are testing have excellent flow throughout the lift range.
The Eliminator 195
So much for our engine spec. now let’s look at the heads we are going to test in a little more detail. You can get most of the pertinent facts pertaining to these heads at
AFR Cylinder Heads Inc: Aluminum cylinder head manufacturing and flow dynamics so I won’t dump the specs on you here. What we will do however is focus on the most important issues as they relate to power generation.
Here is an example of an AFR 195 Eliminator head. There is a lot more to it’s ability to make power than just a favorable set of port cc’c and flow numbers.
First valve sizes; these heads are equipped not with the usual 2.02/1.6 valves sizes but 2.05/1.6. Granted that’s just a relatively small increase for the intake but it’s a step in the right direction and can have more of a positive effect than might be otherwise supposed on an air starved engine. Historically AFR have been good at dimensioning valve seats that allow good flow from just a mite off the seat right through to 0.700 lift. When we talk of seats here that really encompasses the area about a quarter of an inch before and after the actual contacting seat area.
These heads are CNC ported – that’s the only way you can get them. For the most part CNC porting costs quite a bit more than as-cast heads – but do enough of them and the price starts to drop through economy of numbers. But that will only take you so far. Every pass along whatever axis the CNC mill is working on costs money. The more passes made the finer the finish and the more costly the end product becomes. The question is knowing where to apply a fine finish and knowing where a course finish will do or even be better. Part of the equation here is answered by the wet flow work the late Ken Sperling (Rick’s dad) did starting way back in the mid 80’s. Take a look at the Eliminators port finish and you will see a surface that in many respects is analogous of the surface of a golf ball. A note for those who maybe did not know but without the dimples a golf ball will fly only about 2/3rds the distance on a drive. At the end of the day AFR expertise and volume production bring us a cylinder head for an extremely (and I do not use that word lightly) cost effective price.
What you see here is the combustion chamber (available in either 65 or 75 cc form), the 195 cc intake port and the 65 cc exhaust. Note the fine finish in the combustion chamber but the coarser finish in the ports.
OK lets get down to tests on the flow bench. Because of the short time scale I had to work with here the flow tests had to be run on T&L’s flow bench. For what it’s worth Lloyd had me calibrate it about 6 weeks previously so I doubt it was much off the mark.
The graph below shows the numbers we achieved. They were quite a bit higher than AFR advertises so we ran them a second time and pretty much got the same results. At worst I would say that our numbers are within 3% so the 295 cfm at 0.700 lift we measured could be as low as 286 cfm but that is still better than the 280 AFR claim. But I am sure that few of us would complain about getting more airflow than we had bargained for.
Our figures were somewhat better than AFR advertise but even allowing for measuring tolerances our numbers would still be better than theirs. Assuming that to be the case we can hardly fault them for selling us too much airflow!
Sections through the intake and exhaust port reveal much about the form AFR uses to get the results seen. Note the traditional ‘ski-jump’ intake form on the port floor is all but gone.
Along with good flow the swirl was also good so this combined with a relatively small port one a large displacement engine should result is excellent port velocity which hopefully, will lead to good torque.
As good as the flow figures were for a moderate size of runner there is a lot more to making power when ever a hydraulic roller cam is to be used. The dynamics and the vector forces involved with a hydraulic roller are such that for a given spring load it is more likely to collapse (see the hydraulic roller oil story at
Hydraulic roller race oil - what's it worth??? ). What this means is that no matter how good the heads may be if the valves are not opened and closed as they should be the whole engine will be a failure as a big power producer. I am using the word big because that is exactly what I mean here. I have seen a 100 hp lost through poor spring selection and the use of a set of factory lifters that were barely good enough for a stock application let alone something intended to go deep into the 500 hp zone.
To counter the possible effects of indifferent lifter performance when a hydraulic roller cam is used AFR have taken some drastic steps. First the valves. Instead of being the regular 11/32 (0.341) stem diameter those used are 8 mm (0.315). Along with this the valve heads are thoughtfully trimmed of excess mass. A typical intake/exhaust valve weight is 129/115 grams. The AFR ones weighed in at 105 and 99 respectively.
More effort than usual has gone into the shaping of these valves. Not only has the weight been minimized but also the form for good flow has been given a high priority. Not the radius on the edge of the exhaust valve. Not only does this aid low lift flow but also cuts the valve temperature.
Saving valve weight is good but at the end of the day it is the spring that finally dictates whether or not the valve will open to the dictates of a hydraulic roller cam profile transmitting motion through a less than perfect lifter. This is where the AFR heads score big. The spring used is not particularly strong but because it has low mass and consequently a high natural resonant frequency it gets the job done without requiring as much delivered force to do so. When I first check out the springs used I was very surprised at the apparent quality of the spring. For a valve spring like this to be installed on a head costing as little as these AFR heads do I can only surmise that they must be buying them in huge numbers or stealing them!
Both the springs shown here deliver about the same poundage seated and over-the-nose. Note how much smaller in diameter the AFR spring and retainer are.
A typical quality spring for a valve train such as we are dealing with here is a 1.450 (or so) diameter item that weighs in at about 110 to 115 grams. The retainer for this weighs in at another 33 grams. By comparison the very high grade AFR spring, at 1.240 diameter, weighs in at 92 grams and uses an 18 gram retainer. From experience I can say that this move can be worth anything up to as much as 500 rpm more before a hydraulic lifter gives up with 300 or so being relatively common. In most case where the heads are being used with a stock or near stock cam AFR efforts in the spring and valve department seem to get the job done and rpm’s as high as can be expected for a stock cammed engine are more than satisfactorily achieved. For bigger cammed motors AFR has another move up their sleeve. This is their hydraulic roller rev kit.
This is AFR’s hydraulic roller rev kit. It’s overall effect on the valve train depends very much on the starting point. Even with the best lifters it is worth extra RPM.
Before getting into this Eliminator test an opportunity to run a before and after test on the AFR rev kit presented itself. Dusty had to tear down one of our 383 mule motors which had a set of Dart heads on and had stock factory hydraulic roller lifters installed. With no other change the rev kit was worth a solid 500 rpm.
Here is how the AFR rev kit installs. It is easy with the heads off and a gross inconvenience with them on.
We also tried this rev kit with some decent aftermarket hydraulic rollers and found less of an increase in rpm to valve float although with this rev kit and some beehive springs we did run a 0.566 lift valve train to around 7350 rpm.
In the case of our first test engine we had a Comp valve train from cam to rockers. With a similar setup employed (beehive springs) we had already run to about 6500 rpm so expected to be just fine without a rev kit as peak power on this engine was expected at about 6100 rpm.
After bolting on all the usual stuff like the oil pan (Moroso) intake manifold, carb etc the engine was ready to go
With that we hit the dyno and broke our stroker 408 in. Once broken in and serviced a real effort was put into calibrating the carb and ignition for a smooth idle. This proved a worthwhile exercise as even the moderately big numbers of our Comp hydraulic roller cam were tamed by the 408 inches within. The result was a totally stock like idle. After idle calibration pulls were made to calibrate the carb. This was another one of those incidents where the AED carb was all but on the money. I say ‘all but’ with good reason as we tend to look at minor carb calibration tricks learned over the years. The result was individual calibration of the corner of the carb feeding cylinders 5 and 7. The results were as per the graph below.
To get a better handle as to what these heads might be worth compared to a set of as-cast heads the average output of three similar spec 408’s with as-cast heads are shown on the graph. But before jumping to any concrete conclusions here let’s be sure that we fully understand the limitations of this comparison.
The three test engines were done some time before the test of our AFR heads. Since doing these tests the spec of some of the brands used has moved on and will produce better results if tested today. This means we have to put things into prospective here. The peak numbers involved were 510 lbs-ft and 494 hp for the average of the as-cast heads. For the AFR Eliminator heads these numbers were 532 lbs-ft and 548 hp. On the face of it the AFR heads look to be good for peaks as much as 22 lbs-ft and 54 hp over a set of as cast heads. And maybe they are this much better than some of the brands out there but, from subsequent tests, we know that at least one brand of as-cast head will make about 20 hp more these days than when our original tests for the above curves were done. So, while not a true back-to-back test what we have shown here is the realm of improvement the CNC AFR heads can delivered. These gains are such that even if our as-cast head output numbers were off by fifteen to twenty hp it hardly matters as the AFR heads are still a big step forward.
The forgoing may look like a damper on AFR’s bragging rights to claim, from our tests, a 54 hp increase over their competition’s as-cast heads. However there is still one factor here that has yet to be discussed. With all the as-cast heads we had plenty of time to extract the best by virtue of the rest of the engines spec and by a good dyno setup. When these AFR tests were run we (T&L) had about 5 days total to build the engine, dyno it and ship it. There was no time to experiment with intakes to find what these heads liked best. Nor time to swap carbs or try different cams. In round numbers the AFR heads allowed our 408 to deliver, from a 10.5/1 CR, 1.3 lbs-ft per cube and 1.34 hp per cube from what proved to be a relatively small cam for this displacement of engine. By any standard these are great numbers for what is essentially a bolt it together, low cost engine (this very motor shipped complete, turn key, for a little over $6200).
The numbers produced are more than sufficiently good for us to say ‘next time we build and test for the best combo for these AFR heads’. That, more or less, is just what we did.
First a 383 utilizing a Scat cast crank and their budget stroker rods plus a set of Mahle pistons made up the bottom end. For a cam we used a custom grind single pattern Comp Xtreme Energy solid street roller grind # 4875. This profile has a duration of 248 degrees at 0.050 and an advertised duration of 286 degrees. Duration at lash is some 12 degrees longer than the advertised duration. The observed valve lift (after lash) with 1.6/1 Comp rockers all the way around was just shy of 0.600. We tested with both a single and two plane intake and the results are as you see below.
Peak torque was good in both instances. If drag strip times were the principle requirement the single plane would be the fastest. But take a look at the curve produced by these heads paired with a Professional Products Crosswind – that’s the black curve on the above tests. What we see here is a totally streetable 535 hp and a great torque curve topping out at 503 lbs-ft.
Solid Cam 408.
Our next project was a 408 build much along the lines of the one discussed earlier but, instead of a hydraulic, it used a slightly bigger Comp solid street roller cam. The profiles chosen were 286XSR (grind # 4875) for the intake and 292XSR (grind # 4876) for the exhaust. These profiles are 248 and 254 for the intake and exhaust respectively. With Comps 1.65 rocker (#1006) on the intake and 1.5 (#1004) on the exhaust net valve lift after lash was 0.620 and 0.588 respectively. The results achieved were as shown below.
Again we see that a two plane works well low down but this 408 is stretching a two plane intake, even one as good as the Edelbrock Performer Air Gap, to the limit. Note that because the engine wanted to produce it’s torque peak up around the 5200-5300 mark it was already taxing the flow capabilities of both the carb and the two plane intake. The result is that peak torque with the two plane suffered somewhat. I am sure that some of this could be made up for by porting the intake and using as much carb CFM as can be had from a 4150 based carb. Maybe that’s a project for another time. The bottom line is that given an intake that could breath these 195 AFR heads allowed our 408 to produce almost 600 hp. Have you any idea how frustrating it is to be just 1% off a key output number like 600! Still 595 is an impressive number to achieve with what is essentially a street drivable small block that still is not costing an arm and a leg. As for torque this 195 Eliminator equipped 408 seemd to really like the port and plenum matched Holley Strip Domionator and the AED 950 carb used. A specific torque output of 1.32 lbs-ft per cube from a 10.5/1 pump gas engine is good by any standard.
For more information on these heads check out AFR site at
http://www.gofastnews.com/board/tech...owresearch.com