Choosing Cylinder Heads for More Power | Project LS 350Z

Four years ago, as we were sitting in engine building class assembling the LS for our LS-swapped 350Z project, we weren’t exactly sure what we were going to do with the aluminum V8. We were sure of one thing, though: We knew we only wanted to build it once. That’s why we chose lightweight JE Pistons, as well as a forged crankshaft and connecting rods from K1 Technologies. If we had to replace our stock crank anyway, we figured why not make it a good one, right?

After assembling that strong (and expensive) bottom end, we mostly closed the aftermarket parts catalog and finished our build with a fairly mild cam and a set of 243-code OEM heads. Why? Because we knew we could always upgrade our heads and cam later, and we were curious how well our budget option would work. Plus, we couldn’t really make good decisions and match our heads and cam to our application without knowing our application. Remember, in order to sign up for engine building school, we’d broken the cardinal rule of engine building, and built an engine without having a plan for what we’d install it in, or how we’d be using it.

In the years since, our LS1’s future has become much more clear. We finished it, swapped it into a 350Z at  LOJ Conversions, and were now adding safety gear to go road racing. Our LS’s future was on track, which had us thinking about those compromises we’d made when building it.

And it wasn’t just that we felt bad hamstringing that shiny rotating assembly: We had classing to consider, too. We wanted to compete in NASA’s power-to-weight based TT2 competition, which meant we were shooting for an 8:1 maximum average horsepower to weight ratio, as measured on a Dynojet chassis dyno. With a theoretical competition weight of 3150 pounds, that meant we were limited to 393 maximum average horsepower.

What’s maximum average horsepower? This is a simplification, but basically it’s the average of the highest horsepower readings within a 4000 rpm range centered on the peak horsepower RPM and measured in 250 RPM increments. The number of points averaged changes based on engine redline, with higher-redline engines including more points (and thus more of the power curve) in the average. The highest bracket, engines with redlines higher than 7000 rpm, averages the eight highest points, including the maximum horsepower number. The lowest bracket, engines with a redline below 6000 rpm, only averages four.

What’s this mean in English? Basically, the higher our redline, the more points on the horsepower curve that are averaged against our peak power number. We believe the intent is to give peaky, high-revving engines (Hello, VTECH Hondas) a fighting chance against torquey V8s like ours. But we can use this rule to gain a slight advantage thanks to our lightweight rotating assembly. We’ll set our redline above 7000 rpm in order to produce the broadest (and therefore the lowest) maximum average horsepower calculation.

We’d dyno tested the 350Z on Very Cool Parts’ Mustang dyno shortly after finishing it, and only managed 318 peak horsepower to the wheels, with a redline of 6500 rpm. Even on a higher-reading Dynojet dyno (NASA’s required brand for calculating power-to-weight ratios), odds are we wouldn’t have even made enough peak power for the limit of our class, nevermind maximum average power.

And we weren’t merely trying to meet that power limit. In addition to raising our redline to have a more-favorable formula applied, we can also use our electronic throttle body to get an edge on the competition. Our goal is to exceed the horsepower limit, then tune the car with HP Tuners to close its throttle body when necessary to produce an artificially low peak horsepower number, and a flatter horsepower curve. The side effect should be faster laptimes. Take this approach a step further, and we’ll be able to add or remove weight and power to tailor the car to different tracks.

All this scheming pointed to our most pressing issue: We needed more power. And we knew from our dyno chart, and its huge falloff at high RPM, that we needed to get more air flowing through our V8 to make that power. That meant starting with the low hanging fruit on our LS1: Those 243 heads and that mild camshaft. And while we could have just swapped cams and picked up power, talking to our experts made it clear that we’d need a better set of heads to make the power we were hoping for. We already had some of the best-flowing factory heads available, which meant it was time to enter the world of CNC ported heads.

What’s CNC porting? Simple in theory: Take a cylinder head, clamp it down in a CNC machine, and run a computerized milling operation to remove material and make it flow better. In practice, though, CNC porting produces a broad range of quality, with each shop having their own secret recipe that claims to be “the best.” Some make tons of power. Others might not. With CNC machinery getting cheaper every year, it’s never been easier to buy a piece of equipment and open your own porting operation. Our advice is to look at an outfit’s reputation—not just their price—when shopping.

Besides the program run in the machine, there’s another huge factor in the price and quality of CNC-ported heads: The base material that’s being ported. You’ll find three types of CNC heads for sale: Ported OEM heads, ported aftermarket cast heads, and billet heads. What’s the difference? Here’s a basic rundown:

  • Ported OEM Heads: The builder starts with a new or used OEM casting, often one the customer supplies, then puts it in the porting machine and removes material to improve flow. Because so much of the work—making the head itself—is done before the shop ever touches it, these are often the least expensive way to get a great-flowing head. There are downsides, though: The shop is at the mercy of the parts supply, which usually means you’ll need to find a good set of core heads to send in. Brand new castings, if they’re available, will usually cost so much that it’s more cost-effective to jump right to ported aftermarket heads anyway. And, because the porting must be based on what the OEM has already cast, the builder has to work around those original compromises or weaknesses. Plan on spending around $1000 to have your set of OEM heads ported, plus the cost of a fresh valve job while you’re in there.
  • Ported Aftermarket Cast Heads: These are the most common type of aftermarket heads. Rather than start with an OEM casting, these are cast by the builder, then they undergo the CNC porting process. Because the builder doesn’t have to work within the OEM design constraints, they can often better optimize the porting program or even make drastic changes to things like combustion chamber design, compression ratio or valve angle. They can also add other niceties that wouldn’t make sense for mass production, like thicker decks to prevent warping and add strength, and extra clearance for pushrods. Assembled CNC-ported aftermarket cast heads can be found for a little over $2000. Several manufacturers offer aftermarket cast heads for sale without any additional CNC machining as a budget option, too.
  • Billet Heads: It’s rare to not be able to accomplish your goals with a cast aluminum head, but for the highest horsepower automotive engines in the world nothing else will do. These are made from one or many pieces of solid billet aluminum, with material removed until all that’s left is a cylinder head. With prices ranging from the cost of a new LS3 crate motor to the cost of a whole Corvette, these heads aren’t really a part of our world, but they’re still cool to look at. What do billet heads get you that cast heads can’t? Mostly strength, which starts to matter if you’re making 10,000 supercharged nitrous-fueled horsepower at the drag strip. Otherwise, though, we don’t think billet heads have a role in amateur road racing.

With the choices laid out, we zeroed in on our path forward: Ported aftermarket cast heads seemed like they’d have the most potential to make power and be the most durable, and it was just gravy that they also looked like the best deal by the time we factored in the cost of cores, valves, and shipping. Of course, we are still curious about budget-oriented ported OEM castings; so we’ve got a separate story planned on that.

Choice made, we needed to find some heads. And we’ll be perfectly frank: We’re not LS engine experts. We don’t have every head option and combination memorized, and only have a surface-level understanding of where we’d left power on the table. Fortunately we have a telephone, though, so we used it to call up people who are LS experts: Air Flow Research, who have been building aluminum racing heads for nearly five decades. AFR’s name might be more familiar in the drag racing world than it is in ours, but one big force in the company is Tim Torrecarion, a GRM reader and our spirit guide through our LS head selection journey.

Before we went any further on the call, we started with a caveat: We didn’t want any special treatment, meaning AFR shouldn’t give us any advice that they weren’t willing to give everybody who picked up the phone and called them. Surprisingly, they said that wasn’t a problem; in fact they actually encourage it, and have a dedicated tech line staffed by experts who can help racers like us make more power. Their knowledge doesn’t end at heads, either, as we found AFR advising us on camshaft selection, gearing, tuning, and even other companies’ heads and how they might compare to AFR’s options. The price for this? Absolutely free, though obviously the goal is for you to end up with a set of heads in your shopping cart.

After a few hours of discussion over several calls, we ended up with a set of AFR’s 215cc mongoose cylinder heads, with 215cc referring to the intake runner volume. AFR describes them like this:

AFR's 215cc mongoose cylinder heads are recommended for moderate street or street/strip builds with displacements up to 408 cubic inches operating at or below 6500 RPM. This medium sized, high velocity torque monster flows 312 CFM @0.600" of lift.”

No, this isn’t the highest-flowing head available (AFR makes LS1 heads with runners as large as 245cc). We chose these heads in order to get the top-end power increase we were after without decimating the broad powerband we knew would help lower our lap times. It’s important to size heads to your goals and displacement, and AFR felt that this was the best choice for our “small” 5.7-liter LS. We checked the box for a $76 high-RPM valve spring upgrade, too, so we can raise our redline and take full advantage of that bulletproof rotating assembly (and the favorable break in the rulebook) on track. Total price? $2537. That’s not cheap, but it’s a great deal for five-axis CNC-ported aftermarket heads that came fully assembled and ready to bolt onto the car. Those playing along at home should remember that unlike having a set of OEM heads ported, this option means we got to keep our original heads for the next project (or an eBay listing) to further offset the cost of the AFR parts.

These new heads would gain us a lot of airflow, making the new weak link the street-oriented camshaft we’d originally installed when we built the engine. We’ll cover camshaft selection—as well as upgrades elsewhere in the valvetrain—in our next update.

Like what you're reading? We rely on your financial support. For as little as $3, you can support Grassroots Motorsports by becoming a Patron today. 

Become a Patron!

Join Free Join our community to easily find more project updates.
View comments on the GRM forums
Colin Wood
Colin Wood Associate Editor
1/28/21 8:40 a.m.

It's cool that you can actually see the scoring from the CNC machine.

nocones UberDork
1/28/21 8:48 a.m.

This car vs your project Z06 (Or a more stock C5 Z06) would be a very interesting article.  Compare Total Build Costs, Driving feel, Operating costs during track use (Rate of consumption for Tires/Brakes), Laptimes (With Data analysis to see where each chassis does better), Track Durability (Can you lap it all day or does either need breaks to cool off), Hours invested in the builds.

It's just 3150 lbs/390 HP, 2 seat sports car seems like a comparison begging to be made.  Also Modified vs Buy it upfront.

And yes I'm well aware that projects like this are about way more then the pace of the end result.  As someone with thousands of hours and about $9000 into a car that is about as fast as a Well driven Spec Miata at a track.  It's still a compelling comparison to make.

84FSP UltraDork
1/28/21 9:10 a.m.

Nice head choice that focus on velocity vs just max cfm's.  You will be right in the range where you are now fully restricted by your intake and tbody.  There are really nice options out there form the OEM catalog.  BlackbirdTA on LS1Tech is a pro head builder for Nascar teams.  He did an exhaustive flow best test across available intakes and heads.  Fantastic info here.  

LS intake flow testing

350z247 New Reader
1/29/21 2:26 p.m.

In reply to nocones :

Agreed, this would be a very interesting comparison. I'd guess a stock C5 is a much better choice than an LS swapped 350Z, but there are SO many variables, the biggest probably being how much work you can do yourself.

Sponsored by

American Powertrain

Precision Transmission Center

LOJ Conversions

Our Preferred Partners