The ASE billet intake manifold line took roughly two to three years to reach production. It did not start from a blank screen, and it was not designed by guesswork on a dyno.
This article documents the development in three stages — initial design, flow bench R&D, and final production — matching the three videos filmed along the way. If you want to understand what you are buying and why it is shaped the way it is, this is the whole story.
Part 1 — Where the design came from
This is not a ground-up product. It is the evolution of the ASE DIY manifold kit — a plenum base, a set of billet runners, and an intake flange — which had been in the field for around five years and running on serious race programs for three of them.
What changed was capability. Bringing a five-axis machine in-house made a one-piece billet manifold possible, so the question became whether the proven kit could be taken further as a single unit.
The input came from outside as much as inside: racers consistently making 1,200 to 1,300 horsepower on front-wheel-drive and all-wheel-drive combinations, the tuners actually calibrating those cars, and specialists in airflow, cylinder head flow, and cam design reached through twenty years of industry contacts. Plenty of design principles had been absorbed over years of tuning and fabrication without a full understanding of why they worked. That gap is what the outside input closed.
What changed from the DIY kit
- Plenum port entry went from oval to round, maximizing airflow into the runner.
- The oval-to-round transition was stretched over a longer distance. On the DIY kit that transition happened abruptly over a short span. Spreading it out helps air speed and velocity, which keeps the charge off the runner walls — and air that is not scrubbing the walls runs cooler.
- The internal surface finish is deliberately not polished. Rough enough to aid air speed and reduce intake air temperature, smooth enough not to restrict. Most people assume hyper-smooth is the goal. It is not, and crossing that line costs power, reliability, and IAT.
The first machined article was not about performance at all. It existed to answer manufacturing questions — how to program the part, how to fixture it, how to make it repeatable, what tooling changes were needed, how much harmonic vibration showed up, and what surface finish came off the tool. Software will not tell you that. You have to cut one.
Part 2 — Proving it on the flow bench
Rather than machine test articles out of billet at hundreds of dollars apiece, the candidate runner designs were 3D printed and taken to the SuperFlow bench at EIC Motorsports in Lakeland, with Gabby — an engineer who designs cylinder heads bespoke per application.
The test was deliberately narrow. Four runners, identical in length, shape, and cross-sectional area. The only variables were the inlet radiuses and the entry shape. Fixing everything else is what made the comparison mean anything.
What the bench showed
- Roughly 20 CFM separated best from worst — from radius and entry shape alone, with every other dimension held constant.
- Runners do not flow the same in both directions. Common assumption, and the bench disproved it. Flowed from the plenum side versus the cylinder head side gave consistently different numbers, and the spread was larger still when flowed the way the runner actually sees air in service.
- Smaller radiuses make the runner look longer to the engine than it physically is — a lever for shifting where the engine wants to make power.
- None of the four designs was a restriction. Every variant kept up with the best-flowing race heads available for these engines. That was the pass/fail question, and all four passed before refinement even started.
The framing that stuck: the engine sees the runner as an extension of the cylinder head. The transition should be seamless, because functionally there is no boundary there.
Why it was designed around naturally aspirated numbers
In all-motor classes the top five qualifiers are often separated by hundredths of a second, running shared tuners and shared builders with access to the same equipment. Finding two or three horsepower in five different places is what moves a car from fifth to first.
A good naturally aspirated design carries over to a boosted application. The reverse does not hold — boost can paper over a mediocre manifold, but it will not fix one.
Part 3 — What made it into production
Curved runners
The single biggest departure from the DIY kit. Straight runners on a front-wheel-drive car, particularly with spacers, tend to put the plenum into the underside of the hood unless the engine is leaned over. Curving them solves hood clearance, eases charge piping routing, and clears the transmission bracket. It was not attempted on the older equipment — possible, but the cost would have been prohibitive.
Sealing
The gasket mounting surface is wider and thicker than the original design, with the bolt holes brought inboard. Sixteen bolts hold the plenum down. The problem being solved is specific: not everyone re-torques their car between rounds, and finding a boost leak in third-round eliminations because two bolts backed out is a bad afternoon.
Bolted, not welded
Some builders weld their multi-piece manifolds together. It works until an engine lets go — then debris is trapped inside something you cannot open. The one-piece plenum bolts down so it can be separated and cleaned.
Injector angle
The injection angle stays close to Honda's original rather than being reoriented for machining convenience. That preserves atomization at low air speed, which matters at idle and cruise and makes fuel trims far easier to dial in. Roughly 90% of these end up on street cars, and that shaped the decision.
Plenum
The back plate is bowed rather than flat, increasing plenum volume for the higher-power combinations while settling on a volume that works across a wide range of builds.
Fixed dimensions
Runner length carried straight over from the DIY kits — it was already proven, so there was no reason to relitigate it. Stack heights were tested at 10, 15, 20, and 25 mm. Floor finishes and both compound and single floor radiuses were evaluated on printed articles before anything was committed to billet.
Throttle body
A standardized 90 mm Ford-style flange, so AEB, K-Tuned, Skunk2, and anything else on that bolt pattern will mount.
Machining
Seven to eight hours in the machine per manifold. Producing it as multiple pieces would have been considerably easier — reaching down past a large plenum base to cut injector features means a long tool on a tilted five-axis setup. The concession was made on machining time rather than on the design.
A note on the numbers
The CFM figures discussed in the flow bench video are readings from 3D printed test articles taken during development, not published specifications for the finished manifolds. They are shown to illustrate how much difference small geometry changes make, not as performance claims.
The manifolds
Available for B Series, K Series, and H Series applications, including rear-wheel-drive specific configurations.
Still prefer to build your own? The DIY manifold kits that started all of this remain available for B, H, and F/S2000 applications.