A diesel does not make power by fuel alone. It makes power when the air motion in the cylinder is right, the fuel stays in the bowl, and the camshaft events support that process instead of fighting it. That is why we start with swirl and squish. Swirl keeps the charge moving, squish drives dense air into the combustion chamber near top dead center, and together they help the fuel mix with oxygen instead of turning into soot. This is also why slick-top pistons matter. Valve reliefs may buy clearance for a more aggressive cam, but they interrupt swirl, weaken squish, and create pockets where combustion quality falls off. The result is more soot, higher exhaust temperature, and a greater chance of piston and cylinder-wall damage when injection runs too long. Our rule of thumb is to keep injection duration under control and keep the event centered around top dead center, because once fuel breaks out of the bowl, parts start paying for it. From there, cam design becomes a balancing act between cylinder fill, blowdown, overlap, valve cooling time, and valve-to-piston clearance. On the L5P, the head’s stronger early intake flow and much higher swirl give us a better foundation than earlier Duramax heads. That lets us push for more high-rpm airflow without giving up the combustion quality that keeps the engine alive. The goal is simple: make more power without soot, without runaway EGT, and without hurting the stock parts just to chase a number.
With the new camshaft installed in the Duramax, Gale Banks explains that the engine is nearly ready for its first startup. The project previously made 912 horsepower with a stock cam, and the latest phase has focused on developing a new camshaft with Comp Cams over roughly six weeks. The target is to gain another 100 horsepower and push the engine past the 1,000-horsepower mark.
Banks frames the entire effort around maintaining proper diesel combustion behavior rather than simply adding lift and duration. His central point is that power in this engine depends on preserving squish and swirl, then using cam timing to improve cylinder filling and exhaust blowdown without creating soot, excessive exhaust gas temperature, or piston damage.
Banks begins with the intake process and the role of swirl in a diesel. As the piston moves down on the intake stroke, air entering through the intake valves forms a high-speed spiral in the cylinder. In the Duramax, that swirl is essential because it helps mix injected fuel with air far more effectively than a simple straight-flow arrangement.
Comparing the L5P to earlier Duramax heads, he notes that the intake and exhaust valve sizes remain the same, but the L5P head uses a redesigned intake port. Earlier engines brought the intake ports all the way to the head surface, while the L5P shortens them and reshapes them to improve swirl substantially, with a smaller improvement in intake flow. These are not gasoline-style ports intended only to maximize raw airflow. They are shaped to twist the incoming air, which improves combustion efficiency. Banks warns that many diesel head-porting jobs increase flow while destroying swirl, and that this tradeoff is disastrous for combustion quality.
To show why swirl matters, Banks describes the combustion chamber from above: injector centered over a bowl-in-piston chamber, with a slick-top diesel piston that has no valve reliefs. Without swirl, fuel injected near top dead center does not mix properly with the air. Instead, it forms soot, which he describes as unburned fuel heated in the absence of oxygen. That soot exits the engine as black smoke and represents fuel that did not make power.
The problem gets worse if injection continues too long. As the piston moves down after top dead center, the spray can break out of the combustion chamber bowl and strike the piston crown and cylinder wall. That creates severe hot spots, piston damage, and fuel impingement on the cylinder wall. Banks shows this as the mechanism behind damaged pistons and what he calls metal-sprayed cylinder walls.
With proper swirl and squish, the process changes. As the piston approaches top dead center, the piston-to-head clearance tightens to a little less than 0.030 inch. That squeezes the rotating air from the perimeter into the combustion chamber bowl. Banks calls this squish, and says it also creates tumble. The result is that the high-density air previously stacked against the cylinder wall is forced into the bowl where the fuel is being injected. That produces a much more uniform air-fuel mixture, with oxygen available where it is needed. Because diesel fuel does not vaporize and homogenize as easily as gasoline, this air motion is critical.
Banks then explains why aggressive racing pistons with valve reliefs can hurt a diesel. Valve reliefs interrupt the rotating air as the piston rises, acting like a brake on swirl and reducing swirl rpm. When injection begins, the disrupted air motion weakens squish into the bowl and leaves pockets where additional air never reaches the fuel. Those become soot pockets, and the overall combustion process deteriorates.
His solution, when valve reliefs are unavoidable, is more air. He cites the nominal stoichiometric diesel air-fuel ratio as 14.5:1, but says a live engine needs 15 to 25 percent excess air because the mixture is never perfectly homogeneous. That is why he prefers a larger turbocharger compressor than normal. More air density helps compensate for compromised in-cylinder motion. He notes that he has used this approach successfully in drag racing and at Bonneville. Nitrous oxide can also cure many mixture-distribution problems because it spreads oxygen throughout the cylinder, but it is impractical for long-duration wide-open-throttle use such as Bonneville. In the Killin' A Duramax project, Banks is using slick-top pistons and relatively modest valve lift, then relying on heavy turbocharging to make the combination work. In his view, the engine has now run out of camshaft rather than airflow.
Banks next turns to valve timing and cylinder filling. He reviews the four-stroke cycle and focuses on overlap, the brief period near top dead center between the exhaust and intake strokes when both valves can be open. On the stock-cam, stock-turbo example, he says the engine made 568 horsepower at 3,000 rpm with intake manifold pressure at 30.9 PSIG and exhaust manifold pressure at 62.4 PSIG.
He argues that comparing those pressures directly is not enough, so Banks Instrument developed what he calls engine scavenge ratio: intake manifold pressure divided by exhaust manifold pressure, minus one, times 100. In this case the result is negative 52.5 percent, meaning intake pressure is 52.5 percent lower than exhaust pressure. Under those conditions, opening the intake valve too early would encourage exhaust backflow into the intake port rather than improve cylinder filling. That is why the stock intake valve does not begin opening until 3 degrees before top dead center at 0.001 inch valve lift.
He also emphasizes that dirty tuning creates additional problems. If soot is present, some of it gets past the rings and contaminates the oil, degrading the oil package and lubrication quality for heavily loaded components. He says black oil is a warning sign of this problem. Dirty tunes may feel stronger, but they increase soot, backpressure, DPF loading, and wear throughout the engine.
Banks walks through the stock cam in detail. On the intake side, the lash ramp begins at 34 degrees before top dead center, but actual valve movement at 0.001 inch starts at 3 degrees before top dead center. At 0.050 inch tappet lift, the intake event is measured for comparison purposes, and maximum lift centerline occurs at 100 degrees after top dead center. He notes that the piston is already well down the bore by then and moving very quickly. Because the project is retaining stock valve springs, additional lift is limited by coil bind, so the camshaft must do its work through timing and lobe design rather than brute lift.
The intake valve reaches 0.050 inch on closing near bottom dead center and is fully closed at 30 degrees after bottom dead center. The opening lash ramp is 31 degrees and the closing lash ramp is 36 degrees. Total duration from first valve movement to final closing is 213 degrees. At 0.050 inch tappet lift, the intake duration is 155 degrees. Banks also tracks valve-seat duration carefully, measured from 0.001 inch opening to 0.001 inch closing, because that is when the valve transfers heat into the seat and water jacket.
On the exhaust side, the lash ramp begins 76 degrees before bottom dead center and actual valve movement starts 41 degrees before bottom dead center. This is the beginning of blowdown, where cylinder pressure is reduced before the piston starts pushing exhaust out on the exhaust stroke. Banks says the goal is to reduce cylinder pressure to exhaust-manifold pressure by bottom dead center so the piston rises against minimum resistance. Otherwise, crankshaft power is wasted driving the exhaust stroke.
The stock exhaust reaches maximum lift centerline at 105 degrees before top dead center, then closes to 0.050 inch tappet lift at 22 degrees before top dead center. At top dead center, the exhaust valve is only 0.014 inch open, which helps prevent backflow into the intake. It finally closes at 23 degrees after top dead center, and the lash ramp ends at 57 degrees after top dead center. Exhaust opening and closing ramps are 35 and 34 degrees respectively. Exhaust duration is 244 degrees at 0.001 inch tappet lift and 166 degrees at 0.050 inch. The exhaust valve seat duration is 476 degrees, and Banks considers that cooling time critical because the exhaust valve runs much hotter than the intake valve.
A major part of the new cam program is Comp Cams' LST, or Low Shock Technology. Banks says Billy at Comp focuses heavily on the first 0.006 inch of tappet lift and on how the lobe shape controls what happens after the valve starts moving. The objective is to reduce shock loading on the valve springs while still increasing effective camshaft aggressiveness.
Using Spintron footage, Banks compares a conventional profile to an LST profile on the same engine. Without LST, the rocker opens the valve and the spring shows substantial surge both during opening and after seating. The retainer movement also reveals valve rotation. With LST, the spring remains much more stable, with far less surge during the event. Banks says this stability allows more camshaft and higher rpm while still using stock valve springs. That is a key enabler for this project, since the goal is to push the stock valvetrain farther without immediately resorting to aftermarket springs and other hardware.
Banks then ties cam timing back to injection timing. His rule of thumb is to keep injection within a 40-degree crank-angle window centered roughly around top dead center, ideally within 20 degrees before and 20 degrees after. He wants peak cylinder pressure to occur about 8 to 12 degrees after top dead center.
At an earlier 711-horsepower stage, start of injection was 18 degrees before top dead center and end of injection was 8 degrees after top dead center, for a total of 28 degrees of injector-open time. At top dead center, piston-to-head clearance was about 0.028 inch, while by 8 degrees after top dead center the piston had moved to about 0.180 inch from the head. Banks says that once injection duration stretches beyond 40 degrees, piston damage and general combustion breakdown become likely. He has experimented into the mid-30-degree range and even around 38 degrees, but says things were already going wrong there. His goal at 1,000 horsepower is to keep injection duration well controlled rather than simply extending it to make power.
He also stresses how fast all of this happens. At 3,800 rpm, the combustion event is occurring 33 times per second, and the useful injection window is only a tiny fraction of the full 720-degree four-stroke cycle.
To support the cam design, Banks uses flow-bench testing at 28 inches of water pressure differential. The cylinder head is mounted on a bore matching the engine, and airflow is measured through the intake and exhaust ports as valve lift is increased in steps. Although the bench data extends to 0.600 inch lift, the relevant range for this engine is around 0.375 inch maximum valve lift.
Comparing an older LMM-style head to the L5P, the intake side shows only modest gains in total flow, but the L5P is stronger early in the lift curve. Banks cites roughly 4 percent more flow at 0.050 inch, about 16 percent at 0.100 inch, and about 5.5 percent at 0.250 inch. He says this early-flow advantage is especially important because it accelerates swirl development. The older head peaks around 0.300 inch lift and then degrades slightly by 0.375 inch, while the L5P reaches a higher peak and then plateaus rather than falling off.
The exhaust side shows a much larger L5P advantage. Early blowdown flow improves immediately, and unlike the older head, the L5P exhaust port keeps climbing at higher lifts. By 0.375 inch lift, Banks says the L5P has about 20 percent more exhaust flow. He believes there is room to increase exhaust valve diameter from 31 mm to 33 mm while still clearing the cylinder wall, and expects that would improve flow from roughly 0.150 inch lift upward, but he does not plan to make that change unless the project reaches a point where the heads must come off anyway.
Swirl testing confirms the L5P advantage. Using an anemometer at bottom dead center during intake flow testing, Banks measures swirl rpm versus valve lift. At 0.250 inch lift, the L5P reaches about 3,000 rpm of swirl, while the older LMM head is only around 1,200 rpm. That large increase in swirl is one of the reasons the L5P head supports higher fueling without losing combustion quality.
Banks closes by showing how the new cam is intended to reshape airflow relative to crank angle. Traditional cam profile plots show lift versus crank position, but he overlays those with actual airflow traces from the flow bench to show what the engine really sees. On the stock cam, the LMM intake trace flattens at the top because the head stops gaining flow, and the exhaust trace is especially disappointing. The L5P improves both, particularly on the exhaust side, where the airflow curve becomes much more favorable.
The Banks cam is designed to exploit that better head. On the intake side, he wants the valve to open sooner, increase area under the curve, add some overlap, reach maximum practical lift, and hold more flow on the closing side to improve cylinder filling. At the same time, he wants to close the intake as near as possible to the stock closing point so the engine retains good heat of compression and cold-start behavior.
On the exhaust side, he wants to begin blowdown earlier because the engine speed is increasing from about 2,800 to 3,800 rpm, leaving less time for pressure release. That earlier opening may cost some torque, but this is a high-rpm diesel combination. He also wants more lift through the top of the exhaust event and a later closing point, limited by valve-to-piston clearance during overlap.
Banks studies those clearances directly. The new cam greatly enlarges the overlap triangle compared with the stock cam, but still stays within the available piston-to-valve clearance. The exhaust side has more lash than the intake, 0.016 inch versus 0.012 inch, because the exhaust valve grows more with heat and closes up lash in operation. He expects more exhaust-valve growth at 1,000 horsepower than in the stock 445-horsepower engine, so that extra clearance is intentional.
With the stock cam, overlap at 0.001 inch lift is 26 degrees, while overlap at 0.050 inch tappet lift is negative 45 degrees, meaning both valves are effectively closed through that interval. The new cam opens the intake sooner, closes it later, opens the exhaust earlier, and likely closes the exhaust near its current point. Banks describes it as the most camshaft he can fit into this engine while preserving the stock hardware strategy. The next step is to fire the engine and see whether that combination is enough to reach 1,000 horsepower.