Recently, we posted an article discussing bore vs. stroke, and the advantages of each. That article used largely hypothetical figures and measurements as examples. Now, on a very similar topic, we have a video from the dyno-test king himself, Richard Holdener. In this video, Holdener actually dyno tests a destroked LS3 to gather real data.
For the video, Holdener starts with a “destroked LS3” short-block, which consists of an LS3 block, with an OEM 4.8-liter crankshaft (3.268-inch stroke), along with 6.300-inch Lunati H-beam rods, and a set of custom 4.065-inch slugs from JE Pistons. For the first round of tests, Holdener uses a set of Air Flow Research 230cc LS1 cylinder heads atop the 339 cubic-inch short-block.
The camshaft was from Brian Tooley Racing with .627-inch of gross lift, intake, and .596-inch exhaust and a 243-degree intake, 262-degree exhaust duration with a 110-degree LSA, and a Holley hi-ram intake with dual 750 cfm UltraHP carburetors was used. Once dialed in, the baseline combination made 607 horsepower at 7,800rpm, and 466 lb-ft of torque at 6,200rpm
The combination makes for a significant power and RPM gain over a stock LS3 run in a similar fashion. The combo is good for over 100 peak horsepower coming on 2,000rpm later while losing significant power below 5,800rpm to the stock LS3.
Switching out camshafts on the destroked combo to a Comp Cams grind with .637 inch of lift and 253 degrees of duration, intake and 269 degrees, exhaust with a 118-degree lobe-separation angle, the combo made 633 horsepower at 8,000rpm, while trading-off power below 7,300rpm.
Next, Holdener swaps out the cathedral-port heads and intake for a set of Trick Flow 255cc CNC rectangular-port LS3 heads with a Holley rectangular-port Hi-Ram intake. However, that change came along with a 10cc drop in combustion chamber volume, giving up some compression and ended up costing a significant amount of power, and not really providing a fair cathedral-vs.-rectangle-port comparison.
One thing Holdener points out in the video — besides the apples-to-oranges comparison of cylinder heads — is that the reason the engine was able to spin to 8,000rpm isn’t because of the shorter stroke offered by the 4.8L crankshaft. “This engine would make more power if I put the 3.622-inch crankshaft in it,” says Holdener.
He points out that even at 8,000rpm, the engine isn’t near the limits for that stroke. When running the numbers on this combination, we come up with a mean piston speed of 4,357 feet per minute (or 22ish meters per second, if you want to relate to the Engineering Explained video). That is significantly below the 5,000fpm we often hear referenced as a practical limit for mere mortals using common, modern rods and pistons.
Holdener points to the valvetrain as the limiting factor on the average LS engine’s ability to turn RPM. “If you get the cam right, with the right lobe design, and get the valvespring right — not just lots of pressure — you’ll be able to run the RPM,” Holdener says. With advances in modern components, piston speed isn’t the limiting factor it once was for engines of this caliber and configuration.
To verify his point that you could take the same engine, and turn the same RPM with a 5.3L or 6.0L (3.622-inch) even a 7.0L (4.000-inch) stroke, we decided to run the calculations for mean piston speed. Using the formula of “2 x stroke x RPM / 12”, we get a mean piston speed of 4,829fpm for the 3.622-inch stroke and 5,333fpm for the 4.0-inch stroke. While not impossible by any means, we’d probably be looking at more robust components for the 4.0-inch stroke at 8,000rpm.
Make sure to watch to the end of the video for Holdener’s thoughts on bore vs. stroke and displacement in general. Really, what it comes down to in the real world, is if you are a fan of RPM, modern aftermarket components are capable of handling the piston speeds and associated stresses for most small-block combinations.
So, as long as you get your valvetrain right, you can spin an engine to speeds once considered unattainable. While getting the valvetrain right isn’t always the simplest of tasks, it’s far from the herculean task it was in decades past.