How to Build LS Gen IV Performance on the Dyno. Richard Holdener

       The test engine was an LS3 crate engine supplied by Gandrud Chevrolet. It was run in stock trim with no changes to the cam, compression, or induction system.

       For dyno use, I installed a set of long-tube headers, FAST throttle body, and XFI management system. The crate engine produced 493 hp and 484 ft-lbs of torque with the stock heads.

      Stock LS3 vs Chevy Performance CNC L92 on a Stock LS3 (Horsepower)

       Stock LS3 Heads: 493 hp @ 5,700 rpm

       Chevy Performance CNC L92 Heads: 503 hp @ 5,900 rpm

       Largest Gain: 13 hp @ 6,400 rpm

       Replacing the stock LS3 heads with a set of CNC-ported L92 from Chevy Performance netted only minor gains on this otherwise stock LS3 crate engine. The CNC-ported heads flowed considerably more than the stocks, but the mild stock engine simply didn’t need any more flow to support the existing power level. Test 3 illustrates the amount of power that the extra airflow can be worth on the right application.

      Stock LS3 vs Chevy Performance CNC L92 on a Stock LS3 (Torque)

       Stock LS3 Heads: 484 ft-lbs @ 4,800 rpm

       Chevy Performance CNC L92 Heads: 497 ft-lbs @ 4,600 rpm

       Largest Gain: 16 ft-lbs @ 4,400 rpm

       The torque curve tells the same story because the mild LS3 simply couldn’t use the additional airflow. Because the stock LS3 heads support well over 600 hp, they were more than enough for this stock crate engine. The head swap would be worth much more than 16 ft-lbs on the right application.

Test 2: Stock LS3 vs AFR 245 on a 408 Stroker

      Since both the AFR 245 cathedral-port and factory LS3 heads used for this test flowed so well, I made sure I had a solid test engine for the comparison. Starting with a 6.0 block, the 408 was assembled using components from Speedmaster and Wiseco. The stroker assembly included a 4.0-inch forged-steel crank combined with 6.125-inch 4340 forged steel connecting rods and 10-cc dished pistons.

      The forged pistons from Wiseco featured valve reliefs to allow for the hydraulic roller (PN 289LRR HR14) cam (.624 lift, a 239/255 duration split, and 114 LSA) from Comp Cams. Designed for a rectangular-port head application (which may not have favored the cathedral-port heads), the cam was combined with a set of standard-travel lifters and hardened pushrods from Comp Cams.

      The .030-over 408 stroker short-block also featured a new timing chain and oil pump from Speed Pro, a set of ARP head studs and MLS head gaskets from Fel Pro, and a Moroso oil pan and windage tray.

      Comparing the rectangular-port LS3 and the cathedral-port AFR heads also required an intake swap (to match the respective head ports). To keep the test as even as possible, I selected FAST LSXR intakes for both applications. Both were also run with the same 102-mm Big Mouth throttle body. As always, both heads were run with the same air/fuel ratio and timing values.

      Equipped with the stock LS3 heads, the 408 produced 581 hp and 543 ft-lbs of torque. After installation of the AFR 245 heads, the peak numbers jumped to 604 hp and 665 ft-lbs of torque. It must be pointed out that the head swap also included a change in static compression ratio because the chamber volume on the two heads differed by 5 cc (64 cc vs 69 cc). This meant that in addition to the increased airflow offered by the AFR heads (349 cfm vs 316 cfm), they also increased the static compression ratio by .5 points. With the exception of a short, 250-rpm range (from 4,100 to 4,250 rpm), the AFR heads improved the power output from 3,000 to 6,700 rpm. Having more peak power is good, but having extra power everywhere is even better.