Today we got a chance to dyno our first 2012 Si. A coupe with Navi, the K24Z equipped Civic performed pretty much as expected. Below, we've posted dyno plots comparing this press fleet car with our original Project Si car as it was delivered back in late 2005.
Click to enlarge
Note that the new Si produces slightly more torque below 3000 rpm. From 3000 to 6000 rpm there is a substantial increase in torque over the old K20, as you would expect with an additional 18% displacement. However, from 6000 rpm onward, where the K20 is in VTEC, the curves are virtually identical with the K20 actually producing slightly more power at peak.
VTEC on the new Si was at approximately 5000-5200 rpm. The new car also ran extremely rich above 6000 rpm. It was probably running in cat protection mode. The old Si's would do this as well, but only after extended running. Every single run from start to finish on the new Si saw it run rich at high rpms, probably a side effect of having to deal with the close coupled catalyst and the higher exhaust gas temps (EGTs)it will see so close to the exhaust ports.
Another interesting note was that the 2012 car showed extreme variability on the dyno. Some runs produced substantially more low end torque (10-12 lbs-ft), but would fall short on the high end. Some would produce less torque below VTEC, but then produce a few more hp on the high end. Datalogging showed the car running very conservative ignition timing (on the order of 12-14 degrees where we would normally run 20-24 deg on a K20). This data indicates, to me, that there is some potential for filling out the torque curve and improving consistency. With no other changes, provided one is willing to risk long term cat problems, tuning should allow this engine to match the K20 from 6000 to 7500 rpm or so, produce a little more torque below 3000 rpm than it already is (and do so more consistently) and improve midrange output a little bit more.