LS engine torque curve dip from 3000-3500 rpm

[racin69z]

I've been reading and looking at a bunch of dyno graphs of LS engines. Most all of the test start at 3000 rpm. There is almost always a dip from 3000 to 3500 and then it goes up from there. Ive seen a couple of graphs

Is that a byproduct of how the dyno loads the engine when the pull starts, or is it really a loss in torque. If it is a loss of torque, does anyone know what causes it and how to change that?

Lynn

[CGT]

One or the other or both. Getting an engine to pull clean at 3000 can be part dyno equipment, part engine, or part operator. And even still the first several hundred rpm of those pulls can be questionable. I tend to be skeptical of those numbers unless I was present to hear the engine or watch how the pull was made .

Also, most of things I've seen pulled in that 3 to 6500 to 7ish range generally have a dip through an area below peak torque. The engine is just not "in tune" through that range. Valve timing, header sizing type and lengths, intake sizing, lengths etc.

[Newold1]

Go online maybe a pull up some dyno runs from someone like Steve Brule at Westech and you will see that a dyno operator who does dyno pulls all day and all week long can really make a difference on how a dyno is calibrated and set up for incrementally better data collection and printouts. There is no fixing an engine that has weak areas of output on a dyno other than tuning and adjustments, but dyno's can point out the Jewels and Junk in an engine build!

However as previously said here a dyno can also show the performance dips and holes in an engine build with respect to parts that just are not quite right! This is one of the more important aspects of dyno's they show performance over an entire rpm band and not just total HP and torque.

[PackardV8]

OTOH, when running computer dyno simulations, the graph will often show dips in the 2500 -3500 range which don't seem to respond to changes in intake, compression or cam timing. Sort of like the horsepower and torque curves crossing at 5252 RPMs, maybe it's just the math in the formula.

jack vines

[CGT]

OTOH, when running computer dyno simulations, the graph will often show dips in the 2500 -3500 range which don't seem to respond to changes in intake, compression or cam timing. Sort of like the horsepower and torque curves crossing at 5252 RPMs, maybe it's just the math in the formula.

jack vines

They dont respond real well in real life either....through that area

[rebelrouser]

OTOH, when running computer dyno simulations, the graph will often show dips in the 2500 -3500 range which don't seem to respond to changes in intake, compression or cam timing. Sort of like the horsepower and torque curves crossing at 5252 RPMs, maybe it's just the math in the formula.

jack vines

My race car engine has one of those dips on the performance trends software. Played with it a little and no matter what cam I put in the software it stayed. I had Jones spec me a camshaft and when I plugged it in, it went away, one of the reason I went ahead and purchased it. The car ran good, in the 9's, but always wanting more.

[BradH]

This isn't LS-based data, but some engine dyno tests I ran years back on a 452" BBM were from 3500-7000. We tried four different intake manifolds, along with some different spacer & carb combinations. There was a dip in the torque curve between 3500-4000 that -- for my application -- seemed influenced mostly by the intake being used. The range between the best HP & torque peaks for all the intakes was about 10#s (about 615-625 HP & 570-580+ tq respectively), but that dip between 3500-4000 varied by close to 50 lbs-ft. When it drops that much, it ain't a dip, it's a hole.

The worst offender was the intake w/ the largest plenum & largest average CSA of the runners (also, the most tapered runners), which made as much peak HP as the one other intake that made about 625, but was never as strong at peak torque. The "torque king" intake only made about 615 HP, but was also the only one to exceed 580 tq, and had virtually no dip in that same RPM range. The other 2 intakes' torque curves were basically "flat" in that range.

Not saying it's definitely "an intake thing", only trying to point out one of the things the LS engines could also be sensitive to. Another build of mine with the same basic bottom end, almost identical valve events (but more lift), and better heads had no dip at all in the torque in that range using an intake that was very similar to one that was "flat" with the earlier combination, and made more torque & HP across the entire RPM range. Thanks - Brad

[pcnsd]

I am not an LS guy, but I have seen the 1/2 peak HP RPM dip repeatedly in various singles and twins I have worked on. I think is a common phenomenon once camshaft overlap period gets above ~50* duration with tuned length exhaust systems. The effect appears to get worst as the overlap period grows in duration. I have a theory as to the root cause and that theory seems to be supported by the things, I have seen work to minimize but still not eliminate the dip.
My root cause theory:
Tuned length exhaust systems are tuned for peak HP RPM, typically using the 2nd wave (3rd harmonic for PipeMax folks) as the primary length. This same length is also tuned at other RPMs depending on wave cycle (3, 4, 5, etc.). The 4th wave is resonant on such a pipe at ½ the tuned RPM, but the while the wave cycles at the speed of sound, the valve events have decreased their relative speed allowing a much larger time window during overlap. This increased time window provides the opportunity for things to go wrong on the 4th wave.
Only lower relative exhaust pressure is beneficial during the overlap, but there is now twice the relative time window available. The negative pressure arriving at the valve interface is followed by positive pressure which decreases the cylinder fill efficiency and lowers the torque.

Things I have seen that mitigate the dip:
1. Stepped exhaust pipe system with a one size smaller pipe than previous off the head.
2. Anti-reversion chamber placed at the first step location.
3. More lift, less duration camshaft.
4. Advancing the camshaft slightly.

I have a tuned length intake. I can’t say that it helped with the torque dip during tuning but it is targeted to support RPM between PT and PHP so that should be no surprise. That said, I absolutely think it logical that an intake could be built that could counter the fill efficiency loss to some degree as suggested by Brad H. It is just countering one wave with another.