exhaust port to resonator length measured
 

I measured the distance from the exhaust ports to the resonator while I was under the car this weekend. The distance is about 6 feet from the front exhaust port to the front of the resonator. According to Harvey, the sound wave from an exhaust pulse should come back to the exhaust port at about 180 degrees after the exhaust valve opens. The rpm at which the resonator accomplishes this is:

rpm = 15*v/d

where

v = speed of sound
d = distance from exhaust port to the resonator

If the exhaust gases are pretty hot, then the speed of sound may be around 1600 ft/sec, and for the stock resonator, this gives

rpm = 15*1600/6 = 4000 rpm

The third cat is about 4.2 feet from the exhaust ports. If this thing can actually cause a reflection, then it causes effective scavenging at:

rpm = 15*1600/4.2 = 5400 rpm

So, it might be doing some good at 5400 rpm.

Mychailo,
Will this equation really hold, given that the cats (all three of them!!) aren't true circular cross sections? I can see how it would work if we had straight pipe all the way back to the resonator.
-Bill

Theory, and what actually happens, can be miles apart. So many things can change the final length that is used. The temp of the exhaust gas has the most effect. The exhaust shielding and the catalytic converters will keep the temp high. I think 1800 ft/sec is more like the speed of sound at the front section.

The point that the valve has lifted enough, to produce the sound wave, will change with engine speed. The point that they want to bring the reflected wave back, may be different to what we think due to the way the inlet system works.

This is only finally decided by building the complete system, with all the cats and muffler fitted, then running the set-up on the dyno, to find where the resonate peak occurs in the cycle. The system is then altered to bring the resonance to the desired point. It may be altered to remove any resonate pulses that are working against the engine at a different rpm.

The one thing that we can be sure of, is that the high speed resonance will occur at maximum torque revs, along with the inlet tract resonance. The two systems will work together, the exhaust starting the inlet flow by creating a negative pressure at the start of the inlet phase. The inlet resonance producing a positive pressure at the end of the inlet phase to force more air in.

If the two systems don't coincide, the engine won't produce the maximum torque that it should. Subaru have done a great job on this engine, with a very wide torque curve. So in my opinion, changing the front end of the system, will only loose top end power.

Harvey.;)

Yeah, I wonder about this too. Any time the diameter changes, this can cause some reflection. I read that for a strong reflection to occur, the exhaust gas wave has to encounter a sharp increase in cross sectional area, and the area increase has to be at least a factor of 8. The primary cats has a maximum ID of about 105 mm. The exhaust tubing before the cat has a diameter of about 43 mm. This is about a factor of 6. Pretty close. Also, the cats have a thick (several mm) screen in the exhaust flow. I would think this would cause some dampening of the sound wave. So, my thought is that the cats reduce the effectiveness of this resonance action.

As Harvey said, Subaru probably put the whole mess on the dyno, and through trial and error, got the most out of it with the limiting conditions.

My one thought at this point is whether or not it would be wise to replace the center cat with a merged collector.

Non catalitic model.
 

In the Australian RHD WSM, there is a model that does not have catalitic converters, exhaust gas recirculation or O2 sensors. It also runs leaded petrol. I am sure it was not sold here.

The output quoted is 230 hp @ 5600 leaded
220 hp @ 5600 unleaded

Torque quoted is 228ft/lbs @ 4800 leaded
224ft/lbs @ 4400 unleaded.

The interesting point is the higher rpm that torque is developed, is this the result of no cats, or some other tuning changes.

I don't know where this model would have be sold. Though it would be good to check it out, to see what the exhaust differences are, without the cats. What is used in place of the 3rd cat to join the two header pipes? More questions then answers.:)

Harvey.;)