Exhaust valve timing: rpm dependency 1

[Xplode]

I have a question regarding the timing of closing the exhaust valve.

From a P-V diagram, you can see that closing the exhaust valve too early would raise the exhaust pressure, causing higher pump losses, or in other words decreasing the internal efficiency. A part of the exhaust gases remains in the cylinder and also the volumetric efficiency is lowered.

If you would close the valve too late, some of the exhaust gases would re-enter the cylinder, again lowering the volumetric efficiency. But now the pump losses are slightly reduced, so the internal efficiency is raised.

You can conclude there is an optimal closing angle.

My question is how this optimal closing angle changes with engine speed?

My first thought was that the exhaust gases leave the cylinder at sonic speed through the valve, so the time needed is inversely proportional to the engine speed. As the crank angle is 360 x n x t, this would lead to a constant optimal closing angle.
Is this correct?

 

[ivymike]

I think the answer will depend in part on the specifics of the exhaust port, valve, and exhaust system geometry. You can change the "breathing" of an engine significantly by tuning the exhaust syatem, and the tuning interacts with valve timing vs rpm. Engine performance simulation software (GT Power, WAVE, etc) can help you answer the question for a particular configuration.

 

[BrianPetersen]

In general, "no", because of complex pressure wave interactions in the exhaust manifold. The positive pressure wave that was started at the beginning of the exhaust stroke sends back a negative reflection wave every time it encounters an expansion or junction in the pipe, and a positive reflection wave every time it encounters a restriction or contraction in the pipe (or the end of an adjacent header pipe that has a closed exhaust valve at the end of it). These pressure waves move at a speed that depends on the temperature and the intensity of the pressure wave (in simple terms, the speed of sound - although with strong pressure waves, the speed of sound is no longer "fixed").

At various engine speeds, the situation during the overlap period varies. Typically the exhaust system will be tuned to give a negative pressure in the exhaust system at a certain RPM range to scavenge the cylinder. Outside of that range, there could be less scavenging or in some cases reverse flow due to positive pressure there.

Consequently, there's no simple single optimal closing angle. It will work correctly at one speed (if it has been optimized - not all are), it will work decently over a certain range of speeds, and work against you at other speeds.

All of this complication is why we have engine simulation software.

 

[reidh]

Xplode,

Just a quick correction to your OP: the exhaust gasses do not travel at sonic speed, rather as Brian noted it is the pressure waves that travel at sonic speed.

-Reidh

 

[Marquis]

To answer the original posters question: I've found the optimum closing angle as found on 1 D cycle simulation and on extensive dyno testing, is that it changes with engine speed, an early Exhuast valve opening is required for higher engine speeds, as there is less real time around to vacate the gases during blow down, and at lower speeds a later EVO (nearer BDC) is usually optimum to prevent reversion and to get maximum work out of the cycle.
The actual numbers change with engine type , exhaust port configuration and whether a tuned length exhaust manifold is used or not, expansion ratio, cyliner pressure and cylinder size etc etc. The trends rarely do. Engines with dual variable cam phasing sometimes show a slight benefit over their single VCP counterparts in the torque curve department, but not by much, and more if the exhaust VCT is used in conjunction with a tuned length exhaust manifold system a la Ferrari F 355 or BMW S54 M3 engine (although some of this low speed torque benefit is undoubtably down to better scaveneging which is more to do with overlap than EVO)
The biggest benefit of exhaust cam phasing, regardless of what kind of exhaust manifold design is used is at part load and the impact of pumping losses and therefore fuel economy.

 

[Tmoose]

I think the intake valve closing point is more likely to define the lower rpms that an engine will run well. Delaying the intake closing for a particular (low) rpm will really inhibit how much mixture is captured, and even the carburetor's signal. I think The loping idle and high idle speed required of carbureted cars with radical cams is an intake phenomenon.

 

[Bribyk]

Dodge elected to run variable exhaust timing on the new Viper motor, instead of intake timing. This was because the V10 already has plenty of torque and they wanted to improve idle and low-speed characteristics.

 

[Tmoose]

Hi Bribyk,

you wrote - "Dodge elected to run variable exhaust timing on the new Viper motor, instead of intake timing. This was because the V10 already has plenty of torque and they wanted to improve idle and low-speed characteristics."

This SAE info includes the benefits you mention, but suggests variable intake timing's "prime benefit would be improved low-end torque." timing

http://www.sae.org/automag/technewsletter/070402Powertrain/04.htm

I think that greater torque (by definition BMEP) would result mostly from better cylinder trapping/filling, since the reduced charge dilution provided by ex valve adjustment alledgedly does does not improve BMEP.

The original post was concerned with various pumping and volumetric efficiencies. I'm thinking the SAE explanation of variable ex timing's benefit as emission improvement with no improvement in torque, unlike intake timing adjustment (almost certainly earlier intake closing) which would have boosted torque, speaks for increased volumetric efficiency via charge retention as a result of earlier intake closing (than is required for 6000 rpm operation).

If this dyno test is legit that is a mighty impressive torque curve from 2000 rpm up for a 2 valve motor.

http://www.dragtimes.com/2008-Dodge-Viper-Dyno-Results-Graphs-14116.html

Gotta wonder what would happen if the Viper's intake consisted of carburetors and non-variable intake plumbing. Port Fuel INjection does a nice job of providing equal fuel to each cylnder. Then, long intake runners and a plenum protect the air flow sensor from the reverse flow that fools a carburetor into the double stuffing carburetion effect at low rpm with delayed intake closing necessary for good 6000 power. And unlike a big carburetor, FI reads low rpm cfm with good understanding.

Road and Track may have put a slightly different spin on Dodge's variable exhaust timing rationale.

http://www.roadandtrack.com/article.asp?section_id=3&article_id=5891

"To meet future emissions requirements, a variable valve timing system was designed that employs a 2-piece camshaft. A hollow camshaft drives the intake valves, while a solid shaft mounted inside it can be rotated independently up to 40 degrees to vary the exhaust timing. This helps optimize valve overlap at low and high engine speeds. "

http://www.roadandtrack.com/assets/download/1007_Viper_SRT10.pdf.pdf

I wonder if 1100 rpm @ 60 mph in 6th is an operating condition, and where the $2100 gas guzzler tax goes?

 

[SomptingGuy]

Exhaust timing was the be all and end all on my old RD350 YPVS. Those were the days!

- Steve