Diesel vs miller cycle

[Chris Matheny]

High compression sets a higher intake velocity sooner in the intake stroke allowing for higher volumetric efficiency at lower usable rpm (Ram tuning). The intake valve close angle sets the rpm at which the most trapped charge will happen. Unless there is a restriction in the intake tract, a higher compression ratio will show higher trapped VE all else equal.

 

[Vigilant1]

High compression sets a higher intake velocity sooner in the intake stroke allowing for higher volumetric efficiency at lower usable rpm (Ram tuning). The intake valve close angle sets the rpm at which the most trapped charge will happen. Unless there is a restriction in the intake tract, a higher compression ratio will show higher trapped VE all else equal.

Chris, how does this work in practice? Let's say my engine has a CR of 8. I shave the head to reduce the size of the combustion chamber and now I have a CR of 10. The swept volume is the same, the cam is unchanged, the induction stroke lasts the same time and the valve opens and closes at the same points in the piston travel. With all that, how does the shaved head change the amount of air (by mass) that gets taken in (i.e the volumetric efficiency)?

It means the oxygen is more concentrated (smaller space)

Sure, but that's not "more oxygen."

By the way ... gasoline will not burn without oxygen ... but oxygen will burn without gasoline.

Oxygen will not burn without fuel of some kind.

Mark

 

[rv6ejguy]

VE doesn't change with CR. Thermal efficiency increases which shows as more power since more energy has been extracted from the fuel. At some point, the energy losses in compressing the charge offset the diminishing gains in TE as CR is increased. This limits the highest useful compression ratios on SI engines to about 16 to 1 effectively.

 

[Chris Matheny]

Compression ratio sets the intake velocity in motion. Think of having a syringe with 50cc of air in it statically, plug the end and draw another 50cc in, think of the force it took to do so. Now reduce starting cc to 10cc and draw another 50cc into it and measure the force. This is how a higher static compression ratio forces a higher intake charge velocity and can induce ram effect at lower rpm. The piston creates a greater pressure drop across the intake valve on a higher compression ratio engine. Increased charge velocity can increase VE and trap more charge, even to trapping over 100% due to air and fuel being mass in motion. More charge equals more O2, fuel and the real workhorse of the equation Nitrogen. You can burn gasoline vapors and oxygen all you want but the heat from their burning expands the nitrogen (and other trace inert gases) and their thermal expansion is the force on the piston. Without them (working fluid as they are called) we get a flash flame with no expansion to push the piston down and turn the crank.

 

[PMD]

More charge equals more O2, fuel and the real workhorse of the equation Nitrogen. You can burn gasoline vapors and oxygen all you want but the heat from their burning expands the nitrogen (and other trace inert gases)

I do hope you realize Nitrogen is NOT an "inert" gas, it is reasonably reactive (and of course why we have NOx emission issues).

You physical analysis is correct, the change in volume from going from 6:1 to 10:1 CR in a 100cc displacement cylinder means the starting volume of the void you are going to expand goes from 20cc down to 11 cc, so the initial drop in pressure from descending piston COULD be significant except that there are other variables that come into play. Your reasoning assumes that there is some condition of pressure in the cylinder at TDC that can take advantage by having a valve that blocks flow on one side and is fully flowing on the other side of TDC. Reality is, valve timing has overlap and the exhaust valve is still open and the intake open as well during the transition from the exhaust stroke to the inlet stroke. One must consider the resulting pressure (thus why exhaust tuning CAN be so critical) during the part of the overlap of cycles when the reduced chamber volume could make any significant difference in volumetric efficiency by providing a lower pressure on the cylinder side of the valve and port.

The takeaway from this should be that if your exhaust tuning is effective, it may be significantly beneficial to have a larger volume at TDC at a lower-than-atmosphere (actually lower-than-port) than the benefits of the smaller starting volume would give at 1 bar and TDC opening. Before one dismisses this out of hand, just remember how incredibly effective exhaust tuning was and is in 2 cycle engines - where you need to manage a significant portion of time with both inlet and exhaust ports wide open right at BDC.

Now, I don't dispute that there IS some difference, but it will be much less than assumed due to the nature of cylinder pressure during valve overlap. I might point out that during my 4 cycle SI days, I spent a LOT of time working of valve heads, seats and both the port AND combustion chamber side of the surface beside the seat since IMHO flow at low lift really DOES make a difference in flow later in the cycle as what you point out is probably true - the earlier and more effectively you can start the mass of gasses in the port moving the way you want them to go the more efficiently (yes, HIGHER VE is the goal) the overall inlet cycle (or exhaust) is going to be.

 

[rv6ejguy]

In my earlier days, I built my own flow bench and engine dyno and spent many years doing professional cylinder head and engine development for road racing engines- mainly on Toyota and Nissan engines. I've done thousands of flow bench and dyno runs on several different dynos.

The flow bench is a very useful tool but gains there don't always translate to gains on the dyno which is always the real proof that you're going in the right direction.

I see lots of folks worried about flow at .025 or even .050 valve lift however the amount of flow down there is a small percentage of half or full lift valve flow which constitutes the bulk of the valve open time (duration). Low lifts contribute minimally to filling or scavenging the cylinder, especially it you look at how long the valve is open at these small lifts. Total flow is a product of flow and duration. Flow is highly affected by delta across the valve which changes with piston position. The piston moves slowly near TDC and BDC where intake and exhaust valves are open only small amounts. Delta is low, flow is therefore low.

I pulled this out of my flow bench archives to illustrate:

.050 valve lift intake flow 14 [email protected] 10 inches H2O
.600 valve lift intake flow 129.0 [email protected] 10 inches H2O

.050 valve lift exhaust flow 8 [email protected] 10 inches H2O
.600 valve lift exhaust flow 89 [email protected] inches H2O

Simple flow benches simulate steady state flow and X pressure differential but that's not what happens inside a running engine where volumes and pressure deltas have infinite changes.

VE describes how much air an engine can induct in relation to it's cylinder displacement and is always highest at torque peak rpm.

Since flow is so low at low lifts, this translates into slow flow and slow velocity changes within the intake and exhaust passages when you look at the volumes you are filling and emptying.

The impact of slightly less chamber volumes going from 10 to 12 to 1 CRs is very minimal in how rapidly the pressure deltas change with the piston near TDC and low valve lifts there. Therefore it's impact on cylinder filling rates, hence VEs, are minimal. The bulk of the inducted air is moved during the half/ full/ half valve open period where exposed valve aperture is large and pressure deltas are large. Charge inertial effects are huge at high rpms and you need big and rapid Delta P changes to affect the charge inertia. This isn't happening at small valve lifts near TDC.

Most VE increases come from optimizing "wet" flow in the head (fuel droplet size and movement important), camshaft design and intake/exhaust optimization. CR is determined usually more by the fuel octane which will be used in a particular engine and that may be one of the very first considerations when developing any new engine for stock or racing applications. In other words you pick CR first and that is fixed for the rest of the design decisions you make down the line. Not much you can do about it.

 

[PMD]

Ross: Glad you point out that the flow bench is only part of the story as it does not include the whole range of dynamics going on in a running engine. I realize that low lift flow is a small thing on a flow bench but I firmly believe in not leaving ANYTHING on the table. Three things I feel important about low lift flow: the velocity at low lift (especially of an opening exhaust valve or inlet on a boosted engine) can be extremely high and that small "jet" of flow can do the Bernoulli thing and start a much more massive port or chamber full of gasses moving in the right direction. Secondly, it will have some affect on the accoustics of what is going on both in the chamber and in the port. Finally: If you can get the flow cleaned up at low lift the much high mass/lower velocity flow through a more open valve should suffer less turbulence spilling of of the (no longer sharp) edges of the valve and chamber around the seat. Something I always wanted to try, but never figured out how to do is to see streamlines in flow through a valve/port to see how/where turbulence originates, expands or decays.

 

[rv6ejguy]

My point was and I don't think I made it clear, I never found any useful gains on the dyno regarding low valve lift flow improvements shown on my flow bench and there are way more productive areas to unleash a lot more hidden hp- camshaft, chamber mods, windage, friction, exhaust, intake etc.

Wet flow benches have been around for over 2 decades to allow flow visualization and reshaping of ports and chambers.

 

[PMD]

[QUOTE="rv6ejguy, post: 632561,Wet flow benches have been around for over 2 decades to allow flow visualization and reshaping of ports and chambers.[/QUOTE]Sadly, I went on from building engines to building many other things about 30 years ago. Will have to look it up and see how that is done...thanks.

 

[lelievre12]

.
Keep in mind that high compression itself doesn't really do anything .

It is the fact that more oxygen has been brought in to aid fuel burn.
.

Once the Lambda >1 then more oxygen doesn't really help "burning". What helps efficiency is the better system exergy resulting from igniting the fuel at a higher starting temperature and pressure.

Using ideal gas law for Otto Cycle we can see the benefit in the following exergy flow;;

Compression Ratio=6.2:1


Note the isentropic expansion from 3 to 4 produces 1.33 MJ of expansion work output per gram of fuel.

Now all things remaining equal, look at the same cycle at Compression Ratio=9:1


Using the same amount of fuel and the same amount of oxygen the isentropic expansion from 3 to 4 produces 1.60 MJ/g of expansion work output. Thats 20% more power for the same fuel and air.

Fossil fuel is a high exergy fuel. It can burn at very high temperatures if gas pressure is high enough. By elevating the starting temperature and pressure before burning the fuel we "unlock" this exergy benefit.

There are many other factors at play which for the so inclined can be read in more detail at: PII: 0196-8904(88)90054-4 (upenn.edu)

 

[Chris Matheny]

I too have built my own flow bench and found where and if it can be of help. It has its limitations but so does a dyno, you can make a great dyno engine that is a turd on acceleration, ( we aren't worried about this in planes but its an example of testing limitations). A small combustion chamber can also be a big benefit during the 5th cycle as I call it (overlap) where a smaller space can allow a way more efficient scavenging of the cylinder, and with correct exhaust tuning create a pressure drop before the piston even leaves TDC making for a cleaner less diluted charge and also less pumping losses in the system as a whole. Everything is a compromise in an engine. Like the thick antiquated rings they use. I spoke with a race team where I did my cylinder head and induction classes as they spun a short block with stock 5/32 rings and one with .040 rings and found just ring drag at 7000 RPM took 150HP to turn with the thicker rings and around 60HP with the thinner rings that actually sealed better due to higher conformity to the cylinder walls. Unfortunately in a big air cooled engine the rings transfer a large portion of the heat from the piston to the cylinder for cooling. Granted we are not spinning these rpm but piston speed may not be as far off as you think when dealing with the long 4.375 stroke. We are unfortunately forced to "optimize the compromise" to have an engine that makes good power but also lives a long life.

 

[PMD]

I spoke with a race team where I did my cylinder head and induction classes as they spun a short block with stock 5/32 rings and one with .040 rings and found just ring drag at 7000 RPM took 150HP to turn with the thicker rings and around 60HP with the thinner rings that actually sealed better due to higher conformity to the cylinder walls.

Any idea how they measured/derived/calculated the ring energy losses? Finding 90 HP just sitting right under their noses is pretty impressive. Also: do you have any reference to published studies of this same topic? I have trouble imagining that an 0.040" top ring would not turn into a large diameter Belville washer under combustion loads and the contribution of combustion pressure to ring pressure against cylinder wall would be minor. You have my undivided attention on this one!!!

 

[Chris Matheny]

They spun the engine with an electric motor like used on a spintron. The thin rings like a very flat ring land machined after all other machining. .040 isn't a thin ring these days either. I have an engine here I'm assembling with a .8mm ring pack and .6mm are becoming a lot more common. Most autos today run a 1mm top ring in them and go for many hundred thousand miles. I'll look in my induction class reference to see if I can find the example today when I'm at the shop.

 

[PMD]

Thanks. Yes, aware that there are some very thin rings in the SI world, just that everything I have had in my hands for some time is usually a big diesel. I though 1/8 was "thin" for compression (oil control rings have been very thin for some time, so I can see they work in a well lubricated part of the cylinder) but still have a picture in my mind of "L" compression rings that use cylinder pressure to seat. Simply did not occur to me there could be THAT much difference in frictional losses.

I did find one paper, and it is quite interesting: SAGE Journals: Your gateway to world-class research journals
My concern with rings that narrow would be the pressure against asperities during reversal at TDC and the resulting "park marks", but as you point out, there seems to be pretty good experience with this at the 1mm range today. BTW: in this paper note the energy losses assumed are in the 3-6% range for top ring.

 

[Chris Matheny]

Looked through my class books and didn't find the study. Emailed the instructor and he is no longer with that engine builder so doesn't have access to the test results anymore. We did converse and he said at low RPM like a traditional aircraft engine there was very little to gain with thinner rings but in auto conversions and Rotax like rpm they become a lot more of an advantage. He also re-iterated that the big rings in the low rpm engines transferred a lot of the heat to the air cooled cylinders. He was involved with a reno racer a couple years ago and was fairly successful with adding new approaches to the old tech engine. Look up Justin Meaders 2018 Reno gold class. I feel like we've veered off topic.