New split-cycle concept to control diesel HCCI combustion
25 October 2010
|
|
| Turbocharged HCPC engine scheme. Click to enlarge. |
A team from Universita degli Studi di Pisa (Italy) and Rolf Reitz at the University of Wisconsin-Madison are proposing a novel combustion concept—Homogenous Charge Progressive Combustion (HCPC)—based on a split-cycle principle to control HCCI combustion in diesel-fueled engines. Ettore Musu from the University of Pisa presented a CFD study of concept at the SAE 2010 Powertrains, Fuels and Lubricants Meeting in San Diego.
The HCPC concept consists of forming a precompressed homogeneous mixture outside the cylinder and then gradually admitting the charge into the cylinder during the combustion process. The turbocharged concept allows reaching engine speeds of 6,000 rpm, with a high indicated efficiency of 45% along with clean combustion; power density at 6,000 rpm is 64 kW, with 300 kPa intake pressure.
The basic idea of the new Diesel combustion illustrated in this paper is to control the heat release rate by a gradual supply of an almost homogeneous charge, without relying on exhaust gas recirculation to moderate the reaction. Thus the HCPC range can extend to all engine operating conditions, including very high loads.
—Musu et al., SAE 2010-01-2107
The intake and compression phases are performed in a reciprocating external compressor, which drives the air into the combustor cylinder during the combustion process, through a transfer duct. A transfer valve is positioned between the compressor cylinder and the transfer duct.
The compressor piston has a fixed delay, in terms of crank-angle degrees, with respect to the combustor piston. The combustion takes place after combustor TDC; during the combustion process, the combustor piston moves downwards whereas the compressor piston moves upwards. As a consequence, the air moves from the compressor cylinder to the combustor cylinder. Contemporary with the air transfer, fuel is injected into the transfer duct, evaporating and mixing with the air, thereby bringing about the conditions needed for homogeneous combustion.
| HCPC Combustor and Compressor | ||||||
|---|---|---|---|---|---|---|
| Combustor | Compressor | |||||
| Displacement | 598 cm3 | 250 cm3 | ||||
| Bore | 86 mm | 68.3 mm | ||||
| Stroke | 103 mm | 68.3 mm | ||||
| Geometric compression ratio | 85:1 | 79:1 | ||||
| Squish height | 0.5 mm | 0.5 mm | ||||
In the study presented at SAE PFL, a turbocharging system with overall efficiency of 50% was considered for the inlet and outlet pressure. A pilot injection was used to achieve stable ignition timing, and the injector was a 7-hole injector.
Increasing engine speed from 2000 to 6000 rpm, indicated thermal efficiency decreases from 47% to 45% for the turbocharged case and from 46% to 44% for the naturally aspirated one due to higher pressure losses in the transfer phase. These results are very meaningful and represent an important step forward in the HCPC engine development. As a matter of fact, the new HCPC engine with smaller compressor is able to run at the even better ISFC of diesel engines at speeds that are typical of SI engines for passenger cars.
...Unlike in conventional diesel engines, the speed of the combustion process increases along with the engine speed, keeping almost constant combustion quality and duration in terms of crank angle, like in SI engines.
Pollutant emission behavior vs. engine speed...confirms the validity of the HCPC combustion concept. Emissions are almost independent of the engine speed. Soot and HC are orders of magnitude lower than the ones of a diesel engine. The low CO emission trend confirms the high combustion efficiency even at high speeds. NOx emissions are almost constant and similar to those of a diesel engine.
—Musu et al.
Resources
Ettore Musu, Riccardo Rossi, Roberto Gentili and Rolf Reitz (2010) CFD Study of HCPC Turbocharged Engine (SAE 2010-01-2107)
Tamagna, D., Musu, E., Gentili, R. (2007) A Preliminary Study Towards an Innovative Diesel HCCI Combustion ASME Paper No. ICEF2007-1743 (2007).
Power density is 64 Kw/.........?
Could this engine be scaled up/down enough for high efficiency gensets for extended highway speed PHEVs?
Posted by: HarveyD | 25 October 2010 at 12:55 PM
This looks like a Scuderi engine with a turbo on it. The only significant difference in the engine itself is that the transfer valve. I've always been skeptical of such designs because of the pumping losses incurred.
@HarveyD
Don't know but it's sort of irrelevant. High power density gensets are possible with existing (and cheap/proven) ICE technology. The ICE that was chosen for the Volt is a dinosaur. It's almost as if they didn't want the Volt to be too good out of the gate. Why they wouldn't use a highly boosted atkinson cycle engine is beyond me.
Posted by: GreenPlease | 25 October 2010 at 02:10 PM
I was about to say the same thing about the Scuderi resemblance, GreenPlease.
The expander is 0.6 liters, so 64 kW output means a power density of about 107 kW/liter. Two sets with 1 liter total expander displacement should be enough for most passenger vehicles, and it looks ideally suited to a compact "V" engine arrangement. A single pair would be more than enough for the Volt.
Posted by: Engineer-Poet | 25 October 2010 at 04:23 PM
Is this just a paper study? It isn't clear from the above whether anything has actually been built, and you can't read the original SAE paper online without payment.
Regardless, split-cycle ICEs have to solve the materials/cooling problem caused by never running the cool intake charge into the expansion cylinder. I'll be impressed when I see actual test results.
Posted by: Nick Lyons | 25 October 2010 at 08:08 PM
Agree, Nick. This engine will face the same issues as the Scuderi as we have discussed before.
BTW, the Lund university and Prof. Bengt Johansson have shown 57% indicated thermal efficiency with pretty much a conventional Otto-Diesel engine setup...pretty much a hybrid between the two cycles. See
http://www.greencarcongress.com/2010/09/ppc-20100928.html
At 45% indicated efficiency, this team has a lot of work to do, especially with a totally new and unproven valves actuation and engine design and setup, etc.
Posted by: Roger Pham | 25 October 2010 at 09:36 PM
Very large piston diesel engines get very high efficiency, 50 percent.(more than some fuel cells) The efficiency of an automobile engine is not very important because people operate the vehicles in an inefficient manner and buy large vehicles that are inherently inefficient.
Plug-in-hybrid vehicles are not as easy to drive inefficiently and computers can be programmed to keep them more efficient.
Very simple single piston engines are good enough for emergency range extenders for most automobiles. Exhaust filters and converters can keep the exhaust clean enough and on the average there will be no exhaust. ..HG..
Posted by: Henry Gibson | 25 October 2010 at 11:54 PM
This split cycle has the advantage that it doesn't require the finicky control systems of the Partially Premixed Combustion system.
This might play well with the Transonic supercritical fuel injection system. Exhaust heat would be recycled, and there would be no issues of incomplete evaporation of fuel droplets leading to particulate emissions.
Posted by: Engineer-Poet | 26 October 2010 at 10:27 AM
Also, there needn't be any problems with cooling. Since there is no compression of an unburned fuel charge in the expansion cylinder, the cylinder head and piston crown can be made of refractories such as silicon nitride. This cuts heat loss, increases expansion work, and yields hotter exhaust gas for improved turbocharger performance.
Posted by: Engineer-Poet | 26 October 2010 at 05:51 PM
@E-P,
You would think so, intuitively, but Woschni et al in 1988 published the findings of their direct experiment in actual combustion chamber, measuring heat flux in insulated combustion chamber...and the results are not very encouraging...See the following link:
http://scitation.aip.org/getabs/servlet/GetabsServlet?prog=normal&id=JETPEZ000110000003000482000001&idtype=cvips&gifs=yes&ref=no
Let me post here an abstract of this paper:
"Recently great expectations were put into the insulation of combustion chamber walls. A considerable reduction in fuel consumption, a marked reduction of the heat flow to the cooling water, and a significant increase of exhaust gas energy were predicted. In the meantime there exists an increasing number of publications reporting on significant increase of fuel consumption with total or partial insulation of the combustion chamber walls. In [1] a physical explanation of this effect is given: Simultaneously with the decrease of the temperature difference between gas and wall as a result of insulation, the heat transfer coefficient between gas and wall increases rapidly due to increasing wall temperature, thus overcompensating for the decrease in temperature difference between gas and wall. Hence a modified equation for calculation of the heat transfer coefficient was presented [1]. In the paper to be presented here, recent experimental results are reported that confirm the effects demonstrated in [1], including the influence of the heat transfer coefficient, which depends on the wall temperature, on the performance of naturally aspirated and turbocharged engines. "
Posted by: Roger Pham | 26 October 2010 at 07:46 PM
Ford's experimental adiabatic diesel (silicon nitride cylinder and ringless piston, which I've held in my hands) showed the effect of volumetric efficiency loss. Induction of charge air into an uncooled cylinder results in premature heating of the air charge, lower volumetric efficiency and increased compression work. Gains in such engines came from turbocompounding, recovering the energy in the hotter exhaust gas. However, a split-cycle engine does not suffer such losses because it does not induct or compress the air charge against the uncooled walls. They are ideal to exploit insulated combustion chambers.
Posted by: Engineer-Poet | 26 October 2010 at 08:55 PM
@E-P,
You're discussing another phenomenon, not heat transfer rate at very high wall temperatures from the hot gas to the wall. Here, Woschni et al 1988 reported strictly heat transfer rate during combustion, not other types of heat transfer during compression nor expansion. They have heat sensors embedded in various positions inside the cylinder and piston to record instantaneous temperatures at the moment of combustion. This was an actual experimentation with data gathering in a well-controlled laboratory, not a speculation.
Let me remind you that Prof. Woschni is considered an authority in in-cylinder process, whose 1967 heat-transfer model and equation have been widely in engine calculation and simulation.
Results of insulated-combustion engines later have shown only very modest gain in BTE of a few percentage points the most, and only in turbo-charged diesel engines. The post-combustion heat recyling by the turbocharger could have explained this modest gain.
Posted by: Roger Pham | 26 October 2010 at 10:25 PM
Then tell me if Woschni or anyone else is still using that 22-year-old heat-transfer equation (people make mistakes, after all, and they do get past peer review), and if it has any relevance at all to a uniflow engine.
Ah, at last: you admit it increases efficiency. It also decreases the heat load on the cooling system and materials beyond the insulation itself, so the bogeyman of a melting engine is banished. In this split-cycle engine, it would probably have a greater effect on efficiency because the expander could be larger relative to the compressor.Posted by: Engineer-Poet | 28 October 2010 at 05:54 AM
Before discrediting a long-time authority in the subject, please recall the difference between heat transfer coefficient and heat transfer rate.
The latter depends on the temperature gradient. Succintly, the higher heat transfer coefficient means that the use of combustion chamber insulation will be less effective than if the heat-transfer coefficient would have remained the same, that's all.
The much lower indicated efficiency of split-cycle of 45% vs. 57% of regular Diesel-Otto cycle as shown by the Lund University will make this a moot point. Even going thru all the complexity of split cycle and insulation, it would be difficult to top the Lund's result.
Posted by: Roger Pham | 28 October 2010 at 01:34 PM
If it's only the coefficient that goes up, we can bring the actual net transfer to near-zero by letting the surface of the insulation heat up to the average chamber temperature.
This has been done in quite a few adiabatic engines.
Unless the split-cycle can- Achieve better than 50% turbocharger efficiency, allowing more work to be extracted either via a larger expander or turbocompounding.
- Achieve greater mechanical efficiency due to higher volumetric efficiency
- Insulate the expansion cylinder to achieve greater work output and transfer more energy to the turbocharger.
It looks to me as if all of those are quite feasible, and there are none of the finicky mixture-control issues of Lund-style stratified charge systems.Posted by: Engineer-Poet | 30 October 2010 at 12:19 PM
Agree.
Let's consider also an OPOC engine without cylinder head, therefore no heat loss from the cylinder head. This means no need for insulation, since the trouble of insulation would yield diminishing returns.
Now, if you would inject gasoline twice, like the Lund University's method. Inject gasoline after the ports have closed, and inject it again near TDC in order to ignite the already mixed charge. Without heat loss from the cylinder head, you can get even higher than 57% indicated efficiency...and then, an electric turbocompounder (TIGERS) to get squeeze out even more energy from the exhaust. An electrically boosted turbocharging system for rapid transient operation. No engine can top the volumetric efficiency of a 2-stroke OPOC...without heat transfer from the cylinder head, since there is none!
Posted by: Roger Pham | 30 October 2010 at 07:17 PM
OPOC is uniflow, but the exhaust-side piston can still benefit from insulation.
Posted by: Engineer-Poet | 30 October 2010 at 08:49 PM
@E-P,
WRT diabatic engine, I'm afraid that ceramic is still too brittle for engine use, unless there are very recent advancements that I'm not awared of. Metal engine has requirement for lubrication, which cannot withstand too much heat stress. Also, metals would melt or severely weaken at much lower temperature than the average temp of uncooled cylinders. Lowering overall combustion temperature and power density would really suffer, since combustion would be too slow and your engine rpm and hence output would plummet, and you may have emission problems also at higher speeds.
The romance with adiabatic engine is pretty much a work-in-progress after many decades.
Posted by: Roger Pham | 30 October 2010 at 08:57 PM
You don't have to insulate the cylinder walls; the bulk of the heat loss is through the head and the piston (which have the most area exposed to hot gas early in expansion), so using ceramic for the facing surfaces will yield most of the benefits available. The cylinder walls, rings, etc. remain conventional. Caterpillar was working on this back in the 1980's, looking to beat 50% thermal efficiency.
Posted by: Engineer-Poet | 31 October 2010 at 02:56 AM
@Roger & EP,
Did either of you go to the High Efficiency Engine Symposium or the Fluids and Lube meetings last month?
Either way, what is your feeling on how these single cylinder lab engine results will translate into real measures of efficiency (BTE) in multi-cylinder format?
I've yet to hear Reitz, Johanson or anyone else playing with XXXX (flavor of the week Low Temperature Combustion) strategy address how their concepts will avoid the significant drop usually seen if the concepts every make it that far.
As an air-handling expert, what worries me is the ability to produce the in-cylinder conditions that are manufactured in the laboratory to achieve their results. The requirements usually involve fairly lean AFR with very high EGR rates requiring high boost levels and low engine backpressure. The resulting low temperature exhaust (~400C) creates a very challenging problem statement for the turbocharger to create the required boost. Long story short, i don't think they can do it if they unplugged their supercharger and actually let the engine backpressure, AFR, and EGR rate equalize with real tubocharger hardware.
Whenever i've seen them challenged, they avoid they stonewall and rely on arrogant responses like "we've checked our math and it should work".
Anyhow, i'd be interested to hear your thoughts on this subject.
UA
Posted by: UnnaturallyAspirated | 06 November 2010 at 07:54 PM