HCCI variable compression high gas flow ORBOX engine

[TheFelixWankel]

Wonderful design! i must say!

Howerver; with not fully understand in the motor, i can see this engine may suffer from friction and sealing issues. It is very difficult to seal abnormal shapes when exposed to these kinds of pressures.

As well i am worried about heat. The rotating cam can be cooled easily, yes, but what of the two shifting veins on the top and bottom? Currently the only sure-fire method would be cooling via fuel, but this would lead to poor fuel consumption. What are your ideas?

[orboxbe]

@TheFelixWankel

Indeed, it's all about sealing, lubrication and cooling, before any combustion engine can be considered.

Sealing: Actually, the shape is not "abnormal", but like a cylinder, rather "well behaved". The glideplates on the pivots (shifting veins) control sideway sealing, a top seal acts as axial seal (here separate, but could be rotor shaped). Actually more tight than a (double) piston seal.

[orboxbe]

@TheFelixWankel

Friction: About 4% is lost in piston engines. Often you will see wear created at TDC on the cylinder wall, as pressure here presses the sealing rings out. At the area with highest thermal stress we have: hIgh pressure, small area. Not to good for lubrication and friction. Please consider the force on each tooth in a gearbox, and than compare it to the "rotating cam". If you do the calculations, the forces are not so great. And so is the friction.

[orboxbe]

@TheFelixWankel

Heat: Actually, containing heat is very important to get thermodynamic efficiency. So actually the inside combustion chambers in the pivots should be isolated. The great thing here is that these chambers do not have to be lubricated and can be very hot!! On the other side, to prevent the walls from melting, cooling has to be supplied. Most simple, like in the rotor, by axial channels (out of center). Or just cooling the backside of the "shifting veins".

[orboxbe]

Example, take t=12, s=5, we have rotor-pivot distance=13, and area 169-(3.14*5*6) = 75.

If you want a compression ratio of 10 and circular combustion chamber, the area should be 7.5 or a diameter of 1.5.

A rotor thickness of 2 would now give a minimum volume of 15, and a maximum volume of 150. Scale as you like (x 1cm?).

[orboxbe]

You can almost completely fill the minimum volume for high compression. The max volume is approx. (t*t+s*s) - (pi*t*s)/2 times the thickness of the rotor. The maximum volume can be easily calculated considering the ellipse has a surface of pi*t*s, with t and s half the max and min diameter.

The distance between rotor and pivot axis is sqrt(t*t+s*s), so the square as max expansion is 2(t*t+s*s). Substract the rotor surface, and divide by 2, to get the surface at only one side of the rotor.

[orboxbe]

If you are considering making a prototype, I would suggest to start very simple with an elliptical rotor. A template/stencil is easily printed (even Powerpoint). A combustion chamber can be made in the pivots at the cross of the straightedge.

See also my other youtube video on an elliptical compressor.

[Marc6331]

my understanding is that EGR is also useful in controlling ignition timing in HCCI engines, is this correct? how do you control ignition timing in your design?

[orboxbe]

@Marc6331

EGR has, within limits, some advantages, but certainly also disadvantages, For HCCI it is much worse compared to variable compression ratio (VCR) control using a plunjer, as in our design. EGR has problems in creating reliable combustion in SI engines, so image how it will effect auto-ignition.

[orboxbe]

@orboxbe

For instance, HCCI with EGR will have mixing limited by heat exchange between fuel-mix and exhaust gases. Mixing being a semi-chaotic process, pockets will precombust irregularly, and timing is difficult to control cycle-to-cycle.

[orboxbe]

@orboxbe

VCR does not have this problem, and has been proven a dependable HCCI control method. A pressure sensor measures the exact time of combustion, and adjusts the plunjer to advance or delay the combustion on the next cycle. Independent of fuel quality (diesel or petrol or whatever burns properly) and fuel-air mix. This means you run full-throttle all the time (so no throttle needed), so no pumping losses here.

[orboxbe]

@orboxbe

The amount of fuel getting injected in the inlet controls the amount of power generated, and especially at low loads, very high compression ratio's, and thus efficiencies, can be achieved. And of course with our much larger intake and exhaust ports, gas can be moved in and out much more easily, with again higher efficiency, and higher speeds.

[Marc6331]

Which tool did you use to create the animations?



You might want to think about adding an explosion effect, to show clearly the combustion chambers as well as the intake and exhaust process. (the combustion chambers are the "L" shaped forms in the corners of the rectangle right?)

[orboxbe]

@Marc6331 used Autocad Inventor to make the animations. But apparently it was not clear the main combustion chambers are between the rotor and the pivots, not the L-shaped secondary chambers, which could be used for precompression ('turbo"), but do not have an optimal shape for efficient combustion (but if you only want more power, probably they could be used for combustion also.). This is also where the VCR plunjers are, and control minimum volume at combustion.

[orboxbe]

@orboxbe

Gases can only enter through the blue inlet from below, and exit through the red exhaust on the top. If you watch the rotation, you can see all 4 cycles (intake, compression, expansion, exhaust) happening at the same time, one in each chamber.

[Marc6331]

I've taken another look at your web site. doesn't an elliptical design suffer from the same problems as a rotary? ie :

difficulties making seals

controlling lubrication

high unburnt hydrocarbons

manufacturing problems relating to synchronizing parts of the engine

keeping the engine cool

[orboxbe]

@Marc6331

rotary sealing:

I have most of the NSU and Honda reports on sealing a Wankel engine. The reason they fail is because the seals are pushed and pulled! This wears them out! Just see any animation on Youtube, and see how the tip is dragged and pushed during the epitrochoidal motion. Our design does not have that problem, as the seals are created by the glideplates.

[orboxbe]

@orboxbe

rotary lubrication:

Actually the Wankel is pretty well lubricated. The problem is that a lot of sealing oil is pushed into the exhaust creating pollution. This is one of the reasons the RX8 Renesis engines has side ports. We have the continuous lubrication after the exhaust. No oil is pushed in the exhaust.

[orboxbe]

@orboxbe

rotary HC unburnt hydrocarbons:

as mentioned above, in a Wankel, some oil is exhausted, contributing to the pollution. But more detrimental is the shape of the combustion chamber, which resembles 2 cups at combustion. When the rotor moves forward, one cup is still being compressed, while the other is already expanding.

[orboxbe]

@orboxbe Apart from these 2 banana shapes having a poor surface/volume ratio, the flame front on ignition can hardly reach the crevices, and unburnt fuel is moved along the surface. Now look at our design, where after combustion, on the other one side of the seal we have exhaust, on the other compression. Any fuel leaking towards the compression side, would be combusted the next cycle. Any fuel trying to leak towards the exhaust cycle would be overtaken by the moving seal.

[orboxbe]

@orboxbe

No crevices (which also in piston engines are the main cause for UHC (unburnt hydro carbons), moving seals, and HCCI will make make it lowest in pollution.

[orboxbe]

@orboxbe

rotory manufacturing:

drilling an epitrochoidal Wankel cylinder is very costly. Our rotor shape can be easily extruded aluminium, the pivots simple staightedges. The simplicity of the geometry allowing a low cost engine was the first thought. A Wankel actually is self synchronised as well, but needs a gear. There is no gear needed in our design, as the rotor makes its own (2 teeth) gear.

[orboxbe]

@orboxbe

If you would consider an outside gear, it could have the same shape, only in an oil bath. But I do not think this is needed. If you do the math, the forces on the glide plates and rotor, are not so high.

[orboxbe]

@orboxbe

rotory cooling:

Actually also in a Wankel the rotor is cooled through the axis, but it is more complicated as it runs excentrically on the crankshaft. Cooling the (hot) rotor through the axis is real simple in our design. The wankel has specific sides on the housing for combustion and exhaust, creating strong temperature stress. In our design, the housing will get uniformely heated.

[Marc6331]

There seems to have been some success building HCCI engines but only for low loads it appears. Is this the reason why HCCI are not yet commercially available?

When people refer to "low loads" are they refering purely to RPM? do you know what RPM range current HCCI engines are effective? would they operate effectively say up to 3000 rpm?

In this case your engine under the below configuration should work a treat?

[orboxbe]

@Marc6331

Actually, HCCI does not work very well in a valve controlled piston engine, which is one reason why research is marginal. Another is the very high peak pressure, as all fuel is burned instantaneously. A third is the difficulty to integrate a plunger/piston to create VCR, so less viable ways are researched. Adapting regular piston engines for HCCI has proven not economical. A different geometry is mandatory.

[orboxbe]

@orboxbe Important:

1) no valves:

With auto-ignition HCCI, the most important thing is uniform heat distribution. One big issue is the (very) hot exhaust valve, which commonly pre-ignites a HCCI mixture at higher loads. Our design does not have valves.



[orboxbe]

@orboxbe

2) combustion chamber wall: High peak pressures require thick walls, and instant heat release good cooling. Considering the very thick amounts of metal enclosing the combustion chamber at ignition, and large bearings, we will not have that problem.

[orboxbe]

@orboxbe

3) Real VCR, not VVT/EGR:

On a regular cylinder head, which is full with valves to maximize gas flow, there is no space for HCCI pistons/plungers, so the most common way to create HCCI (GM and others) is by variable valve timing(VVT). One way is by exhaust gas recirculation (EGR), where exhaust gas is kept inside by closing the exhaust valve early (or alternatively sucked back into the inlet). The amount of hot gasses determines how early the HCCI ignition starts.

[orboxbe]

@orboxbe

Problem with EGR is proper mixing to prevent pockets of hot gas. Another way is by leaving the inlet valve open longer, the engine now basically runs in an Miller cycle, with a longer power than compression stroke, but less power is generated now, unwanted at high loads. Proven, in many reports, working at high loads, is a VCR plunjer, ..as in our design.

[Marc6331]

do you have by any chance have any graphs showing (theoretical):

fuel consumption

power output

rpm

torque

compresion ratio

efficiency 

[orboxbe]

@Marc6331

Theoretical graphs I have seen many. Easy to make. But useless without validation. Just some indications

fuel consumption 200 g/kWh

power output 16 kW/kg, 14kW/liter, 800 kW/l (engine displacement)

rpm 30,000

torque 260 Nm (=Power/(2*pi*rpm/60) (ungeared) = 800,000/3,000=260 Nm)

MEP (mean effective pressure) 16 bar (2*pi*260/0.001 = 1.6 MPa= 16 bar)

compresion ratio 1:8 to 1:24

efficiency 40%

[Marc6331]

let me tell you where i'm at. i'm looking for a simple, light, compact, fuel efficient, easy to produce HCCI engine that could produce between 20 and 30 KW. when you say your 1000cc engine can produce up to 800KW, i'm thinking a downsized version. what cc and how many cylinders would i need using your design to get this sort of power and do you know what the fuel consumption would be?

[orboxbe]

@Marc6331

Properly engineered, using the right materials etc. a 1 liter engine could produce 800 kW. It would be flaming hot, and run 30,000 rpm. Run at a still respectable 3,000 rpm (compared to diesels) it would only be 80kW, the power of a "normal" passenger car. Scale down by half, (length, width, depth) and you would get about 10kW. So about 20x25x15 cm, about 10 kg.

[orboxbe]

@orboxbe

specific power: swirl is important to get a fast burning fuel mix. This is why the top of pistons are oddly shaped in engines. In our engine swirl is naturally occuring and increases on compression. Faster ignition again allows higher rpms. With less intake restrictions, there is no reason a 4 x 250cc could not run 30,000 rpm. At 4 ignitions per rotation (versus 2 for a 4 cycl. 1 liter 4 stroke engine)

So about 8x the specific power of a regular engine.

[orboxbe]

@orboxbe

So about 800 kW from a 1 liter engine. At 50 kg a specific weight of 16 kW/kg (about double a F1 engine). Considering cooling, lube and auxillary equipment also 50kg, still 8kW/kg (5x the RX8 Renesis 1.3 liter Wankel engine at 184 kW @ 125kg)

[Marc6331]

i am greatly interested by your concept using the Fermat-Apollonius geometry.

do you know what the power output would be for a 4 * 250cc HCCI engine? and what the fuel consumption would be in grams / KW hour?

[orboxbe]

@Marc6331

about 800 kW from a 1 liter engine. At 50 kg a specific weight of 16 kW/kg , and about 200 g/kWh .

Any engine will have to live with the Carnot limitations, but regular engines are limited in power and efficiency largely by air restriction. If you do the calculations, the intake air actually reaches the speed of sound in the intake valve (outlet is hot, so speed of sound is higher, so it can be smaller!) at max speed of about 8,000 rpm.

[orboxbe]

@orboxbe

Average piston speed is about 25 m/s for both F1 engines and commercial (BMW/Audi) engines, with F1 having about half the stroke, and double the bore (so bigger valves!) and double the rpm and power. Max piston speed is about 40 m/s and max acceleration of the piston about 100,000 m/s2 or about 10,000 g! But F1 engines are rather expensive (I'll explain if you like). So "square" engines with about bore=stroke are the standard for commercial engines.

[orboxbe]

@orboxbe

efficiency: less restrictions, lower pumping losses, continuous airflow will increase efficiency. Keeping compression ratio just below knocking (or in HCCI mode, just at that level), allows for high efficiency. Basically we will reach Diesel efficiency (45%) versus (30%) petrol engines, but with a cheaper and simpler design, running much higher rpms. So 50% higher efficiency is not unreal. 40% (30% higher) should be a realistic goal. (200 g/kWh (80/200=40%) vs 270 g/kWh (=30%))

[orboxbe]

@orboxbe

But also thermodynamically, the shape of the (pre-)combustion chamber is important (one reason Wankel is inefficient), more confined giving less heat-loss. In a regular engine it is a pan-cake on combustion, which limits flame propagation and increases high heat-exchange to the piston head. Our design has a better surface/volume ratio.

[Marc6331]

I love the idea of your high gas flow HCCI rotary engine.

Do you have a working prototype already built? where are you based?

i am interested in developing this in belgium, could you let me know more about your shareholder program and licencing oportunities.

[orboxbe]

@Marc6331

I had the idea for a very long time, and I just had some time to make a computer model based on the geometry, just to show it can be done to confine a properly sealed variable volume in this way. I have a background in physics and mathematics, that's why. Nobody seemed to have ever used the Fermat-Apollonius geometry, and it seemed a better way than epitrochoids (Wankel). And where most "inventions" fail on sealing, lubrication and cooling, this geometry has a lot of advantages.

[orboxbe]

@Marc6331

I have little time or skills to build a prototype, so I just never came about to do that. If you would like to take up development, please feel free to do so. If you like a license, that's not a problem either. None are needed for any development anyway. My problem is that for any commerical success, a serious and costly test program has to be done and financed.

[orboxbe]

@Marc6331 I have worked in the Germany in the automotive industry for many years, and anything "not-invented-here", or not from a university, is difficult to promote. And the way large (automotive) companies are organised, they have seen too many of their own "inventions" fail, to be interested even to look around (and do claim anything useful would be invented by now).

[ollieoniel]

would the oscillatory movement of the padlles cause shaking. you might need counter weights but i dont know how you would fit them in.

[orboxbe]

@ollieoniel : thanks for the question!

The elaborate answer is at the FAQ section of the orbox website. But in short: an obvious way to cancel vibrations would be to put 2 or more units in series, much like 4,6 or 8 cylinders instead of 1 in piston engines. But it also can be done within 1 unit by deliberately unbalancing the "paddles". Some more calculation is needed here, but the general idea is in the FAQ section.

[orboxbe]

please comment if you see flaws, we did not find any.

it is simple, and has proper sealing, cooling and lubrication other designs often lack

no gears, no valves, high gas flow, high efficiency, high specific power