Numerics Technology Ltd. is developing a rotary internal combustion turbine (RICT), which will have general usage, but will be initially targeted at both MW and Micro size CHP applications. A RICT petrol / gas fired prototype of designed 1.4 MW (1,900 BHP), has been built and is undergoing testing. The engine is a constant volume pulse turbine compared to normal continuous burn constant pressure units. Features of the RICT are its potential for high output, high efficiency, high power to weight and compactness of design. Being a turbine it should be reliable. Exhaust heat recovery is not needed and the design includes a simple ‘energy top up’ assistance turbine. The unit, having one rotating part, is designed to be inexpensive to manufacture. A ‘diesel turbine’ is planned with units of a few KW’s for domestic usage also.
1. INTRODUCTION
The ‘RICT’, which is a constant volume (CV) pulse turbine, is the latest of a long history of designs aimed at making a successful working engine of this type. A pulse turbine utilises much higher pressures and instantaneous temperatures than a normal constant pressure turbine. The power density of a RICT type device can be up to ten times that of a conventional turbine with the result of a correspondingly higher power to weight ratio. Additionally the CV cycle lends itself to compression ignition, thus creating the opportunity for higher efficiency and fuel diversity as well.
This potential efficiency advantage, compact size and fuel diversity has relevance to micro energy systems and CHP applications probably more so than power to weight ratio. The problem, as early researchers found out, is one of building a practical example of such a machine that does not suffer from high gas leakage, or large thermal and aerodynamic losses. This latter issue is deeply coupled to the fundamental physics and practical design of the rotor / stator ‘core’ of the engine. Numerics Technology Ltd. was formed to develop a practical constant volume pulse turbine in order to realise the potential benefits that such an engine could bring. This paper deals with the underpinning theory, design, progress to date and the envisaged areas of application for the engine.
2. BASIC DESIGN
The engine basically has a separate compression stage and expansion stage, like a ‘normal’ gas turbine, plus a ‘toroidal shaped’ intercooler in between. This is provided to improve the volumetric efficiency of the engine. A longitudinal cross section of the engine is shown in figure 1. Referring to numbers in figure 1 the compressor (110) in the current design is two staged. The expansion stage is made up of two units the rotor / stator ‘core’ (210 & 310), which produces the bulk of the power and a large shaft based impeller (118), the ‘turbine assistance’ unit. This adds further power derived from the exhaust gases. The core of the engine is primarily air-cooled. The cooling air after passing over the intercooler (112), passes both over the stator and through the rotor. It is then mixed with the exhaust gases downstream of the core, before entering the turbine assistance unit. (Ignore other numbers in the figure, which apply to the patent).
Fig 1)
|
Figure 2 is a transverse cross-section of the rotor / stator core. The stator is made in two halves (310 & 320) that are moved by a servo controlled cam mechanism to accommodate diametric expansion of the rotor. This occurs due to centripetal forces and temperature differentials between the rotor and stator halves. Two combustion chambers (316) are positioned in the stator halves. These are fed via peripheral transfer ports at both ends of the rotor (230). The ports in turn are fed from the stator halves (314), the whole making a rotary valve. The gas expansion process is by a series of steps through increasing volume chambers in both the rotor and stator. At each expansion step the rotor receives an impulse couple. Having chambers in both the rotor and stator has the effect of increasing the number of impulses that the rotor receives by the square of the chamber count, or an ‘n2’ effect. There are two horizontally opposed expansion chamber sets, or ‘cylinder equivalents’. The gases flow radially not axially, as in conventional turbines. The additional cutouts in the centre of the rotor (236) are designed to provide air-cooling. There are also fins (334) on the stator exterior for the same purpose.
While the core units may seem complex it has proved possible to fabricate these using local sub-contractors and currently available manufacturing techniques. Neither are any very specialised materials required, e.g. ceramics or non-standard alloys.
 |
3. BASIC THEORY
The expansion process is a stepped adiabatic, or more correctly, polytropic one, that has been coined ‘cadence recursive’ expansion. The governing formula for this is:
Pn,m = Pn-1,m-1 + (Pn-1,m – Pn-1,m-1).Vmj………………………eqtn1
(Vm + Rn-m+1)j
P’s are pressure
V’s are stator volumes, R’s are rotor volumes
m is the ‘step count
n is the stator chamber number
j is the polytropic ratio of specific heats for the working fluid (gas).
This equation when plotted gives the ‘P-Step’ diagrams for the rotor and stator. The stator P-Step diagram is as shown in figure 3. A similar one for the rotor also exists. (PV diagrams tend to be ‘cramped’ at initial volume expansions and hence less clear). As normal, the area under the PV diagram gives the nett power available from the expansion process, excluding other factors (see below).
The power impulses are given by the equation:
I = CosA.K1.V2.(K2.P)1/2………………………………………………….eqtn 2
A is the angle of the tangent to the rotor periphery.
K1 and K2 are constants.
P and V are derived from equation 1.
K1 is fixed by the physical characteristics of the engine for a given chamber, including chamber nozzle area and deceleration distance. K2 is notionally constant, but is in reality a complex variable involving gas density, which both changes during the gas expansion into the chamber during an impulse and also over all subsequent expansions. The total power output for the core is the sum of all of the impulses, or:
Itotal = Σ n1.n2.I ………………………………………………………….eqtn 3
n1 is the rotor chamber count and n2 the stator chamber count.
It should be noted that none of the above equations account for heat loss, which may be simply accounted for in equation 1; or losses due to turbulence or gas leakage, which are not so simply dealt with. Apart from the three ‘primary’ equations above, there are other ‘secondary’ engine equations. These define the best ‘shape’ for the chambers and the optimum chamber sequence. Also the fundamental relationship of power output to rotor diameter and hence engine size. Plus simple factors such as ‘working ratio’, which is the expansion ratio over the compression ratio and is not fixed at one, as in a piston engine.
Dealing with each of the secondary equations in turn, it turns out that the optimum chamber ‘shape’ is simply where the chamber length in millimetres equals the volume in millilitres. This means that for an optimum series of chambers, the lengths should increase with the chamber volume. This tends to lead to a sub optimum and more complex design of rotor. Hence a fixed chamber length compromise, set by the combustion chamber volume, was chosen for the prototype.
The optimum theoretical chamber expansion sequence turns out to be a simple equation. It is however, not totally practical as it results in too many chambers needing to be accommodated, which also increases heat loss. The practical optimum is the theoretical sequence modified by one or more ‘chamber sequence transition break points’ set by the rotor periphery ‘real estate’. Conveniently this also reduces heat loss
The power output to diameter for this class of engine is defined by the relationship:
Power = K.L.R3.5……………………………………………………..…….eqtn 4.
K is a constant
L is the rotor (or engine) length and R is the rotor (or engine) radius.
This equation is defined for equal tensile strength rotor material run to the maximum rotor speed limit. Thus for an engine of 420mm diameter (the prototype) running at 30k rpm, the scale sized engine of 210mm (8.25”) diameter run at 42k rpm will produce 4.5% of the larger unit’s power (62 KW or 83 BHP).
Fig 3)
4. NOMENCLATURE
A new nomenclature to describe all types of gas turbines has been developed. This has been necessary as the RICT has many possible variants, which could become confusing without a formal description methodology. The prototype’s configuration is {C.I,T.T}. { } are the engine delinearators. ‘C’ means a compressor, ‘T’ means a turbine, I an inter-cooler. Flow sequence is left to right. Commas denote a compressor / turbine set. Full stops are dot products. Italicisation means auxiliary units and bold type means power production units. Lower case, when applied, provides supplementary information.
5. PROGRESS TO DATE
The basic engine was completed August / September 2001 after starting work in March 1999. Some eighteen subcontractors were involved. Figure 4 shows the general appearance of the prototype at this time, prior to mounting in its test frame and connection of its auxiliary equipment. The unit is approximately 650mm long (25”) by 420mm in diameter (17”). Since completion, testing has proceeded and various foreseen and unforeseen technical hurdles have had to be overcome. The most important of these has been compressor performance followed by rotor speed capability (although the prototype was not planned to be run at full design speeds). Also the proprietary engine control unit (ECU) as used in the engine, had to have detail configuration and interface issues resolved. The purpose designed engine protection unit (EPU), which provides the stator position control, various trip and other specialised functions, has been tested and brought on-line in stages. Other auxiliary units include plenum boxes, oilers and motor systems.
The engine was tested initially by running it up with an electric motor to 3,000 rpm. Fuel vapour tests were conducted prior to injection tests (the prototype is petrol fuelled). Due to the low-pressure ratio delivery from the compressor, these were unsuccessful. The engine is now fitted with an external compressor that can deliver up to 10 Bar and tests have been conducted with this. The engine exhaust note was heard for the first time during these tests. (A sound recording may be heard on the web site, ref 1).
The engine’s fuel injection system has been connected up and with combustion air provided from the external compressor, testing is about to begin again. Following tests will be with the engine’s compressor converted to four-stages. This should confirm self-sustaining operation, power output potential and most importantly, efficiency. The initial design targets are 30% for spark ignition, 40% for compression ignition.
Fig 4)
6. AREAS OF APPLICATION
Theoretically the RICT engine can be scaled up or down over a wide size range without significant change in thermal efficiency. Many application areas are possible subject to proving performance and reliability. We are aiming initially at the CHP market and related stationary power systems. We see single units capable of scaling up to well over a megawatt and down to 50 KW. There may be two or three sizes in this range. Secondly there is a potential range of from around 50 KW down to a few KW and this is the principle area for domestic and housing scheme applications, currently served by piston engines. The RICT design avoids problems identified, such as the frequent head maintenance needed with such piston engines in the domestic market fired with sulphur bearing North Sea gas.
Additionally the development of a compression ignition (diesel) variant and its potential much higher efficiency will add further financial advantage.
The speed of rotation of the RICT is not seen as an issue since there are many examples of high speed generating systems in the micro market now. Providing the RICT engine cost is kept down (which should be possible as it has few moving parts), multiple engine redundant systems with guaranteed ‘up times’ are possible. The self-sufficient, low maintenance domestic CHP plant running on oil, LPG, or gas, with no mains connection, is the obvious outcome.
Returning to the prototype, which is a designed 1.4MW unit at full rpm specification of 30k, the area of application for this unit will be in the range 200 KW to over 1 MW. The design allows for dual cores to nearly double the power with a suitable compressor. Also various injection modes, using water injection directly onto the rotor and a hybrid ‘after burning’ variant are possible, to further increase power levels. The water injection variant, or rotary internal combustion steam turbine ‘RICST’, should allow water injection up to 100% saturation levels, unlike piston engines, which can only tolerate up to 15%. This is because the combustion process is not directly involved, so no dousing will occur. In the hybrid variant ‘RICHT’, after-burners can be placed in the downstream oxygen rich cooling air stream, to add extra power output via the turbine assistance unit. These variants may be useful separately or together ‘RICHST’, for accommodating short-term peak output power requirements in electricity, CHP and other applications, (some are mentioned in 6 below), possibly even such as VTOL.
7. THE FUTURE
Once the ‘large’ petrol fuelled prototype has been fully proven, work will commence on both a scaled down ‘kilowatt’ micro unit for domestic CHP applications and collaterally compression ignition variants for both large and micro units. The water injection, hybrid and other fuel (e.g. gas) variants will also be developed. The initial target markets of micro and medium CHP will be extended to marine (requires marinisation), road and rail haulage and aeronautical (particularly diesel). The automobile market is not of immediate interest.
It is planned that the efficiency level, particularly of the compression ignition variant, will be developed to the extent that the RICT could make an important contribution to the reduction of greenhouse gases.
References:
1. Web site: www.numerics.co.uk
2. ‘The New Rotarian’ IEE Review July 2001
3. Analysis of the C-V Turbine – John Whurr, Cranfield School of Mechanical Engineering