Despite its many attractions as the best working fluid for Stirling engines, hydrogen has the disadvantage of extremely wide flanimability limits in air ranging from 5 to 75 per cent mixtures of hydrogen and air. Other gases, methane, for example, have much closer flanimability limits of 6 to 14 per cent methane in air by volume. Moreover, hydrogen has a high affinity to oxygen and the enthalpy of reaction (heat released in burning) 129000kJ/kg (30 960 Btu/lbJ of hydrogen, compares with 50 143 kJ/kg (12 034 Biu/lbJ for methane.

A uiomotive en gin as

It is therefore difficult to escape a feeling of apprehension when contemplating Stirling-cycle vehicle engines pressurized to 100 to 200 MN/'m (29 007 lbs per sq in). From the material presented above there is no doubt that the automotive engine must use hydrogen; there really is no alternative for engines ol high power density and efficiency. Vet. the prospect of high pressure hydrogen systems, generally available for public use, is a daunting prospect to those concerned with matters of safely.

It is inevitable that some leakage will occur from the working space to

. i i • . .t--- j—I.I. ——---;--------:------1 ...:.!_----1----; — i favoured by Philips it is inevitable that sooner or later material failures will occur with consequent leakage of hydrogen from the working space.


Hydrogen is. moreover, a difficult gas to contain. It is so light and fluid that it passes readily through materials that are apparently solid. At high temperatures and gas pressures the effect is enhanced and most materials including metals are porous to some degree to hydrogen. In an engine with a direct-heating fossil-fuel combustion system any leakage through the hot parts of engine pose no safety hazards; hydrogen will simply burn (to water vapour). Por indirectly heated systems with a liquid-metal heat-transfer loop, some means of venting the hydrogen gas must be incorporated.

Percival <1974) cited hydrogen permeation of materials as one of the principal unresolved problems in Stirling engines. He indicated that one avenue of approach adopted by Philips was to provide an impermeable ceramic liner in the heater tubes but provided no details and the matter has never been mentioned in any of the several papers published by Philips.

Hydrogen emhrittlement

Another important effect to be considered in the use of hydrogen as ¡1 working fluid is the embrittlement which metals experience when exposed to hydrogen, particularly at high temperatures and pressures. The effect of hydrogen on the mechanical properties of metals and other materials is extraordinarily complex and profound. An excellent compilation of material on the subject of hydrogen effects was assembled bv Beachem (1977).

These important matters have noi been sufficiently addressed in the public literature. They are expected to receive due attention in the Stirling engine research program supported by public funds now being developed in the United Stales by the Department of Energy.

General Motors Corporation's research

A significant statement regarding safety in the use of hydrogen as the working fluid in Stirling engines was contained in PercivaPs (1974) historical record of the General Motors development program. In that work Percival reproduced a memorandum to the G.M. laboratory management in which he investigated possible restrictions on the use of hydrogen working fluid in the 110 kW (150 hp) JL23 General Motors double-acting Stirling engine for bus installation. He examined the possible restrictions on public highways, and in tunnels, determined the use of hydrogen with various regulatory authorities who might have been responsible for restrictions on the use of hydrogen but no existing regulations or limitations on its use were found.

The total amount of hydrogen contained in the GM 4L23 bus engine and reservoir was estimated, with the engine pressurized to the normal value oi 10.3 MN/nr (1500 lbs per sq in), to be a total of 0.022 kg (0.050 lbs), equivalent to a free volume at 21.11 °C (70 °F) of 0.2724 m3 (9.62 cu ft) This volume of gas escaping from the engine and mixing with air to a 5 per cent mixture would need to be confined to a volume of 5.435 m' (1.52 m X 1.52 mx 2.44 m) (192 cult: 5 ft x 5 ft x 8 ft) to assure flame propagation on ignition.

Assuming ignition and the reaction of at least one half of the hydrogen the heat released would be 1370 kJ (1300 Btu). This was said by Percival to be equivalent to 0.032 kg (-¡'i lb) of gasoline or I, of the smallest available can of propane. Another factor favourable to hydrogen cited by Percival was the relatively low flammability limits for hydrocarbon gases: 2.2 per cent for propane, 1.6-2 per cent for butane, 1.4 per cent for gasoline, 5-6 per cent for methane. For hydrogen, the lower flame propagation limit ranged from 4 per cent, for combustion vertically upward, to 9 per cent, for combustion vertically downward. For hydrocarbon gases, on the other hand, the flammability limit was more or less independent of the direction of propagation so they were more likely to be entirely consumed. The extended upper flammability limit for hydrogen, 74 per cent compared with 6 to 14 per cent for hydrocarbon gases, was dismissed by Percival rather summarily as being of serious concern only in the case of non-habitable spaces. He failed to address the question of explosion of a rich hydrogen mixture in a closed engine compartment, thought by many to be a serious problem.

Percival pointed out that hydrogen diffused at a rate 4 to 8 limes that of hydrocarbon gases. In the open air or in a large room an instantaneous leakage of all the hydrogen from a 110 kW (150 hp) engine would be dissipated to a non-flammable mixture in a few seconds. On the other hand, hydrocarbon gases disperse more slowly and, being heavier than air, tend to accumulate or remain concentrated in low lying areas.

In summary, Percival presented a convincing case that the potential danger of lire or explosion from hydrogen leakage in the bus installation was small and substantially less than the possibility of lire from a hydrocarbon fuel source.

Font Motor Company's research

Further work on the hazards of hydrogen as the working fluid in automotive Stirling engines was carried out by Goodale and Walter (1976) at Stanford Research Institute on behalf of the Ford Motor

Company. Goodalc's report is contained as an appendix in the comprehensive report by Kitzner (1977a).

No other firm data is available, but it is said that United Stirling have approached the Swedish regulatory authorities and have gained approval for the use of hydrogen as the working fluid in their automotive engine. It is further understood that their development work has embraced the experimental study of lire or explosion hazards with hydrogen engines in vehicles in garages or confined spaces. No details have been published.

No doubt, all the companies presently involved are pursuing similar work and are engaged in discussions with the regulatory authorities. The development of the so-called 'hydrogen economy' in the future, particularly in the United States, will result in hydrogen becoming far more familiar than it is today. Safety measures and handling procedures will be developed that will, no doubt, do much to allay the apprehensions expressed in the early part of this section. There are no reports of accidents involving Stirling engines in the literature, except a brief reference. made almost in passing by Pcrcival (1974) in discussing experiences with an experimental GPU-3 engine. I iearsay has it that one fatal accident occurred at Philips in the early days of their work on hydrogen engines (mid-1950s) but no details of this are known.

compound working fluids


Compound working fluids in Stirling engines were investigated in a preliminary way by Walker and Agbi (1974). They assumed a compound working fluid to have two components: the gaseous carrier and the phase change component. The phase change component experienced a change of phase from liquid to vapour in moving from the cold space, through the regenerator to the hot space.

I he principal attraction of the compound working fluid was the possibility of achieving a high specific output at a moderate mean pressure level. Secondary advantages were anticipated in terms of improved heat transfer arising from the boiling and condensing processes as well as some relief in the critical problem of reciprocating seals.

Walker and Agbi (1974) carried out the preliminary study of compound working fluids by comparison of a series of idealized Schimidt-type thermodynamic cycles. The gaseous working fluid had the characteristics of air, and mixtures of air and water having some mass ratio ft - mv/ma particular to the cycle were prescribed.

It was found that for the same limits of maximum and minimum temperature (cost) and volume (size) and for the same maximum pressure (weight) the area of the work diagram for the cycle having a compound working fluid was substantially larger than that for the simple gaseous working fluid. In other words, in an engine of the same size, cost and weight, die use of a compound working fluid increased the work output.

A physical explanation for this improvement was found in the recognition that the phase change from liquid to vapour of one component of the working fluid caused, in effect, art increase in the volume compression ratio of the cycle with consequent benefit to the pressure ratio and net cycle work.

Improvement in Ihe volume compression ratio was a highly desirable-effect. In the design ol practical Stirling engines, it was rarely possible to attain a volume compression ratio greater than 2.5 without undue sacrifice ol heat transfer surface area or introducing unacceptably high lluid friction effects. By increasing the effective volume compression ratio, an increase in the amplitude of the cyclic pressure excursion was obtained with consequent beneficial results on the area of the work diagrams for the various spaces in the engine.

Isothermal analysis with compound working fluid

The Schmidt analysis (see Chapter 41 was found amenable, with relatively minor modification, for application with a two-phase, two-component working lluid.

'Hie compound working lluid was assumed to consist of two components. one behaving at all times as a perfect gas, and the other in the compression space existing in the liquid state at low temperatures and, in the expansion space in the vapour state at high temperatures where it was further assumed to behave as a perfect gas. The specified mass ratio ¡3 = tnjma was assumed to prevail throughout the system and to be unaffected by the phase change from liquid to vapour of one component.

In Ihe conventional Schmidt analysis, a linear temperature profile was assumed for the regenerator. With a compound working fluid the arbitrary assumption was made that the regenerator dead space was divided into two volumes, one maintained at temperature TE, the other at temperature Tc and of such size that the total mass of working fluid jm the regenerator was divided equally between the two spaces. At the interface it was assumed that a step change in temperature occurred, accompanied by a change in phase of one component, and that the interface moved as required to maintain the equality of mass in each regime. There is no physical justification for this assumption. It was adopted simply for the sake of computational convenience and will probably need to be improved for future work.

The total pressure p was taken to be the sum of the partial pressures of the two components and, further, to be instantaneously constant through-nut th<» i.e.

Then pvc, the partial pressure of one component in the liquid state, was taken to he always zero, and the volume of the liquid was assumed to be negligible, so that in the compression space p ~ pac.

It was recognized that the assumption of a constant mass ratio /J was probably not attainable in practice because the vapour component would tend to migrate from the hot expansion space to the cold compression space. This could perhaps be compensated by gravitational pumping, i.e. putting the cold space above the hot space. A constant mass ratio, /3, might represent one extreme limit achieved by an engine running fast. The contrasting limit achieved in an engine running slowly might occur in the case where the vapour pressure pvc = pvc has a negligible value so that all the phase change component exists in the liquid state in the compression space. In this case, ft in the expansion space is effectively zero and the results correspond to that of the cycle with a single component gaseous working fluid.

Principal assumptions

I The regenerative process was perfect.

2. The process of compression and expansion occurred isothermally at temperatures Tc and Tr., respectively.

3. Volume variations in the compression and expansion spaces occurred sinusoklally.

A. The vapour component at temperature T'c existed in the liquid slate with negligible volume and vapour pressure.

5. The vapour component at temperature TB existed in the superheated vapour state and behaved as a perfect gas according lo the ideal gas equation pv V ~ mvRvT.

6. The gaseous component behaved as a perfect gas at all times and obeyed the equation paV = mMRnT.

7. The total mass of the working fluid distributed between the compression, expansion, and dead spaces remained constant.

K. The ratio of mass of vapour component to gaseous component was constant and uniform throughout the machine.

9. The instantaneous total pressure of the working fluid was uniform throughout the working space.

10. The mass of working fluid in the dead space was divided equally into two parts, one at temperature TC; and the other at temperature rE. At the interface of these two masses there was a phase change of ihe vapour component from superheated vapour to liquid.

11. The rotational speed of the engine was constant.

1 -» c. — a.. -----v.r.^.—c— -t. -------1'---- ■ ••

variation only and equilibrium existed between components and phases at all times.

Summary of analysis

The process of development of the analysis was precisely similar to that given in Chapter 4 and yielded the following equations:

1. Volume

(a) Expansion space

(b) Compression space

(c) Dead space

2. Pressure

(a) Instantaneous pressure p 1-5

(b) Pressure ratio pmax= 1 + * P:n»n I ~ 8 where /\ - (K ~ - k " + 2kK cos a) B — K 'r K ~2S fi = A/B

Solar Stirling Engine Basics Explained

Solar Stirling Engine Basics Explained

The solar Stirling engine is progressively becoming a viable alternative to solar panels for its higher efficiency. Stirling engines might be the best way to harvest the power provided by the sun. This is an easy-to-understand explanation of how Stirling engines work, the different types, and why they are more efficient than steam engines.

Get My Free Ebook

Post a comment