Philips Technical Review - Research | Philips Bound... · 246 PHILIPS TECHNICAL REVIEW VOLUME 20 which also possesses thefeature that the crankcase need not he pressurized. This was

VOLUME 20, 1958/59, No. 9 pp. 24.5-276 Published 8th May 1959

Philips Technical ReviewDEALING ~H TEC~CAL PROBLEMS

RELATING TO THE PRODUCTS, PROCESSES AND INVESTIGATIONS OFTHE PHILIPS INDUSTRIES

THE PHILIPS HOT-GAS ENGINE WITH RHOMBIC DRIVE MECHANISM

by R. J. MEIJER. 621.412-231.312

Since the first reports on the Philips hot-gas engine, published in this Review in 1946 and1947, further research and development on the engine has taken place along two different lines.These investigations have resulted on the one hand in the construction of the cold-gasrefrigeratingmachine, now in production for some years, and on the other hand they have led to the designof a new type of drive mechanism which is of special importance for the hot-gas engine proper(particularly large machines). By the application ofthis drive mechanism it has been possibleto drop the idea of using one piston to perform both power and gas-transfer functions (the"double-acting" principle) and return to the thermodynamically more efficient system ofseparate power and transfer pistons. In this way and by the incorporation of a number of im-provements in the design of regenerator, heater etc., the engine has now been given a form thatpromises well for future development. Measurements carried out on a 40 H.P. experimentalengine built according to the new design have demonstrated that, as far as efficiency and specificpower are concerned, the new engine can compete with the best amongst familiar forms of primemover, besides possessing all the virtues inherent in the hot-gas cycle.

With the help of modern materials and with newknowledge of flow and heat-transfer phenomena,the hot-air or hot-gas cycle, which has been knownsince the early part of the last century, can be madeto take place with high efficiency. This was madeclear in a series of articles that appeared in earliervolumes of this Review 1,2,3,4,5).

Contrary to expectation, development work onthe cycle in the Philips laboratories did not in thefirst place lead to a hot-gas engine, but to a gasrefrigerating machine - which, in the meantime,has come to occupy an important place in refrig-eration practice 4,5). There are several reasons forthe slower development of the engine. The principalamong these were the practical difficulties encoun-tered when it was tried to apply what was termedthe "double-acting" or single-piston principle.

1) H. Rinia and F. K. du Prê, Air engines, Philips tech.Rev. 8, 129-136, 194·6. .

2) H. de Brey, H. Ri~a and F. L. van Weenen, Fundamentalsfor the development of the Philips air engine, Philips tech.Rev. 9, 97-104, 1947/48.

3) F. L. van Weenen, The construction ofthe Philips air engine,Philips tech. Rev. 9, 125-134, 1947/48.

4) J. W. L. Köhler and C. O. Jonkers, Fundamentals ofthe gasrefrigerating machine, Philips tech. Rev.16, 69-78, 1954/55.

5) J. W. L. Köhler and C. O. Jonkers, Construction of a gasrefrigerating machine, Philips tech. Rev. 16, 105-115,1954/55.

This was indeed an elegant principle. The basictype of hot-gas engine, the Stirling engine, hasseparate power and transfer pistons, while in our"double-acting" engine one moving body performedthc functions of both, this constituting a very con-siderable mechanical simplification. Moreover, indistinction to the small transfer-piston (or displacer-piston) engines made at the time (which workedvery well), the single-piston engines did not requirethe crankcase to be pressurized. This promised con-siderable advantages for higher-power engines wherea pressure crankcase would necessarily involve large-weight penalties. However, these advantages hadto be paid for. In the first place an exceptionallyintractable lubrication problem now arose: a pistonhad to act as a moving gas-tight seal between a hotand a cold space between which a large periodicpressure difference occurs. Also, both thermodynam-ically and aerodynamically, this type of enginewas inferior to the displacer-piston engines, owingto the fact that the volume variations of the hotand cold spaces could no longer be freely chosen asregards their relative magnitude and phase ..

With further development it has proved possibleto drop the idea of a single "double-acting" pistonand to design an engine of the displacer-piston type

246 PHILIPS TECHNICAL REVIEW VOLUME 20

which also possesses the feature that the crankcaseneed not he pressurized. This was made possibleby the embodiment of a new kind of drive mecha-nism. The latter offers the, additional advantagethat even a one-cylinder engine can be perfectlybalanced. Of the various engines built according tothis design, a single-cylinder 40 H.P. machine willbe described here and some result~ of measurementso~ the output and efficiency will he given.ro save the reader the trouble of consulting ear-

lier' volumes of this Review, we shall now verybriefly recapitulate the principles of the hot-gas

\engine, restricting ourselves to an engine providedwith 1\ displacer piston.

Brief account of the hot-gas cycle

An internal-combustion engine provides a surplusof work in virtue of the compression at low temper-ature of a certain quantity of air, to which atomizedfuel is added either before or after the compression,the subsequent heating of the mixture by rapidcombustion, and its expansion at high temperature.The hot-gas engine is based on the same principle,

i.e. the compression at low temperature and ex-pansion at high temperature of a given quantity ofgas. The heating takes place, however, in an entirelydifferent manner, the heat being supplied to the gasfrom outside, through a wall. For this reason thedescription "external-combustion engine" is appro-priate. Owing to the high thermal capacity of thewall, it is not of course possible to heat and cool thegas simply by rapid heating and cooling of the wall.Stirling had realized as far back as 1817, however,that the gas temperature could be made to changeperiodically by causing a "displacer piston" totransfer the gas back and forth between two spaces,one at a fixed high temperature and the other at afixed.low temperature - see fig. 1. If we raise thedisplacer piston in fig. 1, the gas will flow from thehot space via the heater and cooler ducts into thecold space. If now the displacer piston is moveddownwards the gas will return to the hot spacealong the same path. During the first transfer stroke.the gas has to yield up a large quantity of heat; anequal quantity of heat has to be taken up duringthe second stroke. The regenerator shown in fig. 1is inserted between the heater duct and cooler ductin order to prevent unnecessary wastage of this hcat.h is a space filled with porous material to whichthe hot gas yields heat before entering the cooler;when the gas streams back, it takes up the storedheat again prior to its entry into the heater.

'I'he displacer system, which serves to heat and

cool the gas periodically, is combined with a powerpiston that compresses the gas while it is in the coldspace and allows it to expand while in the hotspace (all dead spaces in cooler, heater etc. beingdisregarded). Since compression takes place at a

---- Hot space

---- Displacer

Cold space

96937

Fig. 1. Principle of the displacer-piston system. Moving thispiston up and down causes the gas to be transferred back andforth between the hot and cold spaces, via heater, regeneratorand cooler.

lower temperature than expansion, a surplus ofwork results. Fig. 2 shows four phases of the cyclethrough which the whole system passes if a discon-tinuous movement of power piston and displacerpiston is presupposed. The displacements they areassumed to undergo are plotted as functions of timeinfig. 3; the ordinates in band E represent the varia-tion in the volume of the hot space, and those inband C the variation in the volume of the cold space.The volume variations are plotted separately in thelower part ofthe diagram. Fig. 4 is the p, V diagramof the cycle (V is the total volume of the gas).

In a practical version of the engine the movementsof power and displacer pistons must of course becontinuous, not discontinuous, as they have beenassumed to be in these figures; the continuous move-ments will be obtained with the aid of some kind ofcrank and connecting rod mechanism. It will notthen be possible to distinguish any sharp transitionsbetween the four phases, but this will not alter theprinciple of the cycle (or detract from its efficiency- see below). The movements of power piston anddisplacer might now be as indicated in fig. 5, in whichthe volume variations of the cold and hot spaceshave again been plotted separately. The only es-sential condition for obtaining a surplus of work isthat the volume variation of the hot space should

"

HOT-GAS ENGINE WITH RHOMBIC DRIVE MEcIÄ~f~llIPS' GLO£IU ,~~[2~tt t f.1958/59, No. 9

I IJ 1II N 96938

Fig. 2. Diagrams to illustrate the hot-gas cycle. For the sake of clarity the power pistonand the displacer piston are supposed to move discontinuously; it is then possible to dis-tinguish four phases that make up the complete cycle.I Power piston is in lowest, displacer piston in highest position. All the gas is in the

cold space.II The dispJacer piston is still in the highest position. The power piston has compressed

the gas while i t is at low temperature.I I I The power piston is still in its highest position. The displacer piston has pushed the gas

through the cooler, regenerator and heater into the hot space.IV The hot gas has expanded, and the power and displacer pistons have together returned

to the lowest position. While the power piston remains there, the displacer pistonwill now push the gas through heater, regenerator and cooler into the cold space,whereupon situation I will have been restored.

have a phase lead with respect to that of the coldspace; this is equivalent to requiring that the ap-propriate p, V diagram, shown in Jig. 6, should betraced out in the clockwise direction.

(1--

'='-- III

p

t

96940 -vFig. 4. The p,V diagram of the hot-gas cycle represented byfig. 3.

For the subsequent discussion of specific powerand efficiency it will be necessary to recall the moreimportant of the formulae worked out in the articlescited. They are based on the assuniption that VE,

the volume of the hot (expansion) space, and Vc,the volume of the cold (compression) space, varywith the crank angle a in a purely sinusoidal fashion:

96939

-tFig. 3. The discontinuous movements of power piston (Z)and displacer piston (V) assumed in fig. 2, plotted as functionsof the time. Band E represents the volume variatons of thehot space, band C those of the cold space. These variations areplotted separately lower in the diagram.


IjlHt--

96941I

(;(=0 _(;(

Fig. 5. As for fig. 3, except that now the piston and displacermovements are continuous (the abscissa represents the crankangle a). It is no longer possihle to distinguish sharp transitionsbetween the phases of the cycle.

VE = ~Vo(l + cos a) , (

Vc = ~wVo[1 + cos (a-tp)].)

The crank angle a (for constant angular velocity w,proportional to the time: a = wt) is measured fromthe position at which VE has its maximum valueVo; tp is the phase angle between the volume varia-tions of the hot and cold spaces, and w is the ratiobetween their maximum volumes. In reality thevariations in VE and Vc will certainly not be purelysinusoidal, but the calculations show that the effectof the higher harmonics can generally be neglected.A further assumption underlies the derivation of theformulae, namely that the cycle is "ideal", charac-terized by constant gas temperatures TE and Tcin the hot and cold spaces respectively, and negli-gible flow and other losses (to which we shall re-turn later on). Another important quantity intro-duced is the gas temperature ratio,

(i < 1).

The condition that the mass of the working fluidremains constant throughout the cycle now leadsto the formula for the pressure p as a function of

the crank angle:

l-ö(3)p = pmax 1+ Ö cos (a- e)'

in which pmax is the maximum pressure occurringduring the cycle and

1/ i2 + w2 + 2rw cos tpÖ =----------------~

r + w + 2s(4)

w sin tptan e = --------

i + w cos tp

(s is the relative volume of the "reduced dead space").From the above we obtain the average pressure,

(5)

l/l=bp = pmax V 1+ ö '

and the power output

(6)

ÖP=~wVop(l-i) sine. (7)

1+ 11-ö2

We are concerned here with a reversible cyclicprocess in which, in accordance with the "idealized"conditions assumed (isothermal behaviour in coldand hot spaces and 100% efficient regenerator ac-tion), the supply of heat takes place at only onetemperature TE and the removal of heat at only

(1)p

r

96942__ v

Fig. 6. The p, V diagram of the cycle represented by fig. 5.

one temperature Tc. There is a theorem in thermo-dynamics which states that under these conditionsthe efficiency with which heat is converted intowork (the thermal efficiency) is that of the Carnotcycle:

(8)

(2)From this we may obtain the quantity of heatsupplied per second,

P Ö

qE = -:;;= ~w Vo P 1+ Y1- ö2 sin e. (9)

1958/59, No. 9 HOT-GAS ENGINE WITH RHOMBIC DRIVE MECHANISM 249

This recapitulation will suffice for our presentpurposes. We shall now go on to draw certain con-clusions which have been essential in deciding themanner in which the cycle must be made to takeplace in a practical hot-gas engine.

Specific power, losses and overall efficiency

The main considerations governing the design ofan engine will in general be to see that its specificpower and overall efficiency are as high as possible(consistent, of course, with certain conditions, suchas a reasonable working life). The former term refersto the power output per unit volume swept outby the power piston. Overall efficiency is definedas the ratio of work performed per unit time dividedby the heat content of the fuel consumed per unittime. These two aims are in some respects contra-dictory.We see from formula (7) that the output of a given

engine (for which certain values of the parametersw, cp and s have been decided upon) can be steppedup by increasing the mean pressure p and the shaftrevolutions n (angular velocity w) and by reducingthe ratio 7:. A reduction in 7: also brings about adirect improvement in the efficiency; formula (8)shows that, theoretically, 1] is dependent on 7: alone.The coolant temperature, which determines Tc,should therefore be made as low as possible, andthe temperature TE at which heat is supplied tothe heater element should be made as high as pos-sible, i.e. as high as the properties of the heaterwall material (e.g. resistance to creep and oxidation)will permit. To find out the effect on the efficiency,if any, of the other two factors, pand n, we shallhave to pay attention to the various kinds of losses,which are not 'taken account of the in formulae.The following are the most important 6).a) Losses due to mechanical friction in the moving

parts.b) "Adiabatic losses". Within the spaces whose

volumes vary, expansion and compression takeplace more or less adiabatically, not isothermally,with the result that TEn' the mean temperaturein the hot space, will be lower than TE, the heatertemperature and that at which heat is supplied.Similarly, Tcn in the cold space is higher thanTc. Internally, therefore, the engine has a tem-perature ratio 7:a = Tèn/TEn which is greaterthan the external 7:, the implication being a lowerefficiency.

6) Losses (a) to (e) inclusive were discussed at rathergreater length with reference to the cold-gasrefrigeratingmachine 6), in which they likewise occur; the losses under(f) are of course absent in the refrigerating machine.

c) Fluid friction or flow losses, representing thework done in forcing the gas to flow back andforth through cooler, regenerator and heater.

d) Regenerator losses. Since in practice a regenera-tor can never be 100% effective, the temperatureof the gas leaving it is too high on entry into thecooler and too low on entry into the heater. Thusan additional amount of heat is removed in thecooler, and lost to the engine.

e) Heat-transfer losses at the walls of the elementsresponsible for heat exchange between the work-ing fluid on the one hand and the coolant andcombustion gases on the other. As with the"adiabatic losses", these losses have the conse-quence that the ratio 1: is effectively increased.

f) Flue losses, which are inevitable in any kind ofburner. The hot burning gases can yield up heatto the heater body only while their temperatureis yet above that of the latter. When they havecooled to this temperature the remaining heatcontent of the gases will- if no special measuresare taken - be lost via the flue. These flue lossescan, however, be considerably reduced by furthercooling the flue gases by heat exchange with theincoming fresh air, which is consequently pre-heated.Now, closer consideration of the influence of ]i

and n on the losses shows that the former factorhas little effect on their' relative magnitude 7).Hence it is always advantageous to make p as highas possible; the experimental engine that will bedescribed below was designed for an average pres-sure p of up to about 100 kg/cm2• On the other hand,an increase in the engine speed does affect efficiencyadversely, because the flow losses increase with 'n.

at more than a strictly proportional rate. Theselosses, however, do 'not only depend on the enginespeed (and, of course, on the dimensioning of theducts); they differ considerably according to the gasused as the working fluid. Since the hot-gas cycletakes place within a closed system, the designer hasa free choice as to the gas so used. We decided onhydrogen as the working fluid for the new engine.Owing to its low density, flow losses are, not exces-sive even at considerable engine speeds. Moreover,with hydrogen good heat-transfer characteristicsare attainable. In the machine -described here thenominal running speed is 1500 r.p.m.

7) However, the relative magnitude of the losses (e) doesdepend somewhat on the power developed by the engineand therefore also on p (and n too): the greater the rateat which it is desired to pass heat through a given areaof heat-exchanger wall, the greater will be the temperaturedrop across the wall, and hence the smaller is the internaltemperature difference, i.e. the greater is the ratio •.

250 PHILlPS TECHNICAL REVIEW VOLUME 20

Helium mayalso be used as the working fluidinstead of hydrogen. The overall efficiency is thenfound to bé somewhat lower.The further details of the engine we shall now

proceed to describe are mainly concerned with thenew drive mechanism to which reference has al-ready been made, the design of regenerator, heater,cooler and air pre-heater and finally a completecontrol system for reguJating the power of theengine.

The drive mechanism of the displacer-piston machine

Earlier versions

There are various ways of making the power anddisplacer pistons perform the required movements.Fig. 7 shows a drive mechanism of a kind that usedto be embodied in small hot-gas engines. The powerpiston is linked to the crankshaft K by a connectingrod (con-rod) D in the normal way; the movementof the displacer is derived from that of the powerpiston with the aid of the rocker T, which is linkedby arm SI to the displacer-piston rod V and byarm S2 to a certain point on the power-pistoncon-rod. The latter has to be forked in order toleave free passage for the arm SI and the displacer-piston rod, which passes centrally through the powerpiston. A different form of drive mechanism has

Fig. 7. Drive mechanism of the kind formerly used in smallhot-gas engines. D = forked piston con-rod. K = crankshaft.T = rocker linked hy arms SI and S2 to the displacer-pistonrod V and to a certain point on D.

v

Fig. 8. Drive mechanism used in the gas refrigerating machine.The piston has two parallel con-rodsDl and D2 which link itto cranks Kl and K2; between the latter is a third crank K',which con-rod V' links to the displacer-piston rod V.

been adopted for the Philips gas refrigeratingmachine - see fig. 8. There are three cranks in theshaft, and the two outside ones, which lie at thesame angle, are linked to the piston by two parallelcon-rods; the displacer piston is actuated from themiddle crank, which is offset by a certain anglefrom the other two.

We alluded in the introduetion to the. drawbackof mechanisms such as those of figs. 7 and 8, espe-cially for large engines: the crankcase has to hefilled with the working gas under high pressure.This is necessary for two reasons: first; to limit theleakage of gas past the piston; second, to precludeunnecessarily large downward forces on the drivemechanism. In the absence of any compensatingpressure (buffer pressure) in the crankcase there isa difference of pressure across the piston fluctuatingbetween pmin and pmax, the minimum and maximumpressures occurring within the cylinder during thecycle. Introducing. a buffer pressure of ~alue pmin,for example, reduces the maximum pressure dif-ference across the piston to pmax - pmin.It is clear that with such a pressurized crankcase

considerable weight penalties are involved. Particu-larly troublesome in large engines, this drawbackbecomes even more of an embarrassment if it is

(

1958/59, No. 9 HOT-GAS ENGINE WITH RI-IOMBIC DRIVE MECHANISM 251

desired to operate at very high pressures in orderto increase the specific power in the way explainedabove.

The new drive mechanism (rhombic drive)

The above-mentioned drawbacks do not arrsein the new drive, which we now propose to describe.Briefly, it comprises twin cranks and con-rodmechanisms, identical in design and offset from thecentral axis of the engine; the cranks rotate in op-posite senses and are coupled by two gearwheels.

Fig. 9 is a schematic diagram of the system. Fixedto the power piston 1 by way of piston rod 2 is ayoke 3. One end of the yoke is linked by con-rod 4to crank 5, the other end by con-rod 4' to crank 5'.The displacer piston is actuated by a precisely simi-lar arrangement: the displacer-piston rod 7, whichpasses through the hollow rod 2, has fixed to it ayoke 8 which is linked to cranks 5 and 5' by con-rods 9 and 9' respectively. If 9 and 9' are given thesame length as 4 and 4', the two pairs of con-rodswill form a rhombus, only the angles of which vary

71- __

73 --_12 - __

10 _

96945

Fig. 9. Schematic diagram of rhombic drive mechanism. 1 =power piston. 6 = displacer piston. 5_5' = cranks in two shaftsrotating in opposite senses and coupled by gears 10-10'.4_4' = con-rods pivoted from ends of yoke 3 fixed to the hollowpower-piston rod 2. 9_9' = con-rods pivoted from ends ofyoke 8 fixed to displacer-piston rod, which runs through thehollow power-piston rod. 11 and 12 = gas-tight stuffing-boxes.13 = buffer space containing gas at high buffer pressure.

Fig. 10. Model of an actual rhombic drive mechanism, in twopositions (cranks and gearwheels not present in this model).

when the system is in motion; it is for that reasonthat we have adopted the name "rhombic drive".Gearwheels 10-1 0' ensure exact symmetry of thesystem at all times. The two crankshafts being gear-ed together, the entire shaft output can be takenoff either.

Fig. 10 shows a practical form of such a crankmechanism. The motion is illustrated by photo-graphing the mechanism in two different positions(the cranks and the gearwheels are absent in thismodel).The symmetry of the system and the coaxial

arrangement of power-piston and displacer-pistonrods make it an easy matter to avoid putting thecrankcase under high pressure. The stuffing-box 11for the displacer-piston rod is inside the hollowpower-piston rod. One more seal, namely the stuff-ing-box round the power-piston rod (12 in fig. 9),is all that is necessary to form a comparatively smallcylindrical chamber 13 under the piston, separatefrom the crankcase; this "buffer space" can befilled with gas at the desired buffer pressure. Thesuccess of this arrangement depends essentially onthe fact that the power-piston rod stuffing-box issubject to no lateral thrust, as the horizontal com-ponents of the forces exerted byeach pair of con-rodsare exactly balanced at each yoke. (That they do so,and that frictional losses are low in consequence,has the further advantage of enhancing the mechan-ical efficiency of the system.) The minimum per-missible volume of the buffer space is determined


only by the range within which it is desired to limitthe pressure variations inside the chamber. In amulti-cylinder engine the buffer chambers can beinterconnected; this allows the volume of the indi-vidual spaces to be made even smaller.

In principle a space of this kind can be walled off under thepiston in engines embodying the crank systems shown in figs. 7and 8, but it is then necessary to have the rods running in acrosshead guide in order to prevent their gas-tight stuffing-boxes being acted upon by the horizontal components of theconnecting-rod forces. This leads to considerable structuralcomplications.

The power piston and displacer movements result-ing from the new drive are displayed graphically infig. 11, in the same way as in figs. 3 and 5. It willbe seen that if the direction of rotation is as indi-cated in fig.9, the volume variation of the hot space

_C( 96946

Fig. 11. Curves similar to those in figs. 3 and 5, showing thedisplacements undergone by the power and displacer pistonsin a rhombic drive system such as that of fig. 9.

will have a phase lead with respect to that of thecold space, as is required. The power and displacerpistons do not by any means move in simple har-monic motion, but it is found that a very goodversion of the Stirling cycle is obtainable. Thismight seem surprising, for on the face of it the useof the rhombic drive may appear to have restrictedthe freedom of choice with regard to parameters wand q; (the amplitude ratio and relative phase ofthe variations in volumes VE and Vc - see above),which play a large part in the design of a hot-gasengine. However, more exact analysis of the newdrive shows that it is in fact possible, by alteringthe offset of the crankshafts, the proportions ofcranks and con-rods and the ratio of power anddisplacer piston diameters, to vary these parametersover quite a wide range.

The balancing of the drive mechanismThe rhombic drive has the important property of

allowing the designer of an engine even with onlyone cylinder to obtain complete dynamic balancingof the forces due to the inertia of moving parts, andof the moments of these forces - "complete" in thesense that the fundamentals and all the higherharmonics are balanced. We shall demonstrate thisfor the simple case where the con-rods are of equallength tas in fig. 9).The configuration is given infig. 12. On grounds of

symmetry it is clear that the sum of all inertialforces acting horizontally is zero at any given in-stant. The same applies to all inertial-force momentsabout an axis perpendicular to the plane of the draw-ing. Hence all we need consider are the inertialforces acting in the vertical direction. We begin byresolving the circular motion of each crankpinT- T' into a vertical and a horizontal component.The vertical movements that the power and dis-placer pistons make as a result of the horizontalmovement of the crankpins are equal and oppositebecause the con-rods are of equallength (symmetri-cal deformation of the "rhombus"). Hence, if themasses of the two pistons (including the rods etc.attached thereto) are made equal:

(10)

the sum of the vertical inertial forces correspondingto horizontal crankpin movement is always zero.There remain the vertical movements of power anddisplacer pistons as a result of the vertical compo-nent of the crankpin movement. These movementsare exactly equal and, moreover, identical with thevertical movement of the crankpins. One can there-fore imagine the moving mass of the two pistons,

Fig. 12. Configuration of rhombic drive in which yokes 3 and 8are of equal length, and all four con-rods 4-4' and 9-9' areequal. For discussing the inertial forces, the circular motionof crankpins T-T' can be resolved into vertical and horizontalcomponents.

\

't


mz + mv, as localized in the crankpins, t(mz +mv)in each. By means of two counterweights mc mount-ed opposite each crankpin at a suitable radius rc,vertical inertial forces can be created which exactlybalance those of (mz + mv) - and this is true evenif the rotation of the crankshafts is not exactlyuniform ..

In fact, the mass and offset of each counterweightare chosen such that the latter also serves to balancethe crank and other eccentrically positioned movingparts. If mk is the effective mass of all these parts,and rk the distance of their centre of gravity from .the crankshaft axis, we have the following con-dition for balancing:

rcmc = lr(mz + mv) + rkmk. (11)

From the foregoing it will be clear that the reasonwhy it is possible in this way to balance the inertialforces in a displacer type of engine is that, even forsingle-cylinder machines, there are two reciprocatingmasses whose movements differ in phase 8). Inorder to satisfy the condition mz = mv it willgenerally be necessary to make the masses of thetwo pistons themselves different. This should notinvolve any difficulty, for the two bodies havedifferent shapes and different functions' and are notinterchangeable.

The configuration of fig. 12 is a special case of a muchmore general form of the ~ew drive. We obtain the more generalform by abandoning the restrietion that the pairs of connectingrods are equal in length, and by allowing each displacer-pistoncon-rod to hinge not on the crankpin but about any arbitraryfixed point on the power-piston con-rod (see fig. 13). In thisgeneral form, which design considerations may render moreattractive than the special one of figs. 9 and 11, the powerand displacer piston rods continue to move coaxially withoutthe presence of any lateral thrust. This means that the expe-dient of the buffer chamber with gas under pressure can againbe applied without complications. Moreover - and perhaps.surprisingly - it can be shown that the inertial forces of themoving parts of the general form of this drive mechanism canbe completely balanced as before, provided certain straight-forward conditions are fulfilled. We shall state these conditionshere without proof. For balancing to be possible the parametersof th~ mechanism as indicated in fig. 13 must satisfy theequations:

LIe = 0,

c = 111-21 cos fJ+ (Ira

ml-m4 +ï (m3 + 2m4) cos fJ = o.l. . (12)

The net inertial force can then be completely balanced by

8) It is clear that the rhombic drive could also be used forattaining complete dynamic balancing in internal-com-bustion engines, but it would probably be a propositiononly in engineswith at least two cylinders, i.e. in which atleast two pistons are in reciprocating motion,

fitting counterweights whosepositions and masses are given by

. aï (m3+2m4) sin fJy=n-tan-l----~-------

ml+m2+m3+m'I+~mÓ rand

I~---'------~'---~~--~~--"ma= ~ 11 (mt+máma+má~mór +H (ma+2m4) sinfJr

. (13)

respectively, It will be seen that when a is zero, as it is in thespecial case first discussed, these conditions transform intothe simple ones given earlier.The particular version of the hot-gas engine with which we

shall concern ourselves in this article has the simple form ofcrank mechanism shown in fig. 9.

In addition to inertia forces, an olternating moment actson the foundation of every piston engine when under loadas a consequence of the periodic variation of the torqueproduced on the shaft. With the rhombic drive mechanism(either the special or the general form) a relatively simpleexpedient makes it possible to balance out a large part ofthis alternating moment. It is only necessary to fit twoflywheels instead of 'the usualone, one on each of the twooppositely-rotating crankshafts.

.-+-- '<,-: I -,/ . .. I! .~.:._._.

\.

96948

Fig. 13.General form of the newdrive: mi... ma are the massesof the moving parts, imagined localized at various points forthe purpose of deriving the conditions for the balancing ofinertial forces and moments.

t..

Further constructional details of the engine

The general appearance and layout of the experi-mental engine may be seen in jig. 14..Fig. 15 is across-sectional sketch, burner and pre-heat er notbeing shown; for the sake of clarity some of theparts have not been drawn exactly to scale. Thenotation of the drive mechanism is the same as inpreceding diagrams; the counterweights are alsoshown (14.).


Fig. 14. Experimental model of the Philips hot-gas engine withrhombic drive, in position on the test bench.

As in previous designs described in this Review 3),the cooler, regenerator and heater all have an annu-lar configuration, and are mounted around theexpansion-compression cylinder. This arrangementmakes for compactness, and low gas-flow losses.The displacer piston consists of the piston base 6aand a thermally insulating dome 6b. The piston basefits like an ordinary piston into the cylinder 16at the level of the cooler. There is a small clearancebetween the dome 6b and the cylinder wall 15 ofthe hot space, just sufficient that the two nevercome into contact. Underneath the power pistonthe buffer chamber 13 can be seen.The design of the regenerating casing has under-

gone a good deal of modification. In the first versionof the experirnerrtal engine the regenerator casingwas a comparatively flat annular body (jig. 16);the cylinder wall15 formed the inner wall, the outerwall a being joined to the cylinder via perforatedflange b. Considerable stresses were set up in thewalls of the casing owing to expansion of the upperside as its temperature rises to the high value TE,the lower side remaining at the low temperature Tc.Efforts to step up the output of the hot-gas enginewere soon brought to a halt by the thermal stresses

ansmg m this way, these becoming excessive es-pecially as the working pressure of the engine wasraised. The difficulty has been avoided by dividingup the annular regenerator space into a number of

Q6Q4Q

Fig. 15. Cross-section of the engine shown in fig. 14, muchsimplified and, for the sake of clarity, not entirely drawn toscale. The burner and air pre-heater are absent; they are shownseparately in fig. 18. Spaces filled with hydrogen are shaded.1 = power piston. 6a and 6b = displacer piston and insulatingdome. 12 = power-piston stuffing-box. 13 = buffer chamber.14 = counterweights. 15, 16, 17 = cylinder in which powerpiston and displacer piston move. 18 = regenerator compart-ment. 19 = cooler compartment, 20, 21, 22 = heater tuhes.23 = fins. 26 = tube for temperatnre probe. Other symbolsas in fig. 9.


small chambers ("cups") not contiguous with thecylinder wall; see fig. 17. Thermal stresses in theregenerator wall then ceased to be a problem. Fur-thermore, it was now possible to ensure that thelines of principle stress ran substantially straightfrom the head of the cylinder 15 via the cylinder

Fig. 16. Earlier design of regenerator compartment, with cy-lindrical outer wall a attached to cylinder wall 15 by meansof perforated flange b.

wall 16, 17 to the base of the latter cylinder (seefig. 15). In earlier designs embodying the flangementioned above, these lines tended to curve out-wards, which meant that the flange had to withstanda strong bending moment. The new design allowsthe whole assembly to be kept comparatively lightdespite the high maximum pressure of the gas.The design of the cooler has been adapted to that

of the regenerator. The cooler housing 19 containsa number of tubes bunched together in groups(cooler "units"), the end of each unit protrudinginto one of the regenerator cups. The tubes of eachcooler unit are mounted with a sliding fit in thecooler housing; in this way the regenerator cups areleft free to undergo smalllongitudinal displacementsconsequent on the expansion of the heater tubes.The heater consists of two sets of pipes, 20 and 21,

in close-packed configuration but with small clear-ances. The hot gases from the burner are blownbetween these tubes in a direction perpendicularto their length. The tubes of set 20 terminate insidethe regenerator cups (see figs. 15 and 17); gas fromthe regenerator streams upward through these tubesinto the annulus 22, whence it is carried downwardsby the tubes of set 21, which are brazed into thehot end 15 of the cylinder. To improve the heattransfer between the combustion gases and the work-ing gas, a series of fins 23 are brazed to the pipesof set 20.The heater is surrounded by the air pre-heeter

(27 in fig. IBa), which functions in the followingway. The air necessary for combustion enters the

Fig. 17. New design of regenerator compartment and heater.The regenerator (below) is now split up into a series of cupssurrounding but not touching the displacer cylinder. Thenew design solved the thermal-stress problem. It will be seenthat three tubes issue from each regenerator cup, their exten-sions forming the vertical heater tubes that carry the upward-streaming gas. Alternating with these, and connected to themvia the annulus at the top, are downward-flow tubes that leadto the head of the cylinder. The lower part of this tube systemis fitted with fins in order to improve the hea t transfer fromthe combustion gases to the tubes.

lower end of channels Ai situated around theouter perimeter of the air pre-heater (fig. 18b)and then flows via the spaces A between thespirally curved partitions into the channels A2

around the inner perimeter and thence, upwards,to the burner. The exhaust gases, which leavethe burner with a temperature somewhat higherthan that of the heater, enter channels Biaround the inner perimeter and flow via B, asecond set of spiral spaces lying between the Aspaces, in a direction opposite to that of the in-coming air, to the channels B2 around the outerperimeter of the air pre-heater. It has proved pos-


27-

a

b96951

sible on this principle to build a very compactpre-heater (fig. 18c) with low resistance to airflow. Moreover, the pre-heater "insulates itself";all the channels around its outer perimeter (AI andB2) are near the ambient temperature.

The burner (24 in fig. 18a) is of the "swirl-cham-ber" type, which can be used in conjunction witha large number of fuels possessing widely differingproperties, such as coal gas, propane, butane,kerosene, diesel oil and light fuel oil. If liquid fuelsare to be used, the burner is fitted with an atomizer25, which sends a fine spray of the liquid into theupper part of the burner.

Regulation of engine output

It will be seen from formula (7) that the poweroutput of the hot-gas engine can be controlled veryconveniently by altering p, the mean pressure ofthe working gas; changing p has no effect on the

cFig. 18. a) Cylinder head of engine, surmounted by burner 24,atomizer 25 and surrounded by air pre-beater 27. Other sym-bols as in fig. 15.b) Cross-section through air pre-beater, showing the narrowA and B spaces which lie between the spirally curved parti-tions and tbrough which pass, in opposite directions, freshair for and spent gases from the burner. The air enters viathe lower end of the channels Al and leaves via the upper endof A2' both sets of channels running parallel to the cylinderaxis. Similarly the burnt gases enter via the lower end ofchannels BI and leave via the upper end of channels B2' like-wise parallel to the cylinder axis.c) View of air pre-heater, which fits over the heater shownin fig. 17.

thermal efficiency (eq. 8) of the engine, providedthe cooler anel heater temperatures remain un-altered. To fulfil this proviso a separate automaticregulator is required. (The overall efficiency is infact somewhat affecteel by a pressure change be-cause, in addition to the effect mentioned under 7),the absolute magnitude of some -less important-losses remains practically constant; the effect canbe deduced from fig. 21 below.)

Accordingly, the control system of the hot-gasengine falls into two parts:1) A system for admitting gas to and withdrawing

it from the engine in order to adjust the pressureto give the output required (i.e. the brakingtorque), while maintaining the (pre-set) speedof the engine constant.

2) A system for controlling the fuel supply to theburner in such a way that the temperature ofthe heater remains steady when the power out-put (anel hence the rate at which heat has to be

1958/59, No. 9 HOT.GAS ENGINE WITH RHOMBIC DRIVE MECHANISM 257

supplied) is changed. For this purpose, controlaction is exercised by a thermostat, the tempera·ture-sensitive element of which is inserted, forexample, in the stream of gas through one of theheater pipes (through tube 26 in fig. IS; in theheater of fig. 17 two such tubes are incorporated).The control system for regulating the pressure

will be discussed here.Regulation of the pressure in the engine described

here is effected by a set of hydraulically operatedvalves. A diagram of the system is given in fig. 19.The governor 1, which is driven by the engine shaft,ensures that, at the nominal speed to which the

governor causes the oil pressure in pipes 2 and 3to drop, with the result that valve 9 of the regulator10 opens up, whereby allowing gas from the cylinderto flow via non-return valve 11 to the inlet of a smallauxiliary compressor 12, whence it passes into thetank 6. The release of gas from the engine continuesuntil the nominal speed is attained (whereupon valve9 closes).As stated above, the buffer space 13 under the

piston (see also fig. IS) is filled with gas at a bufferpressure; this pressure must alsoundergo adjustmentin accordance with changes in the load. The engineunder discussion has been designed for a buffer pres·

2

Fig. 19. Hydraulic mechanism for regulating the output of the hot-gas engine. By alteringthe oil pressure in pipes 2 and 3, governor 1 controls the operation of valves 4 and 9. If theengine speed falls below the pre-set value, gas from the tank 6 flows via valves 4 and 7into the cylinder 8; if the speed setting is exceeded, gas passes from 8 via valves 11 and 9and compressor 12 back to 6. The function of the slide throttle 16, which connects the cy·linder with the buffer space 13, is to ensure a rapid response when the load is reducedsuddenly.

6

engine has been adjusted to run, a certain oil pres-sure is maintained in pipes 2 and 3. If the engineslows down slightly in consequence of an increasein the braking torque, the governor raises the oilpressure in those pipes, with the result that valve4 in the feed device 5 opens up, and gas from tank6 passes via feed valve 7 into the space 8 whichrepresents the cylinder of the hot-gas engine. Pres-sure inside the engine having risen accordingly,its power output increases. The injection of addi-tional gas continues until the engine speed has re-verted to the original setting (whereupon valve 4closes). If the speed increases slightly in corise-quence of a decrease in the braking torque, the

38

7

74

96952

sure equal to p, the mean pressure in the cylinder.ln principle this buffer pressure is obtained auto-matically, by means of a capillary tube connectingthe buffer space with the cylinder. We shall returnpresently to other provisions made for this purpose.The auxiliary compressor gives a gas-tank pres-

sure rather higher than the minimum pressurein the cylinder under conditions of full load. Forincreasing the engine output, therefore, it is possi-ble to inject additional gas via valve 7 until thefull power is developed. This adjustment takesplace very rapidly. When it is a question of reducingthe output, for example from full power, theauxiliary compressor initially performs no function

[" '


because gas from the cylinder returns to the t~nkvia valves 11 and 9 under the excess pressure inthe cylinder; this continues until the maximumpressure in the cylinder has dropped to the tank-pressure value. This too happens very quickly. If,however, the cylinder pressure is to be lowered stillfurther, then the auxiliary compressor must function,and thereafter it is this that determines the speedof response. If a sudden decrease in the' load (e.g.sudden removal of the whole braking torque) is notto result in racing of the engine, the compressor re-quired would have to be very large. This difficultyhas be'en overcome by incorporating an additionalregulating mechanism. Besides opening and closingvalve 9, regulator 10 actuates a slide throttle 16which offers the gas a direct path between cylinderand buffer chamber; the escape of gas will be largeor small in accordance with the magnitude of thechange in the load. It is clear that this gas escapepath is equivalent to a leak past the power piston:it causes a certain decrease of the useful power de-livered by the engine (which is now indeed associat-ed with a decrease in efficiency; whence the term"loss-regulation", coined for this control). Becausethe action of the slide throttle has an almost in-stantaneous effect, a very small auxiliary compressoris adequate 9).

In consequence of the direct connection betweencylinder and buffer space when the power of theengine is being reduced, the buffer pressure likewiseadjusts itself almost immediately to the correct value.This quick readjustment is highly desirable, sinceit limits unnecessary asymmetrical thrust on thedrive mechanism. A similar purpose is served byvalve 15. It prevents the buffer pressure laggingtoo much when output is suddenly raised to a muchhigher value: if the value from which it is raisedis so low that the mean pressure in the cylinder isless than the tank pressure, gas from the tank willpass directly into the buffer chamber via valves 4-and 15.

Engine data and test results

We now give the more important data relating toa single-cylinder hot-gas engine designed and builtin Philips Research Laboratories, Eindhoven, alongthe lines described above.

D) In a multi-cylinder engine we arrange for the leakage totake place in a rather different manner: the slide throttleopens a gas path from one cylinder to the next. This isequally effectivebecause neighbouring cylinders must inany case differ in phase. Thus the one throttle providesinstantaneous "loss regulation" for two cylinders.

.Cylinder bore 88mm

Power-piston stroke 60 mm

Nominal engine speed 1500 r.p.m.

Mean piston velocity at1500 r.p.m. . . . . . 3 m/sec

Highest permissible valueof pmax . . . . 140 kg/cm2 (",-,140atm)

Mean pressure p corre-sponding to the abovevalue of Pmax 105 kgfcm2 (",-,lOS atm)

Compression ratioPmaxfPmin. . . 2.0

Nominal cooling-watertemperature . . . . . 15°C ("",60 OF)

Nominal heater tempera-ture 700°C (",-,1290OF)

Hydrogen

Light fuel oil

Working Huid

Fuel ...

We have already alluded to the advantages ofhydrogen as the working Huid; the How losses arerelatively small and good heat transfer is attainable,the efficiency being thereby greatly enhanced. Inaddition, the speed of response of the engine tochanges in the load or controls, which is ultimatelydetermined by the effective cross-sectional areasof ducts and valves in the regulating system, isimproved by using a low-density gas.

The results of measurements of the brake horse-power and specific fuel consumption of the enginewill now be dealt with. The efficiencies derived fromthese figures relate to the engine without auxiliarydevices 10) and are based on a fuel of calorific value(low heat value) 10000 kcal/kg (~18000 B.T.U.flb).The measurements were made at four differentpressure levels corresponding to Pmax values of50, 80, no and 140 kgfcm2, and at six differentengine speeds for each of these pressure levels, na-mely at 250, 500, 1000, 1500, 2000 and 2500 r.p.m.Fig. 20 is a general view of the equipment used forthe measurements. In jig. 21 the measured powerand overall efficiency are plotted as functions ofthe engine speed in r.p.m. It will he noted that apower of up to 40 H.P. can be developed and thatthe overall efficiency has a maximum value of38%. In addition, the measured torque is plotted

10) Apart from the usual pumps and fans for forcedoil feed andcooling,the auxiliaries comprise the regulating compr!lssor .already mentioned and a blower for the burner. Of allthese the latter absorbs the most power, viz. 1 to 2% of

. the engine output. . .


Fig. 20. Equipment used for measuring the torque, shaft power and efficiency of the hot-gas engine under widely differing conditions.

in jig. 22 for the highest working pressure, for whichPmax = 140 kg/cm2• The unusual flexibility of theengine is apparent from the small variation in thetorque (11.5 to 15 kg.m or 80 to 110 lb.ft) at greatly

I] Pe

i30~--~~--__~-----*----~~~~]ot

L_~ __ ~ ~ ~~ L_ __ ~O500 7000 7500 2000 2500

_n r.p.m.77325

Fig. 21. Measured shaft power Pe and efficiency rJ of the ex-perimental version of the new Philips hot-gas engine withrhombic drive, plotted as functions of the engine speed forvarious values of pmax.

different speeds. Equally remarkable is the factthat the overall efficiency varies only slightly overa considerable range of speeds and pressures.It is the usual practice to display the results of

76kgm

/'V ---r-....

7 Prrox='40kgjcm2 <,/ <,

<,

74

70

8

6

4

2

oo 1000 2000 2500

r.p.m.9b954

7500_n

500

Fig. 22. Measured torque M as a function of the engine speed,for the highest admissible pressure level.

260 PHILlPS TECHNICAL REVIEW VOLUl\iE 20

Pe=/020 40 60 80 lOO120 14OH.P.20

7 kgfcm2

Pe grams175 per

t 6 ' 180H.P.hourPe

~'9015 fs]/200

4 V210

10 /220,3 /240,

/260/

25

engine tests in a graph such as that of jig. 23. Herethe coordinates are the engine speed and pe, the"mean effective pressure" (m.e.p.]. This latter isnot the mean gas pressure actually occurring in theengine but merely a formal ratio, being defined asthe work done by the shaft per revolution dividedby the volume swept by the piston; hence the prod-

Pe=70 20 JOH.P.

o~----~~ __~~ ~~ ~ __~~o 500 1000 1500 2000 2500

_n r.p.m:g

.Fig. 23a shows these curves for the experimentalengine. The upper boundary to this family of curves,indicated by a broken line, is due to the ceiling onthe gas pressure in the cylinder; in the present casethe highest permissible pressure is 140 kg/cm2, andconsequently Pc for a given engine speed cannotexceed the value corresponding to that pressure.

o 400 800 1200 1600r.p.m._n

b 97326

Fig. 23. a) Curvesof constant specificfuel consumption (in grams of fuel per H.P. hour;for lb per H.P. hour divide by 453) are drawn here in the pc,n diagram for the hot-gasengine. Pc is the mean effective pressure (m.e.p.) defined as the useful work done by theshaft per revolution divided by the volume swept out by the piston. Accordingly, pc. n oePc,which means that, for a given engine, every point in the diagram corresponds to a definitevalue of the power Pc and that the curves of constant Pc (lightly drawn in the diagram)are hyperbolae. The broken line representing the upper limit to the measurements corre-sponds to the pressure ceiling of pm." = 140 kgfcm2 (this curve is derived from the Pccurve for this pressure, given in fig. 21,simply by dividing each ordinate by the correspond-ing abscissa). Thebrokenline representing the lower limit corresponds to the lowest pressurelevel at which measurements were made, corresponding to pm." = 50 kgfcm2•

b) pc,n diagram for a diesel engine (after J. M. Kuijper, Dieselmotoren, Stam, Haarlem1954, page 81, fig. 52).

uct of pe and n, the speed in r.p.m., is the specificpower output. For a given engine, therefore, everypoint in the pf"n diagram corresponds to a definitevalue of the power developed Pe, and the curves ofconstant Pe are hyperbolae, in view of the fact that

(14)

In such a diagram contours of constant specific'fuel consumption (grams of fuel per horsepower-hour, output measured at the shaft) are now drawn.

(The lower boundary to the curves is not an actualrestrietion on the engine; it corresponds to the lowestpressure at which measurements were carried out,namely Pmax = 50 kg/cm2• The broken-line por-tions of the curves of constant specific fuel consump-tion, which Iie beneath the lower limit, have beenextrapolated. )The striking thing about the diagram is that the

curves do not form complete loops, as they do inthe corresponding diagram for a diesel engine, for

'..


IIIIIIIII

40% -9--- 'YJ 40%I ----0 --<>---

JO'! I 30 rzIt I20 t20 I

I5 I

la I 10I

0 0 I 000 400 600 700 800°C 0 10 20 JO 40 50 60 70°C

a b 97327

Fig. 24. Shaft power and efficiency of the experimental hot-gas engine as functions of (a)the heater temperature and (b) the cooler temperature. The curves are those appropriateto an engine speed of 1500 r.p.m. and a pm." of 140 kg/cm2•

example - see fig. 23b. This means that the hot-gasengine achieves its highest efficiency at an m.e.p.which lies close to the highest permissible value.This highest permissible value is here not determinedby a "smoke-limit" as in a diesel engine, but simplyby the strength of the engine casing. In makingsuch a comparison, however, it should be,emphasized

. that m.e.p. values alone are quite useless (and so isthe specific power output) as a yardstick for com-paring the hot-gas engine with internal-combustiontypes from the standpoint of bulk. The fundamentaldifferences are too great. A better yardstick is thespecific weight, that is, the weight per horsepower.

Although few special measures were taken to keepthe weight of the present experimental engine low,the specific weight is in fact only 5 kg per H.P.(11 lb per H.P.).The following are the extreme values found for

various quantities.

Maximum shaft output. Pe *) .Maximum mean èffective

pressure »« **)Maximum torque * *)Maximum specific power

output *) (H.P. per litresweptvolume ofpiston)

Minimum specific fuelconsumption * **) . .

40H.P.

26 kgfcm2("" 26 atm)15 kg.m (",,110 lh.ft)

120 H.P .flitre

165 gramsfH.P.hour(0.36IbfH.P.hour)

*)**)..~) Occurs at n = 2500 r.p.m. and pm." = 140 kg/cm2•

Occurs at n = approx. 800 r.p.m, and pm." = 140 kg/cm2•

Occurs at n = approx. 1350 r.p.m, and pm.x = 140kg/cm2•

All the above values were measured with thenominal heater and cooler temperatures given ear-.lier. The effect of changes in these temperatures onoutput and efficiency may be seen in figs. 24aand b, The curves given are appropriate to an enginespeed of 1500 r.p.m. and a pmax of 140 kgfcm2•

Finally, figs. 25a and b are graphs showing theheat balance of the engine plotted as a functionof Pe (at n = 1500 r.p.m.) and as a function of n(at pmax = 140 kgfcm2) respectively. The narrow-ness of the band representing Hue losses shows theeffectiveness of the air pre-heater: over a large rangeof n and pc values the burner has' an efficiency of

over 90%. Chief amongst what are shown as "otherlosses" are those due to mechanical friction. Theitem mayalso include losses due to incomplete fuelcombustion ("non-detectable" Hue losses). One can-not expect these "other losses" to be specified withany great accuracy, but from the good reproducibil-ity of the results, we may reasonably conclude thatthe mechanical efficiency of this rhombic-driveengine is high.

Concluding remarks

To end with, we may recall here the essentialfeatures of the hot-gas engine in general (cf. 2»:a) Liquid and gaseous fuels of greatly differing

properties can be used, little modificätion of theengine being necessary.

b) Lubricating-oil consumption is low and there islittle wear, owing to the absence of corrosivesubstances and to the relatively low pistonvelocity.

c) Noise level is very low.

l1


g

0 Other losses

»>j.--- Flue losses-J...--... 1-

Cooling-waferlosses-i-

.-»>I--

Useful hea

80

60

20

oo 205 TO 15

-Pe25

kgjc",

100% Other losses- JFlue losses--I---

I- Cooling-waterlosses

/V-- .--

useful heaf

80

60

40

20

oo 1000 7500

_n500 2000 2500r.p.m.97328

Fig. 25. Heat balance for the experimental hot-gas engine (a) as a function of pc at n =1500 r.p.m., and (b) as a function of n at pm." = 140 kgfcm2•

d) The torque varies only slightly over a large rangeof engine speeds.

e) The torque variation as a function of the crankangle is relatively small. In this respect a single-cylinder hot-gas engine is comparable with afour-cylinder internal-combustion engine.In addition to these general features, the experi-

mental engine described here converts heat intomechanical energy with a high efficiency, as theresults of our measurements illustrate. Moreover it

Summary. After a brief recapitulation of the hot-gas cycle asapplied to an engine provided with a displacer piston, certainnew design features of the Philips hot-gas engine are described.An account is given of a new kind of drive mechanism in whichthe required power and displacer piston movements are ob-tained by means of twin crank and con-rod mechanisms offsetfrom the engine axis; the crankshafts are coupled by two gear-wheels, and rotate in opposite senses. Where such a system isused, the buffer pressure that is desirable under the powerpiston can be contained in a relatively small cylindrical bufferchamber - there is no need to pressurize the whole crankcase.The new drive mechanism also permits complete dynamicbalancing of the inertial forces of moving parts, even in asingle-cylinder engine. Other innovations concern the designof the regenerator compartment, the cooler and the heater: anentirely new approach has cleared up the problem of thermalstress in these components, whose weight can be kept quitelow despite the high gas pressure. The pre-heater has also been

has been shown that good values of the weight perhorsepower are attainable. These favourable quali-ties are mainly to be attributed to three features ofthe design: the use of a displacer piston with itsattendent aerodynamic and thermodynamic advan-tages, the use of hydrogen as the working fluid, andthe use of high working pressures. A further advan-tage of the newly developed engine is its freedom fromvibration, in consequence of the complete dynamicbalancing of the inertial forces of moving parts.

redesigned, with the ·result that burner efficiencies of up to90% are now reached. In addition, a fast-acting system hasbeen developed for regulating the engine output, this beingcontrolled by altering the mean working pressure of the gasin the cylinder.

These various features have been applied in the design andconstruction of an experimental single-cylinder engine ofabout 40 H.P., containing hydrogen (helium could also beused) as the working flnid and burning light fuel oil. The highestpressure permissible in the cylinder is 140 kgfcm2• The nominalspeed of the engine is 1500 r.p.m. Extensive tests on thisengine, the results of which are given in the article, show thatwith the heater temperature at 700 oe it attains an overall ef-ficiency ofup to about 38% (engine alone, without auxiliaries).The maximum value of the specific power was found to be120 H.P. per litre swept volume and the best specificfuel con-sumption 165 grams (0.36 lb) per H.P. hour. The specificweight of the engine is 5 kg (11 lb) per H.P.

Philips Technical Review - Research | Philips Bound... · 246 PHILIPS TECHNICAL REVIEW VOLUME 20 which also possesses thefeature that the crankcase need not he pressurized. This was

Documents