Considering that at the moderate angular speed of 3,000 revolutions per minute (rpm), a 6-cylinder engine in one hour only carried out more than 2 million strokes by means of pistons only a few mils (10 inches) [25.4 mic- rons; 10 x 2.54 cm] less in diameter than the cylinders in which they move at peak linear speeds of about 35 miles per hour (mph) [56.4 kilometers per hour], also undergoing peak decelerations of about 600 g's (at top-dead-center, or TDC), further being subjected to peak pressures as high as 40 atmospheres or more, considering also that the side-thrusts — namely the horizontal forces imported to the piston by the crankrod reactions — may reach many hundreds of pounds [ 1 lb = 4.4N], that the temperatures in the piston-cylinder environ¬ ment are high, about 80 C at the bottom of the piston and well above 200 C at the top, that such conventional piston-cylinder systems effectively seal the high pressures in the combustion chamber above the piston from the ambient pressure in the crankcase below, simultaneously also sealing the same spa- ces against oil-pumping from below to above, and that such internal-combus¬ tion (IC) engines satisfactorily perform over very long periods of duty, it is obvious that the conventional IC engine represents an astonishing (if taken for granted) feat of engineering.

In view of the severity of operation and excellence of performance just indicated, it is widely felt, and agreed with by this inventor, that the per¬ formance of the conventional engine, in particular that of the piston-cylinder system, as such is probably as effective as can be practically expected. The invention however recognizes that the conventional piston, for instance the four-stroke type used in automobile IC engines, is appreciably more frictio- nal than the theoretical minimum recognized by the invention, and that accor¬ dingly significant improvements, proposed hereunder, can still be achieved, resulting in higher efficiencies, that is, higher fuel economy, and in ancil¬ lary advantages such as longer engine life.

The representative conventional piston of the spark-ignited IC en- gine with four strokes comprises essentially a lower part, the skirt, through which the horizontal components of the crankrod forces acting or reacting on the piston are transmitted transversely to the cylinder wall, and an upper part, the crown, which holds and houses a number of rings for oil-control and for sealing the combustion chamber above the piston from the oil case below,

to prevent both gas-leakage and oil-pumping. Typically there are two sealing rings, cal led compression rings , at the top of the crown, and one or more oil- control rings below them, in representative automobile engines . The comp- ression rings do not form a solid circle but instead are finely split and pre¬ stressed to expand from the grooves in which they are housed to form inti¬ mate contact with the cylinder wal l in order to achieve the desired sealing.

In view of the high number of strokes performed by such an engine , seal ing against oil-pumping into the combustion chamber must be rigorous , — since even one drop per cycle pumped into said chamber for the operational conditions assumed above would mount into a loss of 500,000 drops an hour. Whi le the two compression rings typical ly may and in fact do assume a great many shapes and profi les , the upper one at least is made to physical ly scrape the oil from the cylinder wall downward. The oil acting as a lubricant between piston and cylinder wal l is provided from below and oh the whole works its way up to adequately unti l overcome by the del iberate scraping action of the upper¬ most compression ring, said action being most drastic near TDC. Accor¬ dingly there is little if any oil on the cyl inder wal l -opposite said uppermost ring when the piston is in or near the TDC position , and this seal ing ring also being prestressed to expand, in that region rubs against the inadequately lub¬ ricated cylinder wal l , whereby, at least intermittently, high friction is gene¬ rated in the piston cycle , namely near TDC, entailing abrasion both of the cylinder wall and of the ring. As a result fuel consumption is higher than it would be in the absence of such abrasion , and in the long run, the upper cylin- der flares , the upper compression ring wears, until engine operation becomes excessively degraded.

Another factor is that ignition in the combustion chamber raises the gas temperature therein to the order of 5 ,000 F [2 ,760 C] and that this fire reaches that part of the cylinder wall uncovered by the piston in proportion to its distance ^* from TDC; if furthermore the upper groove housing the top com¬ pression ring is eroded enough by said ring having long spun and twisted in it on account of the strong frictional forces applied to it in the conventional case, the fire may also leak through the passage so abraded and reach the lo¬ wer groove. I t is commonly held that unless the oil on the cylinder wal l near TDC be kept at a minimum, as in fact it is by the deliberate scraping of the upper compression ring in that area, deleterious effects such as burning, carbon-formation and ring-sticking wil l be excessive, and that if the fire were to reach the lower groove, it might or would destroy the oil-film, or seal , there.

Final ly the conventional piston within the sl ight clearance left to it inside the cylinder is known to tilt in general with respect to the cyl inder axis . This is because no provision is made to control the attitude of a conven¬ tional piston, which can be observed to slightly tilt one way or another depending on the di rection of the transverse force applied to it and on that of it s longi¬ tudinal motion. As a consequence , even though the conventional piston skirt always rides on oil , this skirt friction too is higher than the minimum recog¬ nized by the invention.

As already stated, the object of the invention is to create a low- friction piston in order to achieve higher fuel-economy, for instance in auto¬ motive vehicles. The invention recognizes that on one hand it is possible to keep with impunity an oi l-film as high on the cylinder-wal l as the TDC position of the upper sealing ring, and on the other that said sealing ring can be desig¬ ned to be a hydrodynamic reaction means , more particularly a hydrodynamic reaction surface means to expand by pressure-actuation and to be balanced by hydrodynamic forces or reactions or reach its end position against a limit means , in order to always maintain an adequate oil film as the seal between the combustion chamber and the crankcase or oil case , whereby abrasion is substantial ly entirely eliminated except perhaps with respect to the grit ϊnhe- rent in oil , further that the piston can be provided with specifical ly designed and cooperating hydrodynamic .reaction means , more particularly hydrodynarric reaction surface means , or hydrodynamic bearing surfaces, in particular at the skirt , to balance the above-mentioned horizont al or transverse thrusts , whereby the friction is reduced stil l further while simultaneously an attitude control of the piston with respect to the cylinder axis is achieved, so that on the whole significantly increased fuel-economy is obtained, and as an ancil¬ lary advantage , longer engine l ife is real ized.

The various elements of this invention cooperating towards the pur¬ pose of reducing piston friction as wel l as conventional means and methods not claimed per se but desirable in the implementation of this invention are discus¬ sed in detail below in relation to the drawings and examples. For the sake of overview , a complete qual itative discussion of the invention relating to the drawings is provided first, fol lowed by numerical examples at the end.

Fig. 1 shows an embodiment of the piston of the invention partly in profile and section; fig. 2-is a horizontal cross-section of squeeze-effect bearing sur¬ faces used in the piston of the invention; fig. 3 is a vertical cross-section showing one embodiment of a sealing ring of the piston of the invention;

fig. 4 is a vertical cross-section showing another embodiment of a sealing ring of the invention; fig. 5a, 5b represent one-sided profiles of compact embodiments of the piston of the invention; fig. 6a shows a symbolic two-sided hydrodynamic surface; fig. 6b.,

6c are graphs of the hydrodynamic parameter Δg used in the examples of this description;

' fig. 7 is a graph showing the transverse thrust T(0) used in the examples of this description, and fig. 8 is a reproduction of the performance graph , of a conventio¬ nal engine, used for a theoretical comparison with the performance of the pis¬ ton of the invention.

Fig. 1 shows essential ly a vertical profile of one embodiment of the piston 90 of the present invention. The gaps between piston and cylinder wal l are much exaggerated for clarity. The piston 90 moves within a cylinder 60 which is capped by an annular splash guard 62, which is optional . The piston is actuated by a conventional crankrod 40 with conventional wrist-pin 41 . Simi¬ lar to conventional pistons , piston 90 of the invention also consists of a crown and a skirt , but the crown 20 and skirt 30 of the invention are much different from the conventional design. I n general the piston of the invention also com¬ prises a heat-barrier at the top of the crown , said heat-barrier however being of conventional design and not an object of this invention.

The conventional heat-barrier 10 symbolical ly shown in fig. 1 with its temperature-dropping chamber 14 or equivalent is general ly desirable and sometimes mandatory in order to maintain adequate hydrodynamic properties in the temperature-sensitive oil film acting as the bearing for the various surfa¬ ces and components of the piston of the invention . For a represent ative auto¬ mobi le spark-ignited I C engine as considered herein, the temperature of the conventonal upper compression ring is about 200 C or more; the purpose of heat-barrier 10 is to drop this temperature to about 100 C or less.

Below heat-barrier 10, there is a land 21 of asymmetrical shape which controls the oi l flow by being down-pumping into the oil case if the film opposite on cylinder wal l 61 becomes excessively thick; reservoir R ¹ inter¬ mittently stores and dispenses oil displaced, during the periodic meotϊon of the piston. Lands 23 and 27 are sufficiently recessed from wall 61 to generate only little hydrodynamic friction, but are hydrodynamical ly profiled to induce high and sufficient hydrodynamic reactions from the oil on wall 61 in case of improper approach, say malfunction; in another embodiment of the invention,

OMPI

however, this part of the crown is telescoped into the skirt of the invention, and the hydrodynamic surfaces of these lands 23 and 27 are then merged into one' of the hydrodynamic surfaces of the skirt.

In the embodiment of fig. 1 , a sealing ring 70' is seated in a groove 25 between the lands 23 and 27. In the absence of significant pressure in the combustion chamber above, this ring 70' remains prestressed snug against groove 25; under the effect of pressure however, it will expand and form a narrower gap between ^" itself and wall 61, thereby providing the required and increasing sealing, as shown in further detaϊ. below in relation to fig.3 and 4. While only one sealing ring 70' is shown, obviously several can be used if desired.

The crown 20 is separated from the skirt 30 by a groove 28 with an optional oil drain 31 for return to the crankcase or oil sump. Such optio¬ nal oil drain may be associated with a conventional oil ring, of which the frtc- tion is relatively but not trivially small, which would be seated in groove 28. The skirt 30 of the embodiment of fig. 1 consists of two, well- localized, mirror-symmetrical hydrodynamic bearing surfaces 32 (32a, 32b, 32c) and 34 (34a, 34b, 34c) which are symmetrically located resp. above and below the horizontal plane 36 in which acts the horizontal component T of the crankrod reaction. They are furthermore separated by a groove 33 with oil- drain 35 to the crankcase, again a conventional oil-ring being permissible, if desired, in said groove 33. For the conditions of operation indicated in fig. 1 , in which the velocity U is downward and the thrust T to the left, dynamic equilibrium is obtained from the hydrodynamic reactions shown. The major hydrodynamic bearing reactions are the "stretch effect" (which is roughly the water-ski effect), termed W herein, and the "squeeze effect", arising from the fact that for instance a film of oil trapped between two parallel and app¬ roaching surfaces requires time to escape, termed W herein. As indica¬ ted in fig. 1 , the stretch effect acts on both sides of the piston bearing sur- faces, whereas the squeeze effect acts only where the piston approaches the cylinder wall. Surfaces 32 and 34 are made bi-directional because of the reci¬ procating nature of the piston motion. Because surfaces 32 and 34 subtend slants with respect to wall 61 , namely surfaces 32a, 32c and 34c, 34a, they generate a stretch effect , parallel surfaces 32b and 34b generating a squeeze effect. When the geometry of the various surfaces of bearings 32 and 34 are properly selected for each and with respect to one another, the forces and moments of these hydrodynamic reactions can be made to balance the trans¬ verse thrust T for all longitudinal positions of the piston within the cylinder

so as to control the piston attitude with respect to the cylinder axis, for inst¬ ance to keep it paral lel to said axis , and furthermore such selection of geo¬ metric and other parameters (for instance temperature and viscosity control) can be such that the hydrodynamic friction itself shal l be minimized. The atti- tude control just cited is obtained by the cooperation of the two bearing sur¬ faces 32 and 34 shown in fig. 1 . These wel (-local ized bearing surfaces there¬ fore fil l a doubly antϊfrϊctiona! role: they minimize the friction compared with that encountered in the conventional case at the skirt , and by the attitude con¬ trol they make possible they eliminate or at least reduce the friction which in the conventional case arises at the conventional crown because of the un¬ controlled tilt in the prior state of the art. Finally, while two bearing sur¬ faces with considerable symmetry are shown in fig. 1 in illustrative manner, it should be clear that a different number of bearings with different geomet¬ ries can also be resorted to in order to obtain similar effects. Fig. 1 further shows oil 50 below the skirt and against wall 61 ; this oil is assumed fil l ing the space between the piston and the said wal l at least as high ring 70' , but is omitted from most of the drawings for the sake of cla¬ rity. Again fig. 1 shows a conventional crankrod 40 with wrist pin 41 , and the transverse thrust T(θ) . I t is furthermore assumed herein that the top of the crown is symmet¬ rical enough that the transverse forces exerted on it by the pressure in the combustion chamber will substantial ly cancel out and therefore can be neglec¬ ted from consideration.

Fig. 2 shows a horizontal cross-section through surfaces 34 along line 3-3 of fig. 1 . In this case the skirt's hydrodynamic bearing surfaces are unequal arcs of circle because a smaller surface is adequate to take care of the minor thrust , here shown occurring on the right-hand side. This allows a sl ight saving in hydrodynamic friction. Where desired, any or all the skirt bearing surfaces may be made 360 . For the sake of simpl icity, fig. 2 omits al l other cross-sectional views of the components except that of cylinder 60.

Fig. 3 and 4 show specific embodi ments of the sealing means 70' of fig. 1 . In fig. 3, the sealing ring 70 is approximately T-shaped in its vertical cross-section and rests in a corresponding groove 25. Said T-sha e consists of a head 701 and a stem 702. Ring 70 preferably is pre-stressed to slightly contract in the absence of pressure in the combustion chamber; it will expand as the combustion chamber pressure builds up, as explained hereunder. A sl ight play is provided between the height z„ of said stem and the height of groove 25, and again between height H of the said head and the vertical dimen¬ sion subtended by the groove faces of lands 23 and 27 further, the length s

of stem 702 is sl ightly less than the depth of groove 25; finally fluting 72 is provided at the upper end, at the higher base of of head 701 , and along one side of stem 702 for the purpose of allowing automatic sealing control between the combustion-chamber and the crankcase by providing easy propagation of the combustion-chamber pressure to those surfaces on which it should act , in par¬ ticular during compression and combustion.

Because of the various clearances just indicated, two effects are obtained when pressure appears in the combustion chamber. . I n the first place, lower side 702b of stem 702 is pressed into a hermetic seal against groove surface 25a, so that there is a (mathematical ly) discontinuous pressure drop between the bottom of groove 25 and land 27, and a(mathematϊcal ly ) continuous pressure drop across the height H of the head 701 of the ring. The second effect is that a force per unit length p z ₂ , where p is the pressure in the com¬ bustion chamber, pushes the bottom 702a of the stem toward wal l 61 , whereby the seal formed by the oil between the surface 71 of the head and wal l 61 be¬ comes tighter, and in fact , tight enough. As shown later under example 2 , this automatic seal ing behavior is feasible provided the pertinent parameters are properly selected. The dashed lines for lands 23 and 27 indicate other pos¬ sible designs . The embodiment of fig. 3 is particularly suitable where relatively substantial thermal expansions or other changes in size take place , since the pressure wil l force the ring to expand until balanced by a thin enough film of instance oil . As the ring is necessarily split , large expansions may degrade for/blow- by seal ing, which might be remedied by using two sealing rings of the inven- tϊon. Whi le manifestly it is general ly desirable to use the fewest possible parts , the invention places no restriction on the number of its seal ing rings, whether of the embodi ment just shown in fig. 3 or that of fig. 4 discussed fur¬ ther below . Final ly it is instructive to compare the behavior of the ring of fig. 3 (and also that of fig. 4) with the conventional case . Conventional ly, the compression ring expands because prestressed and near TDC comes into at least intermittently abrasive contact with wal l 61 , whereas in the invention the seal ing ring always rides on oil . Even at the very high peak pressures occurring in the combustion chamber of an automobi le engine , say 50 atmosphe¬ res under some conditions , the hydrodynamic reactions provided by the oil film of the invention are sufficient to keep the ring surface 71 apart from wal l 61 . Accordi ngly the conventional abrasive friction is eliminated and repla¬ ced by hydrodynamic friction, which is much lower. Considering that the crown friction in the conventional piston is about 4/5 of the total piston fric-

0Λ.PI

tion , the reduction .in friction achieved by the invention is accordingly subs¬ tantial . An ancil lary advantage , itself of great weight , is that the wear taking place in the conventional piston system at the cylinder wal l and compression ring is eliminated by the invention , whereby much longer engine l ife is achieved. Fig. 4 shows another embodiment of a seal ing ring of the invention , which is applicable where dimensional changes in piston and/or cyl inder-wal l are relatively smal l and where a fixed excursion of the ring 80, which is also pressure-actuated , may be desirable . I n this case the seal ing ring 80 comp¬ rises a lower sliding surface 80a which makes an acute angle with the vertical and which in operation glides on the corresponding slanted surface 27'a of land 27' . Ring 80 at its inward base is provided with a foot 801 general ly at right angl e to body 802 , face 801a of foot 801 being some distance S' apart from face 25'a of land 27' in the non-actuated case. Furthermore faces 801 b and 25'b as wel l as 27'a and 80a are apart by a cl earance d in the unloaded . case , so that there be no jamming of ring 80 in groove 25' . A fluting 81 is pro¬ vided at the top of ring 80 so pressure from the combustion chamber can easily communicate into groove 25' , simi larly to the provisions made for ring 70 of fig. 3. General ly ring 80 is substantial ly prestressed to contract in order to assume the position shown in fig. 4, namely abutting face 23'a and 23'b unti l overcome by a certain , preselected threshold pressure. This appreciable prestressing is desirable because ring 80 is susceptible to the inertial forces exerted on the piston because of its reciprocating motion , the said ring other¬ wise expanding when approaching bottom-dead-center (BDC) . The prestressing imparted to ring 80 therefore ought to approximately correspond to the magni- tude of the inertial forces it is exposed to . Accordingly a definite pressure threshold must be appl ied to ring 80 before it wi l l expand. Once that thres¬ hold is exceeded, ring 80 expands as the pressure increases further until it reaches a stop means , represented here by l imit-surface 25'a . Further inc¬ reases in pressure obviously remain ineffective as regards the excursion of ring 80. Ring 80 therefore offers the characteristics of both a substantial pressure threshold and of a finite excursion.

I n the conventional case , the compression rings spin azϊmuthal ly in their grooves , and furthermore twist in them, on which account these grooves ultimately wear out . The seal ing means of the invention however , being shiel- ded from abrasive contact with the wal l 61 , thereby are also protected from forces causing substantial spinning and/or twisting. Because the pressure- actuated motion between rings 70 and 80 resp. shown in fig. 3 and 4 and

_3 their seating grooves 25 and 25' resp. is minute , i . e. , only a few mils [10 x

_0.V.P λ>- VVi

25 mm] , and as this motion furthermore is lubricated by the residual oil between these surfaces in contact , both very low friction and hermetic contact is provided between faces 25a and 702b in fig. 3 and 27'a and 801 a in fig. 4. I t should be noted that even if there were some wear, the seal ing bet- ween these surfaces would remain unaffected. Both surfaces 25a and 702 could wear considerably before any deterioration would be felt . Surfaces 80a and 27'a again could be considerably eroded without other effects than inc¬ reasing the level at which the pressure ceases being a factor in the expan¬ sion of ring 80. Ring 80 in any event might be designed in fai l-safe manner for such remote possibility to be balanced then by hydrodynamic reactions , simi larly to ring 70 of fig. 3. Again, the dashed lines for lands 23' and 27' denote other possible dimensions of these lands- Fig. 5a, 5b are embodiments of the piston of the invention in which the ^• crown and the skirt as previously shown in fig. 1 are partly telescoped into one another. I n the embodiments of fig. 5a,b the lower hydrodynamic bearing surface 34 or 34' may be the same as or similar to the bearing 34 of fig. 1 . But the hydrodynamic bearing surface 32' (32'a ,32'b,32'c) above wrist pin 41 may be either the same as or similar to surface 32 of fig. 1 , as shown in fig. 5b, or else the paral lel surface 32'b may be entirely replaced by a face such as 71 in fig. 3 of a ring 70 seated in a groove 22" , as shown in fig. 5a.

In the first embodiment relating to fig. 5b and for which no seal ing ring is used, bearing 32' clearly requires some feature preventing oil- pumping into the combusion chamber. One way of obtaining this feature is to make surface 32'a sweep back more than surface 32'c from wal l 61 . I n the second embodiment of fig. 5a, using a seal ing ring 70 such as shown in fig. 3 , the sweep-back feature just mentioned may also be combined with a dee¬ per groove 22' acting as an intermittent oi l storage similarly to groove 22 of fig. 1 . I t is evident from inspection that the embodiments relating to fig. 5a , b are quite compact; their actual size would be comparable to conventional partial skirt pistons.

If desired, the piston of the invention can be supplemented by a splash guard 62 in the form of an annular disk placed on, or sunken into the top of cylinder 60 as shown in fig. 1 . Otherwise, if at all , the piston of the invention requires no modification of the remaining engine. As the dimensions of the piston of the invention are approximately the same as those of conven¬ tional pistons, substitution can be effected without affecting the engine,

possibly selecting a crankrod of slightly different size on occasion may be convenient. Furthermore the piston of the invention can be made of the same materials as its conventional counterpart, and conventional oils can be used. Accordingly the gain in fuel economy and in engine life afforded by this inven- tion is achieved by only modifying the piston itself, that is, economical ly. Because of l ittle extant knowledge about the processes of wear, (tribology) , it is not real ly possible to predict how much longer-lived both the piston of the invention and the cylinder in which it moves wil l be. I t seems reasonable that mechanical creep will not occur because of insufficiently severe operational conditions , and that the life of the piston of the invention (and of the cylinder) will be determined by the other remaining degrading fac¬ tors also affecting the conventional engine, namely corrosion , grit etc. I t is known that in conventional engines, the skirt of the piston always rides on oil and furthermore wears relatively little, so that substantial extension of life can be expected for the piston of this invention.

Steady-state operation has been implied so far. Some thought how¬ ever must also be given to strongly aperiodic motion. I f for instance the en¬ gine is left shut off overnight , some of the oil film on wal l 61 of fig. 1 obvious¬ ly wil l drain into the oil case. However enough of it will remain, whether trap- ped by capil larity, adhesion or stored in the various grooves , to allow the piston of the invention, when started again, to immediately take the path of least resistance away from the wal l against which it had inevitably come to rest in intimate contact during the long previous period of inactivity . The viscosity of the cold oil being very high , correspondingly little oil only is needed to prevent any abrasion. Furthermore, the proportion of strongly aperiodic piston motions to those of steady state is vanishingly smal l in practice, so that even if there were some abrasion in such transient conditions — as con¬ ceivably might be the case for an engine being restarted when still hot — it would be equal ly trivial . The problem, if any, of oil presence near the combustion chamber wi l l now be considered. For all the raging fire's disquieting propinquity to the oil film recommended by this invention on the cylinder wal l even near the TDC position of the piston, the commonly feared burning or destruction by fire can only take place if the oil is in the gaseous state; in other words, first the oil must evaporate. The heating which might cause such evaporation can arise from two sources , namely from the hot and/or burning gases during compres¬ sion and combustion resp. , and from the hot metal of the piston crown . As regards the latter case, oil burning in conventional engines is kept consϊde-

rably cooler than the conventional one — at least where the preferred bar¬ rier is employed — then also there will be even less evaporation of oil from contact with the piston of the invention. Accordingly only the proximity of fire to the oil film is left to be considered. Be it assumed the piston is past TDC during the combustion phase, so that there is fire in the volume of the cylinder already vacated by the piston in its descent, and that there is the desired film of oil on the wall 61 so bared. Some of the heat being generated will flow out of the bulk of the burning gas, which is a poor thermal conductor, through the oil film, which is also a poor thermal conductor but exceedingly thin, and then through the thickness of the cylinder, which is a metal and a very good thermal conductor, into the radia¬ tor coolant, which can be considered an infinitely good thermal conductor and in fact an infinite heat sink. It is intuitively clear that the infinite heat sink smack against the metal of the cylinder also will keep the oil-film temperature low. A tremendous temperature drop from the inside of the fire to the boun¬ dary between oil film and combustion gas is made possible by the thermal transfer coefficient h.. at said boundary on one hand and by the transient, though recurring nature of the caloric combustion pulse on the other, as the thermal time-constant of the cylinder metal-thickness much exceeds the lengths of said caloric pulse. Computations, omitted for the sake of brevity, indicate that the temperature of the oil film at the gas boundary might rise about 30 C above that at the metal boundary. Such rise obviously is wholly harmless as regards evaporation. Furthermore, this rise is wiped out the moment the relatively cool piston again makes contact with the outer surface of the oil film and thereby reestablishes also the desirable viscosity. On the whole therefore, even though more oil is exposed in the case of the invention, it is kept cooler than conventionally, whereby too the conventionally feared factors of carbon formation, ring sticking etc. , rather than being enhanced, on the contrary are reduced. Finally, the minimum oil-film thickness between the bearing surfaces and the ring faces on one hand and the cylinder wall on the other will now be briefly considered. The bearing capacity from the hydrodynamic stretch and squeeze effects, and the friction all increase as the oil film thickness decrea¬ ses. In addition, the film thickness may have to take into account the rough- ness of the surfaces opposite each other. Again machining tolerances also must be taken into account. Illustratively, it can be assumed that machining tolerances for the skirt bearing surfaces and the cylinder walls can be kept to within +/- 50 mϊcroinches [ 1.3 microns], and that the roughness of

these surfaces is kept to within +/- 25 microϊnches [0.64 microns]; then a nearest nominal approach of 300 microinches ( 0.3 mils) [ 7.6 microns ] would leave at least 0.15 mils [ 3.8 microns] ^" of oil film. No abra¬ sion would take place under such conditions, all the more that hydrodynamic reactions for such a thin film are very high for the conditions considered herein. As shown by one of the examples below, the gap left at closest app¬ roach in a typical assumed case is considerably larger. The closest approach is one of the design criteria of the piston of the invention. It ϊs"affected by tha dynamic conditions of operation, by the number and type of bearing surfaces and ring faces, and can be traded off to some extent against such parameters to optimize the particular operation. As a rule of thumb, the invention pre¬ fers a closest approach of at least 0.3 mils [3.8 microns] in general.

EXAMPLE 1: Sk i rt ' s closest approach to cylinder wall

Symbols and parameters are resp. defined and assumed as follows: W = hydrodynamic reaction

2 3 2 L^= viscosity in reyns (Ib-sec/in ) [7 x 10 Newton-sec/m ] t.= width of slider into plane of paper (fig. 1), assumed to be l inch [38 mm]

-2 ^~~ - ₃ <y - slope of slider surface 32c, assumed to be 10 .- , = slope of slider surface 34a, assumed to be ,J2 x 10 a = h-i ho (f'9- D. assumed the same for all surfaces, left side of fig. 1 a' = parameter a on the right side of fig. 1

U = piston speed in inch/seconds [0.025m/second] g(a)= ln(a)-2(a-1)/(a+1)

H«/H- (fig. 1), assumed being 2 H = height of squeeze effect surfaces 32b, 34b, assumed being inch [12i mm] h,, h. = final and initial gaps between wall 61 and surfaces 32b, 34b (fig. 1) piston weight: assumed 1 lb [454 g] crankrod/crankshaft-radius: assumed 4:1 clearance between diameters of wall 61 and surfaces 32b, 34b: assumed 3 mils [0.076 mm] engine speed: assumed 5,000 rpm (i.e., very high). From inspection of fig.1 ," T(0) = A W (32c) + 2 W + A (34a) (1) where the subscripts "st" and "sq" resp. refer to "stretch effect" and "squee- ze effect". •

The expressions for these hydrodynamic reactions can be found in general form for instance in the "Standard Handbook of Lubrication Engineering"

OA'

O'Connor & ^■ Boyd, McGraw-Hil l , 1968, chapters 3 and 7, namely

W = ^LUJLg(a)/m ²

W sq = - tH ³ L(dh'dt)/h ³ where m is the slope of the hydrodynamic surface , or sl ider, and where h is the size of the squeeze gap, whereby eq. (1) can be rewritten as

J τm ^dQ _{= 0#41 g(Udt) +} 0.56 x 10 ^~6 [1/H? - 1/h? ] (2) w J _1 ' ' where dt = dO/w, w = angular speed (sec ), Udt = ds, s = the distance tra¬ veled by the piston under consideration for the integration under consideration., & where the right-hand side after multipl ication by[4.4]will be in MKS .

Fig. 6b, 6c show that Ag = g(a) - g(a') is reasonably linear, so that Δg = g(-_r[h.+ h ) and can be taken out of the integral sign in eq. (2) . Fig. 7 shows the function T(θ)dθ as the area under the curve for an engine speed of 5.000 rpm ; for other engine speeds, the shown area is multiplied

2 by (N/ 5000) , where N is the particular rpm. Letting h. = 1 .5 mils [0.038 mm], eq. (2) for the maximum absolute integrated area, — namely the negative one — is approximately

0.67 = 0.81Ag + 0.56x10 ^_6 /h^ where again left and rjght if multiplied by [4.4] will be in MKS .

As both terms on the right are functions of h,, hr can be solved for, • in this instance easily by trϊal-and-error numerical substitution . The closest approach so computed is approximately h = 0.98 miles [0.025 mm] . While this result was obtained considering only the transverse thrusts due to inertial forces, one should keep in mind that the magnitude of the transverse thrust from the power stroke is of the same order as for said inertial forces assumed abo¬ ve , and that furthermore . varies relatively slowly in view of the sharp inc¬ rease in hydrodynamic forces as it decreases .

An extreme case is now postulated at an engine speed of 500 rpm but with a power stroke equivalent to the inertial forces encountered at 5000 rpm, whereby eq. (2) becomes

4.5 = 0.81Δg + 0.56x 10~ ⁶ /h^ which again, if multiplied by [4.4] on left and right , wil l then be in MKS, and the solution yields approximately h _f = 0.38 mils [0.0095 mm]. Accordingly the two bearing surfaces of the skirt of fig. 1 shal l al¬ ways ride safely on a film of oil throughout the entire range of operation. Ob¬ viously other results , specifical ly tailored to particular conditions , can be obtained by varying the diverse parameters involved.

-BU RE4 ^ _0MPI

EXAMPLE 2: Design parameters for ring 70; its closest approach to wall 61. The surface 71 of ring 70 is meant to stay parallel to wall 61 during operation. Ring 70 being prestressed to contract, it will expand only if there is some pressure p in the combustion chamber, that is, during compression and power. Ring 70 is so designed that said expansion is counteracted by the sque¬ eze effect. For the sake of analytical simplicity, the lower surface of stem 702 is considered resting on surface 25a only at two points, where it is sup¬ ported by reactions N-, and N„, (fig.3). As it is undesirable that ring 70 should tilt or twist in groove 25, each of the reactions N. and N„ must be lar¬ ger than zero. It is assumed for simplicity that 7.. = z = z ₃ = H/3 = 1/6 inch [4.2 mm] h. = 1.5 mils [0.038 mm]

' —6 9

= 3x 10 reyns [0.021 Newton-sec/m ] The balancing force and moment equations can be written (fig.3)

here the first and second 1 ϊnes if multiplied by 4.4 wϊl 1 be in MKS and where the third line if multiplied by 0.098 also will be in MKS.

Accordingly the requirements that N. and N„ each be larger than zero can be shown from the solution of the above simultaneous equations to be resp.

[x + χ ² /2s + (s ² - H ² )/2s] and [s ² + H ² - x ² ], the former being the more stringent. Different constraints obviously will be obtained if the magnitudes z. , z„ and z, are varied from the assumptions above.

The combustion chamber pressure p may be approximated as p = p sin (20) between 0 = 0 amd 0 = 90 . While in an engine 0 is related to the crankshaft angle 0, for the purposes at hand, namely the expansion of the ring under combustion-chamber pressure, it can be treated independently since such expansion would take place any time such pressure would be applied.

Using the expression for the squeeze effect of example 1 in conjun¬ ction with that above for the combustion chamber pressure, namely

W = - H ³ L(dh/dt)/h ³ = p LH/6 = LHp° sϊn(2øy6 sq ' cc ^cc which upon multiplication by[4.4] will be in MKS, and letting dt = dø/w, where w is s tthhee aanngguullaarr ffrreeqquueennccyy ((sseecc ), then integration results in

OΛ ι-

which will be given in MKS when multiplied by 1.6 x 10 and which for engine speeds ^' of 5000 and 500 rpm and for p° of 600 and 100 lb/in [ 40 and 6.8

CC atn.] resp. leads to h _f (5000 rpm) = 1.0 mil [0.025 mm] h _f (500 rpm) = 0.88 mils [0.022 mm]

EXAMPLE 3: Sealing adequacy of ring 70

The invention relies on the oil film between the ring or sealing means 70' and wall 61 to provide sealing between the combustion chamber and the crankcase. Ring 70 of fig.3 is considered hereunder for its sealing pro¬ perties. Because of the high pressures temporarily generated in the combus- tion chamber, there must be enough oil adhesion between the face 71 of said ring and wall 61 to prevent the oil film in-between from being blown clear into the crankcase. The rate dQ/dt at which oil is forced through ducts or paral¬ lel surfaces (see for instance Streeter & Wylie, Fluid Mechanics, 6th ed. 1975, McGraw-Hill, p 242, eq.5-13), is given by -dQ/dt = [ LCdp^/dHjh ³ ]/^^ where dp /dH = p sin(20)/H and the relation applicable to any system of units.

Sealing adequacy can be ascertained by assuming there was film of -oil H high between wall 61 and surface 71 , and by determining the magnitude of Hi at the end of the pressure pulse, the piston being considered stationary. The amount of oil present at any time is merely _f

Q = LhH and in view of the results of example 2, a mean value h may be used, whereby

where w = angular frequency (sec ) and where the epxression applies to any system of units. Letting the initial H. = ^~ inch [12 - mm], and integrating,

H ² = -[h^ δ^] which again applies to any system of units. If engine speeds and combustion-chamber peak pressures of 5000

2 rpm and 600 lb/in [ 40 atm.] are resp. assumed, and remembering that

H _j = inch [12 mm],

H _f = 0.38 and 0.47 inches [9.4 and 12 mm] resp.

As the film of oil essentially remains in the gap, sealing is ade¬ quate. Replenishment of the fraction lost is provided by such means of the in- vention as the asymmetrical land 21.

- ^{QI :} 'PL_

EXAMPLE 4: F rϊctϊon of crown 20 holding one ring 70

Provided lands 21 , 23 and 27 are sufficiently recessed from wall

61 , their hydrodynamic friction can be neglected. This is assumed the case here, and only the friction of ring 70, assumed to the sole sealing ring, will be considered here. As surface 71 is parallel to wall 61 , only shear friction is involved, which is given by

F , = jjLH / [in any system of units] where L = circumference = 10 inches [0.25 m]

H = i inch [0.013 m] Now

9 9 n 9

1/h = 1/h. + p [1 - cos20]/[6twH ], [in any system of units] where w(5000 rpm) = 5001T/3, U = U(θ) = 0-20°.

To obtain the frϊctional equivalent in horsepower P, P = FU, and U being in inches/second [0.025 m/s],-

—9 P = 2.3x 10 'u 2 U ² /h [horsepowers being essentially the same in

\ either system of units].

The results for F , and P are more easily obtained by numerical methods than by integration, and table 1 below lists both the procedure and its results. These results, together with the friction for the skirt obtained fur- ther below, later will be compared with the experimentally ascertained total friction of a convent ϊonal piston.

Table 1 let; w = w(5000 rpm) = 500 ^" »T/3; h. = li mils [0.038 mm]; t = 3 x 10 ^~ reyns [0.021 Newton-sec/m ² ]; Use 1/h ² = 1/h ² + [p° _c ( 1 - cos20)/[6 _/ μ-wH ² ]

= 0.44x 10 ⁶ (1- cos20) [multiply by 1 600 for MKS]

0 .26 x lO ⁶ 0.44x10 ⁶ + 0, .26xl0 ⁶

0° 20° cos20 : x(1- •cos20) x(1-cos20) h,(mϊls) h, [mm] ^" 10 20 .94 0 .2 x 10 ⁶ .46 x ^' I0 ⁶ 1.48 [0.038]

20 40 .77 0.06 .50 1.41 [0.036]

30 60 .50 .13 .57 1.33 [0.033]

40 80 .17 .22 .66 1.23 [0.031]

50 100 -.17 .30 .74 1.16 [0.029] 60 120 -.50 .39 .83 1.10 [0.028]

70 140 -.77 .46 .90 1.05 [0.027]

80 160 -.94 .50 .94 1.03 [0.026]

90 180 - 1 .52 .96 1.02 [0.026]

- -

3 -4

P = FU/6.6x10 = 1.5x10 FU horsepower [multiply numbers by

0.11 to obtain MKS]; P = 2.3x10~ ⁹ U ² /h _f [multiply- ^' numbers by 0.11 for,MKS]. n ' θ U(0) U ² (θ) h _f (θ) U ² /h _f 2'.3x10 ^"9 xU ² /h _f

-10° 220 0.5x10 ⁶ 1.5x10 ^~3 0.03x10 ⁹ 0.07 ° 0 0 1.4 0 0 10 220 .05 1.3 .04 .09

20 440 .19 1.2 .16 .37

30 650 .42 1.2 .35 .80

40 800 .64 1.1 .58 1.3 50 950 .90 1.1 .82 1.9

60 1050 1.1 1.0 1.1 2.5

70 1100 1.2 1.0 1.2 2.7

9.8 hp

[9.8 hp/9]x[90°/720°] = 0.14 hp mean.

EXAMPLE 5: Friction of skirt 30 (fig. 1) The same parameters as for example 1 are used, namely

H = height of squeeze surfaces = ^~ inch [12| mm] skϊpt The friction F is the sum of that from shear and from the stretch-effect (see O'Connor & Boyd, loc. cit., page 3.5), namely

F ^Skϊrt =2W,L[H/h + 2(q(a) + q(a')) (1 m _32c + 1/m^ )]U ski PΓ and the power P' equivalent to such friction is merely UF , which, if U is inch/second [0.025 m/s], must be divided by 6.6x10 to be expressed in horsepower [or by 750 if the MKS system is used], and where the term q(a) + q(a') , with a and a' resp. referring to the left-hand and right-hand si - des of the bearing surfaces of fig. 1 , is almost constant considering that h va¬ ries only slowly for the operational conditions considered, said term being about 1.1 for 5000 rpm; h in the expression above is taken as h= 1.2 mils [0.030 mm]. The mean power P' can be computed by integrating with respect to 0 from zero to U , as U is a known function of 0, (see L. Lichty, Combus-

2 tion Engine Processes, 1967, McGraw-Hill, p 585), whereby U can be appro¬ ximately written as

U ² = [4 N/60]x[sin ² θ + _- sin ² θcosθ ] Upon substituting the numerical values above and integrating with respect to θ, one obtains

P' =(1.2/τ) f (sin ² 0 + isin ² θcos0)d9 = 0.6hp Before comparing the overall friction of the piston of the invention with a conventional one, it is worth noting that the skirt of the piston of the inven¬ tion , besides providing an attitude control (fig. 1), also is less frictional than the conventional skirt. To that end fig.8 (from E.F. Obert, "internal combustion engines and air pollution", 1973, intex, page 486), shows the empi¬ rical friction curve of an engine with cylinders and strokes comparable to the engine considered above. At 5000 rpm the total measured friction is about 54 hp or 9 hp per piston. It is generally assumed that the conventional skirt frϊc- tion is about 1/5 - 1/4 of the total friction, so that the skirts of fig.8 would be frictional by about 1.8 - 2.3 hp compared with the figure of 0.6 hp determi¬ ned above, in this example for the piston of the invention.

EXAMPLE 6: Power-gain and Fuel-Economy of the Invention C Coonnssiiddeerriinngg tthhee ccrroowwnn ffrriiccttiioonn ffiirrsstt,, iiff pp //ww iiss ccoonnssiiddeerreedd eessssen-

CCC tially constant — a reasonable ; aassssuummppttiioonn —— ,, tthheenn., tthhe crown frictional power

2 lloossss PP'' iiss proportional to N , where N = rpm, whereby, using the results frcm example 5,

^(5000- rpm)= 0.61 hp/pϊston ^(2500 rpm) = 0.15 hp/pϊston ^~ P- (1000 rpm) = 0.024 hp/piston.

Considering next the power loss P in the skirt, which is also es-

2 sentially proportional to N , and using the results from example 4, we obtain

P(5000 rpm) = 0.14 hp/pϊston P(2500 rpm) = 0.04 hp/piston P (1000 rpm) = 0.005 hp/pϊston.

The power losses due to friction in the engine equipped with the pistons of the invention can now be compared with those experimentally obser¬ ved for a conventional engine, for instance the V-6 of fig.8. The P' and P values listed above must be added for each rpm and then be multiplied by 6, the number of pistons. Table 2 immediately below provides this comparison:

Table 2 Engine speed 5000 rpm 2500 rpm 1000 rpm conventional power loss 54 hp 23 hp 3.8 hp invention's power loss 4^ hp 1.2 hp 0.2 hp invention's power saving 50 hp 22 hp 3.6 hp conventional brake power 130 hp 95 hp 30 hp power gain (% fuel gain) 38% 23% 12%

{ O

The total useful power gain , as a percentage , is shown in dash-dot lines in fig. 8.

The improvements shown in table 2 apply to a hot engine. A cold engine when started remains inefficient for about 5 miles or more in the conventional case, to the point of using twice the fuel of what it would when hot . i f commu¬ ting travel is assumed, say 20 mi les [32 km] one way, further that the hot-engine fuel consumption in such stop-and-go traffic is 15 mi les/gal lon [6.4 km/l iter] whi le only 7 ^~ \ [3.2] for a cold engine , and if further it is assumed that the heat- barrier recommended by this invention in addition to its primary function also cuts the inefficient start-up phase in two , whereby only 2 ^~ mi les [4 km] would be travel led under such adverse conditions , the gain in fuel economy from that fact alone would be about 12%; as the engine speed for such traffic would be somewhere between 1000 and 2500 rpm , there would also be a gain in efficiency somewhere between 12% and 26% (see table 2) , say 19%, from the reduction in friction made possible by the invention , and therefore the total gain in commu¬ ting travel efficiency would be about 31 %t end description

_0-V.PI λf- WIPO κV