Axial flow cooling for air-cooled engines
Cooling of existing air cooled engines uses fins and air flow essentially perpendicular to the axis of piston travel. By orienting the fins and air flow to be largely parallel to the axis of piston travel, far superior cooling can be achieved. The air flows thru grooves of relatively constant cross section for the entire flow path. All parts of the fins are reached about equally by the passing air. Both features contribute to improved heat conduction into the cooling air. Because air velocity over the fins is reasonably constant everywhere, less power is required to force the air thru the fin structure. Since the cooling power is ultimately taken from the engine output, this results in greater power being available to do productive work.
This application claims priority to co-pending U.S. provisional patent application No. 60/505,683, filed on 25 Sep. 2003, and incorporated herein by reference in its entirety.
RELATED APPLICATIONSNONE
FEDERALLY SPONSORED RESEARCH OR DEVELOPMENTNONE
BACKGROUNDIn general, air cooled engines have fins that require the cooling air to flow in a direction perpendicular to the axis of the cylinder. In a single cylinder engine, or an engine with a single bank of cylinders (such as a V2, a horizontally opposed 2, or a single bank radial), this is not a bad configuration. There is adequate space for fins and cooling air. Even a double bank radial works fairly well because the cylinders of the rear bank are oriented between the cylinders in the front bank, so there is adequate access for the cooling air to reach the aft cylinders. In the horizontally opposed air-cooled engines with 4 or more cylinders that are commonly used in automotive and aircraft applications, having the fins oriented perpendicular to the axis of the cylinders is a serious disadvantage. The air flow must be oriented parallel to the fins. That means either (1) the air flow thru the fins must be parallel to the axis of the crankshaft, which provides less cooling for cylinders behind the front cylinder, or (2) it must be perpendicular to both the crank axis and the cylinder axis, which is the orientation used in all modern applications.
It takes power to drive cooling air thru the fin structure. Ultimately this power must come from the engine being cooled. This reduces the useful power output of the engine, and the net efficiency of the engine. There has been a lot of effort for more than a century in designing efficient fin configurations for cooling engines with a minimum of lost power. To get efficient cooling, it is desirable to have the cross section of any given gap between fins to stay a constant area as the air passes thru the engine. With this configuration, the air moves with a constant velocity and a minimum of power is required to provide a given amount of cooling. In addition, the path length thru the engine should be minimized to maintain a thin boundary layer between the fin and the moving air.
Now consider the situation in standard down-draft (or up-draft) cooling where essentially all the air must pass between the cylinders. The air enters above the cylinder and head, which are typically 10 to 20 cm wide. Then it passes thru the gap between the cylinders, typically 1 to 2 cm wide. Then it is blown out beneath the cylinders and heads, again 10 to 20 cm wide. With careful duct design, the cooling air can be guided around the engine to pass over most of the fins, but the restriction at the passage between cylinders always increases the pressure required to force sufficient air thru the engine. The power required is the product of pressure times volume flow rate. The volume flow rate is fixed by the cooling requirements of the engine. If there is a restriction in the flow path that has half the area of the rest of the path, the flow velocity at that point will be twice as high as the velocity over the rest of the fin. Since pressure drop increases approximately with the square of flow velocity, the pressure drop per unit distance of air travel in the restriction will be four times as high as in the rest of the engine. With the air flow passing down between the cylinders, this is unavoidable. The result is excessive power required for cooling (very undesirable), and the possibility insufficient cooling under some or all operating conditions (even more undesirable).
It does little good to try to cool the engine from the “top” (further from the crank shaft). The rocker arms sit on top of the engine and that assembly introduces so much thermal impedance that it is impractical to cool the heads by using fins over the rocker arms. Porsche has developed a head in which the two valves are one above the other, as opposed to side by side, giving more space for fins and air passages between the heads. This requires a tricky valve linkage, and does nothing for the flow restriction between cylinders and the base of the heads.
For a specific example of present cooling problems, consider the Jabiru engine, built in Australia. The Jabiru has several desirable characteristics. It is a very compact engine for its power rating. Largely as a result of this, it is considerably lighter than other engines of similar power. Also, the small size makes the structure strong. Size and weight are important in many applications, and critical in aircraft. Strength is always desirable. A 6 cylinder Jabiru rated at 130 horsepower (100 kW) is essentially the same size as, and lighter than, a 4 cylinder Volkswagen producing half the power. There is no free lunch. The cost of the reduced size and weight of the Jabiru engine is that the compact design makes it essentially impossible to cool the engine when operated at rated power. Thru the remainder of this discussion, the Jabiru engine will serve as the model. However, all the results from this analysis of the Jabiru engine are obviously applicable to other in-line and horizontally-opposed air cooled engines.
Now consider the situation faced by the cooling air. The air typically enters at the top of the engine and flows down over the fins of the head and cylinder (downdraft cooling). The argument does not change much if the direction of flow is up from below the engine (updraft cooling). The air enters the fin structure in a region where the fins are typically 30 mm high and is squeezed between the cylinders where, in the case of the Jabiru engine, the fins are only 5 mm high. Thus, the air has to travel 6 times as fast while it is between the cylinders, which requires 36 times as much pressure drop per unit distance traveled, and 36 times the power per unit distance of flow. Ultimately, this power comes from the engine, and decreases the power available to do useful work. The problem is intensified in the Jabiru engine, where the use of six head bolts means that there is a long path length where the air must travel at high velocity. Also, when forcing air to flow around a cylindrical obstacle, the air flow tends to leave a dead air zone ahead of the center of the cylinder, and a much bigger dead air zone behind the center of the cylinder. Careful use of ducts to guide the air will reduce the sizes of these dead air regions, but it cannot eliminate them entirely. Another problem is that the conductivity of heat from the the metal fin to the air increases with increasing air velocity. Where the air moves slowly, a thick boundary layer forms, and conductivity into the air is low. In the situation shown in
Now consider the path length of the flow thru typical fins. In round numbers, this path length will be π times the average radius of the cylinder fins. If the cylinder has a bore of 100 mm, that is a radius of 50 mm. The cylinder wall has a thickness of about 5 mm, and the head must surround that by about an additional 5 mm. Thus, the radius to the base of the fins will be about 60 mm. If the fins are 30 mm high, the average radius of the fins becomes 75 mm. That gives a flow path length of 235 mm. This is a much longer path length than is desirable from purely thermodynamic considerations. Typical automotive radiators have path lengths of under 50 mm, and they usually have staggered fins within that distance. Aircraft oil coolers typically have air path lengths of 15 to 20 mm, with staggered fins within that distance. A flow path length of 235 mm is asking for thick boundary layers and lousy conductivity from the fin to the air.
BRIEF DESCRIPTION OF THE ILLUSTRATIONS
All other figures depict engines, or parts thereof, using axial flow cooling.
Most of the problems of both updraft cooling and downdraft cooling are eliminated by a novel cooling configuration, henceforth referred to as axial cooling, with the flow of cooling air traveling essentially parallel to the axis of the cylinder. Fins on the cylinder and head are oriented so they are essentially parallel to the axis of the cylinder. Cooling air is injected into the fin structure near the rocker arms. From there it flows over the head fins, then the cylinder fins toward the crank case. Ducts contain the air within the fin structures. The ducts may terminate some distance from the crank case, allowing the warmed air to escape. Better, the ducts may guide the warmed air toward the cooling air outlet, where it may be accelerated out of the engine compartment using exhaust augmentation. It is not necessary for the fins on the head to be aligned with the fins on the cylinder. In fact, it is undesirable for the two sets of fins to be aligned. Having a discontinuity in the fins between the head and cylinder disrupts the boundary layer, yielding improved heat transfer from the cylinder fins to the air. It is possible to pump the air in the opposite direction, from near the crankcase to the rocker arms. But, thermodynamically it is better to pass the coldest air over the region with the highest heat loading, the exhaust ports, near the rocker arms. Also, the ducts are easier to make and install if the air flow is from the rocker arms toward the crank case.
DETAILED DESCRIPTION OF THIS INVENTIONFor purposes of illustration, the figures show axial cooling adapted to the Jabiru engine. Similar adaptations can be made to other air cooled engines. The compact design of the Jabiru engine provides a severe test for any cooling scheme. If it will work with the Jabiru engine, it will work with any engine.
The entire periphery of head (20) is covered by fins (30), said fins being divided into several functional groups. Combustion chamber fins (32) cool the dome and upper edge of the combustion chamber. Exhaust side fins (33) cool the very high heat load of the exhaust port and the exhaust side edge of the combustion chamber. Intake side fins (34) primarily cool the intake side edge of the combustion chamber, the intake port not generating any heat. Bottom fins (35) cool the bottom edge of the combustion chamber. Notice that bottom fins (35) are cut out in two regions (24) to provide clearance for the push rod tubes (not shown). This leaves the fins in the corner (36) which provide some cooling to that edge of the combustion chamber but whose primary purpose is to provide a suitable quantity of cooling air to the region of cylinder (10 in
In addition to the fins shown here, it is entirely possible, and desirable, to drill a set of holes vertically thru the metal separating the intake and exhaust ports. Such holes, properly aligned, can provide the air flow to the central couple grooves between combustion chamber fins (32). Blocking, or partially blocking, the entry to the grooves between these central fins, near rocker arm housing (21 in
Since the exhaust port quadrant of the head has about twice the thermal loading of any other quadrant, it might seem reasonable to make the head nonsymmetrical around the axis of the cylinder, with longer fins on the exhaust port side than on the intake port side. In fact, early Jabiru engines were made that way. Apparently Jabiru learned that this did not work well. Actually, that approach is counter productive. It results in little or no cooling on the edge of the combustion chamber at the intake port side. In addition, in an axial flow cooling system, as will soon be described, it results in lower air speeds over the fins on the the exhaust port side, with little or no increase in heat dissipation.
It is blatantly obvious that intake side fins (34) have no direct access to the cooling air input. This is a significant part of the design, not an unforeseen problem. The function becomes obvious in the discussion of
Bottom fins (35) have a flow path length of only about 20 mm. They cool little more than the bottom edge of the combustion chamber. The volume between bottom fins (35) and rocker arm housing (21) is occupied by the intake pipe. Bottom fins (35) cannot be made longer on this side. There is no real need for the cooling on this side anyhow. On the exhaust side, where more cooling would be very desirable, the exhaust pipe does not allow a longer flow path thru bottom fins (35). In the middle, hidden behind rocker arm housing (21), resides the sixth head bolt. Access for machining that area and installing that bolt does not allow a longer flow path length for bottom fins (35) in that region. Although bottom fins do little cooling, they are necessary for delivering a proper quantity of air to cool the bottom of cylinder (10 in
The dramatic thing shown in
The air paths are more restricted where the air must flow around the cylinder bolts (hidden within the fins) at the bottoms of head bolt cutouts (23). As exhaust side fins (32) pass head bolt cutouts (23), the fin height is small to nonexistent. Closer to cylinder (10 in
The air path thru the side fins is clearly shown in
At one end of the engine, exhaust side fins (32) will not have mating intake side fins (33) to carry cooling air past the edge of combustion chamber (42). At the other end of the engine, there will be no exhaust side fins (32) to supply intake side fins (33) with cooling air. If the cooling air duct fits tightly to the heads at the ends of the engine, then exhaust side fins (32) suffer a severe restriction in their air flow path, and intake side fins (33) will receive no cooling air. It is a simple matter to shape the cooling air ducts to provide suitable passages and air flow to resolve this situation.
Note that both the head (as shown in
Comparing
To greatly reduce this effect, the leading edges of cylinder fins (11) are sharpened, as shown in
It is also desirable to sharpen the trailing edges of the head fins. Similarly, it is desirable to offset and sharpen the edges of exhaust side fins (32) and intake side fins (33) to break up the boundary layer and improve flow between them. However, both these steps are significantly more difficult to implement, and the improvement is significantly less than is gained with sharpened cylinder fins. Consequently it is probably not worthwhile except in extreme conditions such as airplane racing.
Ducts for containing cooling air within the axial flow fins are considerably simpler than ducts presently used in low drag down-draft cooling installations. A three piece duct is adequate for each side of the engine. One piece extends over the cylinders and heads from the crank case to the spark plugs. A second piece extends under the cylinders from the intake and exhaust pipes to a few cm from the crank case, leaving ample space for the hot air to exit downward. Ideally, this second piece is connected to a plenum under the engine which guides the heated air toward the outlet port. The third piece of duct extends from the spark plugs, over the rocker covers, to the intake and exhaust pipes, connecting to the first and second pieces. The third piece also incorporates the nozzle that picks up the intake cooling air. A wide variety of duct configurations will work with an axially cooled engine. This is just one example of a simple, effective duct that allows easy access to the engine.
Cooling ducts for axial flow cooling are not shown. The required shape of such cooling ducts is obvious. An optimized duct will contain a diffuser and turning vanes. These features are well known in duct design.
DefinitionFin Efficiency: The heat flow rate from a given fin divided by the heat flow rate that would occur if the fin had an infinite thermal conductivity, all other parameters remaining unchanged.
High fin efficiency sounds good, but it requires a lot of fin material and large volumes of space in which to locate it. In real situations, for almost any definition of “optimum”, the most desirable fin efficiencies are in the vicinity of 70%.
Experimental ResultsThermal measurements on a Jabiru engine in a laboratory indicate that it is probably impossible to operate the engine continuously at rated power in any existing airplane. These measurements, with very good ducts distributing cooling air over the heads, indicate a quiescent temperature rise of 300 C when operated in a normal small plane. There is not enough pressure available at speeds small airplanes can realistically achieve to force enough air thru the engine to keep it cool, especially on a warm day. It will barely survive at 70% cruise power. The condition of used heads from Jabiru engines is testimony to the difficulty of keeping the engine at a reasonable operating temperature. There are two major problems, fin area and air flow velocity. A Jabiru head has about 1500 cm2 of surface area actually exposed to moving air, and over much of that area the air movement is sluggish. The fin efficiency of the Jabiru engine is very high, above 95%, but this does little good because there is so little fin area and the cooling air cannot effectively reach much of the area that does exist.
By comparison, the head demonstrated herein has a fin surface area of about 2700 cm2, 1.8 times as much, all fit within the same volume. More important, the air flows evenly over essentially all of that area cutting thermal resistance from the fin to the air to a fraction of the value in the present Jabiru engine. This is important, but even more important is the fact that the pressure required to drive air thru the head fin assembly is dramatically lower. Tests on a flow bench with a mockup of the head fins demonstrated herein show that for any given volume flow rate, the pressure drop of this design is well under 10% that of the present Jabiru design. Thus there is adequate air pressure available to cool the head to a comfortable temperature even at full power and low speeds, as in a prolonged, steep climb. Not only does this promote engine reliability, it dramatically reduces cooling drag, which takes a significant fraction of the total engine output power. The result is higher speed and better fuel economy.
The fin efficiency in this design is over 75%. This is a little higher than optimum, but cutting thinner fins and making more of them would be difficult.
Range of ApplicationsThis patent is a detailed description of axial cooling applied to the Jabiru engine. The Jabiru was selected as a model because its compact design exacerbates the cooling problems present in all air cooled engines. Clearly a similar design process will result in improved cooling for other existing engines, and new engines that may be designed in the future.
Axial cooling is most applicable to horizontally opposed engines of more than two cylinders. It can be used in 2-cylinder horizontally opposed engines, but the advantages are limited. At this time, there are very few in-line air cooled engines. Axial cooling is certainly appropriate for in-line engines, if anyone is interested. Most V engines of more than two cylinders are water cooled, but an axial-flow, air-cooled V engine is certainly possible, with air entering above the engine, flowing down thru the head and cylinder fins, and exiting at the two sides of the crank case. In radial engines, as in 2-cylinder horizontally opposed engines, axial flow cooling could be used, but the advantages are limited. The real advantages of axial flow cooling occur in configurations where there is limited space for air to pass between adjacent heads and cylinders, and it is advantageous in any such engine.
Axial flow cooling is applicable to stationary installations with a fan providing the driving power to the cooling air, to mobile installations where the motion of the vehicle causes the cooling air to flow over the engine, and to mobile installations where a fan (or propeller) is used to augment the airflow caused by the motion of the vehicle. As mentioned above, Exhaust augmentation of the cooling flow is also possible and desirable with axial flow cooling.
Claims
1 An air cooled engine with at least two cylinders, in which more than half of the head cooling fins are oriented within 30° of being parallel to the direction of piston travel.
2 An air cooled engine as in claim 1, in which more than half of the head cooling fins are oriented within 20° of being parallel to the direction of piston travel.
3 An air cooled engine as in claim 1, in which more than half of the head cooling fins are oriented within 10° of being parallel to the direction of piston travel.
4 An air cooled engine as in claim 1, in which more than half of the head cooling fins are oriented within 5° of being parallel to the direction of piston travel.
5 An air cooled engine as in claim 1, in which more than half of the head cooling fins are oriented within 2° of being parallel to the direction of piston travel.
6 An air cooled engine as in claim 1, in which more than half of the head cooling fins are oriented within 1° of being parallel to the direction of piston travel.
7 An air cooled engine with at least two cylinders, in which more than half of the cylinder cooling fins are oriented within 30° of being parallel to the direction of piston travel.
8 An air cooled engine as in claim 7, in which more than half of the cylinder cooling fins are oriented within 20° of being parallel to the direction of piston travel.
9 An air cooled engine as in claim 7, in which more than half of the cylinder cooling fins are oriented within 10° of being parallel to the direction of piston travel.
10 An air cooled engine as in claim 7, in which more than half of the cylinder cooling fins are oriented within 5° of being parallel to the direction of piston travel.
11 An air cooled engine as in claim 7, in which more than half of the cylinder cooling fins are oriented within 2° of being parallel to the direction of piston travel.
12 An air cooled engine as in claim 7, in which more than half of the cylinder cooling fins are oriented within 1° of being parallel to the direction of piston travel.
Type: Application
Filed: Sep 27, 2004
Publication Date: Mar 31, 2005
Inventor: Clifford Cordy (Reno, NV)
Application Number: 10/950,371