and some more
ROD LENGTH RELATIONSHIPS*
(with Crank @ 90 deg ATDC)
Piston Position Crankpin/Rod AngleStrokeRod LengthRod Anglefrom TDCATDC3.55.7017.882.025184.108.40.20617.402.01872.593.56.0 016.962.01173.043.56.2016.392.00273.60
Table 2StrokeRod LengthRod AnglePiston DistanceCrank AnglePiston Accel3.55.7017.071.48772.932728.353.55.8516.651.49473.3 52504.723.56.0016.261.50073.742324.263.56.2015.761 .50874.242097.27
ROD LENGTH RELATIONSHIPS with CRANKPIN/ROD centerline @ 90o @ 7500 rpm
*data from Jere Stahl Another concern in selecting the rod length is the effects of mechanical stress imposed by increasing engine speed. Typically, the concept of mean piston speed is used to express the level of mechanical stress. However, the word "mean" refers to the average speed of the piston in going from the top of the bore to the bottom of the bore and back to the top of the bore. This distance is a linear distance and is a function of the engine stroke and engine speed, not rod length. Therefore, the mean piston speed would be the same for each rod length listed in Table 1.
Empirical experience; however, indicates that the mechanical stress is less with the longer rod length. There are two reasons for these results. Probably the primary reason for these results is that the profile of the instantaneous velocity of the piston changes with rod length. The longer rod allows the piston to come to a stop at the top of the bore and accelerate away much more slowly than a short rod engine. This slower motion translates into a lower instantaneous velocity and hence lower stresses on the piston. Another strong effect on mechanical stress levels is the angle of the connecting rod with the bore centerline during the engine cycle. The smaller the centerline angle, the less the side loading on the cylinder wall. The longer rod will have less centerline angle for the same crank angle than the shorter rod and therefore has lower side loadings.
Classical textbooks by Obert ( ) and C.F. Taylor ( ) provide little guidance on the rod length selection for passenger or commercial vehicles other than to list the ratios of rod length to crank radiuses that have been used by various engine designs. Race engine builders using production blocks have done quite a bit of experimentation and have found many drivers are capable of telling the difference and making clear choices along with similar results from motorcycle flat track racers/builders.
Because of recent developments in computer modeling of the engine cycle by R.D. Rabbitt ( ), another factor may be critical in selecting a given connecting rod length. This new factor is the cylinder head flow capability versus connecting rod length over stroke ratio (l/r) versus engine speed. To understand this relationship, let us first review previous techniques used to model air flow during the engine cycle which as Rabbitt points out is founded on principles initiated in 1862 and refined in 1920. These theories are documented in Taylor's textbook ( ). To calculate air flow throughout the cycle these models use such parameters as mean or average inlet mach number for the port velocity and an average inlet valve discharge coefficient which compensate for valve lift and duration. In these models a control volume is used to define the boundaries of the combustion chamber. The air flow determined by the previous parameters crosses this boundary to provide air (and fuel) for the combustion process within the control volume.
However, this control volume has historically been drawn in a manner that defines the boundaries of the combustion chamber in the area of the inlet and exhaust valves as if the valves were removed from the cylinder head (ie. a straight line across the port). With the valves effectively removed, the previously mentioned average port flow and valve discharge coefficient (ie. valve restriction) are multiplied within current computer models to quantify the air flow (and fuel) delivered for each intake stroke. But, as Rabbitt points out, this approach totally ignores the effect of the air flow direction and the real effect of valve lift on the total air flow that can be ingested on each intake stroke.
Rabbitt reaches two important conclusions from his study. One, because of the direction of the air flow (angle and swirl) entering the combustion chamber, three dimentional vorticies are set up during the intake stroke. Two, that above a certain piston speed, density of the mixture at the piston face is a function of valve geometry and valve speed. Rabbitt further discusses the effect of the first conclusion as it relates to the mass of air that is allowed to flow through the port and by the valve. Vorticies can exhibit different characteristics and in general conform to two general types--large scale bulk vorticies that could be described as smooth in nature and small scale eddies that are highly turbulant.
If one can consider that the vacuum produced by the piston on its downward travel to be the energy that causes the air to flow through the port when energy losses throughout the intake tract (including losses at the valve) are at a minimum, the flow delivered to the chamber will be maximized. If the area between the piston face and the valve is also included in the consideration of flow losses, the effect of the type of vorticies created can be more easily understood. Large scale bulk vorticies comsume less energy than highly turbulent eddy vorticies. Thus, more of the initial energy from the piston's downward movement is available at the port-valve-combustion chamber interface with which to draw the intake charge into the chamber. Small scale eddies eat up energy which reduces the amount of the initial energy that reaches the port-valve-combustion chamber interface which in turn, reduces the port flow.
Rabbitt's second conclusion follows that at some higher piston speed, the vorticies within the combustion chamber (which are assumed to be large scale bulk type at low speeds) transition from the bulk type to the small scale eddy type. At this point the flow into the combustion chamber ceases to increase in proportion to increases in engine speed. It is theorized that this flow transition point can be observed on the engine power curve as the point at which the power curve begins to fall off with increasing engine speed.
As indicated earlier, piston speed is normally viewed as mean or average piston speed. Thus for a given engine, the mean piston speed increases as the rotational engine speed increases. However, in Rabbitt's model the piston speed of concern is the instantaneous piston speed during the intake stroke near TDC. For any given engine, changing the rod length to stroke (l/r) ratio changes the instantaneous piston speed near TDC. For the purposes of flow visualization, the type of vortex formed should not care whether a given instantaneous piston speed had been achieved by a given rotational speed or changing the (l/r) ratio and operating at a new rotational speed. As long as the instantaneous piston velocity is the same, the type of vorticies formed should be the same and the amount of air inducted into the cylinder should be the same.
If other factors influenced by rotational speed such as the time distance between slug of intake air flow and valve opening rates relative to the acceleration of the air slugs were ignored, one should be able to predict the location (RPM) of the peak power as a result of a change in the (l/r) ratio. Note, that even though power is a funtion of air flow and air flow should be roughly constant for the same instantaneous piston speed (neglecting the afore mentioned factors), the power may not be the same because of the lever arm effect between the crank radius and the connecting rod. (As we noted earlier, the shorter rod should have the advantage in the lever arm effect.)
In reality, the analysis must be viewed by stroke (ie intake, compression, exhaust, power) the selection of exhaust valve opening time combined with the exhaust system backpressure and degree of turbulance the exhaust port experiences. If the exhaust port has good turbulance control then you may run a shorter rod which allows you to use more exhaust lobe which reduces pumping losses on the exhaust stroke.