An in-depth look at rotating & reciprocating assemblies

Engine speed limits come from two sources primarily. Either you're limited by the bottom end, bearings rods and pistons, or the cylinder head valve train components etc.

It's a bit easier for me to start with the bottom end as the parts are cheaper, so I bought a used piston/rod assembly from a K24A2 RBB motor, and a matching one from a K20 (PRB). To know some of the forces and loads involved here, I needed weights, so I cleaned the carbon and residues off the parts and weighed them. Measurements are in grams. Anything below 50g used a very fine resolution scale, but my scale for heavier measurements only goes to 1g resolution.



Interesting to note that the K24 piston (316g) is lighter than the K20 (335g). The K20 was in much worse shape with respect to carbon though, so there could be a few grams coming from what I couldn't get out of the ring grooves. The K20 piston also has a much larger dome, so there's probably a bit of mass in that.

The split weights for the conrods are pretty rough approximations, I haven't made a split weight rig yet, just using a clock stand to support the big end and eyeballing the level of the small end resting on the scale like this:



All the weights represent only one data point, so I'd expect some variation engine to engine, but it's representative at least of what would be running in the engine.

Looking at the wrist pins, both are 2.050" long with a 0.150" wall thickness, comparable to something on the shorter and lighter end from JE.

For the sake of the load calculations, I'm throwing the advertised rev limits into a calculator I've built. 7100 rpm for the K24, and 8600 rpm for the K20 (highest I could find). I wanted to start here because it's a known value, how fast can the K24 be run until it matches the loads seen by the K20. My comparisons are at several places. I want to look at the wrist pin load, stresses in the connecting rod, rod bearing loads, and finally the main bearing loads.

For reference, 0 and 360 degrees represent TDC. I'm looking specifically at the TDC event during overlap, between the exhaust and intake strokes. This is where the connecting rod is most heavily loaded in tension. TDC between compression and combustion always has gas forces which reduce the tension load on the connecting rod, these forces aren't in my calculations yet. I can get into that at a later date.



Corresponding pin loads:


So the K20 sees a load of 20,344N (4,573 lbs), and the K24 is down at 15,505N. Increasing the redline of the K24 to 8,132rpm gives the same pin loading as the K20.



The differences in pin load forces around BDC are due to the geometry differences, stroke and rod length.

Looking at the main bearing loads, they are higher because the pin loads only have to accelerate the piston, pin, circlips and ring pack. The main bearings have to worry about both the rotating and reciprocating mass of the conrod and the side forces caused by the angle of the conrod relative to the bore. The forces on the main bearings due to the rotating part (big end) of the conrod can be offset somewhat by the counterweights, but I don't have a crankshaft to cut up and measure how much counter weighting a K20 or K24 has so I'll look at a 0% balance factor and 100% balance factor.



You can see that if the rotating component of the rods is balanced with crankshaft mass on the counterweight (100% balance factor, or BF), the loads on the main bearings are significantly reduced.

So if the crankshaft has a 0% balance factor, the K24 can spin up to 7740 rpm before the main bearing loads equal:

With a 0% balance factor, the K24 can spin up to 7740 rpm before the peak loads are the same:


Or with a 100% balance factor, the K24 can spin to 8000 rpm before the main bearing loads are equal:


Coincidentally, the peak forces that the connecting rod big end bearings experience are the same as the main bearings would if the crankshaft were not counter weighted (0% BF). This is really the only number of concern for rod bearings because it will correspond with the smallest oil film thickness.

So now we know that in terms of bearing oil film thickness and red line, the K24 at 7740 rpm is the same as a K20 at 8600 rpm.

I've seen reference to "A K20 at 9000 is the same as a K24 at 8000 rpm". Well I tested that too and it's actually 8100 rpm with the component weights I measured.

I'm still thinking about how to approach conrod stresses. Both appear to be nice forged pieces, but the K24 rod has much less material meeting the big end, but overall a thicker and wider I-beam profile.