Discussions, even debates, have taken place amongst engine builders on the topic of rocker arm geometry. Some friendships have undoubtedly been strained from the resulting disagreements on the best way to proceed. This series of technical articles will present the problem in a way that I’ve never seen presented in any technical article before. We will consider the question of: Just what is proper rocker arm geometry for a pushrod-actuated valvetrain? Specifically, we will look at the Big Block Chevrolet as used in high performance powerboats. A similar study could of course be done for any engine, and the results would be slightly different from those presented below.
Where is the Force?
In any engine, not just one for a go-fast boat, we have to ask ourselves the question: how is my customer going to use this engine? What is his duty cycle going to look like? Where is he going to spend most of his time? I think most performance marine engine builders would agree that, if they’ve done their jobs well, the customer will be either idling in the no-wake zone or above 3000rpm. In the engine valvetrain, the forces at each condition are very different, and they occur at a different point in the cycle. I’ve made this point to a few colleagues and they seemed interested in learning more. In order to support that discussion, I had to generate some graphs in my favorite spreadsheet utility to aid the discussion.
What you will see below is a simple physics-based model of valve-side loads (as opposed to the cam side of the rocker arm) that considers a typical Big-Block Chevrolet valvetrain, a valvespring that would be typical of one used in a go-fast boat engine, and a mechanical roller cam profile that is on the “aggressive” side for something with 225° of duration @ .050 and 0.700” valve lift with a 1.7:1 rocker.
| System Equivalent Mass (g) | 454.2 |
| Spring Seated Load (lbf) | 140 |
| Spring Rate (lbf/in) | 600 |
| Lash (in) | .018 |
| Intake Centerline Angle (° crank) | 108 |
The variable for this study will be engine speed, and we’ll look at several different ones to get a clearer picture of how the load changes location in crank angle as we proceed through the RPM range. For a discussion on System Equivalent Mass and how to calculate it, have a look at the series on valvetrain frequency.
Key to understand is that valvetrain forces are a function of two opposing forces:
- The force generated with the cam profile acceleration multiplied by the equivalent mass of the system.
- In some parts of the profile, this acceleration is positive, and adds to the force of the valvespring. In others, the acceleration is negative (We can call this the “nose” of the profile). This is where we rely on the valvespring to hold the valvetrain together.
- The valvespring force
To be clear, this is a simple kinematic analysis, which means stiffness is not considered here. There are effects happening (driven by what we talk about in the series on valvetrain frequency) that are beyond the scope of this discussion, and they are significant to getting a good functioning system. Tread with care!
Case 1: Idle
At idle speeds (700-1200rpm), the forces in the valvetrain are driven by the valvespring, which has a peak load of 550lbs at peak valve lift. You can see that inertia force, plotted in tan, is nearly zero. The spring force, in light green, is hard to see underneath the Net Valvetrain Force, in black. In the background, I’ve plotted valve lift because this is something most of us can identify with. The valve lift curve will give us a reference for where in the valve event we are. Of course, everything is plotted in the crank angle domain.
Case 2: 2500rpm
At speeds around where a boat would be just coming on plane, we can see that the valvespring is starting to work a bit. The net valvetrain force is 75lbs under the spring force, while we’re just starting to see the acceleration peaks of the profile show at the beginning and end of the flanks of the profile.
An off-topic side note: If you’ve ever wondered why flat-tappet cams are broken in off of idle speed, the answer is in these first two plots. The sweet spot between 2500-3000rpm is the “sweet spot” for valvetrain loading, where peak load is at a minimum.
Case 3: 3500rpm
This is a potential cruising engine speed, so somewhere the customer might spend a good bit of time. For the first time, the inertia force (mass times acceleration) is starting to dominate the load. The peak valvetrain loads occur at about 20% of max lift, on both the opening and closing sides. It peaks at 476lbs. Over the nose, there is 406lbs of load on the valvetrain, a 144lb reduction of load in this area versus idle speed.
Case 4: 4500rpm
At this speed, the valvetrain load is again inertia dominated, and the net force is now in excess of the valvespring load. The peak load here is around 640lbs. Again, max load is approximately at 20% of max lift.
Case 5: 5500rpm
As we push the sticks further forward, the inertia force grows and brings the net valvetrain force to 843lbs, and it’s still at 20% of max lift. Now, the load at max lift has reduced to 195lbs, and the minimum load is 133lbs. It’s here when we’re close to maxing things out were this a hydraulic valvetrain, as we need to maintain some load on the leakdown plunger in the lifter for it to function properly. Since this is a mechanical valvetrain, we’ll push it a bit further.
Case 6: 6000rpm
This is a speed few spend much time at. Peak valvetrain loading is 960lbs, again at 20% of max valve lift. Minimum loading over the nose is around 78lbs. At this speed, lifter pumpup would either have already happened, or would be imminent. Our valvespring is about maxed out, considering that by now, it is “surging” which reduces the load it has available to control the valvetrain. The situation is critical.
Case 7: 6600rpm
At this speed, we’re done without doing development. The net valvetrain load reaches a minimum of zero, and a maximum of over 1100lbs. Valvetrain engineers call this the “Loft” point. Running reliably at this speed and above takes many hours of engineering, spin rig development, and confirmation testing. Not to mention many thousands of dollars spent. This is where the experts live.
Takeaway:
This is the complete picture through the RPM range. The takeaway is this: the max loading on the valvetrain occurs at 20%-30% of max valve lift. This is far away from both the max lift point and where the popular 50% of max lift “mid-lift” geometry is at a 90° angle to the valve. Note: We’ve looked at an intake valvetrain here. If there is interest, I can offer some insight on the exhaust side as well, which will have higher loading yet.
In our next segment, we’ll dive into the implications this has for where an engine builder might want to set up his rocker arms.
