Suppose you are designing a new engine, so one of the things you must specify is the rocker arm ratio. Do you go numerically high (say, above 1.5), or numerically low (below 1.5)? Here are my thoughts:

Numerically low ratio

• lifter and pushrod will move more distance (I'm thinking of inertia and momentum)
• less spring force on the pushrod and lifter
• reduces friction
• cam must have more lobe lift to achieve the same valve lift

Numerically high ratio

• lifter and pushrod will move less distance
• more spring force on the pushrod and lifter
• increased friction
• cam doesn't need as much lobe lift to achieve the same valve lift

I have to believe reducing friction is always good, even with roller lifters. However, it seems we have a catch 22 with lobe lift and spring pressure: more pushrod and lifter movement with less spring pressure to control them if you choose a low ratio rocker, and less pushrod and lifter movement with more spring pressure if you choose a high ratio rocker. Perhaps a low ratio is better suited to lightweight pushrods and lifters, and vice versa?

 

Thoughts:

Job One is valve motion, including lift, duration and timing of events. This determines how the engine will use the intake and exhaust systems to produce the torque and power the engine design wants. Everything else is subserviant to this.

Rocker Arm Ratio (RAR) is just one of the tools used to get the valve motion. Consider it a secondary condition, not a primary one.

No matter how you achieve the desired valve motion, 1:1 RAR and lots of lobe lift, or 2.2:1 RAR with little lobe lift, the valve springs still have to control the valve, as well as the rest of the bits. The forces resulting from the valve accelerations (F=Ma) are more important than the rubbing and rolling friction in the lifter and cam/lifter/lifter bore interface.

Limitations in available lift at the cam lobe also help determine RAR. With flat tappets (lifters), the lift/degree of rotation of the lobe (Cam Velocity) is mechanically determined by lifter diameter and lobe diameter. Nextel Cup engines are an excellent example. They sometimes need more than twice the available lobe lift at the valve. High RAR as well as valve lofting are used.

Everything is a compromise: as you increase the RAR, you decrease the cam velocity and the resulting side loads on a roller lifter, but you increase the normal (axial) load, so you need a stiffer pushrod to minimize its deflection and compression. That makes for a heavier pushrod, but pushrod mass is a small portion of the total valvetrain mass, and as you said, with a high RAR it's moving less so it creates less inertia force, but there is more spring pressure acting on it. There's no free lunch. The good thing is that stiff pushrods are readily available and relatively inexpensive. It's a bad place to cheap out in a performance engine design.

Modern pushrod engines (not an oxymoron, IMO) tend toward higher, not lower RAR. The original SBC had under .400 valve lift and used 1.5 RAR. We're at twice or 2.5 times that (in extreme applications) now, and 1.25 times that in production engines with 1.7-1.8 or so RAR. We may see more RAR in the future on OEM and "street" engines, not less. I suggest it's the lesser of the evils resulting from lots of valve lift.

My $.02

 

OldSStroker, I've noticed too that OEMs are going to larger RARs, aren't they moving to larger base circle cams too? I think i read somewhere that there's a close correlation between lobe cross-sectional area (or just shape, can't remember) and an engine's torque curve, so, from an engineering perspective, how would that "blend in" with using higher RARs?

 

With higher lift lobes and more aggressive profiles larger base circles reduce lifter loads. Remember the Gen I SBC was an early 50s design, and the LSX a early-mid 90s design.

I believe you meant area under the lift curve helps pump air (torque). More area there means valve is open longer and/or higher and/or quicker to allow more air in. Higher RARs get more lift at the valve from a given lobe.

None of this matters much unless the ports flow well at the higher lifts, which modern engines do. Funny how all these things work together, isn't it?

 

Thanks. It is fascinating to see how all this stuff works.

 

Suppose we know the exact lift curve we want for the valve. Can't the cam be designed to produce that lift curve no matter what RAR we select (within reasonable limits)?

Interesting... are you assuming roller lifters, or flat tappets too? Certainly there must be some compromise between rpm capability and durability, even with roller lifters.

Wouldn't you need a strong pushrod in either case? With a high RAR, the pushrod must resist being crushed, and with a low RAR, the pushrod must resist bending, right? It seems that perhaps it would be easier to design a pushrod to resist crushing rather than flexing.

 

Yes, if we don't exceed cam velocity limits on a flat tappet. The optimum design should take into account roller diameter, base circle diameter as well as rocker arm geometry (not just nominal ratio). We are most concerned with what the valve does and we design the lobe to achieve that motion.


Both, but for different reasons. Flat tappets have almost point contact with the cam so there is very little side loading, but lots of pressure (lbs/sq. in.) at the lifter/lobe contact point. Using Diamond Like Carbon (DLC) coating on the lifter foot and/or the cam lobe is a big help in minimizing friction at the contact point.

There's always a compromise for durability. Endurance engines are a lot different from drag engines in this respect. Cup engines may run 1.5 million revs on a engine, but a (conservative) engine running at 5500 average for 24 hrs (like C6R) turns about 8 million revs in anger. A Pro Stock engine turns fewer than 1000 revs per run.

Pushrods theoretically see low bending loads because they have ball joints at each end. However there is some friction between the ***** and the cups in the lifter and rocker, so some torque is put into the pushrod from either end. Cup engines probably use DLC coated pushrods.

The pushrod is a thin column so axial load tends first to compress(shorten) it, but since they are not perfectly straight or round and wall thickness can vary minutely, and bending loads are applied at each end, they begin to bow.

Once they start to bow it becomes easier and easier to increase that bow. That's why larger diameter and especially tapered pushrods are used. Of course bow causes the pushrod to act like a vaulting pole which may not be what you want, or maybe it is if you need loft to get the lift you need at high rpm. That's another story.

Remember that with a low RAR the valve induced loads (spring and inertia) are multiplied to the pushrod, but the lifter induced inertia loads are LOWER because the lifter is accelerating less. With high RAR the opposite occurs. Somewhere, somehow, the pushrod has to take the loads. There's no free lunch.

 

Here's where the thread starts to branch out. I'm guessing DLC is not suited for all applications, since you need some friction to spin the lifter. Would DLC be applied to flat lifters after break-in?

Even ignoring friction, I visualize it sort of like waving a stick. High RAR is like waving it with small amplitude, and low RAR is like waving it with large amplitude... except both ends are constrained, so the waving happens in the middle.

I know, but sometimes one lunch is a better deal than the other(s).

 

No. Remember that the lobe is ground on a slight angle so that there is a small rotating moment to spin the lifter. DLC has to be applied to a VERY smooth surface or you just make the surface into a diamond file.

I don't think that's necessarily a good analogy. Assume .600 lift with 1.6 an 2.0 rockers: the difference in lifter movement is .075 inches which isn't a lot. The pushrod doesn't "wave" as much as it bends or bows because of the load. Tapered pushrods, especially double tapered ones, have the stiffest part of the pushrod in the middle where it tries to bow the most.

A simulation on EA Pro showed that pushrod forces more than doubled from 1000 rpm to 7000 rpm (330# > 770#) due to inertia forces. There was very little difference in the forces with 2.0 RAR vs. 1.6 RAR. That makes sense because the less movement (and therefore acceleration and force) with the 2.0 means less inetia forces on the pushrod which offset the higher spring-induced load.

The largest forces aren't when the lifter goes over the nose at high rpm like they are at idle. In the case I simulated the load over the nose was almost zero because the lifter was just about lofting, which is often the case. The largest load (@7000) was as the lifter was just onto the flank (after the ramp) and accelerating itself and the valve a lot.

My $.02

 

No to which question? If the surface is too smooth, how will the lifter spin? And how can DLC be applied to just one surface? My understanding is that the lifter face and the cam lobe must be in the same neighborhood of hardness, otherwise the harder surface may wear out the softer surface.

I was thinking more of the movement of the pushrod tip in the rocker arm. Heads with pushrod slots must often have the slots elongated even when just switching from 1.5:1 to 1.6:1 rockers. I thought that indicated a significant change in side-to-side movement of the pushrod, which would be even greater when comparing 1.5:1 to 1.7:1 (or even 1.8:1). *scratches head*

 

No.(I only saw one question mark in the quote.) Break-in would mess up the smooth surface that was being coated.

1) Small angle and a moment arm and 300-700+ lbs of force with a mu > zero should do it. You can't have a frictionless surface (mu = zero).
2) On a flat lifter the materials of lifter and cam need to be compatible, but not necessarily the same hardness. Lifter foot hardness is generally harder than the cam, especially on cast cams. The lifter is loaded in a very small area, but the cam works all around the lobe.

I believe some Cup flat lifters and cam lobes are DLC coated. With DLC coated piston pins the steel rod isn't coated.

Slots are lengthened to accomodate the pushrod end of the lifter getting closer to the pivot stud with higher ratio because the distance from the pivot to the valve remains the same, but the location of the pushrod cup in the rocker moves toward and away from the valve as the rocker rocks.

Pushrod deflection due to load is an entirely different can of worms. Inertia and spring loads bow the pushrod as discussed earlier. Yep, sometimes the pushrod will bow enough to rub the clearance hole in the head. Some folks may keep this clearance to a minimum, say .010 over the pushrods unloaded geometric travel to crutch a bending pushrod. Some folks don't.

 

OK, brain fart on my part there.

This is exactly what I have been thinking about. With a low RAR, the pushrod tip is farther from the rocker fulcrum. Couple that with the increased lobe lift, and the pushrod tip in the rocker travels through a big arc. With a high RAR, the arc should be quite a bit smaller. That's the basis of my stick waving analogy, which I'm bound and determined to validate.

 

I kinda figured that.


Good luck to ya'.

 

LOL, I see you remembered my BBC solid flat tappet thread.