mgmshar
03-28-2004, 11:18 AM
I'd like to offer some comments on all of this. Before I begin, I've been working in Detroit for the automotive industry for about eight years now. The last six years, I've been working on valve train. So I've been working around the design and development of valve train systems (valve springs, rocker arms, etc.) for a little while. I'm not a "know-it-all" expert or anything, but maybe I can dispel a few myths. I will admit that most of the valve trains that I've been around have been OHC "Type-2" (a.k.a. roller finger follower) valve trains, but a lot of the principles still apply.
Why do we care about dampers and bee-hives? Well, the main reason is valve train stability. All valve springs, regardless of design, have natural frequencies at which they resonate. Ever see a slinky going back and forth? Valve springs do the same thing, although they do it at much higher frequencies. The problem is that those frequencies often correspond to certain rpm's in the engine. If a valve spring "gets excited" at a certain engine speed and starts to resonate, then bad things can happen. Valves will tend to bounce off the seat after they close. In extreme cases, valve train separation can occur. I have seen high-speed video of this happening, and it's not pretty. In addition, valve spring resonance causes the stresses in the valve spring to increase, which of course reduces its life. A certain engine that is in production at this time (won't say which one) can be made to misfire at WOT at exactly 5,200rpm, because the valve springs on the exhaust valves hit a resonant frequency. The exhaust valves don't stay closed, and the engine will start to miss.
Ever wonder what the most highly stressed component in an engine is? Believe it or not, it's the valve spring. That's why modern engines use special "super-clean" steel with high alloy content (chrome, silicon, vanadium, etc.) in their valve springs. Even the smallest microscopic "inclusion" (non-metallic particle) can cause a spring to fail very quickly - and I have the broken parts to prove it.
So, how do valve train designers make valve springs better? Well, there are a few basic options:
1. Increase valve spring loads. In other words, make the valve spring pressure at valve closed and valve open higher. This is the oldest trick in the book, and still is used in by racers everywhere. Why does this work? Springs with higher pressures generally (but not always) have higher natural frequencies, which raises the rpm at which they start to resonate. In addition, the higher "over the nose" loads delay valve train separation until a higher engine speed. The downside is that everything in the valve train, especially the cam in flat tappet applications, will see more wear. Many people on this board have experienced this.
2. Add dampers to the spring. Dampers come in two styles: internal ribbon dampers (like the Kirban's springs), or external "cup" dampers (like the Comp Cams springs). The principle is simple. When the spring gets excited, and the coils start to resonate (like a slinky), the friction between the coils and the damper work to quickly stop the resonating. In other words, the friction "damps out" the coil movement. Dampers are very effective, and are in fact still used in some engine applications. The downside with dampers is that they add mass to the valve spring (if they are internal ribbon type). Any mass in the reciprocating parts of a valve train generally lower its peak speed capability. The other downside of dampers, at least for production engines, is that there is a perception that the dampers lose their effectiveness over time. As the damper "relaxes" its grip on the spring coils, and as the surfaces of the dampers get polished, there is less friction between the damper and the spring. This reduces its ability to dampen out spring resonance. This is not a big deal, unless you're trying to make your engine last 150,000 miles (like GM, Ford, and Chrysler do).
3. Alternate spring designs, such as "bee-hive" and multi-frequency springs. I can say that "bee-hive" springs are definitely in vogue in modern engines. A few examples include the Ford 4.6L/5.4L OHC (2-valve and 4-valve), the Chrysler 3.7L/4.7L OHC, the Chrysler 5.7L "Hemi", and the GM Gen-III (LS-1, LS-6, etc.) engines. Bee-hives are used because they allow a smaller valve retainer, which reduces valve train mass and increases packaging space for the rocker arm. In addition, most beehives are designed to be "multi-frequency" springs. If you look carefully at the windings of the spring, you will see that some coils are closer together (usually at the bottom of the spring) than others. This coil spacing is very carefully chosen, so that certain coils go completely solid at certain valve lifts (usually near valve closed and valve open positions). When a coil in a valve spring goes solid, meaning that it touches the coil above/below it, the natural frequency of the entire valve spring changes. By changing the natural frequency of the valve spring during the valve lift, resonances in the spring can be effectively reduced. The only downside of bee-hive springs is that the coils that remain active (do not go solid) through the entire valve life event are usually VERY highly stressed. The steel used to make these springs has to be the very best available. However, the benefits, and the lower cost compared to dampers, are why most modern engines use this type of spring.
One note about bee-hives - they MUST be designed for the application. Coils must be spaced to go solid at the right valve lifts, otherwise the spring is no better than a simple, un-damped valve spring. The way to check this is to measure the spring on an Instron machine (or something similar), and make sure that the spring rate changes (coils go dead) at spring heights just short of valve closed position, and preferably just long of valve open position. You can't just slap a bee-hive from some engine into another engine and expect it to work well - it won't.
One other note - in extreme applications (top fuel dragsters, NASCAR engines, etc.), people use "double wound" springs. These are valve springs that are actually two valve springs packaged inside each other. Because the two springs are designed with different spring rates (and natural frequencies), one of the two springs is always not resonating. This approach is used when very high spring loads are needed (high valve lifts and rpm).
Well, that's my download on valve springs, taken from a valve train engineer's point-of-view. Hopefully this will be helpful to someone.
Why do we care about dampers and bee-hives? Well, the main reason is valve train stability. All valve springs, regardless of design, have natural frequencies at which they resonate. Ever see a slinky going back and forth? Valve springs do the same thing, although they do it at much higher frequencies. The problem is that those frequencies often correspond to certain rpm's in the engine. If a valve spring "gets excited" at a certain engine speed and starts to resonate, then bad things can happen. Valves will tend to bounce off the seat after they close. In extreme cases, valve train separation can occur. I have seen high-speed video of this happening, and it's not pretty. In addition, valve spring resonance causes the stresses in the valve spring to increase, which of course reduces its life. A certain engine that is in production at this time (won't say which one) can be made to misfire at WOT at exactly 5,200rpm, because the valve springs on the exhaust valves hit a resonant frequency. The exhaust valves don't stay closed, and the engine will start to miss.
Ever wonder what the most highly stressed component in an engine is? Believe it or not, it's the valve spring. That's why modern engines use special "super-clean" steel with high alloy content (chrome, silicon, vanadium, etc.) in their valve springs. Even the smallest microscopic "inclusion" (non-metallic particle) can cause a spring to fail very quickly - and I have the broken parts to prove it.
So, how do valve train designers make valve springs better? Well, there are a few basic options:
1. Increase valve spring loads. In other words, make the valve spring pressure at valve closed and valve open higher. This is the oldest trick in the book, and still is used in by racers everywhere. Why does this work? Springs with higher pressures generally (but not always) have higher natural frequencies, which raises the rpm at which they start to resonate. In addition, the higher "over the nose" loads delay valve train separation until a higher engine speed. The downside is that everything in the valve train, especially the cam in flat tappet applications, will see more wear. Many people on this board have experienced this.
2. Add dampers to the spring. Dampers come in two styles: internal ribbon dampers (like the Kirban's springs), or external "cup" dampers (like the Comp Cams springs). The principle is simple. When the spring gets excited, and the coils start to resonate (like a slinky), the friction between the coils and the damper work to quickly stop the resonating. In other words, the friction "damps out" the coil movement. Dampers are very effective, and are in fact still used in some engine applications. The downside with dampers is that they add mass to the valve spring (if they are internal ribbon type). Any mass in the reciprocating parts of a valve train generally lower its peak speed capability. The other downside of dampers, at least for production engines, is that there is a perception that the dampers lose their effectiveness over time. As the damper "relaxes" its grip on the spring coils, and as the surfaces of the dampers get polished, there is less friction between the damper and the spring. This reduces its ability to dampen out spring resonance. This is not a big deal, unless you're trying to make your engine last 150,000 miles (like GM, Ford, and Chrysler do).
3. Alternate spring designs, such as "bee-hive" and multi-frequency springs. I can say that "bee-hive" springs are definitely in vogue in modern engines. A few examples include the Ford 4.6L/5.4L OHC (2-valve and 4-valve), the Chrysler 3.7L/4.7L OHC, the Chrysler 5.7L "Hemi", and the GM Gen-III (LS-1, LS-6, etc.) engines. Bee-hives are used because they allow a smaller valve retainer, which reduces valve train mass and increases packaging space for the rocker arm. In addition, most beehives are designed to be "multi-frequency" springs. If you look carefully at the windings of the spring, you will see that some coils are closer together (usually at the bottom of the spring) than others. This coil spacing is very carefully chosen, so that certain coils go completely solid at certain valve lifts (usually near valve closed and valve open positions). When a coil in a valve spring goes solid, meaning that it touches the coil above/below it, the natural frequency of the entire valve spring changes. By changing the natural frequency of the valve spring during the valve lift, resonances in the spring can be effectively reduced. The only downside of bee-hive springs is that the coils that remain active (do not go solid) through the entire valve life event are usually VERY highly stressed. The steel used to make these springs has to be the very best available. However, the benefits, and the lower cost compared to dampers, are why most modern engines use this type of spring.
One note about bee-hives - they MUST be designed for the application. Coils must be spaced to go solid at the right valve lifts, otherwise the spring is no better than a simple, un-damped valve spring. The way to check this is to measure the spring on an Instron machine (or something similar), and make sure that the spring rate changes (coils go dead) at spring heights just short of valve closed position, and preferably just long of valve open position. You can't just slap a bee-hive from some engine into another engine and expect it to work well - it won't.
One other note - in extreme applications (top fuel dragsters, NASCAR engines, etc.), people use "double wound" springs. These are valve springs that are actually two valve springs packaged inside each other. Because the two springs are designed with different spring rates (and natural frequencies), one of the two springs is always not resonating. This approach is used when very high spring loads are needed (high valve lifts and rpm).
Well, that's my download on valve springs, taken from a valve train engineer's point-of-view. Hopefully this will be helpful to someone.