My $0.02...
Back in the late-90's and 2000, I was a valvetrain release engineer at one of the OEM's. The engine we were developing at the time was having problems with valve springs breaking (and other problems related to valve spring harmonics), so I got to know valve springs and the process used to make them very well. During that time, I spent time at all three of the major US valve spring manufacturers (Peterson, Associated, and Mubea). At that time, if you bought a valve spring from anybody (Comp, Crane, or an OEM), it was almost certainly made by one of these three companies, and likely made in the USA or Europe. Now, a lot of the manufacturing has been moved to Mexico and other countries, and I would guess that there are now Chinese companies putting their products in the market.
A couple of things I learned that are worth passing on:
1. If the valve spring is made on an automated, modern process, it will not "sag" or lose load over time. If the spring is made on a good process, it will be "heat set", meaning it will be compressed to a certain height and then baked in an oven (simplified description). A spring that has gone through this process will lose very little if any load over its life in an engine. Note that heat-setting was a relatively new process to use in OEM spring manufacturing even in the 90's - the springs on our original 1987 engines probably didn't have that process, which is why they tend to sag over time.
2. In most automotive industries, stuff sold to the "aftermarket" is not made on the same processes as is used in OEM. I currently work for a bearing company, and I can tell you that the specs for a bearing used for an aftermarket alternator (for example) are different from a bearing to be used in an OEM alternator. Valve springs are no different. So, I can't guarantee that those Comp or Crane or whatever valve springs are produced to the same, modern processes and specs as an OEM valve spring would be made.
3. Proper valve spring design does not depend only on load. At a certain RPM, a certain amount of valve spring load is needed to ensure that the valve doesn't "float" (lifter loses contact with the cam) as the valve changes direction from opening to closing. That amount of force is a function of how heavy the valvetrain is (valve, locks, retainer, rocker arm, pushrod, lifter), how aggressive the cam is, and how fast the cam is rotating. However, as previously suggested, look up some videos of valve spring behavior at high RPM (they are out there on YouTube). All valvetrain systems have harmonics, meaning they resonate at natural frequencies. Think about ringing a bell - when you strike the bell, it vibrates at a specific frequency, producing a tone. Valvetrains and valve springs do the same thing. If the valve spring is "ringing" (resonating) at a frequency that corresponds to the cam rotational frequency, bad things can happen. Increasing spring pressure can help mask these problems. Other ways to help are to increase valve train stiffness (larger diameter pushrods, stiffer rocker arms, etc.) or decrease valvetrain weight (hollow valve stems, lighter rocker arms, bee-hive springs that allow for a smaller retainer, etc.). Either the above will tend to move the natural frequencies of the valve train higher, hopefully above the operating speeds of the engine. Of course, the natural frequency of the valve spring itself isn't affected by any of the above - that's simply a function of the geometry (size and shape of the wire) used to make the valve spring. The natural frequency of the valve spring itself can be mitigated by using thinner wire and/or a multi-rate spring (coil spacing is different at the bottom than the top), but this has the side effect of increasing the stress in the wire. For modern, lightweight, multi-rate bee-hive springs, special materials like chrome-silicon-vanadium wire are often used If you don't have complicated software to model valvetrain behavior (who does, really?), then sometimes good ole' trial-and-error is needed to pick the right valve spring for a given cam and valvetrain system.
4. For most engines, the valve spring is the most highly stressed component. When a valve spring is compressed, the stress in the wire is very high - higher than in the connecting rods, crankshaft, or anything else in the engine. That's why super-clean exotic steels like chrome-silicon and chrome-silicon-vanadium are used. A valve spring sees what we call "cyclical fatigue" - every time the cam comes around, the stress in the wire goes up and then down. (Think of bending a coat hanger and then releasing it over and over again). Because of this, all valve springs will eventually break if they run long enough (rarely is the stress level low enough for the steel wire to have "infinite life", especially at high RPM's). Properly designed, the valve spring should easily last "the life of the engine" in an OEM application. However, for the big cams that we sometimes run, I wouldn't expect 100,000-mile durability.
5. Valve spring dampers can be your friend - On the stock valve springs, there is an internal ribbon-looking spring inside the main spring (IIRC, haven't seen mine in ten years). That is not a "helper spring" - it's a damper. It's purpose is to help remove some of the coil "ringing" that occurs when a valve spring is running near one of its natural frequencies. If the coils start to "ring" and vibrate back and forth after the valve is closed, the friction between the coils and the damper will dampen the vibrations out. The goal is to remove all of the coil "ringing" before the cam comes around for the next lift event. Dampers can do this quite effectively. On the downside, dampers increase valve train friction and can themselves break if not designed properly. There are different types of valve spring dampers. Some are the "ribbons" I described above. Some are "cups" that go inside the valve spring and expand outward against the coils. Some are "cups" that go outside the valve spring and expand inward onto the coils. BTW, the engine I mentioned at the beginning of this post eventually went into mass-production with internal "cup" style dampers made of plastic. Interesting.
Here's a typical video of valve spring behavior at high-RPM. Notice how after the valve closes, the coils in the valve spring are still "ringing".
If the ringing is bad enough, then the valve can bounce back off the seat after it's closed or float over the cam nose when it's open. Both are bad and can cause erratic engine operation, particularly at high RPM. The two basic ways to "fix" this are to increase valve spring load (higher valve spring load will keep the valve planted even if the coils are ringing) or to "tune" the valvetrain better (with different valve spring design, lighter weight, higher component stiffness, valve spring dampeners, etc.). In the perfect world, "tuning" is better, since higher valve spring loads come with their own problems (such as wiped cam lobes or worn-out roller bearings on lifters). However, since we don't usually have access to high-speed cameras and valvetrain modeling software, we have to make do.
OK, the onslaught of relatives looking for Christmas dinner is coming. Hopefully somebody learned something, and Merry Christmas!
