Let's try this again... Valvespring tech, 101!

Over the last few months, there have been some discussions about valvesprings, much without data or hard information. So anyway, I thought I'd share and article and an offer to get the ball rolling here.

First some background. I'm a physicist, so I'll start from springs and masses. The point of a valve spring is to keep the valve following the cam profile throughout the RPM range of interest. Pretty simple, yes? So then why are springs rated at things like 80 lbs seat pressure closed, and 200+ lbs seat pressure open to just keep a little valve that weights at most a couple ounces where it's supposed to be? Don't those numbers sound excessive?

So, let's think about the valve train. There is a camshaft with it's lobes. There is some form of rocker that rides the lobes, pivots somewhere, and presses on the end of the valve. This is for overhead cams (standard V8s have lifters and pushrods, but the idea is the same). What pushes back on all this stuff to keep it from rattling is the valvespring. It sits on the top of the head, and is kept in place by the retainer and lock. So this spring has to be able to push all this hardware back against the cam lobe.

The real killer hear is that as things go fast, it's not the basic static weight of the parts that comes into play. The spring has to overcome the inertia of the valve, the retainer, the lock, the rocker, as well as the spring (or springs) themselves. At 7000+ RPM, this takes a lot more force than is needed to just hold the parts against the force of gravity. Thousands of times the force. It can get pretty staggaring.

So, what does all this matter? For our Tritec engines, the choices for valvesprings have, until recently, been pretty limited. Every head I've seen has used some form of convetional spring. No one was offering beehive geometry. Some used Ti retainers, some used the stock steel ones. Some used single springs, some double. Some of the springs needed pocket work cause they were too long, and some needed spacers cause they were too short. None of this is bad per se, but it's all the consiquence of using parts pretty much designed for other applications and used because they were "close enough".

So, what's different? Well, nothing really. The PAC Racing Springs in the NS Valvespring and Retainer kit aren't custom for the MINI. They are, in fact, "close enough" springs found in the PAC spring catalog. What is different with the set is that the retainers are custom to the MINI so that a spring that is a touch too long can be used without machining a deeper pocket. What is new is that this is the first spring set for the Tritec that uses ovate wire (oval spring wire) and beehive geometry. The result of this is a lower mass spring and retainer set that has reduced harmonics compared to conventional springs.

Now, in some of the PMs that I've gotten from people, some who were not really happy when about the debate about springs, there was one quote that stuck in my mind: "You know that no one can feel the difference between one valve spring and another." And to be honest, I think that if the springs are properly selected, that may even be true. But we have a group of people in MINIs who are pushing ever higher RPM, and using aftermarket cams with more lift and aggressive profiles (this is called "ramp" to the cam guys). The faster the ramp, the faster the acceleration of the parts. Agressive ramp is the key to the NS cams benefits with lower overall lift (but that's another story). Anyway, we're demanding more and more out of the valvespring and retainer. And some are doing this with stock, or stock equivalent springs (for example the Cosworth single spring setup was a stock-equivalent spring. Same strength as the stock, just more consitent in specifications.)

Think this is all just marketing BS? Well, this is why I love Hot Rod Magazine. Sure it caters to the V8 Big Block crowd and American chassis and the like, but engines are engines, and they do some really great tech articles. The November issue was no exception and they have this article All About Valvesprings. I scanned it for your reading pleasure. It's seven pages (a lot for a mag article) and the high points of it are:

• In the past, valve train performance was limited by spring technology.
• Times have changed, and new geometries and materials have come into play.
• One V8 engine had significant high RPM HP increases when going to a SOFTER beehive from a dual spring set up!

So read away, and learn from the experts......

I've also got an offer to those that are interested in our spring and retainer set. The set retails for $385 on our site here. I want to get some of these kits into the hands (and onto the engines) of those that think they can benefit from them. So for the first 5 kits that are ordered in this offer, I'll let them go for $335 a set. To get this price, you have to do PayPal to [email protected]. If you want to do a telephone credit card transaction, e-mail me at [email protected].

If you have a stock motor, no tune, no cam, no nothing, these aren't for you. If you have a raised red-line and/or aftermarket cam, these may be just the ticket for really getting all the HP your engine has to offer as you bang the rev limiter!

Matt

ps, I'm pretty sure that in the coming weeks and months, you will see others offer Beehives for the Tritec. Why not? SuperTech, Ferrea, Comp Cams and others are using them for more and more applications! We use PAC Racing springs and are proud of it! Read all about PAC Racing at thier web site.

pps, there was a thread here that appears to be gone. While it did have more than it's fair share of BS in it, it also had a lot of good information. They are available for V8s, Hondas, Evos... Subies even Harleys.... And now, (actually for about 6 months), they have been available for the MINI. Instead of complaining that the MINI is the baastard child of the automotive aftermarket due to it's small size, we have access to some of the latest tech out there in valvetrain technology. Others will follow. FES and NS were first.

Constructive comments and questions are more than welcome. BS and the rest of the craap that got the last thread deleted is not.

 

Couple of things on springs....

for those that don't like the basic principles or think they are a waste of time, skip this post.....

So, more on how springs work..... the basic ideal spring is massless, and has a constant spring rate, conventionally called "k". The force that a spring exerts is -kx, where x is the displacement. The minus sign indicates that the spring pushes back against the direction of displacement. This is why the spring is called a linear spring, as the force curve is actually a straight line that goes through zero force at zero displacement.

When this spring is attached to a mass (typically called "m" in all the textbooks, imaginitive it's not), the once displaced the system will oscillated with the mass going back and forth. In technical terms the stored potential energy of the compressed spring is put into the kinetice energy of the mass and out again and in again and on and on and on. There is a natural frequency to this, that is equal to the square root of (k/m). So the stiffer the spring, the higher the natural frequency, the heavier the mass, the lower the natural frequency. A little nerdy, yes, but we need to understand this to deal with spring surge later....

So, in no particular order, let's deal with how actual linear springs aren't ideal... Let's look at a standard linear spring. Think in your suspension or on your valves, doesn't really matter which. Look at the end of the spring. On struts, there is usually some change in the winding so that the end of the spring is flat. On valve springs, they are typically ground flat (this results in a shorter "bound" or "locked" state, where all the coils compress and the spring becomes solid. Well, why does this matter? It matters because the ends where it's not like the rest of the spring introduce what's known in the physics world as non-linearities. So that nice straight line that goes through zero for the ideal spring isn't really that straight at the origen any more. This in itself isn't that big a deal. On your car, the rest position is somewhat compressed, and same with the valve spring. But one effect of the non-linearity stays behind, kind of like a nagging headache. This is what physicists call "mode coupling". You can think of mode coupling as the same thing as overtones on a guitar or piano string, or over tones on a pipe organ tube, or anything like that. while a little nerdy again, this is some of the stuff that induces spring surge as well. Save these concepts for later.

Now, real springs are not massless. They have the weight of the wire distributed along the length of the spring. So while in freshmann physics we say "consider a perfect spring with constant K fixed at one end pushing against a mass m...." (sorry for all those horrific flashbacks to freshman physics, if you took it), in the real world there is no massless spring and the mass of the spring has to be considered. So what does this do? This actually REDUCES the force exerted by the spring as it goes faster and faster! Wha???? Think of it this way, at every slow rates, very little of the stored energy in the spring goes into kinetic energy of the mass of the spring moving! At very high speeds, a real, non-zero amount of the stored energy in the spring goes into the kinetic energy of the spring wire moving. This is why in the article referenced in the first post, the author quoted an engine builder who said that he can often take the third spring out with no loss in performance. The added push of the third spring in this case is mostly going into just moving the mass of the spring..... So no real gain in valve control. (For those that want to cry BS, I'm sure that there are cases where the third spring is usefull, this isn't a blanket statement that all three spring apps don't work etc etc etc.... So put down the flamethrowers....). But simple little springs are like Ginsu knives... But wait, there's more.....

Now, the valve, retainer, lock and rocker arm are all much stiffer than the spring (as is the cam lobe) so for now, let's just imagine that they have some fixed effective mass that the spring is working on. Let's say this mass is somewhere between 4-8 oz. So why are we talking about spring forces of (For the stock spring) a bit over 200 lbs seat pressure when the valve is open? Well, when your just holding the effective mass in your hand, this is how much force it takes to counter gravity. But as it oscillates more and more, the force required to make the mass move in it's desired path scales linearly with frequency. So to move this stuff takes an ever increasing amount of force. So now we have a situation where a close to ideal spring with mass is working less and less hard against a constant mass that requires a higher and higher amount of force to push! Couple of "take aways" here:

• You have to choose a spring that still can push hard enough at the peak RPM to move the mass.
• Using a lighter wire wins all over for any of these cases cause less of the stored energy goes into the moving mass of the spring itself.

So, a lighter spring of same spring constant will work to a higher RPM than a heavier spring that has an equal spring constant. Also, this is why PAC and others that use them are so hot on the "Ovate" (or oval or multi-radius) spring wire. By changing the shape of the wire, they can use less mass to do more work. This graphic is a simulation of stress distribution in the ovate geometry.

It's a subtle but important point. A somewhat related point is that metallurgy comes into play here as well. Better materials control allows all the creation of alloys that will hold temper better for longer than in the past. So you can use a tailored alloy (PAC Racing calls thier PACaloy) and treatment to lower the mass of the spring as well.

While I'm sure many are asleep now, we still haven't asked Erics question yet.... We'll get there Eric, don't worry....

Now, there is a bitchen high speed video taken on something called a "SpinTron" on high speed motions of valve springs.
[ame=http://www.youtube.com/watch?v=yfmb-tCo2yA&feature=player_embedded]YouTube - High Perfomance valve spring testing 10K RPM[/ame]
This gets directly to Erics question above. First thing you notice is that the spring doesn't move like one would expect. All the coils don't compress equally, and some sort of wired oscillation is happening. Remember the non-linearities and mode coupling? Those are what drive these non-ideal movements. This is spring surge. This can get so bad, that the end of the spring will bounce off the head or worse, the retainer! Not only that, but all this motion of spring mass will slam up and down inbetween the retainer and the head, doing all sorts of mischief. Now in the video above, things are moving pretty well, but you can see the non-ideal motion. This is also the source of the non-linear force curve as a function of RPM for the force between the spring and the valve (or between the valve stem and the rocker arm etc). This is the root cause of valve float or bounce. This is bad.....

But we're not done yet. (remember, there's more!). The smaller top of the beehive means the part of the spring moving fastest has the least mass, and that the retainer is lighter too. This is good for all parts of your valve train. Every engine builder or person in the industry who races, builds, designs or whatever will tell you less mass at the top of the spring is good (as long as the whole system can do what is needed.)

Remember our friend ressonance? The Square Root of (k/m)? Well, beehives are kind of like a progressive spring. The force curve for them isn't exactly linear so that means that k isn't a constant, but it's a function of displacement. So k=k(x). What this means is that the ressonant frequency of the spring changes as the spring is compressed. How much? I don't honestly know, but it's real and it's what is the science behind the reduced surge in a beehive spring. What happens when the ressonant frequency isn't constant is that even if the energy of the ressonance is the same, it's spread out over a wider frequence (read RPM range) and has a lower amplitude (the surge force wave has a smaller peak force). This too is good. This means that the surge waves going through the valvetrain are smaller in magnitude. This smaller surge wave coupled with the lighter mass is why a beehive spring of the same force as other springs will rev to a higher RPM. Non of the is marketing BS. Non of this is spin. This is real, and any engine builder with any real understanding of valvetrains won't argue any of these points. Consider it all explanations of accepted fact.

Now, criticsm has been brought up about testing. Fact is, this tech has been tested. Here is PACs testing environment. One of the instruments I want to point out is called the SpinTron:

Anyway, a spintron is an instrument (not a cheap one either) where you bolt up a block with no pistons or con rods or crank, and it spins up the valve train to ungodly speeds. To see what's going on you put laser distance system or high speed cameras where they can see the valves (sometimes with big holes machined in the block, head or valve cover) to gain proper access. This is the machine with a high speed camera that was used to create the video of the spring above. (read all about SpinTrons here). PAC uses SpinTron testing in creating their products. Also x-ray stress analysis etc etc etc. Read about it on the PAC Engineering and Testing links... Anyway, the springs have been tested. Up the wazoo. Ti retainers have been tested... Up the wazoo here as well. This isn't new tech. It's new to MINI tech. We, as is usual for small market cars, are just treading down the very same path that others have gone down already.....

As far as on car testing, they are on cars and work! So take that testing criticism, put it in your pipe and smoke it! That's about all it's good for, but with it's total lack of content, it probably won't pack much of a buzz in a pipe either! But to be fair, there are no dyno plots of the springs on MINIs. That is true. So do they offer benefit to the MINI? Are they even needed? First off, let's look at when you would see and when you wouldn't see anything on a dyno plot.

In the article linked to above, there is an example of where you would see it. At high RPM, the valve is starting to float with the dual spring/damper combo, and that goes away with the beehive. So if the engine is already suffering from valve float, it will show up on a dyno plot each and every time, as the higher RPM range accessable via the lighter valvespring mass, retainer mass and lower surge forces all come into play and the valve is well controlled. Great. Obvious. Would be a slam dunk if we had mountians of dyno runs showing that this is indeed a big problem. I don't have them. But something is missing from the discussion that all good engine builders know full well....

So what about when you CAN'T see it on a dyno plot? Is there any benefit there? And the answer is you bet there is! But it's not a short term win, it's a long term win. And this is less wear and tear on the valvetrain. Lower surge forces mean that there is less abuse to the valves, the rockers, the seat of the head, the retainer, the cam itself. All this speaks to durability. All the parts will last longer with a lower chance of failure because of the known, proven and well accepted technology (pretty much everywhere but in MINIdom) and benefits that they deliver.

But what is really interesting, is that this tech doesn't cost any more! So you don't have to pay a premium to actuall benefit from it. It's out there. It's available, and it's competitively priced compared all the alternatives that you can buy for the application. So I don't see any good reason to NOT choose the better tech for the job!

That said, I'd love it if everyone would run out to thier cars, place an order with me, and swap out thier valvesprings for the kit Lynn developed, and I and others sell. But I don't expect it. If you have a working valvespring/retainer combo in your car, and are happy, I'm happy! If you're building an engine or a head for high RPM work, you should definantly think hard about taking advantage of the benefits of Beehives with Ti retainers.

Like other head retailers have said, they've been waiting for this tech for 2 years. It's already here, has been for several months. NS did it first. Others will follow. Others should follow. It's just good tech for the car.

Matt

 

OK guys...

one thread went down the crapper already! Please don't make me cut and paste this stuff into yet another!

Also, I find it interesting when the explanations are accurate and full of verifiable fact, then the normal pisssing contests are absent.....

Matt

 

Some examples of float....

so what's all the "valve float" that everyone keeps talking about? All that stuff that I described comes together, and there just isn't enough force to keep the valve held against the seat in any meaningful way. Here's a look at it happening in real time......
[ame=http://www.youtube.com/watch?v=_REQ1PUM0rY]YouTube - Valve Float[/ame]
Anyway, the motions you can see are pretty dramatic: The valvestem is wobbling, and this is even before it starts to float. The wobble is a "mode coupling" from the very stiff spring that when it's compressed it wants to bend a bit. This is present to some degree in all wound springs, but is less with more turns of thinner wires.

When it's floating the whole assembly starts to rotate! (If you wonder why this is, take a slinky and pull apart a a bit of it. You'll feel it want to twist along it's axis. This "mode coupling" of axial and rotational motion is present in all wound springs.) Normally when the valve is closed, this rotational motion is quenched.

At the end of the vid, you see something related to, but different than, valve float. This is called bounce, for obvious reason! Anyway, the spring assy isn't doing it's job, and the valve isn't well controled as it comes down to the valve seat. It slams into it and bounces off! Yikes! This is also a dramatic example of the spring actually having less clamping force when in motion due to the work required to move it's own mass, mentioned in one of the lengthy posts above.

A comment on how these vids are made. The Spintron outputs a sync signal to a high speed camera. So it's not a "super slow motion" movie, but really more of a sequence of strobe pictures timed to take a pic at the aquisition speed of the camera (I think it can go about 4000 frames per second) to get the valve at the same instant (or close to it) in time. You can see in the vid where the RPM is changing quickly and the system looses it's sync. You can also see in the last part of the vid where the bounce is fillmed that they milled a big freakin' hole in the side of the engine & head to get access to this for the movie. This engine is NEVER going to see the inside of an engine bay again!

So to beat the (almost) dead horse of testing again. We all test, just to varying degrees. The NS valve springs have been dimensionally tested (all the parts fit), they have been tested in cars (yes the car runs). They have not been dyno tested by me or anyone I know of. And while the materials and design of the springs have been lab tested by PAC. They have not been tested for longevity in a fleet of test cars or on accelerated lifetime test rigs for the MINI. But then, the guy who was complaining about no testing doesn't do that either! Maybe you'll get some dyno tests before and after, but as described before that may or may not show any benefit, depending on use case. And you may get some short testing on some local cars that are convinient to test on. That's all good. But don't kid yourself. All of the parts makers for MINI do the durability testing on maybe a house race car for a bit of time, but none do it to industry or car manufacturer standards. For all of us parts builders and suppliers, the customers are the durability testers. Anyone who tells you different is just blowing smoke.

Matt

 

When I wanted to get some quantitative MINI data...

I started digging into the SpinTron. I was hoping that I'd be able to run just a head on with different spring and valve setups along it's length to do testing in parallel. I learned that there is no effin way I could foot the bill for a whole engine and all the parts (I'm sure I'd have to pay for a mounting plate and maybe a crank mandrel as well), and that the Spintron is a really, really bitchen piece of equipement! I can see how it's credited with really revolutionizing valvetrain development and testing....

Matt

 

I'm always amazed...

at what was done before computers. I remember my Dad doing work with sliderules. I have a couple of them as pieces of history, but can't do much more than basic math with them. I remember when he got his first HP Pocket Calculator! Over $900, back when the dollar was worth a heck of a lot more than it is now!

Matt

 

Some numbers...

Just checked a couple of things...

Stock Springs weigh 51.5 grams each
Stock Retainers weigh 12.1 grams each
Stock Intake Valve weigh 44.5 grams each
Stock Exhaust Valve weigh 38.3 grams each
Don't have locks to measure.

So the stock spring weighs MORE than the stock valve, intake or exhaust! A lot of the spring force is moving it's own mass!

Cosworth
Single spring weighs 49.5 g each (but is stock spring rate)

NS PAC Beehives with custom retainers
Spring weighs 48.8 g
Retainer weighs 5.9 g

I don't have any of the aftermarket valves to weigh, nor do I have Supertech or Ferrea springs to weigh either. If any of you do, send me the numbers and I'll add them to the table. The beehive is a slight beehive. I don't know how much of the weight reduction is better spring material, and how much is the smaller diameter where it counts. (and look, I've got no problem admitting when I've hit the end of my knowledge either....)

Matt