The Inertia Files (partial)

[mazdarati]

I thought i would post this up here for anyone who is interested in learning about the basics of cam timing, duration, breathing, etc. this is from part of a writeup called The Inertia Files written by someone who is no longer around the car community and the original source is no where to be found, this is the only copy i have been able to find.

without any further delay i present to you, the inertia files....

Introduction

I continue to see around here some serious nonsense and misinformation about the most fundamental of all aspects of internal combustion engine performance -- breathing. Herewith, I hope to address some of the stuff necessary to understand -- really understand the relationships between camshafts, breathing and performance. I apologize in advance for being too elementary, too advanced or too pedantic. I'm doing my best.

The Latin word for "I breathe" is "spiro". There are a lot of English words which derive from that Latin word:

* respiration (breathing again),
* perspiration (breathing through),
* inspiration (breathing in),
* expiration (breathing out)

The word which is of particular interest to us is aspiration (breathing to or towards).

A normally-aspirated engine is one which is dependent upon atmospheric pressure to breathe, to provide the air required for combustion of our fuel. For now we will talk about normally aspirated engines, but pay attention, turbo guys -- some of this stuff will apply to you in surprisingly novel ways.

In virtually every 4-stroke automotive piston engine ever built (I believe the exceptions were sleeve-valve Willy Knights) the breathing events are controlled by poppet valves. Rotary engines and 2-stroke engines use a different method of controlling the breathing events. We might address these types later.

Valves

A poppet valve is, in the simplest terms, a flat disc affixed to a round thin shaft -- looking very much like a nail. The flat disc (valve head) blocks an opening (port) to the combustion chamber when the valve is closed.


The closed valve seals the combustion chamber, permitting the expansion of combustion gases to turn the crankshaft.

When the valve is open, it allows either combustion gases to leave the combustion chamber (exhaust valve) or a fresh charge of air/fuel mixture to enter the combustion chamber (intake valve).

I will avoid describing how pistons, rods and crankshafts work in hope that all here have that part conquered. If not, maybe a quick trip to http://auto.howstuffworks.com/engine.htm might be in order.

OK, now that we have the fundamental stuff all in one place, we now need to teach our engine to breathe. Air is necessary for our engine to run. We can provide fuel with a garden hose, but without air (or more importantly the 20.946% of air that is oxygen), nothing will happen.

The camshaft

How we get the air in and out is the job of the camshaft. How we increase performance and at what RPM range is highly dependent upon the camshaft and how it manages the valves.

In the next exciting episode of this series, we will discuss the need for different cam profiles at different RPM ranges, what the costs are and what we give up in order to get something else in return.

Caveat

If you feel all this is too elementary for you, feel free not to read. Do not feel free to complain about my ability to teach. As mentioned above, I'm doing my best.

[mazdarati]

Understanding inertia

OK, let's try and make this sucker run.

Extremely important in understanding performance aspects of camshaft & breathing is the concept of inertia. A lot of people are acquainted with inertia through Newton's First Law.




Originally Posted by Isaac Newton
Every object in a state of uniform motion tends to remain in that state of motion unless an external force is applied to it.

Of course we all know that. When your ride is sailing down the boulevard (that'd be a state of uniform motion) and the light turns red, you step on the pedal next to the throttle (this would be our external force effecting a change to the uniform motion thing).

While you're waiting for the light to change, you're also in a state of uniform motion (i.e. stationary) and once the light changes you add a little force (using your right foot) to change your state of motion.

To disregard either the red light or Newton's First Law is to invite punitive insurance rates, bent valves or both.

Filling our cylinder

In simplest terms, air enters our cylinder in an attempt to fill the space created when the piston moves downward (remember, we're talkin' NA here -- turbo stuff comes later). However, none of this is really that simple.

At low speeds (say a Mazda 2.2 idle speed of 750 rpm) we are experiencing 12½ engine revolutions every second. At red line (and this is a SLOW engine by any measure) it's rotating 100 times per second.

If we consider (for the moment) our intake stroke to be ¼ of our four stroke cycle (some 180° of crank rotation), then we need to complete our intake stroke in anywhere from 40 milliseconds (at idle) to 5 milliseconds (at redline). A millisecond is a thousandth of a second, so we have a lot to do in very little time. Five milliseconds is the duration of an eyeblink. We need to open our intake valve, move our piston from TDC to BDC and close our intake valve, all in the time it takes to blink.

Because we have a flywheel (as well as 3 other cylinders) keeping things going (inertia, remember?), the part about getting the piston from TDC to BDC is pretty well automatic. The part about opening and closing the valve is not as easy.

Opening our valves

In almost every modern 4 stroke engine valve opening is accomplished by a cam (a rotating wheel with a bump on it, if you will). A cam looks like this:



and when you hang a bunch of cams together on a shaft, it gets called a camshaft.



Valve closing is performed by a spring or springs acting in the opposite direction of the cam. There may be all manner of intermediate buckets, shims, lifters, pushrods or rocker arms, but ultimately the cam opens the valve and the valve spring (or springs) closes the valve.


The only current exceptions of which I'm aware are Ducati motorcycles which use cams to open and to close the valves (desmodromic valve gear) and a few F1 engine manufacturers (Renault?) which are experimenting with pneumatic (air-powered) valve actuation.

Anyway, We need to get something like a half-liter of air into our cylinder (and keep it in) in something like 5 milliseconds at redline.

To get that much air in there, we need to get that valve open and we need to do it quickly. To get the valve open we need to overcome (all together now!) inertia.

How quickly can we open the valve?

As we saw above, we have very little time to open our valve. However, if we make our cam too radical (open the valve too quickly), our valve springs won't be able to overcome the inertia of the rapidly opening valve and the valve will float (keep moving further open -- wider than the cam intended) and likely contact an unsuspecting piston (doing grievous damage to both valve and piston). When valves open wider or longer than what the design of the cam intended, it is called "valve float"

Valve float is not a good thing.

Of course valve float can be reduced/eliminated by increasing valve spring tension, but with that comes more breakage, higher wear rates and greater lubrication problems.

In practice, a well designed camshaft will gradually start moving the valve, then accelerate it increasingly briskly until it is about halfway open, then begin to decelerate as it reaches maximum lift, sorta like a sine wave.

Here's a little chart showing (stock F2 intake cam -- sorta) valve lift plotted vertically and crank rotation (degrees of duration) plotted horizontally.



The area under the curve is roughly equivalent to the amount of air ingested by our engine (dependent upon engine rpms). Once we understand that, then it become obvious that we can enlarge our "breathing window" by

* opening our valve longer (widening the window)
* opening the valve more (heightening the window)
* opening the valve more quickly (making the flanks of the curve steeper)


However, there is a price to be paid for each of those strategies. As mentioned above, to accomplish any of them while avoiding valve float stronger valve springs are required. Stronger springs bring a whole new set of problems.


Then we have to close the valve, so we go through the same thing in reverse. All in 5 milliseconds -- an eyeblink. And in the meantime, we must try to fill the cylinder with air.

To get the air into the cylinder, we must overcome (you know what to say!) inertia.

The air that's hanging around our air cleaner has inertia. It would just as soon stay there. But when that engine starts swallowing air (2.2 liter engine demands about 233 cubic feet per minute), it has to come from someplace.

Because we need to overcome that inertia of that air loitering around our air filter, it's gonna be difficult to completely fill that cylinder, but we're gonna try. Once we get that air movin' through our intake port, chasin' that piston down there, we want to take advantage of the inertia we've built up. Even when the piston has passed BDC, it still hasn't begun moving upward very quickly. Our column of air bombing down the intake port is still filling the cylinder, even though the piston is starting to rise. So why don't we delay that intake valve closing for a little bit? Let a few more molecules or air enter (which will allow us to burn a little more fuel).

Good idea! Now we're using inertia to make performance!

However, it's not all roses. There is always a downside.

Tune in for the next exciting episode of 4-stroke breathing.

[mazdarati]

Packing in more air

In our last episode, we were holding our intake valve open a little later, because our intake charge was, well, charging down the port, past the valve into the cylinder, even though the piston was starting to rise.

The first few degrees after BDC do not cause very much piston movement -- if the piston is sitting on the top of your wall clock, how much vertical movement will occur from the time the hour hand moves from 6 o'clock (BDC) to 7 o'clock (30° ABDC)? Have a look at your clock on the wall (not the digital one, Jack, the analog clock) to see how little vertical movement is achieved.

The trouble is, how long should we hold that valve open? It depends upon how much inertia that incoming charge has. At low RPMs, the intake charge is travelling fairly slowly; at high RPMs, the intake charge is really hustling along.

The rising piston can overcome the comparatively modest inertia of the low-RPM intake charge more easily. That is, if the valve stays open too late, intake charge gets pushed back into the port (some call this reversion, I don't like that term), resulting in less air-fuel mixture available to be burned (and to make torque).

This helps explain why a "radical" cam has such an uneven idle and such pukey low-speed performance. It trades off a little low-speed performance to enhance the high-speed torque range. It holds the intake valve open a little longer so that when the air is hurtling down the ports at high speed, it keeps a barrelin' past the valve for longer, even though the piston is trying to start its compression stroke. Ideally, our intake valve will close at the point when the two forces (the inertia of the incoming charge and the pressure of the rising piston) are in equilibrium (in balance) and the intake charge has slowed down and stopped but has not yet begun to head back upsteam into the intake port.

Unfortunately, this equilibrium is achieved at different speeds for different valve closing event timing. When everything works right (valve closes just as pressures equalize), you can pack in more air than a stationary cylinder of that size would contain.


The ratio of the actual amount of air inhaled divided by static cylinder displacement yields a number known as volumetric efficiency usually expressed as a percentage.

You can't have it all!

Write this down:

======================================
=== Volumetric efficiency changes with RPM ===
======================================

You pays your money and you takes your choice. A screamer with lopey idle and no low end grunt or stump-puller tractor-like low end performance but run out of breath at high Rs. Volumetric efficiency at one end or the other (or the middle, if you like) of the RPM spectrum at the expense of other RPM range(s).

Yes there are solutions, but not for our 2.2

While there are high-tech workarounds (VTEC changes the valve timing at different RPMs -- VICS changes the intake tract length, which changes the amount of inertia the intake charge has), for our comparatively low-tech 2.2, there are few such choices.

As we learned in lesson 1, the more air we ingest, the more fuel we can burn and the more torque we can generate. Increasing volumetric effiency is just a fancy way of saying we ingest more air.

Next episode -- exhaust duration and timing.

You! In the back! Wake up! Even if there's no test at the end of this, your snoring is disturbing others!

Let's turn the crank for a bit

In our last episode, looked at ways of increasing volumetric efficiency and how the timing of the intake valve closing event influenced volumetric efficiency in different RPM ranges.

Today we're gonna skip right past our compression stroke, even though that's the defining part of the Otto cycle engine and move on to the power stroke. The "power" of the power stroke is what creates torque, so we all want to maximize the pressure pushing on the piston in order to maximise torque.

However, as we mentioned back in episode #2 (I think it was), we have to deal with some serious inertia issues, both in moving (that is, opening and closing) our intake valve and in moving our fresh charge into the cylinder within a VERY short period of time (as little as 5 milliseconds -- an eyeblink -- at redline).

The exhausting part

Those same inertia issues affect our ability to remove the spent gases from the cylinder (exhaust stroke). Those same time constraints that encouraged us to leave our intake valve open a little while longer to increase volumetric efficiency are now encouraging us to open our exhaust valve a little (actual a good bit) early, as well.

The exhaust valve in our stock 2.2 F2 engine opens 55°BBDC. That is, if our crank throw is the hour hand on a clock (unless your engine is Honda or Corvair, it turns clockwise as seen from the "front" or pulley end), it's just barely past 4 o'clock when the exhaust valve starts to open. The gases are still burning and they make a fair bit 'o' noise once that valve starts opening, so you need a muffler to keep your neighbours from getting too pissed when you head off to work in the early dawn.

Why did Mazda decide to open it so early? Well, the expanding gases have pretty much done most of their expansion by then. As the piston gets lower in the bore, the pressure drops and also the temperature drops (that's either Boyle's Law or Charles' Law, I can never keep them straight), further lowering pressure. The residual pressure can be used to push itself out past our exhaust valve, helping to overcome (you know what comes here, doncha?) the inertia of our spent gases. By the time our exhaust valve is fully open (about 8:30 on our "clock" crankpin, that outrushing gas pressure has dropped to nearly atmospheric, but because it now has inertia as it heads down our header/manifold/whatever, it creates a kind of vacuum behind it, which helps to suck the remaining gases our of the cylinder (the V8 guys call this "scavenging" the cylinder).

As well, the rising piston helps to push out this residual gases as our exhaust valve starts to close.

As our piston gets close to the top of the stroke, it slows down (see how little vertical displacement on the clock face as we go from 11 to 12 o'clock), even though our crank is still turning the same speed.

Both (all) valves (slightly) open at the same time??

At this time, even though our exhaust valve is still partially open, we will start to open our intake valve(s). The period when both/all valves are open is called "overlap", and is one of the fundamental aspects of 4-stroke performance.

Next episode, we talk overlap.

As an aside, I notice that our audience has dwindled from the 22 eager faces at the start of the show to just you 4 diehards.

That's all right. At the end of the race, those folk who left early be a couple of laps down for their lack of curiosity and lack of persistence, and you guys who are sticking it out will be contending for the checker. 'Way to hang in there!

[mazdarati]

In our last exciting episode, we saw the exhaust valve open and (partly through the pressure created by the burning charge, partly through the pressure exerted by the rising piston and partly by means of inertia), we saw our spent gases heading out the exhaust port and down the manifold/header. As the exhaust valve began to close, we began to open the intake valve, so that both valves are partially open at the same time. This portion of the cycle is called

OVERLAP

A few (actually quite a few) bad things are associated with overlap -- the incoming charge is diluted by the residual exhaust gases remaining in the combustion chamber and trying to escape through the "wrong" port (hurting power), raw unburned fuel escapes down the exhaust port (hurting fuel economy and emissions, not to mention power output) and because of those first two things, idle is rough and low speed response is uneven.

So why should we have overlap at all?

If we're designing a engine to be operated at one speed only, like a portable welder or a generator (I guess they're both the same, eh?) we can design our valve events and our camshaft to work optimally for the governed RPM -- to make optimum use of the inertia of the intake and exhaust gases at the design speed. That is, aside from starting up or shutting down, the intake and exhaust gases are travelling through the manifolds and ports always at the same speed (sorta -- we won't discuss the exception here) because the engine is rotating at the same PRM.

But for an automotive engine, we need to make power at varying speeds/RPMs.

At varying speeds, the gases entering and exiting our combustion chamber exhibit differing degrees of inertia, depending upon engine speed. A low-overlap, low-duration cam that allows good low-speed response will be woefully gutless at high RPM ranges (but great for trailer towing and emissions). A cam that will make full use of the inertia of incoming air and exiting exhaust at high RPMs (by using long duration and lots of overlap) will be barely capable of running under say, 1500 RPM and will pollute to beat the band, due to raw fuel/air mixture exiting the exhaust port during overlap and due to the air/fuel mixture that stays in the cylinder, being diluted by exhaust combustion gases which do not escape out the exhaust port.

All in all, camshaft design is a compromise. You can have your power at low RPM or at high RPM ranges, but not both (see this comparative dyno showing how a stock-cammed, 22° overlap, lightly modded 2.2 turbo outshines a heavily-modded, cammed, 89° overlap, 2.2 turbo all the way up to 4200 RPM), although, as we learned back in episode #3, you can stretch the "good" RPM range with things like VTEC (changes duration) or VICS (changes intake runner length to "fool" with the inertia.

We can also, as we learned previously, open and close our valves more quickly (to get to the same lift quicker, or to get more lift in the same amount of time).

In order to "control" the valves (not let them float) we said we must use a higher spring rate to overcome the inertia of the valves and keep them from contacting pistons. That is not strictly true. We can also lighten the valves to reduce their inertia.

Reducing valve weight

The easiest way of reducing valve weight is to make our valves smaller. In and of itself, this can be sorta self-defeating. Smaller valves make for worse breathing, not better.

But wait! What if we make the valves smaller, but use more of them!!

By using smaller valves, we can avoid the disadvantages (camshaft and lifter/rocker wear, breakage) associated with stronger valve springs and still accelerate the valve off the seat quicker without the same fear of valve float, because the valves now have less inerita.

Why did Mazda use 2 intakes and 1 exhaust valve per cylinder?

In most NA engines, exhaust valves are smaller (usually around 80-85% of the size of intakes). The main reason for this is that we have only atmospheric pressure (14.7 psi) to push the air into the (NA) engine, but we have much higher (maybe 500-600 psi at exhaust valve opening time) combustion pressures pushing the spent charge out. We don't need as big an opening for exhaust with all that help from residual combusiton pressure. Remember what Isaac Newton said about being acted upon by an external force? Well, all that exhaust gas pressure is an "external force" which helps overcome the inertia of the the combustion byproducts (exhaust gas). As well, a bigger exhaust valve would be harder to open with all that cylinder pressure bearing on the other side of the valve head, again resulting in higher wear rates for camshaft lobes and rockers.

On the intake side, however, the larger, heavier valves will float sooner (given an equivalent spring rate) so in order to minimize wear and breakage, they put in 2 smaller, lighter intake valves.

By the way, those same issues about how much pressure is available to overcome inertia on the intake and exhaust sides also applies to port size as well as valve size -- you can get away with smaller exhaust ports than intake, because of the higher exhaust side pressures.

Next time around, we'll look a little closer at how to optimize (theoretically at least) the valve opening and closing events and why the rod length/stroke ratio is so important in determining when to open and close valves.

No homework tonight, boys and girls. Thanks for being patient.

[mazdarati]

In our last exciting episode, we examined the overlap period and how critical it was to which part of the RPM range our engine will make good torque.

Today, we're gonna have a look at when we want to do the most breathing during our four-stroke cycle.

The thing is, we want the valves open the widest when our pistons are travelling the fastest. Put another way, when piston speed is at its maximum, so is the need to move air. To move this air we need the valves WFO.

So when, exactly, is piston speed at its maximum? While it might be tempting to guess that, because the piston is stopped at TDC and at BDC, the point at which maximum speed is attained should be at 90°ATDC, right?

Actually, no. Maximum piston speed is achieved much closer to TDC than that, typically around 70-75° BTDC and the same ATDC. The exact point is dependent upon the stroke and the rod length and occurs when the angle formed by two lines -- the one joining the centre of the wrist pin to the centre of the rod journal (in essence, the connecting rod) and the other joining the centre of the rod journal to the centre of the main journal -- is a right angle.

Maybe to explain better, lets look at the home-made graphic (I was really, REALLY gona try to do this all in text, but I wimped out; sorry). Here's the picture:



Here's the bearing surfaces we care about:


Here is the bearing superimposed on the right triangle:


And here's the representation of the relationship showing the 90° angle:


At that instant (for an F2/F2T with a stroke of 94 mm and a rod length of 158 mm, it occurs around 73.43° ATDC -- if you know your stroke and rod length and a little trig, you can figure out your value) piston speed is at its maximum and (except for our old friend inertia) air flow past the valve should be at its greatest.

To get the most out of our engine (to maximize volumetric efficiency) our valve should be wide open, or nearly so, at this point. In reality, in order to keep overlap at an acceptable level and to achieve acceptable lift with realistic valve acceleration rates, our intake valve is still opening at that point and our exhaust valves are already starting to close when the equivalent point on the other side (73.43°BTDC) is reached.

The point is, we need to recognize that moving our cam lobes around, whether it is done by the camshaft grinder or by adjustable cam sprockets is done for the purpose of getting the air in and out more effectively at the desired RPM range while attempting to minimize the negative side effects at other RPM ranges.

I hope the point is getting through here that there are many, many factors at work here (all of them interdependent) and the people who equate "regrind" with "performance" need to do a little more homework. The combination that the auto manufacturer gave us is not bad. When we change things, we might improve things at some RPM but we will for sure lose at another RPM. Generally (for NA engine), the higher the maximum torque output, the narrower RPM range it will sustain for. You can have a flat torque curve (but not very high), or a high,sharp peak (but over a very narrow rRPM range).

Next time around, we'll examine just how our cam timing events differ in a forced induction (turbo- or super-charged engine but NOT nitrous -- no matter what the supposed illuminati say).

If we have time, we'll also look at the relative importance of valve timing events and valve lift. Class is out.

[mazdarati]

We no longer need to hold our valves open longer or open them wider to get more air into the cylinder -- our pump is feeding it under pressure. To use an analogy, think of how much more fuel comes out of your injectors when a fuel pump is pushing it out and how little would come out if we depended upon gravity to let it dribble out.

Reducing overlap

Well, we can now pack the air in without needing more duration and/or more valve lift. And we should. Why? Because of all the negative aspects of overlap as we discussed back in chapter 5, especially with regard to low speed response and torque. Less overlap means better low speed response, better fuel economy and better torque in the RPM range where our turbocharger is least able to assist us.

When we reduce overlap, we necessarily reduce duration. Does this hurt performance? No, because our intake charge is under pressure. We can put more air into the cylinder with less duration, just by pumping it in.

The folly of "big" duration

Could we induce more air (and therefore increase torque) by increasing the duration and boosting? Probably not. Most of the pressurized air would get blown straight in the intake and immediately out the exhaust during overlap. Not only would this kick hell out of fuel economy, but would result in rich mixture and lower power.

Overlap is unnecessary and counterproductive in a turbocharged engine. Folks who yearn for a "bigger" cam in order to upgrade their turbo engine (without even knowing what "bigger" consists of) will usually be disappointed.

If the "bigger" pertains to lift only, we can easily change a non-interference engine into an interference one. Also, more lift without increasing duration means faster valve opening and closing rates, resultin in earlier valve float. If we add stronger springs to raise the valve float RPM limit, we start running into breakage (both breaking springs and also sucking valve stems right off the head -- an event that will hole your piston better than any amount of detonation).

As well, the only aftermarket camshafts available for the F2 engine are regrinds, meaning the manufacturer takes a stock camshaft and removes some lobe material to change the profile. When this happens, rocker geometry goes all to hell and valve guide wear and valve seal wear escalate like crazy. It also takes more horsepower away from the flywheel needed to push the valves that much more sideways once we've wonkered up the rocker geometry.

Finally, most factory camshafts have some surface treatment, phosphating, nitriding or induction hardening to reduce wear on the lobes. Guess what happens to that surface treatment when a cam is reground?

All in all, regrinds are not the hot tip, 'specially in a turbo engine.

Event importance

Too late for this now, maybe next week. School's out!

this is where the information stops, many more pages were lost to data corruption

[Cattman]

good god, great info bro. thanks