HOHN

Hi guys-- it's been awhile.

I thought I would post my ramblings on what a "good" cam approach for our engines would look like-- and more importantly, WHY.

First, some basics in camshaft definitions. Here are some reference terms many of you will already know:

Duration-- this is the degrees of angle over which the tappet is lifted above some defined point. It is stated in crankshaft degrees. For example, you might see "230 @ .050" as a duration spec, which simply means that the cam lifts the tappet above .050" for a total of 230 crankshaft degrees. In the gasser world, you'll commonly see "advertised duration" quotes at .006" (for hydraulic cams) and .020" (for mechanical cams). Because many people don't think a valve really flows much below .050" tappet lift, you'll commonly see duration quoted at that point. Another common reference is at .200". In general, these lift points refer to TAPPET lift, not valve lift so that rocker arm ratios are elimated as a variable.

Lift-- Quite simply, this is the maximum excursion of the tappet from the "base circle"-- how far the cam lobe will push the tappet before heading back down.

Base circle-- the portion of the camshaft where the there is no lobe-- the tappet is just along for the ride and the valve is closed.

Lobe centerline-- an imaginary line that starts at the center of the camshaft and radiates outward, perfectly bisecting a cam lobe. This marks the point of maximum tappet lift, and serves as an angular reference point. There are both intake centerlines and exhaust centerlines for the respective cam lobes.

Lobe separation angle, or Lobe Centerline angle-- this is the angle (in cam degrees) between the intake and exhaust centerlines. I will use the term Lobe Separation Angle or LSA as it is more descriptive and avoid confusion with the intake centerline.

Installed centerline: This compares the time phasing of the camshaft relative to a fixed reference. Generally, the intake side is used, if the installed centerline is the same as the lobe separation angle, then the cam is said to be installed "straight up". If the Installed centerline is lower than the lobe separation angle, then it's said to be "advanced". If the installed centerline is higher, then it's said to be "retarded."

For example, a cam with 110 degree LSA installed at 106 is said to be advanced 4 degrees.

Overlap-- the period where both intake and exhaust valves are slightly open because the intake valve opens before the exhaust valve closes near TDC. It is stated in crankshaft degrees, typically in the same vein as duration.


~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

So let's start with the THEORY of camming a performance diesel.


First, let's state that for any given combination of variables, only ONE optimal sequence of cam events exists. For example, if a change to a larger turbo, my "perfect" cam is not perfect anymore. If I dial in my cam for 2500rpm, it won't be optimal at 2900rpm. So we have to establish exactly what set of variables we are hoping to optmize.

Second, a turbocharged diesel engine has vastly different needs than a naturally aspirated gasoline engine. A naturally aspirated gas engine relies heavily on the “overlap” period to breathe effectively. The escaping exhaust in a performance NA gasser engine can “suck” on the intake five times harder than can the descending piston, so relying just on piston motion without optimizing overlap will cost you a lot of power. A turbo diesel doesn’t have this dependence. A turbodiesel also doesn’t need to worry about a large overlap period letting fuel escape and causing rough idle and high emissions.

Finally, cam specs only matter to the degree that they manipulate the engine’s airflow quantity and timing relative to what the engine requires. In other words, if the rocker arm ratio changes, the valve diameter changes, the compression ratio changes—ALL can cause a need to re-evaluate and alter camshaft specifications.

Another factor in a turbodiesel is the operating boost pressure relative to turbine inlet pressure. A cam that is optimized for having boost greater than or equal to TIP will not be very tolerant of turning the tables and running TIP higher than boost. The opposite is also true in that a cam that more tolerant of excessive TIP will not perform as well when the boost/tip balance is good. There is a tradeoff here to be made.

So where to begin in generating a cam spec? The first step in my opinion is valve curtain area relative to cylinder displacement. The larger the displacement is relative to the valve area available to feed it, the longer duration is needed for a given RPM.

Also related to valve curtain areas is the proportion of intake to exhaust areas. Most of the time we see the “75% rule” that exhaust curtain area should be 75% of the intake curtain area. It’s not that simple though, as this is for NA gas engines. An engine that has a more restrictive exhaust path will need a larger window for exhaust to flow through in the same amount of time. Similarly, an engine using nitrous oxide or other such “chemistry” will need a larger exhaust flow capability.

I don’t know that actual B series valve diameters, but if we peg it at, say 35mm, we get the equivalence of one large valve roughly 49mm in diameter. So we have a Liter of displacement (5.9L/6)trying to breathe through essentially at 49mm valve.

So is this “big” relative to the valve or “small”? Well, let’s compare it to a typical small block chevy 350, the most common engine I know of. With a 2.02” intake valve (51mm), the Chevy is feeding only .71L of displacement. The B series is feeding a full Liter cylinder with a smaller 49mm valve!

So, we can say that relative to this reference, the B series is “undervalved” and will need a longer duration for a given RPM to let the engine breathe effectively. This could be a good thing, because a longer duration allows the tappet time to travel further, meaning that we can lift the valves farther open and get better breathing.

So now we start to come to a ballpark for valve event timing. Since this is an iterative process, let’s start at a baseline and tweak it based on our analysis of it.

Let’s think of the four valve events (Intake Open, Exhaust close, Intake Close, Exhaust Open) and where we might want to peg them.

First is intake opening. We want to open the valve before TDC to make sure the fresh air has a chance to get in motion before the piston start down. More importantly, we want the valve to be near its maximum flow when the piston is at its fastest acceleration (max suction). In order to get to the intake centerline close to the point of max piston speed, we need to start opening the valve well before TDC. If boost is greater than TIP, we can open the valve a lot earlier without harm. If TIP is greater than boost, then we will be allowing reversion of hot exhaust back up into the intake tract. I’m assuming TIP is always below boost, so we’ll peg this somewhat early at around 25 deg BTDC.

Exhaust closing is a somewhat insignificant event in a Turbodiesel. This is because we aren’t relying on escaping exhaust for intake suction nearly as much, and we don’t have much to worry about losing a little incoming fresh air out of the exhaust (in a gasser this would be fuel, and hence power wastefully going out the exhaust). Just the same, the cylinder during overlap is under very low pressure, and adjoining cylinders are under higher pressure. For example, when Cyl #1 is at TDC with both valves open, cyl #6 is at BDC with the exhaust valve open at much higher pressure, meaning we are at risk #6’s exhaust ending up in #1 at overlap if we leave #1s valves open too long after TDC. Let’s peg this at 18 deg ATDC for now.

Intake closing is one of the most critical valve timing events because it determines how much fresh air we trap in the cylinder at a given RPM. Later intake valve closure will reduce bottom end torque and increase tq at higher RPM for a given valve compliment. Again, this is relative to both RPM and the displacement the valve has to feed. Since the B series has a somewhat short rod/stroke ratio, it is tolerant of later valve closure, and because of being “undervalved” it actually needs a little later closure. We’ll start with this event occurring at 60 deg ABDC.

Exhaust opening is another very important event because it strikes the balance between power extraction from the charge and effective exhaust blowdown. If we open the valve too early, we let pressurized gasses escape that would be better served pushing down the piston. If we open the valve too late, then we won’t have a good exhaust blowdown and the piston will be doing negative work on the exhaust stroke trying to “push out” the exhaust. Ideally, we don’t want the piston to have to push much at all—we’ll hopefully have the pressure mostly bled down before the piston starts up on the exhaust stroke.
Our engines have two qualities that point towards a need for an early valve opening. First, they are very high compression ratio, which means they also have a correspondingly high expansion ratio. That means most of the useful power is made earlier in the power stroke and we don’t lose much cylinder pressure opening the valve early.
Secondly, they have a turbocharger hanging in the exhaust that greatly restricts exhaust flow. This means that the “blow down” event will need more time, so we have to open the valve early to give it a head start.
One other interesting thing is that the turbine’s restriction allows us to recover to pressure energy back to the piston. If we had a super-duper high flow exhaust, the escaping gases would exert almost no pressure on the piston once the valve was opened because they could easily escape. But because the turbine restricts flow, the gases have something to “push against” to still do work on the piston going downward. So we’re going to try popping the exhaust valve open rather early to help blowdown and drive the turbine. We’ll set this event at 75 degrees before BDC.

Putting these valve events in order, we have this:
IO- 25 BTDC
EC- 18 ATDC
IC- 60 ABDC
EO- 75 BTDC

Because of how we quote cam specs, these events give us the following specs:
Intake Duration: 265
Exhaust Duration: 273
Intake Centerline: 107.5
Exhaust Centerline: 119
Lobe Separation Angle (LSA): 113
Overlap: 43

We find the installation offset by subtracting the intake centerline from the LSA. Since LSA is greater, this means the cam is installed “advanced” by (113-107.5) 5.5 degrees. If the LSA was smaller then the intake centerline, then it would be “retarded” by the difference between them.

You may have noticed that I’ve basically ignored valve lift as a spec. That is because it’s really one of those rare things were more is always better. Of course, I am assuming you have the appropriate valve springs, clearances, etc needed to run higher lift. Lift is something that really is balanced against durability, as higher lift cams increased tappet/cam wear, but especially deteriorate valve springs.

This is partly why the stock B series cam lift is so pathetically low. It is designed to let the springs and tappets last a LONG time.


Now the cam spec I generated earlier is sort of a generic performance spec that assumes you are staying within the stock RPM range. If you are shooting got the 3K-4K range, then the spec would change drastically. All the valves would be opening earlier and closing later.

Some possible tweaks to this cam spec that would be worth considering for towing and such might be closing the intake sooner (improves bottom end grunt), opening the exhaust even earlier (gives more boost and helps get heat out of the engine) and opening the intake a little sooner to help get cooler air in sooner. This would basically be “advancing” the intake lobe, which would reduce the LSA and intake centerline both.

Discussion, anyone?

JH

pickupman96

I just had to read that 4 times to understand it all. I think I just learned more that I did in 4 years of high school. That sounds like a awesome cam you came with specs for and if you want I will one if you would like to grind it for me...

kawi600

Im in the process of picking a cam for my truck and Ive figured out its not so simple as it looks. I have to pick a cam for the power goals on the truck. What RPM range I want to make power in, etc.
I could get a cam for top end and some good twins, but im assuming my low end will suffer.
Even gearing and tire sizes play into where that sweet spot will be.
I think Im going to have to find a shop that drag races and see what they use. The guy I have working on it now likes big singles and nitrous to spool. I dont know how happy Im going to be with no low end power.

HOHN

The problem with cam specs is that they are very heavy on the trial-and-error. Even Cummins with all its fancy CFD and FEA software still has to run a couple variants to get a cam spec that does what they want.

My proposed specs above are really just used to illustrate the thinking behind the effects of different cam specs. They are probably in the ballpark, if the ballpark is fairly large. They could be off as much as 12-15 degrees on either end, which is significant since cams spec changes of as little as 2-3 degrees will make a difference.

If I were filthy rich, I'd order up a couple dozen cams ground to different specs and test them with a heavily instrumented test truck over the course of a year or two.

But it's just me and my armchair engineering throwing out a SWAG for discussion.

HOHN

It may be presumptive of me to say, but I suspect that a lot of the diesel racing outfits probably don't have that strong a grasp of cam-specing for a CTD.

The degree to which a cam much be compromised correlates strongly with RPM range involved. If someone is following the high-rpm approach, then you will be giving up a good bit of bottom end to optimize the top end-- just like the hot rodders who have big hp and crappy idle and weak bottom end. Those guys optimize their gassers for 6K rpm and as a result they don't really have much going on below 2500 rpm.

Anyone hoping to make power at RPM higher than the stock 2700 (and most are) really should have a different cam. The factory cam is OK at 2700 rpm, but as RPM sweeps up past 3K and 3500, the factory cam is completely out of its league. You can make power up there with enough boost and and/or nitrous, but it's similar to how you can perform dentistry with a chisel and sledge--- you can, but it's far from optimal. The stock cam is just not a performance cam-- its mild lift doesn't let the head breathe well. Not that the head has a ton of flow to offer..

All that to say that unless you are shooting for optimizing RPM levels around 3500 and up, you need not give up much bottom end if any, imo. This is the kind of application where the Helix 2 would be good, imo. I think it might be a great cam for what I would want personally.

If you are going to shoot a generous amount of nitrous, then add a little more duration to that exhaust lobe and open it even earlier. This will show up as more duration and a wider lobe separation angle.

JH

Crimedog

Good topic Hohn, and very logical theory on the camshaft design.

Now, going off my ideas and what I have taken from your post, it is necessary to increase either lift duration or maximum lift to optimize the cam for higher RPM power. I see the main cam companies advertising higher RPM power, but wonder how they are doing it on re-grinds? To me it seems that you would have to add material to get a larger duration or greater lift. Anybody have any insight on this?

1-5-3-6-2-4

cams are or can be spray welded to build up material then reground to suit.

Flatbed24v

great post, I have always wonder how cams work. I don't know much if anything about the inside of these engines, but 60* abdc seems like you won't get as much air in as say you would at 30-40* abdc because the piston is on it's way up, so less volume for the air to go into? Unless the compressed air from the turbo is strong enough to fill the cylinder with the same/more amount of air as it would at bdc. Not sure if our turbos are compressing the air in or if it is just forcing it in but not really compressing it. I know that we can run any where between 20-60 psi but is that enough to compress the air ( at 60*abdc)enough to achieve the full amount of air that you would get at bdc? I may be way off on this, cause I am not sure if 60* is considered a "big" move. I am asuming that, that is 60 out of 360 of a full rotation. That seem like a lot to me. Other then that I like your theory on cams. When you say you have to give up low end to achieve the best high end high rpm, how much are you giving up? Is it enough to make you NOT want to drive it on a daily basis? To achieve a higher lift from the cam could you just put a spacer under the valve cover or is there some thing else that will be in the way? So if I understand this cam talk, then one would have to choose the power down low or up high in the rpm?

BC847

It seems I've heard or read often that useful flow doesn't really start till the valve is 50 thousandths of a inch off the seat (and vice-versa). Perhaps this could/can skew the original numbers. To be sure it could effect ramps and such.

Perhaps it applies more to a naturally aspirated engine.

HOHN

Regrinds need not be spray welded. The spray welding can be done, but it's not preferred as it produces a weaker cam for a given diameter.

Of course, the other main process used also produce a weaker cam as well, and that's simply grinding the cam to a smaller base circle.

If the concept of base circle is clear to everyone, then you can probably imaging how making only the non-cam part smaller can make the camlobe proportionally bigger-- PRESTO-- larger cam.

The negatives of a regrind are significant, though. First, you are quite limited in how much you can alter the cam lobes. If you don 't significantly reduce the base circle, you can only enhance the basic lobe design. You're not going to be able to redically shift LSA, for example. Second, you need to reaccomplish the heat treat if you regrind a cam because most cams are softer on the core and very hard at the surface, whether from induction hardening, nitriding, or what have you. If you re-grind the cam, you either make this hardened surface much much thinner or you grind completely through it. Either way needs completely new heat treating.

A regrind makes the makes the lobes more aggressive more than it increases performance. What I mean by that is while it gives a small increase in performance, it gives a much larger increase in the stress on the valvetrain. The ramp speed on the tappet is now more aggressive, the valves are getting bounced around more violently, etc.

For a milder peformance enhancement, I'd consider a regrind to be a band-aid at best. I personally think the cost-benefit is not worthwhile. To me (jmo), the hassle of cam installation is a do-it-once-do-it-right kind of affair. Others may not mind cam swaps so I'm sure their calculation can differ.

HOHN

You are correct that closing the valve earlier will trap more air-- under certain conditions. Under other conditions, [edit:CLOSING] the valve later will trap more air.

There is a ramming effect that occurs when the air starts rushing into the cylinder. This ramming affect can allow the cylinder to trap more air even whent he piston is heading up and theoretically pushing the air out of the cylinder.

For each engine RPM for a given engine there is an optimal time to close the intake valve. Close it too early and you slam the door on all that flow that was finally up to speed and rushing in. Close it too late and you close the door after the flow had already slowed down and was starting to get pushed back out the intake.

There are two main factors that push the "perfect" point either earlier or later. The first is the size of the cylinder the valve area has to feed. The more "undervalved' the engine is, the later you can leave the valves open because the ramming effect is more pronounced (or put another way, it NEEDS more time to fill the cylinder). So if you have a very tiny cylinder with lots of valve area (say, a 4-valve crotch rocket engine), then you would close the intake valves earlier because the engine breathes well. Conversely, a stroker big block engine would be comparably undervalved and would shift this optimal valve closure point later.

The second main factor is RPM. The point of intake valve closure is the main determinant of where tq RPM will occur in a NA application. (and thus we should optimize this before accounting for the turbo-- then optimize the turbo for this same rpm). The later I close the intake valve, the higher the RPM where it becomes "perfect" for a given combination.

Let's play with this using gasser examples to illustrate, since most of us are somewhat familiar with a muscle car style V8.

Let's say I have two engines that both represent sort of extremes. Let's say on one hand I have a Ford 302 with hi-performance cylinder heads with large valves. On the other hand, I have a stock-headed Mopar 440. Let's assume they have otherwise identical cam specs and we are only changing the point where the intake valve closes.

What we would see is that closing the intake valve at a given point (let's say 50 deg ABDC) would produce very different results. The Ford might make peak tq at 4800rpm, while the Mopar might make peak torque around 3000rpm-- with the SAME intake valve closure point.

Moreover, the cam is proportionally much more aggressive in the smaller Ford than in the big Mopar. The Mopar will have a much smoother idle, higher idle vacuum, etc. All because the same cam is proportionally miuch milder in the bigger engine.

A bigger cylinder needs a more aggressive cam-- longer duration, etc-- and also can tolerate it.

Returning to the B series, closing the valve at 60 deg ABDC would somewhat radical in a smaller cylinder, but with a proportionally large cylinder to feed, this is actually not that aggressive. On paper, it seems big because at that point, we've gone from a nominal displacement of 359 cubic inches to a mere 278 inches by reducing the effective stroke of the engine. But in the real world, it's not what would seem.

The largish cylinder for the small valve area pushes the rpm down for where the "sweet spot" of intake closure to occur. I would expect peak tq with my theoretical cam to occur somewhere between 2000 and 2500 rpm (again, considerin just NA-- no turbo).

Justin

Crimedog

That makes a lot more sense now. I had always been leery of their term "re-grind" because of the lost hardening properties you mentioned, but never knew exacty what part they would be grinding.

I guess this question would be more geared towards a vendor, but are the advertised "new" performance cams just stock Cummins cams that are ground down, or are they actually machined directly to the vendor's spec? I imagine it would be kind of spendy to machine your own cam from scratch...

I can see how making the base circle smaller to gain lift and duration would cause some crazy accelerations within the valve train. Good timing on the post, by the way, we just happen to be studying camshafts in our Machinery Design class

Flatbed24v

so if we are considering our valves to be small for our displacement, then wouldn't it be easier to bore out our valves seat and put a bigger valve in? With bigger valves and our stock cam it would make our cam seem more aggressive, right? Yet you wouldn't have to go threw as much R&D it get the right cam( maybe!!). I wonder what it would do to the torque and hp on the rpm curve?

Mike Holmen

Hohn, do you know any of the aftermarket grind specs? From what I know these engine like more overlap than stock (basically holding the valves open longer). You don't need lots of lift as the head don't flow tons of air. The centerline of both the intake vs exhaust should similar to you spec'd, just put more centerline on the exhaust. The reason that cummins don't do this is emissions. An aftermarket cam will allow more cylinder breathing clean fresh fuel and clean air (scavenging of the cylinder). I've messed with a few cams, but anything is better than the stocker. Head porting also helps on the top end, higher HP trucks.

dodgeman01

I can't wait to see where this goes. I'm looking into getting a cam right now. I bought a PDR street cam from a guy for 150 bucks but I'm still not sure its what I want. It was cheap and I can sell it for more than I got in it if I want. What I'm doing with my truck is sled pulling in the stock turbo classes. What I was is as many RPMS as I can get and of course as much HP as I can get. I might be going to the 2.6 turbo class becasue I have a turbo that will work in the 2.6 class. I would love to know how well my PDR street cam will work before I go through all the hassel of putting it in. anybody run one?
thanks guys this thread is great!!!
DM01