Possible insight into the TDB "cam wars"

[Joe R.]

Hello All:

Like many on this TDB, I have followed with interest the periodic debates that flare up here regarding the optimal cam specifications for our cars. Duke Williams believes that most of the aftermarket cams have too much overlap and have LSAs (lobe separation angles) that are too small.

Others question, rightfully so, how Duke knows something that none of the aftermarket cam manufacturers know. The aftermarket cam industry is very competitive and it has several major players. It's hard to believe that all of them are wrong.

I'd like to suggest that possibly both sides of this debate are correct, but they have been talking past each other. I believe the key difference lies in what assumptions are made about the exhaust system.

I'm using Engine Analyzer (same program that Duke uses) to simulate various options for an "L79 clone" 383 stroker motor I am building. I have simulated my engine using stock 2.5 inch cast exhaust manifolds and a stock 2.5 inch midyear Corvette dual exhaust with an estimated 500 CFM flow capacity. The simulation also includes my pocket ported heads. My goal is to have a wide torque band for street driving, with peak power at about 5500 RPM.

After choosing a stock 280XR Comp Cams cam (224 intake duration, 224 exhaust duration, 110 LSA) that seemed to come the closest to my goals, I decided to experiment with some of the cam parameters that Duke says are not optimal in the aftermarket cams.

I was surprised to find that by delaying the intake centerline and further delaying the exhaust centerline, the EA simulation showed significant improvements in average torque and power. I ended up with an LSA of 118 degrees, far larger than the 110 degrees of the catalog cam.

Note that for this analysis, I left the lobe profiles unchanged. The only parameters I altered were the lobe centerlines, which indirectly affected the LSA.

On the surface, it appears that either the EA simulations or the aftermarket cam vendors are wrong. As a sanity check, I changed the simulation back to the catalog 280XR cam and then changed the exhaust to headers with open exhausts.

Wow, what a difference! Compared to the exact same engine with a stock exhaust system, the open headers produced an additional 50 peak HP and 50 peak foot-pounds of torque. The improvement was nicely spread over most of the RPM range.

Now here is the interesting thing. With headers and open exhaust, I could not "improve" the power curve in any significant way by modifying the lobe centerlines from the values used in the catalog 280XR cam. Every change I tried made it worse. This suggests that the catalog 280 XR cam has been well optimized for open headers.

So, I submit that just maybe both Duke and the aftermarket cam manufacturers are "correct," but the difference lies in the assumptions about what exhaust system is being used.

For a vintage Corvette that has a stock exhaust system, it appears that maybe Duke is onto something. For the same car with open headers, the aftermarket cams (at least the one I simulated) seem to be just about optimal. The key here is that for a high overlap cam, even the modest back pressure of a stock midyear exhaust system is enough to dramatically impair the performance. Some (but not all) of this loss can apparently be bought back by making the type of changes that Duke has been advocating.

[Clem Z.]

Re: Possible insight into the TDB "cam wars"

chevy high performance solid lifter cams were designed to be run with open exhaust because back then most all racing associations required the racers to use stock cams but did allow open exhaust. road racers had to use the stock exhaust manifolds but drag racers were allowed headers. chevy wanted to make sure their car were out front at the finish line.

[Duke W.]

Re: Possible insight into the TDB "cam wars"

I've made the point many times in the past that aftermarket cams are optimized for headers and open exhaust, not the OE manifolds/production exhaust systems. This point seems to have gone over some peoples' heads', and I'm glad to see than someone else has figured it out.

I design my "restoration cams" specifically for the production manifolds and exhaust systems, which is what 99 percent of us run, and they turn out to be much different than both vintage OE and aftermaket designs, but they look very similar to modern LSX camshaft designs in terms of lobe phasing. Like I said in the Tale of Two Camshafts article in the Fall 2010 Corvette Restorer, there is a message there!

Look at the summary table of cam data. Both my designs are closer to the current LS7 cam in terms of POMLs than the vintage OE cams or typical aftermarket cams. The primary difference is the relative duration.

Typical massaged vintage big port heads (461X and later) yield an E/I flow ratio of about 0.80 compared to about 0.65-0.70 as machined by Flint. The LS7 head design intentionally compromises the exhaust port to get as big an inlet port/valve as possible, which yields an E/I flow ratio of about 0.65, and that's an okay design philosophy because it can be mostly compensated with a fairly early opening exhaust valve without hurting the bottom end torque too much, which yields signficantly longer exhaust than inlet duration.

In the case of massaged vintage big port small block heads the high E/I flow ratio is just the nature of the beast, so optimizing the valve timing to achieve maximum average torque/power across the rev range with the high E/I flow ratio and production manifolds/exhaust system yields shorter exhaust than inlet duration.

Also, in the article I summarized how I approached the design by starting with the same EVC and IVO points of the 300 HP cam in order to maintain OE overlap so as to not affect idle behavior. That was a primary design requirement. I then optimzed IVC and followed by optimizing EVO, and "optimizing" means achieving maximum average torque/power from 1500 to 6000 with a minimum low end torque spec for each design. After a couple of IVC and EVC optimization iterations, the final step was advancing/retarding the whole cam until the last ounce of optimization was achieved.

Engines are not one dimensional. There is much more to them than peak power, which most guys fixate on. A good road engine needs broad torque bandwidth to be responsive from off idle to peak revs, so for a road engine I establish a low end torque spec, then design to achieve maximum torque/power from off idle to peak revs within that low end torque boundary condition, and both designs required maintaining the OE idle behavior so they could pass a PV test.

For a vintage racing engine with a typical racing exhaust system, we (the owner and me) establish a "design speed" (essentially the redline) and a low end torque spec - typically 80 percent peak torque at no less than 3000-3500 revs, which is what revs typically drop to on the slowest second gear corner. Then, within this boundary condition I design for maximum average torque/power in the upper 30 percent of the rev range, which is where the engine will operate, shifting at the redline with a vintage close ratio Muncie four-speed; and aftermarket cam vendors like Crane and Crower usually have an off-the-shelf design that is very close to my specs, so the owner can buy the off-the-shelf design rather than having to pay the additional cost of a "custom" design.

Duke

[George J.]

Re: Possible insight into the TDB "cam wars"

Duke,
so would the 30-30 cam have been optimized for stock manifolds and open exhaust, or headers and an open exhaust? IE, road racing, or drag racing? Has that simulation been run?

George
31887

[Domenic T.]

Re: Possible insight into the TDB "cam wars"

Joe & Duke,
This is not pointed at either one of you but tat a couple guys that used these simulators?

I have had a lot of my time wasted by a few guys in the past that designed by simulator. One VERY pushey guy paid us to do it his way and have his 72 vette engine done in our shop the way he designed it on some simulator program he had. It was cam, compression and valve size with some porting and matching.

His 480 simulated HP turned out to be about 300 when he had it tested on the dyno. He traded the vette in before he had a thousand miles on it for a newer one because of his mistake that he tried to find blame for.

I thought his design was much like a truck engine and he said the 400 + HP would be made just under 5,000 RPM. He was well educated ( to educated) and understood what every thing was about but had NO practical experiance building engines, just a big mouth.

The other guy was not so smart , he built his engine , (simulatedof over 500 HP) and would add the % gained by headers, compreddion, cam, you name it to his best advantage. I heard it ran good but not near the HP on paper.

My point is that I don't think these programs are all that good. You really can't believe it till you run it.

Some paper work is good but build it, test it, then publish the results. That's the real HP gain or loss.

DOM

[Duke W.]

Re: Possible insight into the TDB "cam wars"

Duke,
so would the 30-30 cam have been optimized for stock manifolds and open exhaust, or headers and an open exhaust? IE, road racing, or drag racing? Has that simulation been run?

George
31887

I'm not sure what the 30-30 cam was optimized for. I think it was a desperate attempt to get more power out of the 327 for the SCCA wars against the Cobra, but no camshaft can make up for 1000 pound weight difference.

The 30-30 cam has too much duration and WAAAAAY too much overlap to work well with manifolds and mufflers, which is why I (and GM) recommend the LT-1 cam to replace the 30-30 when the engine is rebuilt.

As far as racing is concerned, the 30-30 responds very well to racing exhaust systems because of all the overlap while being very easy on the valvetrain, and I've recommended it to vintage racers, although "pure" racing cams with similar duration and more lift work better, but they can beat up the valvetrain.

Duke

[Duke W.]

Re: Possible insight into the TDB "cam wars"

Joe & Duke,
This is not pointed at either one of you but tat a couple guys that used these simulators?

I have had a lot of my time wasted by a few guys in the past that designed by simulator. One VERY pushey guy paid us to do it his way and have his 72 vette engine done in our shop the way he designed it on some simulator program he had. It was cam, compression and valve size with some porting and matching.

His 480 simulated HP turned out to be about 300 when he had it tested on the dyno. He traded the vette in before he had a thousand miles on it for a newer one because of his mistake that he tried to find blame for.

I thought his design was much like a truck engine and he said the 400 + HP would be made just under 5,000 RPM. He was well educated ( to educated) and understood what every thing was about but had NO practical experiance building engines, just a big mouth.

The other guy was not so smart , he built his engine , (simulatedof over 500 HP) and would add the % gained by headers, compreddion, cam, you name it to his best advantage. I heard it ran good but not near the HP on paper.

My point is that I don't think these programs are all that good. You really can't believe it till you run it.

Some paper work is good but build it, test it, then publish the results. That's the real HP gain or loss.

DOM

Most guys use a program (and its several clones sold under other names) called Desktop Dyno. I used it too years ago, but it will only simulate SAE gross (with STP correction the only choice) on a lab dyno. I switched to Engine Analyzer because can simulate SAE net (including SAE air density correction) with front end accessories and a reasonable exhaust backpressure model. From there you have to make the driveline/tire loss adjustment which is typically 0.85 for rear drive cars with 1:1 top gear manual transmissions to predict SAE corrected rear wheel power.

So bubba puts a big, high overlap cam into the simulation along with headers and open exhaust to get his magic 500 HP SAE gross, but then it will barely break 300 SAE corrected RWHP on a chassis dyno.

Once you put the engine in the car with a reasonable exhaust system, even with headers, the combination of high overlap and exhaust backpressure just kills VE and power - even more on the low end than the top end. Contributing to this are the losses from front end accessories (including bubba's junky flex fan) and the lower SAE air density correction which reduces everything another 4.5 percent.

And since this bubba motor has such weak low end torque it needs a 2500-3000 stall converter (yeah- bubba doesn't really want to deal with clutching and shifting) and he ends up with a not particulary powerful engine with no low end torque that gets 6-8 MPG around town. But it sounds real mean idling through the parking lot on cruise night.

For most configurations I have found that EA 3.0 is fairly accurate on horsepower. It's high on peak torque, but low on low end torque, so overall the real torque/power curves are flatter than predicted.

Simulating and testing a racing engine to SAE gross is okay because in the real world most racing engines do run headers and open exhaust, and the only extra front end accessory in the car is usually just an alternator.

You have to use the right tool and reasonable models, and I design to optimize performance across the range as the engine is configured in the vehicle. I don't just come up with some pie-in-the-sky peak power number. I system engineer the entire engine based on what's really important. In the case of a Corvette engine restoration the requirement is to maintain idle quality (which requires maintaining OE overlap), have a reasonable low end torque spec then work the internals for better head flow and the highest possible CR, which will maximize horsepower, but, I never establish a peak power requirement, even for a racing engine. For a racing engine I want maximum average power in the upper 30 percent of the rev range. If you design for absolute peak power the average in the upper 30 percent range may actually drop a bit.

Duke

[Ken A.]

Re: Possible insight into the TDB "cam wars"

Joe & Duke,
This is not pointed at either one of you but tat a couple guys that used these simulators?

I have had a lot of my time wasted by a few guys in the past that designed by simulator. One VERY pushey guy paid us to do it his way and have his 72 vette engine done in our shop the way he designed it on some simulator program he had. It was cam, compression and valve size with some porting and matching.

His 480 simulated HP turned out to be about 300 when he had it tested on the dyno. He traded the vette in before he had a thousand miles on it for a newer one because of his mistake that he tried to find blame for.

I thought his design was much like a truck engine and he said the 400 + HP would be made just under 5,000 RPM. He was well educated ( to educated) and understood what every thing was about but had NO practical experiance building engines, just a big mouth.

The other guy was not so smart , he built his engine , (simulatedof over 500 HP) and would add the % gained by headers, compreddion, cam, you name it to his best advantage. I heard it ran good but not near the HP on paper.

My point is that I don't think these programs are all that good. You really can't believe it till you run it.

Some paper work is good but build it, test it, then publish the results. That's the real HP gain or loss.

DOM

Well said! Too much info on the TDB is hearsay or something someone else printed & then is regurgitated as a "fact". I am all for real world dyno/ strip & street testing- just like "the General did it" back then.

[Michael H.]

Re: Possible insight into the TDB "cam wars"

I'm not sure what the 30-30 cam was optimized for. I think it was a desperate attempt to get more power out of the 327 for the SCCA wars against the Cobra, but no camshaft can make up for 1000 pound weight difference.

The 30-30 cam has too much duration and WAAAAAY too much overlap to work well with manifolds and mufflers, which is why I (and GM) recommend the LT-1 cam to replace the 30-30 when the engine is rebuilt.

As far as racing is concerned, the 30-30 responds very well to racing exhaust systems because of all the overlap while being very easy on the valvetrain, and I've recommended it to vintage racers, although "pure" racing cams with similar duration and more lift work better, but they can beat up the valvetrain.

Duke

Yes, the 30-30 cam was an attempt to get more power. And it did, as long as the owner/driver knew how to use the power. These engines were NOT designed to be used with low numerical differential ratios though. (I know this from experience. My very first brand new Corvette was a 64 coupe with a 365 HP, 2.20 1st gear and a 3.08 differential. What a miserable hopeless combination!)

In 64 and 65, there weren't many other production cars that produced the same kind of muscle inch for inch on the street, even through mufflers.

GM didn't replace the 30-30 cam with the LT1 cam because it made more power. This change occured in the early 70's because emissions were becoming more and more important and HP was becoming less and less important. The LT1 cam would be correct for the LT1 engine and could also be used to service the past Corvette/Camaro models that used the 30-30 cam. It was a cost reduction move.

[Duke W.]

Re: Possible insight into the TDB "cam wars"

Shortly after the LT-1 cam went into production, the 30-30 was discontinued, and the LT-1 cam was specified in the P&A catalogs for all previous OE 30-30 applications.

The Duntov cam continued to be offered into the early ninties.

It wasn't a cost issue. The cost savings from eliminating one part number is miniscule. The LT-1 cam replaced the 30-30 because the 30-30 was a lousy cam for a reasonable road engine application. It just killed the torque curve below 3000-3500.

Duke

[Michael H.]

Re: Possible insight into the TDB "cam wars"

Shortly after the LT-1 cam went into production, the 30-30 was discontinued, and the LT-1 cam was specified in the P&A catalogs for all previous OE 30-30 applications.

The Duntov cam continued to be offered into the early ninties.

It wasn't a cost issue. The cost savings from eliminating one part number is miniscule. The LT-1 cam replaced the 30-30 because the 30-30 was a lousy cam for a reasonable road engine application. It just killed the torque curve below 3000-3500.

Duke

Duke....

Why do I always get the feeling that you get a commission check for every LT1 cam sold?

Have you ever actually owned a car that had it's original 30-30 cam replaced with the LT1 cam? I'm talking about real life, not on a table top computer dyno.

By the way, you obviously have no idea what it costs GM for every part number in the system. Not just the parts, but the accounting and everything else that goes along with it. It's big dollars.

The period around 1971 was a time when GM was dropping a lot performance parts from the system. Lots of them. I still have most of the lists from the late 60's through the early/mid 70's and you would be surprised what was discontinued in the early 70's. That was the worst of the new emissions era and GM was no longer keeping parts that were not necessary or good sellers.
The LT1 cam was used in vehicles that were newer and still in warranty. GM HAD to choose between the 30-30 and LT1 cam. It was an easy choice. They couldn't install 30-30 cams in newer cars that had to pass emissions tests.
The 30-30 cam was only used for Corvette for two years and Z28 for three years. Combined, that would still be a very low percentage of total production.

The old Duntov 097, on the other hand, had a lot more applications and sales, including passenger cars, so the bean counters decided to keep that number in the system.

As others have mentioned, cam designs from computers don't always prove to be worth the time invested. Crane Cams had LOTS of proof of this. Wanna see some of the failed designs? What looks good on paper or on screen doesn't necessarily work well in real life.

[Duke W.]

Re: Possible insight into the TDB "cam wars"

No why would I need to own a car with an original 30-30 cam. I have guided two L-76 owners through the converstion to a "327 LT-1". The dyno tests on on the internet and both felt that the "327 LT-1" was a better configuration due to both better low end torque and more top end power than the Flint-built L-76.

Sure, it costs millions to maintain inventory, but ONE LINE ITEM. As far as volume is concerned, the last application for the Duntov cam was seven years old when the 30-30 was discontinued, and the number of '57 to '63 Corvettes with them was probably in the same ballpark as the number of Corvettes and Z-28s that were originally equipped with the 30-30.

By the early seventies there weren't too many who were buying the Duntov cam over the counter because the aftermarket and GM's own mechanical lifter cams has surpassed it in terms of power.

So, given your logic they should have discontinued the Duntov cam when they discontinued the 30-30, but the Duntov remained available for over 20 more years, and I doubt too many were sold after 1970. I know I didn't buy one when I rebuilt my 340 HP engine. I used... I'm sure everyone knows the answer, and at the time I made the change based on pure engineering intuition because I didn't have access to engine simulation programs back then. It took another 25 years to prove that my intuition was correct.

As I've said before, the LT-1 cam uses the L-72 lobe on the inlet side and the 30-30 lobe on the exhaust side, so if you want to hate the LT-1 cam, you have to hate the other two, also.

Duke

[Gene M.]

Re: Possible insight into the TDB "cam wars"

Hello All:


................ After choosing a stock 280XR Comp Cams cam (224 intake duration, 224 exhaust duration, 110 LSA) that seemed to come the closest to my goals,...............

Joe, I think the Comp Cam number you are pertaining to is 270H with a 270/270 overall duration and 224/224 @.050" with .470"/.470" lift, 110 lobe centers. This is a fast rise cam with only 46 degrees to split between rise and fall ramps below .050" lift. Understand this cam will greatly increase cylinder pressure due to shorter overall duration compared to an L79 (222/222) with 290 degrees. The lazy ramps of an L79 yields 68 degrees to split between rise and fall ramps below .050" lift.

Now on your "cam game boy" try the L79 specs with the fast rise/fall rates of the Comp 270H at 268 overall duration. I think you will find performance improvements.

[Joe R.]

Re: Possible insight into the TDB "cam wars"

Joe & Duke,
This is not pointed at either one of you but tat a couple guys that used these simulators?

I have had a lot of my time wasted by a few guys in the past that designed by simulator.....
DOM

Hi Dom:

I completely agree about the risks of using simulators. I'm an electrical engineer and I have been designing analog circuits for 30 years. I frequenlty encounter cases where less experienced engineers fall in love with their circuit simulations, and then fail to actually test their designs in the lab before committing to a production design. Sometimes the simulations are wrong.

That being said, the EA simulator is generally well regarded, so it's results shouldn't be dismissed without some careful consideration. On the other hand, some of the aftermarket cam companies are also well regarded, and their lead designers are very smart and very experienced. So, they can't be easily dismissed either.

This apparent contradiction has bothered me for some time as I have read the various TDB postings in what I refer to as the "cam wars." I found it hard to believe that all the major aftermarket cam manufacturers were wrong.

For me, the key insight was that perhaps both sides are right if you take into account the differences between open headers and stock Corvette exhaust systems.

As I noted in my original posting, I could not "improve" the catalog Comp Cams cam in my EA simulations as long as it was running through open headers. However, once I added a stock exhaust system with only modest back pressure, the catalog cam lost a lot of power. Cam timing changes in the direction that Duke has been advocating helped to buy back some of that power loss.

So, I suggest that in future debates about what sort of cam timing is "best," the speaker should clearly state whether he is referring to performance with open headers or with a stock exhaust system. It appears that the choice of exhaust system has a big effect on what cam timing is "best." In particular, even a small amount of back pressure seems to just kill the performance of a high-overlap cam.

In addition, as Duke has noted many times, the definition of "best" also depends on whether you are a racer who is mainly interested in peak power at high rpm, or a street driver who is more interested in average power over a broad rpm range.

Maybe we just need to be clear about the constraints and the goals when we debate which cam is "best."

[Ronald L.]

Re: Possible insight into the TDB "cam wars"

Coupled to a 4.11 I've never found a Lt1 I didn't like

The HP was always important - GM nor any OE was really up to the emissions legislation requirements thus it suffered to meet the law. They weren't asking customers if they wanted a smog equipment choked car or not with what they were pumping out the doors in the 70's...what was the first thing you did if you were not under the threat they'd pull you over and do a smog test???