Jim G’s SLURP Project – Posting 15: Selecting a camshaft
This posting is one in a series. If you have not yet read the earlier postings in the series, you’ll need to, as this posting won’t make much sense without having done so! Just do a search using the word “SLURP”.
In my last posting, I described Red’s new engine: a salvaged LS series “LQ9” engine from a Cadillac Escalade. In postings before that, I described how this engine would, in place of its normal powertrain controller module (PCM) computer, be equipped with a dual carbureted Edelbrock intake, and an MSD ignition box that would fire the factory coils based on signals from the factory crankshaft and camshaft sensors. This gives me the advantages of modern ignition without the attendant powertrain computer and huge factory wiring harness with its army of vehicle mounted sensors.
I also mentioned how the engine made 345 crank hp in factory stock trim, but how it is remarkably easy to hotrod it via cam and exhaust changes.
In this posting, I am going to describe the camshaft selection process I used. Note that the general approach is applicable to any engine camshaft selection process, and particularly applicable to the cam selection process for any of the LS series engines, like the LM4 5.3 liter in the SSR I just sold, and the LS2 6.0 liter in the 05/06 SSRs and in Corvettes. In fact, LS series engines are now used in ALL Chevy V8 engine applications, across all car and truck lines.
The process for choosing a camshaft is lengthy and involved, but can be conceptually broken down into the following steps:
1. Decide on how the vehicle, and thus its engine, will be used
2. List all the key vehicle and powertrain attributes relevant to the camshaft selection process
3. Specify the engine powerband parameters desired, based on the engine usage decided above
4. Recognize the constraints and opportunities within the physical engine design & construction
5. Recognize the constraints, opportunities, and unknowns/uncertainties within the vehicle with the engine installed
6. Set a budget range
7. Based on the above factors, make one list that describes “the ideal camshaft”
8. Either get specific advise from camshaft suppliers, or use engine simulation software that includes camshafts election functionality, to zero in on either a specific camshaft or range of acceptable camshafts, that satisfies the list made in step 8
9. Model the engine in engine simulation software, as errors, shortfalls, or impacts of uncertainties are handled much less expensively NOW than after buying the camshaft and its supporting other valvetrain components!
10. Refine the choice of camshaft based on the engine simulations
11. Buy the camshaft and all the required “supporting components” for a successful implementation
Phew!
Sounds like a lot of work, doesn’t it. But, it is that upfront planning work that helps to ensure a successful outcome versus an expensive disappointment. Camshafts are NOT easy and not simple. Even what appear to be rather “small” differences in things like duration, timing, overlap, and lobe separation angle have profound effects on the way that the engine behaves.
Here is how I executed each of these steps in an effort to make a good cam selection for “Red”:
Decide on how the vehicle, and thus its engine, will be used:Red is intended to be an “almost daily driver” (I work out of a home office, so no daily commuting usually). The engine needs to be predictable, easy to start, and have a dependable idle. There is no power steering or power brake loading on it, but there is going to be air conditioning, so again, we need a sensible idle. 92 octane fuel is reliably available locally, fuel cost is not terribly important (I only put on 650 miles per month currently) and so we assume that 92 octane fuel will be used. Highway cruising speed rarely exceeds 75 mph, and is typically at 65 mph.
List all the key vehicle and powertrain attributes relevant to the camshaft selection process:Red weighs only 3000 pounds – this is important to point out to any camshaft advisor, as most assume that a pickup truck weighs 4500 to 6500 pounds – because most modern pickups do. He also has a numerically high axle ratio of 3.90. This, coupled with his light weight, means that the engine need not be a “torquer” to get Red moving and to keep him feeling lively. This frees me up to go for some high rpm power.
The transmission will be a 700R4 automatic custom built by Phoenix Transmission for Red, with a torque converter that stalls at just 1800 to 2000 rpm. Greg Ducato, the owner of Phoenix Transmission, recommended this stall speed for reasons that will be explained in a future posting.
The induction system is an Edelbrock RPM Air-Gap intake, with two Edelbrock Thunder carburetors, each of 500 CFM flow capacity. This intake system, according to multiple tests I have read about, performs better than the stock fuel injection system at low rpm and at very high rpm, but may be somewhat inferior at midrange rpms. Its fuel mileage will be inferior to that of fuel injection.
Red is a pre-emissions vehicle, so he need not comply with emissions regs, and is not subject to any testing beyond a basic safety test at renewal time (every 6 years).
Specify the engine powerband parameters desired, based on the engine usage decided above:The 3.90 gearing, the tire size, and the transmission’s overdrive ratio all combine to result in just under 2000 rpm at 60 mph in overdrive. Because I live in “Texas Hill Country”, I want enough power at 2000 rpm = 60 mph to ensure that no downshifting is required during highway cruising. I have calculated that to mean that I ideally would like about 130 crank horsepower at 2000 rpm, which, after drivetrain losses, will still deliver about 100 rear wheel horsepower at that speed.
Beyond that, another important powerband characteristic is that there be no abrupt change in the power curve at some specific rpm point, as Red’s light weight and his front end weigh bias, coupled with marginal traction on a dusty or wet road, could make for some excitement I don’t need.
Recognize the constraints and opportunities within the physical engine design & construction:The LQ9 engine is basically a Corvette LS2 engine but with 10.08 versus 10.9 compression ratio, and a much milder cam. By changing cams, I am eliminating the key difference, although the lower compression ratio is also nice (The 10.9 on the LS2 requires detonation sensors that ARE activated all to easily, and I will not be running the computer necessary to react to them).
The engine is easily, and safely good to at least 6500 rpm, provided that the valve springs are changed to ones that exceed the capabilities of the stock LQ9 springs and also the “lukewarm” ones used in the LS2, and that the pushrods are changed to ones that are reliably stiffer.
Beehive valve springs naturally do not resonate, are rather lightweight, and have smaller tops that require smaller retainers to hold them in place. All these characteristics make them ideal for very high rpm use. Using titanium retainers in place of the stock steel ones adds even more insurance for stable and true high rpm operation.
Fully stock LS2 engines and the earlier LS6 Corvette engines have both been certified by GM for operation to 6500 rpm. LQ9 engines with properly selected valvesprings and pushrods have been spun to about 7200 rpm in a couple of tests I have read, but I intend to observe 6500 rpm as my ceiling. The point is that at 6500 rpm my engine will be safe, conservative, and still plenty powerful if I specify the right components.
The LQ9 engine also uses the same heads as the LS2 and Corvette LS6 engines, except with larger chamber to get the lower 10.08 compression ratio. These heads feature fully replicated ports (all cylinders have ports that are identical in size and shape, unlike on the earlier Chevy Smallblock engine), and a greatly superior valve angle. They breather like gangbusters even in stock form, generating tremendous high rpm power.
The carbureted intake system will help me achieve the low rpm (2000 rpm) power I want, and will also give me a boost at very high rpm (roughly 6000 rpm and above). GM Performance Parts Division now certifies in its own crate engine offerings that the carbureted version of the crate LS2 they offer is significantly more powerful than the injected crate version (see their website and do the dyno chart comparison yourself).
The Edelbrock RPM Air-Gap intake I am using is a relatively “high rise” intake for a street intake and is in fact optimized to run between 2000 rpm and 6500 rpm. It is also a “dual plane” design instead of a “single plane”, which sacrifices a bit of very high rpm power in order to enhance low and mid range power. It’s a great match for my target rpm range.
The LQ9 engine can safely handle at LEAST 0.590 inches of lift at the valves, but I set the “comfort” range for me as 0.550” to 0.575” maximum lift, so as to limit the valvetrain acceleration forces and promote durability.
Recognize the constraints, opportunities, and unknowns/uncertainties within the vehicle with the engine installed:The big opportunity here is that Red, with his high and boxy shape, has so much vertical engine compartment space that I can have just about any sensible intake system I would like. The Edelbrock RPM Air-Gap intake is a relatively “high rise” intake for a street intake (about 6 inches high). The carburetors add another 3 ¼ inches of height. The mammoth air cleaner assembly I chose add another 4 inches. ALL “modern” vehicles would cry “uncle” after just the intake manifold, but for Red, this 13 ½ inches of pure intake sysem fits just fine under his gargantuan hood! I could probably even, as two people have now suggested, stack an extra K&N air filter on top of the single element, and STILL fit! This is a huge plus for intake capabilities.
The big constraint is Red’s steering box. It is located right on top of the driver’s side chassis rail, and, unfortunately, protrudes about 2 ½ inches into the space between the rails, and even has one bolt that sticks out another 1.2 inch or so beyond that at one point. One of my suppliers, Chassis Engineering, has warned me that the steering box will likely need to be spaced outward a fraction of an inch to reliably clear their aftermarket engine mount. Equally important, and more relevant to this camshaft discussion, we may be a bit limited in our ability to install either large tube or small tube headers because of it.
The DEGREE of steering box interference is a big “unknown” as far as headers (you don’t go out on The Web and look up “LQ9 or LS2 engine headers for 1955 Chevy pickup”  ). Therefore, one of my constraints HAD to be that whatever cam I select must not be critically dependent upon large tube headers or even small tube headers. It must work with even the stock cast exhaust manifolds if necessary, without suffering grievious (as opposed to minor) power losses.
Set a budget range:I was pretty concerned about the overall (not just camshaft and valvetrain) budget for Red, but somehow everything seems to be falling miraculously into place. For the camshaft, valves springs, and pushrods, I had originally budgeted $600.
That was before addressing titanium retainers . . .
But, the more I thought about it, the more I realized it was shortsighted to be actually installing new valve springs and pushrods, and NOT to at the same time change to titanium retainers, since there is so much careful labor involved. So, I reluctantly asked one of my suppliers about the cost of titanium retainers, and it DID take my breath away.
So did the cost of the camshaft I ultimately zeroed in on, as it turns out to be a cataloged, but not “stocked” cam, that costs more than many others.
The final budget for the camshaft, valve springs, Ti retainers, and pushrods is about $850. Ouch! But, that valvetrain will be pretty durable, and very stable at very high rpm.
Based on the above factors make one list that describes “the ideal camshaft”:Summarizing all the above, what I went looking for was a cam with the following characteristics:
• Operates well in the 2000 to 6800 rpm range (bit of cushion for safety at the top end)
• Maximizes “area under the power curve” instead of peak power alone
• Ideal valve lift range is 0.550” to 0.575”
• Moderate valvetrain acceleration forces
• Work well with beehive springs that can handle 0.600” lift, and have over 300 pounds closing force at maximum valve opening
• Stable idle
• Ideally generate 130 crank hp at 2000 engine rpm
• Work well with a torque converter that stalls between 1800 and 2000 rpm
• Provide enough vacuum signal to correctly operate one of the two carbs at idle and low rpm (we are using a progressive linkage)
• Provide a progressive, versus sudden growth in power with rising rpm
• Exploit the LQ9’s high rpm capabilities and sensationally good stock breathing
• NOT be acutely sensitive to exhaust configuration (since we have unknowns)
zero in on either a specific camshaft or range of acceptable camshafts, that satisfies the list made in step 8:This part turned out to be harder than I thought it would be for two reasons:
1. There are a LOT of camshafts out there in the marketplace
2. Most manufacturers, and virtually all magazine test reports, list only the “duration at 0.050 inches”, and not also the “seat-to-seat duration (typically measured at 0.004” lift)
This second point is actually quite vital, as to be able to ACCURATELY model a camshaft’s performance in a specific engine, you need to be able to deduce its entire operating cycle behavior. That is not possible without having BOTH the 0.050 and seat-to-seat durations, PLUS all the other details of valve events (e.g. intake lobe centerline, intake valve opening point, etc).
I found one cam manufacturer that provided the required level of detail, Crane Cams, and that full disclosure policy is what earned them my business, despite the fact that their cam was one of the costliest ones I found.
I actually examined at least a dozen and a half of their cams, modeling them each in turn in my new engine simulation software, dialing closer and closer to the list of attributes I showed earlier in this posting.
Model the engine in engine simulation software:The software modeling turned out to be very educational, despite the fact that the software I could afford is limited in its abilities and degree of precision. (Really good engine simulation software costs tens of thousands of dollars!).
Even though you must remember at all times that the software modeling is VERY rough and subject to many uncontrolled or unknown or unspecified variables, I was amazed at how well it enabled me to at least eliminate cams that would be clearly in the wrong rpm range, or that would be too acutely sensitive to intake type or to exhaust type.
One Crane cam part number kept “winning” the comparisons I was able to do, so I finally called Crane’s technical support department, described what I had done, and asked how close or how far off the engine simulation software was.
The technician, “Tony” evaluated it in his own software for a couple of minutes, and then said that my current “best candidate” was actually quite close. But, Tony was still just a bit unhappy with the low rpm characteristics, since that 130 crank hp is so important to me, and he suggested another cam, that was just a few degrees shorter on both intake and exhaust duration. He showed me where the specs were, and suggested I try that one myself as well.
Good call on his part. Despite the milder durations, Tony’s suggested cam made only about 3 or 4 less peak horsepower, but added order-of-magnitude 10 hp virtually everywhere else. This meant that in an actual acceleration run, Tony’s cam would result in Red covering more distance in less time with the “milder” cam.
Refine the choice of camshaft based on the engine simulations:In running 1,850 (!!) further automated simulations after that, using the entire cam library provided in an add-on package to my software, I was able to only ever beat its “area under the curve” calculation by only a maximum of 2%, and only by sacrificing some low end power at the expense of very tiny gains at top end. That would be a negative for daily use, so Tony’s cam won the search competition!
I guess this might just prove that those cam manufacturer techies really do know their stuff!
Buy the camshaft and all the required “supporting components” for a successful implementation:This also turned out to be much harder than I planned.
First, the cam itself was about the most costly choice I could have made (I guess that’s WHY it performs so well). Secondly, those titanium retainers are easily the costliest items I have recently bought, on a dollars per ounce basis. Thirdly, that cam is a “special” cataloged but not stocked cam, that will take Crane 3 or 4 days to produce to order, and then shipping time gets added to that. Fourthly, the Crane dealer technician who is assembling the entire cam, pushrod, valve spring, and retainer order for me got sick, so we lost an entire day. I am hoping to get everything finally ordered and in transit very soon, but it realistically looks like it could be close to 2 weeks before all the pieces are actually here.
I keep reminding myself, the cam and other valvetrain components make or break the engine, so this is all worth it!
Summary:So, this is how I did it in my case. I’m sure that there are others who have done it better, especially if it is something they do frequently. I just tried to ensure that I did the very best job that I could do on a set of components that I regard as absolutely key to the real world performance and durability of an engine.
If it helps others, that would make me feel better about all the time and energy expended!
Here are two screenshots, showing the “likely worst case” and “likely best case” for the power curve for my engine, depending upon how well that intake actually works, and how many choices we actually have on exhaust when we fit the engine into Red.
http://i252.photobucket.com/albums/hh16/JimGnitecki/RedLQ9withCranecam-worst.jpghttp://i252.photobucket.com/albums/hh16/JimGnitecki/RedLQ9withCranecam-best.jpgJim G
