posted 05-16-2000 08:39 PM            

Can I pick your brain on camshafts for my Cleveland?

I currently run the Ford Motorsports M-6250-A341 cam, .510"/.536" 214/224 {292/302} and I really like it.

But I've been considering changing to a roller cam because I want to get more torque under the curve, which you know is real handy for a 351C 4V.

So considering my car will weigh 3400-3600 lbs when I put it back in street configuration, with my 4.30s and a C4 with a 3000 stall converter, what would you suggest for a good roller?

I know you mentioned before to not bother with a hydraulic roller.

I'd like to keep it under 7000 to save my valvetrain, but I want power!

By the way, in case you don't remember, I have '70 quench heads, 10.7:1 compression, Boss studs and guideplates, Torker manifold, and I'll be changing to a 750 or 800 double pumper.

posted 05-17-2000 07:53 PM            

------------------
Alex Denysenko Co-Administrator and Moderator
Fleet of FoMoCo's too long to list!

posted 05-17-2000 09:57 PM            

But thanks for the recommendation of the Comp Cams 32-771-8. .634-.634 lift 110 lobe separation, 246-246 duration @ .050 and 288 advertised.

I don't like the fact that it's not a split cam. The Cleveland needs the help with the exhaust that a split grind provides.

Do you really think I can run that kind of lift on the street?

posted 05-17-2000 10:58 PM            

There are no restrictor kits in my cleveland, and I've never noticed cam/lifer/or lifter bore wear. It does drain the sump though, I ran the oil just below the dipstick lower mark and ran into oil pressure problems 'till I filled it up again.

Valve guide wear is another issue... I've gone through 2 sets of bronze k-liners in that many miles (intakes are worse for some reason?) and it's still not clear what is the problem (help! Alex?). High lift and heavy springs can't be helping? The valves seem to survive.

After my experience with a hydraulic roller I'd be looking to a solid as well ...(noise aside?). Rollers need heavy springs - which hydraulics don't like? Somebody (who might know) told me that the internals of a hydraulic roller follower are the same as a conventional high performace flat tappet hydraulic. Can anyone confirm or dispute that comment?

Re: the arguement about the Boss 351 cam - I suspect compared to the cams that Alex uses the two cams described are "same ballpark".

posted 05-17-2000 11:51 PM            

when the guides flogged out the first time (intakes only) the shop I got to do the heads also said "geometry". After freshening up and replacing the heads we checked the valve to rocker contact by putting bearing blue on the valves and turning the motor over. The intakes were loading to one side, so we installed shims (60 thou from memory) under the intake rockers, which seemed to get the contact into the middle of the valve. My rockers are adjustable (screw and locknut on the pushrod ball), so no need to change push rods?

That was about 12 monts ago. Since then the guides have worn again, but not so badly (about 10 thou) and seems to be both intakes and exhaust (supporting the geometry diagnosis, yes?).

How would you check the geometry???? Does brg blue sound right?

What about lubrication? Are some stem seals better than others? Some of the stem seals are coming adrift and working their way up to the retainer.

Thanks in advance for you suggestions..... it'd be great to get this thing sorted and working properly.

posted 05-18-2000 07:56 AM            

I believe the technology is probably better today, both with the reliability of roller lifters, and designs of cam profiles, thanks to the computer. I know the roller design allows larger lifts, but Alex, you shocked the hell out of me with the .644" rec. That's stout! {Oh-oh, get out the Play-Doh!}

posted 05-18-2000 11:55 AM            

Lighten up Alex. This forum is for an open exchange of ideas and experience.
My reply was not meant as a personal attack. I was simply pointing out the
difference in the specs of the cams.

>Anyone that does not use a bronze gear with a billet roller camshaft per "ALL" the manufacturers
>recomendations is a "FOOL" and deserves what he gets! A pile of greasy powdered metal!

The cam gear was designed and specifically recommended for the solid roller by the camshaft
manufacturer (Crane). It may have simply been a quality control problem. The OEM's make
the steel gears live with the hydraulic roller cams so I suspect the materials technology is
available.

>You will only get into lubrication problems if you restrict the oil to the cam bearings too
>much. That's the only "secret". Everyone that builds a Cleveland is so pre-occupied with
>restricting the oil to keep it on the mains, that they get carried away.

That's a good point.

>6000 miles??? How many people put that many miles on a "HOT" S&S engine a year anyway?

I'd certainly like to be able to. Last year, I put 5000 miles on the car that I want
to build the engine for.

>I have only been using roller cams for about 24 years now so I realize that I'm a little "NEW"
>to this, but bare with me. If properly lubricated and adjusted a quality billet cam and lifters
>will last indefinaly. Roller bearing lifters will only fail from lack of lubrication. I have
>put well over 20K miles on a Crane roller equiped vehicle over a 9 year period with the same
>Crane bronze gear also. The oil gets changes too often, and the lash gets set too often, but
>it keeps on running fine.

Interesting. That's much better life out of the bronze gear than anyone else I've
talked to.

>I'm sorry to say, ask the guy who's "been there, and done that" for the real deal straight story,
>not someone who's "heard" about it, or "read" about it! NUFF SAID!

I did ask the guy who did it and saw the aftermath. It wasn't pretty. In all three cases,
I spoke directly to the source. I've also spoke with a couple of guys, including one with
the Crane gear, that seem to be making them live on the street but the miles are still low.

>With the size of the 351C ports we (and the cam companies) have found that the larger exhaust
>duration actually hurt performance and caused more reversion.

I've wondered about that. With the large port size, does the flow slow down too much to
effectively scavenge the exhaust port? Is the effect the same for lift and duration or
would you still want to run more lift on the exhaust side, even if duration is similar?
With milder street type cams, I would think the split pattern would still be desirable.
Something to test through simulation, I suppose.

>Remember we did not design cams with computers back then and could not simulate conditions
>like we do now. You have to trust me on this one Buddy!

What simulation program are you using?

>Alex, I can only say that you absolutely have a valve train geometrry problem.

That brings up a good question. Assuming proper valve train geometry, what sort of
valve guide and valve seat life would you expect from a cam like the one you described
(0.634" lift)?

>Dan, those failures you've seen were caused by the steel distributor gear, I'm sure. Everything
>I've ever read said you never use a steel gear on a steel camshaft.

Are you sure that wasn't iron gear on steel camshaft? If I'm not mistaken, the OEM's use
steel distributor gears with their steel hydraulic roller cams.

>Alex, please enlighten us {me?} on CFD.

Computational fluid dynamics, a method of predicting "fluid" (including air) flow. The flow
of fluids can be described by a set of partial differential equations known as the (unsteady)
Navier-Stokes equations. Except for a few trivial cases, the N-S equations cannot be solved
directly. The general solution approach is to define a mesh of nodes over the surface being
investigated. On the mesh, the N-S equations can be decomposed into a large set of
simultaneous nonlinear equations which can be iteratively solved. For any sort of non-trivial
surface, the number of nodes required and thus the number of equations to be solved is enormous
so very powerful computers (we use Cray mainframes) and sophisticated methods of solution are
required. CFD is used for certain problems in aircraft aerodynamics but it is still only a
supplement to wind tunnel testing and other theoretical approaches.

I don't know what the OEM's are using but the engine simulation software that I use, Dynomation,
uses an approach called the method of characteristics which is what we use for in our hydraulic
systems transient analyses.

posted 05-18-2000 02:09 PM            

Do you feel your computer simulation program is accurate?

posted 05-18-2000 02:28 PM            

They recommend spring # 929-16, which has 150 lbs seat, 415 lbs open. That's not that bad.

posted 05-18-2000 06:04 PM            

It seems to be accurate given accurate input information, otherwise it's
GIGO (garbage in garbage out) just like any other piece of software.
For an accurate simulation, you need lots of information like head flow
versus lift, cam profile (lobe lift versus angle), induction and exhaust
path lengths, carb flow, header and collector dimensions, valve sizes and
minimum port areas, intake manifold inlet and exit areas, bore and stroke,
rod length, compression, etc. It will also simulate forced induction.

Certain assumptions appear to have been made. There's no way for the
program to know if the carb chosen is atomizing or not, for instance.
So you can increase carb size beyond all reasonable bounds without
incurring any low speed metering penalties. When using a plenum style
intake (independent runner is also an option), it assumes an optimized
single plane. There is no way to vary the plenum volume or to simulate
a dual plane manifold. You can input cam card opening and closing
information (oddly only at 0.006" lift) as a first cut but for accurate
results, you need a lift versus angle profile. There's also no way to
vary any friction or thermodynamic parameters.

I view the output as sort of an upper bound. I've been trying to gather
dyno sheets from friends and club members to guage the software but few
of them have the sort of detailed information I require.

Unfortunately, Dynomation is expensive at $600 a copy and, at that price,
you think they could update the DOS interface. Actually, the software is
free but you have to pay the $$$ to get the printer port key (apparently
an active device) that allows the software to operate in anything but
demo mode. Most of the other commercially available programs are just
computerized guessers. The only real competitor to Dynomation that I'm
aware of is Engine Analyzer Pro at around $460. If any one is using Pro
(or any other high fidelity simulation), I'd like to compare notes

posted 05-18-2000 11:01 PM            

Alex,
I have a Crane roller hydraulic cam that I thought I would use in the 393 Boss stroker I am building. It will have .644/.644 lift, 238/238 duration at .050" lift, 278 advertised and 112 lobe separation. This is just a bit 'smaller' than the minimum you suggested to Steve for use in his 351 C/vehicle. I'm planning primarily street use for the engine/vehicle. Your thoughts ... is it 'acceptable' or would you suggest that I go for a solid lifter roller with tighter lobe separation and more duration?
Thanks
Gene

posted 05-23-2000 02:46 AM            

Your an engineer! I was kinda cautious about wading into CFD because it's heavy and I didn't want to put anyone off, (or try to suggest that I understand it except in a very general way). Actually, my impression is that you guys in the US are pretty open to this kinda stuff and don't mind discussing it?

I was wondering about simulating the airflow through ports/ past valves to get some idea of how the inertial terms are behaving and get some feel for "scavenging" where the inertia of the gas in the port keeps the gas flowing even though the cylinder is already "full". Ditto exhaust. . It might also be interesting to look at port/manifold missmatch to see what kind of turbulence (air/fuel mixing) that gives. Actually the possibilities are enless.... but like you say very non-trivial to code and solve (I suspect larger auto companies already have proprietry software, possibly simulating different parts of the motor separately?). Imagine parameter estimation for different cam profiles through a rpm range - or even simulating the behaviour in a V8 intake!

You say you have a CFD package for only $600? And it will do all that stuff? Sounds like it's already programmed for a motor - you just drop in you constants? Or am I getting confused and you're not talking CFD anymore?

Back to the rocker geometry... Alex - what about if I lift the heads and remove teh valves. I should be able to see how the guides are wearing - ie valve pushed out at the top, or pulled in? Then I can install lash caps or rocker spacers respectively? Because the rocker tip follows an arc centred at the fulcrum, the tip starts to the inside of the valve centreline. As the valve opens the tip contact moves across to the far side of the valve centreline, and then back to the inside again at full lift (pull your rocker covers off and have a look at the contact as you spin the motor over, not you Alex! - I can't imagine this is news to you!). Anyway, there is no way to keep teh tip centred in the middle of the valve as the rocker moves up and down, so some (?) compromise must be prefferable. Sorry to bombard you with questions Alex, but I (we?) don't get many oppertunities to pick the brain of someone with so much experience with performance motors. Mostly guys just want to sell you stuff!

posted 05-23-2000 06:52 PM            

Yup, undergraduate degree in aerospace engineering and graduate degree in control systems engineering. I work for Boeing (formerly McDonnel Aircraft) in aerodynamics and flight controls. Peace on Earth and death from above.

>I was kinda cautious about wading into CFD because it's heavy and I didn't want to
>put anyone off, (or try to suggest that I understand it except in a very general way).

Anything's fair game if it'll make your car go faster.

>You say you have a CFD package for only $600? And it will do all that stuff? Sounds
>like it's already programmed for a motor - you just drop in you constants? Or am I
>getting confused and you're not talking CFD anymore?

We're not talking CFD anymore. A CFD code that would accurately simulate engine
dynamics would be very complex and not likely to run on a PC. The program I
mentioned is based on the method of characteristics. It is indeed programmed for
an engine and makes some assumptions on it's layout.

>I was wondering about simulating the airflow through ports/ past valves to get some
>idea of how the inertial terms are behaving and get some feel for "scavenging" where
>the inertia of the gas in the port keeps the gas flowing even though the cylinder is
>already "full". Ditto exhaust. .

That can be done with Dynomation. For instance, you can vary intake manifold runner
length and port area. For the port area, you enter an inlet area (the area where the
runner opens into the plenum) and the minimum area. The program then calculates a
taper angle. It turns out a small taper angle (from a larger to smaller area) usually
makes best power with a wet flow manifold. Going from a larger to smaller area speeds
the flow up as it travels down the port which helps keep fuel in suspension. This is
one reasons why the Strip Dominator is a better intake for a 351C-4V than a Torker.
The Torker has the opposite area change, so flow slows down as you travel down the port.

Valve diameters, exhaust pipe and collector length and diameter, cam timing, compression
ratio, fuel-air mixture, boost pressure, and much more can likewise be varied.

>It might also be interesting to look at port/manifold missmatch to see what kind of
>turbulence (air/fuel mixing) that gives.

Dynomation does not have that flexibility. On the exhaust side, it will simulate
stepped headers, though. Our hydraulic transient anlyses programs here at work
do have the flexibilty of mismatched areas and are also based upon the method of
characteristics, so the theory supports its. Dynomation just doesn't implement it.
Dynomation assumes the geometry of the intake manifolding, either a single plane,
open plenum, or an independent runner configuration.

>Actually the possibilities are enless.... but like you say very non-trivial to code and
>solve (I suspect larger auto companies already have proprietry software, possibly
>simulating different parts of the motor separately?). Imagine parameter estimation for
>different cam profiles through a rpm range -or even simulating the behaviour in a V8 intake!

Oh yes, you can be sure the OEM's have invested vast sums of money in that sort of R&D.

posted 05-24-2000 03:43 AM            

Your comment about manifolds is interesting, and not what I'd imagined (fluid dynamics is a ripper topic huh?). I'd imagined it would be better to go from small to large area (reducing the flow velocity), thereby taking advantage of the gas inertia to carry it through the port.

I hadn't considered atomisation, but I'm pretty interested in that too. In another post I mentioned I've been having problems with the exhaust crossover in the intake. I've tried blanking it off a couple of times now - as per the "magazine" articles that suggest it to increase the intake gas density and therefore power. Doesn't work at all for my application. Goes slower with no heat. The cause appears to be fuel atomisation (vapourisation), because I end up with black plugs, especially in the rear pots (presumably raw fuel runs "down hill" when the car's nose is up on hard acceleration). Jetting makes no difference, and as soon as I open the heat passage the problem goes away. I've been told that the 3310 750 vac I run has "atomisation problems" anyway, and I should fork out a bunch of bucks for a new carb with annular discharge boosters. Leaving the manifold heater connected seems a slightly cheaper option at the moment....

What does your program say about valve shrouding? My 2V heads have shaved down 4V intake valves (2.125). These were reccommended to reduce the combustion chamber shrouding, which seemed a plausible suggestion at the time. I guess that's a major advantage of hemi (and to some extent canted) combustion chambers?

Stepped headers - this really is a fantastic thread!!!!!!!!!!!!!!!!! I read a magazine article (please don't stone me!!!! anyway I doubt you can throw this far!) that had a set of headers made up that continually went to a larger bore (lots of welding!). Is that what you're talking about? A tri-y design must do somthing similar yes? What's actually happening in terms of the gas flow? I've always imagined that as the pipe steps up a size it causes a reversion of the gas pulse, leading to a slight vacuum pulse going back to the port and encouraging scavenging? (I'd like somebody to define "scavenging" before I display anymore ignorance!!) So, if you have a stepped header you could expect several reversion pulses, and get effective scavenging over a range of rpms???

What effect does air/fuel mix have on your program? Are you talking combustion thermodynamics or gas density, or both, or maybe somthing else entirely.....what an interesting topic!

What does the program say about port size? The 2V vs 4V arguement seems to go on endlessly...... Interestingly no other manufacturer has gone to such huge ports? But then I would say that wouldn't I???

Actually, I'd really like to see a thread on porting. Or did I miss it? My understanding (or probably lack of it...) is that you open up the port under the valve (prefereably suiting bigger valves) so that the seat contact is at the outer edge of the valve. Also smooth the transition around the valve (ie 3 angle valve job). Also grinding the short turn radius seems to be popular, as does port matching - especially intakes. But fluid dynamics is so damn complex huh? The reason I reckon this might be a pile of XXXX (there you go Steve, I did the x's myself this time!) is that I once put a steamcleaner nossle into a diesel engine intake port (no valve obviously). The resulting spray of water out the other end was fascinating, it sort of came out in a vortex just above teh seat. Ie if you put the valve in, but lifted it about 1/2 inch above teh seat, then all of the water would pass through the gap, and not touch the valve. I looked into the port, and basically teh port entered at a tangent to the valve area, creating the vortex. Not sure if that would work on a gas engine and more rpms? The car heads I've looked at pretty much just head directly at the underside of the valve, more so after grinding out the short turn radius. The ports in my 2V heads get very narrow around the short turn, but I'm not completely convinced that it's just to clear all the valve gear. Perhaps Ford knew what they were doing? and the expansion after the restriction causes teh flow to spread out, thereby missing hte valve head??? The head shop I use wants to grind this area, and they supposedly have a flow bench so they can actually measure whether they are improving the flow or making it worse.....

Which reminds me of the flow-bench question I had the other day..... I assume (not having actually seen one) that a flow bench is basically an air pump that maintians a constant vacuum under the valve, while simultaneously measuring the normalised gas flow? That allows the flow to be measured at various valve openings, giving you the typical flow vs valve opening curve (in CFM per port).

But, and this seems to me the big but, a flow bench must then measure the steady-state condition when the valve opening and pressure drop are stable and constant. No engine actually works like that, the valve open and shut (rapidly!) and the vacuum depends on what the piston is doing. So how can a flow bench even hope to approximate the real dynamic situation? In particular the inertial or ram effect of a good manifold?

Nearly done!

The diesel truck motors that I used to work on almost always had solid roller lifters. Not terribly surprising since each lifter had to open two valves through a rocker bridge arrangement. Anyway, those motors generally went about 3/4 million miles between rebuilds. And when they did need re-building, the cam and lifters just went back in with a cursury visual inspection. Ie solid roller lifters can survive for ages under favourable condition (Dan you were expressing doubts about reliability of the needle bearings?) Of course diesel truck engines only go about 2200 rpm max, and probably don't have the spring pressures you'd need for a hot gasoline engine with roller cam. Distributor gear wear wasn't a real issue either!