Michael

Kevin - Quote: “...not so simple as it involves a Crank Angle Chart.” Actually, this is pretty simple – it’s just the crank-slider model. Sophomore engineering Dynamics class. Quote: “ ...peak power numbers will continue to rise or fall in the rpm range based on the point in time that your airflow goes supersonic.” This is not true. The airflow in the intake tract is NEVER supersonic. With a large enough pressure difference (around 8 psi at standard atmospheric conditions, for isentropic expansion; you’ll need an ever larger pressure difference for the real-life conditions) the flow will be choked (sonic) at worse. For it to actually go supersonic, a converging-diverging nozzle is necessary. Careful flow measurements in actual running engines have shown that the maximum instantaneous velocity attainable inside the intake port is limited to a Mach number of around 0.5. Jim McFarland’s articles (“Performance Professor” on the web, former Hot Rod contributor) point out that about 240 ft/s (about 0.2 M) MEAN flow rate is a good practical upper bound for flow rate at the peak torque (peak volumetric efficiency) point. Heat addition (which happens as the flow goes toward the combustion chamber) does result in an acceleration. Total (stagnation) temperature is increased, which for a calorically perfect gas in subsonic conditions will tend to drive the mean flow toward sonic. Of course, an air-fuel mixture (with trace water vapor and combustion products from reversion into the intake) is not a calorically perfect gas, so the equations of motion have to be integrated numerically to arrive at the “proper” result (simply put, the constitutive equation for anything but a calorically perfect gas is too complex). The full answer comes from integration of the Navier-Stokes equations – for multiple species, with the “appropriate” turbulence model, and proper description of inflow and outflow boundary conditions. Specific models have been developed for internal flows, but they’re basically just educated guesswork. Commercial CFD houses such as Fluent will sell you a software package that attempts to calculate this, but it’s still largely voodoo mathematics – which they will themselves admit. Characteristics (infinitesimally weak rarefaction and compression waves) move at the local drift velocity, plus or minus the speed of sound. So in a coordinate frame relative to the local drift, they are actually sonic. It is management of the characteristics – understanding their geometry in an x-t diagram – that’s responsible for phenomena like ram tuning – but that’s another topic. Quote: “3) When the sound barrier... occurs at 0.55 to 0.60 of the speed of sound”. The statements about sound barrier are not accurate, but you’re right that the practical “speed limit” inside a port is around 0.5 of the local speed of sound. Quote: “LPV = (0.0353 x RPM x S x B^2)/CA” Well, if CA is an area, then the units for LPV come out to velocity, not volume. Quote: “Your intake manifold’s cross sectional volume needs to be 0.80 of your LPV of your cylinder head’s intake port cross sectional volume...” “Cross sectional volume”? Perhaps you really meant “cross sectional area”? Quote: “CA = ... = 1.93 sq. inches” Using your formula, the units for CA now come out to a length cubed per unit time – in other words, a volumetric flow rate – and NOT an area. However, consider the following: we want to relate the engine’s airflow needs to port cross sectional area and port averaged velocity, right? The engine ingests air only when the intake valve is open. Since we’re talking about port velocity and port area, and have left the valve curtain area outside of the discussion, we need to make some assumptions. Ideally we would calculate the area under the curve for the valve opening event (if we had data from the cam manufacturers on valve opening amount vs. cam angle, and typically we don’t). That area is conceptually equivalent to holding the valve open at a larger, constant lift, but over a much shorter duration. That larger lift would give a valve curtain area larger than the port cross sectional area. Calculating the right equivalent duration is difficult, because it’s not enough to consider steady-state effects; we need to consider unsteady gasdynamics. But we’ll guesstimate. My guesstimate is an effective “duration”, for a street cam, of around 70 degrees, or roughly one fifth of a revolution. An experienced engine builder will have a more accurate guesstimate, taking into account precise details of the cam lobe shape! THIS IS PRECISELY WHERE EXPERIENCE IS SO IMPORTANT, and theory is insufficient! Anyway, to continue – a V8 with 460 cubic inches at 5000 rpm (2500 intake strokes per minute) and 85% volumetric efficiency ingests about 1.18 cubic feet of air/fuel mixture per second. But our fictitious valve is only open for one fifth of a revolution per revolution, so that 1.18 cubic ft/s is equivalent to around 5.9 ft^3/s during the fictitious equivalent valve-open event. So let’s conserve mass: volumetric flow rate equals cross-sectional area times mean (integrated) velocity. A typical moderate oval-port head will have, as you point out, about 3.5 square-inch port cross sectional area. So we get 242 ft/s. That happens, by pure serendipity, to coincide with McFarland’s advice; meaning, if I want a torque peak at 5000 rpm, and IF my guesstimate for effective equivalent valve opening duration is any good, then the 3.5 square-inch ports are just right. But in all honesty, my guesstimate is too crude to be of direct practical use for head selection. But the real lesson here is that the main limitation really isn’t dealing with too fast a flow speed through the intake port; it’s about keeping for flow velocity high enough. Low flow velocity leads to poor intake conditions in the combustion chamber, and poor VE. At low rpm, those heads give a really low intake velocity – at some point, too low of a velocity. But how low is too low? Ah, here again we need experience! Also keep in mind that port shapes are different! A longer, more contorted port might have a higher port volume than a shorter, straighter port with a larger cross-sectional area. Port volume data, by itself, is not very useful. And unless you cut up a cylinder head on a band saw into sections, you really won’t know the internal port shape or the area at the narrowest point inside the port. Big block heads have larger port volumes that small blocks, only in part because they have larger port cross sections; they have much longer runner lengths, too.

I'll take the risk, and suggest an actual course of action: definitely get the RX-7! Why? Because you know already know RX-7's well, and it's always best to go with what you know. You already know what parts work, what parts are ill-matched, what the car's weaknesses are, what's best left alone, what needs total rework. I also think that Z's look better than RX-7's, but what matters more is which car you're more familiar with - because skills and experience (and NOT time or money!) will be the main factors that determine whether the car gets built, and how good it will turn out.

Grumpy – Thanks for the advice. One issue that comes to mind is the tradeoff between better-flowing heads and a smaller cam, vs. worse-flowing heads and a larger cam – assuming comparable port volume and flow path geometry in all cases. The advantage of the former, in my view, is the more benign valvetrain dynamics. Also, as the instantaneous valve curtain area starts to approach the size of the port cross sectional area, flow rate begins to really taper off, in which case the difference between static (flow bench) and unsteady (in the actual engine) flow data gets even larger than “usual”. So, what I’m really asking is, why the recommendation for Brodix heads, which have among the worst flow bench data on the market, and a relatively large cam (0.610” lift)? BTW, my comment about diesels was mostly figurative, to stretch the point that my preferences lean toward the absolute opposite of those who need or want an engine build that emphasizes high-rpm hp over a narrow rpm band. Kevin – I do value your contribution to the discussion, but would like to mention that elementary arithmetic isn’t always the best avenue of advice. There is a tremendous difference between theory and practice. Lack of a firm footing in the workshop – that is, the practical experience – does not imply naivete in basic engine theory. Practical choices for what to buy are especially difficult in today’s market, where so many similarly-spec’d and similarly priced products are available, and every manufacturer makes great effort to tout his product as superior. And as Grumpy noted, amateur software simulations and paperback textbook rules-of-thumb have their limitations. Two-phase viscous compressible flow with unsteady boundary conditions and curvature that makes the quasi-one-dimensional assumption inappropriate is, I think, a complex enough problem that even good familiarity with the theory is of minor help with making practical choices. So, I am looking to duplicate, as far as feasible, an existing engine build. Those peanut port heads, by the way, are history - they got cracked during the process of pressing in new valve guides.

Grumpy, I was wondering if you could please post an analogous set of specs for a 454. You've posted links to 454 builds in the past, but most were either 496 strokers, boat engines, had custom-ported stock-casting heads, or some combination of the above. There's also a list of "big block combos that work" over on Chevytalk, but those are almost all race motors. For instance, last year you (if I recall correctly) posted this: Displacement: 468ci Compression: 10.0:1 Static Comp.Ratio Heads: World Products Merlin Iron, Oval Port Intakes Intake Manifold: Weiand Team G Carb: Demon 850cfm Cam: Crane PN# 139651 Hydraulic Roller Advertised Duration: 306/318 In/Ex Duration @ .050": 244/256 In/Ex Lift: .632"/.632" In/Ex Lobe Separation Angle: 114 That's something to think about, but way too high-rpm-oriented for me. Right now I'm looking at the following: * stock 454 2-bolt block, bored 0.030", decked 0.022" * stock 6.135" rods, resized, with ARP bolts * Speed-Pro hypereutectic (cast) pistons, 0.090" dome * stock 4.000" cast crank With the machining to-date, with this combo the pistons stick out 0.001" above the deck. Compression ratio with 119cc chambers and 0.040" head gasket comes out to around 9.12:1. The rotating/reciprocating assembly has been balanced with my damper, flywheel, and clutch. So that set of components is now fixed. I also have a Performer RPM (oval) intake, 4160-style 750 cfm carb, and an exhaust tract based on slightly modified Hooker block-hugger headers. What I need to buy is: cam, lifters and the rest of the valvetrain, and the heads. I've run numerous simulations on Desktop Dyno, but I am profoundly disappointed with the program's assumptions and results. So I would rather rely on the advice of experience. My current cam choice is something like the following (custom grind, hydraulic roller): 0.557"/0.557" lift, 227 deg exhaust, 221 deg intake (at 0.050"). The purpose of the roller cam is to avoid lobe-wiping - that's what killed my original engine 3 years ago. I would certainly be open to the idea of a mechanical roller, if I could find a cam small enough for my application. And that is, diesel-like torque below 3000 rpm and a never-exceed rpm of around 5200 (I hate high rpm!). Yes, I'm aware that this is too conservative in terms of piston speed limitations, but I drive like an old farmer. Thanks.

I’ve had similar frustrations with Desktop Dyno.... It begins its calculation only at 2000 rpm, and for various reasons is pretty unreliable at or below 3000 rpm. So if you’re more interested in a high-torque street motor than a race motor, the entire DD simulation is questionable for your rpm range of interest. It accounts reasonably well (perhaps) for changes in cam duration, but not cam lift. As in, increase the cam lift enormously (say, add 0.100” to the lift!), and my 454 picks up maybe 5 hp at the peak! It implies that going to a larger carb, larger diameter headers, etc., will raise the entire torque curve, instead of mostly just helping at the top end – an implication that’s completely false. This is because DD assumes “perfect atomization in the carb”, for example – regardless of jet size, venturi size, vacuum signal, etc. It is remarkably insensitive to changes in cylinder head flow rates, especially to increasing the flow rate. Try doubling the flow rate of your existing heads – the torque curve should change greatly, right? This isn’t a physical scenario, but it’s useful for sanity check of the code. Well, on my engine combo, DD shows almost no improvement. The cam calculation based on duration at 0.050” is inaccurate. DD really needs duration numbers as-advertised (at 0.006” lift). If I built my engine according to Desktop Dyno recommendations, I’d end up with a large-duration but low-lift solid-roller cam, a pair of 1200 cfm Dominators on a tunnel-ram manifold, open headers, a compression ratio of 16:1 and small-port heads with tiny valves. It would spin up to 11,000 rpm, still make great torque at 2000 rpm and run on 87-octane pump gas. Yeah.

I'm in a similar situation. My block (Mark IV 454) is drilled only for the offset-mount starter. I have the 168-tooth flywheel, and a Lakewood bellhousing. The stock full-size GM starter fits the flywheel and bellhousing, and even clears the headers, but the spacing between the headers and starter is awfully close (heat soak problems, etc.). I was considering an aftermarket starter with adjustable mounting block, so that the starter housing could be rotated to find the optimum position. The extra torque of an aftermarket starter would also be welcome. The Powermaster 9526 ($220 at Summit) fits, but it's doesn't have an adjustable mount, and looks pretty wimpy. Summit's house brand, "Nippondenso style", PN 820323-OS, is similarly priced, and apparently has more torque. Best (and most expensive) is a CSI starter, 100PSBP. (all of this stuff is on page 98 of the Nov-Dec 2003 Summit catelog). Anyway, I have two questions: (1) roughly speaking, how much torque would a 9.5:1 CR 454 need on a mild day; that is, is ~200 ft-lb enough? (2) do folks have any particular input, pro or con, about the Summit house-brand starters?

I saw a 2005 Mustang concept car at the D.C. Auto Show last week. While hardly a beauty, it looks considerably better in "real life" than in photographs, especially from the side view. My main gripe is about the front end; the forward rake can't be good for airflow over the hood. To me, it's a definite improvement aesthetically over the current model. One advantage is fewer fake scoops, vents and bulges. But no doubt there will be bad news about the weight. "Retro" designs are invariably heavy, and this one promised to turn out like the Thunderbird. By the way, the new GT40 is stretched by about 12" over the original. Apparently the sort of customer envisioned for this car would not have been able to fit in the original.

About a half dozen members on this site own or have owned a big block Chevy-powered Z. Most of those cars were specialized toward drag racing, though a few (such as mine) are multi-purpose. Whether or not the complexity of the swap is justified depends, as most things, on your skill level and your intended level of performance. Though estimates will vary, the concensus is that about 500 "streetable" hp is the threshold beyond which BBC becomes more cost-effective than SBC - assuming that the Z chassis has been suitably modified. Unfortunately (for me, at least) big blocks don't have much of a following on this site; such threads rarely develop beyond, "wow, this is cool...." If you do choose a big block, select the larger displacements - at least 454. Below 400 cubic inches, the small block makes more sense.

Doubtless this one came up before, but our trusty “search” function uncovered only peripheral mention of timing covers (for example, http://www.hybridz.org/phpBB2/viewtopic.php?t=24353&highlight=timing+cover). My 454 is getting a “mild” hydraulic roller (221/227 deg, 0.557 lift). Hence, issues of thrust buttons, Torrington bearings, all that good stuff - and how the timing cover holds everything in place. The question is, is a cast aluminum timing cover – a > $200 item – so greatly superior over a stamped-steel timing cover (a $20 item, if that) with a thrust button? Does the stamped steel timing cover need reinforcement (welding) – and if so, is that mostly a matter of avoiding spark scatter as the cam dances back and forth, pulling the distributor drive gear along with it? Granted, a 2-piece timing cover sounds very useful. And in the Chevy big block world, there are 2-piece and 3-piece timing covers (the latter a $270 item) – so what are the benefits of the 3-piece over a 2-piece? Trouble is, I’m only familiar with the concepts and terminology from catalogs and books – I’ve never seen a roller cam installed.

Do you have a hydraulic throwout bearing, or a conventional clutch fork? If it’s the latter, it’s likely that your clutch fork is bent. The hydraulics (master and slave cylinder) are trying to do their job, but with a bent clutch fork, there’s nothing to "push with" – so, the clutch never fully engages and shifting is difficult. This happened on my mother’s 1995 Saturn (I know, very different type of car, but the basic concept is similar). Shifting into first or reverse was very difficult, shifting into second was easier, third was yet easier, and so forth.

hmm... the machine shop's claim was that excessive porting of the heads might have "weakened them in critical places", so the formation of cracks upon pressing in new valve guides was a problem waiting to happen. Grumpy, what is the "correct" procedure to install new valve guides - I mean, what methods should the machine shop have used? Your offer of a donation is very generous - thanks! But as you point out, the lure of aftermarket aluminum heads is too great to forego. With that in mind, what is your impression of the new AFR oval-port heads? As far as I know, they have not yet been featured in the automotive press.

Bad news! One of the heads is cracked. I took them to the machine shop for evaluation, decking, new valve seats and guides. First, the shop pressure tested the heads at 35 psi (probably more than necessary, but evidently that's their standard procedure). Then they installed new seats and guides. After pressing in the new guides, they repeated the pressure test. And this time one of the heads failed - in two valve bowls. So after spending $600 on machine work, one of the heads is worthless. And if I can't find another 236-head, the set is worthless! But even if I do find another head (so to speak), 40 hours of port work (I work slowly) is ruined. On the bright side, AFR just came out with oval-port heads for big blocks. Their rectangular-port heads were out for around a year now, but those are too large for my application (305 cc ports and up). Their oval port heads are 265 cc. But aluminum BBC heads... $2000... ouch! So what do you think, guys - should I force the machine shop to eat their $600, since technically the heads were OK until they messed with them?

There are around 6 BBC-powered Z's on this list. The ex-pink drag car mentioned earlier belonged to Ron Jones; I believe that he recently sold it. Another example was a blown, deep blue-colored drag car, which (as far as I recall) has been converted to a small block after changing owners. "Ratsun's" Z is probably the most storied big-block car on this site. And then there's my 280Z, with a 454 from a '78 Suburban. It's probably the only non-drag car with a big block, and the only one with a stock rear suspension. A somewhat out-of-date but lengthy description is available on pparaska's site. The car briefly ran in the summer of 2000, but has been off the street ever since, with engine trouble. This fall/winter I'm finally getting around to rebuilding the engine.

Welding the roll bar to the map light area of the "roof" (or more precisely, the sheet metal enclosure that houses the hatch hinge mounting points) is an excellent idea. The roll bar in my 280Z's cage is welded to this sheet metal enclosure in two places, one just behind the driver's head, and the other behind the passenger's head. It's also a good idea to weld the roll bar to the points (passenger side and driver side) on the unibody sheet metal near the front bottom corner of the quarter-windows. That turns the roll bar into a sort of B-pillar. This approach is useful, in my opinion, regardless of whether you go with a full cage or limit yourself to a roll bar and strut tower braces. In either case, the theme is: tie the tubing into the unibody wherever possible.

I removed the crash beams in the doors of my '78 280Z; the car has a roll cage with X-bars, so gutting the doors to save weight felt reasonable. Grinding/cutting/chiseling the crash beams and the surrounding metal took considerable effort. When I was done, I took the pile of metal chips and shavings from both doors, put everything in a bag, and weighed the bag. Result (drumroll, please!) - 12 lbs. 12 lousy pounds, for all that work. Two lessons here, I suppose: (1) removal of metal in "non critical" spots will indeed save weight, but (2) the weight savings is very minor. Real weight savings requires a combination of aggressive cutting - and the wisdom to know where to cut!

I'm also in the Dayton area. Or rather, in the direction toward Dayton, off of exit 50 on I-71

Back when my roll cage was being buit, I was warned about the drawbacks of chromoly. Even if the welding is done properly, with heat-relieving the welds etc., the material is still brittle. DOM mild steel was the suggested material for all but the most weight-conscious applications, because it will deform plastically before failure. But this is a long-running debate. Incidentally, I heard that 0.120" DOM tubing is typically around 0.118"-0.119" from most suppliers. That sounds like a silly difference, except that NHRA inspectors have been known to reject nominally 0.120" cages when the main-hoop wall thickness was anything less than 0.120", even by one thousandth! So, many cage builders will use the next larger size, 0.134", just to be safe.

If you have a friend/relative who’s savvy with a particular brand of engine, that is a tremendously important consideration regarding which brand to use for the swap! Keep in mind that figuring out the swap itself – the engine mounts, the exhaust routing, the transmission crossmember, the driveshaft, the throttle linkage, and so forth – is only half the battle. The other half is the care and feeding of your engine. In my case, I had some one help me with the swap. The swap-related issues have been sorted out, the car was driven on the road, but then it’s been sitting for years because I’m stuck with the engine rebuild. Hey, if I had a helpful neighbor who really knew flathead Packard V8’s, I’d do a Packard swap. The point is, your experience and familiarity with a particular engine brand is enormously more important than any intrinsic advantages that a particular engine brand or family might have. Go with what you know. There may be superior engine choices, but what matters is how well YOUR engine runs, not how well it runs in the magazines.

mas280, was your brake-bleeding problem solved by installing the ZX components? By the way, what’s the procedure for bleeding the master cylinder itself? I had a similar problem, though somewhat more benign: after reconnecting the brake lines on all 4 corners (they were disconnected for some suspension work), I studiously bled the 4 brake slave cylinders. Now the pedal functions passably, but it’s still softer than I remember it being before the “repair”, and upon first pressing the brake pedal there’s a guttural wail (air rushing by/compressing) coming from the area of the master cylinder. Throughout the car, most of the brake lines are new, and the proportioning block has been relocated. Calipers and slave cylinders are all stock.

If a refined GT is a priority, it’s hard to argue against a later-model non-hybrid. While remarkable performance gains can be obtained from swaps, especially big-engine-into-small-car swaps, even maintaining the stock level of refinement is a challenge, let alone improving it to the point of being comparable to modern vehicles. As in most situations, probably the best approach is two cars – the GT cruiser and the hybrid bruiser. Personally I decided that comfort, reliability and economy are too difficult to combine with serious performance. The result of that is driving sub-$3000 boring econoboxes on a daily basis, enjoying the >30 mpg and the cheap insurance, while leaving the performance aspirations to the Z.

Mike, Glad to hear that you have reached the decision to keep your Z's! Maybe now I'll be inspired to resume the saga with mine....

I think that much of the confusion about the structural “benefit” of ladder-frames comes from old hot rodding lore, where it was perceived that ladder frames are a better base for roll cages, as compared to unibodies. Unibodies do need local reinforcement (steel pads welded to the sheet metal) to properly integrate with a roll cage. Whereas a traditional ladder frame can support roll cage tubing without such local reinforcement. And unibody sheet metal has lots of compound curves, which complicate the making of flush joints with tubing. But if the designer is mindful of these restrictions, there’s no question that a cage-unibody united structure will be stiffer than a ladder frame-cage, especially in torsion. Going with a separate ladder frame and “dropping” Z sheet metal over would become a reasonable alternative if you want to use a different car’s suspension, but don’t want to cut and weld the Z unibody for the new suspension pickup points.

The 3.36 is attractive for those folks who have high-torque engines but don't have overdrive transmissions. This scenario is relatively common, as the transmissions that can reliably handle prodigious amounts of torque tend to not have an overdrive top gear; for example, the Powerglide (automatic), Doug Nash and Muncie M-22 (manual). I would love to find a 3.36 - or even better, the mythical 3.15 - but despite years of regular searching in Los Angeles-area pick-a-part yards, I never did come across a 1979 2+2 ZX. Though somewhere I heard that aftermarket suppliers are selling 3.36's in the $1K range.

My car was one the road... briefly... about 3 years ago. Since it's now off the road, does that mean that I got renoobed?

I forgot to mention - that photograph is on page 50.