middleagedcrazy

Its amazing to me that I seem to be the only person that knows what an "estimate" is. Maybe I should blame the schools or pocket calculators or something. Estimates are approximate answers, and there are always exceptions when something is different than the estimate is based on. Lets look at Gollum's counter examples. The 98 Corvette actually agrees with my rule of thumb (10:1 CR -> 60 ft-lb/L). All the examples that don't agree are 4-valve DOHC motors. The Honda S2000 is interesting because if you look at dyno charts you see a two-step torque curve; its a pretty mild 60 ft-lb/L (or less) up to 5800RPM, where the VTEC shift the cam lobes and - shazam - 75 ft-lb/L. What might that tell us? Maybe that radical cam timing cam make some good numbers at high-RPM, but it really stinks at low-RPMs (otherwise why go to all the complexity of VTEC?) This might be really interesting if I could build a VTEC L6. The new Ford V8 is similarly interesting. It doesn't have a traditional nearly flat torque curve; it starts relatively low (60 ft-lbs/L) at 2000 RPM and climbs steadily to a peak of 78 ft-lbs/L around 4500RPM, and then tapers back into the low 60s by 7000RPM. Turns out the Coyote can independently control the timing of the intake and exhaust cams; cool - you can put an aggressive grind on the cam and then advance/retard them to fine tune the valve timing. I couldn't find any dyno charts for a stock 300ZX; my rule of thumb would have been around 62 or 63 ft-lb/L, Gollum says max torque is 66 ft-lb/L. I'll hazard a guess that the torque curve has a bump in the mid-range where the cam/intakes are tuned to make max power. Works fine, but only at that rev-range. LS2 Chevy's are interesting because they are push-rod V8s with no VTEC-like technology and make close to 80 ft-lb/L with an 11:1 CR (thats a crate motor, the ones that Chevy put in Corvettes were only in the 60-70 ft-lb/L range). I don't know how they do it - maybe the engine designer made a deal with the devil or something. I'm guessing an aggressive cam and they cover up the peakiness with EFI and sheer displacement. Bottom line - my "model' isn't perfect, but its not all that far off, and its really pretty good for old-timey 2V fixed-cam-timing engines. Lets take a look the example Tony gave of an L28 with ported heads, low compression, SUs and 182 whp@6500. Lets call that 225 crank-hp, which works out to 182 crank ft-lb and 65 ft-lb/L, admittedly more than I would have guessed.Tony said all this motor would need is triples to hit 200 whp. I agree, but at what RPM? I don't think bigger carbs are going to add torque; they're just going to let this engine rev higher. (200/182) x 6500 = 7150RPM. I still don't see me being too far off about needing 7500RPM to hit the 200 whp mark. As to the SYSTEM jab, I never said that all you needed was displacement and compression. The whole point of coming up with the model was to just find a starting point - an approximate displacement and CR and redline - to help make decisions based on real world experience. If you think trying to use math to understand how an engine works is a bunch of egg-head-theory stuff, its your loss. I'm happy to listen and learn, but if you want me to kiss your ring or something first, then don't bother to share all your valuable knowledge. Don

Not to belabor this, but you took a general statement for a production engine (that is: an engine built by a manufacturer for every day use, with a CR around 9-10:1) and tried to apply it to a what is essentially a racing engine. The rest of my text says CR is the major factor affecting torque/liter. 13.6 should give about 12-15% higher torque/L - and all the little tweaks done to a high-performance engine adds a bit more. Call it 70 ft-lb/l, and your 3L should make about 210 ft-lb torque, and 7200 RPM works out to 288 hp (thats at the crank). I don't understand how you got 275 ft-lb at any RPM, but I'd like to learn. What I'm thinking is that a high-overlap cam has a sweet-spot in the RPM range where it combines with the manifold length to give a Veff > 1 and so generates a bump in the torque curve. If you look at Z-ya's vintage-racer-build-up (awesome thread by the way) you see the same thing - the engine is a 2.85L making a fairly flat 72 ft-lb/l, except for a big bump in torque around 6000RPM. I'll be the first to agree, my approximation doesn't work all that well for borderline race engines. I still think the approximation is pretty good for mild-street engines (up to around 11:1). I guess you can drive a 13:1 motor on the street. Back in the muscle car days I knew lots of guys who threw a big cam and bigger Holley on a SBC and they made awesome power - from 5000 to 5500 RPM... Keeping them running at a red light required setting the idle somewhere around 1500RPM and frequent throttle-blips to "clean out" the carb. Not what I want to spend $3000 to achieve. With FI and a good fuel map you can no doubt straighten that all out - I'm still open to FI. I should probably chase down someone in the area with Megasquirt and ITBs and see how they run, I could see being really happy with such a setup. Tony - 205 (crank HP) and 9:1 compression (I'll guess 57 ft-lb/l) works out to 9450 RPM - bet it sounded pretty! Finally, if my estimates are low, the worst thing that happens is I build a motor that makes more power than I had hoped for. That ain't so bad... Thanks again for the info - its what I asked for - I'll let you know what I learn along the way. Don

Tony, I think I said it several times; I made some estimates to come up with a starting point, but those estimates are based on a fair amount of real observation. Once a cylinder of a given volume has compressed a given mass of air/fuel mix and the spark plug fires, how the air/fuel got there doesn't matter much. Cylinder pressure is going to go up and force the piston down, transmitting power to the wheels in the process. How much of the heat energy gets turned into work depends largely on the expansion ratio of the cylinder (which also happens to be the CR). I've been reading car magazines for 30 years and looking at dyno sheets posted on the web a few more - 60 ft-lb/liter (1 ft-lb per inch) is a pretty good guess for a production auto engine. Well tuned engines - where fuel delivery and ignition is just right - seem to make a tiny bit (1-2%) more. Increasing the CR gives relatively small increases; I've looked at theoretical efficiency charts in my old thermodynamics book that predict about 4% difference between 10:1 and 11:1, or about 63 ft-lb/liter for 11:1. Project out to 13:1 and its about 10% more than 10:1; call it 66 ftl-lb/l, a well tuned 13:1 motor might make it to 68. The few 13:1 engines I can find data for fit this prediction fairly well: L88 chevy: 7L x 67 ft-lb/L = 469 ft-lb. At 6000 RPM, that works out to 535 hp, which agrees reasonably well with published info (for whatever its worth). I can see how a poorly designed or tuned engine can make less torque/liter than this, but not really more. A tuned intake can bump up the torque at a particular RPM, but its at most a few percent. The one thing that screws up this simple view of the world is VTEC engines with multi-path intakes. By playing games with valve-timing and plenum-length/volume they can put a big torque/liter bump at mid-range RPMs - much more than than my model predicts - but the torque at the power peak is still pretty close to that predicted here. Lets look at your 11:1 example. I'll assume its well tuned and guess a little on the high-side: 64 ft-lb/L; thats a max torque of 128 ft-lb, and the algebra tells us (205/128) * 5252 = 8400 RPM. How close did I get? I understand these are estimates, and they ignore a lot. Just throwing in forged pistons won't get an L28 to 7500 RPM; there has to be headwork, the right cam and carbs/FI else it doesn't matter how long you hold the gas down - you're not going to top 6K. If these parts aren't right the engine won't make anything close to my estimates. These are the details I'm hoping to work out. Don

I am the original poster, and I didn't just pull numbers out of the air; I actually did a lot of thinking about what I wanted and my options for how to get there. I wanted to avoid getting into debates about torque-vs-hp and whether dynamic compression is real and such other rants, so I went light on the theory. I'll try to explain my thinking without killing any sacred cows... Before I go too far, let me say this is all approximation; its back-of-the-envelope numbers to get to a starting point. The 200 whp goal is based on having noticeably more power than the RSX-S I drive to work every morning. I'd be happy to get more, but I don't want to sacrifice too much drivability by going too big on cams or carbs. Max crank torque from an NA engine depends mostly on displacement and a little on CR. In a well behaved engine, torque is pretty flat from around 3000RPM up to the power peak where torque starts to drop. A 10:1 CR will generate about 60 ft-lbs/liter (my apologies to the metric purists for those units). Increasing CR by 1 point will increase torque by about 4% (to 63 ft-lb/l) , and dropping it 1 point decreases torque about 4% (to 57 ft-lb/l). You shouldn't assume that trend is linear - as CR goes up it yields smaller and smaller increases in max torque. Again, this is an approximation, but its close for lots of real world engines. Note that the cam doesn't have much impact on max-torque; only on the shape of the torque curve. A big cam can move the point where torque drops off to a higher RPM, and so makes more power. I want to run on pump gas, so a CR around 10 is as high as I want to go. 60 ft-lb/liter * 2.8 liter = 168 ft-lbs . The relation between torque and hp is well known, HP = torque * RPM / 5252. Nothing magic here, just physics. Use some algebra: (250 hp / 168 ft-lb) * 5252 = 7800 RPM! This is a ridiculous redline, but its what you need to turn to actually make 250 crank hp on a 10:1 L28. I don't consider that realistic, so if I want to get close, I need more displacement and maybe a little more compresson. Clearly a stroker build would meet my needs. 60 ft-lb/l * 3.1 liter = 186 ft-lb. (250/186) * 5252 = 7000 RPM. Bump up the compression on a stroker to 11:1 and you can get 250 crank hp at 6500 RPM. I'd prefer to split the difference and shoot for 2.9 liters displacement, a 10.3 CR and a redline somewhere around 7-7.5K RPM. What I didn't say was that a 3.1 can't rev; what I was trying to say was at this hp level (250 crank) the stroker is flowing all the air it needs to around 6500RPM. Reving higher can make even more power, but also requires more flow capability. I suspect the big reason most strokers don't rev past 6500 RPM has nothing to do with long crank throws or rod/stroke ratios; its that they're running 40mm triples and small cams that can't flow enough air to feed 3.1 liters past 6500 rpm. No matter what combination of displacement and RPM you take to get to that hp level, you need a head that can flow the air needed for that power level. Anyways, lots of things to think about - thanks again for the real world feedback. Don

I don't plan to cruise at 7500RPM; I just want to be able to wind 2nd and 3rd a bit during spirited driving or the occasional track-day and not have to worry about breaking something. I'm not obsessed with 200HP - its just a stake-in-the-ground to help focus my thinking. If I put a motor together that makes 210 its not like I'm going to back off the timing to hit that "magic" number. On the other hand I don't want to miss that target and spend lots of money for no gain. I've seen lots of stories on the web where someone throws a big cam and triples and a header on a stock L28 and ends up with 150 whp and "a cool lopity idle"; that's what I want to avoid doing... I do believe that all things being equal, EFI is a better technical solution and I would be just as happy with ITBs as carbs. I'm a little put-off by the complexity: I don't think there is a turnkey setup with the throttle-bodies, control-unit, sensors, pumps, cable etc. needed to put FI on an originally carbureted car. Chasing down parts from different sources and getting them to work together can be a hassle I'd prefer to avoid; that and having triples being something to cross off the bucket list... If there is a turnkey setup out there I might well change my mind. Thanks for all the info. Don

Tony, I'm sure you can get 200 whp from a stock displacement L28, but I'm thinking it takes an 11:1 CR and a big cam. I'm not quite sure how to read your example; if you're saying someone pulled 256 HP at 6500 RPM from a stock displacement L28 I'm impressed - that works out to a LOT of torque and a ridiculous CR (13:1?). If they pulled the 256 at some higher RPM (after the AFR went rich) then its still impressive but a little more believable. The "just starts to pull hard at 6K" is what I'd like to avoid. I'd like to pull hard from roughly 4K to 7K, and I'm willing to trade some absolute top-end power for mid-range. I know I could get the whp easily with a 3.1 build, but they all seem to max out around 6500 RPM or less (a lot of them seem to run 40mm triples which seems a little small for 3.1L, maybe they just need bigger carbs?) As I said, I have this totally irrational perception that its not a sports car if you can't wind it to 7000. I know its not going to be the fastest car on the road; me being happy with the way it drives is more important than a bit more acceleration I'll only notice if I drive around with a stop-watch. I thought the L28 crank, rods and cam were reliable to 7-7.5K - the expensive part being forged pistons? That seems like at most an extra $600 - a little less if comparing to cast pistons. Doesn't seem bad for peace of mind and not a lot in the grand scheme of an engine build. Or am I missing some other money-black-hole? John, if there are L24s out there making an easy 200 whp I'm reading the wrong books and web pages... that seems like race-car territory (100+ bhp/l). Can you outline the build? You're right about the weight of an RSX - its closer to the weight of a 280Z, the 260 is lighter and that helps a bit; I could lower my hp goal by about 15% which would be a lot easier to reach - but once the engine is done I'll probably be thinking R200 and 5 speed, which will erase most of the weight advantage. I think I'll stick with my 200 whp target. I did briefly consider a frankenmotor: L24 crank and rods in an L28 block with 88.5mm bores (I think it works out 2750 cc) - good for style points - but given my goals I'd rather have extra displacement. Don

I've got a mostly stock 260Z that is a typical "sunny day" car - but lately its been upstaged by my new daily driver - an RSX type S. Fortunately I just dropped my daughter off at college for her last year, so come spring I'll finally have a few bucks to sink into the Z. My goals are more subjective than logical: Feel faster than my RSX. Run triples. I know I'll probably get tired of tuning carbs and swap to ITBs or something, but I've lusted after Webers (or Mikunis, Dels, etc) for a long time. tHave a usable redline north of 7K RPM. I know its more efficient to make power at lower RPMs but I want to hear an exhaust note that makes me feel like I'm Paul Newman at Limerock. A smooth idle and linear throttle response. Run on 93 octane. The RSX weighs about the same as an S30 and according to the internet puts down around 165-170 whp, so I'm thinking I want about 200-210 whp, or approx. 250 crank hp. I've been searching the internet and doing back-of-the-envelope estimates to figure out how to do that. What I'm leaning towards is (roughly) 88.5mm bore L28, stock L28 crank and rods, high-silicon forged flattops, 10:1 CR, quench head, 45mm triples, mild cam (aprox .490 lift, 270 duration). Mild cleanup port on the head - unshroud the valves - but don't go too big on the port to keep velocity up. Basically a mild stroker build without the long crank (to give me my 7500 RPM redline without radical porting). Things I couldn't find good answers for: Best way to get a 10ish CR with a quench head? Can an MN47 be opened up to 46cc combustion chambers? I'm nervous about the shim-n-shave mode for the P79; all those shims stacked under the cam-towers seems a little scary at 7500 RPM... Maybe go with the Kameari idler gears and adjustable cam gear? Is there any benefit to using L24 rods with the L28 crank? I've plotted the piston velocity curves with both rod-lengths and the extra 3mm just doesn't seem to amount to much? Keeping the shorter rods might actually help by creating more intake "vacuum" early in the intake stroke to squeeze a little more flow out of the carbs (45mm seems a little on the small side for a 250 hp motor). Any real world experience here? Undoubtedly lots of things i haven't thought of yet... I'm trying to settle on a block and head before I get into more details. Opinions, ideas, experience welcome. Don