Date: Mon, 13 Feb 1995 10:33:40 -0500 (EST) From: "Kenneth D. Rolt" Subject: summer engine project To: JeffC55@aol.com Cc: Virtual Vairs Jeff, I reposted this to the entire group. There are many people who would have an interest in this project, and a bunch that can give you some advice in both theory and practice. - Ken Rolt ---------------------------------------------- On Sun, 12 Feb 1995 JeffC55@aol.com wrote: > Ok guys,possibly this summer or next I need a new engine for my 68 conv . > Goal is to have a verrrry fast, somewhat different, streetable,corvair based > power plant. > Here's the idea.140 based/dual turbo/intercooled/fuel injected/with 1or 2 > verticle cooling fans.All components too be located for minimum tube distance > / maximum response. > I plan to mount 2 garrett T-2 turbos about where heat ducts exit.just > behind axel (turbo has internal waste gate/very small). Turbo axis is to be > about= to piston center line.1or2 intercoolers mounted flat with blowers on > top or mounted verticle with blowers sucking through (whichever is > better). this will be a powerful engine, and so it is crucial you control heat production and elimination. the exhaust manifolds (cast iron will last much longer on a turbo car than headers BTW) will get quite hot under boost and they are very close to the head(s) and va;lve cover(s)... you need a way to control the amount of heat getting to these. the heat wants to stay in the exhaust, so it can be used as any energy supply to the turbo (via hot exahust gas). if you aren't carful, you will melt the head. i strongly suggest you call or write the owner of the World's Fastest Corvair, which is turbocharged. I think some of our California members in the group know Tom Keosababian (spelling?) and can direct you to a phone # or address. the engine will have to be built to be overkill bulletproof. forged pistons, high qulaity valve guides, alloy valves, Corsa or Spyder 1964-1966 forged nitrided crankshaft, low friction rocker arms, four spider gears in the differential, etc.. > questions are > 1) is this enough "drop" to get the oil to drain out of the turbos or > will i have to use scavging oil pump. I don't know. you want to make sure that the turbo bearing(s) is *never* oil starved, so I think an additional reservior and separate cooler post-turbo would be useful. I don't know if gravity feed is enough. Tom (or someone else) may be able to answer this better. > 2)should i use modified log maniflods or headers and will the "volume" and > exhaust pulses be sufficent & conducive to maximize exhaust impeller speed > for a wide speed range. it's supposed to be a street car, therefore it should be reliable. I would use either stock logs or use a casting which resembles the stock item. i would not use headers, but others have disagreed with me on the use of headers on a street car. > I dont plan on using a balance tube. the balance tube helps make the engine think it is connected in more ways than just the crankshaft, by giving it a feedback connection on the intake side. you *might* still want some form of balance tube; experimentation will tell you yes or no. > 3) heads / pistons will be modified to reduce compression to 8-8.5to1 and > still keep squish chamber. Can you recomend pistons to use & perfered > method of acheiving c/r. Bill Fisher's book has a very good section on this. It's dated but it makes great sense, and the physics haven't changed since it was written. > 4)what fuel injection can handle boost conditions & where/how would you > mount the injectors? don't know what systems; you should check to see what turbo'd engines have 6 cylinders and have a c.i.d. in the 2.6 to 2.8 liter range. injector mounting wants to be near the intake ports, and aimed at the port. The nuances of exactly where they point (do you point them at the valve neck, or the valve opening?) might be an rpm-dependent thing, so you would hacve to read some books on f.i. systems. the fact that there is boost just increases the velocity (and total mass or volume of air), so you will need more fuel with more boost, but the pointing of the injectors is subtle and may not be that significant except at very high speeds. go read and see what the practicioners say. 5) will this work ??? I have no desire to build a fancy > street "dog" ! this will be a heck of a lot of work. i think building a very thorough f.i. system would be a fine feat by itself. the turbo added might pose too many small nagging details... i.e. you are trying to get two complex add-ons working correctly at the same time. doing a single one *might* be hard enough. how good are you as an engineering tinkerer? you may need to do a substantial amount of tinkering to get everything working right. > any input/ suggestions welcomed > thanks JeffC55 ----------------------------------------------------------------------- Kenneth D. Rolt kdrolt@sanders.com Lockheed Sanders / ASW MAN 6-2100 tel: 603 / 645-5750 P.O. Box 868 Nashua, NH 03061-0868 U.S.A. fax: 603 / 645-5731 ----------------------------------------------------------------------- From: Rad Davis Subject: Re: summer engine project To: corvair@shadow.mitre.org Date: Mon, 13 Feb 1995 18:29:43 -0500 (EST) Kenneth D. Rolt wrote: > > Jeff, I reposted this to the entire group. There are many people who > would have an interest in this project, and a bunch that can give you some > advice in both theory and practice. - Ken Rolt > ---------------------------------------------- > > On Sun, 12 Feb 1995 JeffC55@aol.com wrote: > > > Ok guys,possibly this summer or next I need a new engine for my 68 conv . > > Goal is to have a verrrry fast, somewhat different, streetable,corvair based > > power plant. > > Here's the idea.140 based/dual turbo/intercooled/fuel injected/with 1or 2 > > verticle cooling fans. Number of fans doesn't matter a rat's ass. What does matter is the CFM volume of air you can put over the engine and the evenness of distribution. You will see max heat generation at the highest RPM at which you can get max mixture flow through the engine. On a turbo engine this is closer to the HP peak than the torque peak. According to Chevrolet you need 18 CFM/HP. (SAE 140C "The Chevrolet Corvair). This works out to 80 hp and 1800 cfm on the '60, which represents a 20% safety factor. Your problem is that even if you manage to put 4000 cfm through a fan system, you're going to have a fun time trying to stuff it through the fins on the engine. This means that you have a steady state power limitation equal to the maximum airflow you can force over the engine. I hope you like your power in short bursts... > > All components too be located for minimum tube distance > > / maximum response. > > I plan to mount 2 garrett T-2 turbos about where heat ducts exit.just > > behind axel (turbo has internal waste gate/very small). Turbo axis is to be > > about= to piston center line.1or2 intercoolers mounted flat with blowers on > > top or mounted verticle with blowers sucking through (whichever is > > better). Intercooler mounting probably should be on top of the engine over cyl # 5 and 6 to minimize tubing runs in the intake tract. You'll have a plumbing problem if you try to use air-air intercoolers--there's got to be a lot of really cool air going into the engine or you'll have expensive problems. The intercoolers preheat the air very well. I therefore suggest that you do it like the high performance WWII fighters and the racing Porsches and use air-water intercooling. Mount a small circulating pump somewhere and put the water radiator up front. For that matter, use water cooled turbos too, and plumb them into the same thing. If the turbos are water cooled, you don't have to worry about waste oil collected in the center section coking up, which should relieve your worries about mounting the turbos so low on the engine. > this will be a powerful engine, and so it is crucial you control > heat production and elimination. the exhaust manifolds (cast iron > will last much longer on a turbo car than headers BTW) will get > quite hot under boost and they are very close to the head(s) and > va;lve cover(s)... you need a way to control the amount of heat > getting to these. the heat wants to stay in the exhaust, so it can > be used as any energy supply to the turbo (via hot exahust gas). > if you aren't carful, you will melt the head. Heh. You just shot your own argument out of the water, Ken. The most effective single thing you can do to reduce heat load on your cylinder heads is to go to wrapped headers. If you use thick-wall stainless tubing you should be able to avoid burn-through. Don't worry about equal length or other silliness. Just get the tubes away from the heads and valve covers and into the turbos as soon as possible. Heat shields for the turbos and wrapping the turbo exhaust pipes is also a good idea to reduce heat problems around the valve covers, which are a major oil-cooling component. I will warn you in advance that even a headered 140 with no turbo and a good cam can get into a destructive thermal cycle under sustained high-load conditions--my father experienced this in his air-conditioned 140'd rampside with the a/c running and a headwind. The a/c system, btw, does not dump hot air on the engine. The head temperature got up to over 500 indicated and the oil temp kept climbing, so he backed out of it. He is, incidentally running dual folded-fin coolers... > i strongly suggest you > call or write the owner of the World's Fastest Corvair, which is > turbocharged. I think some of our California members in the group > know Tom Keosababian (spelling?) and can direct you to a phone # > or address. Seconded. Another guy you need to talk to is Warren LeVeque. See the Communique-- I think he has an ad in there. Louis Pinard of Red River Corvairs in Louisiana is also somebody you should talk to. All these guys build strong engines that live. > the engine will have to be built to be overkill bulletproof. forged > pistons, high qulaity valve guides, alloy valves, Corsa or Spyder > 1964-1966 forged nitrided crankshaft, low friction rocker arms, > four spider gears in the differential, etc.. Little known (apparently...) fact #1: All Corvair Cranks are forged from the same steel alloy in the same forge. Only the stroke and surface hardening vary. I presume that they magnafluxed the cranks that were nitrided, but then maybe they didn't... Regardless, I recommend you do so to yours. Little known fact #2: Nitriding (or any other bearing surface treatment) doesn't make a crank one whit stronger. Not a bit. Indeed, nitriding theoretically could make the crank more brittle. What it *DOES* do is increase the surface hardness on the journals. This is useful if a) your engine has something abrasive in the oil in the oil or, b) you're getting metal to metal contact between the journal and its associated bearing. The harder journal surface ensures that the bearing gets chewed up, instead of the crank. Why, then, is nitriding so popular? Because the stock-block racers who popularized it (and its ancestor hard chrome plating) frequently did suffer from condition a (in the form of track dirt and disintegrating valve gear) and condition b (in the form of lubricant breakdown, insufficient bearing area, lubricant overheat, detonation, insufficient oil pressure, or a combination of these) and could ill afford to replace cranks before every Saturday night's or Sunday's racing. For them it was cheap insurance--you can't win if you don't finish. For a street duty car that won't be run at full throttle for 500 miles on a 90 degree day, it's probably a waste of money. Even if you're doing what you're talking about doing to this engine. What *ISN'T* a waste of money is to have the crank magnafluxed and shot-peened to relieve surface stresses and case-harden the throws. Corvair cranks are known to have a cracking problem, and this will do more to improve engine longevity than nitriding will. > > questions are > > 1) is this enough "drop" to get the oil to drain out of the turbos or > > will i have to use scavging oil pump. > > I don't know. you want to make sure that the turbo bearing(s) > is *never* oil starved, so I think an additional reservior > and separate cooler post-turbo would be useful. I don't > know if gravity feed is enough. Tom (or someone else) may be able > to answer this better. Um, Ken-- he wasn't asking that. He was asking about getting the "used" oil OUT of the turbos after it has cooled and lubricated them. I presume he knows that you have to feed the turbos straight from the oil pump, and a supply pressure of 50-70 psi doesn't hurt a bit, btw. I think that if you use a water-cooled turbo (see above), and you run a pan extention, deep pickup, and the stock quantity of oil (you'll need a longer dipstick) that you probably would be OK like this. With an oil-cooled turbo, you have to make sure that oil doesn't stay in the turbo very long, or it will turn to carbon--not good. I don't know if there's reliable enough drain to do without a pump or not. I suspect not. > > > 2)should i use modified log maniflods or headers and will the "volume" and > > exhaust pulses be sufficent & conducive to maximize exhaust impeller speed > > for a wide speed range. > > it's supposed to be a street car, therefore it should be reliable. > I would use either stock logs or use a casting which resembles > the stock item. i would not use headers, but others have disagreed > with me on the use of headers on a street car. None of the options will cause pulse damping or have excessive volume. Your problem is that *ALL* of the stock options have insufficient cross-sectional area for a turboed 140. The stock 140 plumbing without the muffler runs out of breath at about 5600 rpm with normal aspiration. With a turbo it's only going to get worse. Anything you can to do improve the exhaust breathing will only help. There is also the heating issue discussed above. I would strongly suggest that you consider going to angled exhaust ports and tubular headers. > > I dont plan on using a balance tube. > > the balance tube helps make the engine think it is connected in > more ways than just the crankshaft, by giving it a feedback connection > on the intake side. you *might* still want some form of balance tube; > experimentation will tell you yes or no. This all depends on the throttle-body setup you choose. If you use two throttle bodies, you'll either need a very precise throttle linkage (and throttle bodies) or you'll need a balance tube. The thing won't idle right without it. One throttle body, of course, acts as a balance tube itself. I suggest that you use the tube. It doesn't have to be any bigger than the stock one, and isn't much of a problem to plumb around. > > 3) heads / pistons will be modified to reduce compression to 8-8.5 to 1 and > > still keep squish chamber. Can you recomend pistons to use & perfered > > method of acheiving c/r. Heh. What are you going to use for fuel? Pure toluene? Not much else will let you run much boost with this setup. Are you aware that *STOCK* actual compression for the 140 is about 8.6:1? They quoted a number above nine to impress people. But the real value is almost a full count lower. You need the squish chamber, true. Which means that you want one gasket clearance between the squish part of the head and the piston. You're going to have to bathtub out the chamber and maybe mill the part of the piston under the chamber to drop compression into the 7-7.5:1 range if you want to run fuel of less than 105 research octane with anything resembling boost. Louis Pinard gave up on squish all together when he built Billy Bruce's turbo '63 drag coupe. He used a pair of 95 hp smog heads. Billy's turning high 12's... > Bill Fisher's book has a very good section on this. It's dated but > it makes great sense, and the physics haven't changed since it was > written. True, sort of. We know a bunch more about combustion inside the engine than we did in those days. > > 4)what fuel injection can handle boost conditions & where/how would you > > mount the injectors? > > don't know what systems; you should check to see what turbo'd > engines have 6 cylinders and have a c.i.d. in the 2.6 to 2.8 liter > range. injector mounting wants to be near the intake ports, and aimed > at the port. The nuances of exactly where they point (do you point > them at the valve neck, or the valve opening?) might be an > rpm-dependent thing, so you would have to read some books on f.i. > systems. the fact that there is boost just increases the velocity > (and total mass or volume of air), so you will need more fuel with > more boost, but the pointing of the injectors is subtle and may not > be that significant except at very high speeds. go read and see > what the practicioners say. The only possible stock match I know of is the turbo 280 Z/300 Z injection. It's based on Bosch L-jetronic or Motronic depending on the year, and there are several aftermarket tuners who tweak Z-cars well who you could contact for help. > 5) will this work ??? I have no desire to build a fancy > > street "dog" ! > > this will be a heck of a lot of work. i think building a very > thorough f.i. system would be a fine feat by itself. the turbo > added might pose too many small nagging details... i.e. you are > trying to get two complex add-ons working correctly at the same > time. doing a single one *might* be hard enough. how good are > you as an engineering tinkerer? you may need to do a substantial > amount of tinkering to get everything working right. There are problems you don't know about yet. You'd better build the strongest rearend you can manage, because a full-power upshift WILL tear all the teeth off the spider gears in a stock diff. The corvair diff was designed for a 160 lb-foot engine. The stock turbos were pretty rough on them--your setup will make expensive scrap. You probably should have somebody mig weld the crank gear/output stub to the rest of the crankshaft. It is easy to rip the keyway out of the output stub at about 260 lb-ft of torque. You'll also have to tap the stub for larger flywheel bolts. Considered the flywheel/clutch you'll need yet? Another thing you're going to have to have is programmed boost and load sensitive ignition, probably as part of the injection computer system. A little pressure retard capsule won't cut it. You'll also need some sort of integrated water-alcohol injection. > > any input/ suggestions welcomed > > thanks JeffC55 Well, I guess you got what you asked for! Hope it does you some good. But I warn you that a car with 180 reliable HP is a lot more fun than a car with 250 unreliable HP. Rad Davis