Interesting interview with GM chief engineer
Thread Starter
Moderator
Joined: Dec 1969
Posts: 10,745
From: Buffalo, New York
Interesting interview with GM chief engineer
From an interview in Car and Driver:
"A pushrod is basically a metal tube designed to transfer the reciprocating motion of a valve lifter, riding on the camshaft, to the valve.
I found the pushrod fascinating 20 years ago, back when I was a high-school car geek devouring car magazines. It was about the time the Japanese car invasion began and Detroit started to worry about loss of market share here. In the mid-'80s, Detroit stuck by the pushrod, and nearly every available engine used it. Domestic cars were for the most part, however, mediocre contraptions that allowed foreign automakers to make inroads into the U.S. market. Back then our 10Best Cars competition had five spots for domestics and five for imports. In 1983, the pickings were so slim that the Chevrolet Caprice Classic made the list.
When asked what made Japanese cars better, critics gave a variety of answers, but one stuck in my mind: While the Americans produced crude, underperforming pushrod engines, the Japanese were turning out sophisticated four-cylinder, single- and double-overhead-cam engines with twice the number of valves per cylinder.
The idea of locating the camshafts in the cylinder head was not new, but to a populace still unaware that a 1929 Duesenberg Model J, for example, had double overhead cams and four valves per cylinder, the overhead-cam engine sounded advanced, exotic.
The pushrod was the scapegoat for all that was wrong with Detroit. "No doubt about it. In the '80s, the pushrod was a hated component, a symbol of the uncompetitiveness of the domestic industry," says Sam Winegarden, GM's chief engineer for small-block V-8s.
I called Winegarden because I've always been fascinated by GM's decision to stick with the pushrod when it introduced a new small-block V-8 in 1996.
By the '90s, many domestic four- and six-cylinder pushrod engines had been scuttled in favor of overhead-cam designs. In 1995, Ford axed its venerable pushrod 16-valve V-8 for a V-8 that still had just 16 valves but now ran with a pair of overhead cams.
Keeping the pushrod seemed like yet another example of the General's hanging on to its glory years. But since then, I've come to see that, for a mass-produced V-8 engine, the pushrod layout has more advantages than disadvantages.
"Power is a function of airflow, pure and simple," says Winegarden. "If you're not going to fill the cylinder with valves, why have an overhead cam?"
Although you could design a V-8 engine that uses pushrods to operate four valves per cylinder, the layout lends itself to using just two. Four valves generally have greater total valve area than two, thus creating greater airflow and more power. In the case of the Chevrolet Corvette Z06 and its pushrod two-valves-per-cylinder 405-hp, 5.7-liter V-8, Winegarden says, "We've been able to meet the performance requirements by using more displacement. Two more valves per cylinder would get us another 10 percent in total valve area."
A modern engine is a dizzying array of compromises as designers strive to meet power, cost, reliability, and emissions demands. The Vette V-8 would ultimately have more power if it had four valves per cylinder, but it would lose ground in cost, complexity, and physical size.
"On the Vette, performance density is only beaten by some of the big Ferrari or heavily boosted engines," Winegarden continues. One of the huge advantages of the pushrod V-8 layout is its clever use of space. The camshaft and the pushrods lie in the unused area between the cylinder banks. A DOHC or SOHC V-8 is much wider and a little taller because the camshafts are on top of the cylinder heads.
Enthusiasts often talk about how much power an engine produces as a function of the combined volume of the cylinders (power per liter). But what's often overlooked is the power produced compared with the outside dimensions of a motor.
The small-block is a compact engine. I did some rough measurements of the Z06's V-8, the DOHC V-8 in the Porsche Cayenne S (335 hp, 4.5 liters), and the Northstar V-8 in the Cadillac SRX (320 hp, 4.6 liters). The small-block was about six inches narrower than the Porsche V-8, two inches shorter than the Caddy's V-8, and close in length.
That may not sound like much, but it gives designers more freedom to position the engine for good weight distribution and leaves ample room for suspension and steering components.
Don't forget about the weight advantages. The Vette's engine has one camshaft and a small drive chain that's connected to the crank. A DOHC V-8 has four camshafts and drive gears, beefy heads to hold the shafts, and two long chains. "Our V-8 is a simple and elegant design. It's pretty easy to put together, which helps the reliability, and costs," says Winegarden. He estimates a $400 saving over a DOHC motor. And, of course, with fewer parts comes less weight. (Winegarden says the aluminum small-block is 44 pounds lighter than the Northstar.)
GM also used this architecture for its truck engines. Spreading the tooling and design costs over a large number of vehicles reduced the per-unit engine cost. That's one reason the $52,385 Z06 can compete with sports cars costing twice as much.
So if the pushrod design makes such a good V-8, why does GM make a DOHC V-8 Northstar? "I'm not going to touch that one," laughs Winegarden. GM's party line is that some customers want what it calls "high-feature engines." Winegarden does admit there are some refinement benefits to the DOHC layout, but personally, I don't find the Vette's engine to be a bit unruly.
GM is no longer the pushrod's sole champion. Dodge has reintroduced the pushrod Hemi V-8. And this year, Toyota is running in the NASCAR Craftsman Truck Series with a newly developed pushrod V-8. Of course, that's a motor used only for racing, but is it so far-fetched to wonder if Toyota would make a production pushrod V-8?
Such an engine would be cheaper to build than the complex DOHC V-8 that's now in the Tundra pickup, and Toyota would also realize some marketing benefits from the huge racing investment. The irony is almost too sweet to imagine."
Rich Krause
"A pushrod is basically a metal tube designed to transfer the reciprocating motion of a valve lifter, riding on the camshaft, to the valve.
I found the pushrod fascinating 20 years ago, back when I was a high-school car geek devouring car magazines. It was about the time the Japanese car invasion began and Detroit started to worry about loss of market share here. In the mid-'80s, Detroit stuck by the pushrod, and nearly every available engine used it. Domestic cars were for the most part, however, mediocre contraptions that allowed foreign automakers to make inroads into the U.S. market. Back then our 10Best Cars competition had five spots for domestics and five for imports. In 1983, the pickings were so slim that the Chevrolet Caprice Classic made the list.
When asked what made Japanese cars better, critics gave a variety of answers, but one stuck in my mind: While the Americans produced crude, underperforming pushrod engines, the Japanese were turning out sophisticated four-cylinder, single- and double-overhead-cam engines with twice the number of valves per cylinder.
The idea of locating the camshafts in the cylinder head was not new, but to a populace still unaware that a 1929 Duesenberg Model J, for example, had double overhead cams and four valves per cylinder, the overhead-cam engine sounded advanced, exotic.
The pushrod was the scapegoat for all that was wrong with Detroit. "No doubt about it. In the '80s, the pushrod was a hated component, a symbol of the uncompetitiveness of the domestic industry," says Sam Winegarden, GM's chief engineer for small-block V-8s.
I called Winegarden because I've always been fascinated by GM's decision to stick with the pushrod when it introduced a new small-block V-8 in 1996.
By the '90s, many domestic four- and six-cylinder pushrod engines had been scuttled in favor of overhead-cam designs. In 1995, Ford axed its venerable pushrod 16-valve V-8 for a V-8 that still had just 16 valves but now ran with a pair of overhead cams.
Keeping the pushrod seemed like yet another example of the General's hanging on to its glory years. But since then, I've come to see that, for a mass-produced V-8 engine, the pushrod layout has more advantages than disadvantages.
"Power is a function of airflow, pure and simple," says Winegarden. "If you're not going to fill the cylinder with valves, why have an overhead cam?"
Although you could design a V-8 engine that uses pushrods to operate four valves per cylinder, the layout lends itself to using just two. Four valves generally have greater total valve area than two, thus creating greater airflow and more power. In the case of the Chevrolet Corvette Z06 and its pushrod two-valves-per-cylinder 405-hp, 5.7-liter V-8, Winegarden says, "We've been able to meet the performance requirements by using more displacement. Two more valves per cylinder would get us another 10 percent in total valve area."
A modern engine is a dizzying array of compromises as designers strive to meet power, cost, reliability, and emissions demands. The Vette V-8 would ultimately have more power if it had four valves per cylinder, but it would lose ground in cost, complexity, and physical size.
"On the Vette, performance density is only beaten by some of the big Ferrari or heavily boosted engines," Winegarden continues. One of the huge advantages of the pushrod V-8 layout is its clever use of space. The camshaft and the pushrods lie in the unused area between the cylinder banks. A DOHC or SOHC V-8 is much wider and a little taller because the camshafts are on top of the cylinder heads.
Enthusiasts often talk about how much power an engine produces as a function of the combined volume of the cylinders (power per liter). But what's often overlooked is the power produced compared with the outside dimensions of a motor.
The small-block is a compact engine. I did some rough measurements of the Z06's V-8, the DOHC V-8 in the Porsche Cayenne S (335 hp, 4.5 liters), and the Northstar V-8 in the Cadillac SRX (320 hp, 4.6 liters). The small-block was about six inches narrower than the Porsche V-8, two inches shorter than the Caddy's V-8, and close in length.
That may not sound like much, but it gives designers more freedom to position the engine for good weight distribution and leaves ample room for suspension and steering components.
Don't forget about the weight advantages. The Vette's engine has one camshaft and a small drive chain that's connected to the crank. A DOHC V-8 has four camshafts and drive gears, beefy heads to hold the shafts, and two long chains. "Our V-8 is a simple and elegant design. It's pretty easy to put together, which helps the reliability, and costs," says Winegarden. He estimates a $400 saving over a DOHC motor. And, of course, with fewer parts comes less weight. (Winegarden says the aluminum small-block is 44 pounds lighter than the Northstar.)
GM also used this architecture for its truck engines. Spreading the tooling and design costs over a large number of vehicles reduced the per-unit engine cost. That's one reason the $52,385 Z06 can compete with sports cars costing twice as much.
So if the pushrod design makes such a good V-8, why does GM make a DOHC V-8 Northstar? "I'm not going to touch that one," laughs Winegarden. GM's party line is that some customers want what it calls "high-feature engines." Winegarden does admit there are some refinement benefits to the DOHC layout, but personally, I don't find the Vette's engine to be a bit unruly.
GM is no longer the pushrod's sole champion. Dodge has reintroduced the pushrod Hemi V-8. And this year, Toyota is running in the NASCAR Craftsman Truck Series with a newly developed pushrod V-8. Of course, that's a motor used only for racing, but is it so far-fetched to wonder if Toyota would make a production pushrod V-8?
Such an engine would be cheaper to build than the complex DOHC V-8 that's now in the Tundra pickup, and Toyota would also realize some marketing benefits from the huge racing investment. The irony is almost too sweet to imagine."
Rich Krause
Registered User
Joined: Sep 2002
Posts: 655
Nice reading.
I'll have to admit, I've been one to want GM to move toward the
over-head cam in the F and Y Body (minus LT5) for years.
After reading that, it's not so favorable to want the over-head design.
I guess having a pushrod design in a motor that lives under 6000
RPM can't really use the rotating and reciprocating mass argument
to justify the over-head cam alternative.
With respect to engine efficiency, how much better would the
mult-valve head compare to the twin valve head?
Does anyone have the VE% and BSFC values for naturally aspirated 350's? Northstar motor vs. LS6, or LS1?
P.S. Can someone design a pushrod motor that allows me to
swap cams in one hour like an overhead design?
I'll have to admit, I've been one to want GM to move toward the
over-head cam in the F and Y Body (minus LT5) for years.
After reading that, it's not so favorable to want the over-head design.
I guess having a pushrod design in a motor that lives under 6000
RPM can't really use the rotating and reciprocating mass argument
to justify the over-head cam alternative.
With respect to engine efficiency, how much better would the
mult-valve head compare to the twin valve head?
Does anyone have the VE% and BSFC values for naturally aspirated 350's? Northstar motor vs. LS6, or LS1?
P.S. Can someone design a pushrod motor that allows me to
swap cams in one hour like an overhead design?
Registered User
Joined: Nov 2001
Posts: 2,985
From: In a house by the bay
Long live the Pushrod!
Some of the most beautiful designs are also the simplest. Mikhail Kalashnikov's AK-47 is the perfect example. Not as flashy as an M-16 but tenfold more reliable.
Good article and I agree with Winegarden, tis indeed ironic.
Zero,
Don't know exact numbers for VE on those motors but it stands to reason that the potential is there when you look at it in terms of valve-area in relation to piston-area.
As for changing cams.... all a matter of packaging. I can swap a cam in about an hour..... just depends on which V8 you're talking about.
-Mindgame
Some of the most beautiful designs are also the simplest. Mikhail Kalashnikov's AK-47 is the perfect example. Not as flashy as an M-16 but tenfold more reliable.
Good article and I agree with Winegarden, tis indeed ironic.
Zero,
Don't know exact numbers for VE on those motors but it stands to reason that the potential is there when you look at it in terms of valve-area in relation to piston-area.
As for changing cams.... all a matter of packaging. I can swap a cam in about an hour..... just depends on which V8 you're talking about.
-Mindgame
Registered User
Joined: Jun 2002
Posts: 2,456
From: sarasota, fl
Originally posted by Zero_to_69
P.S. Can someone design a pushrod motor that allows me to
swap cams in one hour like an overhead design?
P.S. Can someone design a pushrod motor that allows me to
swap cams in one hour like an overhead design?
sure....just change cam when motor is pulled!
atleast we dont have to buy 4 cams, 32 new valves, 32 springs, 32 locks/retainers !
Registered User
Joined: Oct 2002
Posts: 2,931
From: Upstate NY
Good read, Rich.
I still like the idea of two cams stacked in the LS block with pushrods to allow variable phasing (sometimes called "variable timing"). If the upper cam was the intake, the pushrods woud be very SHORT, and rpm limits could jump.
Ironic that the racing LS-6 Cadillac CTS-V was recently handicapped by the governing body (SCCA maybe??) by adding 200 lbs to the car, reducing intake restrictor size and limiting rpm on its "low-tech" pushrod engine so the cars from Yurrip (remember Satch anyone?) can compete. Heinrichy could only manage a third last week.
I still like the idea of two cams stacked in the LS block with pushrods to allow variable phasing (sometimes called "variable timing"). If the upper cam was the intake, the pushrods woud be very SHORT, and rpm limits could jump.
Ironic that the racing LS-6 Cadillac CTS-V was recently handicapped by the governing body (SCCA maybe??) by adding 200 lbs to the car, reducing intake restrictor size and limiting rpm on its "low-tech" pushrod engine so the cars from Yurrip (remember Satch anyone?) can compete.
Heinrichy could only manage a third last week.
Registered User
Joined: Sep 2002
Posts: 655
Mindgame,
We're talking about the first gen SB (L98). Pulling waterpumps,
rads, timing covers, intake manifolds, etc.
As for the VE% specs and BSFC, I'm just wondering how much
"better" the SOHC/DOHC design is over the pushrod.
Power per cube and $$$/HP, is it really worth the switch?
The Viper and Vette are still pushing rods...
We're talking about the first gen SB (L98). Pulling waterpumps,
rads, timing covers, intake manifolds, etc.
As for the VE% specs and BSFC, I'm just wondering how much
"better" the SOHC/DOHC design is over the pushrod.
Power per cube and $$$/HP, is it really worth the switch?
The Viper and Vette are still pushing rods...
Registered User
Joined: Oct 2002
Posts: 2,931
From: Upstate NY
Originally posted by Zero_to_69
As for the VE% specs and BSFC, I'm just wondering how much
"better" the SOHC/DOHC design is over the pushrod.
Power per cube and $$$/HP, is it really worth the switch?
As for the VE% specs and BSFC, I'm just wondering how much
"better" the SOHC/DOHC design is over the pushrod.
Power per cube and $$$/HP, is it really worth the switch?
Let's see: (LS6) 405 hp/ 346 CID = 1.17 hp/cube
4.6 L = 281 cu. in. @ 1.17 that's about 329 hp. Where's the DOHC 4.6 Ford? 320 in the highest NA 4.6, I believe.
LS2's 400 hp base 6.0 L engine power/cube equates to a 307 hp 4.6 DOHC which is right where Ford is in the "standard" 4.6 4 cammer.
Of course "performance density" which one might equate to "vehicle weight per hp" is interesting if the LS6 and 4.6 DOHC engines mentioned above are compared in the same weight vehicles:
3400 lb/405hp = 8.4 lb/hp
3400/329 = 10.3 lb/hp or a 22% advantage for the LS6.
Of course when the 4.6 gets a superchanger the numbers improve to 3400/390 = 8.7 . Now the (NA, 2 valve pushrod)LS6 only has a 3.5% advantage.
I'm not sure if I believe the $400 cost advantage Winegarden claims unless he is factoring in warranty costs
, but the weight and balance advantage of the LS engine can't be denied. Even $200/engine is HUGE when you are making millions of engines.Inspite of the valve area/piston area advantage for a 4-valve, it's more what you do with what you have than what equipment you have. That applies to other things in life also.
I also find it ironic that the SOHC 5.4 L truck engine has a torque peak at 2500 and power peak at 4500 and the 5.3 pushrod LS truck engine is at 4000 and 5200. Where is the OHCs high rpm capability being used here? Remember GM & Ford both sell way more of these V8s in trucks than in cars...WAY more. I'm guessing that when it comes to a 100,000+ mile rebuild a SOHC or DOHC engine is going to need a lot more $ of wear parts just in the cam drive than a pushrod engine.
My $.02
Thread Starter
Moderator
Joined: Dec 1969
Posts: 10,745
From: Buffalo, New York
Originally posted by OldSStroker
HP/cube is probably a fairly good measure of VE.
Let's see: (LS6) 405 hp/ 346 CID = 1.17 hp/cube
4.6 L = 281 cu. in. @ 1.17 that's about 329 hp. Where's the DOHC 4.6 Ford? 320 in the highest NA 4.6, I believe.
LS2's 400 hp base 6.0 L engine power/cube equates to a 307 hp 4.6 DOHC which is right where Ford is in the "standard" 4.6 4 cammer.
Of course "performance density" which one might equate to "vehicle weight per hp" is interesting if the LS6 and 4.6 DOHC engines mentioned above are compared in the same weight vehicles:
3400 lb/405hp = 8.4 lb/hp
3400/329 = 10.3 lb/hp or a 22% advantage for the LS6.
Of course when the 4.6 gets a superchanger the numbers improve to 3400/390 = 8.7 . Now the (NA, 2 valve pushrod)LS6 only has a 3.5% advantage.
I'm not sure if I believe the $400 cost advantage Winegarden claims unless he is factoring in warranty costs , but the weight and balance advantage of the LS engine can't be denied. Even $200/engine is HUGE when you are making millions of engines.
Inspite of the valve area/piston area advantage for a 4-valve, it's more what you do with what you have than what equipment you have. That applies to other things in life also.
I also find it ironic that the SOHC 5.4 L truck engine has a torque peak at 2500 and power peak at 4500 and the 5.3 pushrod LS truck engine is at 4000 and 5200. Where is the OHCs high rpm capability being used here? Remember GM & Ford both sell way more of these V8s in trucks than in cars...WAY more. I'm guessing that when it comes to a 100,000+ mile rebuild a SOHC or DOHC engine is going to need a lot more $ of wear parts just in the cam drive than a pushrod engine.
My $.02
HP/cube is probably a fairly good measure of VE.
Let's see: (LS6) 405 hp/ 346 CID = 1.17 hp/cube
4.6 L = 281 cu. in. @ 1.17 that's about 329 hp. Where's the DOHC 4.6 Ford? 320 in the highest NA 4.6, I believe.
LS2's 400 hp base 6.0 L engine power/cube equates to a 307 hp 4.6 DOHC which is right where Ford is in the "standard" 4.6 4 cammer.
Of course "performance density" which one might equate to "vehicle weight per hp" is interesting if the LS6 and 4.6 DOHC engines mentioned above are compared in the same weight vehicles:
3400 lb/405hp = 8.4 lb/hp
3400/329 = 10.3 lb/hp or a 22% advantage for the LS6.
Of course when the 4.6 gets a superchanger the numbers improve to 3400/390 = 8.7 . Now the (NA, 2 valve pushrod)LS6 only has a 3.5% advantage.
I'm not sure if I believe the $400 cost advantage Winegarden claims unless he is factoring in warranty costs
, but the weight and balance advantage of the LS engine can't be denied. Even $200/engine is HUGE when you are making millions of engines.Inspite of the valve area/piston area advantage for a 4-valve, it's more what you do with what you have than what equipment you have. That applies to other things in life also.
I also find it ironic that the SOHC 5.4 L truck engine has a torque peak at 2500 and power peak at 4500 and the 5.3 pushrod LS truck engine is at 4000 and 5200. Where is the OHCs high rpm capability being used here? Remember GM & Ford both sell way more of these V8s in trucks than in cars...WAY more. I'm guessing that when it comes to a 100,000+ mile rebuild a SOHC or DOHC engine is going to need a lot more $ of wear parts just in the cam drive than a pushrod engine.
My $.02
The "power density" concept makes a lot of sense. I was looking at the engine from a Honda S2000 recently. Very impressive hp/L and therefore VE, to be sure. But it freakin' HUGE for a 2L. It looks to me like you could shoehorn an LS6 into the S2000's engine compartment. Now THAT would be an interesting swap.
As far as cost, I am sure the number of units has a major effect on cost. And I understand that retail prices and cost don't have a very close relationship. But what's the retail cost of an LS6 longblock? Maybe $6K. A 400hp longblock from Porsche is probably $20K and more like $40K if it has a prancing horse on the valve covers.
Don't get me wrong, I like all kinds of powerplants. But what I like most of all is something that does the job better and cheaper because of better engineering, not just complexity for its' own sake.
Big kudos to GM for the Gen III.
Rich Krause
Banned
Joined: Oct 2002
Posts: 6,518
Originally posted by Mindgame
Some of the most beautiful designs are also the simplest. Mikhail Kalashnikov's AK-47 is the perfect example. Not as flashy as an M-16 but tenfold more reliable.
Some of the most beautiful designs are also the simplest. Mikhail Kalashnikov's AK-47 is the perfect example. Not as flashy as an M-16 but tenfold more reliable.
As for Pushrods, I just say put them higher in the block. Give me a shorter PR and a longer timing chain any day. Or even better a shorter deck height, and short PR's.
Pile on that ROCKER ARMS! yes OHC can do them but 1.7's are usually not the norm there. .100 more lift on a production motor is a big deal.
Still really don't think that the Ford Mod motor is worth much, much rather have a SN95 with a 5.0L in it and 2000+ body work on it. Hell I want a 5.0L with a 408 W swap in it for my next toy car. (Yes I know even OldSStroker thinks that's scary!)
Bret
Registered User
Joined: Feb 2000
Posts: 1,445
From: Seattle, WA
Isn't the LS1 cam a lot higher than in a small block? I'd imagine a TALLER deck-height would be better for R/S ratios and strokers... depends how many cubes you actually want I guess.
I gotta tell you though... it's tough not to fantasize about a 10" deck height and a dry sump oil pan on the T/A. Granted the sump's easy enough to convert, but dropping the mounts 2 or 3" to actually use it, and then find someway to get a taller deckheight in there for a nasty stroker would be a real fun project. Crazy expensive... but fun. Looking at the displacement vs. packaging issues it's not hard to see why big blocks rule on the strip but have problems getting into production vehicles (does ANYONE still make a big block in production vehicles?)
I gotta tell you though... it's tough not to fantasize about a 10" deck height and a dry sump oil pan on the T/A.
Granted the sump's easy enough to convert, but dropping the mounts 2 or 3" to actually use it, and then find someway to get a taller deckheight in there for a nasty stroker would be a real fun project. Crazy expensive... but fun. Looking at the displacement vs. packaging issues it's not hard to see why big blocks rule on the strip but have problems getting into production vehicles (does ANYONE still make a big block in production vehicles?)
