It seems that many (most?) older engine designs rely on having a single cam down near the crank and a series of pushrods to operate the valves. OHC involves having a timing chain that can stretch or break and completely ruin your valvetrain. What advantage does an OHC design offer over the more traditional pushrod design?
Also why do some higer-performance cars have dual overhead cams? This seems to result in more power, but how?
The overhead cam opens the valve more precisely (it pushes directly on the valve). The further away the cam, the less precise the transmission.
dual cam -> two valves -> breathe in, breath out.
rwj
Less mass in the valve train (no pushrods and sometimes no rockerarms) means less valve float. This means the engine could be run at higher rpm or could have more abrupt valve opening and closing. In any case it provides better control over the valve timing.
BTW, many (if not most) pushrod engines have cam chains, they’re just much shorter and maybe stronger.
Interesting–does anyone have any timelines for these things? I.e., when OHC eclipsed pushrod engines, and when cam chains replaced timing gears?
Overhead camshafts actually date back to before world war one. Peugeot actually built a race car that had a DOHC motor in 1912. Most production cars produced between the two world wars actually were flathead engines because they were the least complicated engine to build. Overhead valve engines didn’t become common until after WWII. The latest V8s from both Chrysler and GM are still overhead valve engines, though they were recently redesigned from the ground up and are otherwise modern engines.
The reasons they used overhead valve engines are quite simple. They are easier to package, since they are shorter and narrower. They are somewhat cheaper, since they only have one camshaft, and they produce more torque than the equivalent overhead cam motor. This is especially helpful in trucks, which is the main use for those large V8 engines.
What’s the difference between “flathead” and “overhead valve”?
Flathead engines don’t have their valves in the top of the cylinder head, rather in the block beside the piston.
The valve train on a flat head is contained entirely in the engine block.
See T-head picture here http://abbysenior.com/mechanics/valve.htm as an example of flat head design. There are more photos of other valve train designs down the page.
I was just talking to someone about this yesterday. Why do some cheaper cars have 4 valve designs but not DOHC? I assume there’s an advantage in having the additional valves, but that DOHC is more expensive and not considered worth the cost in cheaper cars.
Er, I agree with the rest of your statements, but the claim in bold is gonna require some additional rationale. Why do you think OHC engines inherently produce less torque than pushrod configurations?
Stranger
The way I understand the 4 valves per cylinder thing is that fluids will flow at a flaster rate through several smaller holes than one bigger hole with the same cross section. So, having 2 intake valves with total cross section of, say, 4 square inches will result in faster airflow than one intake valve with the same cross section. This is also the reason why on motorcycles and old cars having more carburetors = more power.
This is the way it was explained to me. I could be way off. But I still don’t understand why having DOHC is better than just plain ol’ OHC.
With one overhead cam you still have to have rocker arms, with two you don’t.
No, on a 2 valve per cylinder OHC engine you can direct act on the valves without any rockers, you just design the engine that all the valves are in a straight line. An example of this design would be a Volvo B21, B23, B230 engine (same basic design different displacements, produced from 1976-1995 in 200-and 700 series cars)
One big advantage of OHC engines is lower maintence. Back in the day, you needed to adjust OHV engines every 12,000 miles or so. Some modern engines with OHCs and mechincal cam followers have an adjustment interval of the life of the engine. In other words, as long as you don’t fuck with it, you don’t have to fuck with it. 
This cuts the maintence cost way down, and also lower emissions in the real world where not every consumer has all of the required maintence done, and not every mechanic does a maintence job right.
Funny you should mention Volvos…I was thinking of my '74 Volvo 144’s B20B and how simple that design was when I was writing my OP.
Ehrm, how critical is this?
I have an older OHV engine in a tractor and I’ve never adjusted anything valve-related (although I’ve done just about every other standard maintenance chore) on it. What are the consequences of not doing the adjustment, and what exactly is adjusted?
Here’s a Wikipedia article on pushrod engines. Really, it’s common knowledge that pushrod engines make more torque. This article also mentions it, and also mentions that the power usually comes on at a lower rpm.
On adjusting valves, most if not all cars have hydraulic lifters since I can remember. The only cars I can think of that had solid lifters that would have needed adjustment were the high performance versions of the Chevy 350 V8 in the sixties. Of course with solid lifters you can rev the engine much higher than with hydraulic lifters.
In Formula One engines they’ve done away with camshafts entirely.
When it’s time for the engine to start, multiple sensors on the crankshat indicate to the engine’s digital management system where the crankshaft is, in relation to the engine’s firing order. A software map then activates pneumatic compressed air via solenoids to open every valve in the engine precisely the right amount per cylinder per compression stroke and voila! She kicks over.
From that point onwards until the session is finished, the valves are being blown open or blown shut by solenoids and highly compressed air stored in pressure tanks almost identical to a scuba tank. If your scuba tank springs a leak and runs low, you’re in trouble!
The system was first pioneered by Renault in 1986. Apparently Ayrtton Senna used to say that he could hold a given gear if he wished by an extra 2,000 rpm. The redline back then was 15,500 rpm. That was VERY impressive for the era. In that time, the engines are now nudging 19,000 rpm quite successfully for an entire race.
There’s no way that springs could handle such rpm without creating valve bounce. Well, in 250cc four stroke motorbikes they can, but when you get into 3 litre plus engines, your moving mass gets much higher and longer and the springs needed to spin 19,000 rpm would be gargantuanly big and heavy. So, a digital pneumatic system is the way to go.
Getting back to the OP however, Rick summed it up best. A nicely designed SOHC system can be a very, very reliable machine. It’s a nice compromise which produces good revs and good horsepower without a limiting degree of unreliability brought on by complexity. And there’s the rub you see. Engines like the Cosworth DFV were magnificent masterpieces, but they required regular rebuilds. That was what, 1968 and some 165 Grand Prix wins ago?
The technology isn’t new - with the exception of the pneumatic solenoid system - but the compromise has been easier to justify in terms of greater complexity these past 20 years due to greater levels of quality control.
Methinks they are comparing apples to oranges in that article; I am not an engine designer, but I don’t see that OHC vs OHV configurations make any difference in terms of peak torque, which has more to do with stroke vs. bore and the number of cylinders and firing order/cylinder geometry/crank design than the valve actuation method. Chevrolet’s vaunted small-block V8 (an OHV pushrod design) has high torque on the low end not because it is a pushrod but because of the angle between the banks and the particular selection of stroke length vs. bore; the VW low angle V6 (with 15 degrees between the banks, making it almost an inline 6) has relatively low torque but high power output at elevated RPMs because of its geometry. (Similiarly for inline and boxer-type engines.)
OHV allows for a simplier, cleaner design at the expense of greater inertia and less control (as compared with modern variable timing systems). I don’t see how this mandates greater torque for the same engine geometry and stroke/bore configuration.
Stranger
To be fair however, that has as much to do with cam profiles and stroke length as it does with valvetrain design. It’s not quite accurate to assert that the valvetrain itself is responsible for such torque curves.
I own a hi-po 66 Mustang Fastback with the K-code engine in her. She has the press in studs and Crane roller rocker arms etc. Solid lifters on the cam. A unique sound to be sure. From what I’m reading, the jury’s still out on whether a new tech roller cam or old tech solid lifters produce the most revs. Whatever, you need good rocker studs to handle the pivot speeds. But that’s another discussion for another thread.
Dual overhead cams allow optimal valve placement and angle for most combustion chamber designs. If you’ve got to operate both intake and exhaust valves with the same cam, you either compromise on valve position or add extra hardware to get the job done.
Engines with hydraulic lifters (virtually all American-car engines of the last 50+ years, but by no means virtually all engines worldwide) are sef-adjusting in terms of valve lash. No maintenance needed. Engines with shim-type lash adjustment should be checked regularly (typically 60,000 miles for recent designs, more often for 60’s-70’s era designs), but often maintain acceptable lash beyond the checking interval. Engines with screw-type lash adjusters (most European and Asian cars of 60’s through 80’s, less since then) typically specify adjustment every 15,000 miles. In the real world, a 30,000 mile interval is adequate for the great majority of them.