Lighter valve springs - an OEM efficiency tactic
 

I was sure this topic was discussed before, but couldn't turn it up in a search.

Anyone who's ever worked on a head for the first time may have been surprised at the amount of pressure valve springs exert on the cam. Try turning a lubricated cam in a head - there's an unexpected amount of resistance!

The idea of going with valve springs with less resistance was suggested - as something that would likely return a very small improvement in efficiency.

I happened across a reference that suggests using lighter springs is actually a tactic automakers have used to reduce fuel consumption:

Source is a discussion of putting an Altima engine in a Sentra (and then building/tuning it for more power).

2002 Nissan Sentra SE-R Spec-V - Sport Compact Car Magazine

Not that this is particularly useful info. How are you going to get lighter valve springs? Or decide "how light is too light?"

I suppose "too light" would be when the springs don't exert enough force to shut the valve completely before the next stroke. If for instance the intake valve isn't closed when the compression stroke starts, it'd shove some of the fuel/air back out the intake...

On old, very low-speed engines, you may find no valve springs at all. Air opens and closes the valve. The main job of the valve springs is to avoid "float" -they overcome the inertia of the valve and lifter and keep them in contact with the cam as they start to close. That force goes up as the square of speed, so it pays to avoid high revs. If you are going to lower the red-line, a cam with less dwell and overlap is better, as there is less inertia in the manifold gasses.

I think you are thinking of engines with no cam on the intake. Cylinder vacuum opens the intake and cylinder pressure closes it.

There are engines with no springs, cams open and close the valves. Ducatti is famous for there valves without a valve spring proper.

Ducati Desmo Valves

Do any cars have valves closed by a mechanical linkage?

Like Ducati: Ducati.com || Bikes || Tech cafè

Mercedes sports-racing cars from the 50s had them, and perhaps a few other exotics. It would be nice to see that combined with variable valve timing. Perhaps hydraulic lifters plus hydraulic closers would cure the clearance woes.

A particular problem on an interference engine. Snap, crackle, pop.

Here's a question: do valve springs get significantly weaker over time/use?

I happen to have a complete, worn out 1.0L engine...

most springs are just steel. Depending on the strength of the steel and the amount of flex they're under will determine the life before unacceptable margin of stress makes them un-useful. That is a very spring specific thing. BUT! The springs will most likely die long after piston rings or bearings fail. Diesels last what, 1-2-4 million miles? I don't work on industrial diesels, but I'm betting the springs in OTR trucks aren't ever replaced unless they are stretched or compressed out of spec.

A friend of mine does local deliveries, home every night; anyways, his truck just had its first serious engine service at 1 million miles. Sure it's had the weekly and monthly oil, fluids, leakdown etc tests, but never the engine cracked open. At the 1 mil mark they put new crank bearings in, new piston rings, new turbo seals, and put it back together (he's a car nut, so he asked the mech what was done). He says it runs a little stronger now, probably has slightly less friction and higher compression. Netted him about .3mpg more too.

Placibo? maybe, but the .3mpg is proof its running better. When you average 6.85mpg over 800 miles in a day, .3 is a huge savings.

Bottom line I'm saying: Steel springs last forever. Titanium lasts even longer (less pron to fatigue). Go with lighter springs if you want slightly better FE, but keep the rpms down to prevent valve float. Don't want any head crashes...

Steel springs are run below their fatigue limit, but only became really reliable after shot-peening became common. That was the secret of Toledo swords, too, BTW. Anyway, if you were to have your springs etched down, you could get any strength you want, and the loss of the surface treatment would probably be compensated by the reduction in strain, if the change is substantial. Springs, valve train weight, RPM and lobe shape all interact. The reduced duration on an XFI cam gives it a peakier lobe, even with less lift, so the need for spring strength is not reduced along with the power band. If you are willing to be religious about a low red-line, there are definite gains available, though.

Fortunately, the Suzuki 993cc engine in the Metros are non-interference.

One of the things I am planning when I finally get around to the head swap in my car was getting weaker springs. With the way the engine is tuned right now I can't really get the engine to redline anyway the top end is totally weak. I get the best power shifting at 3500 so if I find a set of springs that will float at 3750 I am set :)

I have a pile of Metro springs sitting around I just have never measured one to see what size they are so I know what to order to replace them. If it ever gets warm here I might get out there and work on something.

Even so, spring bounce is a bad thing. Pains must be taken to hold revs below that threshold. That probably means an electronic rev limiter of some sort.

Desmodromic valve actuation is a maintenance nightmare. Also running two valve trains probably exerts more drag than stiff valve springs.

You probably could lighten up the springs if you slowed the engines down from the howling RPM most small engines turn.

That raises the question: how much internal friction or HP savings is there with lighter springs?

That would depend on if were talking flat tappet lifters or roller lifters.

OK even so...

The Ducati Desmodromic system can be spun on the test bench with a snap of the fingers...literally.
Inertia keeps the valves moving quickly enough to seat fully even at idle speeds to eliminate the need for springs which the Mercedes cars used in the late 1950's.

Maintenance isn't a headache BUT you must know what you are doing , it take a lot of time AND it is very easy to get it wrong.
For the consumer this means higher service costs which not all are prepared to pay.

For Ducati it is as much a sales tool as it is an efficiency concern.
Look at the number of bikes not using it and that becomes very clear.

At the moment Ducati are the only builders using that system.

Pete.

Frank ,

I recall reading a statement giving the numbers around 25% of engine power but I am unable to find the source.

Pete.

OK Found it after all.
The numbers are lower than 25% and are reported as 19%
A quick copy / paste of the abstract is below:

Engine friction lubricant sensitivities: A comparison of modern diesel and gasoline engines
R. I. Taylor
Shell Research and Technology Centre, Chester, UK

Abstract
Engine friction models have been developed that take account of the variations in lubricants with temperature, shear rate, and pressure. These models have been used to study the lubricant sensitivities of modern diesel and gasoline engines. Total engine friction losses for a Perkins Phaser four-cylinder, 4.0 l, turbocharged, inter-cooled diesel engine, operating at 1300 rpm, with an SAE 15W-40 lubricant, were estimated at approximately 2 kW, with the piston assembly contributing 46%, the bearings 49%, and the valve train 5%. Total engine friction losses for a Mercedes Benz M111 2.0 l gasoline engine (used in CEC sludge and fuel economy engine tests) operating at 2500 rpm, and medium load, for an SAE 15W-40 lubricant, were estimated at 1.5 kW, with the piston assembly contributing 42%, the bearings 39%, and the valve train 19%.

Pete.

Good find, Pete. Nice to see some numbers on the subject.

They make it clear why Suzuki went with a 2-ring piston on the higher-efficiency Metro XFi rather than the 3-ring pistons in the garden variety engine.

I've never seen any mention of valve spring differences between the two though.

Yes, good find. Still wondering about the savings or potential savings from spring changes. Also should find out more about those two engines- one with 5% valvetrain losses, one with 19%. What's the diff?

Diesel vs gas engine Frank. Makes sense the gasser would have an higher percentage of losses from the valve train since they rev higher and have lower compression.

19% of 1.5 kw is roughly .38 hp. Suppose you halve the valve train losses, and your car uses 15 hp going down the road at 55 mph, that's 2.5% better FE.

Making the valve train lighter can also help reduce the forces needed to keep adequate control of the valves. If you know that you're never going to run the motor over 3000 RPM, for instance, you can likely get away with an awful lot of lightening of the rocker arms, retainers, and so on. Going to a lighter material (e.g., titanium) for the valves themselves will pay potentially large benefits for weight reduction.

That can let you reduce the spring pressures even further...

I don't know enough about the subject to begin to guess at how light you can go with the parts, or with the springs. But note that often times the super lightweight materials are super expensive... Titanium valves in particular!

-soD

I just had a thought.

The spring resist rotation when opening, but aids rotation when closing.

Not sure what that means just yet.

I think it's a zero sum situation, because there'd always be another lobe resisting rotation while one lobe is aiding. The net effect is just a lot of friction on the cam. Moreso without rollers (as is the case in the Suzuki motor).

I think that depends somewhat on the number of Cylinders (and arraignment

For example I was adjusting valves on a 7m (I6) last week and it "feels" easier to turn the cams than a 5sfe (I4). Very very not scientific, I'd guess it is a case of smoother, not less energy, just thinking . . . .

Let's remember that valve train losses are pretty much linear with RPM, while other internal friction varies with load. There seem to be many benefits to lower speed engines. Too bad racing formulas were based on displacement, rather than weight or fuel.

Frank ,
Apart from the obvious RPM differences for the two engines mentioned (1300 for the Perkins and 2500 for the Mercedes) I a guessing the detail design differences (finger roller valve operation versus bucket tappet operation etc) may account for the rest.

The comments about the actual values of the valve springs themselves being basically a zero sum item is essentially correct as Metro MPG stated above.
The force needed to compress the valve spring is returned as the compressed spring releases the energy stored in it.
There is a small amount of internal friction within the steel and this is released as heat.

The differences in spring pressure add to the friction component of the engine by the added pressures they apply to the items they contact like camshaft bearings , rocker bearings and pushrod contact points.

Cheers , Pete.

AHA! I wasn't crazy. I had already started a thread on this topic a while ago. Different forum though:

valve springs - GasSavers.org - Helping You Save at the Pump

:D

A quick copy / past from another board discussing the same topic:
"Work is defined as Force * Distance. Say we have a valve spring with a seat pressure of 245 lbs. and installed height of 2". We also have a valve opening of 0.700 which gives an open spring pressure of 600 lbs. The work done is the area under the spring curve between the 2.000" installed height and the 1.300" open height. In this case about 296 in-lbs or about 25 ft-lbs.

Power is defined as the time rate of doing work, or Work/Time. At 7200 RPM the cam is spinning 3600 RPM, or 60 RPS, or 21600 cam degrees per second. For our cam, 264 @ 0.050, there are 66 cam degrees from 0.050 to max lift. Let's add another 24 degrees to from the base circle to max lift and call it 90 degrees.

This means it takes about 0.004 seconds to go from the lift ramp to max lift. So if we take our 25FtLb/.004sec = 6250 FtLb/Sec, and with 1 Hp=550 FtLb/sec, it takes 11.36Hp to open the valve at 7200 RPM. We can see that if the engine were only running at 3600 RPM, it would take twice as long to open the valve, so it would only take half the horsepower to open the valve.

Now that's just the power required to open the valve. It really doesn't answer the question about power consumed. That same energy stored in the spring at max lift has to go somewhere, so it delivers it to the cam on the back side. Theoretically, the net power should = zero in my mind. But there will be some loss by heat generated by the spring. Also, at higher RPM, as the cam tosses the lifter for a fraction of a second after max lift, the energy released by the spring doesn't return that power to the cam since the lifter isn't touching the lobe.

So I can see how there are some losses, but I just wonder how the claims of 20-30% came about. When you hear that stuff from several respected builders over the years, you just kinda take it at face value and don't question it."

Cheers , Pete.

2016 Prius: lighter valve springs
 

Zombie thread alert!

I thought of this discussion when I saw this slide from the U.S. press introduction of the 2016 Prius:

http://ecomodder.com/forum/attachmen...1&d=1448245300

From: The 2016 Toyota Prius Is An Enthusiast's Car

The Honda CRX HF valve springs are much weaker than the regular Civic/CRX valve springs. And the HF did have a lower redline. 5000 RPM versus 6500 RPM.

You have an incredible memory Metro.

The new gen IV Prius does produce less HP and torque from the same engine that is in the gen III. I wonder if peak revs were lowered along with lighter springs to achieve this?

Lower horsepower? Do the villagers know? They will want to get their torches and pitchforks.

19%
 

Would the 19% include:
*cam follower
*pushrods/rotators/ ends
*tappets/rockers
*springs
*valve guides
?

lighter springs
 

I a late 1970s SAE Paper about the Shell Mileage Marathon of earlier times,they showcased a 1920s Chevy which used lighter valve springs for their engine strategy.It did lower the useful rpm of the engine,but they topped out at around 55-mph coasting downhill and averaged in the 20-mph range over the course.
A metric which would be interesting about springs would be their hysteresis.
Unlike the rubber of a tire on a paved road,a steel valve spring may have virtually zero physical distortion during service and whatever power is used to open the valve may be almost completely liberated when the valve closes.
A stronger spring requires more effort to compress,but it returns most of that back?
PS I was shocked at how much effort it took to hand rotate the camshaft on the CRX.

Maybe they should go back to the suction operated intake valves. Talk about low spring tension.

regards
mech

If reducing friction is the goal, what about roller rockers? I understand they aren't available/ won't work for Geo Metros (and OHC engines), but they would achieve the desired results, no?

F1 engines use pneumatic valves. No springs.

Also, in my Fiat, the valves are hydraulically opened, the springs only close the valves, essentially halving the amount of time the springs are exerting friction on the cam (and friction is going to be minimal during the closing phase).

roller rockers
 

They're expensive,but they make a measurable reduction in engine friction.
Ford chose them for their Mustang engine many years ago as part of their friction reducing strategy.
*low tension piston rings also
*streamlined water pump impeller