How does Porsche determine redline RPM on various engines?
 

I was wondering if anyone here with engine building experience knows how and/or what determines the redline RPM Porsche designates for their various air-cooled engine combinations? I've had extensive experience building small block Chevrolet stock car racing engines but I'm a newbie with my first air-cooled 1986 3.2. I have to say it's the smoothest, most well-balanced engine I've ever felt by far though. I've built 600 hp SBC engines with shaft mounted rocker arms, forged pistons, racing valves/springs, etc. and our self-determined redline numbers were determined mostly from talking with other racers running similar engine combinations. It was commonly known that you didn't really want to turn the engine combination we were using more than about 8200 RPM for very long or you were going to have problems but we never blew one up due to over revving. In fact, we had a problem on one dyno run and an engine went to 9000 but we never had any problems out of that engine.

One of the reasons I'm curious about this is because it seems like a lot of people have the top end of their engines done around 100K - 150K and I'm in that range. I'm pretty sure my valve guides are worn because I see a little blue smoke from time to time and I'll go through about a quart of oil every 300 miles of normal driving or 100 miles of hard driving. With my engine building experience that amount of oil usage doesn't concern me because we used to run without any valve seals at all and would easily use 1 - 2 qts in a 100 miles of racing but if I could get a little more performance a top end rebuild at this point might be justifiable.

It feels like my 3.2 just loves RPM. It's quite a disappointment to have to shift at 6250. The engine seems like it really just "gets on the cam" at about 5000 and feels like it could pull hard to at least 6750 if I was to take it that far. I guess I'm wondering if valve springs and/or valves are the main RPM limiting factor and if having my heads rebuilt with some performance springs and/or valves would allow me to safely move my redline up? From the limited knowledge I have so far, it seems like some of the early 70's RSR engines have a redline quite a bit higher than mine.

If you are an engine builder you understand the valve springs can't move the large valves in the 911 head really fast. The stock rod bolts can stretch if they are over stressed.

Part of the issue is the breathing inside the case. If you do a total rebuild, and do some fly-cutting on the case, and stronger connecting rod bolts, and other expensive things like titanium valves and such, you can get the revs up. Just open your wallet and empty it out.

I can certainly see where rod bolts might be the weak link. Do the higher revving earlier engines have better bolts? With my limited experience on these engines so far I can see where a wallet can empty very quickly without a lot of HP gain if you don't use discretion. Titanium valves aren't out of the question though if I could move the redline up with just a head rebuild.

From my limited exposure to engine building, I believe the answer is that it depends on a lot of things such as valve float, lubrication delivery, cooling, loading on moving parts (rods, valves, etc.), and so forth.

careful with the worn valve guides too -- as built/spec'ed, the valve to guide clearance helps dissipate heat from the valves to the head --too big a gap between the worn guide (not seal) and valve can allow the valve to overheat and (worst case) drop a head --

I think rod bolts first thing to do then titanium stuff and race springs. A chip and you should be able to rev to 7 grand

Level 1: Ti valve keepers; ''race' valve springs; ARP rod bolts on stock rods; pistons if you want

- I am at level 1 (good to 8k spins) & have it twin-plugged too

windage, etc. is not very imp. on a dry sump motor

* * *

Level 87: electro- valve "train"

Thanks for the input! I'm thinking now I may have asked the wrong questions though. I may should have asked first if the stock intake runners, valve size, cam, etc. will actually allow the engine to keep pulling hard after 6250 or does it fall flat? If it's just an illusion that it wants to keep pulling then I probably don't need to waste the money on eliminating the weak link in the valvetrain or rotating assembly.

3,2s can gain quite a bit from an exhaust upgrade, usually meaning headers, and a chip to match. Spinning it faster is a completely different ball game.

Cam selection is also important. The stock 3.2 cams make peak power (in a stock motor) @ 6100 rpm I think (maybe less); 964 or 20/21 cams, with longer duration, move the peak higher (~6500+). Revving above 7000 really matters most if you have power to go with it.

yup, cams & exhaust, Steve Wong chip

what state are you in? what emissions regs. do they have?

just be glad you don't have CIS

RW Webb... I'm in Alabama. We're 49th in everything... thank God for Mississippi. The good thing about this though is that we don't have any emissions or inspections. Run what you brung and hope you brung enough. ;)

“Speed costs. How fast do you want to spend?”
- Donsco

Interesting responses.

Lets say your tach shows a redline of 6250. My guess is the engine makes peak power around 6000. So anything past that your losing power so why not shift. Now if your racing and your in the middle of a turn then your might stay and let it over rev some. As mentioned that comes back to how strong the engine is built.

For example I have a twin plug 3.0L in my 72T. Built to turn over 8000. I had a special cam cut for track day use. The peak power is at 6900 so I set the rev limiter to 7200. My gearing is such that when I shift a 7000 it drops to 4000 and the peak torque is 4500. Torque is what pulls you out of a turn not rpm above the peak where your losing power.

When I built the engine (with Mike Bruns help in Florida) I wanted to be able to run the daylights out of it and once a year change the oil and adjust the valve. Several years ago I had a missed shift. I saw the tach go past 8000. Everything was fine, no damage and I've put another 10,000 miles on her. Thank you Mike.

I do believe you are confusing Alabama with Louisiana - and you stole our state motto.:D

But feel free to duke it out with Arky'saw for 48th...


anyway you can either put on dual free flow cats or straight SSIs to muffler (depending on your level of personal care for smog contributions; or plant some trees to mitigate)

If you are upshifting for maximum acceleration, the ideal upshift point is clearly beyond the peak power point.

Ideally, the upshift should take place when power in the new gear is as high as power in the old gear.

An easy way to see why this is the case consists in writing down the formula for car acceleration as a function of power and see what's the car acceleration if you upshift at peak power or if you keep the old gear for a little longer.

RWebb... I actually know a guy with a Tesla... will that cancel out the carbon footprint of my oil-burning 911 that is usually found with my right foot planted firmly on the floorboard? If that's not enough maybe I could start recycling like my neighbors. That should help a lot to make up for my contribution to mild winters. They put out their little basket of plastic bottles and milk cartons every Tuesday morning for the 50,000 lb. diesel-burning-hydraulic-monster to roar up and collect. I've avoided the conversation about the reality of how that truck uses more energy sitting there idling for 30 seconds while picking up their plastic bottles than they ever contribute. I think I'll go hug a tree now to make up for my terrible attitude. :D

WPOZZ... That makes sense. I think I've read something about "power under the curve" before where it talks about maximizing the area under the dyno curve to determine the best gearing and shift points. If most dyno curves look like a mountain peak then your maximum area under the curve (which would equate to the most hp) would be evenly distributed before and after the peak.

The cams are on a 113* lobe center for emissions reasons. There just isn't enough valve overlap for breathing above 6700 rpm.

Agreed on the power under the curve philosophy on redline. I’ve understood that a 3.2 can breathe well enough to rev to 7000 if you put headers and a good performance muffler. The stock valvetrain and bottom end will both tolerate 7000. That does put you into a riskier zone if you miss a shift of course - you’ll be that much higher and more likely to bend valves. Better rod bolts are a great investment if you have the engine apart. A top end rebuild will make that accessible.

Thanks Josh & Townsen. That's the kind of information I was looking for. These flat sixes just feel so well balanced internally compared to a V8 it seems like they should be able to turn more RPM. It makes sense that the margin for error would be smaller though. The best way to describe to sweep up through the RPM band is 'effortless' compared to any other engine I've had experience with. There's no discernible harmonic like your get with V8 engines. Sure, a big V8 can make lots of power and lots of noise when forced to but my 911 seems like it's "comfortable" at any RPM in the range and I don't feel even slightly guilty when winding it out. I'm the first to want to enjoy a car at the limits but I don't want to feel like I'm abusing it and my 3.2 seems "happier" at high RPM than lugging around town at 2500.

For the 3.3/3.2 motors, rod bolts are a weak link. They sometimes just undo for no apparent reason. They're well-known to do exactly that if you overrev, usually on a mech overrev from a bad shift. The factory reverted to the larger bolts with the 964.

The factory bolts permanently deform when installed, so are one-use only. This is standard practice & unremarkable - but factory bolts are close to their limit when installed correctly; if they stretch more, you lose tension on the bolt and spin a big end bearing.

ARP or Raceware rod bolts are made from better metal (and can be re-used, because when installed correctly, they don't exceed their elastic limit). In fact, if you're serious about revving high, look into aftermarket rods like Pauter or Carrillo, which are both lighter and stronger - and come standard with better rod bolts.

Revving an SC or 3.2 beyond redline doesn't get you any more power on its own. The cam profile (which gives good low/mid-range response) is already out of its sweet spot. To actually take advantage of the higher redline, you need to change cams.

A chip remap/exhaust on a 3.2 might get you 15-20HP I'm told. That's about it for "easy". It takes significant work/money for a 3.2 to gain 50-60HP - and those changes might reasonably be expected to result in a motor less flexible than stock that will wear out faster.

This just doesn't make sense; forced induction is cheaper and makes more power. A low-boost turbo conversion would get you 300 at the wheels (or more than a 993RS) with a wide powerband and great flexibility for a fraction the cost, all of the mods being external/bolt-on.

A more detailed/comprehensive conversion would make a lot more power. Would still be cheaper than building a highly tuned 3.2 N/A motor.

Thanks Supggy... I didn't even know there was the option of a low boost bolt-on turbo until reading about it a couple of days ago on another post. I'm definitely going to look into that also.

The 3.2s rod bolts are "stretch to torque" and are the biggest weak link. But there isn't any "mystery" to them letting go. It is just a delayed reaction. The bolts have yielded at their install torque so if you over rev the motor past the bolt's holding limit anything extra will yield the bolt farther. And every time you do that the clamp load is lowered. With the lowered clamp load the RPM that it takes to open the parting line is reduced. So your safe redline is reduced and you don't know it. And each time you exceed it you are hammering on the bearings and reducing the bolt's clamp load.
End result is you have one good over rev and you are possibly on a trail to a failure quite a ways down the road when you have completely forgotten about the original over rev.

I suspect that the absolute limit on the bolts is right around the ClubSport chip's redline which is 6800 (or 6850).
Porsche didn't leave much on the table in the way of easy power. The one spot is the exhaust. There is about 20 HP available if you get a good set of 1-5/8 primary headers. With that it will pull hard to 6500. (I haven't changed the 6500 redline in my car so I can't tell you about how it pulls past that.)

- Increasing the RPM reduces a engine's life exponentially. The rule of thumb I was taught about 30 years ago was "Every time you increase the RPM by 1000 the engine life is reduced by a factor of 10." (It probably isn't accurate with modern materials and enginer design but it helps keep it in perspective.)

If you want to go past these limits you need to open up you wallet (and motor) and completely go through it. Valve train, rotating assembly, cam, induction, twin plug, Ps & Cs, ...

Or you you could buy a 3.6 and save some money...

Thanks Quicksilver! Very good explanation. That really answers my question about whether the limiting factor is in the valve train or the bottom end. I think I'll keep gathering information at this point in my quest for the "ultimate daily-driver-torque-monster-instant-throttle-response-streetable-well-mannered" air cooled flat six.

Yep, the only "easy" power is the exhaust. All of your neighbors will hate your car, and you will hate it on any highway trip with all the racket. But if you are just looking for fun back woods drives dodging deer, exhaust is the way to start. Before you dump a ton of money in the car remember is is all just gone, and the next buyer may hate it all, or love the free upgrades you paid for.

Any of the water cooled 911s will have the power you crave. And the brakes and suspension to handle it all.

In the end, it is YOUR car, do to it want you want.

What is really wonderful about the 3.2 valvesprings are you can run some aggressive cam profiles that get you to 7K (or somewhere abouts) without needing to change them.

If you put in ARP rodbolts, I wouldn't be scared to turn up a 3.2 to 7K all day long.

The rods are extremely short and wide girth wise compared to V8's that regularly get turned up to 7-8K (I created a thread about V8 rods to ours long ago).

I think why we keep these motors to such a low RPM is simply the expense of blowing out the cases with a rod failure.

Very, very costly, so we are genuinely very conservative on RPM when in fact, the low end should EASILY take it.

valve springs are the first limiting factor in the redline.
rod bolts would be the second limiting factor
finally... air/fuel mix. EFI has programmed limit and would need to be modified.

I don't want to speak for Steve Weiner, but he has a lot more faith in stock valvesprings than most on here.

People overload the valvetrain with unnecessary spring rates causing premature cam failure for nothing.

I think they have less mass in the pistons and slower piston speeds so the rods/bolts aren't under as much stress at the same RPM as say a stock 3.2.

Again I say when you are looking for power and rpm don't forget to look at the torque curve. For instance if you have an early 911S you have to turn it up to near 7500 rpm to get that power. A 911T with 40 less power with stay with the S for the first 100 yards because of it's peak torque (around 4000) is much lower the the S (around 5000).

My 930 with CIS and factory wastegate made 80% of peak torque @ 3000 RPM (where boost was all-in @ .8 bar). Peak torque around 3500 - and enough of it that transmission longevity is a consideration. My hybrid K27 made 6 PSI boost at 2400 RPM - that, with SC cams, meant it scooted pretty smartly from low RPM.

Power band done by 6K (cams), but a wall of torque is very useful/flexible - and very accessible in most driving conditions.

Mechanical vs. accelleration red line
 

I think it helps to consider two "redlines." The mechanical redline is the RPM above which things might break. Valve float, for instance, possibly leading to valve to piston contact. In the 3.2, the rod bolts are not as strong as they were on the earlier engines - Porsche reduced the diameter by 1mm to deal with the 74.4mm crank. The fix for this is to use aftermarket super rod bolts when rebuilding. Short of a missed shift overrev, a stock 3.2 is apt to run out of steam well before it reaches this point.

If you really have valve float, you should feel it. I once had most of my outer valve springs break, and where I had formerly reved to 8,000 rpm, now the motor wouldn't go above 7,600. So I shifted there and finished the race and didn't seem to lose anything, but it was very obvious that the valves were floating. I guess I had enough clearance, as there was no damage to valves or pistons.

The other "redline" is the upshift point - the RPM for each upshift, beyond which you are getting less acceleration than you could if you had shifted earlier (but not too early, if maximum acceleration is what you want).

As a practical matter, this one doesn't matter for street driving - other than the occasional pass on the highway, how often are you at wide open throttle - and upshifting. So it is kind of academic, but for the track you want to maximize acceleration with what you have.

Where to upshift for max accell is determined by your gears, and your torque curve. If you multiply torque at a given RPM by the gear ratio, you have what amounts to thrust at the rear wheels. The ring and pinion and the tire radius are technically part of this, but for the purposes of deciding where the optimum upshift point is, they can be ignored, because they are constant (absolutely for the R&P, basically so for the tire rolling radius), and you aren't calculating a lunar orbit.

You can make a chart of this using graph paper and your torque curve from a dyno. Here it doesn't matter if the dyno was flywheel or rear wheel, the curve is the same for these purposes. I suppose you could use a factory curve if you have a fully stock motor and exhaust.

Porsche calculated this stuff, especially for its race cars, and Pano had an article on it, saved in one of the Upfixin's.

Here is one made using a spread sheet:


http://forums.pelicanparts.com/uploa...1568433753.JPG

This US 1982 3.0 stock motored SC, with a good exhaust with stingers, is best shifted from 2 to 3 at about 6,250, and 3 to 4 and 4 to 5 at 6,000 rpm.
Porsche says redline is 6,700 rpm.

The speeds shown are irrelevant to these shift points (assuming you use your tach), because they use an arbitrary tire size. They will be at the same RPM no matter what tire is used.

It is easy to see that if you shift early, you lose a bit because you now are at a lower thrust than if you had reached the optimum point. Similarly, if you shift late (stretch it), you lose a bit because you spent some time at a lower thrust before the shift. If you have on track data of sufficient granularity, you can see these effects - if the acceleration number is different between the lower and higher gear at upshift, you missed the optimum.

Note this has nothing to do with peak horsepower directly. The shift point is apt to be above that RPM, but that isn't useful. It is way above max torque.

I'd be careful about thinking the engine is still pulling well, etc., above the optimum shift point. My 8,000 rpm race motor felt like it could keep pulling above that figure, which was where I shifted. Then I put it on the dyno, and 7,600 turned out to be where to upshift (other than stretching it on a track approaching a braking zone).

Very interesting Walt. Thanks for the info!

My point exactly Walt. If you just want to twist your engine high and hear it scream 500 rpm over the redline is probably safe, if not done to regular. Above that is dangerous territory.

If you want max acceleration, racing, autoxing or a stop light drag race, then as Walt's post shows you need to consider torque as well as horsepower.

40 years ago I raced corvettes with my old friend in SCCA. We found out that bigger is not always better, carbs and exhaust. We found that if we could tune and gear for the best pull out of each turn we could get jump out of the turn and hold them off at the end of each straight.

Walt's explanation is really nice and clear.

It could be made just a bit clearer by talking about the power curve instead of the torque curve. The cool thing about thinking in terms of power is that the expression linking power to car acceleration does not require the gear ratios and tire diameter. Then, solving the maximum acceleration problem is simple: at any given car speed, one needs to choose the gear that results in the engine delivering the most power.

As a consequence, it is almost always the case that it is necessary to drive beyond the peak power point in order to have maximum acceleration. The peak torque point of the engine is irrelevant to the choice of gear for max acceleration.

http://forums.pelicanparts.com/uploads10/rolling+road+run+May+19_20071179576551.jpg

Maybe we should use the term performance rather than power. Power equates to HP and torque is the pulling power to move the vehicle weight. You can see where this engine makes around 250 hp but it doesn't do that till near 6000 rpm. Look back at the torque curve and see when you have only 150-200 hp available the torque has an almost steady pull until the horsepower catches up.

Also after the horsepower peaks it drops off rapidly. My experience running on a dyno with 3 different cars showed the operator ran the engine wide open until the horsepower peaked, then let it run another 2-300 rpm to confirm the power was indeed going down and then they took their foot off the throttle.

In my message above I've used the word "power" very strictly. It could be measured in hp or, as Porsche likes to do it, in kW.

I have to disagree, HP is derived from torque and rpm, and does indicate the ability to move the car. But what actually moves the car is torque and what you monitor(besides the seat of your pants) is rpm.

To fully understand the performance potential you do need to know torque, + rpm + gearing(including tires) You can gloss over some w/ hp but gearing et al is still critical to full understanding.

You also need to include vehicle all up weight and aero to really see what's going on

here are some examples
say i want to know the performance potential and when to shift a '73 2.7RS. It is spec'd at a nominal 210hp @6300, what does that tell you? Suppose you want to compare it to say a 993RS speced 300hp @6500

if you know the torque curve, gearing weight & aero you can predict the acceleration in each gear, you can also see that the 2.7 will be as fast or faster up to ~60mph where the edge shifts to the 993
2.7RS
http://forums.pelicanparts.com/uploa...1568487019.jpg

993RS
http://forums.pelicanparts.com/uploa...1568487019.jpg

Even easier to the 2.7 will out accelerate the 993 in 1 & 2 but lose from there up.
http://forums.pelicanparts.com/uploa...1568488129.jpg


Looking closely at the 2.7 curve in 4 & 5 w/ & w/o aero you can see where to shift for best acceleration
http://forums.pelicanparts.com/uploa...1568487019.jpg

I wonder what Dr. Max Thrust would think of this discussion