LastMezger

I'm just wondering, given the obsession with overrev reports, if there's any engineering report or evidence that someone can point to that supports this concern?

I mean if the car drives well, passes a PPI, etc...why does it matter?

sandwedge

Main reason it matters is that it can be used as an excuse to deny a warranty claim. A new engine will set you back about $20,000 if they find a good reason to say it's your fault and not theirs.

acao

Just last week, I posed this question to three mechanical engineers. The professor is the chair of the automotive propulsion systems lab at a major university and on retainer by one of the big three. The second is the professor's consultant, former student, and lead engineer of a big three engine that is currently the basis for his former employer's modern muscle car (and truck) offerings. The consultant left Detroit and now works for a startup. The third is the consultant's friend who still works in Detroit for one of the big three.
***
Me: On engines for consumer automobiles, what's the typical safety margin for determining redline and rev limiter? If a manufacturer sets redline at 7k, and limiter at 8k (or 7.5k if that's more realistic), what is the number that makes the engine go kaboom?

Professor: Modern engines can regularly hit 10k and above. But, if you look at an engine's power curve, it drops off dramatically. In your example of a 7k redline, the curve may peak at 6800. The main reason for determining a redline is that there is no reason to keep spinning faster, there's no power. The engine can spin to a point of failure, but it's usually well above redline.

Consultant: It's complicated. It could be a function of the valvetrain or the cranktrain (i.e., crankshaft, con-rods, pistons). From a valvetrain perspective, the valve springs must have enough force to close the valves. The faster you spin the engine, the more spring force you need to overcome the inertia of the moving valve. If you spin too fast, you can get what is (or was) called valve float. Generally, it won't make the engine go "kaboom", but it's still not good. On the crankshaft side, there's a lot of mass accelerating and the bearings need to be sized to react it. I remember looking as the relative contribution of accelerating mass and cylinder pressure on main bearing loads a long time ago. It turned out the the loads from combustion pressure dominated at low engine speeds, while the loads accelerating the masses (pistons, con-rods and crank arms) dominated at higher engine speeds.

Consultant's friend: There is no general rule that establishes the over-speed capabilities of an engine, or which components will fail first. In a perfect world all the components would have the same over-speed capability. Often, there are historical reasons why one component might be the “canary”. I’ve seen where the crankshaft and bearings are overdesigned and the valvetrain is the limit. I’ve also seen where the valvetrain is more capable than the bearings. 20 years ago the engine would start to lose power (Valvetrain going out of control) and it was obvious that things were being pushed too far. Today’s hardware is designed to a higher requirements and the power won’t drop off like it used to.

The rpm limits that are put into the engine calibration is generally 200 to 400 hp above the max “shift speed” of the transmission. The engines are designed to sustain that condition for many hours. I’m not aware of OEM’s testing much above those speeds since they are regulated by the engine controller, and therefore are not part of the durability schedules. Going above these speeds is uncharted territory; an engine might be able to run above the max rpm for a while, but then again, it might not! My recommendation to anyone attempting this is to make sure they have new, high viscosity (40 weight) oil in the engine, and keep it cool ( below 280F) it will go a long way towards protecting the bearings.
***
So there you have it. Implicit in this discussion is that the overrev occurs on orderly acceleration. It did not not seem to cover an overrev caused by a 5-2 shift.

simsgw

In a reversal of the usual case where I'm the one offering the detailed engineering explanation, let me simplify that one ACAO offered:

• More rpm creates more force.
• Things break with enough force. All things, except neutrons I suppose.
• Before they break, they bend. (As a metaphor for a dozen types of pre-yield distortions.)

Go ahead. Ignore the rev reports, but don't be shocked when you find out some part of that new-to-you car is bent.

Gary

avader906

over-rev reports are there - on manual cars - to check if the engine+transmission had suffered "driver error" from incorrect DOWNSHIFTS. Ie, from 110mph in 5th to 2nd (instead of 4th). That one would be severe and over-rev the engine proper - likely inducing a spin and breaking a few things. However range 1-3 over-revs are usually there because the car had bounced on a rev limiter a few times, has little if none mechanical implication.

Macster

The problem is that the car may drive well, given a short test drive that so many buyers rely upon. But after an extended test drive the engine controller may register misfires. These in the case of overrevs may arise from one or more bent valves that are not sealing properly. This will only get worse. The valve will burn or even eventually the head will separate from the stem.

IIRC there are restraints upon what overrevs an engine can have logged and still qualify for a PPI.

Additionally the presence of a lot (whatever that is) overrevs even if they do not extend into the danger ranges give some indication as to how the car was driven/treated by its previous owner(s).

Redline is arrived at by the engine's ability to flow air at higher speeds. Reving the engine beyond this gains one nothing but a higher stressed engine.

There are real margins/variations between for instance valve springs that may have one engine able to tolerate considerable overrev conditions while another otherwise identical engine may not tolerate those so well.

But the engine maker has to consider the worst case not the best case.

Side note: Years ago I came across a report of a Carrera GT engine failing during a test drive by a car mag writer. He misses a shift and the engine blew up. Post engine blow up tear down/analysis found the engine speed reached around 14K rpms before something let go. These seems like a lot of margin -- given the red line is IIRC around 8K for this engine -- but remember the redline doesn't drop with miles. That 8K redline is there at 1 mile, at 1000 miles, at 10,000 miles and at 100K miles.

Sincerely,

Macster.

LastMezger

Wow guys...great answers! Thanks.

It's not related to buying a car or anything...just curious.

alexstjo

This might be a stupid question but can you over rev a PDK?

No HTwo O

Given that there are hundreds of Porsche for sale at any given time, why not do your due diligence, and buy one that has:

1) The least number of over-revs in range 1-2
3) Hopefully none in rage 3 (grey area)
3) And absolutely NONE in rage 4-6

LastMezger

It won't let you.

I'm not in the market. Was just wondering if there was any actual evidence to suggest it matters.

sin911

I blew my old 997.1 C2S's engine at Sebring. Nothing past Range 4 (I think it is around 8,200 RPMs), but apparently that was enough to blow it up... It was due to mis-shifting, I am not denying it. I didn't think at the moment it was enough to bend and then brake a connecting rod and cause it to shoot through the case.

The rev limiter is there to protect the engine and the components, as mentioned above more rev means more force and along with it more wear and tear.

No HTwo O

I bet you were well above 8,200 RPM. The gauge is not as fast as the actual occurance, it lags behind.

FYI.

simsgw

I think he was speaking of over-rev's recorded by the DME. No engine can be faster than that.

Gary

ralt12

What is the date on the Motor Over-Rev Ignition Range chart (i.e., is it the same for 997.1 and 997.2??)?

No HTwo O

It's a few years old. Probably only for the 997.1. Does a 997.2 have a higher redline than a 997.1? If not, the chart may still be accurate for both.