1 2
DaewooOfDeath
DaewooOfDeath SuperDork
10/28/24 11:46 p.m.

Thanks guys, this is really informative, and it brings to mind another question.

It seems that the two reasons for wanting a balanced engine are NVH and reliability/rpm capability. In an extreme example, let's say a seven liter 180 V4, the NVH would be a disaster, but it would also shake the engine apart at relatively low rpm as all those weird twisting and bucking forces try to break transmission input shafts and twist crankshafts apart, etc. 

However, there seem to be a lot of performance platforms that have NVH problems without having the reliability/rpm problems. I'm thinking of the Porsche V4 LeMans engines, the weird shared pin 90 degree V6 F1 engines, Audi inline 5 engines, etc. Is this simply a matter of NVH problems starting at a much lower threshold than reliability/rpm problems. Ie, is the point at which a 2 liter I5 starts shaking itself apart only a little lower than the point at which a 2 liter I6 does so, or are they both so high it doesn't matter?

Pete. (l33t FS)
Pete. (l33t FS) MegaDork
10/29/24 6:48 a.m.

I am not sure where you get the idea that an I5 is not necessarily a balanced engine or has inherent issues.

 

For RPM, there are many other things in play, too.  The longer the crankshaft, the more torsional harmonics are a damaging factor.  Take an 034/035 Audi five cylinder over 8000rpm for much and you will wreck oil pumps.  It's on the snout of the crank and torsional vibrations destroy the gears.  BMW sixes have similar limitations, but I am not as familiar with that platform.  Harmonic dampers on the end of the crank, both ends nowadays (dual mass flywheel anyone?), are critical.

 

This is a place where short crankshaft strokes have a nice benefit.  Lower strokes mean the crankshaft is stiffer.

DaewooOfDeath
DaewooOfDeath SuperDork
10/29/24 8:03 a.m.
Pete. (l33t FS) said:

I am not sure where you get the idea that an I5 is not necessarily a balanced engine or has inherent issues.

 

For RPM, there are many other things in play, too.  The longer the crankshaft, the more torsional harmonics are a damaging factor.  Take an 034/035 Audi five cylinder over 8000rpm for much and you will wreck oil pumps.  It's on the snout of the crank and torsional vibrations destroy the gears.  BMW sixes have similar limitations, but I am not as familiar with that platform.  Harmonic dampers on the end of the crank, both ends nowadays (dual mass flywheel anyone?), are critical.

 

This is a place where short crankshaft strokes have a nice benefit.  Lower strokes mean the crankshaft is stiffer.

I5s require balance shafts to cancel out inherent rocking motions assuming a 144 degree even fire, I think. 

I guess an easier way to ask this would be as follows:
1. Rotational mass per cylinder is identical between an I6 (perfect balance) and an I5 (rocking imbalance). Stroke is also identical.
2. Crankshaft length, rigidity, weight, etc is the same. Connecting rods are identically strong.  
3. Oil pumps, harmonic balancers, etc, are of the exact same effectiveness. 

Doesn't the I5's rocking imbalance cause it to fail from rpm before the I6, all else equal?

Pete. (l33t FS)
Pete. (l33t FS) MegaDork
10/29/24 2:41 p.m.

I have literally never seen a five cylinder engine with balance shafts.  My Audi felt fine at 6000 and my much longer stroke Volvo feels great at 6750.

I think you are confusing harmonics with imbalance.  Harmonics is what damages parts.  I blame people calling the thing on the front of the crank a "harmonic balancer" instead of "harmonic damper" smiley

Driven5
Driven5 PowerDork
10/29/24 4:09 p.m.

Inline-5 are imbalanced the same way Inline-3 (and cross-plane I4) are. They technically have 'perfect' primary and secondary balance, but have a primary end-to-end rocking imbalance. Like pretty much any other engine imbalance, they can be designed to be perfectly acceptable without a balance shaft (or shafts) if you keep them within certain thresholds. Most have fallen into this category. The GM Atlas 5-cyl though, at 3.5/3.7 displacement, is notably balance shaft equipped.

If all else is equal, then the primary durability difference would seem to be the I5 having higher (mostly first and last) main bearing loads than the I6. If that's also the weak link in the chain, then the I5 would likely fail sooner. If not, then not.

iansane
iansane SuperDork
10/29/24 4:53 p.m.
Pete. (l33t FS) said:

I have literally never seen a five cylinder engine with balance shafts.  My Audi felt fine at 6000 and my much longer stroke Volvo feels great at 6750.

The newer 5 cyl 07k has a balance shaft. Obviously you know the older MC type stuff doesn't

Pete. (l33t FS)
Pete. (l33t FS) MegaDork
10/29/24 7:45 p.m.

In reply to iansane :

Good to know.  I didn't know that there were different generations of 07K like that.  Time to do some reading!

 

I don't know that I ever saw a 5cyl Colorado.  The frames on them rusted frightfully fast here.  I did some electrical work on one but it was a four cylinder.  And I couldn't rack it because the frame was so bad, so I was crawling around under it after driving up onto ramps.  He didn't care because he only used it for taking the trash out and never really drove it on the street.

 

Regarding bearing wear, anecdotes aren't data but the 2.4l Volvo that I rebuilt a few years ago (six main bearings) had pretty heavy wear on the mains at 2, 4, and 5.  The rod bearings looked like new so I reused them.

DaewooOfDeath
DaewooOfDeath SuperDork
10/29/24 9:45 p.m.
Driven5 said:

Inline-5 are imbalanced the same way Inline-3 (and cross-plane I4) are. They technically have 'perfect' primary and secondary balance, but have a primary end-to-end rocking imbalance. Like pretty much any other engine imbalance, they can be designed to be perfectly acceptable without a balance shaft (or shafts) if you keep them within certain thresholds. Most have fallen into this category. The GM Atlas 5-cyl though, at 3.5/3.7 displacement, is notably balance shaft equipped.

If all else is equal, then the primary durability difference would seem to be the I5 having higher (mostly first and last) main bearing loads than the I6. If that's also the weak link in the chain, then the I5 would likely fail sooner. If not, then not.

It sounds like this would be an edge case and that NVH becomes a problem much sooner than RPM tolerance, though still instructive since the rocking would translate to bearing loads. Would I be right in assume that a 180 degree V4 would actually shake itself apart at low RPM because of imbalances, probably right after it vibrated its unfortunate operator into unconsciousness?

Pete. (l33t FS)
Pete. (l33t FS) MegaDork
10/29/24 10:15 p.m.

In reply to DaewooOfDeath :

Two opposite corners would be at TDC while the other two corners would be at BDC.  The reciprocating mass's CG would be rotating around the middle of the engine twice per revolution.  It'd be entertaining to watch smiley

 

The older VW-looking Subarus (EA engines) were definitely not very smooth despite being boxers.  They had a particular warble to their vibrations that was almost pleasant.  I suspect that these engines had a much higher bank offset than the newer engines in part due to their having only three main bearings, so they probably needed a lot of depth between the opposing cylinders for stiffness mass.  They also were/evolved from pushrod engines, so they may have needed a higher offset for cam lobe placement reasons.

EJ and later engines have five main bearings and the Fx engines have so little cylinder offset that the counterweights look like paper plates.  The newer engines are a LOT smoother than those old EAs and the weird soothing thrummy vibrations are mostly gone.

Driven5
Driven5 PowerDork
10/30/24 12:30 p.m.

In reply to Pete. (l33t FS) :

I think you're right that the move from 3 to 5 mains likely helped smooth the newer H4's out, but rough comparison of pics shows going from a roughly crank pin width strap between opposing crank pins, to a narrow main with two thin crank webs that combined is noticeably wider than the crank pin width. So if anything, as far as I can tell, the cylinder offset looks to have actually increased with the additional mains. 

However, what it has also done is get rid of the crank bending caused by the semi-cantilevered crank pins when not at TDC or BDC. That's the big reason I4, H4, and V8 have gone from 3 to 5 mains and I6 have gone from 4 to 7 mains, and I suspect is the primary reason increasing the number of mains to 'fully' (double shear) support the crank pins would have an NVH impact.

The other interesting thin I noticed about the EA crankshafts, are that they simply don't look to have much of anything for dedicated counterweights. Presumably they're just relying on the equal and opposing (but misaligned) crank pins for balancing forces then. Meanwhile, the newer ones do extend those thin crank webs to be 'fully' counterweighted as well. I believe this would also likely serve to similarly reduce bending forces in the crank.

Pete. (l33t FS)
Pete. (l33t FS) MegaDork
10/30/24 1:10 p.m.

As mentioned a lot of 4s don't even have significant counterweighting.   Or more than three mains.

 

Look at a Model T crank smiley

ShawnG
ShawnG MegaDork
10/30/24 1:52 p.m.

I've driven large displacement 4-cylinders with and without balance shafts. I've ridden motorcycles which have had the balance shafts removed.

I've never felt a significant difference with or without.

paddygarcia
paddygarcia HalfDork
10/30/24 3:06 p.m.
Pete. (l33t FS) said:

As mentioned a lot of 4s don't even have significant counterweighting.   Or more than three mains.

 

Look at a Model T crank smiley

Or an O-360 6 liter boxer 4...

Pete. (l33t FS)
Pete. (l33t FS) MegaDork
10/30/24 7:05 p.m.

I can't find it, but I do have the PDF on one of these computers... it was an engineering paper written by Rolls-Royce or one of the engineers, about the development of the crankshaft on the inline 6 that they used for a few decades, as the engine continually increased in displacement, power output, and RPM capability.  It started out (in the 1920s?) as a simple counterweightless design as was common practice back then.  Then it evolved into something that we would recognize as more appropriate.

 

Neat stuff.  It was a maybe 44 page long PDF.

1 2

You'll need to log in to post.

Our Preferred Partners
2HAXDrGMFBqbDhSXb7INULS5683cKhmjEnRXGj4gG5rBT9sGZ4JEECn6NyHaPAcR