I think you’re missing the principle of how a taper works.

It MUST be tight fitting. It should be so tight that it requires a pulley puller, some heat, and a few brass hammer bangs to remove the hub after it has been installed.

This

Though I generally don’t want to be hammering / impacting main bearings you get the point.

It should be a a bear to remove it after a proper install with well matched tapers.



After the install, the bolt is doing very little to hold the joint together.

That key looks like a specially designed offset key, not a damaged stocker.
That's weird!
Normally special keys like that are used to offset cam timing and not offset the hub alone...

For the reason why all of your crank hardware appears to loosen over time.

This is a bit of a guess on my part but it's worth thinking about.

The Type 4 engine has no vibration damper on it.

In its stock form it probably works fine without a damper.

Each power stroke applies individual power "spikes" into the crank. As these spikes increase in intensity and cycle rate this hammering can induce resonance into the crank assembly.

Making the rotating mass heavier reduces this tendency. All of the dual mass flywheel that are in use today are stupid heavy.

Slowing the rotating mass down helps combat this as well but that's not "sports car".

Adding some type of elastomer in an external pully is how these issues have been delt with in the past by most manufactures.

A complete BS "power upgrade" for many cars is a solid crank hub and lightweight smaller diameter drive pully. Sometimes these were listed as a "power pully."

What you get with one of the puppies is accessories that don't work worth a shit at idle and excessive crankshaft resonance.
You end up with broken locating keys, torn up crankshafts, loose crank bolts, and crank driven oil pump explosions (this is very common!!!).

I REMOVE "power pullies" whenever I find one...
A stock balancer is vastly superior to one of these.
For street cars I try to use a "Fluid Damper" if it's available for that engine.
Fluid Damper is awesome at eliminating harmonics and resonance below 6500-7500 rpm.
For an engine constantly operating at or above those numbers I prefer a tunable elastomer balancer (you can get different O-rings for these) or a specific balancer for high rpm use. It will be listed as a "tuned" balancer and it should spec out an RPM range where it is effective.

The T4 has no harmonic balancer anyway.

Most performance oriented T4s are using much lighter pistons and rods and being spun to higher engine speeds.
It looks like most decently built T4 have at least 50% more power than stock and some double the stock power levels.

I agree with Jeff.

I'll be spending extra time on the crank snout and pully interface.
30lbs sounds good too.
New hardware (maybe upgraded) should be incorporated.
Medically clean threads and blue Loctite will be used.
Once a year inspections (maybe more often?) will be done.

My thoughts...
Maybe wrong?

And now SuperHawk will produce a white paper showing something else...

SH is right about the crank bolt not being the reason stuffs loosening up...

I got a core 2.0 from 914Sixer.
It was complete stock engine in pretty good shape.

During disassembly the crank bolt removed easily.

The crank hub not so much.

I pimp-juiced it and fitted a harmonic balancer puller.

Went tighter on that puller than I normally like too.

Long 1/2" breaker bar tight

Tapping on it with a hammer didn't do shit.

It stayed "torqued up" for a couple of days.

Impacting the flywheel bolts off (at the other end of the crank) popped that bitch right off.

The taper connection was TIGHT!

Horizontally opposed engine has nearly perfectly balanced 1st and 2nd order harmonics. Therefore doesn’t need or get a harmonic balancer.

And that my friend is why I love air cooled horizontally opposed engines. Simple, light, and smooth running vs inline 4, V6, and V8. I would die for a V12 but fortunately Tygaboy hasn’t fit one in a 914 . . . Yet.

If fitting correctly a tapered connection should not come apart easily. Generally a bit of heat is required to open it up a bit.

If this connection does come loose when in service, is there a telltale noise or other symptom us novices could heed? I know this winter I will be checking mine.

OK Master...

Why are SO many high-performance balancers available for the Subaru 4 cylinders?

There's a bunch of different Fluid dampers for them as well.

I cannot remember a Subie with a solid balancer as stock equipment ever.
I've been repairing Subies since the early 80s.
Seen WAY TOO MANY.



I didn't say I was right...

I'm trying to help determine why one member (a freaking knowledgeable one!) has 3 of the same problem.

I'm in agreement with Jeff.

3 is NOT a fluke.

SH, what do you think is the reason for Jeff seeing as many of these as he has?

I will put my $5. on they were never properly fit, checked for a high percentage of contact and torqued correctly.

Yes, but I suggest by then it's too late.

Either end of last driving season (October?) or beginning of this year's (April?) I began to hear a knocking noise at idle, frequency changing with the RPM, a lot like a light rod knock. It was coming from the fan end of the engine. Car drove OK, and the noise went away off-idle. I ignored it.

I suspect this was the key and keyway getting beat to s**t. If I hadn't been such an idiot I would not have ignored it.

However, it was very soon after that, like maybe 2 hours' driving time, that the engine began to run like crap and thus post #1. So I may had staved off some damage by investigating the noise sooner but I probably would have had to rework the crank anyway.

I do admit to giving the fan/hub a light tap-tap with my small ball peen hammer prior to tightening it down, both cars.

GA

Greg, sorry to hear about the layoff. After reading all the posts, I think I am going to remove the fan hub bolt, loc tight it and re-torq it for peace of mind.

I’m no expert expert on Subaru’s for sure but I’ll give you my speculation and what came up with after reading a couple of Subaru SAE papers.


Subaru does indeed call it a harmonic balancer but they don’t say what frequency they are tuning it for. I did find a waterfall plot of their 1st order and 2nd order frequencies and indeed they are much lower than typical 2nd order harmonics from an inline 4 cylinder as would be expected. There does seem the be a weird set of vibrations at 1.5 engine order.

Click to view attachment

This leads to two points of speculation as to why they have it.

1) It is likely tuned for other engine NVH harmonics coming from A/C, power steering, or something else that would be objectionable to customers? A/C would be the prime suspect since all compressors impart objectionable vibrations on the engine especially at idle. Could these be the 1.5x engine order vibrations that show up in their NVH waterfall plot?

2) the Subaru crankshaft is longer than a VW T4 crank. The Subi is a 5 bearing crank vs the T4 which is effectively a 3 bearing crank (4th nose bearing doesn’t support much). The T4 crank has much thicker webs between journals and it may be a stiffer crank.

So the Subi may have some crankshaft torsional vibration issues as you were alluding to. They also spin the crank to higher RPMs than the T4 and forces go up as the square of crank RPM. Maybe they are using the harmonic damper to damp crankshaft torsional vibrations that the T4 isn’t subject to.

I know VW and Porsche air cooled engines far better than Subaru’s and the only thing I can say is that none of the VW air cooled engines nor the early Porsche 6 cylinders used a harmonic balancer. And the Porsche six cylinder crank is easily as long or longer than a Subi four cylinder. Porsche six’s were similar or higher power output per liter than Subi (naturally aspirated) so it seems odd to me that they would need the harmonic balancer due to crankshaft torsional vibrations related to durability, but the devil is in the details and I certainly don’t have all the details on Subi’s. Perhaps the Subi crank is not as stiff as the Porsche six or VW T4?

@technicalninja

As you elegantly stated . . . I may not be right. I’m always open to be schooled and I’ll keep digging to see if I can better understand why Subaru is using a harmonic damper and what frequency they have it tuned for.

duplicate

You BEASTS!

Just throw Greg under the bus will ya!

Greg, always remember, I'll have your back!

The hawk and the cabman are subversive...

Now for serious:

SH- I'd be a pissed off engineer if the NVH resonance graphs were taken with ANY accessories running at all.

Subaru has almost exclusively used either vane style compressors (exactly like a mechanical air pump-just sealed up better) or scroll style compressors.
Both designs do not have reciprocating pistons and are super smooth in operation.

I may use a small scroll for a retrofit in a 914 mainly for the smoothness and lack of cycle torque change. If you are forced to use a cycling compressor the scroll design (which is supposed to be the most efficient) has almost zero engagement torque.
It's damn hard to tell when it engages. Often no idle compensation is needed.
I've kept 3 baby Subie scrolls (dead ones) as test fitters.

Subie compressors are probably not a source of serious NVH noise...

I do have a question. You mentioned EARLY 911 motors did not have harmonic balancers. Does this mean later ones do?
If so, when did they arrive?
What other changes occurred along with the balancer?

Balancer is really not what this device should be called.
Vibration shunt or eliminator sounds better in my book.


Ya'll be easy on Greg now...

Not trying to throw anyone under the bus so to speak.

Checking torque after the fact with a typical clicker wrench isn’t necessarily an indication that the hub is loose or that it wasn’t torqued properly on assembly. I fully believe that the race car engine in particular was built properly.

I don’t want to go into a deep dive on how fastener toque specs and subsequent audit torque checks are established but it will suffice to say that we would need way more than a sample size of two or three fasteners.

Likewise simply jacking up the torque may or may not help. If the fastener torque spec is close to yield, increasing it arbitrarily can take it over yield and there will certainly be some loss of clamp load if the fastener wasn’t designed for a torque to yield application. I don’t think that is the case here but I don’t have the fastener data to know one way or the other since I don’t know what grade the hub fastener was as OEM or if someone swapped it out for something lesser along the way (doubt it).

At this point I’m deep in the weeds and simply hoping that the fix holds up for the long haul.

Not sure when later engines got a harmonic damper and/or which variants get it. Modern build complexity is beyond belief. What engine variant, PDK vs manual, dual mass flywheels vs single mass light weight flywheels, 993, 996, 997, etc. Late model 911 stuff isn’t my thing.

The proper engineering term is tuned mass damper.

@technicalninja

Greg, when I first saw the damage on the crank snout, I thought it was trashed and un-usable.

Your repair appears to be excellent!

I thought "WOW did that clean up nice!" and I'd expect what you massaged to work fine for years.

Really nice repair work there IMO...


I'm surprised that no-one has mentioned a sleeve retaining liquid.

I've used this on non-tapered crank snouts to hold on loose crank gears when the crank should have been replaced. This is a "Hail Mary", last ditch, southern engineering fix...

https://www.amazon.com/Retaining-Compound-C...096406&th=1

I've never had a "glued" together assembly come apart, but I wouldn't warranty that for 10 minutes.

You might not be able to get a tapered joint to separate after using this stuff.
Fire would probably be a requirement.

@technicalninja

Found some interesting data for the FA engine

Click to view attachment

Harmonic damper is tuned for 2nd order - for about a 150hz resonance.

I’ve overlaid in red what the crankshaft resonance would probably look like without a damper.

When a tuned mass damper (harmonic damper) is used it cannot reduce the energy input; it can only redistribute the energy to other frequencies. When it does this it usually creates two side bands - one above and one below the resonant peak. I’ve artfully illustrated this in green.

These plots appear to have been created from data supplied by Fluiddampr so there is the comparison of OEM Harmonic Damper to their product.

While you can see they (Fluiddampr) knock the resonance down further, they do this by pushing the energy out to higher frequency (RPM) range. This would not be a good thing for extended racing / track day use where the engine would spend more time at those higher RPMs.

The bottom line: FA motor making a whole lot more HP and Torque than our measly T4s and it would appear that perhaps the Subaru cranks has some propensity for a 150Hz resonance.

Another interesting dilemma is that just because it has this resonant frequency doesn’t mean that it is a durability issue. It just may be that Subaru is using the damper to calm NVH and make the car more enjoyable to drive. Without access to Subaru durability or warranty data we can’t know for sure which reason it is that Subaru is using the harmonic damper.

Obviously, we would need similar data for the VW T4 to draw any conclusions about whether the T4 would benefit from a harmonic damper and of course that data would be needed to tune it appropriately.

Source:
https://www.ft86club.com/forums/showthread.php?t=139127


PS: Thanks to you I’ve learned a little more about subi’s than I knew a few days ago

Once again, your tenacity is impressive!

Thanks for the time you spent researching this.

I need to think up some un-answerable question for you...

@technicalninja



Also keep in mind, that the 2nd order harmonics of the boxer engine are already quite low as compared to an in-line 4 per the data in post #61. The more I think about that, the more likely it is that Subaru is using the harmonic damper for NVH reasons but we’ll never know since I don’t know anyone that works for Subaru to get any insight.

@Superhawk996

I didn't intend to insinuate that you were wrong.

I was referring to me!

You are an absolutely top-notch source for knowledge.

You will go through the trouble of looking stuff up like no-one else I know.

I really appreciate your input.



On the loosening up of the crank bolt BS.

In 40 years of building stuff (very limited VW and Porsche) I've found as you improve horsepower you commonly need more vibration attenuation than less.

In the Miata world the stock balancer is SO bad many tuners want an aftermarket balancer BEFORE a tuning session. The stock balancer is not good enough for double or triple the horsepower.
Now, I'm tripling power on these, most of the T4 work on here doesn't triple the power so the need for increased balancing will be less.

I've seen multiple threads in the short time I've been on here regarding crank noses being damaged. Many of these could be from improper mounting but there are enough of them that makes me think "maybe there's more to this than we realize".

I like my street-based vehicles to have excessive damping.
I'm also a believer in full weight flywheels to increase transmission life...

Thanks again for a scientific look at the balancing requirements for an H layout engine.
It has been informative!

Good luck with that, Mr SH is most often right on target and we should all be thankful for his input.

I think part of that is just due to your short time on the site.

Like you - I’m new too, having only been here since late 2018.

There have been two instances of crank nose loosening that got a lot of traffic in the last 9 months or so which makes it seem like it occurring often. But honestly, I don’t think I’ve seen more than 3 or four instances since 2019. Have there been more? Certainly. But, probably not a regular occurrence by any stretch of the imagination.

Like so many things, it’s partly an issue of people not understanding how things are designed to work.

What do I mean by that?

In the case of a taper, people simply think tightening the joint to spec is good enough. But that is 100% contingent on having a taper that is clean, dry, and has good interference fit between the two parts as they would have been when fresh back in 1970s when taken out of the dunnage for the engines initial build.

I think as parts get older and more beat up, we will see this issue more often. Especially with the home built engines where good intentions may not be enough to make up for a lack of attention to detail (not aimed at anyone in particular - especially OP).

For some reason, lots of people don’t have a real appreciation for how precise fits in an engine are. Likewise many don’t have an appreciation for how easily steel can be damaged.

I’ve worked in engine plants and I can tell you horror stories - mostly related to new hires that don’t have an appreciation for the precision machining that parts have and how they need to be protected from damage. There is a lot of work and training done to impart knowledge of how important it is to handle parts with care, yet you still see instances where someone isn’t thinking and damage can happen.

To some extent we all have to fight the urge to just get a job done and we need to focus on doing it right. I know I’m guilty of this sometimes.

Anyway, just a point of view for consideration about how common this issue may or may not be.

Here’s the thing. The Miata harmonic damper is PERFECTLY adequate for a stock engine - which is what Mazda designed it for.

It’s not fair to say the damper is inadequate for an engine that has been modified for 2x or 3x the horsepower. Many things in the engine are going to be inadequate at that point.

I had a stock 91’ Miata that was regularly autocrossed and ran weekend track days. That Miata went to 228k before I finally had to let it go due to rust that was getting too dicey. No issues what so ever with the stock harmonic damper or engine durability. I gave it away with the engine still running strongly.

The video below is from my car - used to collect a bet for a case of beer that the engine wouldn’t last to 100k without a rebuild the way I was driving it. I easily won the 100k bet at which point he wanted to go 200k - double or nothing.

The beer was great


https://www.youtube.com/watch?v=6-m2IT4Xojo?si=jJjnsojGL3tlJ9w9

No offense taken at all nor did I think that was what you implied. I certainly don’t mind being called out when i am wrong. That’s how we learn new things !

@technicalninja

In stock form I'd have bet 250K...

Bet your engine is NOT what took the 91 down and 91s still had the short nose crank from the original 323 motor.

I had one of the earliest 323s with a crankshaft failure.

I wrote the "service procedure" for Mazda as they had "remove engine from car" as the first step and if it's an automatic you could change a crank in-chassis with a couple of LONG bolts to push the auto trans back far enough to clear the flexplate.
You don't disconnect anything!

Shaved over 4 hours off of the job...


Mazda blew the doors off of all the previous iterations of the small 2 place roadsters that came before it.

The only one that was close was the Fiat Spider and the Miata body-slammed that...

It was the first RELAIBLE sports car IMO.

I was having a dinner conversation with a couple buddies last night about this problem, how the tapers came loose and damaged the crankshaft.

One buddy - a Chevy LS guy - started making fun of old Germans with wonky engineering ideas and why didn't they just do like 'Merica and put a press fit and key there.

The other buddy drives a late-model BMW E-something (M4? I don't know those cars) just a few years old. He noted that the BMW line is still using that taper fit for driving the cam belts, without a keyway, and is having problem with guys modding their engines to the point where when suddenly downshifting the crank hubs are spinning and destroying multi-$10Kengines. The aftermarket "fix" is apparently to drill the crankshaft for a pin/key to stop that.

And that got my brain to remembering when I did my camshafts belt on my '00 Audi S4 at 100k miles, some 14 years ago. I remember that job with dread. The cam pulleys were taper-fit only, no drive keys. You'd set the engine to TDC and remove the tensioners and the belt, then loosen the bolts holding on the gears; you'd put the new belt and idler pulleys on and then setup this elaborate (rented) alignment tool with bars and pins that would locate all the engine drive components - crankshaft, camshafts, whatever was in there - at a specific rotation. And then you'd tighten the pulleys down to spec.

I remember thinking at the time that my entire investment in that car was wholly dependent on a bunch of taper-fit pulleys being properly secured by bolts...and if anything slipped I'd be funked...but nothing ever did.

Amazing how thoughts like that get buried away in your head, only to be exhumed at interesting times...

So ze Germans are still onboard with this taper fit thing even today. - GA


Click to view attachment

Exactly why myself and many more just love the simplicity of our 1/2+ century old sporting machines.

The later model Audi with the direct injected V6 changed to chains!

They moved them to the back of the engine, and should you need to do anything you get to remove the FREAKING transaxle!
The tensioners and rails are a source of problems with this motor.
Poor oil change intervals dooms these to an early demise.

Sometimes newer is not necessarily better...