Imbalanced Flywheel 6

[stragenism]

Apologies for asking a possibly common flywheel question: my car forums of choice are vastly under-qualified to help. My question is twofold: How likely is it that a 23 lb (mild steel?) flywheel turned on a lathe (improper tool) or ground (proper tool) would become out of balance significantly enough to cause the entire driveline to vibrate? How much imbalance would be necessary to cause this situation?

The car is a 2.5L H4 AWD (Subaru WRX).

I recently had my clutch changed at a shop which sublet the flywheel resurfacing to another shop of unknowable quality. The service manager inadvertently informed me that no shops in the area are capable of balancing flywheels of "that size." I assume they must have ground the surface because the clutch does not slip.

I've done a great deal of googling trying to understand the issue better. I don't have access to a vibrations text, only my notes from the class. So I have the equations of motion, but lack the experimental capability to characterize the constants. Ultimately, I'd like to use Shigley's to make some prediction on the increased fatigue life of the crank. (I've yet to apply ISO 1940/1 which was suggested in another forum for predicting amplitude of vibration of a pump rotor.)

Thank you for any information!

 

[thruthefence]

I would think a poor set up, (lathe or surface grinder), may lead to this. ie the friction surface not perfectly in plane with the crank.

or assuming a workman-like product (correctly done); possibly a poor installation, such as the flywheel not perfectly seated on the crank, or the clutch pressure plate not perfectly seated (not likely).

Not being 'current' on Subarus, this is a wild guess.

 

[stragenism]

Thanks thruthefence. No possibilities are ruled out, at this point, even after the shop pulled it back down "found nothing wrong" and put it back. Meanwhile, my driveline continues to vibrate, and I believe from the greatly increased clutch chatter that a damping spring has probably broken on the clutch plate.

 

[Tmoose]

thruthefence is correct about the "wedge" grinding being one main suspect.

Here is a picture of a WRX flywheel of unknown vintage.

http://303lurch.com/wordpress/wp-content/uploads/2009/07/t-leg_wrx_flywheel_before_machining.jpg.

note the 3 dowels to locate the clutch cover/pressure plate

i'd ask what what was "checked". If the "nothing wrong" did not include indicating the face of the flywheel relative to the mounting flange, then that must be done. In the pic above there is a hint of a shadow that suggests the surface does not touch the benchtop, which would be the case if the crank flange surface is in a recess. If the machine shop just set such a flywheel on the grinder's table it would not reference the proper surface. If the ring gear is what touches table first, a similar un-true surface would result. An adapter (precise ring) is need to "pick up" the correct surface. An experienced or well trained machinist would not grind it improperly because she would (1) use the right adapter and (2) recognize that if the grinder was taking off a lot of material all on one edge there is a problem.

Rather than let the shop take it apart again, I'd make an effort to indicate the flywheel surface with the starter removed, rotating the crank by hand. there could be fore and aft motion within the crank thrust clearance, but if you applied axial pressure to the crank while turning I'd expect to get a meaningful reading. 0.002 inch is OK, 0.020 is not.

Another way to get in trouble is to install the clutch with the dowels missing. then the entire weight of the pressure plate can be offset a mm or so by the clearance between bolts and clutch.

Here is a pic of a subaru crank and flywheel.

http://www.autopartslib.com/wp-cont...preza-Crankshaft-and-Piston-Parts-Diagram.png

It looks like there is no register in the FW for the crank flange, and that the pilot bearing might extend into the crank to provide flywheel radial centering

 

[stragenism]

Tmoose, thank you for the wealth of considerations! I assume the install shop's technicians lacked the know-how or wherewithal to actually measure anything. From what the service manager indicated, it sounded like they were trying to find fault with the clutch/pressure plate (anything to deny responsibility). The possibility of ACT shipping clutches (which they test) with manufacturing defects is very unlikely. Now I have to consider the possibility that the clutch assembly has been damaged or the diaphragm must be adjusted, which the shop will obviously be unqualified to address.

The most frustrating part is that I have no means to verify there was no error in installation. This is coupled with the fact that I have convinced them on the basis of a poor flywheel job to install a brand new flywheel very soon. If they again install the clutch improperly, or it is indeed the clutch itself, I will have egg on my face. It will be months before I have access to the space and tools I'll need to just do it myself.

There is no end of cases of botched flywheel/clutch installations for Subarus, so I'm not looking forward to it.

 

[thruthefence]

On the Lotus Europa Forum there has been a thread concerning persistent transmission input shaft leaks, & driveline vibration. Investigation, after much theorizing, the second teardown revealed the pilot bearing missing from the crankshaft. Apparently came out with the tramsmission shaft at one time and was "lost" before re assembly. Now this is 'way out there', but;

paraphrasing A. Doyle

(altho he believed in Faries)

"when you have eliminated the impossible, whatever remains, however improbable, must be the truth."

 

[evelrod]

To state the obvious...Poor workmanship, improper installation are what jumps out at me given that the problem did not exist prior to the "clutch/flywheel job" !

First, there is NO right machine to resurface a flywheel. I've had flywheels done on a nice new Blanchard grinder and I've done them on my old Sears lathe. Never a problem for me, but I always re ck the balance (a balance shop was my racing sponsor, duh!). Still, I see your problem as stated above. Too bad. Scary part when they "took it apart and found nothing". What in heck were they looking for?

Second, the fact that the gearbox was out could open a few other possibilities that could cause "vibrations" because of a "different" position upon re assembly.

Rod

 

[stragenism]

thruthefence, I've added that to my list of things to mention to the shop technician. I can see this being possible with these people, now.

Rod, you raise good points. I had read many horror stories of slipping clutches occurring simply turning the flywheel on a lathe to resurface it on subarus, so I assumed it wasn't the ideal tool for the job.

As for what they were looking for, that's a good question. I'm not sure why they agreed to comp the labor and spend 5 or 6 hours working on my car if they had no plan!

I had absolutely considered the possibility of mounts or crossmembers being "bound." At the time, I was able to remove and replace them at factory specs with brand new harder mounts. This didn't solve the problem, it just changed the frequency of vibration to the rest of the car thanks to the increased mounting stiffness, haha. Unfortunately for my trim and dashboard, the harmonics happen right around cruising RPM now.

 

[evelrod]

TTF, I've built dozens of Lotus twincams over the years and never left a pilot bushing out. That would appear to be a good thing as they are about an inch and a half OD and pretty hard to miss. Given that they are out and causing vibration, I think such a vibration would be on a much more intense scale than what the OP is looking at... Beside the fact that every time you actuated the clutch the vibrations would change...Substantially!!! I don't do Lotus engines these days, we had our latest done by Dave Vegher as he is an old friend and more up to date on the latest and I'm waaaaay to old to start doing things differently...:-(

Rod

 

[stragenism]

I doubt that could be the problem then, as the vibrations are consistent regardless of clutch engagement... they're dependent only on engine RPM, it seems. That is, revving in neutral, clutch disengaged, or driving in any gear produces the same given amplitude at a given RPM. This is what initially lead me to consider the flywheel.

 

[evelrod]

Well then, there you have it!

New flywheel clutch or, at least, have them balanced by a REPUTABLE SHOP!

Rod

 

[thruthefence]

Rod, the car in question was Renault powered, and apparently had been a basket case with several owners. (many Europas are, including mine, but the baskets are thankfully much reduced in number. Hope I live long enough).

I can see how a part could be misplaced, or even missed during build up, but How in the world did they ever stab the transmission without aligning the disc with the (missing) pilot bearing? (and it is a ball bearing in the Renault.)

Just blind luck I guess.

 

[gruntguru]

Something you can check very quickly.
1. Does the vibration occur if you bring the engine to the problem revs with the trans in neutral? If not your problem is elsewhere.
2. If it does, put your foot on the clutch and select a gear (this will stop the friction plate from turning) then bring the engine to the problem revs. If the vibration is still there, the problem is in the flywheel or pressure plate. If this is the case:
3. Have the mechanic remove the pressure plate and (if possible) start the engine without the pressure plate fitted. If the vibration is still present it must be the flywheel.

 

[thruthefence]

"Have the mechanic remove the pressure plate and (if possible) start the engine without the pressure plate fitted. If the vibration is still present it must be the flywheel. "

I can just imagine the look on the mechanic's face when this is suggested.

Of course, "it all pays the same", I suppose.

IMHO, if this engine/trans must be split again, the whole clutch, pressure plate, and flywheel group needs to go to a competent clutch/driveshaft shop for inspection & balance. Even if it has to be shipped out.

 

[stragenism]

Gruntguru, 1 and 2 are correct as you describe. I already intended to suggest this option to the mechanic, but I don't know as yet what their excuse will be.

Thruthefence, this is not possible for the same reason it was not possible to get the balance on the flywheel checked on my 2nd visit: there are no businesses of which I or the service manager are aware with the requisite equipment in the area, and I don't live there. It's my daily driver, so I can't leave it there for weeks waiting for turnaround on my drivetrain. It's a bad situation.

In the event that I am able to make arrangements, perhaps a rental with discount, what companies might you suggest for this service? Or, alternately, where would I go to search for one? The garage is in Mobile, AL.

Thanks!

 

[thruthefence]

Here's one, (apparently successful enough to open another location):

http://clutchandpowertrain.com/

they are in Mobile, AL

It might be worth a call. They do hot rod stuff, so I would think they could handle a Suburu.

(No financial interest, ect)

 

[evelrod]

EXCUSE ME? Remove the pressure plate and start the engine?

Well, I guess none of you have tried that little bit of nonsense. I'm not that smart and I tried it, ONCE !!! Way back when...When I was experiencing some difficulty or another with the clutch/flywheel. NOT A GOOD IDEA! Especially when the ring gear flies off because it was, primarily, secured to the Al flywheel by the six 5/16" cap screws that mounted the pressure plate !!!!!!

Keep that in mind. It was not a pleasant lesson, but a REALLY good and lasting one.

Rod :-(

 

[Tmoose]

I'm sorry this project has gotten so complicated for you.

I think the flywheel's axial (face) runout needs to be measured.
If the unbalance is the result of poor machining it should be corrected before balancing. And, If the runout is less than 0.002 inch or so, flywheel balance is not likely the issue.

If it was mine and especially if I was in a jam I'd rig up the flywheel supported on the ball bearing pilot bearing with a bolt in a vice expecting I could detect gross imbalance of the flywheel, and then clutch assembly . Might have to remove the seals and dilute the grease for greater sensitivity. That would be a good setup to check flywheel face runout too.

"if I can't measure it, how am I going to fix it?"

An internet search for engine balancing came up with this -
Rick's Machine Shop
3210 Dauphin Island Pkwy
Mobile, AL 36605
251-479-1501

Yellow pages topics like "automobile racing and sports car equipment" "engine balancing" "balancing and vibration analysis" are likely to uncover a few sources. I expect any engine rebuilding shop would either have a balance machine, or know who does. Many shops have "universal" arbors that would be able to handle your Subaru flywheel after the pilot bearing is removed.

 

[stragenism]

Thank you for all of your advice! I'll call up these shops and see if I can find another. I just took the service manager's word (back when I still had some trust in the garage) when she said she called around.

And Rod, I'm glad you learned that lesson for all of the rest of us!

 

[thruthefence]

When the "took it apart & found nothing wrong" assessment was made, to what level was the dis assembly done? Was the flywheel actually removed from the crank & examined, or did someone just say "looks good to me,close 'er up"?

I ask, because I have a collet chuck on my lathe, with quick change cam lock attachment. When I bought it, I spent considerable time 'indicating' the chuck so it ran perfectly true to the spindle, .001 runout; pretty close for a redneck & a Chinese knockoff chuck. (the lathe itself is quite accurate, a very low time Austrian built Emco-Maier)

when making some parts for a project, I noticed severe runout on the work piece. The runout was .008" or so, three inches out from the business end of the chuck. unacceptable!!

Investigation showed I had missed a small piece of swarf, (trapped between the chuck & spindle) when I swapped the three jaw for the collet chuck. You could hardly see it, it was so small.

my point is, that if a piece of trash is under the flywheel, it is possible to have quite a bit of runout at the clutch surface.