TDI swap vibrations in Volvo destroys dual mass flywheels

[cj.surr]

I have had an AHU TDI swapped into a Volvo 240 for about 3 years now. I have tried several flywheel/clutch/transmission combinations without any great results. The main source of the problems seem to be a very strong harmonic resonation/vibration in the 1850-2150 rpm range (as well as lower RPM). It gets worse as load is increased, but is still noticeable on decel. It does not occur when free-revving the engine. It gets worse with increasing gear, with 5th gear being the worst by far. It is not at all a normal diesel vibration, it tries to shake the car apart, and you would not be able to hear the passenger talking at its worst point.

I'll go through the sequence of attempts.

Volvo M47 5 speed transmission with custom adapter (To be fair, the transmission pilot alignment was probably not within spec for this transmission)
Single mass flywheel + sprung hub clutch - Vibration is very pronounced
1.8t Dual mass flywheel + solid hub clutch- Vibration is mostly subdued (still noticeable), but dual mass flywheels do not last. All 3 (new) flywheels that I tried failed in less than 3k miles of gentle driving.

Toyota W56 transmission with DCS adapter plate (I chose this transmission because people have had good luck with it in Toyota trucks)
Single mass flywheel + Sprung hub clutch - Vibration is still very pronounced, no change from Volvo transmission

Tremec T5 transmission with custom adapter bellhousing (I checked transmission pilot alignment and it was within spec)
Euro market Sachs Dual mass for AHU 1.9t Audi A4 + solid hub clutch
- Vibrations similar to 1.8t dual mass. This flywheel has pretty much failed after 100 miles of somewhat hard testing/driving (spending a lot of time in the RPM trouble area)

This last attempt is really what has me baffled, since this dual mass was specified for my exact engine as it came in an A4. I feel that it must be an engine issue that is killing the flywheels since it is the only common denominator. Compression test looked fine. The engine has been running fine and I have put about 60k miles on this swap (mostly with single mass flywheels and avoiding the RPM zone that has issues). It has been getting good mileage and driveability is otherwise fine. I don't think the vibration has gotten worse at all over the course of 60k miles. The engine drove fine and smooth in the donor car before swapping it.

Compression test is fine
Cam looks fine

Other parts I have tried replacing without change:
Injectors
Driveshaft
Harmonic balancer (also tried fluid damper)
Engine mounts (currently using BMW M42 mounts, but I have tried a mk3 tdi rear mount without much change)

I have a spare very high mileage engine that I am considering swapping in just to try, but I have little hope for that effort. I am wondering if anyone has any ideas at this point, because I’m all out. Thanks for reading.


Here is a video of the excessive axial play on the A4 dual mass after 100 miles
https://www.youtube.com/watch?v=FRdsGXdXimU&feature=youtu.be

 

[oilhammer]

If the geometry is wrong, it'll vibrate.

The thing you mention is that it is under load, but at no load in neutral it is fine. That tells me it is a driveline out of alignment issue, and not the engine itself.

The 240 has a live rear axle, but with a carrier bearing on the driveshaft if I recall? Is the "plunge" function of the driveshaft (necessary with a floating live rear axle) working properly?

 

[cj.surr]

If the geometry is wrong, it'll vibrate.
The thing you mention is that it is under load, but at no load in neutral it is fine. That tells me it is a driveline out of alignment issue, and not the engine itself.
The 240 has a live rear axle, but with a carrier bearing on the driveshaft if I recall? Is the "plunge" function of the driveshaft (necessary with a floating live rear axle) working properly?

Driveline angles are all proper, transmission output is parallel with diff flange. I've had 3 different two piece driveshafts (one for each trans), and also tried a single piece with the T5. No plunge issues and the vibration has been there all the same.

 

[oilhammer]

So the rear axle itself has never been messed with? That is pretty much all that is left sounds like.

If you lift the car up in the back, and set the rear axle securely on jack stands, so that the springs are loaded the same as they would be with the wheels on the ground, can you get the vibration to happen? This would of course be a no-load situation, but at least the rotational bits would be the same.

You say it does NOT happen with the engine running, trans in neutral, clutch pedal released, correct? So that should rule out the engine, the flywheel, the transmission's input shaft. Since all that stuff is spinning then.

I would think if it shakes that bad, you'd be able to SEE something with the rear wheels in the air spinning. I know we had a Ford E-van once that had an out of balance driveshaft, and it started shaking so bad that the yoke on the pinion in back would wobble so much it would become a blur... the whole axle was loping up and down violently.

Another option would be a roller dyno, but it may not be very easy to see under the car with one of those, as they usually do not offer much clearance underneath.

 

[cj.surr]

Yes, it can be replicated with the rear axle on jack stands. I don't think smooth free-rev rules out an engine issue. It doesn't vibrate significantly with trailing throttle either, it only gets much worse with increasing engine load.

The rear axle has all new bushings everywhere. It's true that it will shake (as observed with the car in the air) when the vibration is at it's most extreme state. But it seems to be more of an effect than a cause. Like I said, a dual mass smooths things out most of the way. But yet, fails extremely quickly, which tells me the problem is on the engine side of things. What else could cause the dual mass to fail, assuming good transmission alignment?

 

[turbobrick240]

If the alignment is good and the rear axle is good, I'd check the center support bearing(assuming it's still there). I replaced a number of those on my 240's. When they go south the vibrations are horrendous. You might want to try balancing the driveshaft also.

 

[Tin man.]

I would check your centre bearing or propshaft rubber donut it might be giving the spring's in the dmf a hard time.

 

[cj.surr]

Yeah, everything on the two piece driveshaft had been serviced, currently running a single piece with the T5.

 

[Andyinchville1]

HI,

Verify that the driveshaft itself is balanced (no balance weights fell off etc..), the drive shafts are true (not bent) and in proper alignment (i.e. angles from tranny out put to rear end input are correct).

Then about the only thing left to check is that the U-Joints are properly phased if it is a multi piece driveshaft (I think It was mentioned earlier but also make sure the center bearing if present is securely mounted and in good condition also) ..

Hope this helps.

Andrew

 

[casioqv]

I would suspect a problem in the design of your engine mount system... these engines are very unbalanced and shake around *a lot* in an OEM VW, whereas the Volvo engines are all comparatively smooth. Could it be too restrained, and therefore vibrating and twisting the entire engine/transmission in weird directions?

 

[d24tdi]

I think you are looking at the right area focusing on the engine. It would be hard for an issue like driveshaft alignment or rear axle bushings to transmit enough force backwards, through the gearbox and then through the clutch disk all the way to the flywheel, to cause such an intense vibration or tear up a DMF. I think if anything downstream of the trans were causing it, you'd also be breaking U-joints, having differential pinion bearing problems, etc. And I doubt anything about how the engine/trans are mounted could transmit enough force back into the rotating components to cause failures, so I'd rule out anything having to do with the mounting system.

Plus, you have had the same issue despite running three completely different transmission and driveshaft setups and a variety of different mounts. And you say the vibrations are abated when a DMF is used, but then the DMF fails soon. That tells me the DMF is doing a lot of work trying to smooth out torsional pulses, and helping, but that the forces are too strong. I think the motor is the only thing that could cause the kinds of symptoms and damage you're describing. Amplitude of the positive/negative crank torque is too high, for whatever reason, and that's causing the bad vibes with the SMFs or beating your DMFs to death.

What kind of tune is it running? If it's not the stock tune, what kind of SOI timing is it dialing up in the RPM range where the vibration is worst?

When you say compression was fine, what kind of readings did you see? Were they in the middle of acceptable range, or off the high end of the scale?

Know anything about the motor's history? Head ever been off it? What thickness HG does it have?

 

[oilhammer]

I would suspect a problem in the design of your engine mount system... these engines are very unbalanced and shake around *a lot* in an OEM VW, whereas the Volvo engines are all comparatively smooth. Could it be too restrained, and therefore vibrating and twisting the entire engine/transmission in weird directions?

The Volvo engines are most certainly NOT smooth. They are a large displacement (the most common was 2.3L) inline four, with no balance shafts. They got away with them by using HUGE squishy mounts, which failed a lot actually. The manuals also used a pretty chunky flywheel, but most of these were bolted to the Aisin 4sp slushbox.

The AHU in its original form did not even use a DMF, not that it would necessarily hurt anything to have one. I suppose one could check the smooth idle compensation numbers to see any possible hint of a cylinder imbalance, and another thought I had after posting yesterday was to maybe run the engine with the front pulley off and see if the problem is still present. The timing belt should still track OK due to the collar on the tensioner pulley, but you may want to keep a close eye on the front of the crank just in case it starts to walk off, but I have run other VAG engines like that briefly and they never did.

 

[jmodge]

You probably already thought of this, but Has your crank spacer been common during all this?

 

[cj.surr]

I think you are looking at the right area focusing on the engine. It would be hard for an issue like driveshaft alignment or rear axle bushings to transmit enough force backwards, through the gearbox and then through the clutch disk all the way to the flywheel, to cause such an intense vibration or tear up a DMF. I think if anything downstream of the trans were causing it, you'd also be breaking U-joints, having differential pinion bearing problems, etc. And I doubt anything about how the engine/trans are mounted could transmit enough force back into the rotating components to cause failures, so I'd rule out anything having to do with the mounting system.

Plus, you have had the same issue despite running three completely different transmission and driveshaft setups and a variety of different mounts. And you say the vibrations are abated when a DMF is used, but then the DMF fails soon. That tells me the DMF is doing a lot of work trying to smooth out torsional pulses, and helping, but that the forces are too strong. I think the motor is the only thing that could cause the kinds of symptoms and damage you're describing. Amplitude of the positive/negative crank torque is too high, for whatever reason, and that's causing the bad vibes with the SMFs or beating your DMFs to death.

What kind of tune is it running? If it's not the stock tune, what kind of SOI timing is it dialing up in the RPM range where the vibration is worst?

When you say compression was fine, what kind of readings did you see? Were they in the middle of acceptable range, or off the high end of the scale?

Know anything about the motor's history? Head ever been off it? What thickness HG does it have?

Thanks for the reply. I completely agree with your analysis. You bring up some good points.

I have been running some cheap chip tune I bought on ebay for a couple years. I ran the engine briefly on the stock tune when finishing the swap and I'm pretty sure the vibrations were the same. I think I have a stock ecu to throw back in and compare, though. Since it has been a while.

I forget exactly what the compression numbers were but I think they were on the low-middle side of the range. That's probably worth checking again too.

I haven't had the head off, and based on it's previous owner, I doubt anything other than stock HG would have gone back in it.

The Volvo engines are most certainly NOT smooth. They are a large displacement (the most common was 2.3L) inline four, with no balance shafts. They got away with them by using HUGE squishy mounts, which failed a lot actually. The manuals also used a pretty chunky flywheel, but most of these were bolted to the Aisin 4sp slushbox.
The AHU in its original form did not even use a DMF, not that it would necessarily hurt anything to have one. I suppose one could check the smooth idle compensation numbers to see any possible hint of a cylinder imbalance, and another thought I had after posting yesterday was to maybe run the engine with the front pulley off and see if the problem is still present. The timing belt should still track OK due to the collar on the tensioner pulley, but you may want to keep a close eye on the front of the crank just in case it starts to walk off, but I have run other VAG engines like that briefly and they never did.

True, the Volvo engine mounts are what I used originally. They have a ton of flex, but are actually not very large actually. The M42 mounts I'm using now are much larger (still smaller than TDI mounts). The m42 mounts seemed a little smoother and constrained the engine much better (startup and shutdown shaking).

I'm curious why I should try removing the crank pulley? Sounds risky since I need to load the engine and drive it around to simulate the problem. I've tried two different used OEM pulleys and a fluid damper with no difference.

You probably already thought of this, but Has your crank spacer been common during all this?

You mean between the crankshaft and flywheel? I only had a crank spacer for the Toyota setup. The volvo and T5 transmissions, I have just used 1.8t flywheels bolted directly to the crank.

 

[jmodge]

Yes. rwd I take it, if so you seem to have eliminated, by replacement or swap, everything between the crank flange and pinion yoke on the diff?
Have you ever disconnected the driveshaft and powered your diff at the pinion to see what happens. I thought of easing it up with an impact on the pinion nut, but that could cause a problem that doesn't exist. If you could weld a nut straight on a u-joint maybe? Something along those lines

 

[jmodge]

If there is a faceplate on the diff that would work to drive it

 

[turbobrick240]

Sounds like an issue with the adapter plate alignment.

 

[d24tdi]

I have been running some cheap chip tune I bought on ebay for a couple years. I ran the engine briefly on the stock tune when finishing the swap and I'm pretty sure the vibrations were the same. I think I have a stock ecu to throw back in and compare, though. Since it has been a while.
I forget exactly what the compression numbers were but I think they were on the low-middle side of the range. That's probably worth checking again too.
I haven't had the head off, and based on it's previous owner, I doubt anything other than stock HG would have gone back in it.

I think my next move would be to log start of injection under load (Group 4 IIRC though don't quote me) and see what kind of timing it's asking for and achieving. Then I would pop a set of stock chips back in and see what that does to the vibration and timing values.

Some cheap tunes are common about running overly aggressive timing advance. Too much timing can cause a high negative instantaneous crank torque impulse at the start of each power stroke. They do it in order to extend the injection duration window (in terms of crank angle) and thus inject a greater total volume per stroke from the stock size nozzles, a terrible idea, but it allows them to advertise that you can add big HP using their tune without having to touch the injectors.

Then again those same tuning approaches are brutal on pistons and rods and headgaskets, yet you say you have put on 60k miles with no problems other than the vibes and broken flywheels. Probably would have hurt the motor by now if the tune were the issue. So, I wouldn't get hopes up too high on that being the cause. But it's an easy thing to rule out so may as well look at it.

Bottom line my impression is that this problem is related to excessive cyclical variation in crank *torque*, not dynamic balance. I'd conclude that from the symptoms you are seeing and especially from the fact that it is heavily dependent on load, and that the destruction is centered right at the first point that absorbs the crank's torsional forces. The job of the DMF (or of a heavy SMF) is not to balance the internal reciprocating masses in the engine, but rather to dampen the power pulses and torsional harmonics from the crankshaft. I really doubt anything having to do with engine mounts or even the front accessory drive could cause the effects you're seeing. I think somehow either the peaks and valleys of the crank's instantaneous torque output are too high and low, or possibly there is one cylinder whose power contribution is excessively large or too small and setting up a bad torsional harmonic through the crankshaft and flywheel.

Checking the smooth idle injector balance values as OH suggested is a good idea, although at idle the load might be too light to tell you everything. No other obvious engine issues, right? No excess axial movement in the crank thrust bearings? No chance of a bent connecting rod in one cylinder, would have been visible in the compression test results? Any valvetrain noise suggesting a partially collapsed lifter in one cylinder? Or intake manifold clogging that is unevenly affecting the runners to one or two cylinders? Try looking for something strange like those examples, something that could cause the engine's power delivery at elevated RPM and under load to become unequal across the 4 cylinders, yet not be visible in a low speed compression test or at idle, where it sounds like the engine runs fine and vibrations are not a problem.

Or swap your other good running engine in and see if it makes a difference.

 

[d24tdi]

Sorry about the long post, this is an interesting problem and you got me thinking out loud about it. Look forward to hearing what you figure out.

For a reference point, the AHU in my '93 Toyota pickup with an ACME plate and ACME's solid flywheel and R150F trans buzzes the truck a little at idle but is smooth under revs and load. Your SMF and W56 setup should have been the same.

 

[jmodge]

Sounds like an issue with the adapter plate alignment.

That was my first thought but he seems to have eliminated that by the use of multiple adapters and alignment checks

 

[jmodge]

something that could cause the engine's power delivery at elevated RPM and under load to become unequal across the 4 cylinders

Interesting problem and interesting thought

 

[jimbote]

bent crank at flywheel flange?

 

[cj.surr]

Thanks for the replies. I got back around to working on this car. I installed a single mass flywheel setup and it's back vibrating like it did with the other transmissions. I did a warm compression test - 430psi across the board. I tried swapping in a stock ECU, no change.

I poked all of the lifters with a screwdriver, they all would compress about 1/16" from the cam lobe. I think that rules out a bad lifter, also the good compression test seals the deal.

D24tdi, I like how you are thinking about this problem, I very much agree with your analysis.

One thing I have always noticed but I'm paying more attention to now: there is a slight, dull thud every 2 engine rotations (I think) at idle. You can barely feel it in the car but you can hear it. It doesn't seem like part of the typical diesel chatter. It seems like it runs on the same frequency that the problem vibration does. Seems to me like every 2 rotations would have to mean a firing imbalance between cylinders. Gear rattle in this car is also pretty bad in neutral, even with a 22lb flywheel.

I think my next move is to pull the manifolds and see if I can spot anything. Assuming I won't, I'll swap in my spare engine. Really not looking forward to that, it's an especially greasy and high mileage example. It also seemed low on power in the donor car, although I think it was an N75 issue.

 

[turbobrick240]

I'd try driving down a big hill in whatever gear and speed gets things shaking real good and lift off the accelerator but leave it in gear. If the shaking continues, it's not an imbalance in fuel delivery , or ecu/timing issue.

 

[cj.surr]

On decel I will get a light vibration in that low range, but it's not nearly as extreme as under load. I pulled the injection pump off the spare motor, I was actually planning on swapping it on tomorrow. I figure if it has good compression, then any kind of torque imbalance would have to come from fuel delivery.

I also noticed that the rear "flange" (behind last bearing) on the camshaft has a large chunk broken off of it. Doubt it's related, but that was surprising to find.

 

[jmodge]

Can you get at the flywheel with a dial indicator? It sounds like you checked everything between the crank flange and diff flange, maybe one of the two is bent. If it’s a fuel delivery issue may be a good hard double dose of liqui Moli would show some improvement.

 

[jmodge]

One thing I have always noticed but I'm paying more attention to now: there is a slight, dull thud every 2 engine rotations (I think) at idle. You can barely feel it in the car but you can hear it. It doesn't seem like part of the typical diesel chatter. It seems like it runs on the same frequency that the problem vibration does. Seems to me like every 2 rotations would have to mean a firing imbalance between cylinders. Gear rattle in this car is also pretty bad in neutral, even with a 22lb flywheel.

That sounds like cam speed. Maybe a stethoscope at different idle speeds may tell you something.

 

[d24tdi]

I also noticed that the rear "flange" (behind last bearing) on the camshaft has a large chunk broken off of it. Doubt it's related, but that was surprising to find.

Breaking off a chunk of the back of the cam is a pretty common result when someone tries to use the cam timing locking bar as their torque counterhold to loosen the cam sprocket bolt. Likely what happened there. Unfortunate, but usually harmless in terms of the cam's operation since there is not much axial force on that rear cam thrust surface in general. I agree that's probably unrelated to the vibration and flywheel damage symptoms. It does make it a challenge to set cam timing again though, since the slot for the bar is gone. You can do it with care but it's not much fun. Brand new cams for these old TDI engines are under $100 so I would probably be tempted to just throw a new one in along with a set of lifters if I were going to keep running the engine, but YMMV.

Every 2 rotations of the crank does sound like one cylinder has something different going on. Strange it's not showing up in the compression test. One thing that could be interesting to do would be to try a dynamic compression test -- that is, put the compression tester in and start the engine. Of course in order to not blow up the tester you'll also have to have the fuel line to the injector for that cylinder cracked loose so that there's no combustion, just a read of the engine's mechanical compression in each cylinder. Gauge won't be hurt by this. Especially if you can try a few different RPM levels with it, might give you some data points to show if anything in the base engine is creating a cylinder contribution imbalance that the cranking compression test didn't identify. Even if it has good compression at starter speed, there are still ways that dynamic compression could be unevenly affected so the even compression numbers you recorded so far might theoretically not be the whole story in this unusual case.

Changing the IP sounds like a reasonable move to try too. Something like a leaky delivery valve or a flaw in the IP head in theory could create a difference in timing or IQ to one cylinder. Although I would suspect if a problem like that existed, you would be able to see it reflected in the idle smooth running compensation values in VCDS. Did you take a look at those, and if so what do they say?

 

[cj.surr]

I picked up a supposedly good running complete engine today to swap in if necessary. I wasn't feeling good about my previous spare engine for a few reasons.

I did just check smooth idle adjustments with engine warm (nothing for cyl 4, i assume it's the reference cyl)

1) 1.04
2) .66
3) 1.29

If I load the engine at idle (put it in gear and drag the brake), cyl 3 gets up to 1.6, and cyl 1 and 2 get close to zero. I'm not sure if that info is worth anything. Definitely curious what you think about the numbers, swapping the engine was going to be tomorrows project, unless swapping the pump makes more sense.

By the way, the injectors on the engine are Bosio DLC 520, purchased and "calibrated" by Kerma about 10k-15k miles ago.

The idle speed compression test is an interesting idea. Maybe I'll give that a try tomorrow.

 

[cj.surr]

I just tried the idle speed compression test. 500 psi +/- 10 psi across the board. What do you guys think, swap the injection pump based on the good compression test and poor smooth idle values? The injection pump is out of a 390k mile engine, so I'm not really certain that will rule anything out. Otherwise I'll swap in that whole good engine I got yesterday.