Very interesting!
(* The included link is not working any more
18:28 Athens GMT + 02:00 Done)
"The
Passat Transmission Problem Solving Process.
Introduction to Problem Solving
Techniques.
Almost all manufacturing companies will have a formal
problem solving process which engineers will use to find out exactly what a
quality issue is and hence how to prevent re-occurence. Such processes can take
months to complete due to the amount of testing that is sometimes required and
the need to demonstrate the root cause by recreating the problem in the
laboratory.
When working on vehicles at home, we do not usually
have to hand the vast facilities that vehicle manufacturers have and therefore
sometimes repair the symptoms without fully understanding what went wrong. A
good case here is the Triumph valve regression I experienced in 2008. Having
new valve seats fitted got it back on the road, but it is still unknown what
caused it in the first place and hence whether or not the problem will
re-occur. This, however, was an acceptable solution to me as the cost of
analysis in a metallurgy laboratory is far more than the cost of the work
carried out or the value of the bike. It's not even if I'm alone, as garages
and professional workshops are sometimes also guilty of such a lack of
thoroughness, although their experience makes them less likely to do this than
me.
But what if there is re-occurrence? A part is replaced
and it fails again. Cut your losses and scrap the vehicle? Sell it as a
non-runner? When the Passat started destroying rear differentials it was mainly
my curiosity that made me want to investigate, despite the little voice in my
ear saying "scrap it, scrap it". Which, of course, turned out to be Gabi
and I guess this difference is why she studied accountancy and I studied
engineering.
The popular methods for problem solving tend to follow
the path of problem definition, cause and solution. In the car industry, one of
the favourites is the 8D approach, invented in America. (The story goes that
various large companies adopted this approach without the niceties of asking
the inventors. They sued, won lorry loads of money, retired, presumably to some
Caribbean Island, and now allow anyone to use
their methods free of charge). Because the 8D process is my favourite, I will
be using it.
D0 Symptom
Passat Transmission failures
Introduction to the Passat
Transmission Problems.
A few years ago I bought a second hand 1996 Passat VR6
Syncro 35i (facelift) model with 2.9litre engine and 5 speed manual gearbox.
It came with a fully stamped up service book and no reports of major problems.
Problem Sympton.
The Passat has suffered from various transmission
failures.
139997km
LH front halfshaft replaced (inner UJ failed).
FL,
FR and RR wheel bearings replaced.
ca. 158000km
RR tyre puncture. One rear tyre replaced (winter tyre).
158600km
Rear differential and syncro replaced.
All
four wheel bearings replaced.
Front
LH hub replaced.
167000km
RR tyre puncture. Both rear tyres replaced (summer tyres).
167200km
RR halfshaft replaced (outer UJ bearing disintegrated).
168000km
Viscous coupling unit failed (casing cracked). Rear differential and viscous
coupling replaced.
168100km
Rear differential failed (holed outer casing). Previous differential unit
fitted.
168150km
Rear differential failed.
Note that with the second puncture two tyres were
replaced because the tyre fitter had this noted as a requirement for VW
Syncros. Unlike the first tyre fitter it seems.
D0 Emergency Response
Park vehicle on drive and put up on axle stands.
This would usually be overlooked except for safety
related issues. Suppliers tend to write things like "parts sent for
testing" here, but realistically, this would be concerned with recalls or
production stops - both very expensive and generally requiring top management
support.
D1 Create a Team
Frank
John
Okay, an ideal team would have about six members,
including a facilitator. Also worth having a manager in there in case some
expensive testing is required. Never, ever, allow the manager to join the
actual meetings as they all, without fail, think that they alone are blessed
with some fantastic engineering insight and immediately know what the problem,
and solution, are. You then spend three months following this path until it is
shown to be a complete red herring by which time the manager has got bored and
stopped attending anyway.
D2 Problem Statement
Torsional wind up in the rear differential.
Torsional wind up is caused when the there is a
difference between the rotational speed of the bevel gears in the differential.
The input speed is determined by the gearbox and will dictate the output speed
to the wheels. However, if the wheels are being forced to travel at a different
rotational speed (dictated by the road speed), a conflict arises. Normally the weakest
part of the transmission will then fail, which is why half shafts and the
viscous coupling have failed as well as the rear differential.
This is what has
been happening to my differentials. Big hole is where the drain plug has been
removed.
Opening the rear
differential up (with help from an angle grinder) shows the internal damage.
Here, the teeth on the bevel drive have failed as one of the cogs wanted to
rotate at a different speed to the other.
D3 Interim Containment Action
Remove prop shaft
Knowing a little more about the problem, it may now be
possible to introduce some temporary action to protect the customer.
This was suggested by one of the garages I went to.
However I decided not to do this. It should have worked, mind, allowing me to
still drive around whilst trying to figure out what to do.
The next step is to determine the root cause, that is
what actual mechanism has caused the failure. This is perhaps the trickiest
part of the problem solving process and there are, in fact, quite a few methods
which would assist the team. The
following are perhaps the most relevant:
1.
cause and effect diagrams (also known as Fishbone
Diagrams or Ishikawa Diagrams) examines all the factors which may influence a
problem
2.
brainstorming is a team method to generate
ideas, useful when the investigation stalls
3.
is/is not lists examine differences
between what has failed and what could logically have also failed, but has not,
and then examining differences between the two sets.
It took John and I four months to get from the Problem
Description to the Root Cause. This is because we required a full understanding
of how this particular transmission system worked. The information offered on
website forums was a mixture of truth and nonsense, VW dealers did not have any
in-depth knowledge and were not too interested in assisting, the VW dealer
technical information site was a complete dead loss and the workshop manuals
were misleading. However, armed with John's knowledge, a few parts diagrams and
an angle grinder with which to dismantle one of the failed units, we can now offer
the Ultimate Guide to the VW Syncro.
How the VW Syncro system works.
(Blogs alert: the above diagram has an important inaccuracy: the front cv axle shafts are not symmetrical).
The VW all wheel drive system under discussion
consists of three main component parts. The front differential, the viscous
coupling ("syncro unit") and the rear differential.
The front differential (1) drives the front wheels
permanently. The differential is used to allow the two front wheels to rotate
at different speeds, which is important when, for instance, the vehicle is
turning a corner when the inner wheels will rotate slower than the outer wheels
due to relative distance travelled. Sometimes even on a straight road, road
surface changes may also require the wheels to move at different speeds.
If you look at a small capacity quad bike, you may
wonder why they have such big "balloon-y" tyres. The reason is that
they use solid rear axles and hence there will be an amount of tyre scrub when
cornering. The tyres used will facilitate this.
There is a third output (2) from the Passat front
differential which makes it different from a standard front wheel drive vehicle
differential. This is the propeller shaft. The speed at which this rotates is
fixed by the input shaft rotational speed and is geared at 16/21 by the bevel
drive gear drive. The propeller shaft (3) runs under the vehicle to the viscous
coupling.
The viscous coupling (4) is quite a simple mechanical
device. It is made up of alternating circular plates. The plates are mounted in a sealed
drum, and are located very close to each other. The plates have tabs, or
perforations, in them. The drum is filled with a silicone-based oil. When the
two sets of plates are rotating in unison, the oil stays cool and remains in a
liquid state. When the plates start rotating at two different speeds, the shear
effect of the tabs, or perforations, on the fluid will cause it to heat up and
the viscosity will increase, which is the relevant property of the
silicone-oil. The fluid in this state will essentially glue the plates together
and transmit power from one set of plates to the other.
Hence, under normal conditions, the vehicle is front
wheel drive only. If the front wheels lose grip and start to spin, torque will
then be applied to the rear wheels which (hopefully) will find grip and restore
vehicle motion.
Note that the VW have engineered the size of the tab
and perforations to allow a 10% torque transfer at normal conditions (meaning
the vehicle is not front wheel drive only, but split 90/10 front to back). This
is to allow quicker response when the front wheels do slip.
The output from the viscous coupling will then drive,
via a bevel drive (5), the rear differential (6) and is geared at 21/16 to
restore the same rotational speed in the rear half shafts as the front ones.
Now this is where there is some slight confusion. In
my Passat there is a free wheel system (7) which will disconnect drive to the
rear wheels. The workshop manuals I have, one in German and one in English and
both based on original VW assembly sheets state that:
"During braking, a free-wheel mechanism in
the rear axle differential will automatically engage and interrupts the drive
to the rear wheels. This will, therefore, not influence the braking of the rear
wheels and braking takes place in the normal manner".
This is not true. The free wheel operates in the same
way as a motorcycle starter clutch and will allow, for example, shaft A to
drive shaft B but not allow shaft B to drive shaft A. So that when shaft A
spins faster than shaft B, shaft B also spins. When shaft B spins faster than
shaft A, shaft A is not driven, although it can still rotate as long as it
spins slower than shaft B. There is no influence on this system of the braking
system.
It is also not a requirement for ABS, as some forums
state. In fact, not having the free wheel engaged will be an advantage as it
will allow the viscous coupling to assist the ABS by transmitting torque to the
rear axle and hence matching it to the speed of the front axle.
Now "Volklore" has it for the older T3
syncros that tyres need to be swapped from axle to axle to even out wear. If
this is not done, then torsional wind up will occur and the transmission will
fail. These VWs do have some detail differences in design, in fact I understand
that not all used a viscous coupling as the early models used a de-coupler.
Also there was an option of switcheable differential locks for the front and
rear axles.
However, the advice to rotate the tyres is relevant.
This is because if you have two tyres with different amounts of wear, their
effective radii will be different. This would mean that when driving along a
straight road, one (the more worn one) will rotate faster than the less worn
one.
To understand the relevance further, John helpfully
set out two different scenarios:
Front tyres have a greater diameter than rear tyres.
The rear axle is rotating faster than the crown wheel,
so the diff case will be free wheeling around the crown wheel shaft.
Rear tyres have a greater diameter than front tyres.
The viscous coupling will be driving the rear axle,
trying to speed it up to match the front axle speed.
The second scenario here is what happened to my
Passat. Having had punctures, the rear tyres were replaced. With them having a
larger diameter, the half shafts will want to rotate slower than the front half
shafts. The result is that the input plates (connected to the propeller shaft)
of the viscous coupling will be rotating faster than the output plates
(connected, via the locked free wheel, to the rear half shafts). The viscous
coupling is only designed to cope with a small, temporary difference in plate
speeds, so the rotational speed difference between the two plates, likely
exaggerated by the Autobahn speeds the car was travelling at for all failures,
causes the oil bath in the viscous coupling to heat up. This increases the
viscosity, forcing the input and output plates to match speeds. Now we have the
classic torsional wind up scenario where input and output speeds in the rear
differential are different and the weaving occurs as the wheels try to slip
before ultimately the weakest part fails.
As an aside, if you have followed the logic of the
design of the rear differential, you may have been wondering what happens in
reverse, when the free wheel will be engaged. After all, it would be nice to
have four wheel drive if you parked in a muddy field when visiting a traction
engine show or were camping. So VW fitted a direct drive which bypasses the
free wheel in reverse. It is switched by a vacuum operated plunger, the
movement of which is determined by a microswitch at the gear lever.
During the early part of the investigations I found
that the vacuum filter was blocked which was possibly interfering with the
operation of the plunger.
Here is the
(broken) unit removed from the housing, showing damage to the bevel drive gear.
Note the pegs that are used for the reverse drive lock as the large toothed
ring engages with them. To
the right of this is the freewheel mechanism.
Visible here is
the rear differential, which allows the driven rear wheels to rotate at
different speeds when cornering.
Some views of the reverse plunger, showing (left) the vacuum reservoir on the front left wheel arch and (middle) the mechanism itself. The round thing is the diaphragm which moves a small contol rod which engages the large ringed tooth. In the picture on the right I am working out how the mechanism works. The plunger can only engage the drive, it does not engage any free wheel mechanism. When out of reverse, the free wheel mechanism is engaged due to the return spring just visible in the upper picture.
D4 Root cause
The rear wheels have a larger rolling radius that the
front wheels
The root cause is the actual problem. As you have
seen, getting from problem description to root cause is the main engineering
input.
In this scenario, which is true for many problem
solving teams I've been on, I've had to rely on specialist knowledge to look at
cause and effect. This method is very useful if you just want to jump to
conclusions, so it is important to make sure that the root cause matches
exactly the failure mode.
I will have to cheat here as normally I would need to
demonstrate the root cause hypothesis is valid by being able to recreate the
failure, but as I do not have such a large stockpile of differentials to hand,
I'll just need to rely on validating by not having the transmission fail again.
Hence we now have a root cause. In fitting new tyres
to the rear axle, the design of the system is such that torsional wind up
occurred. This makes the following step quite easy.
D5 Chosen Permanent
Corrective Action
D6 Implemented Permanent Corrective Action
Fit tyres with the larger rolling radius on the front
axle at tyre change.
I have bundled D5 and D6 together as there is only one
corrective action and it will be implemented. Sometimes there are a number of
PCAs and, following testing for example, not all are implemented.
So, in future, I will need to control how tyres are
fitted. As is common in Germany,
the Passat has two sets of wheels, one set with winter tyres and the other with
summer tyres. This usefully allows me to compensate for wear due to usage.
This is fitting
the winter tyres (the steel wheels on the left - the summer tyres are on alloy
wheels. The winter tyres' treads have been measured at several points each and
the ones with the deeper treads go on the front.
The cardboard box
is the spare set of wheel bolts as the two sets of wheels have different shaped
seats. These were not supplied to the previous owner by the supplying garage,
incidentally, I bought them after losing three wheel bolts and trying to figure
out why. Nice to know that the main dealers are happy to let you drive around
with incorrectly secured wheels.
D7 Prevent Recurrance Actions
Use tyres of one manufacturer, type and size on all
four wheels.
Good companies always keep records of processes and
requirements for products or services with vehicle manufacturers being
exceptionally good at this, mainly because they buy so many parts in from
suppliers and need to control the quality. So any information learned from a
problem solving exercise can be used to update design or process requirements
or even result in new quality tests being implemented.
D7 Recommendations of Prevention
More prominent advice should be offered to customers
relating to fitting tyres.
Right, I'm going to apologise now for pointing the
"j'accuse" finger now, but I am less than impressed with VW and their
dealers. Manufacturers of consumer goods have a duty (already tested in law I'm
told) to highlight requirements which are not considered as the norm for that
type of goods. And by highlight, this means make sure you cannot miss the instruction.
So a washing machine manufacturer cannot put a small red button on the back of
the washing machine which must be pressed every ten minutes or the motor will
blow up and then tell you the instruction was written on page 54 of the
instruction manual. Because such an requirement is not what "a reasonable
person" would call normal, the manufacturer would be required to stick a
prominent label warning you to carry out the procedure.
So, how do VW deal with the issue of how to fit new
tyres:
Betriebsanleitung
Seite 129
Es empfiehlt sich aber, die Reifen mit der größeren
Profiltiefe vorne zu montieren: Sind die Vorderräder stärker abgenutzt als die
Hinterräder, haben sie einen kleineren Abrollumfang und drehen deshalb
schneller. Dadurch treten Verspannungen im Antriebsstrang auf und die Reifen
verschleißen schneller. Ein erhöhter Verschleiß des Allradantriebes ist jedoch
nicht zu befürchten.
which, roughly translated, says that tyres with more
tread should be fitted to the front because otherwise the faster rotating front
wheels will build up tension in the transmission. The last sentence says that
increased wear in the transmission should not be an issue. Hence it does seem
to miss the point that the differential will explode on the motorway. And
hiding it away on page 129? Unbelievable!
So I think VW failed here. If I was an engineer
working on this issue for VW, I would also recommend
More prominent advice should be offered to customers
relating to fitting tyres.
And in case any VW engineers are reading this and are
saying that the issue is common knowledge, remember that the car was seen by a
VW garage and two independent garages, one of which sells itself as a VW
specialist, and none of them spotted this.
Well, thanks John here for patiently taking me through
the design of the transmission system. I owe you a beer.
Conclusion
Despite all the hassle, I have found the whole
exercise interesting in many ways. I just wish it had happened to someone
else's car. The Passat is now running fine, although there is the occasional
clunk in reverse which I tend to attribute to the freewheel bypass engaging. I
think one of the lessons here (for me) is that whilst internet forums can be a
hugely useful facility when trying to solve problems, not all information is
correct and hence it is always worth verifying in someway. Also, I think it
unforgiveable of VW that not only do they reduce the significance of tyre
fitting, they also allow incorrect information to get out. All it needs is a
sticker near the rear wheel well and I would have saved myself a lot of time
and money."
