Between the years of 1995 to 1998 there was a number of
truck wheel-off instances within the province of Ontario, two of these
wheel-off instances resulted in fatalities.
A public outcry ensued and Government legislators vowed to curtail the
problem. As a result strong legislation
was passed that required all persons that perform tire work to pass a written
examination and licensing requirement.
The use of a torque wrench was mandated when performing tire work. Scale-houses adopted a seven-day a week twenty-four
hour a day schedule. Fines to
commercial vehicles soared to new heights, in order to eliminate the attitude
that a hundred dollar fine was cheaper than fixing it or a cost of doing
business. Fines in Ontario are in the
range of fifty thousand dollars for serious offenses, and wheel-offs are an
absolute liability, in other words wheel-off offenses can have no defense in
court, the operator is guilty and will pay.
The days of the mechanic adjusting wheel bearings by
experience or feel have come to end. A
quantifiable and consistent method of wheel bearing adjustment has taken its
place. There are two basic and
acceptable methods to adjusting trucks wheel bearings. The first method is to follow the
manufacturers recommendations. The
second method of wheel bearing adjustment is to follow the TMC guidelines. TMC
stands for the "The Maintenance Council" and is administered by the American
Trucking Association, with the cooperation of the vehicle manufacturers. It is TMC guidelines for wheel bearing
adjustment that I follow in my place of employment.
The TMC procedure For adjustable type wheel bearings is as follows:
Step 1. Lubricate the wheel bearing with clean lubricant of
the same type used in the axle sump or hub.
Do not use grease on a bearing running in oil. Packing a bearing with grease that ordinarily runs in oil has the
potential of the grease depriving the bearing of oil.
Step 2. While rotating the wheel tighten the adjusting nut
to 200 ft-lbs. This seats the seal and
removes any burrs from the bearing, and is a critical step. Steps 3-5 merely provide for a method of
adjustment that gives the correct amount of final endplay most of the time, and
is not critical. Final endplay is
critical but how you get to the final endplay irrelevant.
Step 3. Back the nut off one full turn.
Step 4. Tighten the adjusting nut to 50 ft-lbs while
rotating the wheel.
Step 5. Back the nut off for a steer axle with single nut
and 12-pitch thread 1/6 of a turn: for 18-pitch ¼ turn. Install the cotter pin.
For a double nut steer with 14-pitch or an 18-pitch threads
back the nut off ½ a turn.
For a drive with 12 or 16 pitch thread back the nut off ¼ of
a turn.
For a trailer with 12 or 16 pitch thread back the nut off ¼
of a turn.
Step 6. Torque the outer nut (jam nut) on a steer with a
less than 2-5/8” nut too 200-300 ft-lbs: 2-5/8” and over nut torque too 300-400
ft-lbs.
Torque the jam nut on a drive with a dowel type washer too
300-400 ft-lbs; Tang type washer too 200-275 ft-lbs.
Torque a trailer jam nut with a nut less than 2-5/8” too
200-300 ft-lbs: with a nut 2-5/8” and larger too 300-400 ft-lbs.
Step 7. Acceptable endplay as measured with a dial indicator
is .001”-.005”.
For a single nut self-locking systems or non-serviceable
axles consult manufacturers
specifications.
Also see : http://www.timken.com/bearings/techtips/tip5.asp
I used to think that attempting to keep to the low end of
the wheel bearing endplay specification was better. My thinking was that any looseness in the wheel-end assembly is
proem to a hammering effect of the bearing and seal. I posed the question to Marvin Swenson. Marvin is a Maintenance Engineer of thirty-three years
experience. I also asked Marvin to calm
my fears to inaccurate readings due to the differences in size and strength of
the person pulling on the wheel.
Marvin’s reply is as follows.
"Your question about the 100-pound man versus the 150-pound
man is a good one!
It turns out that I fall in the 150 pound range and so was concerned about
whether or not I was pulling hard enough to take up all of the slack in
end
play. I think there are merits to your question.
This is my thought:
Lets look at the procedures. Should there be any grease in between the
inner
nut and inner race this would pushed out by the initial torque being
applied. Also by rotating the hub at the same time (both directions any
grease that would be between the rollers and races would be wiped off (except
for a small lubricating film). This leaves us for the most part with clear
spaces everywhere that is called endplay.
Now we get to your question the pull which must be enough to initially move
the mass as a function of any incline surfaces that the mass must travel over
, to include any adhesion of residual grease on the ends of the inner
races
and around the spindle, and any other friction that exists. The force
required will be much less than the total weight of the assembly. If
the
dial indicator moves at all we have overcome the resistance in moving the
assembly to the opposite limits. Once the mass reaches the limits and enough
force is maintained so that no tilt exists (explained later) any
additional
force would not change the dial unless it was enough to start deforming the
spindle hub assembly. (Remember that the lock nut has been torqued at
this
time, so no play exists in the threads of the two spindle nuts.)
Now for the interesting part, once we get to the opposite limit we
must
insure that it stays there while we read the dial gauge. This may not seem
like a problem but the mass of the tire, rim and hub is not centered
around
the spindle and there are small clearances between the race, spindle and race
rollers. This will result in a slight tilting of the
assembly and result
in inaccurate dial readings. I think maintaining a strong pressure in
the
direction of the limits while reading the gauge would improve the accuracy of
the readings.
I think this is supported at least in part by the procedures (If I read
between the lines) in the maintenance manual. It states: "Do not push /
pull
at the top and the bottom of the hub or drum. Pushing or pulling at the top
and the bottom will not give a true reading of the endplay. I think this is
to insure that the slight tilting is not aided in anyway.
I understand your feeling that 0" or near 0" should be a goal for
final endplay, however I
would like to pass on a bit of food for thought on that subject. Please for
give me for letting the engineer come out in this thought process, it’s just
the way we are trained to think.
Lets start by talking about the fact that the two outer races are in a
cone
configuration and that the true roller clearance is not equal) to the end
play (in fact in this case the roller clearance would be less than the end
play. This is because the races are tapered; the endplay measurement is
not
perpendicular or parallel to the bearing race surface (sine or cosine
function
depending on where one would pick the reference for calculations). Then I
think that in the big picture thermal expansion should be considered. If
the
hub temperature increases the distance between the two outer race cones would
get larger thus putting additional pressure on the rollers. This would be
nearly a100% offset by the expansion of the spindle if the hub and spindle
were in operation at the same temperature and if they were made of the
same material (not the case) one is cast iron and the other steel. I
would
just off the top, think that the spindle would operate at a lower temperature
than the hub. This would result in differences in hub and spindle as a
function of temperature. Then there would be another offset to the cone
expansion issue, that the diameter of the cone both (outer races and
inner
races) would increase as the temperature increased but again they
would
not operate at the same temperature. I am sure that the designers had to go
through the math and adjusted the various (of which there more than I have
talked about) dimensions as best they could to minimize the effects of
clearance as function of temperature. In taking a quick look at only
TWO of the
variables, the hub and spindle.
The hub:
The linear expansion coefficient of cast iron is .000011 cm / cm / degree
C.
(steel is .000012 cm/ cm/degree C), so if the drum temperature goes from
70
Degree F to 300 Degree F change in temperature I would
have a 127 Degree C
delta.
If I then assume that the distance between the two outer races were 15
cm (
about 6 ") then I be increasing the outer race cone by .000011 x 15 x 127
=
. 02955 Cm. Or .00825” WOW.
Because the expansion coefficient of steel is larger than that of cast iron
the spindle would get longer by .00075" than the hub if they were
operating
at the same temperature and were the same length. This condition would result
in having an increase in endplay, great so far. However I think that the
operating temperature of the spindle would be less than the drum and in
fact
if the spindle were operation below a 116. 4 degrees C / 241 degrees
F we
would start to decrease the amount of end play. If the spindle were
to drop
below 102.3 degrees C / 216.15 degrees F and the drum stayed at the 300
degrees F the end play would decrease by .001" and we would have 0"
end play
if our initial setting was .001".
This is only the Hub and spindle, many off sets both positive and
negative
will take place but one thing for sure during times of high breaking there
can be significant time periods when all of the components in the
drum,
bearing and spindle assembly can have large temperature differentials. It
takes time to transfer thermal energy from one location to another and
clearances could change a significant amount during that time.
So my conclusion is that sticking to the minimum spec of .001’
would
provide better insurance that the bearings and related component would not go
through period’s insufficient clearances due to thermal related changes".
Marvin Swenson
Las Vegas Nevada
