Steam & Excursion > Plain bearings vs Roller bearing rail car axles….

"Plain bearings" roll with less friction than rollers.
"Why do you say that, Wes?"

A couple of reasons....

1. Plain bearings have only a thin oil film between the rotating axle and the babbited,
   fixed bearing-shell.  (Every automotive & truck engine,  as well as all locomotive
   diesel engines' crank shafts are all spinning on 'plain' bearings and that oil film,
   which is fed under pressure).
    
     Rotating element bearings have limits on their "useful service life",
     typically listed as thousands -of-hours, by bearing part number.
     Rolling element Bearings have maximum RPMs, 

     Plain bearings have NO such RPM limits or 'hours' rating.  
     As long as the lubrication.is provided the useful service life is not specified..

2.  Rotating Roller bearings leave higher temperature readings on thermal-monitoring
    "Hot Box" Detectors.  In today's world, hot box detectors are obsolete.

    [ The old hot box detectors have been, and are being, replaced with 
      high, audio frequency bearing monitors.    These listening devices 
      are better at finding and predicting all rolling-element bearings that
     are in distress, and danger of causing a derailment

      The old hot box devices were spaced too far apart, compared to 
      how quickly a failed, squealing roller bearing will overheat and 
      burn-off the axle end before the next "heat detection" device is reached.
      So, high audio frequency detectors are the only way to catch rollers in distress,
      well in advance of actual squealing or running hot.]

Back when I would occasionally monitor the paper-tape readouts from the "hot box detectors"
at the Receiving Yard entrance tracks to Potomac Yards, in the 1980s, the low peaks were
always from the plain bearings,  The roller bearings were all measurably warmer, the whole
axle end and exposed outer 'cup' of the roller bearing would be warmed to the same 
temperature...even for newly applied roller bearings.

Rolling element bearings used to leave consistently higher heat-peaks on the 
temperature-recording, paper tapes.   Near the end of the use of plain bearings 
in interchange serivice, plain bearings' thermal peaks were consistently 1/4 the height
of virtually all of the rolling element temperature peaks, measured on rolling 
freight trains.

There are more "rolling elements" in each axle bearing assembly, as well as the 
spacer cage that separates the individual rollers from rubbing each other --- all combine to
generate warmer bearing temperatures, when rolling.

On plain bearings, once the oil film is established between the axle and 
the bearing,  the rotating surface and the fixed surface are separated by that oil film.  
The oil film must be 100% uniform, and NO metal to metal contact can occur --- or overheating
will result.   'Friction bearings' is an incorrect name when referring to plain bearings.

Tapered roller bearings are applied In 'facing' pairs.  The pair of tapered rollers are needed 
to control the end-wise ( "axial") thrusts from the tracks and wheels of the rolling freight cars.

Cylindrical ( as opposed to 'taoered') bearings, whether roller equipped ( Hyatt) or plain bearing
are.all configured with simple 'plain-bearing'  thrust-absorbing bearing surfaces.  Plain bearings 
Also use the raised axle diameters at the wheel hubs and the outer end raised collar.
But the restraining ( load-bearing) area is only over the top half of the axle and it's hub.

Tapered roller bearings are far better at controlling axial load forces ....
Taoered roller bearings combine a thrust bearing and a rolling bearing.

Sometimes, when empty freight cars being handled roughly in freight yards, the 
top-only bearing elements could be momentarily separated from the axles, and 
the oil pad pieces ( 'waste-grab') could get caught between the axle and then would
ruin that vital, uniform oil-film between the bearing and the axle.   Later, at higher track 
speed, that 'waste-grab'  incident could over-heat the axle .....ergo, a hot box...

However, steam locomotive driver RPMs are very low compared to internal combustion
 engines, and freight car axles.   That applies to driver crankpins also. ---- low RPMs.

more on a separate essay... 


W.



Edited 9 time(s). Last edit at 05/07/22 20:53 by wcamp1472.

Follow-up on thermal detector "typical spacing...." for thermal detectors.

Thermal detectors had been typically spaced about every 20 to 30 miles apart.
With the wide application of roller bearings on car axles, an "at-risk"
 bearing could pass the first detector without exceeding the 'thermal 
warning'  temperature, then begin to rapidly heat-up, and the axle 
could quickly burn-off ( like in under 5 miles, if at high speed) before
reaching the next detector.

With audible frequency detectors, bearings will emit the characteristic 
frequencies well before even getting warm.   It gives crews and dispatchers
plenty of warning to safely set-out the endangered bearing.  

The warning is announced by radio and a synthetic voice that gives the
defective bearing by "axle-count"----  starting from the lead unit's first axle.
So, locomotives and cars typically. are equipped with either 4-axles or
six axles; however, there are are articulated container cars ( "coupled platforms")
that share a truck center-pin and can have multiple, coupled-platform cars.
Or, there were Diesel engines that had two trucks connected by
a span-bolster --- thus, having 4 trucks per unit.

So, with axle-counting, crews at can count the axles when making the
crippled-car "set out".   The 'squeaking' axle may not give any visible
indication that it is endangered . There typicallly is no smoking,
sensible heat radiating and the "emitting frequency" is above the
audible hearing range of humans.   Also, the set-out process is at
a slow speed, and would not generate the axle RPMs of 60 to 70 mph...
and ( low speed) would not necessarily be an audible signal to someone
on the ground.

in the absence of any physical evidence, Axle-counting is the only way to
determine which car is endangered.   Some crews opt to set-out several
adjacent-cars that might be 'suspect'...

Now, with crews getting smaller and smaller, and trains getting longer and longer,
making any set-outs is getting 'problematical'.....

"Modern Technology" is a wonderful thing, until Reality raises its ugly-head and
 worn bearings begin the high-frequency squealing signalling near the upper range
of their cataloged:  "typical service-hours" ....

W.



Edited 2 time(s). Last edit at 05/08/22 13:54 by wcamp1472.

Now that we discover an "at-risk bearing" and the car goes into at show to replace that axle, will the shop replace all the axles on that car that have the same mileage and type of construction?  This is where AAR rules come into play.



Edited 1 time(s). Last edit at 05/08/22 04:31 by PlyWoody.

PlyWoody, 
I don't know about the rest of the axles and a relevant AAR rule.
At rip tracks, generally, only the defective wheel set will be covered for
"AAR billing"....  Financial responsibility for worn-out wheelsets is typically
billed to the car owner, for payment.

Generally, the bearings out-last the useable wear ( profile) of the 
wheels.

Removed Wheelsets are gathered at the ''cripples track', loaded onto specialized,
dedicated 'wheel-shop cars' ( qualified for interchange service) and sent to AAR-certified
'wheel-shops'.....
Freight wheels are typically 1-W ( 'wear") wheels ---- once they fail the wheel-gauge
measurement in the 'field' (shop).

At a Wheel Shop, the bearings are pushed off the axkes, as a result of pressing-off
the wheels.  The removed bearings are sent to a dedicated AAR-requalifying shop,
and fitted with all new components.  ( seals, rollers, etc)  to the requalified  races.
Then sent to the wheel shops, where they're used on newly assembled wheel sets,
ready for application at rip tracks, as needed.

The axle will be 're-qualified' , and if re-usable it will be fitted with new wheels and
'fresh' bearings.   Fresh bearings can be either 'reconditioned' bearings from the field,
or NEW  bearings from the manufacturers.   Under AAR rules. both are financially treated
as 'new' bearings & are expected to outlast that wheel's usable profile.

I DO know that all bearings the have been 'submerged' are immediately prohibited
in interchange by AAR rules, and must be replaced with qualified axle-sets & bearings.
Typically that would be every wheel-set on the "flooded" car,

If there's a high water, mud-line anywhere on the wheel, that would condemn the
wheel-sets as 'submerged', for AAR billing purposes.... whether the flood water got
to bearing's level or not.  


Typically, the seals on roller bearings will keep out the flood waters; but, that's a dicey
proposition.... AAR wants the bearings "fully usable" before returning to "interchange
service".   "Submerged" bearings cars means that the cars cannot be offered in
interchange to another common carrier, once they have been submerged.  
So the flooded cars are remediated with new wheel sets, or loaded onto an
interchangeable flat car for shipment to a proper facility.

I'm not sure who is "financially liable" for 'Flooded cars', under AAR rules:
'Owner's' ( typically a bank or investor), the 'handling road' at the flood site, 
or other responsible party..

I'll look up the AAR rules...

​W.
 



Edited 5 time(s). Last edit at 05/08/22 08:07 by wcamp1472.

"Plain bearings" roll with less friction than rollers.""Why do you say that, Wes?"

I disagree to above.  A retarder yard was built with a 0.12% down grade as the car would roll at a constant 4 mph for plain bearing cars.

In recent years the retarder yards are built with grade of 0.04% down grade as that is the slope that a roller bearing car will roll at same speed.   The different is the friction of the bearing. 
A 1% grade is one foot per 100 feet. 

At Potomac Yard, in the winter ....below freezing, ---- almost 100% 
of our freight traffic was roller bearings ---- the grease in the rollers 
solidified and it took a slug-set of an SW1500 + a Sw1500 to push a
long train to the hump.   

When released the roller equipped cars, empties especially would
barely get out of the restrders...and they'd stop ...stiff bearing grease.
Loads only rolled about 1/3 of the intended distance..

PY kept the NB hump at the same degree slope as was used when the retarders were 
installed , mid-60s... Flatenning to ent degree would have been a cold weather disaster.
Many of our trains came from upstate NY, Penna , New England and Ohio...
They were way colder than simply have been kept at PY..

Similar (bad) winter exoerienced --- the years at the hump in Silvis, Illinois--- CRI&P
Also flat switching in Oneonta, NY, --- the years with  D&H

Trains left standing for over an hour were especially hard to shove ---- all because
of stiff grease...and all the individual rollers refusing to budge in the stiff grease..

That, plus keeping all the track switches swept,...kept 100% of the non-agreement
stsff personnel on duty,  'round the clock.... but, the food was plentiful and hot,
we rested a few winks at local brand-male motels in Alexandria ..... one to a room...
from before the storm hit, until after the yard was restored to full operstion..

Not all railroading was in the "salad days" of summer..... winters were always
to be dreaded...
 

​Wes.

In theory, the babbit of the bearing should never contact the journal when the oil film is present and keeping them separated. But when I worked plain bearing trucks in the Topeka Shops, or out on the West Rip, every brass I ever saw showed plenty of contact wear, even when still running cool, and bad ordered for something else entirely.

Posted from Android

Oil-film bearings, also known as hydrodynamic bearings, are well understood from an engineering perspective.  A lateral load applied to the spinning shaft, commonly an axle or a crankshaft, moves the shaft off-center inside the bearing toward the loaded side.  Oil is pumped from the clearance on the unloaded side to the smaller clearance on the loaded side by the combination of the shear rate in the oil and the viscosity of the oil.  In a high performance car engine, this can produce a local oil pressure inside the bearing of thousands of pounds per square inch, which prevents metal-to-metal contact between the shaft and the bearing Babbitt.  (With a railroad axle, the unloaded side of the bearing is missing, replaced by a pool of oil combined with a lubricating pad or cotton waste that wipes some of the oil onto the exposed part of the journal.)

The zone inside the bearing clearance where the oil is being pumped to a high pressure is called the hydrodynamic wedge.

The capacity of the bearing to maintain separation is proportional to the rotational speed, the diameter of the bearing, and the viscosity of the oil, and is inversely proportional to the bearing clearance. This describes capacity as pressure, which is multiplied by the bearing area (diameter X width) to determine force capacity. 

The oil film is never sustained when motion is stopped, and is re-established during start-up when there is enough speed to rebuild the hydrodynamic wedge. The cross-over to hydrodynamic action in railroad axles was often considered to be about 10 mph. So there will always be visible wear from metal-on-metal contact (but hopefully not much).

In the days when manual transmissions were more common, people would damage the rod bearings in their cars by applying too much throttle at very low engine speeds.



Edited 1 time(s). Last edit at 05/09/22 11:11 by Elesco.

---

So nobody is confused, I better mention that the above diagram is upside down for an axle bearing.

In the 1945 edition of The Steam Locomotive, Ralph Johnson has a section on train drag forces.  He compared plain bearings with roller bearings.  While he says there is very little difference above 10mph, he indicates that the plain bearing drag is generally not less than that of the rollers. (This may not be true under extreme conditions of cold and stiff grease, but I think he was generalizing for normal conditions.)

So this leaves the question of why temperature sensors were seeing higher temperatures with the roller bearing trucks than with the plain bearings.  As Wes writes,  “The roller bearings were all measurably warmer, the whole axle end and exposed outer 'cup' of the roller bearing would be warmed to the same  temperature...even for newly applied roller bearings.”

I think the obvious answer is that the plain bearings were hidden inside a cast steel journal box, partially filled with oil.  Most of the heating would go directly into the oil as it passed through the bearing, and then it would transfer into the walls of the journal box.  As far as I know, the journal boxes on all or most of the later plain bearing trucks were cast integral with the truck frame.  So the heat would efficiently conduct and dissipate into the frame and the highest temperatures would never see the light of day!

 

Yada, yada, yada, all well and good BUT jornal bearings need attention! Someone needs to be sure that the oil level is sufficient to wet the
pads or stuffing and that said pad or stuffing is in good shape and contacting the axle face. More of that dreaded enemy to the rail industry-
the labor force! Having sold or applied thousands of hydraulic pumps with journal bearings I can say that they work and work well when kept
supplied with clean fluid and not subjected to side loading. We had applications for pumps with journals where we had to filter the bearing
feed to keep the pump alive, thank goodness freight cars were less demanding.

Out 

> In recent years the retarder yards are built with
> grade of 0.04% down grade as that is the slope
> that a roller bearing car will roll at same
> speed.

80 pounds of gravity will keep a 100-ton car rolling?
But maybe 24 pounds won't keep a 30-ton car rolling --
so the computer lets the empty out of the retarders faster?

I serviced high speed three and four stage turbine air compressors with "squeeze film" or hydrodynamic bearings. The oil was injected into the plain bearings at 800 psi and the high speed impellers spun at 90,000 rpm with .5 mil. runout. It alarmed at 1.5 mil and shut down a 2 mil runout. We would run them continuously for 360 days and then take it down for service, oil and filter change, etc. Impressive when it works, catastrophic when it fails.

wcamp1472 Wrote:

> 2.  Rotating Roller bearings leave higher
> temperature readings on thermal-monitoring
>     "Hot Box" Detectors.  In today's world, hot
> box detectors are obsolete.
>
>     [ The old hot box detectors have been, and
> are being, replaced with 
>       high, audio frequency bearing monitors.
>    These listening devices 
>       are better at finding and predicting
> all rolling-element bearings that
>      are in distress, and danger of causing a
> derailment
>
>       The old hot box devices were spaced too
> far apart, compared to 
>       how quickly a failed, squealing roller
> bearing will overheat and 
>       burn-off the axle end before the next
> "heat detection" device is reached.
>       So, high audio frequency detectors are
> the only way to catch rollers in distress,
>       well in advance of actual squealing or
> running hot.]
>
This is not entirely true, audio frequency is certainly not the only way to catch rollers in distress.  I would argue that traditional HBD's are not obsolete based on the use of the data they provide.  UP has used bearing temperature trending along with software algorithms to use standard HBD's to detect bearings going bad long before they reach a temperature critical enough to fire off a HBD alarm.  UP monitors bearing temperature readings at each HBD, and monitors the "trend" of the temperature of each bearing on the train to watch for telltale rises in temperature, long before the bearing reaches a threshold high enough to set off the HBD alarm.  The software development took lots of work to normalize for occasional missed bearing reads as a train passes from detector to detector so that the individual bearing reads can be matched together to monitor the temperature trend for each bearing in a consist.  Once a given threshold is reached, the dispatcher is alerted and they will contact the train crew with instructions on checking and/or setting out the offending car as necessary.  Train crews (and railfans listening in) are totally unaware of all this happening in the background because they never get a traditional HBD alarm.  This is just one part of UP's arsenal of thermal, accoustical and even visual car monitoring systems that all work together in monitoring freight car health.  Temperature trending has long ago proven itself as extremely effective in preventing burnt-off journal issues.  Audio detectors are certainly a very vital part of bearing condition monitoring, but in the past they were relatively expensive, espcially compared to the existing HBD network that was already providing data just waiting to be mined.

How do I know this?  I was the software project manager at UP for this system when it was developed roughly 20 years ago.

MAB

Thanks for a good answer. Do all the big RRs do that now?

timz Wrote:
-------------------------------------------------------
> Thanks for a good answer. Do all the big RRs do
> that now?

I presume they do.  We worked with other railroads on the idea of passing that data along for cars being interchanged.  The thought was that by exchanging data, you wouldn't have to "start green" with a car that may already be having problems, i.e. you wouldn't have to wait for it to pass 3 or 4 detectors before you found a bearing going bad.  In a situation like that, you'd probably already have a burnt off journal before you even got to that third or fourth detector.  I know I presented many times at various industry meetings on the concepts and our work on the project.  It was deemed a major enough safety issue that normal thoughts of competitive advantage weren't much of a consideration.

MAB

Another thought on this from an engineer's perspective, New bearings sometimes run hotter and will set off HBD. I have had it happen, and CSX rules required a set out after the second stop. The conductor would leave a tag on the offending axle so another crew would know it was already tagged. On friction bearing, the oil would run off the top of the axle and it would rust if not moved for several months. I have caught an extra switch engine that all we did for 12 hours was pull and res

Hit wrong button, previous post continued, pull and resport cars at a us government facility in Savannah. They were all friction bearings and had high had brakes