[gpz]

Genma Saotome, on 20 April 2014 - 10:58 PM, said:

Minor nit: The correct term is Solid Bearings. Friction bearings came into common use only after the Roller Bearing salesman spat it out while making their sales pitch.

My specific English to Hungarian engineering dictionary contains the term "sliding bearing", "friction bearing" and "sleeve bearing". The term "solid bearing" is nowhere mentioned.

Edit: After some more search I found the term "solid bearing" only in a totally different meaning: at civil engineering the "full length support" is called as "solid bearing". :)

[copperpen]

Americans speak a different version of English. :)

[mozdonyos]

Thank you for your work guys!
Now completely realistic with the "Low" command.

Again, thank you, sincerely, a Hungarian enginedriver! :)

[Genma Saotome]

All bearings have friction. How much varies by type.


FWIW I've been an active member of each successive version of the Steam Era Freight list since the mid 1990's. The issue of the correct term to use has come up several times and in each case the consensus of opinion was the correct term is Solid Bearing because that was the term used by the Car Builders and their professional publications.

The official standard bearing the American industry followed was documented by first by the American Railroad Association (ARA) and endorsed again by its successor, the American Association of Railroads (AAR). The ARA standards were pretty much whatever was in common use at the time (early 20th century). Each railroad was a full voting member of both organizations and so standards didn't set set w/o a clear majority of railroads saying so. Prior to the widespread acceptance of the roller bearing (in the 1950's) the standard for use on Journals specified the metal that was fit to the truck side-frame and a solid brass end on the axle, to be lubricated by oil. So technically you could say it was an oil bearing as that was, in fact, the material between the two objects.

When roller bearings came into use they replaced the oil and from what I understand what was on the end of the axle was no longer as important.

Now all of this may be unique to the American railroad scene... I know nothing about railroads outside of North America and so it's certainly plausible that other countries used completely different terms, all of which would be, in some sense, equally valid choices to use in the .wag files. That said, what I suggested is at least provided to me by both engineering reference material and persons whose opinions on 20th century freight cars I respect.

[eugenR]

jorgen, on 21 April 2014 - 09:24 AM, said:

Instead for roller bearings I think the new value Low as default for wagon and engines.

LOW will be the default value,
and what are the "not default-value's" for higher friction to writte in the *.eng, *.wag - files?

[copperpen]

To get the best out of the ORTS train friction section you need to add 4 lines of data to each wag and eng file,

ORTSDavis_A ( 482.68 )
ORTSDavis_B ( 11.1877 )
ORTSDavis_C ( 0.4952 )
ORTSBearingType ( Roller )

The three ORTSDavis lines are taken from the Davis calculation of FCalc which can be found in the trainsim.com library. The fourth line is where you put the bearing type.

[Genma Saotome]

copperpen, on 21 April 2014 - 11:37 AM, said:

Americans speak a different version of English. ;)

I give up. Entirely.

[gpz]

Picking up an old thread and Dave's idea, it would be worth looking at FCalc's "FCalc 2.0 Variable Parameters.txt" file, which contains all the Davis A-B-C calculation equations for different type of vehicles, and incorporating them into OpenRails code. These equations are as simple as e.g.:

Type of Rolling Stock              A                            B                   C
Standard Roller Bearings:
Locomotive           (6.3743*weight + 128.998*axles)      (0.32905*weight)  (0.57501*c_drag*area)
Steam Engine         (6.3743*weight + 128.998*axles + D)  (0.32905*weight)  (0.57501*c_drag*area)
Leading Railcar      (6.3743*weight + 128.998*axles)      (0.65811*weight)  (0.57501*c_drag*area)
Middle Railcar       (6.3743*weight + 128.998*axles)      (0.65811*weight)      (0.08146*area)
End Railcar          (6.3743*weight + 128.998*axles)      (0.65811*weight)      (0.14375*area)
Passenger Car        (6.3743*weight + 128.998*axles)      (0.32905*weight)      (0.08146*area)
Standard Freight Car (7.3550*weight + 80.625*axles)       (0.16453*weight)      (0.11979*area)
Empty Hopper         (7.3550*weight + 80.625*axles)       (0.16453*weight)      (0.35938*area)
Autorack             (7.3550*weight + 80.625*axles)       (0.16453*weight)      (0.16771*area)
TOFC                 (2.9420*weight + 88.964*axles)       (0.10968*weight)        (4.45168)
Spine Car            (2.9420*weight + 88.964*axles)       (0.10968*weight)        (4.11780)
COFC(Double Stack)   (2.9420*weight + 88.964*axles)       (0.10968*weight)        (3.89522)

We need 4 parameters: weight, axles, area, c_drag. And of course the type of vehicle and bearing. Although it is not straightforward yet, how to differentiate the various freight cars, but as the first step the standard freight car equation could be used.

Looks like the fact if a car is a leading or a trailing one is also important. Most probably the "C" equation for the locomotives implies that they are leading ones, so the equation might be invalid for helper locomotives. Thus a multiplier should be used for indicating the leading or trailing "property" of the vehicle, applied to passenger and freight cars as well as for locomotives and railcars. ("Middle" locomotives should use the same equation as "middle railcars", and "end" passenger cars/locomotives should use the same as "end railcars".)

By this logic the ORTSDavis_A, ORTSDavis_B and ORTSDavis_C eng parameters should be used just as overrides, for those wanting to fine-tune the values themselves, and not for general use. Anyway, "C" value is inaccurate in this way, because it is much higher for helper locomotives than is needed (7 times higher for middle helpers and 4 times higher for trailing helper, 1.8 times lower for last car).

[gpz]

Some pre-assumptions could be made for avoiding the need for a user to enter any data for Davis equation calculations:

weight: This is available in eng/wag Mass() function.
axles: Reliable data isn't available in eng/wag unfortunately. A new parameter e.g. NumberOfAxles() is needed here, as Dave suggested. A general default could be 4 here.
area: It can be calculated from eng/wag Size() parameter.
c_drag: Defaults to 1, but a new parameter should be defined here.

For eng files Type() parameter tells if the vehicle is steam or other engine. To determine if it is a railcar, the PassengerCapacity() parameter can be used.
For wag files Type() tells if the vehicle is a passenger or freight car.

"C" calculation should use the passenger formula for engines too, and the value of the leading vehicle must be multiplied by 7.06*c_drag, the value of the trailing vehicle must be multiplied by 1.76.

There is the parameter for bearing type already available, should default to low-torque roller bearing type.

And voilà, we eliminated the need for entering ORTSDavis_A, ORTSDavis_B and ORTSDavis_C parameters, thus the users don't need to use the third-party FCalc software for calculating them.

[steamer_ctn]

gpz, on 23 April 2014 - 08:36 AM, said:

And voilà, we eliminated the need for entering ORTSDavis_A, ORTSDavis_B and ORTSDavis_C parameters, thus the users don't need to use the third-party FCalc software for calculating them.

I agree that it would be nice to eventually achieve this goal.

However, to reach this point, I suspect that some additional research, thinking and investigation will be required.

So perhaps it maybe a little while yet.

Cheers