(My below comments are all in the context of using "roller" type bearings in the OR code)
As for as Mullan goes: The decision for us to NOT use the Davis formula friction block was, in fact, because of the fact that when we did use the Davis block, it created what we (the TrainSimulations team) perceived as unrealistically high starting friction forces (0 to 5 MPH). The Davis equation itself may be performing fine, but the fact is that for whatever reason, at speeds below about 5 MPH, the OR code uses some other linear equation to ramp the friction up. Going off memory here, I believe we noted that when starting a loaded coal train from a stop on the 2.2% at Austin on Mullan Pass, each car was seeing in the neighborhood of 2,000 lbs of friction, vs. I think about 1/4th of that when using the older MSTS friction block. The huge starting friction forces made it virtually impossible to start a prototypical train consist without breaking couplers. It should be noted that Mullan is a place where coupler forces are pushed to the max in real life, but they still manage to start heavy unit trains from a stop on this grade on a routine basis. Couplers break, but often that is a result of poor train handling and/or coupler knuckles that have already been weakened due to other factors.
As far as I know, the additional resistance at 0 MPH in the OR roller bearing friction code (which ONLY goes into affect when using a Davis block) was based on an AREMA claim that roller bearing resistance is significantly higher at a stop. Peter, I believe I emailed you some AREMA manual excerpts from a recent AREMA manual where this statement has been changed to say that with roller bearings, there is "little or no" additional resistance when starting from a stop. This is certainly NOT true with solid bearings (aka "friction bearings").
At any rate, I'm assuming that the 5 MPH transition point between the starting resistance equation and the Davis equation is not based on any real research, but instead on one of the OR coder's generic interpretation of what "low speed" means.
As for as Mullan goes: The decision for us to NOT use the Davis formula friction block was, in fact, because of the fact that when we did use the Davis block, it created what we (the TrainSimulations team) perceived as unrealistically high starting friction forces (0 to 5 MPH). The Davis equation itself may be performing fine, but the fact is that for whatever reason, at speeds below about 5 MPH, the OR code uses some other linear equation to ramp the friction up. Going off memory here, I believe we noted that when starting a loaded coal train from a stop on the 2.2% at Austin on Mullan Pass, each car was seeing in the neighborhood of 2,000 lbs of friction, vs. I think about 1/4th of that when using the older MSTS friction block. The huge starting friction forces made it virtually impossible to start a prototypical train consist without breaking couplers. It should be noted that Mullan is a place where coupler forces are pushed to the max in real life, but they still manage to start heavy unit trains from a stop on this grade on a routine basis. Couplers break, but often that is a result of poor train handling and/or coupler knuckles that have already been weakened due to other factors.
As far as I know, the additional resistance at 0 MPH in the OR roller bearing friction code (which ONLY goes into affect when using a Davis block) was based on an AREMA claim that roller bearing resistance is significantly higher at a stop. Peter, I believe I emailed you some AREMA manual excerpts from a recent AREMA manual where this statement has been changed to say that with roller bearings, there is "little or no" additional resistance when starting from a stop. This is certainly NOT true with solid bearings (aka "friction bearings").
At any rate, I'm assuming that the 5 MPH transition point between the starting resistance equation and the Davis equation is not based on any real research, but instead on one of the OR coder's generic interpretation of what "low speed" means.
Comment