[Traindude]

Hello. I am not sure if this is worth simulating in ORTS, but the type of valve gear a steam locomotive has affects its performance. For example, locomotives with Stephenson valve gear have a variable lead--the lead increases as the reverser is moved closer to its neutral position.

On locomotives with Walschaerts or Baker valve gear, the lead is either constant or "crossed" (greater in reverse gear than in forward gear).

Another touch for OR-exclusive steam models is to see the valve gear move when the reverser is moved--for example, the radius rod in the Walshcaerts gear moves up and down as the reverser is moved.

I know I'm getting ahead of myself, but I learned this through the ITC steam textbooks and Dockstader's valve gear sim programs, and I hope it could be simulated.

[captain_bazza]

Quote

Another touch for OR-exclusive steam models is to see the valve gear move when the reverser is moved--for example, the radius rod in the Walshcaerts gear moves up and down as the reverser is moved.

OUCH! :D That one crops up from time to time....it requires a very high level of modeling and engineering skills to get that motion right. The modeler has to model the parts involved and get it right, then animate the action. Considering that most, if not all, current modelers are not steam locomotive engineers and that most use model plans, then it's unlikely such an n'^th degree of accuracy could be achieved. That is not to say that it could never be done, but would it be really worth all the extra effort involved, when it's going to be hardly visible from anything other than very close up.

There is no direct connection between the visual model and the physics model. Long story....but the visual model could just as easily be Mickey Mouse as far as the actual physics code was concerned. :good2:

Cheers Bazza.

[B & O GUY]

Ooh! That should be worth a few thousand key frame animations for the model builder. I've done 16 key frames on 22 parts for a Baker valve gear for my Erie Berkshires with quartering. Not to mention the 4 times that were done for experimentation trying to figure out the geometry. Now just imagine doing that for every position of the reverser both in forward and reverse with changing geometry for every single move of the handle. I did 352 key frames with the Berk and thought I'd never finish.

I think MSTS and probably Open Rails and the Engine file allows for notches of 75 to 100 positions of the reverser in both forward and reverse.

This work would fall on the model builder and his 3D programing as much or more than the Open Rails development team. It would also require the modeler to know the engineering of the real locomotive. And if you think about it. No two engineers would use the exact same notches for every situation. I'll leave it for the new generation to figure out. :)

You could do it in your spare time. :crazy:

Allen :D

[SVRy_Steve]

It would be neat to see, but I suspect, not worth the effort! You'd have to be seriously obsessed to tackle it. Maybe the power reverse on engines so equipped, could move though.

Steve

[steamer_ctn]

 Traindude, on 29 March 2014 - 02:13 PM, said:

I am not sure if this is worth simulating in ORTS, but the type of valve gear a steam locomotive has affects its performance. For example, locomotives with Stephenson valve gear have a variable lead--the lead increases as the reverser is moved closer to its neutral position.

It certainly would be a nice inclusion, but the challenge is to build a simple mathematical model for it. I have seen some models, but they seem quite complex.

Any thoughts?

Cheers

[captain_bazza]

Do a search on youtube for Walschaert valve motion.

Cheers Bazza

[Lindsayts]

 steamer_ctn, on 30 March 2014 - 11:41 AM, said:

It certainly would be a nice inclusion, but the challenge is to build a simple mathematical model for it. I have seen some models, but they seem quite complex.

Any thoughts?

Cheers

Now I have not look at this for quite some time but in the steam code were two look up tables called ForceFactor1 and ForceFactor2. The output of these tables were in effect the overall cylinder effectiveness given the current cuttoff. One should have little difficulty simulating the over all RESULTS of different valve gears if the tables worked.

Unfortunately when I tried them there were two problems. The default data although reasonable produced results that for me were unsatisfatcory, the second problem the data was user definable, this did not work. I tried for quite a while to get user defined data to work, the new data did change the way the loco worked but not in any predictable manner. Both these could have been a single bug.

If required I can give more info on this issue.

Lindsay

[Lindsayts]

 captain_bazza, on 29 March 2014 - 07:34 PM, said:

OUCH! :) That one crops up from time to time....it requires a very high level of modeling and engineering skills to get that motion right. The modeler has to model the parts involved and get it right, then animate the action. Considering that most, if not all, current modelers are not steam locomotive engineers and that most use model plans, then it's unlikely such an n'^th degree of accuracy could be achieved. That is not to say that it could never be done, but would it be really worth all the extra effort involved, when it's going to be hardly visible from anything other than very close up.

There is no direct connection between the visual model and the physics model. Long story....but the visual model could just as easily be Mickey Mouse as far as the actual physics code was concerned. ;)

Cheers Bazza.

Instead of being animated, the movement of the linkages, including the connecting and coupling rods could be mathematicly modeled. I do have a steam engine simulation program which actually already does this, so it can be done. I am currently playing around with the maths to do this in my own program.

I will post the results if I have any success.

I assume this would take a good deal of work off locomotive builders.

Lindsay