Gradients fixed?

You are here: Home > Forum > Simulations > Australian Sims > Moss Vale NSW > Gradients fixed?

Page 1 of 1

Gradients fixed? 07/07/2020 at 13:19 #129210
Joe S
Avatar
128 posts
Hello,

In the recent huge number of updates, I've seen a few threads saying there has been an update to Moss Vale. Great stuff!

Does this mean though, that the gradients have been fixed? I seem to remember the last time I played it a few months ago, that a freight train would arrive at a loop a good hour or two early, rather taking some of the forward planning element out of it?

Cheers,
Joe

Log in to reply
Gradients fixed? 07/07/2020 at 17:36 #129214
GeoffM
Avatar
6274 posts
I believe so. Bear in mind, though, that trains are very generously timed along there.
SimSig Boss
Log in to reply
The following user said thank you: Joe S
Gradients fixed? 07/07/2020 at 21:04 #129236
Joe S
Avatar
128 posts
Understood, thanks Geoff!
Log in to reply
Gradients fixed? 26/03/2021 at 05:35 #138104
drew
Avatar
69 posts
I’ve run the timetable through a number of times now and I think the only glaringly problematic train performance on a grade is on the 1 in 30 section between Unanderra - Dombarton - Summit Tank, on the down (uphill). I appreciate that SimSig isn’t meant to be a train simulator, and that getting it to try and be one isn’t to be reasonably expected.

Almost universally the regular timetabled freight trains ascending the Illawarra mountain from Unanderra to Moss Vale are both bulk commodity trains, and are in their empty state. A train like this, if it has empty wagons but locomotive power for the loaded state, is well over powered on any grade it will encounter (even the perilous 1 in 30’s). I find it odd then that these trains seem to settle at about 14km/h going up the mountain. They would realistically be capable of running to the speed boards. Also a DC traction motored locomotive (like an NR or 81 class used in the SimSig timetable) has a full load balancing speed higher than 14km/h. Low to mid 20’s would be more prototypical. At full load (which the simulated trains are not), at full throttle, at 14km/h, the traction motors are cooking. AC locos might be a little more tolerant of that kind of abuse, but the still wouldn’t plan to load them down to that point. Of course the views up the mountain are very scenic, we could assume the drivers are ambling to take it all in, except they’re losing time in the process.

I don’t know how you’d fix it, if it’s fixable. Moss Vale in SimSig is an amazing achievement, and I’m very grateful. This is just something slightly off that I’ve noticed.

Log in to reply
Gradients fixed? 26/03/2021 at 16:41 #138106
GeoffM
Avatar
6274 posts
What is the weight setting on those trains? It might be inappropriate and can be easily fixed.
SimSig Boss
Log in to reply
The following user said thank you: drew
Gradients fixed? 26/03/2021 at 16:51 #138107
bill_gensheet
Avatar
1309 posts
Some of the train types could be a bit off.
9833 is called 'empty' but is set as Low / Heavy in the train characteristics.
Do trains like 9231 behave better ? It is set as Medium / Normal

Maybe amend a train before entry as 'High / Light' and see how it runs.

Bill

Log in to reply
Gradients fixed? 29/03/2021 at 05:34 #138205
drew
Avatar
69 posts
That could be the answer. I think they’re delineated by empty and loaded but even the empties are classed as heavy in the default timetable. It’s about power to weight ratio. I would try medium first, and see what we get.

I have no experience editing timetables. I guess one would start by making a copy of the default, in case it gets badly mauled. An individual train has to be edited before it enters the sim for changes to be effective, yeah?

Last edited: 29/03/2021 at 05:35 by drew
Reason: None given

Log in to reply
Gradients fixed? 29/03/2021 at 08:11 #138206
postal
Avatar
5189 posts
Online
drew in post 138205 said:
An individual train has to be edited before it enters the sim for changes to be effective, yeah?
Depending on what you are trying to do, you could also edit the train type which would affect all of the trains using that consist which have still to enter. It may be that two train types are required having the same consist but different acceleration rate and weight, one for the loaded trains and the other for the returning empties.

“In life, there is always someone out there, who won’t like you, for whatever reason, don’t let the insecurities in their lives affect yours.” – Rashida Rowe
Log in to reply
The following user said thank you: drew
Gradients fixed? 29/03/2021 at 11:02 #138210
bill_gensheet
Avatar
1309 posts
Yes, certainly save as 'my edit' or something like that, and take a copy of the base timetables elsewhere safely out the way.

To test, it may be better to:
Sort by entry location
Select all those entering at Unanderra
Save as a new timetable *to desktop* - it will only contain those trains and the relevant train types, no rules.

Load a blank new timetable, merge the desktop wtt file and save as say 'Unanderra timing test.wtt'
Convoluted but avoids timetable name clashes that you get otherwise.

Also I suggest adjusting the train types rather than each train. If you edit the train characteristics it becomes a 'custom type'.
The timetable can then be run at 10 or 22x sim speed to see how they keep time.

Then back to the main timetable and alter the train types there. Worth a check to see if any are used elsewhere in the main timetable though.

Log in to reply
The following user said thank you: drew
Gradients fixed? 31/03/2021 at 11:31 #138254
drew
Avatar
69 posts
I was pleasantly surprised that I both managed to follow the steps described, and the results from a quick trial were pleasing. So thank you.

I’ve only run an empty Tahmoor coalie and 9227 (the empty steelworks limestone) up the hill so far. Diesel (Light) set for both, no other changes. 9227 runs the speed board now, and keeps time pretty closely. I’ve consulted a contact closer to the source and the train type could use some editing to get it closer to a typical 9227 (wrong wagons, atypical length, max speed empty should be 80km/h not 90km/h), but that’s for another day.

The empty coalie with the same settings except the length (300m longer), and with the correct 80km/h top speed, settles at a slightly lower speed (27km/h instead of 30km/h) on the ruling grade. That’s much better performance than when it was “Heavy”. I just wonder why they’re not doing the same thing, guessing it factors the length.

At this stage I’m suggesting that setting empty bulk freight with loaded train Loco power to (Light) is more prototypical than the default (Heavy) was.

Log in to reply
The following user said thank you: bill_gensheet
Gradients fixed? 31/03/2021 at 12:00 #138255
Steamer
Avatar
3913 posts
drew in post 138254 said:


The empty coalie with the same settings except the length (300m longer), and with the correct 80km/h top speed, settles at a slightly lower speed (27km/h instead of 30km/h) on the ruling grade. That’s much better performance than when it was “Heavy”. I just wonder why they’re not doing the same thing, guessing it factors the length.
Length is ignored. Do they both have the same acceleration/brake characteristic, as well as weight?

"Don't stress/ relax/ let life roll off your backs./ Except for death and paying taxes/ everything in life.../ is only for now." (Avenue Q)
Log in to reply
Gradients fixed? 01/04/2021 at 07:26 #138262
flabberdacks
Avatar
575 posts
Most of the freight consists on Strathfield and Moss Vale were harvested from what ran on a couple of days in October 2017, as best as we could learn. Beyond that, with regards to Moss Vale, we've decided to add a few more things for the player to do (trying to keep the right hand side of the sim interesting) but without accurate consists, and in a couple of cases, totally incorrect run numbers.

Timetable testing with regards to freight is ensuring that it keeps to the booked time between each location (doesn't become late unless the player makes it late), and doesn't stuff up the rest of the train plan while doing so. If it ticked both of those boxes we put it in.

Hopefully everyone is enjoying the sim as it stands at the moment, and you are more than welcome to have a shot at writing your own timetables, if you have enough information and the time to spare.

Good luck with it

Last edited: 01/04/2021 at 07:29 by flabberdacks
Reason: None given

Log in to reply
The following user said thank you: drew
Gradients fixed? 02/04/2021 at 06:31 #138289
drew
Avatar
69 posts
I might, if I ever get the time, have a go at a Moss Vale time table, but no promises. More likely I’d keep most of what’s there and just amend the few bits that jump out at me as slightly off. The default timetable is playable and enjoyable despite (or sometimes because of) it’s potential inaccuracies.

Back on the Illawarra mountain, I’ve observed that train characteristic maximum train speed causes variations to the speed attained on the ruling gradient. It was indeed nothing to do with length.

Two identical trains, Diesel (Light) (low acceleration & braking). If the Max speed of the train is 80km/h, it’ll do 26km/h on the steepest bits between Dombarton & Summit Tank. If the Max Speed is 100km/h, we get 33km/h.

In reality, the maximum allowable speed for a freight train is determined by the lesser of a number of variables. Most typically by wagon type and restrictions on axle loads. It’s administrative, not a reflection of the potential performance from locomotive power available. Trains of the same power and weight can have different maximum speeds, but will behave identically slogging up a hill at less than their administrative speed limit. SimSig isn’t doing that (and I’m not complaining)(much).

I can accept that SimSig isn’t focused on train performance. I still think setting well powered trains to (Light) is closer to reality than (Heavy) was. Perfect, no. Good enough unless the boffins want to recode the sim, absolutely.

Log in to reply
Gradients fixed? 02/04/2021 at 08:35 #138291
postal
Avatar
5189 posts
Online
drew in post 138289 said:
Two identical trains, Diesel (Light) (low acceleration & braking). If the Max speed of the train is 80km/h, it’ll do 26km/h on the steepest bits between Dombarton & Summit Tank. If the Max Speed is 100km/h, we get 33km/h.
A long time ago, Clive explained to me how the acceleration algorithm worked. He is aware that it is not perfect and it is on his (rather large) to-do list to revise the calculations to try and reflect the actuality with greater accuracy. If I've remembered his explanation correctly the acceleration rate is calculated using an upper limit of the maximum speed so that the rate of acceleration will start off quite sharply and then taper down as the speed approaches the maximum speed of the train. The different acceleration rates selected for the train type alter the slope of the curve but not the underlying principle. Gradients are also factored in to reduce the acceleration rate going uphill and ease it down the other side.

That means that there are cases in SimSig where the real life train performance would not match this approach. For example in the Carlisle 1979 TT there are a lot of double-headed Freightliners with a 75mph speed limit. In real life a single loco could handle that load but with a leisurely acceleration rate. The trains were double headed as a railway decision to improve the acceleration and hill-climbing thus freeing up line capacity. The double-headed trains would have more than enough power to exceed the speed limit for the consist except on the steepest uphill gradients. That is not reflected in SimSig as any acceleration rate is set so that the resulting velocity tends towards the speed limit of the consist. The fudge is to give a higher acceleration rate so the double-headed Freightliners are shown as having a Inter-City passenger acceleration rate and work to somewhere near sectional times.

There is now a new potential way to achieve adherence to sectional times because of the facility to change train type on the fly in the TT. It would be possible to create identical consists with different acceleration rates or weights and at suitable locations in the TT (like maybe Milnthorpe or Tebay for Down trains in Carlisle TT) the consist would be changed to one with a higher acceleration rate or lower weight for the long and steep climbs of Grayrigg and Shap banks with a reversion to the original train type at Shap Summit.

“In life, there is always someone out there, who won’t like you, for whatever reason, don’t let the insecurities in their lives affect yours.” – Rashida Rowe
Last edited: 02/04/2021 at 08:35 by postal
Reason: None given

Log in to reply
The following user said thank you: drew
Gradients fixed? 02/04/2021 at 13:20 #138293
bill_gensheet
Avatar
1309 posts
postal in post 138291 said:

The trains were double headed as a railway decision to improve the acceleration and hill-climbing thus freeing up line capacity. The double-headed trains would have more than enough power to exceed the speed limit for the consist except on the steepest uphill gradients.
Also a factor is wet rails, a single Bo-Bo electric might well not get up Shap or Beattock if it was raining.

Bill

Log in to reply
The following user said thank you: postal
Gradients fixed? 03/04/2021 at 20:22 #138340
clive
Avatar
2736 posts
postal in post 138291 said:

A long time ago, Clive explained to me how the acceleration algorithm worked. He is aware that it is not perfect and it is on his (rather large) to-do list to revise the calculations to try and reflect the actuality with greater accuracy. If I've remembered his explanation correctly the acceleration rate is calculated using an upper limit of the maximum speed so that the rate of acceleration will start off quite sharply and then taper down as the speed approaches the maximum speed of the train. The different acceleration rates selected for the train type alter the slope of the curve but not the underlying principle.
That's about right. And, yes, it's still on the list.

postal in post 138291 said:

Gradients are also factored in to reduce the acceleration rate going uphill and ease it down the other side.
That bit does work correctly: a 1 in 100 gradient alters the acceleration by 1% of gravity and so on.

Log in to reply
The following user said thank you: postal