Dodgy physics

Discuss almost anything about RailWorks.

Re: Dodgy physics

Unread postby TrainMaster1 » Fri Nov 18, 2011 10:39 am

There was indeed an emergency brake valve located in the caboose only for use when the conductor saw something ahead that would cause the need to use it. We even had this is some older passenger cars in our area again for use by a conductor when things just were not going right that day. It was located in the vestibule at the opposite end from where the engineer was running at the time. Actually both ends of the car had one so when they changed ends a crew member other than the engineer had access to the handle.

Remember the conductor is in charge of the train and most of the time spent his day in the caboose away from the head end. The days of head brakeman, fireman, rear end brakeman, conductor and engineer are long gone just like many cabooses in the picture above.

Nick
TrainMaster1
 
Posts: 111
Joined: Thu Feb 17, 2011 11:19 pm

Re: Dodgy physics

Unread postby Sandman2 » Sat Nov 19, 2011 1:01 pm

As a recently retired Locomotive Engineer and someone fairly new to Railworks I've really enjoyed this post. I think it's only a matter of time until Railworks Creators focus on an acccurate physics model. I know it's possible as I ran a Simulator 30 years ago and found the physics to be very accurate of how the handled on a specific subdivision. Speaking on Accurate Railroading, the first part of my career I worked in "Dark" territory in which Train Orders were used in the movement of trains. I found this method very challenging and I would love to recreate it on Railworks. I realize there would be a small minority who would be interested but I feel it would take Train Simulation to the next dimension. Some of the tools required would be a simplified rule book, a Subdivision Timetable and an AI you could manipulate to handle instructions. I envision a Scenario where you're given these tools plus your Train Orders and you would have to decide if your train can leave the initial terminal, how far you can go and whether you take the siding or or hold the main. Those who have used Train Orders will understand when I say that you cannot learn them by reading a book but must experience them real time. I also think this is a way to keep part of Railroad History alive before all us old timers are gone. I hope I haven't strayed to far off topic.
Sandman2
 

Re: Dodgy physics

Unread postby TrainMaster1 » Sat Nov 19, 2011 8:22 pm

I too enjoy dark territory and we do run "dark" territory routes live in our operating sessions. Granted you won't see bunches of AI trains because every train in our sessions have a live crew. Even more challenging are combination routes which have CTC right along side TWC or certain Canadian Pacific routes which use their version of CTC where crews call the DS before entering any siding. Even crews call this "Poor Man's CTC". I agree that this indeed has been an excellent thread. Thanks to everyone who shared their knowledge.

Check out our forum section right here at RWA under VORA operating sessions if you are interested in prototype operations

On 11/26 we will be recreating an actual day from November 1976 on the D&H in northern NY. Alcos, Geeps and SD-45s in EL, D&H, NW, LV and B&M paint will be running. Let me know if you are interested in learning more.

Nick
TrainMaster1
 
Posts: 111
Joined: Thu Feb 17, 2011 11:19 pm

Re: Dodgy physics

Unread postby Kali » Sat Nov 19, 2011 8:33 pm

Well, really, what are signallers ( not signals! ) but the embodiment of train orders? in the really distant past ( 1830s-1860s ) trains were run over here on the time-lapse principle; while you'd have a bobby controlling a track block, he'd determine if a train could go through entirely based on a) how long it was since the last one went by, and b) the timetable. Not quite the same as controlling your own path, but the principle is however the same, there's no absolute knowledge of a train's position.

Now... railworks rather likes having one train in a signal block; it doesn't mind permissive signalling for the player train, but it will refuse for the AI ( and that's a good thing, because the AI has no idea there's another train in front of it until it runs into it! ). So... with some careful burying of permissive signals you could get something that *looked* like the train order system from the player train, even if it's actually working on block principles. If you really wanted you could have canned radio messages for signal aspects assuming there's some cab to cab radio, or maybe you could have an in-cab display of a new train order you just picked up, or probably a couple of other ideas I haven't thought of yet.

The big issue in a train order simulation is just what you said - you can't really go by rules. Computers are all about rules, adaptive decision making is a tricky business.
Kali
 
Posts: 1600
Joined: Mon Mar 14, 2011 1:00 am
Location: England-by-Sea

Re: Dodgy physics

Unread postby Csxgp38-2 » Sat Nov 19, 2011 9:03 pm

I been inside a more modern caboose where it actually is marked "Brakeman's emergency brake valve" next to a lever, so yeah, they have a brake valve like what the others have been saying.
-Harrison
CSX railfans are how tomorrow moves!
Future owner of HSM Co Transportation- Rolling into the future.
"If I could return to those boyhood days of mine, and that green light on the Southern, Southern railroad line."
Csxgp38-2
 
Posts: 732
Joined: Sat Sep 10, 2011 9:18 pm
Location: Apex, NC- Seaboard Country

Re: Dodgy physics

Unread postby Sandman2 » Sun Nov 20, 2011 11:03 am

boleyd wrote:So, if the knowledgeable and experienced people could take an existing route but apply the signals and AI to make it function like a railroad that would become a basis for total immersion in an operation. Today, it is much like a basement railroad where trains run aimlessly around the same route. Interesting but far from real. You have an opportunity to take Railworks to the same level as the very large semi-pro, computer managed, model railroads by applying real day-to-day operations with multiple trains doing proper tasks. Running a 100 car coal train up a hill while raining is indeed fun. It is also fun to see if you can maintain a schedule of passenger stops. Or, it is interesting to shunt about making up trains for future delivery of goods. But if all this were combined into a managed scenario - wow, now you have something way beyond what is done today.


This is exactly what I'm aiming for. I'm in the process of creating a route which I had some experience handling trains. I have all the track gradients and it should be as close to prototype that can be achieved. It's a very challenging Subdivision with mountains, foothills and prairies. Is there many of you out there who are willing to contribute to the effort of bringing Train Orders to life? I'm able to create a Rule Book which I would base on the one I used in the early 1970s. I also have Timetables to use and could design on for use in Railworks. I would need help with creating the signage used in the route as these become very important in a realistic train operation. Also I'd require help with implementing Scenarios. manipulating the AI, creating Instructions,etc.

This is how I envision a Scenario Playing out. You're called for an Extra West at 1900. First, you check your Timetable to see what trains you need to protect against based on Superiority (Train Order,Class and Direction) Since you're an Extra train, you're Fourth Class and running in the inferior direction (West and North are classified as the Inferior Direction) therefore you need to protect against all Scheduled Timetabled trains. After checking the Register to see what trains have arrived, and reviewing your Train Orders you decide you must wait at the Initial Station for one train and then proceed to the first siding and clear the main. This is just the beginning....

There would be a steep learning curve but the end result would be worth the effort. Tutorials could be created to explain rules sand how the Train Order System works. Because of the open interputation of the system, I can see many discussions as I did spending many nights on a drag, passing the time. If you look to the future of Railsims this is a natural evolution especially as multiplayer evolves.
Sandman2
 

Re: Dodgy physics

Unread postby Kali » Sun Nov 20, 2011 1:16 pm

Mmm. Look at what you consider for train orders, and then look at what you have to set in the timetable view :). Anyway a discussion of train order operation would be helpful for me at least, because we've not run anything resembling train orders since about 1870 here and there's obviously bits I won't have any idea about - and really I don't think we've ever run anything like the full system at all since about 1830. Maybe I can dig out some info about the block system we use(d) here if anyone's particularily interested in that.
Kali
 
Posts: 1600
Joined: Mon Mar 14, 2011 1:00 am
Location: England-by-Sea

Re: Dodgy physics

Unread postby TrainMaster1 » Sun Nov 20, 2011 2:10 pm

Yes I guess you could create a system where you can run on your own and have an activity with AI traffic in it...

As a matter of fact that is one of the projects we are developing. This will cover 100% on class one railroads and appropriate for the rules of that road. This is not an easy project and is a direct result of our prototype operating sessions. Yes we are cracking ways of doing this in MSTS, RW and TRS but it is a long slow climb.

We will be releasing activities based upon our sessions in the future with full audio and yes you will see the other train. You and how well you run will be the only variable. This is not around the corner but about a full year away as to do this with the precision we bring to every operating session means we get all the details right.

The biggest hurdle is still the game engine which governs how well each sim represents the real world physics. Not much we can do there as we would have to create a ton of new equipment with the actual gauges that you would use to run (sorry but heads up displays and other keys are cheating and not running a railroad.) Our crews can run by knowing how to read an operating gauge (provided the loco has one). Playing a game is perfectly fine. We decided our group would be more interested in running a railroad each and every session.

So our format when unveiled will have accurate symbols, consists, motive power, track charts (so you can call signals like a real crew), track warrants (which you will have read to you via an actual recorded dispatcher) and the forms you can print to copy your warrants. We do all of this now every operating session. All we are doing is packaging into a format for our shy friends who would rather be a railroad of one rather than run with 20-30 other engineers live in full operating session.

You can be sure that when we release it we will do everything we can to have the physics and controls match the real world (within the limits of the game).

Projects like this are just one of the many benefits of being an active crew member in our group. Check out the VORA located here at RWA.

Nick
TrainMaster1
 
Posts: 111
Joined: Thu Feb 17, 2011 11:19 pm

Re: Dodgy physics

Unread postby Sandman2 » Sun Nov 20, 2011 2:30 pm

Here's a couple links to sites with info relating to Train Orders
http://www.gatewaynmra.org/tto.htm
http://www.theboykos.com/raildocs/cn/trainorders/
Sandman2
 

Re: Dodgy physics

Unread postby Kali » Sun Nov 20, 2011 3:02 pm

Cheers, sandman. Are those particularily detailed?

Ok, so most of the dash on the SD70s doesn't work, ignoring those for now. Are all the dials there on the SD40 now? I presume what's labelled train pipe pressure - or whatever the exact wording is - really ought to be EQ pressure ( actually how fast does EQ rise & fall on something of that ilk if you're in a freight configuration? ) is it missing any other important dials? I can't remember the GP9 too well, but how about that also? ( incidentally have either of those got an overload cutout? )

There are ways of getting better throttle response out of RW which would involve *having* to drive on the dials, because the HUD/F5 display will be showing something rather odd. If I can get this train brake using the handbrakes working then you'd have to rely on dials & feel for that also. If you're that interested in realism that you want this level of detail then I don't see any of this as a problem.
Kali
 
Posts: 1600
Joined: Mon Mar 14, 2011 1:00 am
Location: England-by-Sea

Re: Dodgy physics

Unread postby arizonachris » Sun Nov 20, 2011 3:29 pm

You're right, Kali, the SD70 gauges suck. Only one that seems to work is the speedometer. Good thing I cheat and use the F5 HUD/ F3 Driver's Guide, or I'd be lost in the fog. The new update of the SD40-2 is really nice, at least the one found on Rich Garber's new Ohio Steel 2 route. I dunno if he did it or RSC did it, but in cab controls all seem to work, as well as gauges! (that's the weathered UP SD40-2)
Ryzen 7 2700K, Asus Prime X570P, 32Gb DDR4, 2x 1Tb M.2 SSD's, RTX2060 6Gb, Occulus Rift
Win 10 Pro 64bit, keyboard/ mouse/ wheel/ pedals/ baseball bat
Security Coordinator on the Battleship Iowa
User avatar
arizonachris
 
Posts: 3956
Joined: Sun Mar 21, 2010 10:36 am
Location: Southern California

Re: Dodgy physics

Unread postby Dan » Sun Nov 20, 2011 3:47 pm

Having helped with an MSTS route which was set in dark territory it is worth noting some of the failures.

We tried buried signals but we found that this did not work or create the desired AI behaviour. What we found was that the best way to create a dark territory was to create it as such. We made extensive use of waiting points essentially.

The problems - it is not dynamic. For example, say you are a work train, say you take longer than normal in doing some switching, the place where the meet would take place remains static, the AI train will wait until you get there. In reality, if you were a work train and the AI train had a higher priority you might be told to wait for it to come to you.

We cheated a little by setting up the meets that either the player train or the AI train would have a 20 minute waiting period. If the AI train was in the hole we would give it a long period to sit there so that if a player drove a breakneck speed then the AI train would be there, and likewise, if the player was a grandpa driver the AI train would still be there.

If the player was to take the hole then we would give them plenty of time and tell them in advance - you will be sitting here until xx.xx

As I said what you don't have in this is the dynamic element. That is to say - if you are to meet at place A then you will always meet at place A, if you are early or late. Whereas in reality, if you were running late, early or the AI were running early or late, the meet might be moved to place B.

####

Nick - I think that the only way you can get a truly dynamic system at the moment is by making the way the person works through the scenarios dynamic. So rather than having say 1 2 hour scenario, you end up with 12 10 minute scenarios.

Driver A has a task - pick up wagons from yard. If the driver does it by X time then his next scenario - drive to point B passing AI train at point C is predicated on him doing this is based on his finish time. If he completes it later then he has the same task (drive to point B) but the set up is different so you meet the AI train at point D.

You would effectively have to create a spiders web of micro scenarios - and the succeeding scenario would be the dynamic element. However, by my reckoning your probably end up having to create about 1000 micro scenarios in order to have a dynamic 2 hour 'story'.

Again, like uber-physics it is something that would only appeal to a niche audience. And I can't help but think that some people who demand realism wouldn't actually enjoy realism if they actually had a game that replicated the real world in all its warts and glory.

For example yesterday the train ahead of us broke down (total power failure), so we were stuck behind it for an hour +. Fortunately there was a depot 20 minutes away so they could send something out to rescue the broken down train, but of course we had to wait for control to make the decision that the broken down train was a failed train etc etc. Now I can't imagine there being a huge queue of people wanting to play a game where your train gets stuck behind a broken down AI train and then has to wait for the AI controller to make the decision that the train has failed and then to send another AI train to rescue the broken down AI train.
Dan
 
Posts: 158
Joined: Fri Feb 27, 2009 7:46 am

Re: Dodgy physics

Unread postby Sandman2 » Sun Nov 20, 2011 4:23 pm

Kali wrote:Cheers, sandman. Are those particularily detailed?

Ok, so most of the dash on the SD70s doesn't work, ignoring those for now. Are all the dials there on the SD40 now? I presume what's labelled train pipe pressure - or whatever the exact wording is - really ought to be EQ pressure ( actually how fast does EQ rise & fall on something of that ilk if you're in a freight configuration? ) is it missing any other important dials? I can't remember the GP9 too well, but how about that also? ( incidentally have either of those got an overload cutout? )

There are ways of getting better throttle response out of RW which would involve *having* to drive on the dials, because the HUD/F5 display will be showing something rather odd. If I can get this train brake using the handbrakes working then you'd have to rely on dials & feel for that also. If you're that interested in realism that you want this level of detail then I don't see any of this as a problem.


Yes.the first site gives a very detailed overview of Train Order operations. With regards to the sd40 gauges, the left one is your equalizing reservoir (white needle) and Main reservoir. The right gauge is Brake pipe (white needle) and red is your independent brake. When you apply the brake the equalizing reservoir is used to measure the reduction and the brake pipe will reduce equally.When you release the brake even on a long train (unless there is abnormal leakage the equalizing will come back to normal preesure (90 lbs.) instantly. The main reservoir will drop and slowly increase as the brake pipe charges. You can have 90 lbs on the headend and 75 lbs on the tailend car . Railworks would do well to have a read out of the brake pipe pressure of the tailend car.
Last edited by Sandman2 on Sun Nov 20, 2011 4:48 pm, edited 1 time in total.
Sandman2
 

Re: Dodgy physics

Unread postby Kali » Sun Nov 20, 2011 4:43 pm

Ok, I can make that work properly even for default ones if the animation is there for the EQ needle.

WRT to AI & Dark territories; I have never tested this given I wrestle with the AI enough on a fully signalled route, but I am not sure how well the AI will behave without blocks. Really it likes having a link0 between it and any other train - and in fact that is all it listens to, the rest of RW signalling is at present all for the player's benefit - and it will happily just pile into another train in it's way. However... I can't see why train priorities won't work ( or at least they would work in RW2, RW3's dispatcher has gone backwards a bit ), iirc if you're crossing another train and you're late and it's the same or greater priority than your train, it will take your path. Or they did once, anyway. Trying to get something to cross the player's train is a bit hit or miss right now.
Kali
 
Posts: 1600
Joined: Mon Mar 14, 2011 1:00 am
Location: England-by-Sea

Re: Dodgy physics

Unread postby TrainMaster1 » Sun Nov 20, 2011 6:31 pm

Dan you would then be quite surprised to know that we do indeed see delays in sessions far longer than one hour when traffic, MOW, or equipment crankiness cause a single track railroad to become a zero track railroad. Case In point I was forced to hold a crew in an NS session for over two hours when a coal train ahead broke apart in a single track section. We got him back together but by then I had two north and two south ( including the Amtrak Carolinian) lined up for a long single section of track. Based upon train and NS priority, I instructed the dispatcher to hold the empty train until the higher priority traffic moved.

We have no 10 minute jobs in any of our sessions so not quite sure what a 10 minute scenario actually is. If you pick an assignment for instance this week coming, SU-10, one of the actual traveling switch jobs from November of 1976 that ran on the D&H, you will do the following:

1) Pull your engines off of Mohawk 7.
2) go to Mohawk 5 and pull the first six cars.
3) Take the engine and your six cars and go to mohawk 4
4) Double to the first 13 cars on Mohawk 4.
5) Back to Mohawk 5 and now grab the remaining 11 and your caboose.
6) Now head to Delanson Yard and stop on main 2 at JX Cabin signal.
7) Cut away your engines and pick up two cars off head end off Delanson 1.
8) Head back to train and make your double.
9) Now make cut between cars 8 and 9.
10) Back to Delanson 1 and spot cars 3 to 8 on Delanson for the Altamont Branch Local to work later.
11) Back to Train and double cars 1 and 2>
12) Now continue south to Central Bridge. Remain north of team track siding switch at Central Bridge.
13) Cut between cars 4 and 5
14) Pull ahead south of TT switch.
15) Spot Cars 3 and 4 on siding at Central Bridge.
16) Back to Train and double.
17) Head to Howe's Cave Cement Plant.
18) Stop North of crossing at Howe's.
19) Cut between cars 2 and 3 and run to cement plant.
20) Pick up four loaded cement hoppers off of Howe's Cave 1.
21) Go to Howe's Cave 2 and pick up three empty coal hoppers.
22) Head back to train and make double.
23) Cut between cars 16 and 17.
24) Head back to plant and go to Howe's 1 to spot cars 13 to 16 (empty cement hoppers).
25) Go to Howe's Cave 2 and spot cars 10 to 12 (loaded coal hoppers) at the bunker.
26) Return to train and make double.
27) Head south to Cobleskill Yard. Remain north of yard lead.
28) Make cut between cars 12 and 13. Pull south of Cobleskill lead.
29) Spot cars 10 to 12 on Cobleskill 2.
30) Head back to train and make double.
31) Pull south to Oneonta Yard and Clear main on SB Runner.

This crew is on duty at 1330 and should be in Oneonta Yard (provided all meets with other trains work out) at 1715.

All of the road freights also follow the work pattern that their real world counterparts did back on that day.

Here is what you will hear between the SU-10 and South End Dispatcher:

SU10: SU10- Engine 7314 calling South End Dispatcher.
South DS: South End Dispatcher answering 7314.
SU10: Yes sir, SU-10, crew on board and brake test complete requesting permission to open up at Mohawk South and head south to Delanson.
South DS: I can clear you to south Kelley's Station for right now...have the PB 100 coming north...so here we go...SU-10 with engine 7314 okay to open up at South Mohawk and occupy main 2. Proceed from Mohawk South to CP Campbell on main two and on single main from CP Campbell to KL Cabin and on number two from KL Cabin to Kelley Station.
SU-10: (Repeats that verbatim with no freelancing.)
South DS: OK on the read back 7314...here comes your light Mohawk South and have a safe trip south.
SU-10: Roger that. Waiting on the light at Mohawk South....(once it comes in--and no that does not happen in the sim yet but we are working on pure DS control panels) SU-10 medium clear at Mohawk South from Yard Lead to Main 2 in a south direction, 7314, out.

That gives you a feel for what we do. Yes I do indeed hold instruction classes on CTC, TWC, DTC, DCS Block Control. The CTC and TWC are part of the VORA Intros including how to correctly complete a GCOR track warrant. If an operating session uses anything different we cover that with Operating Bulletins just like actual crews would receive. That covers speed restriction (yes the D&H had many), MOW work (still thinking about adding a crew out there), and anything else that affects the operation of trains over the road. This goes to all crews in advance of the session so they know what's coming their way.

This is the kind of system we are working towards duplicating in our "activity packs" derived from our operating sessions. Nothing beats doing it live though and we have a number of surprises in store there as well.

Nick
TrainMaster1
 
Posts: 111
Joined: Thu Feb 17, 2011 11:19 pm

PreviousNext

Return to General Discussion

Who is online

Users browsing this forum: No registered users and 0 guests