If you want to run a gazillion trains, run rule 251 double track. The only limitation is train spacing, what headway you need.
If you run single track the limitation is the turn time, the time it takes to "turn' the longest stretch of single track. That is the time it takes for a train to leave the siding at one end of the single track, run to clear the other end of the single track and then on opposing train to do the same in the other direction. That time between the when the first train departs the first siding and the 2nd train clears that siding in the opposite direction is the turn time. It sets the time between trains.
That sets how fast you can feed trains onto the layout, which sets the capacity of the layout.
The turn time can be adjusted a bit by fleeting trains, that is running more than one train in a single direction at the same time. Generally that only works in one direction on typical single track, but may raise the capacity slightly. How many trains can be fleeted depends on the length of the signle track, the speed of the trains and how big the sidings are.
We had software to figure out turn times for single tracks, we primarily used it for train spacing and when a two main track railroad had to be single tracked for maintenance. I also used it when things got heavy on the line between St Louis and Jeff City, two main tracks except for two single track stretches over bridges.
Dave H. Painted side goes up. My website : wnbranch.com
gregc ATLANTIC CENTRAL The club layout you posted is a typical old school spaghetti bowl of track in my opinion with no visual continuity of direction of travel. did you look at the schematic on the bottom of the diagram?
ATLANTIC CENTRAL The club layout you posted is a typical old school spaghetti bowl of track in my opinion with no visual continuity of direction of travel.
did you look at the schematic on the bottom of the diagram?
Yes, of course I did. I get it, but it was hard to look at the track plan and relate it to the schematic.
I have never liked layouts that run back thru the same scene or that excessively layer different routes over top of each other.
Yes, I understand that being there and operating the layout, you "learn" it. But big or small I like a more straight forward approach.
Like I have said many times before, there is a big difference between a big layout and a complex one. I'm building a big layout, but it will not be complex.
Really hard for me to relate to the club thing these days.
Sheldon
gregc NVSRR but dont forget the size of your power source can dictate train capacity. things like that can be handled by increasing capacity, adding more power districts and number of circuit breakers but i think you can't increased the capability of the dispatcher, in our case with a single piece of software controlling the layout. i wonder if LCC, which targets very large layouts support multiple dispatchers
NVSRR but dont forget the size of your power source can dictate train capacity.
things like that can be handled by increasing capacity, adding more power districts and number of circuit breakers
but i think you can't increased the capability of the dispatcher, in our case with a single piece of software controlling the layout.
i wonder if LCC, which targets very large layouts support multiple dispatchers
CMRI or Loconet can handle multiple dispatchers, with JMRI or other interface, and have since before the idea of LCC was floated. A physical CTC panel in CMRI is just another node or series of nodes in the netwrok, no reason you can't have two panels, each controlling half the layout. WIth Loconet for a signal and detection bus, using JMRI and software dispatcher panels, there are really two ways to have multiples - two computers, each with a Loconet interface tied to the common Loconet, or one computer tied to Loconet with an interface, and a second (or more) computer tied to the first one using the Loconet over TCP built in to JMRI.
There was an article not so long agom, in either MR or another publication, where layout builder uses multiple tablet computers with JMRI panels on them spread around the room as his local tower panels. I think there were at least a half dozen.
There's a reason I'm not as hot on LCC as some people. CMRI uses dirt cheap RS485 interface chips. Loconet uses a single transistor and a simple comparator which is also a jellybean part. LCC uses CAN, and the CAN chips, despite there being a dozen or more in ever car sold these days, are not nearly so cheap. So much for the idea of using a mass produced product from another market to leverage volume in our fairly niche part of a niche hobby. And some of the touted features have been possible for 30+ years with existing options. The only difference is an official standard - but coming so late in the game, when CMRI has been sort of a defacto standard for controlling layouts that are fully signaled and operated seems rather anti-climatic.
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
dehusmanDispatching is entirely dependent on what the trains do. If you build the layout to handle running 50-60 trains a session then dispatching isn't a problem.
i can see how you can design a layout to support a large number of trains. but i railroads have junctions where trains meet that can become bottle necks
i live near New Brunswick, NJ. east of me is the NE corridor. but west of me, the reading and lehigh valley lines meet in manville a ~1mi from boundbrook where they meet the CNJ line.
i'm more curious in routing/scheduling trains on these single track lines west of me running freight. i assume its not optimal to start a train out of philly(?) that needs to wait in manville, port reading jct, because another train is running thru
i'm guessing on layout, despite any thourough string diagram planning, operators may not run their trains at the expected speed to arive at junctions as expected. if they're too fast they arrive early and have to wait, but if they arrive too late, could they screw up the rest of the schedule and how does that affect the dispatcher
i'm wondering that even if you have an optimal schedule, doesn't you're dispathcer have to be on the ball, ready to switch a turnout and clear the signal once a train moves through a junction. I can see how "fleeting" can help, but you need to know when to use it and when not to.
we've been doing some remote ops and finding 4 remote engineers to be near the limit with 2 remote assistants keeping an eye on things. the club is also doing hybrid ops with most trains runs by engineers at the layout and a few remote. some out of state ex members are interested in remote operating as well as some who would rather not drive at night.
greg - Philadelphia & Reading / Reading
gregci'm more curious in routing/scheduling trains on these single track lines west of me running freight. i assume its not optimal to start a train out of philly(?) that needs to wait in manville, port reading jct, because another train is running thru
You are way over thinking this. Freight railroading isn' the Shinkansen.
Unless there are no sidings left between Phillie and Manville, you would just run that train up to the junction and it would take its turn. Freight operation doesn't have that precision. Its not scheduled with that precision. Freight schedules generally are based on a percentile. For example if the schedules are based on the 70th percentile, they allow the running time that 70% of the trains can make.
gregci'm guessing on layout, despite any thourough string diagram planning, operators may not run their trains at the expected speed to arive at junctions as expected. if they're too fast they arrive early and have to wait, but if they arrive too late, could they screw up the rest of the schedule and how does that affect the dispatcher
The job of the dispatcher is to route the trains. On a real railroad, bulk trains (grain, coal, ore) don't have "schedules" per se and can run at any time. Lower priority manifest can run plus or minus 8 hours from schedule, a higher priority train plus or minus 4 hours from schedule and a really hot freight train from several hours early to a hour or so late. With that variability, nobody sweats having to wait 2 minutes to line a signal at a junction 150 miles away.
The problems on a model railroad is variability of many facets. On layouts, the switching work, unless the owner closely regulates the work, can vary significantly in the number of cars switched/handled from train to train, session to session. The operators vary wildly in how efficient they are, some people can switch 20 cars in 15 min, others can take hours to switch two or three cars. Running speeds are different, some people believe that trains are "all things that creepeth and crawleth", others run at faster speeds. Model yards operate slower and owners tend to overload them, making them a coke point. Model yardmasters seem to think a yard limit board is a hold signal and park trains for long periods. All of those things combined with the fact that layouts have such short runs there is no buffer or slack.
About all I use a string diagram for is to make sure I don't plan to have too many trains out on the layout at any given time (more trains to run than operators) and to make sure the operators will get a reasonable number of trains per session. Trying to schedule the order of trains through a junction I would think would be folly.
One at a time.
You can have 50 different trains waiting in staging, or a train waiting on a passing siding somewhere, but my rules are that only one train at a time moves on the layout.
Some might want to have a train set on auto pilot and run on the layout while they work another train. I can't bifurcate my concentration away from only one train moving, so I wouldn't bother with running the second because I would barely notice it.
Not exactly the same situation as a large club layout.
- Douglas
dehusmanOn a real railroad, bulk trains (grain, coal, ore) don't have "schedules" per se and can run at any time. Lower priority manifest can run plus or minus 8 hours from schedule, a higher priority train plus or minus 4 hours from schedule and a really hot freight train from several hours early to a hour or so late.
thanks for explaining.
i assume no one would be pleased if i were assigned a train and had 5 min conversation with someone before pulling out during an op-session.
the club may have 6+ engineers besides a dispatcher and 3 tower operators. i've had to wait at signals for 5 mins. 2-3 trains will be scheduled over a single track branch, as well; engineers standing around waiting for their train to be cleared out of staging thru the single track while one is coming into staging over the single track
i'm not fond of standing around. i'm not an engineer paid regardless
i think some feel this is part of the realism of a real railroad, the entertainment of operating a model railroad and helps reduce the work load on the dispatcher
perhaps my issue is more related to the scheduling of trains. that a better thought out schedule, using a string diagram, could result in more trains moving on the layout at the same time, keeping operators busy. but then maybe the operating session would end earlier and then what would you do.
i appreciate the comments in this thread. they help me better understand the various aspects that affect/limit the operation of a layout.
1031
gregc i assume no one would be pleased if i were assigned a train and had 5 min conversation with someone before pulling out during an op-session.
That's what happens when you have a friendly operating session. If you had everyone running a train without conversation or friendly banter, no one would want to run trains.
My layout is 20' x 20', 2 levels, double track main with staging on lower level. I dispatch with manual controlled signals. I have one yardmaster and my optimal operating session is to have 4 operators. It can get crowded, and some trains will have both engineer and brakeman.
100 operators? I've seen modular layouts bigger than your original post and they don't have that many operators. How long would an operating session last with that many operators?
Neal
gregcthe club may have 6+ engineers besides a dispatcher and 3 tower operators. i've had to wait at signals for 5 mins. 2-3 trains will be scheduled over a single track branch, as well;
Finally we are getting to your real question.
engineers standing around waiting for their train to be cleared out of staging thru the single track while one is coming into staging over the single track
This harks back to the "turn time" that I mentioned earlier. Also applies to single track helixes. The constraint on the spacing of trains is the turn time into staging.
The question then becomes what do you want the operation to feel like?
Do you expect to have every operator operating a train at every moment? Do you want every operator to have a train to run as soon as the op session starts?
I will operate (after phase 2 is in service) with a crew of 4 yardmaster/yard engines, 3 or 4 road jobs and a dispatcher. After I generate the line up of trains (using a string chart), I then schedule the operators. Which trains will operator A run, which will operator B run, etc. I then adjust the schedules/line up to match the operators.
Real railroads have crews that take 8-12 hours to get someplace. Our crews take way less time so its a completely different dynamic. I have only operated on one layout where I had one train (as opposed to switcher), a turn, that took the entire op session to run.
DoughlessYou can have 50 different trains waiting in staging, or a train waiting on a passing siding somewhere, but my rules are that only one train at a time moves on the layout.
I was on another forum, which I believe GregC is on, and I was getting some advice on setting up my DCC system for a layout I am working on. How many boosters, and how many power districts etc. Layout is to have 11 staging tracks so potentially 10 or 11 trains in staging.
https://model-railroad-hobbyist.com/node/38407
Conclusion was that I could run several trains and probably will only need one booster, or at most two. Part of the topology involved the used of PSX circuit breakers and divide the layout into power districts etc.
I was also recommended that I may want to power block the staging so I can "turn-off" track power to each staging track if I wanted. Reason being DCC engines may be sitting there with sound and pulling power when not in use.
Follow the above link if you want to wade through the discussion but it was very informative.
Rio Grande. The Action Road - Focus 1977-1983
riogrande5761I was also recommended that I may want to power block the staging so I can "turn-off" track power to each staging track if I wanted. Reason being DCC engines may be sitting there with sound and pulling power when not in use.
dehusmanDo you want every operator to have a train to run as soon as the op session starts?
(expect) no
dehusmanDo you expect to have every operator operating a train at every moment?
every operator that would like to, within reason. of course some may only run a train or two and chit chat the rest of the time. but it seems poor scheduling on larger layouts can result in bottle-necks.
gregc dehusman Do you want every operator to have a train to run as soon as the op session starts? (expect) no dehusman Do you expect to have every operator operating a train at every moment? every operator that would like to, within reason. of course some may only run a train or two and chit chat the rest of the time. but it seems poor scheduling on larger layouts can result in bottle-necks.
dehusman Do you want every operator to have a train to run as soon as the op session starts?
dehusman Do you expect to have every operator operating a train at every moment?
OK Greg, the bottleneck problem is the track plan.
Recycling trains at the ends of the mainline looks to be a problem.
Having the staging in the middle of the mainline that short circuits the mainline does not seem like a good plan.
And loop to loop double track mainlines are not really double track in terms of traffic volume.
I have long argued against point to point or even loop to loop layouts as requiring too much setup and turn around time at the ends of the mainline.
In the same layout space, true separate double track continuous double track loops with thru staging would allow trains in both directions to enter and leave staging without waiting for trains traveling in the other direction.
I realize it was typical back in the day, but that layout has the mainline running around and around thru the same scenes a bit too much.
They should have used some of that open floor space........
It can be so simple and so effective, take the mainline once around the room visably, then take it once around the room hidden with staging all along the hidden part.
A west bound train runs the visable length of the main, disappears, parks in staging, another takes its place, reappearing as a new west bound train at the east end of the mainline.
And likewise in the other direction.....
Then you ad in yards, juctions, branch lines, etc.
In my case I only have one visable yard, one major passenger station, one engine terminal, one wye junction, etc, etc.
The only feature that is really modeled more than once are rural commuter stations, there are three of those.
Is there a track plan or any kind of information on the layout we are discussing? If there has been, I haven't seen it. The only thing I've seen is a picture of a decades old layout and string line of a small DRGW layout.
Without any information any discussion is purely in the abstract.
Poor scheduling can bottleneck things.
But then again poor execution of a good schedule can bottle neck things.
A poor track design can bottle neck things.
A poor yard operation can bottle neck things.
A poor communications method can bottle neck things.
Poor operating practices can bottle neck things.
It could be any one, any combination or all of the above. Who knows?
dehusman Is there a track plan or any kind of information on the layout we are discussing? If there has been, I haven't seen it. The only thing I've seen is a picture of a decades old layout and string line of a small DRGW layout. Without any information any discussion is purely in the abstract. Poor scheduling can bottleneck things. But then again poor execution of a good schedule can bottle neck things. A poor track design can bottle neck things. A poor yard operation can bottle neck things. A poor communications method can bottle neck things. Poor operating practices can bottle neck things. It could be any one, any combination or all of the above. Who knows?
It is my understanding that Greg is refering to the Pacific Southern Railway, where I believe he is a member.
dehusmanThe only thing I've seen is a picture of a decades old layout Without any information any discussion is purely in the abstract.
that's the track plan.
ATLANTIC CENTRALIn the same layout space, true separate double track continuous double track loops with thru staging would allow trains in both directions to enter and leave staging without waiting for trains traveling in the other direction.
i believe the layout has examples of all of the above. it has 3 loop staging areas with separate entrance/egress, it has double track mainline but single track to 1 staging area.
i would say it models various realities of existing railroads that have existed for over a 100 years and of course are not optimal for current needs.
gregcthat's the track plan.
Wow. Its reverse loop to reverse loop with a more or less two ovals and a reverse loop dropped in the middle.
I could see where that would be very challenging to operate and if you just ran trains on individual routes, it could get balled up really fast.
If you wanted to operate prototypically you would have to operate the main truck line and the other loopy ovals would have to be connecting lines that feed trains into the main trunk.
Not sure how that would work since it seems like most of the staging is in the middle. If there was staging in each of the end loops, I would try to run it as four major routes: Yard to A-Hyde, Port to A-Hyde, Cliff to F-Hyde, and Yard to Cliff. The routes would overlap on the Dell to Shore portion.
Here's my take on the operation, if I had to overlay some concept on the existing plan. Kind of a Jersey side of New York kinda theme with Port and Yard representing New York City/Jersy City/Port Reading and Cliff kind of a Maybrook type destination, with Hyde representing Wilkes Barre/Scranton and Phillie. The RDG runs Phillie to Port, the CNJ (or LV) runs Scranton to Yard and Port, the NH (or LV) runs Yard to Maybrook.
The constraining factors are the staging at Hyde, the single track, the ability of the end loops to originate and terminate trains and the ability to pump trains between Dell and Shore.
The diagram is so small I can't make out the details of the text description or the individual tracks, so if there is a operating concept, I am unable to read it.
If you want to do west coast, then make substitute LA for New York and assign the routes to the UP, SP and ATSF instead of the RDG, CNJ and LV.
dehusmanHere's my take on the operation, if I had to overlay some concept on the existing plan. Kind of a Jersey side of New York kinda theme with Port and Yard representing New York City/Jersy City/Port Reading and Cliff kind of a Maybrook type destination, with Hyde representing Wilkes Barre/Scranton and Phillie. The RDG runs Phillie to Port, the CNJ (or LV) runs Scranton to Yard and Port, the NH (or LV) runs Yard to Maybrook.
that's an interesting analogy to the extensiveness of the layout.
but i keep thinking of the explanation of car exhaust headers, that hot dogs of gas exit the cylinder and each reaches the header manifold just as the previous hot dog leaves the manifold. i think (model) trains could be scheduled such that when sharing trackage, they arrive at a junction just as it clears from a previous train.
scheduling trains is one challenge, but another is can a dispatcher keep up with such a schedule. yes, this is an unrealistic artifact of a model railroad trying to maximize operations over a limited amount of time.
i'm reminded of gomez adams switching turnouts just as a train completely crosses one and another train reaches it.
it makes sense that the schedule should not only optimize train throughput but dispatcher capacity as well. currently, that workload is shared with tower operators. consequently, the lack of tower operators impacts remote operation.
but there's discussion of integrating tower capability with the dispatcher (more hardware and software). perhaps, as sheldon suggested, a way of having multiple CTC operators managing different parts of the layout.
while i'm sure some think i sound disgruntled, the tread is making me aware of and thinking thru aspects of operation i'm unfamiliar with and may take for granted.
so thanks
As I understand the diagram, one of the problems you have is the many junctions and alternate routes. These create "friction" that makes it hard to move more trains in a given time.
gregc i think (model) trains could be scheduled such that when sharing trackage, they arrive at a junction just as it clears from a previous train.
That probably won't work consistently. All it takes is one club member to decide he needs to go get coffee to make his train a few minute late and jam up the schedule.
If you want to increase capacity, I suggest you look at removing some of the junctions and alternate routes from service temporarily during formal op sessions. That would reduce the number of interactions for each train and ease scheduling. After all, real-life trains can’t teleport directly from New York to Chicago; they run linearly through the intervening country.
In that space, a different track plan could support many more trains in motion, but of course it can’t be changed now. I’ve operated and worked with ops plans for much smaller spaces that handle more trains at once with less friction.
It may be counterintuitive, but fewer trains running at one time may increase the throughput and allow more trains to be run overall in a given time. A jam-packed schedule with lots of dependencies is very hard to maintain and any missed meets and passes will snowball into much bigger problems.
Byron
Layout Design GalleryLayout Design Special Interest Group
cuyamaI suggest you look at removing some of the junctions and alternate routes
i'm not aware of alternate routes?
gregci'm not aware of alternate routes?
There are 4 reversing loops and two "ovals" plus the "main line". I found at least 6 or 7 routes a train can take out of Hyde, plus at least three involving Cliff and three more involving coming off the branch.
dehusmanThere are 4 reversing loops and two "ovals" plus the "main line". I found at least 6 or 7 routes a train can take out of Hyde, plus at least three involving Cliff and three more involving coming off the branch.
+1
13?
an alternate route is separated track between two endpoints, right?
don't understand why you consider a reversing loop an alternate route
yes, trains can originate and return to hyde, but hyde is hidden and is really two separate, westbound and eastbound staging areas with specific tracks for long (80) coal trains.
port complicates things. one (D, E) track is hidden storage for a circus train. i guess the other is an alternate route.
can you be specific?
gregci think (model) trains could be scheduled such that when sharing trackage, they arrive at a junction just as it clears from a previous train.
Yes, but that's not going to have a high degree of success because any pertubation will cause the schedule to fail.
Real railroads don't work that way.
My observation is that you like things wound really tight. Based on other threads you have had, you really, really, really sweat the details. You do not seem to like things "loose".
Unfortunately, this is one of those situations where a "loose" solution might work better than a "tight" solution. You do exactly what the prototype does, work at the 70% percentile. That puts trains early more than late, which gives them a bit of slack, so if they do get delayed its not a big deal.
On the line between Kansas City and St Louis, we had 2 pair of Amtraks a day. The line was conventional double track St Louis to Jeff City and effective double track Jeff City to KC. Trains went west on the Sedalia Sub and east on the River Sub, parallel routes 20 miles apart. Amtrak operated in both directions on the Sedalia Sub. Because of the volume, we had to start putting trains in sidings as far back as Jeff City, 125 miles away from KC, BEFORE Amtrak left KC. The 8th or 9th train from Kansas City would have to wait for Amtrak to run all the way from KC to Jeff City before it could leave Jeff City. Not suggesting that for your club, but just as an example even with a "tightly" scheduled train, delays can happen.
It might also help if you backed off the train density a tad, minimizing the conflicts.
Which, to Cuyama's point, brings us to simplifying the routing. Just use the through route between Yard, Port and Cliff, remove Hyde, Radstock and Richards from service and spike the switches at Dell execpt for the crossovers.
Cliff to Green Mountain becomes staging and you run staging to yard or staging to port. Hugely simplifies the operation. Then you have a straight double track main with a short stretch of single track through Spikes Peak. Simpler to operate.
Routes:
You said you had trains waiting for other trains at junctions and interlockings. That implies that you have trains taking different routes through those interlockings or junctions.
Line every main track switch between Yard and Bank for the normal route. Any time you line a switch (other than a crossover move) is a different route. Yard-Latham-Dell-Shore-Nassau-Bank-Cliff is a "route". Any other combination of statins, or any other order of stations is a different route.
thanks
i understand running railroad like airlines, landing 2 mins apart is unrealistic and unsafe (yes, i'm a firmware guy use to controlling thing in nanosec).
i understand that a tight schedule is totally messed up if theres a model train derailment.
but i think there's room for improvement when you have ~half your engineers waiting for a train out of the same staging yard, but wonder if it gives the dispatcher a break.
i believe a well thought out schedule can results in rational throughput and smooth, uneventful operation. maybe people want some chaos (entertainment)
gregcbut i think there's room for improvement when you have ~half your engineers waiting for a train out of the same staging yard, but wonder if it gives the dispatcher a break.
True.
Next step is to figure out why everybody is bunched at the same place.
Is it that several trains were scheduled to converge on the same place at the same time?
Were they scheduled at different times and they didn't run to schedule?
Or... is the one train everybody is waiting on scheduled at the wrong time?
Or. is the dispatcher prioritizing the wrong train (holding 3 trains to run one)?
Lots of options, lots of ways to adjust the problem.
Dave H. provided good suggestions. In addition, many layouts now stage some trains out on the layout as if "en route". This reduces the congestion in staging and active yards and puts more crews to work at the session start.
cuyama gregc but i think there's room for improvement when you have ~half your engineers waiting for a train out of the same staging yard, but wonder if it gives the dispatcher a break. Dave H. provided good suggestions. In addition, many layouts now stage some trains out on the layout as if "en route". This reduces the congestion in staging and active yards and puts more crews to work at the session start.
gregc but i think there's room for improvement when you have ~half your engineers waiting for a train out of the same staging yard, but wonder if it gives the dispatcher a break.
I do this, been doing it for years.
My operational plan includes one or two freight trains in the yard ready to depart, several passenger trains already in the station, and, there are cutoff tracks that bypass the staging all together, they appear as a junction at each end of the visable mainline and provide two more staging spots during an opps session.
My very first layout, built for me by my father in 1967, had hidden staging. And with power routing TruScale turnouts you could operate it blind. Throw one turnout, drive the train in slowly, it would stop when it hit a dead spot controlled by contacts on the Kemtron switch machine for the turnout. Throw the second turnout, flip the block toggle and pull the other train out.
People were amazed, one train went in the tunnel, a different one came out the other end. Guess my father was ahead of his time in model railroading.......