Quasi-Möbius Coasters

  • 17 replies
  • 792 views
*

Offline Bullethead

  • *
  • 1936
  • shyguy's World Rocks!
    • View Profile
Quasi-Möbius Coasters
« on: July 04, 2018, 02:37 PM »
QUASI-MÖBIUS COASTERS
[/size]

PC has forced me to learn more about real coasters than I ever wanted to.  Recently, I had to learn about Möbius and Quasi-Möbius coasters because the upcoming Vintage Pack has made me want to build something akin to this thing:



I've seen posts where others have made such things and I've always thought they must be too complicated for me to be able to figure out, or I'd never be able to blend the required timing features into a pleasing ride.  But that was before I had the above inspiration.  So, I sat down to learn how to do it.

TERMINOLOGY

As I understand things (and I only just researched this so could well be wrong), there are 3 general types of multi-track coasters.  First, we have the dueling coaster, which is actually a collection of entirely separate, self-contained coasters built side-by-side.  We can do this in PC, and can even set them to launch simultaneously.  However, that's not what I'm talking about here.

A true Möbius coaster, such as I understand the old "Derby Racer" above was, has 2 stations sharing the same track, both of which load and unload.  Riders start in 1 station and get off at the other, only going around the track once on only 1 of the 2 parallel lanes.  This type of coaster is currently impossible to make in PC because coasters can only have 1 station.

A quasi-Möbius coaster, however, has only 1 station.  Riders make 2 laps of the track, once in each lane, and board and exit at the same place.  Real examples include Twisted Colossus.  This type of coaster IS possible in PC---there are many examples on the Workshop.  And it makes for a cooler ride IMHO, as the riders go twice as far as with the true Möbius, and will have different race experiences each lap.

INFORMATION

So, I decided I wanted to build a quasi-Möbius coaster (QMC from now on).  Problem is, despite this having been done by many players already, I couldn't find any how-to info.  Because I assumed such coasters were fiendishly difficult to make, I figured the previous builders were keeping their dark secretes to themselves and I would have to reverse-engineer their creations.  Thus, I downloaded several but they all approached the situation differently, so I learned very little from them, except the important fact that this is NO one way to do it :).

In the end, I just sat down and built one, and experimented with it until I got it working as I wanted.  In this process, I discovered that QMCs aren't at all difficult to make on the technical side, just a bit tedious on the building side.  There are, however, a number of factors you should keep in mind before you start and as you go along.  Therefore, to save others the trouble I had, and to show folks these things aren't any harder to make than other types of coaters, I decided I'd share my notes here.



MUHAHAHAHA, I made a QMC!  That actually "works" in that the trains race along together more often than not (more on that in the Tuning section).  TREMBLE BEFORE ME, MORTALS!  I'm a coaster-building GOD!  Oh wait, it wasn't really that hard  8|  OK, I'll just get on with this "quasi-tutorial".

MAIN COMPONENTS and OPERATIONAL THEORY

This post describes ONE of the apparently MANY ways to make a QMC work, using the simple test coaster I made.  If you want to reverse-engineer it yourself, here's the Workshop link:  https://steamcommunity.com/sharedfiles/filedetails/?id=1431305857

It's actually quite a popular family coaster if you want to use it as such, but I just wanted to see if I could get a QMC to work at all, not make a park-worthy ride.

Here's an overview of this coaster showing the components required to make it work:



This coaster runs 3 trains, each of which does 2 laps before returning to the station.  At any given time, 1 train is in the station and the other 2 are running.  The trains move along the path S-L1-B1-L2-B2.

S = Station
L1 = Lift 1
B1 is Block Section 1
L2 = Lift 2
B2 = Block Section 2

The key to the whole shebang is manipulating how PC handles block sections.  Stations, block brakes, and lift hill all divide blocks of track from each other, and trains will stop on the divider before the next section of track until the train ahead of it passes the next divider.  With 1 station, 2 block brakes, and 2 lift hills, this track is divided into the following sections:  S-L1, L1-B1, B1-L2, L2-B2, B2-S.  So, a train won't leave the station until the train ahead is completely off of L1, and a train won't leave L1 until the train ahead is beyond B1, etc.

This is what makes B1 the most important piece of track.  Without B1, Train 2 will stop at the top of L1 until Train 1 is completely off of L2.  While this does mean they both leave the lift hill more or less together as desired, you get the undesirable side-effect of the 2nd train having to sit there for a full lap, which the peeps find BORING and which ruins throughput.  A LOT of the QMCs on the Workshop have their trains wait like this, sometimes in more than 1 place, which I just don't like at all.  So I wanted to avoid that.

Note, however, that my placement of B1 was pure luck.  Thinking QMCs were all about timing, I figured I'd need a brake there to release a train at the same time another left the station, figuring a wait was inevitable, and then tweak the rest of the track to get the timing right.  When it turned out the trains were racing OK and never stopped on B1, I took it out, and discovered it was totally necessary to avoid waits.  So I put it back in :).

BUILDING THE TRACK

Again, I expected this would be a fight of adjusting lengths, heights, and radii to make everything work out right, but this turns out not to be the case.  There are just a few general guidelines to keep in mind, shown below, plus some general advice:



1.  Overall Layout
If you want to make something like the old "Derby Racer" above, with perfectly parallel tracks for as far as the eye can see, then you don't have many options.  You can't bank the turns so you can only turn at the top of hills when the coaster is slow enough not to make horrible lateral Gs.  This means you're pretty much limited to a simple out-and-back, such as my example, or a LONG 8 or stacked 8, which is mostly out-and-back in the straightaways.  Either type of 8 would give the best ride racing sensation because the trains would alternate being on the faster inside lane through the turns, so there'd be passing.  So basically, with a wooden coaster, you're going to have a very long, skinny footprint.  With hybrid and steel coasters, however, you can do whatever you want, have the tracks twist around each other, etc.

2.  Turn Radii
This is much more of an issue for wooden coasters with unbanked turns than anything else.  When you're building the track, you naturally go around the 1st lap then have to follow it as close as you can for the 2nd lap.  If the 2nd lap is on the inside of a turn, it will have a rather tighter radius than the 1st lap.  If you made the 1st lap's turn too tight, you won't be able to fit the 2nd lap's turn inside it.  So my advice is to make the 1st lap's turns, where they are the outside lane, wider than you think they should be, so you'll have an easier time building the 2nd lap.

The 1st turn in my little coaster here is as tight as is possible to make.  That was with 15m track sections doing 90^ on the 1st, outside, lane.  This made the 2nd lap a real struggle even using 4m sections.

3.  The Crossover
All Möbius coasters have to have a at least 1 crossover.  It is desirable to have the upper train slightly ahead of the lower train at this point, so most of the peeps in the lower train will see the upper train going over their heads.  This is one of the things you can adjust with the various methods of tuning the ride (see Tuning below).

4.  Exit Space and Station Height
I find it desirable to put the exit on the far side of the station, centered on the parked train, as this minimizes unloading time, which accounts for the vast bulk of the time the train is stopped.  Loading takes about 1 second as the peeps teleport from the chutes to the seats and then the restraints come down.  This can't happen, however, until the last of the previous riders have slowly walked completely off the platform and the exit gate has shut behind them.  Putting the exit on the far side in the middle of the train minimizes their walking distance so minimizes the cycle time.  With a 6-car train, the entire unload/reload cycle takes only 25 seconds (28 seconds with 7 cars).

While getting the timing right on a QMC isn't something you need to worry about TOO much during track construction, it's still something you'll have to deal with eventually (see Tuning).  You can always make the loading cycle take longer by adjusting station settings but you can't make it any shorter than the time it takes the last peep to walk off the platform.  Thus, tuning will be easier in the end if you have the shortest possible loading cycle, so you have some wiggle room elsewhere in the ride.  Besides, getting the train moving again as soon as possible makes the peeps happier and increases $/hour.  So I recommend putting the exit on the far side from the entrance.

If you do that, however, the bypass track at the start of the 2nd lap has to be far enough away from the station to give you the necessary room to place at least the start of the exit ramp on the edge of the platform.  In my example, I gave myself plenty of room, way more than I needed, just to illustrate this point.

Then, of course, the peeps are inside the loop of the track and you have to get them back under or over the track to the main path.  Your options here depend largely on how you want the station area to look.  Getting the exit path high enough to go over the track takes a lot of space so only goes well if the station building is already pretty huge.  If you can justify having a tunnel with your theming, great.  But if you have to have the peeps come under the track at ground level, then you have to put the station platform high enough to allow that.  So I recommend beginning construction by laying down a short section of path.  Then put the station on top of it and raise the station until it turns from red to blue.

5.  General Track-Building
This tip applies only to areas where you want the 2 lanes to match up as closely as possible.  So again, this is mostly if you're doing a woody like the "Derby Racer".  When making the 1st lap, try to use fairly long sections of track without abrupt changes of direction either horizontally or vertically.  It also helps if you only change 1 of those at a time as much as possible, again feeding into the long, skinny layout.  When you come around again building the 2nd lap, you have to use many, many very short sections, often 4m (especially in the insides of curves), and line each one up very carefully with the existing track.  This is what makes building a QMC so tedious.

So now the 2nd lap is a bunch of short sections all at slightly different angles to each other.  For instance, you'll find it impossible to stay exactly parallel to the 1st lap, always alternating 0.32^ to the left and 0.35^ to the right each section.  This will look horrible.  However, you can smooth out all these little irregularities without affecting the overall layout because you're only touching 8m of track in total with the smoothing tool at any one time.  So bunny hops don't do away and the radii of turns doesn't change enough to notice, but you get rid of all the little kinks.

6.  Overall Layout Size
I was really worried I have to tweak the track length of each lane constantly to get the timing to work, but this turns out not to be the case.  Just build the 1st lap like you would any other coaster.  You know when the momentum has bled down enough from a given lift hill to be time to take it back to the station.  And you customarily build coasters so that there's no delay at the station or on the brakes leading to the station---as 1 train leaves the station, the next pulls in immediately.

If you build your coasters like this, then you'll have no trouble with QMC timing.  All you do is take the track back to the lift hill at the appropriate time instead of the station, then go around again.  This will make the timing close enough to perfect naturally, on it's own, that you can probably fine-tune it to be exactly what you want with all the tools discussed in the next section.

TUNING

"Tuning" here means adjusting the various controls at your disposal to get the trains running as you want relative to each other.  The desired results are, in order of priority:
  • Having the trains run very close together for the bulk of the track.  This is the whole purpose of making a QMC so if you fail this, you've totally wasted your time.  Everything else is gravy.
  • Having peeps in the lower train easily be able to see the upper train going over them at the crossover.  This has no effect on the ride's stats that I can tell, it's just for style points.
  • Not having a train sit still on the track or in the station at all, if possible.  If not possible, to minimize this, and hopefully have it happen only at the station, not on the track.  This is probably the hardest thing to achieve without undue fiddling with track lengths.

Before you start tuning your ride, keep the following vitally important things in mind:
  • THE RIDE MUST BE OPEN WITH PEEPS FLOWING INTO IT!  Test dummies never unload so testing NEVER shows the correct amount of time the train is in the station.  The ONLY way to determine this is with real live peeps.  So, build the ride in your test lab "park" where you can get some real peeps to go on it without screwing up your real park.  Then save it as a blueprint once everything is good and paste it into your real park.
  • DO NOT start tuning until the Train 1 has gone around the full track at least as many times as there are trains in total.  IOW, if you've got 3 trains, let the ride run 3 complete cycles.  It takes that long for the ride to settle into its natural, long-term rhythm when starting from a cold stop.  And then let it run a bunch more times to really see what's going on before touching anything.  There will always be some slight variations from run to run.  Usually, QMCs will oscillate through a pattern of the racing trains being in slightly different relative positions from 1 run to the next, lasting 4 or 5 runs before repeating.  So watch long enough to learn this pattern and decide whether you like it or not before making changes.  Having some variation from run to run is good as it gives the peeps a less-predictable experience.  And if most of the runs satisfy the desired outcomes above, maybe you shouldn't change anything.
  • Make the train loading cycle constant.  Set the ride to run with ANY load, check both the minimum and maximum station delays, and set them both to the minimum loading cycle time for your coaster and exit configuration.  These settings should ALWAYS be the same time, even if you change them later.  Then set the departure interval to a few seconds less than your load/unload delay time.  This means the train will be in the station for the least amount of time possible---you can always make it longer later if needed.  And don't worry about the "Any load" thing.  If this coaster is worth a damn, it'll always be nearly full.  Except when returning to service after a breakdown.  Then as peeps trickle back in, the ride will have worked back up to its normal rhythm by the time the queue is full again.

1.  Main Tuning Controls
The basic controls, which are probably all you'll need, are shown below.  This pic also shows the starting positions of the 3 trains for this particular coaster.  This is why you have to let the ride run a while before making any judgment on the timing--the trains ain't spaced at all like they will be when the ride has reached its stable operating configuration.



Lift hill speed is probably your best tool.  It'll usually happen that 1 train or the other gets to the bottom of the hill a couple seconds before the other.  If that's the case, slow that hill down and/or speed the other up slightly.  In general, I prefer to start with both hills at max speed as I personally HATE long, slow lifts and adjust one or the other down as needed.  This also used to be a good idea in the game, as slow lifts killed ride excitement.  A few updates back, however, lift hills got removed from the ride stat equations so you can now safely go slow if you want.  But why?  Throughput!

If the lift hills can't solve the problem, or just look silly doing it because one is way faster than the other, you might consider adjusting the length of either the station run or the station bypass run.  Or adding trim brakes to the station bypass.  But if the station side way faster than the bypass side and you can't add track to it, you need to slow the station side down before the lift.  Which means increasing the time in-station.  Adjust both the min and max station delays up a few seconds (keeping them both the same number).  This will usually cause a backup on the station brake run, however, which will then have repercussions all down the line in overall ride timing, so try to avoid building the coaster like this to start with :).  But as I said above, this shouldn't be an issue if you build the QMC essentially the same as you do normal coasters.

2.  Tuning Caveat
After you make any change, you again have to let the coaster run a bunch of full cycles to let it settle into whatever new rhythm it has.  Often, you'll find that whatever you did had unexpected, undesirable results which might make things worse than before.  So always remember how the settings were before you changed them, so you can go back if needed.

BIGGER TRACKS

This collection of notes (NOT a tutorial) was based solely on making a single, small test QMC.  I expect that the bigger the QMC, the more complicated it gets.  For example, the next stage up would be 5 trains, so that 4 are racing and 1 loading.  If you want to go with single lift hills for the whole thing, then you need block brakes in the middle of the runs.  This will affect where trains start, as shown below:



Just based on what I've seen so far, I expect Case 2 would be better as it avoids a grotesquely long brake run coming into the station.  However, it might be easier all around to use 2 lift hills, in which case you'd just use 2 of the 3-train, 1-hill things in a row.  I don't know.  That's a project for another day.

I hope folks find this useful.
« Last Edit: July 04, 2018, 05:25 PM by Bullethead »
-Bullethead
NIHIL INIQVIVS QVAM ÆQVITATEM NIMIS INTENDERE
My Steam Workshop page

Re: Quasi-Möbius Coasters
« Reply #1 on: July 04, 2018, 03:16 PM »
man - nice write up

Re: Quasi-Möbius Coasters
« Reply #2 on: July 04, 2018, 04:15 PM »
ok, wow.  We (I) need a tutorial like this for every type of coaster.  Even when I try to do research I can't find information like this, written at a level I can understand.

I expect we'll start to see a lot of these once the new content comes out.

I do love the look of them and would love to build one, knowing that I probably don't have the skills to create parallel tracks, and also tend to kick peeps out of the park early on to avoid the lag so I'd never be able to test it properly.  :(








*

Offline Fisherman

  • *
  • 1512
  • I'm a nice guy-REALLY!
    • View Profile
Re: Quasi-Möbius Coasters
« Reply #3 on: July 04, 2018, 04:42 PM »
You could do something similar with a dummy station on one side...basically B1 (in your first example) is a fake station...and put the stations side by side.  That would enable you to sorta recreate Kennywood's Racer or Blackpool's Grand National. 

*

Offline Bullethead

  • *
  • 1936
  • shyguy's World Rocks!
    • View Profile
Re: Quasi-Möbius Coasters
« Reply #4 on: July 04, 2018, 06:02 PM »
man - nice write up

Much obliged, Sir  B)

ok, wow.  We (I) need a tutorial like this for every type of coaster.  Even when I try to do research I can't find information like this, written at a level I can understand.

Having spent the last hour trying to correct all the typos and UBB code errors in the original version, which were made while slightly tipsy and then corrected while rather further into the bottle, I'd advise you to drink a large glass of water with a big vitamin B pill and an aspirin, go to bed, and then look at it again tomorrow morning over coffee.  You probably will have a different opinion ;).

Quote
I do love the look of them and would love to build one, knowing that I probably don't have the skills to create parallel tracks, and also tend to kick peeps out of the park early on to avoid the lag so I'd never be able to test it properly.  :(

Well, building parallel tracks is no harder than lining up any building pieces, which you can do quite well.  Turn off angle snap and move the mouse very slowly.  And you only build and test such things on a totally blank slate, nothing in it but the coaster, so no FPS problems.

You could do something similar with a dummy station on one side...basically B1 (in your first example) is a fake station...and put the stations side by side.  That would enable you to sorta recreate Kennywood's Racer or Blackpool's Grand National.

Yeah, the question of how best to fake a true Möbius, alternating-station coaster is a tough one.  The easiest answer is to just go with separate dueling coasters.  That way, you can more easily get the layout built and you have 2 stations.  All you're missing is the crossover which, at least for Grand National, either doesn't exist due to the brake runs and station exits going in opposite directions above each other, or is buried under the stations.  But the drawback is, you have 2 nearly identical rides which are in direct competition with each other for peeps, and 1 will inevitably be slightly more attractive than the other, so will get way more peeps.

If you go QMC, then you only have 1 station.  If you fake it by using an actual coaster station, you'll be spammed by notifications of 1 type or another.  Either that ride isn't complete, or lacks and entrance.  If you instead fake it by building a facsimile, you avoid that.  But in either case, you still have to find a way to separate the station track from the bypass track without being so obvious about it as to spoil the recreation you're trying to make.

It's a tricky problem.  I suspect 1 solution would look better in some situations than others.

Anyway, thanks to all for the feedback.
« Last Edit: July 05, 2018, 05:49 AM by Bullethead »
-Bullethead
NIHIL INIQVIVS QVAM ÆQVITATEM NIMIS INTENDERE
My Steam Workshop page

*

Offline Fisherman

  • *
  • 1512
  • I'm a nice guy-REALLY!
    • View Profile
Re: Quasi-Möbius Coasters
« Reply #5 on: July 04, 2018, 06:42 PM »
Kennywood's Racer leaves the station exactly like Grand National. It's a cool trick to create that Möbius effect.

*

Offline Bullethead

  • *
  • 1936
  • shyguy's World Rocks!
    • View Profile
Re: Quasi-Möbius Coasters
« Reply #6 on: July 05, 2018, 06:33 AM »
Kennywood's Racer leaves the station exactly like Grand National. It's a cool trick to create that Möbius effect.

Hmmm....  I've been giving this some thought....  While the station configuration for Grand National and Racer (which I've recently been spying on via satellite) certainly lends itself to the simple solution of separate dueling coasters, I want to avoid that.  I find the QMC to be a more desirable solution in terms of park management, ride maintenance, and ride quality.  Plus, you get style points for the gadget factor ;).

The main challenge, it seems to me, is that to keep up appearances, you have to keep the bypass so close to the station that you wouldn't be able to put an exit between them.  Thus, you're going to have way longer loading times due to the peeps in the last car having to walk the full length of the platform to the exit chute up front, then the full width of the wide side of the platform to reach the gate.  This would utterly ruin a small QMC like the one here, and also probably a fair amount of medium-sized ones.  However, if the track was long enough, then the overall ride cycle time would be long enough to cover a lot of time in the station without making any train wait for another to load.  You'd just have to run fewer trains than you might want.

I suspect this is what I'll end up attempting when I try my hand at faking the Euclid Derby Racer.  That thing appears to have had a stupidly long track.  From the old postcard above, which appears to have been taken from atop the turn at one end, it looks like a good quarter mile to the far end.  With 6 passes along this length, you're talking 1.5 miles just to go around once.  If you built it QMC, that's a 3-mile ride.   :o.  Surely that's enough track to make up for a loading time of 45-60 seconds....
-Bullethead
NIHIL INIQVIVS QVAM ÆQVITATEM NIMIS INTENDERE
My Steam Workshop page

Re: Quasi-Möbius Coasters
« Reply #7 on: July 05, 2018, 07:38 AM »
Thank you for a terrific non tutorial set of notes!

I have also been toying with an older park recreation, and was trying to figure out how to create a mobius...

*

Offline Fisherman

  • *
  • 1512
  • I'm a nice guy-REALLY!
    • View Profile
Re: Quasi-Möbius Coasters
« Reply #8 on: July 05, 2018, 08:04 AM »
Not separate dueling coasters...because the block brake would be the second station (fake though).

*

Offline Weber

  • *
  • 1515
  • It's not over till it's over.
    • View Profile
    • velvetskyemusic.com
Re: Quasi-Möbius Coasters
« Reply #9 on: July 05, 2018, 04:14 PM »
Wow you're such a nerd!  I love it!

*

Offline Weber

  • *
  • 1515
  • It's not over till it's over.
    • View Profile
    • velvetskyemusic.com
Re: Quasi-Möbius Coasters
« Reply #10 on: July 05, 2018, 04:25 PM »
Kennywood's Racer leaves the station exactly like Grand National. It's a cool trick to create that Möbius effect.


What do you know of my home town coaster?  As a kid I rode it so may times but the Thunderbolt was better at night and then the Steel Phantom appeared which was incredible followed by Phantom's Revenge which IMHO was sub par compared to the original but other's seem to not think so.  Perhaps it was because Steel Phantom was new when I first rode it and wasn't all worn down from years of use, so it was very smooth at first.


HISTORY LESSON TIME:


The idea for Steel Phantom came from Henry Henniger, president of Kennywood Entertainment Company, who wanted a record-breaking steel coaster. Henniger was unsure how to fit the ride into the park, however a layout was worked out in conjunction with Arrow Dynamics.On 27 July 1990, Kennywood announced the name for their new steel roller coaster would be Steel Phantom and that it would use the same station as the out-going Laser Loop.The ride opened in May 1991 as the fastest and steepest roller coaster in the world. It was also the first hyper roller coaster to feature inversions. Laser Loop as it was when I rode it as a kid:
« Last Edit: July 05, 2018, 04:33 PM by Weber »

Re: Quasi-Möbius Coasters
« Reply #11 on: July 05, 2018, 08:17 PM »
was this coaster a pull up and drop launch...or was it some other launch type? The loop and format of this looks very close to The Tidal Wave that I used to ride as a kid at Six Flags Great America. It used a cable pull launch and was super fast for the time. I believe this ride now is a different name at the park in Louisville?

*

Offline Fisherman

  • *
  • 1512
  • I'm a nice guy-REALLY!
    • View Profile
Re: Quasi-Möbius Coasters
« Reply #12 on: July 05, 2018, 09:42 PM »
It was a Schwarzkopf fly wheel shuttle loop... Just like Knotts' Montezuma's Revenge and the lost Astroworld's Greezed Lightnin'.  The Tidal Waves at the two Great America parks were weight drop propelled instead of a fly wheel/clutch catapult. The fly wheel ones went faster... They had more power
« Last Edit: July 05, 2018, 09:45 PM by Fisherman »

*

Offline Weber

  • *
  • 1515
  • It's not over till it's over.
    • View Profile
    • velvetskyemusic.com
Re: Quasi-Möbius Coasters
« Reply #13 on: July 06, 2018, 02:34 PM »
It was a Schwarzkopf fly wheel shuttle loop... Just like Knotts' Montezuma's Revenge and the lost Astroworld's Greezed Lightnin'.  The Tidal Waves at the two Great America parks were weight drop propelled instead of a fly wheel/clutch catapult. The fly wheel ones went faster... They had more power


This is correct!  And it was a very fun coaster I remember.  and if kennywood wasn't so small they would be able to ADD coasters instead of always REPLACING them which is really annoying to say the least.  I am currently 3.9 miles from the park!


More History!

Kennywood is an amusement park located in West Mifflin, Pennsylvania. The park first opened as a "trolley park" attraction at the end of the Mellon family's Monongahela Street Railway on May 30, 1899. It was purchased in 1906 by F. W. Henninger and Andrew McSwigan who later formed the Kennywood Entertainment Company, which owned and operated the park as a closely held family business for over 100 years. Kennywood was purchased by California-based Palace Entertainment in 2007. The amusement park features various structures and rides dating back to the early 1900s. Along with Rye Playland Park, it is one of only two amusement parks designated a National Historic Landmark for its history and historic rides, and is one of only thirteen trolley parks still operating in the United States.


Kennywood is approximately 8 miles (13 km) from Downtown Pittsburgh, in West Mifflin, Pennsylvania. The park is along Pennsylvania Route 837 (Green Belt), known as Kennywood Boulevard as it passes through the borough. The closest Interstate connection is Exit 77 (Edgewood/Swissvale) on Interstate 376. The Mon–Fayette Expressway will eventually go past Kennywood, which will prompt an expansion of the park when it is built.


Historically, the park is on the location of the July 9, 1755 Battle of the Monongahela, where British general Edward Braddock was mortally wounded, ending his expedition to capture the French Fort Duquesne during the French and Indian War. George Washington was a colonel to Braddock, and fought at the battle before they retreated.  Later the land on the bluff above the Monongahela River was part of a farm owned by Anthony Kenny. Starting around the time of the American Civil War, the site was a popular picnic grove for locals, known as "Kenny's Grove".
« Last Edit: July 06, 2018, 02:45 PM by Weber »

*

Offline Bullethead

  • *
  • 1936
  • shyguy's World Rocks!
    • View Profile
Re: Quasi-Möbius Coasters
« Reply #14 on: July 07, 2018, 12:43 PM »
Interesting info, Weber.  I myself, not being a coaster geek, hadn't heard of Kennywood until I started playing PC.  And when I did hear of it, I assumed it was relatively new, built by Kenny Rogers to comply with Dolly Parton's Dollywood ;).  Seriously, I did.  I mean, why not?

But anyway, back to the subject of QMCs...

QUASI-MÖBIUS COASTER RESEARCH, PART 2




As you can see, I have now built a considerably bigger QMC, which is a quasi-copy of the 1913-vintage "Racing Coaster" (nee "Derby Racer") and the long-dead Euclid Beach Park in Cleveland, OH.  The original coaster was a true Möbius, not a QMC, but it looks cool so I wanted to do something like it.  And I learned some more about QMCs, which I will now share.

NOTE:  The above pic was taken from about the same point as the old post card at the top of the OP.  I had originally thought the track on the left of that postcard was part of the same coaster, due to the parallel tracks.  However, research proved this to be a completely separate, single-track coaster called Thiller, a VERY interesting ride in its own right which I hope to replicate someday.

Here's the blueprint for those who want to follow along at home:  https://steamcommunity.com/sharedfiles/filedetails/?id=1434436567

Anyway, on with the show.  Here's an overview of the bigger QMC.  It has the same footprint (250m x 36m) as the real thing, but a taller, pointier lift hill due to friction differences in the game.



As with all QMCs, the trains make 2 laps.  This coaster runs 5 trains so the long straightaways have block brakes on each end, so that 2 sets of trains can be running at once while the 5th is in the station.  All 5 trains, however, start at the station and on the brake run just behind it, under the crossover.  Originally I had only 2 sets of brakes behind the station so the other trains spawned on the mid-course brakes but the coaster NEVER synched up and raced this way.  So I extended the brake run, so that now it makes up nearly all the crossover.

Schematically, the track is arranged like this:



So, the train goes S-L1-B1-B2-B3-L2-B4-B5, and then the brake run B6 through B9.  These latter don't really have any effect on operations.  B1/B2 and B4/B5 divide the main track into sections so multiple trains can run on the 1st and 2nd laps, respectively.  B3 exists for the same reason B1 did in the small, test coaster, to keep a train from stopping on B2 waiting on the train ahead to clear L2.

These trains have only 3 cars because that's what the real coaster used.  Note that the station exit is on the same side as the entrance.  With only 3 cars, however, this still results in a 25-second minimum unload/load cycle per train.

Other than extending the brake run so all trains start at or next to the station, I did no tuning on this coaster to get the timing right.  The station is set to the minimum 25-second cycle type and both lift hills are running full speed.  The way things turned out, it takes a train about 85 seconds to make the 1st lap.  This doesn't quite mesh with the 25-second loading time, but it comes close enough.  Once it stabilizes with full trains loading and unloading after several full cycles, it gets into a rhythm of 2 good races and 1 miss, repeat.  The races are better in different parts of the track due to different relative starting positions of the trains.

It could probably be tuned better to result in more consistent racing.  If anybody wants to take the trouble, feel free :).

I also learned that it's possible to build the 2nd lap with a lot of 8m sections and then do lots of smoothing and tweaking, and still get a pretty good match to the 1st lap.  This was a LOT faster to build than individually fitting huge numbers of 4m pieces, and looks reasonably good.  So the coaster is reasonably park-quality, although it definitely needs a LOT of filler pieces in the supports.  I haven't done that yet.

The ride has the sort of good-but-not-great stats you'd expect for the type of layout, but that means it's totally kid-friendly.  In fact, it's stats make it attractive to about 80% of all peeps, so it should do well in any park with the space for it.



Anyway, I hope folks find this helpful.
-Bullethead
NIHIL INIQVIVS QVAM ÆQVITATEM NIMIS INTENDERE
My Steam Workshop page

*

Offline Weber

  • *
  • 1515
  • It's not over till it's over.
    • View Profile
    • velvetskyemusic.com
Re: Quasi-Möbius Coasters
« Reply #15 on: July 07, 2018, 03:24 PM »
Interesting info, Weber.  I myself, not being a coaster geek, hadn't heard of Kennywood until I started playing PC.  And when I did hear of it, I assumed it was relatively new, built by Kenny Rogers to comply with Dolly Parton's Dollywood ;) .  Seriously, I did.  I mean, why not?


Well because all of the prior info I gave was relevant to your original post.  You see Kennywood has the only existing Mobius Loop rollercoaster in the USA called Racer.  Which means you start out on one side and end up on the other.
« Last Edit: July 07, 2018, 03:27 PM by Weber »

*

Offline Bullethead

  • *
  • 1936
  • shyguy's World Rocks!
    • View Profile
Re: Quasi-Möbius Coasters
« Reply #16 on: July 07, 2018, 09:55 PM »
Well because all of the prior info I gave was relevant to your original post.  You see Kennywood has the only existing Mobius Loop rollercoaster in the USA called Racer.  Which means you start out on one side and end up on the other.

Yeah, I appreciate your experience with the thing.

QUASI-MÖBIUS COASTERS PART 3
[/b]

OK, I've done some more work on this.  It's now PERFECTLY synched and gives excellent races every run.  I also ironed out some of the hastily built janky areas so it's all nice and smooth now.

Updated Workshop link:  https://steamcommunity.com/sharedfiles/filedetails/?id=1434436567

How I Got it in Sync
Recall that the 1st lap took 85 seconds and the minimum station time was 25 seconds.  I figured some math better than just finding their LCM might work better.  I couldn't adjust the lift hills without changing the lap time, but I figured I could slightly increase the station time.  Here's how I did that:

85sec - 25sec = 60sec

60sec / 5 trains = 12sec

So I added 12 seconds to the min/max station delay:  25sec + 12sec = 37sec.  I made both the min and max station delays that long.  AND, because those times can be overridden if the trains fills up before then, I ALSO set the departure interval to 37 sec.  So now the station time is definitely constant regardless of the number of peeps on the train.

Using 37sec for all 3 of these times immediately made the coaster settle into a very consistent pattern, doing the same thing every run.  However, the train coming from the station (the train starting its 1st lap), always hit the lift hill about 2 lengths behind the coaster on its 2nd lap.  Because the 1st lap coaster is on the outside of 2 of the 3 turns, it just got further and further behind over the course of the lap.  So while it could usually see the other train in the distance ahead, it wasn't really racing.

At this point, I had the option of tweaking the load time again or slowing down the 2nd lift, so the 1st-lap train would catch up on the lift hill.  Because the latter option would have changed the 2nd lap time and I was now tweaking the station time assuming the lap times would be constant, I decided to change the station time again.  I reduced it a couple seconds to 35, so the 1st-lap train would start sooner.  This improved things.  Now the 1st-lap train was about 1 car behind the 2nd-lap train going up the lift.  So it still always stayed behind it and still got further behind, but at least it was always fairly close.

Finally, I reduced the station time to 34 seconds.  For whatever reason, this worked out perfectly.



How the Race Works
So now the overall ride works like this once it's run a full 5-train cycle through the station from a cold start.

1,  Train A, starting its 1st lap, leaves the station a split second before Train B and Train C come up behind.  Train B enters the station to unload whille Train C takes the bypass at the start of its 2nd lap to race Train A.

2.  Train A begins about 6 lengths ahead of Train C but is on the outside of the pre-lift turn and moving slower than Train C.  As a result, Train C mostly catches up with Train A during this turn.  When the get to the lift hill, Train A is only about 1 car ahead of Train C.  As they start the lift hill, Trains D and E pass overhead on the diagonal back stretch.



3.  Because Train A gets off the lift chain first, it pulls slightly more ahead of Train C coming down the 1st drop and then back up to the 1st turn.  However, Train C then has the inside lane so during the course of the 1st turn, so by the end of the 1st turn, Train C is about 1 car ahead of Train A.  They maintain this formation all down the front stretch over the station.



NOTE:  Isn't that an awesome pic?  I'm rather proud of it, anyway ;).

4.  Train C also has the inside lane on the 2nd turn, after which it's about 2 lengths ahead of Train A all down the back, diagonal stretch.



5.  But in the final turn, Train A has the inside lane and NEARLY catches back up with Train C.  Train A's nose is about even with Train C's tail all down the home stretch under the lift hill.  Then Train C crosses over Train A still slightly ahead, so everybody on Train A can see this happen.



6.  As Train C slows and stops in the station, Train A catches up and passes just as Train C is stopping, so it's essentially a tie.  Meanwhile, Train D has just left the station.  On its 2nd lap, Train A will race Train D, only with the roles reversed from its 1st lap (IOW, Train A now plays the part Train C just played).



All in all, this is about the best possible outcome.  There's actual racing with passing and varying distances between the trains, ti comes right down to the wire, the upper train crosses just ahead of the lower train, and it's all repeatable for all trains on both laps.  Please forgive me for patting myself on the back :).

HOWEVER....
The overall coaster has "Low" reliability due to all the moving parts:  2 lifts, 9 block brakes, and 5 trains.  It's down for repairs about 6% of the time and needs frequent refurbs.  I tried to improve this by removing the block brakes at the 3rd turn (B2 and B5), which aren't needed once the coaster has warmed up to operating rhythm.  UNFORTUNATELY, those brakes are needed during the initial warm-up cycle.  Without them, the 2nd train stops at B1 and then doesn't have the oomph to get up the hill at the 3rd turn where B2 used to be.  So I had to put these brakes back in.

All in all, this coaster has stats that make it hugely popular to all ages.  It can process about 1500 peeps/hour.  It has 771 prestige without any scenery at all so fully pimped and with a couple of triggered effects, it'll be close to 1100.  So it can make money, but lots of this will go into a mechanic who can't stray too far away because this coaster needs constant attention.  And it takes up LOTS of space.  But hey, I think it's cool anyway ;)

-Bullethead
NIHIL INIQVIVS QVAM ÆQVITATEM NIMIS INTENDERE
My Steam Workshop page

*

Offline Bullethead

  • *
  • 1936
  • shyguy's World Rocks!
    • View Profile
Re: Quasi-Möbius Coasters
« Reply #17 on: July 09, 2018, 05:54 PM »
QMCs:  SUMMARY OF RESULTS SO FAR

The whole point of me messing with QMCs was to try to learn the underlying principles and condense them down to a set of general guidelines.  I think I've learned enough now to do that.  So, here's the TL/DR version:

1.  Choice of Coaster
QMCs can be built using any type of coaster.  And really, going with steel or hybrid is easier and makes a more exciting ride.

If you want to replicate a classic wooden racing coaster, with the tracks edge-to-edge and perfectly parallel all the way around, then construction becomes quite fiddly and time-consuming fitting the 2nd lap to the 1st.  Also, you can't bank the turns so can only turn at the tops of hills where the trains are slow.  This leads to long, skinny track layouts.  Also, the relative positions of the racing trains won't vary along the straightaways as they're both the same.  The race outcome is determined entirely by the relative starting positions on the lift hill and which train is on the inside of turns the most often.

With other types of coasters, you're not tied to having the tracks be perfectly parallel forever.  This means you can have essentially any layout you want, along which the 2 tracks can spiral around each other, etc.  Due to small-scale differences in track length and much more freedom when it comes to arranging turns, the trains will jockey back and forth all along the track, making for a more interesting race.

2.  General Track Layout
This is pretty much building 2 of the same coaster next to each other.  You make the 1st lap but instead of taking it into the station when it runs out of momentum, you take it back to the lift hill and go around again.  Done this way, both laps will take very close to the same amount of time and all will be well.

3.  Number of Trains
This will always be an odd number of 3, 5, 7, etc. because 1 train will be in the station while 1 or more pairs of trains are racing.  The number of pairs is the same as the number of trains the coaster could run if it was just a single lap instead of a QMC.  IOW, if the track only went around once and could run 2 trains (1 on the track and 1 in the station), then a 2-lap QMC version of this coaster would need 3 trains (1 pair on the track and 1 in the station).  A 1-lap coaster able to operate 3 trains (2 on the track, 1 in the station) would be 5 trains as a QMC (2 pairs on the track, 1 in the station).

4.  Train Spawn Locations
The closer trains spawn to the station, the less time it takes for the coaster to warm up to its operating rhythm, usually after about 1 full cycle of all trains once through the station.  If the trains start scattered all over the track on block brakes and lifts, the bigger the track, the longer it will take the races to sync up because the further the coaster starts from its operating condition.  There are 2 problems with having all the trains spawn on the station brake run, however, to keep in mind when designing the track:

* If you have a lot of trains, you'd better make them all pretty short (like 3 cars) or the brake run will be grotesquely long.

* If the track is long enough that you need mid-course brakes  (of which you need at least 2, 1 for each lap), and you have a block brake for each train at the station, you will have a LOT of block brakes.  Plus 2 lifts (at least) and however many trains.  This will hurt ride reliability---expect frequent breakdowns so have the coaster part of a work roster whose semi-dedicated mechanic never strays far from the big QMC.

5.  Track Ingredients
Build the track as you normally would a 1-lap coaster, just go around twice.  The only possible exception is that if there are ZERO mid-course brakes (IOW, a small coaster), you MUST have 1 block brake at the bottom of the 2nd lift hill.  Due to its placement at the start of the 2nd lap, this IS the mid-course brake.  OTOH, if you have mid-course brakes elsewhere, then you PROBABLY don't need the block brake at the bottom of the 2nd lift.

6.  Station Configuration
The trains will only sync up consistently if the time each train spends in the station is constant.  This means 1) setting the min and max delays to the same number AND, because these can be overridden if the train fills up before that time elapses, 2) setting the departure interval to the same time as the delays.

For ANY coaster, QMC or not, the ideal situation is that the in-station time is the minimum amount possible, AND a new train enters the station within 1-2 seconds of the previous train leaving.  You don't want trains to stop anywhere but in the station to maximize throughput and minimize peep boredom.  Thus, you have to make the station to minimize in-station time.  But with a QMC, you can then up the in-station time a bit to tune the timing.

In general, you want to keep the in-station time at 30 seconds or less.  For any trains of 5 or more cars, this requires the exit to be centered on the narrow side of the platform, on the opposite side from the entrance.  If the trains have only 3-4 cars, you can safely have the exit on the same side as the entrance and still be under 30 seconds.

HOWEVER, putting the exit on the opposite side as the entrance means either the station has to be far enough to the side of the bypass run to allow you to place the exit path.  This creates a dogleg in either the track going to the station or in the track bypassing the station, which will make the 1st and 2nd lap times differ by a few seconds.  This will complicate tuning a bit (see below).  Also, entrance and exit on opposite sides of the platform means one or the other will be inside the loop of the track, so you have to provide a means for peeps to get under or over the track back to the main path.  Going under is usually easier as going over means a either a huge staircase structure or a huge station building, take your pick.  Going under MAY allow using a tunnel below grade.  But if that's not an option, you have to have the station and/or nearby track high enough for peeps to walk under.

ALSO, be sure to set the load to ANY.  This will help get the coaster up to operating rhythm faster from a cold start as the number of peeps won't change the time in-station.  If the ride is worth a damn, it'll have a full queue so will always go with a full load in real life.

7.  Tuning the Timing
First, 2 important caveats.  #1:  ONLY test the timing of you QMC with real live peeps.  Test dummies always have a shorter in-station time than you set.  So build the coaster in a test lab empty park where you can feed in real peeps without screwing up your real park.  #2:  ALWAYS let all the trains go through the station at least once before making any judgment on the timing, as it takes that long for the trains to get in proper formation around the track, more times if the trains spawn spread out all over the track. 

In general, the lap times will be essentially equal unless your station configuration requires a dogleg due to the exit being on the opposite side of the platform from the entrance.  Then, the 2nd lap will be several seconds longer than the 1st.

There are 3 numbers that determine a QMC's timing: 1) the in-station time, 2) the time of the 1st lap (measured when a train beginning its 2nd lap passes the stopped position at the station), and 3) the time of the 2nd lap.  The 3rd factor is only significant if there's more than several seconds difference between the durations of the 1st and 2nd laps.  Dealing with 1st 2 factors should be done first because that will sync things up for the 1st lap and also the 2nd if their durations are within a few seconds of each other.   If the coaster isn't synched yet for the 2nd lap, having the 1st lap synched will make it easier to sync the 2nd.

So, AFTER the coaster has run at least all its cars through the station once, with a stopwatch in hand and the coaster open in a test lab park with real peeps, measure the in-station time from the instant a train stops until it moves again (TIS = Time In-Station).  Then measure how long it takes for a train to complete the 1st lap, ending timing when it pulls even with where trains stop in the station (TL1 = Time for Lap 1).

TIS is probably not going to be an even factor of TL1.  But much can be done by averaging that out, as follows:

TL1 - TIS = Amount Of SWAG (AOS) and, for them as don't know, SWAG = Scientific Wild-Ass Guess.

AOS is how much time you need to spread among the loading times of all the trains to get the station to cycle with the ride's 1st lap.  So....

AOS / # of trains = STI (Station Time Increase).  Add that much to all 3 of min and max station delay, and also departure interval.

Then see what happens.  HOPEFULLY, you built the coaster so that increasing the station time by up to 15 seconds or so won't cause trains to wait on the brake run while the train ahead finishes loading.  And the 1st lap of the race should be synched up reasonably well.  It should at least be giving fairly consistent results at this point from train to train.  Then you can tweak the in-station time up or down a few seconds to see if that does the trick.

if the times for the 1st and 2nd laps are significantly different, perhaps due to a major dogleg to make room for a the station exit, you can also adjust lift hill speeds.  I recommend starting them both at max and then tweaking 3 or the other, or both, down from there.



Anyway, that's about all I know.  There are too many variables in design from coaster to coaster to give absolute rules, but the above are things you should be thinking about as you build QMCs.  I recommend building a very small, 3-train QMC as a test article and play with it some to get the feel of things before trying to make a huge ride.

Have fun!
-Bullethead
NIHIL INIQVIVS QVAM ÆQVITATEM NIMIS INTENDERE
My Steam Workshop page