Helpers in the helix...

Dear Dick,

In order of appearance

Um, yep!

Cool, sounds like you've got the core engineering down cold...

Understood, however this almost certainly mandates:
- straight "run-thru" operation, don't try to 2nd-guess yourself with starting and stopping "blind"
- "peekaboo" holes
- "Herniated" loops
(mid-helix scenes with significant scene height, due to "popping out" of the helix footprint)
- CCTV
- LED "worm" diagrams
- or some other form of visible confirmation

Understand and appreciate the preference, but you will create container-loads of frustration without some form of visual feedback over such a long hidden stretch of trackage.

It sounds logical, but in reality is a recipe for string-lining, or alternatively dumping the entire train worth of cars into the centre of the helix (and taking a short/fast/loud trip to the floor).

Helper ops on model RRs really engage the Human operators when the entire train is visible and in-motion. Hiding the train that you're pushing up the hill looses most of it's visual visceral appeal...

How indeed!

- If you are modelling prototype DPU operations,
where the mid and rear locos are controlled by the head-end crew via radio link,
then it would seem using the Consist function on any decent DCC system would fit the bill.
The only "gotcha" is speed matching both the Head-end and helper locos, and having the confidence that the matching is sooo good that you can trust it under "sight unseen trackage" conditions

- If you are modelling literal "head-end with head-end crew, helper locos with seperate helper crew" operations, then sure, setup 2 sets of 12"/1' scale humans with seperate DCC handsets, and head for the "hill". This will test the communications and throttle<> loco response behaviour of both crews,
and yes, with no visible feedback both crews will be "running the hill blind"
(Cue stringlined/dumped-on-the-floor cars again...)

Concievably, yes.

Technically, sure.
(The typical DCC system will consist as many locos as you want,
whether they are physically coupled together or not is technically irrelevant)

Practically, it's the "unknowns"/"unpredictables"/"variables" of the process:
- train slack action
- load of the train on the Head and Helper locos as they start up the grade
(Lead locos slow as they hit the grade, helpers keep pushing,
cars pop out of the middle of the consist, it's a predictable story...)
- any momentary/temporary speed differential as the locos + trains traverse the hidden trackage
​(hump over a track joint, slip on a feeder solder lump that wasn;t perfectly finished, etc etc)

which will cause an otherwise "theoretically electro-mechanically do-able" situation to rapidly "come apart at the seams".

Whether the helix is hidden or entirely open will not change the "practical success rate" of operating in such a fashion over "the hill".

However, an "open visible helix" will allow the operating crew to have a lot faster and accurate realtime way of assessing "is everything OK, and if not, what can we do about it?".While the helix is entirely hidden/in-accessible, 

the first thing the crew will know about any "whoops" is the sound of it being waaaaaay tooo late to do anything about it...

Said another way:
- A visible grade mens the crews are engaged to actually/actively "drive up the hill",
- a hidden grade means the crews "set it in run8 at the hill, and hope for the best"

Seriously, if modelling the helper ops over Bozeman is the focus of the layout,
use the space to model the Bozeman grade "on stage",
and leave the hidden helix to do what it does best,
just make sure the trains "get back to where they came from" as efficiently and quickly as possible...
(while staying upright and on-the-rails, obviously... )

Happy Modelling,
Aim to Improve,
Prof Klyzlr

PS with RailPro's "interactive consisting/speed-matching", it could be very appropriate for such a situation...

PPS can't recall the layout name at this time, but there was an N scale coal-hauler layout in MRP some years ago which mated a ovalised helix with a NoLix to provide "coal hauler helpers" ops challenges. Maybe some relevant lessons to be had RE "which bit needs to be onstage/modelled fully, and which bit is best served by the helix"... 

First off let's assume (I

First off let's assume (I know!) for a moment that when done building the helix, that you will test it!  Right?  That should give you the confidence or lack there of if it's going to work or not.  If it works cover it up and put your helpers at the end, if ti dosen't put the helper at the front and cover it up!  My 2 cents worth, BTW, we have two helix's on30 NCE dcc and we can consist locos anyplace in the train with confidence. Russ.

Thanks for comments....

Well Prof, ALL of what you listed, in all the gory details, has been going thru my addled brain for awhile now. Knowing my own (in)experience, I tend to agree with you. But then I have a built-in engineering-paranoia factor.

I am somewhat limited in space, so I really cannot do the whole hill - or even a major part - in the clear. I don't have access to 'around the walls' in the Bozeman Pass area - in fact it is mostly a hallway. The only way I can get an elevation change of 15 - 20 in. is via some kind of helix.

But stringlining is not something I want to go through, either. One option I do not particularly like is to place two-three 'dummy' engines, one with sound, say, on the back end. This loses realism when the train has to back into the helpers, rather than the opposite.

I do intend to somehow test it, Russ. Thanks for your observations. I am trying to devise a way to test the entire helix (3-4 loops) in place without ripping up the existing track(!). I have some confidence in my quality of laying flex-track - I've had this two-room expanse running for several years with almost zero derailments (not intended to brag, just gives me some confidence). So I'm pretty sure I will be able to build the thing right. BUT - other issues, like coupler alignment, weights, total weight vs engine tractive power, wheel quality (all metal on the helix, I hope), will become far more important. And of course, my own ability (or lack of it) and that of guests'.

It is encouraging to hear Russ' experience with consisting on helices. It gives me confidence to try, anyway. I think there will be a whole lot of more 'quality control' than I am used to, however.

So I'll attempt it. Thanks to you both for ideas and encouragement. And warnings.

marco

what scale?

I'd recommend no live,

My helix is HO scale, 36"

My helix is HO scale, 36" radius and is 8 full loops from top to bottom. I've tested 25 car freights up and down the helix with helpers, no stringlining (even with some really light cars thrown in for testing purposes) or derailments. One thing I think I may do is to have the helpers consisted with the lead units, thereby eliminating any troubles due to throttle mis alignment. If you are going to run with 2 operators (freight and helper) then make sure there is an understanding between the operators that any sign of trouble BOTH ops will hot the Emergency Stop buttons on their throttles. I haven't got my layout fully built yet, so I haven't ran op sessions yet, but I am thinking in a helix that will be enclosed consisting all power in a train is the most trouble free way to go.

-Trevor

Here is another approach

Hello,

my last layout had a 30 inch radius helix.  Two engines were able to pull up 20 cars reliably in HO.  Being in DC only I still managed to use rear train helpers.  I found the trick was to use engines that were speed matching and at the same time were very light.  If I were in your position I would take most of the lead out of the SD45 that would be used as a pusher.  Once the engine is speed matched and  capable of pulling a couple of cars up the helix it would make a good pusher.  If it runs faster then the lead engines it will not be able to push cars off the track, if it runs slower it will be some more weight to the train but hopefully not enough to string line the freight.

Hope it helps

Frank

Ditto - what scale?

Ditto - what scale?

Active speed matching

Hi,

on a more humorous note... Why not actively speed match your pusher engines?

All you need is a coupler that slides a bit lengthwise, two springs to locate the coupler in the center of the slide, two small switches and some speed matching electronics.

If the helper is pushing too hard, the coupler will be pushed into the channel and the helper needs to be adjusted down.

If the helper is too slow, the coupler will be pulled upon and the helper needs to be adjusted up.

Put the active matcher in a dummy loco which is permanently coupled with the "live" helper, but the matcher has to go between the train and the pusher loco.

Have a funny weekend...

Front end helpers

On my CNY&NE RR, I have a 36" diameter helix and plan on using front end helpers when necessary to avoid straightlining derailments on the helix.  I have experimented and even carefully speed matched engines will cause derailments on the helix with enough frequency to make rear end helpers a rule violation.

Dummies?

Is the operation of helpers your main goal, or would the "appearance" of helpers be enough?  Consider your helpers being dummy locomotives.  Duck your train into the helper pocket and pick up the dummy.  The next train down brings it back.  It looses a bit of realism, but exchanges for not worrying about derailing.

Scale is HO. ...sorry

Thanks for all the comments. My omission - my layout is HO.

It seems (to me) that radii of 31 and 27 (redone: now 28) inches are adequate, based on other reports. Bear Creek, why would you consider them too skosh?

My main layout has min. radius 24 inches. Everything works just fine up to and including 72 footers (no 80 foot cars). For the other comment, SD45's have no problem at all on my 24 in. curves.

The main thing I am interested in is the issue of helpers - blind. At least two people on here have expressed concerns similar to my own; these are folks that I have read on other issues and whose opinions I respect.

On the other hand, several others have chimed in that, essentially, speed-matched blind helpers work just fine on their helices. I have to take them seriously as well. Speed-matching is not something I have tried, yet. But I do have Decoder Pro / JMRI and I've read that makes it easier.

I think I'm going to build it and see. Anyway, if it turns out I cannot reasonably run back-end pushers up the helix, I will just have to restrict that, or use (lightened) tail-end dummies with sound. At least I will have (finally) a second deck.

I'm still interested in any other opinions.

marco

Oops - more on 24 in. radius...

I had said above, "My main layout has min. radius 24 inches. Everything works just fine up to and including 72 footers....".

This was not completely accurate. I have noticed uncoupling issues on Walthers 72-ft centerbeams (no actual derailments). But this happens also on larger radii curves (sometimes on straights?), and these things come with brand-X couplers (which need replacing with KD's).

marco

Can you start the helper district at the top of the helix?

Maybe you can build this into your plan in case using helpers on the helix doesn't work out, or if some operators don't feel up to the challenge.

The plan?

   Can you clarify the layout plan a bit? will the helpers run on the visible lower and upper levels or just in the helix?  If they are just a means to shove cars up the hill between levels without being seen maybe a train elevator could be used instead,248 inches could hold a pretty long HO scale train. If the helper operations on the helix are part of the operating plan why not open up the aisle side of the middle level and watch the trains run up at least part of the hill? With such a long oval helix it should be possible to weave the  tracks of the  levels to allow quite a bit of visual interest to the climb..DaveB

Clarification...

Well, Bozeman and Belgrade MT reside in a room about 12' x 15'. Belgrade has a (nonprototype) return loop. The room has two doors in one corner (really inconvenient). One door leads to a hallway about 24' long, of which about 3 ft of width is available, except for a blob at the far end, which is where the helix will go. There really is no room to widen the hallway. The plan has a grade from the room-joint (East Bozeman) to the far end where the helix starts. There are 'signature items' along East Bozeman to the beginning of Bozeman Pass.

The helpers have typically tied on at East Bozeman, which is about where the current 'hole in the wall' between Bozeman and East Bozeman is. I want to model this action, following them down the hall and up the grade to the helix; at the top of the helix is 'West End', a long siding (back down the hall) between the top of grade and the tunnel at the top of the pass. (In reality the top of grade is approx at the mouth of the tunnel, which is about 3/4 mile in length, and travels under I-90 to the East Side, another siding known as Muir.

So I want to follow the helpers from one end of the hall (East Bozeman) to the other end (helix), then they will travel up the helix, become visible again at West End and travel (or decouple) back down the hall to the tunnel. What comes next in the model is TBD.

I KNOW this is 'insincere' according to John Armstrong. 'East' and 'West' should remain consistent to an operator, left or right.

I can't make that work without far worse results. And its MY railroad. So there. Operators will just have to adjust.

Actually, the prototype goes down the East side (Muir) to Livingston MT where there is an enormous engine terminal, the Livingston Rebuild Center. This was NP's major locomotive service center, now MRL's. It would be a terrific thing to model, even with Armstrong's 'Selective Compression' or 'Suggestive Omission', and I have been dreaming for several years how to fit that into my space. So far, unsuccesfully.

However, the furnace room beckons.....watertankfurnacewellpump.ville.

Thanks again for comments.

marco

Speed Matching =/= constant/consistent behaviour

Dear Dick, et al,

There's something which has been bugging me...

One can certainly speed match the loco-mechs/consists on flat straight track
(IE a constant-set-of-conditions) all one wishes.

However, between the semi-predictable effects of 
- curve radii (which may or may not be constant,
indeed with an ovalised helix we can guarantee it won't be!)
- angular movement
(again, check the effects of the ovalised path)
- the increasing load as more and more of the train progressively begins climbing the grade

which will slow the lead locos as they enter the helix/grade,
(while the rear helper "bunch slack" from on the flat, against the increasing gravity/friction resistance,
having not yet begun climbing/curving)

and the range of unpredictable inconsistencies and "gotchas"
(Murphy's Law has not yet been repealed)

IE:
- even the slightest kink at flex<> flex length joints, in an otherwise "perfectly constant radii curve"
- the slightest hump/dip in the rail
​(inc L<> R camber)
- a coupler-set "suddenly" deciding it cannot sustain it's current alignment under Helper-shoves-consist compression, and "flopping over" hard against it's side-stops
- either the lead or helper consists momentarily coming off/online

there are many issues which are completely independent of raw "speed matching",
which can cause various forms of ripple excess/de/recompression of the train between the 2 loco consists,
with predictably disastrous results...
(forget multi-turn helixes,
I've witnessed such issues on HO layouts with "simple" 180-degree 3' radii 2" graded curves...)

Given that in a completely hidden helix there is no "visual payoff"
(the "feel" of standing trackside and witnessing both lead and helper consists give-it-all-they've-got to crest the grade,
which IIRC is what is prompting you to tackle this in the first place),

for all of the engineering (mechanical and electrical/electronic) required to make this system
"set the train on-its-way, set down the throttle and turn your back,
and be absolutely assured that it will appear unscathed out the top first-time-every-time" reliable,
(and to aim for anything less in the reliability stakes is to put your "enjoyment running the layout" quotient at serious risk... plenty of layouts worldwide that look good but are rarely run because they don't run reliably)

I can't help but feel that the current plan is setting you up for a lot of work (and potential pain) without actually delivering "what you really want to model"...

I hope I'm wrong...

Happy Modelling,
Aim to Improve,
Prof Klyzlr

"The helpers have typically

"The helpers have typically tied on at East Bozeman, which is about where the current 'hole in the wall' between Bozeman and East Bozeman is. I want to model this action, following them down the hall and up the grade to the helix; at the top of the helix is 'West End', a long siding (back down the hall) between the top of grade and the tunnel at the top of the pass. (In reality the top of grade is approx at the mouth of the tunnel, which is about 3/4 mile in length, and travels under I-90 to the East Side, another siding known as Muir."

     That sounds like it would work quite well. Leaving East bozeman the train could gain about 6 inches of elevation in the 24 feet before reaching the helix, then circle a couple of helix laps and emerge to climb another 6 inches to the tunnel so as much or more of your climb will be visible than hidden. Sounds like a good plan to me, and I wouldn't worry about the train going east to the right on one level and east to the left on the other level, just design the scenery to be the view from the opposite side of the tracks( viewing from the south side on the lower level and from the north side on the upper level) and it will still have the correct direction from the operator's perspective. ...DaveB

for Prof...

Prof, those are the kinds of ideas that have been bugging me as well. However, based on the comments here, it would seem that some people have had success in similar environments. Of course, we do not know all the details....

My plan is to build it and see. With my typically lightning-fast-workspeed there will be some expanse of time (! probably months !) between completion of helix wood, track, and wiring - and application of 'scenery' to cover the helix. So I should have plenty of time to run trials of various 'speed matched' engines and combinations up *and down* the helix and see what works and what doesn't - while I can see clearly what is going on.

That's the plan. 'Dave B' summed my viewpoint up pretty well. I do plan the long 2% ramp (with vertical easementd) leading to 3 laps (all four vertical gains about 5 in.), so I will be about 20" high. There is a furnace cold-air return in the way....another story.

PS - how does one post .jpg's on here?

marco

Bozeman Pass on Google Earth

Just for info, the current (8/25/14) view of Bozeman Pass on Google Earth shows a train of hoppers Westbound, with what might be a Santa Fe (BNSF) yellow-blue on point and tail ending yet in the tunnel under I-90. I also just now discovered (! on GE) that there is a third spur up there on the south side of the main line, which I've never noticed. It starts at the West End shack and extends about 13 hopper cars to the west. Very short. Probably a 'pocket' for helpers.

Also, looking carefully one can see the concrete footings of the old 2-track standard signal bridges that were up there, one on the West side just East of the West turnout, and one over just down from Muir. There are photos of those in one of the MRL videos.

marco

@ Marco

When you go to test your helix you might want to consider a barrier to hold the train in place should it string-line as the Prof mentioned.  You could use Masonite strips on the inside course just as a safety precaution.  Remember, you are testing and learning.  One oops and you might be picking up parts and pieces.

Another thing I recommend is that you "time" your train climbing the helix. I have read so many articles about layout owners mentioning their operators get nervous when the train is out of sight for so long. Unfortunately some will increase the throttle speed to "speed things up."  This causes that sudden surge that string-lines the train. Or the train comes rocketing out of the top of the helix like it has been fired out of a gun barrel.

So here is another idea.  Once you get the helix built and start your testing, time the climb, etc. put a sheet of card between you and the helix.  This will hide your train from view.  Now for more testing wIthout actually timing it, see how long it takes to reappear in your mind.  Ask yourself, " Did you get anxious?" "Did you get impatient?"  Let us know how you feel about this.

You could build what I call a herniated helix.  It is the same thing as a herniated disc in a way. One level is wider, bulging out than the other levels.  It pops out of one tunnel, runs across the face of the cliff on a ledge to disappear into another tunnel.  You only need a few inches more width to make it work.  It will also break up the amount of time your train spends hidden in the helix.

Den

Comments and some N scale experience.

Marco - just a few comments that may or may not be useful.

1) Somewhere I got the impression that your helix will be hidden track.  Is that correct?

2) I regularly run an N scale 36 hopper coal train with two locomotives on the head end and one on the rear.  I've done this at various times with locomotives from Kato, Fox Valley and even one from Bachmann.  Decoders are from TCS, Digitrax and one "SoundValue" from Soundtraxx.

3) I've never had the opportunity to do this with a grade higher than 1%.

4) The locomotives are all speed matched using the 28 step table. The are within 1mph of each other at the ends and a couple of mid points.

5) I have much better results with BEMF On.  You'll find a lot of advice on the Internet to the contrary.  Most recommendations are to turn BEMF off when speed matching.

6) I have also had better results with longer trains.  It seems the more cars in the train, the more slack is available to take up small differences in speeds between the front and rear locomotives.  You can hear the slack bunching and stretching, especially at the beginning of even the slightest grade. 

7) I have most problems in the first few minutes of running when the motors are cold.  After a few minutes of running, everything warms up and runs rather predictably.

8) I've never had derailments due to the trailing engine going too slow and stretching the train.  Every derailment I have ever had has been due to the trailing unit going a little too fast and turning the train into an accordion.

9) Personally, I would never do this blind.  At least, if it were a blind area, I'd make sure I could easily get into it to clean up the inevitable mess.

10) No matter how well you plan this, you will have a derailment.  Most likely when visitors are watching.  That's the way the decoders work.  They have visitor detectors built in.  I think it's part of the NMRA spec that is only shown to manufacturers.

Just my experience.  It's only one data point so I don't know how useful it will be  Have fun. gs 

Some extra weight in the cars

Some extra weight in the cars may help the flanges stay in contact and keep the train stretched out, worth trying to do if you're using extra power anyhow.

Outcome??

Marco,

Found this post on an MRH search and read about your proposed helper service on a hidden helix with interest. I'm hoping to do the same kind of thing on my current layout under construction.

I was wondering how things turned out for you? Have you done any tests with helper service up the helix yet? Any experiences that you would care to share with the rest of us?

Thanks!