Tehachapi, BC

[ednadolski]

... if you do a lot of speed matching work ahead of time.

And afterwards.  I've found that speed matchings tend to drift over time and occasionally need to be re-calibrated, esp. if the locos have sat unused for a long while.

Ed

[GaryHinshaw]

Gary, you may find that DPUs consisted all together work okay if you do a lot of speed matching work ahead of time.   But I've always run them separately controlled (which is prototypical). 

I have run quite a number of trains that way with my DT402 and Engine Driver throttles, and it generally works fine for me (and it's really fun!).  But I find that my operators are less comfortable with that, at least in the early days, for fear of putting a whole cut on the ground.  I do intend to give crews the option, but I don't want to force it on anyone.

[pdx1955]

But I find that my operators are less comfortable with that, at least in the early days, for fear of putting a whole cut on the ground.  I do intend to give crews the option, but I don't want to force it on anyone.

I'd be too, especially since the ground is "hundreds" of feet down with no adjacent scenery over much of the line. It's really easy to jackknife or stringline cars if your not too careful with helpers. Articulated cars also have a tendency if ones going to go over the whole things going to go. Remembering problems with running helpers on a HO club layout years ago, as dispatcher, its not a pleasant sight to see a whole stack train flip over scattering containers down a slope, instead of making that running meet that you had planned...

[jagged ben]

Well Gary, when I think about it I suppose that's rather logical on the part of your operators.  If they put a cut on the ground using ostensibly speed-matched locos then they can just blame your lack of adequate speed matching!   

[ednadolski]

FWIW, one other trick that perhaps might be of use here is the Mike Danneman 'derating' technique for diesels:   basically removing one of the drive axles from each loco in a DPU set so that they cannot push as hard against the cars in front of them.   It does come with its own tradeoffs, but may be useful in cases where the DPUs tent to 'accordion' the cars in front of them, or drive cars with truck-mounted couplers off the rails due to excess compressive force.

Ed

[mark dance]

No risk; no reward. Some of my thoughts on pusher ops after 6 years of having pushers in ops sessions...

I believe that to fully enjoy independent pusher ops there needs to be some level of train handling challenge for the crew to overcome. Which means there is a risk of operators failing that challenge which brings intrinsic risk to equipment. And over the course of a 3 hour ops session, a challenging pusher op schedule is likely to result in derailments. Net result is that derailments should be expected and accepted.

Note that the derailment risk should *not* come about because of something artifical like an operator not being able to observe (and hear) their train but because the operators are not being adequately observant and responsive.

The handling challenge should not be so hard that it is impossible and repeated attempts should allow learning to occur and success to predictably increase.

On the C&W I estimate about 50% of first time pusher crews put their train on the ground. Nothing disasterous in most cases - just wheels on the ties. This failure rate goes down about 50% with each attempt and never reaches zero. The crews learn to be very observant of their trains and to coordinate closely  with the head end crews so the layout design cannot impede these necessary activity.

Note that my opinions would not apply to DPU pushers. In this case the mid and end train helpers might as well be there just for cosmetic purposes and the objective in that case should be zero derailments.

Just my $0.02

Md

[Cajonpassfan]

I don't have a lot of experience with pusher ops yet, having only tested it a couple of dozen times on my circa '50 Cajon Pass layout, but pusher ops are an integral part of the plan, and I have learned a bit along the way.

My uphill grade is generally 2 in 96, or slightly over 2%, with a few flatter spots along the way. Minimum radius is 18" "compensated", meaning the grade drops to about 1.7% on curves. Train length is limited (generally) to 25 cars, mostly forty and fifty footers, with head end power and pushers ahead of the caboose.

Things I've noted so far:
In pusher ops, "slinky" is not necessarily a bad thing Longer trains with helpers are more forgiving than shorter trains.

Cars need to be somewhat equally weighed; err on the heavy side of the NMRA recommendations.

Thinner wheels and smaller flanges look terrific, but they do find every less-than-perfect spot on the railroad. Not necessarily a bad thing, it is what it is. Fix it. If you can't fix it, get deeper flanges. I'm fixing it...

The road power should "need" the rear end helpers, ie, too much power at the head end can stringline your pusher equipped train in a hurry. The pushers should only have enough power to push themselves and 8-12 cars ahead on the grade to avoid putting cars on the ground. (If the head end stalls for some reason, the pushers should spin in place without pushing the train off the track) This means "detuning" the pusher units as noted in the Danneman articles...
This means more than removing an axle; Mike disconnects an entire truck. In my case with steam helpers it means having a fine running loco with no traction tires that can't pull much (like the original LifeLike mallets with no traction tires).

I like separate helper crews, but it's a special assignment that requires some training; 50% failure rate is not what I would consider okay..
Time will tell...
Otto K.

[Lemosteam]

Thanks Scott.  Sadly it is the same test print that I first tacked up 4 years ago...  However, I do rather like it still, so I probably won't change much when I get it professionally printed, whenever that is. 

-gfh

P.S. In scanning through the archives to find the above link, I came across this pre-Vortex photo where I was trying to picture how the Vortex would fit under the Loop shelf:



Here's how it fits:



(For ops sessions, it is very helpful to have the Vortex visible like this.  I need to move the tall cabinet that currently blocks much of it.)  It's been a long haul, but with track and wiring almost finished, I'll finally be able to concentrate on more cosmetic projects.  That's when the real fun begins.

Nailed it! (Well, likely screwed it)

[SAH]

No risk; no reward. Some of my thoughts on pusher ops after 6 years of having pushers in ops sessions
...

I believe that to fully enjoy independent pusher ops there needs to be some level of train handling challenge for the crew to overcome. Which means there is a risk of operators failing that challenge which brings intrinsic risk to equipment. And over the course of a 3 hour ops session, a challenging pusher op schedule is likely to result in derailments. Net result is that derailments should be expected and accepted. ...

Just my $0.02

Md

Mark's comments are dead on from my experience running on layouts with trains requiring helpers.  Some of the trains are designed to require helpers.  Other stall on the hill creating a "situation" for the train crews, the dispatcher and the railroad in general.  When the helper crew, the train crew and the dispatcher are all working together to get the train over the mountain everyone's attention level is at a peak.  Just like real railroading.  I won't say we never have a derailment but it is not a frequent event.  It's also great fun.

Steve

[GaryHinshaw]

No risk; no reward.

Couldn't agree more with this.  (Of course I have benefitted from Mark's input throughout this build.)  One of my primary goals is to mimic the operating challenges faced by the prototype and Steve's scenario is a perfect example.

Note that my opinions would not apply to DPU pushers. In this case the mid and end train helpers might as well be there just for cosmetic purposes and the objective in that case should be zero derailments.

I'm not yet comfortable running dpu's consisted with the lead units, but I am comfortable with 1 person controlling two sets of independent units.  There is often some amount of fine tuning required on the pusher's throttle.  Independent control also forces you to pay attention.

Nailed it! (Well, likely screwed it)

Thanks John.

[jagged ben]

I'm a little surprised to see people here suggesting that rear DPUs should be slowed or 'derated'.  In my experience, they should be a notch faster on the throttle going uphill, at least say on a 4x2.  Maybe the difference is that my DPUs aren't for show.  I only use them on trains that would stall or stringline without them.

As a personal preference it would be unacceptable to me to modify a loco to play only a particular role like that.

[davefoxx]

I agree with @jagged ben.  While I don't have a large layout, I have messed around with a long train requiring a pusher.  In fact, here is an old video I found with two SD50s on the point of a seventy-five car train with an SD45-2 pushing.  I even documented in the video where the couplers' slack was pushed in.  In my case, with a 12-3/8" minimum radius, the train would stringline without the helper picking up some of the weight of the train.  This kind of ops is why I love mountain railroading. 

DFF

[Cajonpassfan]

I'm a little surprised to see people here suggesting that rear DPUs should be slowed or 'derated'.  In my experience, they should be a notch faster on the throttle going uphill, at least say on a 4x2.  Maybe the difference is that my DPUs aren't for show.  I only use them on trains that would stall or stringline without them.

As a personal preference it would be unacceptable to me to modify a loco to play only a particular role like that.

Umm, I certainly didn't suggest the rear end helper should be slowed; in my experience that's a recipe for disaster. On the hill, it needs to be pushing cars ahead of itself. My experience also tells me that if you have enough power capable of pushing the entire train uphill, you're going to have issues if your front end stalls. This problem diminishes as trains get longer and each set of locos takes care of its share of the train and one simply spins its wheels if the other stalls.

As to having dedicated pushers, this obviously depends on the era and railroad practices. On my transition era rr, dedicated helpers were based at the bottom of the hill, typically older locos too slow for prime time but perfectly capable to help trains over the hill. On Gary's beautiful Tehachapi BC, the practices would be different, but I would think the physics of loco/train/loco behavior on the hill would be similar in principle. Or maybe the answer is that the front end just never stalls....
Fun stuff,
Otto K.

[jagged ben]

Umm, I certainly didn't suggest the rear end helper should be slowed; in my experience that's a recipe for disaster. On the hill, it needs to be pushing cars ahead of itself. My experience also tells me that if you have enough power capable of pushing the entire train uphill, you're going to have issues if your front end stalls. This problem diminishes as trains get longer and each set of locos takes care of its share of the train and one simply spins its wheels if the other stalls.
...
Fun stuff,
Otto K.

So, both you and Ed N mentioned Mike Danneman removing tractive power from one truck.  I realize tractive power isn't exactly the same as throttle speed, but I'm struggling a bit to see the practical difference in this case.   Mind you, I don't think we're disagreeing over anything substantive.  It just seems to me that Danneman's method (which I've only heard about here) only makes sense if you're always in the same certain situation.  Like, for example, you want to model a prototype that would typically have used a pair of engines as helpers, (or you want to have a pair to reduce the chances of a single unit losing electrical contact), but your model train is never long enough to really need two units pushing on the back, which as you say can be bad. 

If Danneman was doing this in analog then I can see how maybe it makes sense.  But at the risk of repeating myself ad naseum, I feel there's a better solution for DCC: control the head end and helpers separately.   If the helpers are pushing too much, notch the throttle down.  If they're pushing too little, notch it up.  As far as I understand, it's what the prototype does.

[Scottl]

If Danneman was doing this in analog then I can see how maybe it makes sense.  But at the risk of repeating myself ad naseum, I feel there's a better solution for DCC: control the head end and helpers separately.   If the helpers are pushing too much, notch the throttle down.  If they're pushing too little, notch it up.  As far as I understand, it's what the prototype does.

I'm certain he still uses DC control.