Maximum Gradients

[Penrhos1920]

What is the maxim downward gradient that can be used on straight track? I'm trying to design a pair of return loops on a double track through fiddle yard.

[andrewnummelin]

All depends on whether you want to be able to control the train on the incline and how much "run out" you have at the bottom. And, of course, on whether the approaches are curved or not. The answers will also depend on the stock - if it's one of your modern multiple units with "continuous breaks" (motorised bogies throughout) you'll probably be OK on 1 in 5 but with 50-100 loaded coal wagons pushing a small 0-6-0 you may want something less steep (unless your stock has working breaks and you stop at the top to pin them down). If one thinks in terms of what could be found on the prototype, and the lower inertia of models, I bet that 1 in 10 should be workable, at least with short trains.
I suspect the really critical aspects will be the curves at the ends of the incline (both horizontal and vertical) and the vehicle suspensions, any bounce or wobble could give serious problems. I'd guess that a circular vertical curve at the top would be OK but I'd want transition curves at the bottom.
I look forward to seeing some wild runs on your roller coaster..... (I thought that the older spelling of break would be appropriate in this context).

More seriously, I look forward to hearing of practical experience from others as to just what is the steepest gradient that has been found to be reliable on a 4mm layout.

[grovenor-2685]

if you are only going down the most critical factor will be the gearing in your locos. Worm drives don't like acting as brakes and will oscillate fro binding on the end thrust then freeing off resulting in jerky motion. You need very good thrust bearings with minimal end play on the worm to avoid this with any significant weight of train. Second problem will be vehicles lifting under buffer compression as just discussed on another topic. All interesting challenges.
I have grades in the 4 to 5% range and am getting there gradually. Some videos on my website.
Regards

[Penrhos1920]

ANDR layout.jpg

I thought I should show what I'm thinking so far.

Red is the scenic section. There will be much more scenic but to make the layout manageable I’m just doing this bit which just happens to run parallel to the rest which is more complex.

This is the nth attempt at working out at a fiddle yard. The requirements are both up and down fiddle yards have 4 tracks for 20 - 30 wagon trains. Plus one track for 2 short passenger trains. The reversing loops are for 3 longer passenger trains shared between them.

So far I think I’ve got the down side done (in green). There’s plenty of space for the up side which I’ve started in orange. The pink reversing loop has 10’ to gain height across the top and 5’ to drop back down. It also has 16’ of level storage on the loop. If the loop is 3” above the main level then the gradient up is 1 in 40 and 1 in 20 down.

Keith in your Jinty load video is the Jinty pushing up 1 in 20 ?

[grovenor-2685]

Keith in your Jinty load video is the Jinty pushing up 1 in 20 ?

At least, I think that bit is actually a bit steeper.
Regards
Keith

[Paul Willis]

Keith in your Jinty load video is the Jinty pushing up 1 in 20 ?

At least, I think that bit is actually a bit steeper.
Regards
Keith

Keith,

This posting reminds me of a question that I've been meaning to ask you for months. It relates to your professional knowledge of the DLR.

Each working day I pass through the Westferry-Canary Wharf-Poplar triangular junction, on the Wf-CW leg as it happens. I'm always amazed about the tightness of the curve and the steepness of the climb.

Any chance that you could share with us what the minimum radius and the gradients are? That would make a decent basis for the minimum conditions for a dockside railway, or similar...

Cheers
Flymo

[grovenor-2685]

On the DLR the minimum radius is 40m on the running lines, 39m in the depot. Speed limit on a 40m turnout is 20kph.
The steepest gradient is 6%, most notably in the Bank tunnel but also elsewhere including the underpass at West India Quay. In open areas of 6% we have to heat the rails to keep the adhesion level adequate in winter.
Climbing the 6% we can push out a failed unit with a good one so it needs a minimum of one third powered axles, normally we have two thirds powered.
Regards
Keith

[Paul Townsend]

The pink reversing loop has 10’ to gain height across the top and 5’ to drop back down. It also has 16’ of level storage on the loop. If the loop is 3” above the main level then the gradient up is 1 in 40 and 1 in 20 down.

I suggest your UP at 1:40 on tight curve is going to be the limiting factor. It certainly is at Highbridge...1:40 on 4ft 6" radius.
I recommend designing for larger radius if possible and 1:50 max.
Then your down grade will be 1:25 and likely to work if you take care with the details others have already suggested.

Whatever the figures are that you adopt, rig a trial lash-up and demo it WITH some of the trains you want to use. Prove it works before committing to final version.

[Guy Rixon]

ANDR layout.jpg

The pink reversing loop has 10’ to gain height across the top and 5’ to drop back down. It also has 16’ of level storage on the loop. If the loop is 3” above the main level then the gradient up is 1 in 40 and 1 in 20 down.

You might have difficulty with the transition from flat to 5% at the ends of the down ramp, given that you're curving sharply in two planes. You might also find that given a reasonable transition-length the plane length of 5% gradient is shorter than you wanted.

Apart from getting in the way of access top the flat yard, is there a reason why the down ramp can't run down the front of the fiddle yard? If it were thus it would be longer and the transitions could be in straight track.

[dal-t]

From my many years of NG modelling (it's alright, I think I'm over it now) I'd strongly endorse what has already been said - it's the curves and transitions that are the critical factors, rather than the grades themselves. Interesting comment about the 'cogging' on descent of worm-and-pinion locos with endplay in the drivetrain, but in practical terms it matters little if the stock won't stay on the track (or make the climb) in the first place. Whatever you do, avoid coincident vertical and horizontal changes; think about vertical 'transition curves' rather than abrupt changes from level to full gradient; and wherever possible, allow a full train length of flat straight between the gradient and any curved track. As to cant, it's your call, but from my experiments (where I'd found myself with a curve on a gradient before appreciating the difficulties that creates) it only makes things worse.

[grovenor-2685]

Whatever you do, avoid coincident vertical and horizontal changes; think about vertical 'transition curves' rather than abrupt changes from level to full gradient; and wherever possible, allow a full train length of flat straight between the gradient and any curved track.

easier said than done. A full train length of straight each end of the gradient would require another 10ft or so which clearly could not be done on the OPs plan, and not in my loft either.
Our track standard on DLR does have "avoid coincident vertical and horizontal changes" as a desirable feature but it could not be a "requirement" as there is not that much space in the docklands. So there are plenty of such cases. The effect is to introduce a bit of twist in the track just like a cant gradient if you use canted curves. What you have to do is keep the twist to the level the rolling stock can tolerate. And, with our models, that tends to come down to 'suck it and see', especially if locos or wagons with rigid chassis are being used.
I have 6 sections of track with steep grades and all 6 have horizontal curves at top and bottom coincident with the vertical curve, otherwise there is no way to fit them in.

So far as the Ops plan is concerned, I agree that it would help to avoid the flyover midway along the sidings thereby aqllowing a more gentle descent. If the gridlines on the plan are at 1ft intervals that seems to make the ruling minimum radius 3ft which is also pretty tight. Seeking advice from Tim Venton will be sensible.
Regards

[Paul Willis]

On the DLR the minimum radius is 40m on the running lines, 39m in the depot. Speed limit on a 40m turnout is 20kph.
The steepest gradient is 6%, most notably in the Bank tunnel but also elsewhere including the underpass at West India Quay. In open areas of 6% we have to heat the rails to keep the adhesion level adequate in winter.
Climbing the 6% we can push out a failed unit with a good one so it needs a minimum of one third powered axles, normally we have two thirds powered.
Regards
Keith

Thanks Keith! I've always wondered about those, and now I have the answer.

I can't think in metric in those sort of terms, so having converted it to imperial and then down to 4mm, I *think* that we're talking about a minimum radius of 20.63 inches... I'm going to rattle something up in Templot using that radius and see what it looks like

The gradient seems to be 1 in 16.67, so pretty daunting but not, if I remember correctly, as steep as the Cromford and High Peak managed to run steam hauled trains up. And I don't think that they heated the rails in the winter, although the sheep would probably have appreciated it if they did.

Thanks again! Now to find some time to doodle some dockside industrial layouts...

Cheers
Flymo

[Chris Mitton]

Having converted it to imperial and then down to 4mm, I *think* that we're talking about a minimum radius of 20.63 inches... I'm going to rattle something up in Templot using that radius and see what it looks like

Now to find some time to doodle some dockside industrial layouts...

Cheers
Flymo

Given those numbers, I've a depressing feeling you're dreaming up a Narrow Gauge Workshop - or even P4 On A Coffee Table Challenge
Regards
Chris

[Paul Willis]

Having converted it to imperial and then down to 4mm, I *think* that we're talking about a minimum radius of 20.63 inches... I'm going to rattle something up in Templot using that radius and see what it looks like

Now to find some time to doodle some dockside industrial layouts...

Cheers
Flymo

Given those numbers, I've a depressing feeling you're dreaming up a Narrow Gauge Workshop - or even P4 On A Coffee Table Challenge
Regards
Chris

Hi Chris,

No, I'm thinking of either something industrial (lots of dockside wharves in the Thames area...) or that curve that I think is at Romford where there is a siding the drops down and sharply curves under the mainline. That would be as much a test of a loco and a few wagons as the Delta Junction (to give it the proper name) is on the DLR.

I can visualise the photo of that track at Romford, but I can't think for the life of me where I saw it. A search on the GERJ for Romford as a term brings up too many hits to follow up each one

Cheers
Flymo

[Noel]

Try Railway Bylines Vol 1 No 1 of December 2003 p24 - an article on the Ind Coope brewery at Romford and its rail connection. The [very sharp indeed] curve is more or less level, at the bottom of the incline down to the goods yard, with a reverse connection from the goods yard. Rail traffic eventually ceased in 1963.

Noel

[Paul Willis]

Try Railway Bylines Vol 1 No 1 of December 2003 p24 - an article on the Ind Coope brewery at Romford and its rail connection. The [very sharp indeed] curve is more or less level, at the bottom of the incline down to the goods yard, with a reverse connection from the goods yard. Rail traffic eventually ceased in 1963.

Noel

Hi Noel,

That's exactly the one that I was thinking of. Thanks for confirming it. I think that it appeared on the cover of one of the issues of GER Journal, which is why it stuck in my mind.

I don't know if it would make a good layout design or not, as one way or the other the actual mainline would be rather a view blocker. We had a good and lively discussion last night at CHEAG about P4 roundy-roundies, so this certainly doesn't fit in the description of that...

Cheers
Flymo

[philip-griffiths]

Richard , have you built the scenic section with a slope? I know a feature of the junction you are building was the gradients of the converging lines, but without realistic braking it will be difficult to control the stock. Have you considered building it on the flat?

Regards.