Gradients: How steep in the model world can they be?

[The Black Hat]

Hello all,

 

I'm looking at building a layout and it is hoped to include a gradient. My question is: How steep can a gradient be in 00 gauge. I am looking to include regular steam use on load 5.

While I know that the steepest main gradient on the main railway network is 1:49 at Grosmont to Goathland and that the line at Foxfield is steeper. What is the gradients others have been able to use on a layout and how reliable has varied stock been on managing the ascent?

 

Thanks in advance.

 

 

[NIK]

5 minutes ago, The Black Hat said:

Hello all,

 

I'm looking at building a layout and it is hoped to include a gradient. My question is: How steep can a gradient be in 00 gauge. I am looking to include regular steam use on load 5.

While I know that the steepest main gradient on the main railway network is 1:49 at Grosmont to Goathland and that the line at Foxfield is steeper. What is the gradients others have been able to use on a layout and how reliable has varied stock been on managing the ascent?

 

Thanks in advance.

 

 

Hi,

 

Lickey Incline in the midlands is 1 in 37.7. Could you explain what load 5 is?.

 

As a general point curves on up gradients make things worse. I've also done some tests on loco pull and stock friction and there was huge variations in both.

 

There is the option of DCC Concepts Powerbase with small very powerful magnets attached under the chassis of locos.

Very thin steel plates hidden under the sleepers facilitate the loco gripping the track more firmly and allow better loco performance on gradients (and curves). A layout at my club has had Powerbase fitted to a combined 36" radius curve and 1 in 40 gradient just in case.

 

 

Regards

 

Nick

[Reorte]

13 minutes ago, The Black Hat said:

Hello all,

 

I'm looking at building a layout and it is hoped to include a gradient. My question is: How steep can a gradient be in 00 gauge. I am looking to include regular steam use on load 5.

While I know that the steepest main gradient on the main railway network is 1:49 at Grosmont to Goathland and that the line at Foxfield is steeper. What is the gradients others have been able to use on a layout and how reliable has varied stock been on managing the ascent?

 The Lickey Incline is steeper (1:38 IIRC), and there might well be shorter stretches elsewhere that are steeper. Lickey is on a fairly major stretch of mainline though. There used to be the Hopton Incline on the Cromford and High Peak at 1:14 - there were steeper ones on that line but they were cable-worked, Hopton ended up the steepest non-cable worked one, but that's about as far from main line as you can get.

 

When it comes to models though what's suitable for you will depend very much on what models you want to run, which will be down to the specific model itself (not sure what you mean by "load 5") and what it's pulling. Best experiment, but at a guess I doubt you want to be unrealistically steep.

[Dungrange]

I'm assuming "load 5" means with five coaches as a trailing load, but of course the answer to your question depends on how heavy these coaches are and how free running they are.  If they are fitted with lighting, then there will be wiper contacts or other pickups which will increase their resistance.

 

There is then the issue of the locomotives and how much the particular locomotives that you have will haul on the flat.  In general, diesel locomotives tend to be capable of hauling heavier loads compared to steam locomotives.  Locomotives with diecast bodies will tend to haul more than plastic bodied locomotives.  Is it going to be possible to add additional weight above the driving wheels of your preferred models to increase traction?

 

As others have said, curves also limit the load that can be pulled up a particular gradient.  1:30 may be okay on the straight, but not if that gradient includes 18" curves or point-work.  I'm not personally a fan of gradients having been put off them in my teenage years (something like a 1:24 gradient seemed to limit me to two couch trains with some locomotives), but if you really want to have gradients, then make them as gentle as possible and like the prototype, aim for something flatter than 1:30 if you can and probably flatter than 1:40. 

[34theletterbetweenB&D]

1 hour ago, Reorte said:

...When it comes to models though what's suitable for you will depend very much on what models you want to run, which will be down to the specific model itself (not sure what you mean by "load 5") and what it's pulling. Best experiment, but at a guess I doubt you want to be unrealistically steep.

Soundest advice. What others can do is a guideline, what you can achieve on your own layout the really important thing.

 

Factors that matter. Make sure both the gradients you test on, and then construct on the layout, are true gradients. By this I mean that if you don't start from a  level datum, it is easy to make a gradient vary significantly from calculated value, and people have come unstuck on this. (Did the test work assuming a level floor, built the layout sections the other way around using the same floor as datum, and when first trial assembled in a school hall with a genuinely level floor, a gradient built to be 1 in 40 was actually 1 in 36. And the data got when the concept was tested was from circa 1 in 45...)

 

Free rolling stock, no heavier than it needs to be. The standard suggested at my first club in my teens still serves me well: should roll away on a  true 1 in 100; at very worst should keep rolling once 'nudged' to start it on this gradient. Way better is possible with both RTR and kit built stock.

 

Ensure locos are weighted with centre of gravity as near centred in the powered wheelbase as possible. Ensure all carrying and tender wheelsets are free rolling and any spring loading for track holding is the minimum required. Be prepared to modify RTR locos, sometimes very significantly. Most RTR OO 4-6-0s are dire, Hornby's B12/3 demonstrates what is possible if the manufacturer does right. The G5 is likely to be a challenge, unless Bachmann have a significantly superior mechanism layout from that used in RTR 0-4-4T from other manufacturers.

 

Mild steel wheels on mild steel rail, near twice the adhesion with nickel silver.

 

Finally. Cheat. Optical illusion to make the gradient look steeper than it really is. Magnets under locos and ferromagnetic gripper strip under the track. Put drives in unpowered vehicles (Heljan's parcels car mechanism in a  coach body is a very good option, see the Stirling single going 80mph with 20 bogies on). Traction tyres/Bullfrog snot (if you really must).

[Harlequin]

For information, Peco recommend no steeper than 1 in 36:

https://peco-uk.com/pages/faqs

 

In the real world, the gradient from Exeter St Davids up to Exeter Central is 1 in 37, all on a curve and 240 yards of that in a tunnel!

 

[JohnR]

Another vote for experiment. Try out various gradients with the likely stock you are going to have. If your upper level is a branch line, no need to test it with a Pacific hauling 5/6 coaches. 

[mdvle]

As stated by others, you really need to test.

 

A couple of years ago there was a guy on here building a huge layout in Spain, and he went through the process of testing his locos and found significant variance between the models.  The only thing that could really be said was (as mentioned above) in general diesels/electrics pull much better than steam.

 

[the results he posted are long gone as he deleted the thread when he was called out on some non-train related issues]

 

Not as comprehensive, but the gentlemen doing Leeds City, Midland Side wanted to have double decked fiddle yards and actually built and tested and found the 1 in 36 gradient to much trouble (though he was looking at operating longer trains than you appear to).  He eventually abandoned the 2 level fiddle yards and instead expanded the shed...

 

 

 

Unless the gradient is absolutely necessary - to change levels - the best advice is simply not to have any given they tend to cause more trouble than really worth.

 

 

[Nearholmer]

I had a bash at laying out the physics of this in very simple terms in another thread (Nether Madder and Green Soudley), although that was met with a noted lack of popular acclaim, because it involves maths!

 

Failing maths; test and experiment.

 

Modern 00 r-t-r, at least representing steam outline, seems plagued by locos where haulage capability (tractive effort) hasn’t been thought-out properly by the designers and, unfortunately, there is no standardised test method.
 

The Railway Modeller has a creditable go at helping potential buyers, by citing haulage capability of all review locos over the same part of the same layout, but that isn’t quite as good as a standardised test.

 

An old rule of thumb used to be  1:100 for prototypical length trains; 1:50 for trains of a decent length (whatever decent means!); and, 1:25 for trains of very restricted length (a couple of coaches). But, I get the impression that a lot of modern locos can’t manage that.

 

 

[Siberian Snooper]

4 hours ago, Harlequin said:

For information, Peco recommend no steeper than 1 in 36:

https://peco-uk.com/pages/faqs

 

In the real world, the gradient from Exeter St Davids up to Exeter Central is 1 in 37, all on a curve and 240 yards of that in a tunnel!

 

 

And bankers used to shove a lot of the trains up the hill,  typically the Z class 0-8-0s, but also smaller engines,  sometimes in multiples.

 

 

[royaloak]

9 hours ago, Harlequin said:

For information, Peco recommend no steeper than 1 in 36:

https://peco-uk.com/pages/faqs

 

In the real world, the gradient from Exeter St Davids up to Exeter Central is 1 in 37, all on a curve and 240 yards of that in a tunnel!

 

Which I once went up with a crush loaded 143 with one engine idling, it wasnt pretty and I thought it was going to stop at one stage but it did get up there, as soon as I pulled away from Central the second engine decided to play again.

[DCB]

The GWR Main line from Newton Abbott to Plymouth had severe gradients around the 1 in 38 mark in both directions,  Freight was banked west from Aller Junction, both ways from Totmes and East from Plymouth.  Many passenger  trains were assisted by pilot locomotives such as Bulldogs and post WW2 by Manors.  However unassisted loads of 12 coaches were taken over these gradients by 10XX County class as well as Castles and Kings. No other UK mainline railway ran such heavy trains unassised on such steep gradients.    Steam locos were limited to 3 coaches unassisted on the  Lickey until almost the end of steam and Banking was just about universal from the GW to Exeter Central

Many model steam locos struggle with these grades.  My cast white metal County with magnadhesion manages  8 Hornby Mk 1s easily on 1 in 36 on steel track and would probably take 12 but the standard out of the Box Hornby County struggles with 3.  Weighted and with tender pickups stripped out it now manages 6 with a struggle.  I have a couple of GWR Prairie Tanks specially adapted for banking duties, capable of pushing a complete  coach train up a 1 in 36. 

[nswgr1855]

From experience assuming free rolling carriages and locomotives with no spring on the leading or trailing trucks, using 3 foot radius curves, you need to ease the gradient by multiplying your prototype gradient  by 1.75 if you want to run the maximum length prototype train length. For example a prototype grade of 1:40, you multiply 40 x 1.75 = 70. Therefore use a 1 in 70 grade on your 3 foot curve. I have found on straights or prototype curves , you can use the prototype grade with plastic bodied RTR, as long as you remove springs from leading and trailing trucks, and adjust the centre of gravity of the loco, adding or removing mass if necessary.

 

Terry Flynn. 

 

 

[Hroth]

In the real world, the Mersey Railway tunnel has a gradient from tunnel bottom to James Street station of 1:27.  EMUs still climb this, though before it was electrified in 1903 it ran with steam power, using rather thuggish Beyer Peacock tanks.

 

 

(Image source: https://www.liverpoolmuseums.org.uk/mol/collections/transport/item-269092.aspx )

 

[peterfgf]

I did some experiments and posted here about 18 months ago:

Basically, as others have said
- 1:100 (0.010 radians) will OK for most steam outline model locomotives; their haulage falls off very quickly at steeper slopes (the dreaded rubber tyres help though).   
- Modern diesel and electric outline models (central motor, drive to both bogies) will have good haulage capabilities at steep slopes.
- I would recommend 1:50 (0,020 radians) as a maximum.

Peterfgf

Edited December 10, 2019 by peterfgf

[Robert Stokes]

I have built a layout with a 4m long straight incline at 1 in 70. My express steam engines pulling six coaches slow noticeably as they go up it. I would echo the above comment about using 1 in 50 ax a maximum.

 

Robert

[Siberian Snooper]

19 minutes ago, Robert Stokes said:

I have built a layout with a 4m long straight incline at 1 in 70. My express steam engines pulling six coaches slow noticeably as they go up it. I would echo the above comment about using 1 in 50 ax a maximum.

 

Robert

 

As do full size steam locos on long banks, diesel and electric locos less so.

 

 

[Reorte]

14 hours ago, royaloak said:

Which I once went up with a crush loaded 143 with one engine idling, it wasnt pretty and I thought it was going to stop at one stage but it did get up there, as soon as I pulled away from Central the second engine decided to play again.

 

So one engine was like the lazy one on a tandem?

[peach james]

Another good thread about grades is "Dent", to be found in the Settle and Carlisle section of the forum.  It originally had a pair of spirals, then of oval spirals, and now has a pair of flat fiddle yards...might tell you something about what will work.  

 

The accidental grades on Long Marton are enough to cause problems with 10 on a Jub, so that should be a good indicator that flat is near a requirement for scale length trains.  (mind, the "accidental" grades are up to about 1:50 in places. I know the floor isn't level, but didn't take the time to get the entire layout level or the track really truly flat when I built it.  It just makes the Jubilee work hard at pulling it's train, everything else manages OK.

OTOH, my lego manages the spiral at home OK- but it requires a fair # of motors to do so.  (basically, power bogies).  To haul the 6 car Santa Fe, I found I needed 4 motors...  thats at about 1: 30 grade.   It varies from 0-2 plates/section, if you understand lego geometry.  The steepest I have ever run Lego trains on is 1 brick 1 plate/section, but that is very crimping of what will or won't run.  

(nevermind the child, he's grown a whole bunch since then...I can't lift him into there any more !)

James

[Gwiwer]

There are plenty of gradients steeper than the oft-quoted 1:37 of Lickey on the national network.   On the Thameslink core the climb from City to Blackfriars is 1:28.5 IIRC and not restricted to units either since locomotives must of necessity pass through on engineer's trains from time to time.  Others have been mentioned above.  Exeter is one of the hardest to work.  All trains begin with a standing start at St. Davids so there is no chance to take a run as there is in some locations.  Curvature, the tunnel, an often-damp rail and the line speed restriction all play their part in making it a tough start for all trains. Even some modern diesels required banking at times; it was not at all uncommon for a class 33 to request a rear-end shove and more than one class 50 has done so on their standard load 9 trains.

 

"Load 5" incidentally refers to five Mk1 carriages or the equivalent of them.  It is, if you like, shorthand which is widely understood when discussing the need for suitable power, platform length and even signal sections.  In recent years the term has been used interchangeably with Mk3 stock which is longer but metre for metre is lighter.  It is therefore not an accurate measure but a generalised description.

 

For modelling purposes please take the advice given above.  Avoid gradients if you can.  Where they must exist keep them as gentle as possible.  Remember though that the Peco-advised maximum of 1:36 means you would need 3.6 metres to rise 10cms and another 3.6 to drop back to ground level.  10cms is a bit more than OO/HO requires for scale clearances but it serves to illustrate the point that you need a lot of space.

 

It is possible to "split" the gradient if one line rises while one drops.  Each line then requires half the distance (1.8m in the example) to achieve the same clearance but the engineering and terraforming required on the layout to achieve this often means it is impractical.  

 

Curvature will increase the drag (friction of wheels against rails) and reduce the weight of train which can be hauled up any given gradient; descending a gradient beware of couplers over-riding and (if curves are involved) buffers locking as the train may attempt to overtake the locomotive.  

 

I had a range of gradients on the big layout which ranged from an unmeasured but approximately 1:40 to an afterthought branch requiring 1:20.  A lot of trains were unable to climb the branch not least the lighter and smaller branch line steam locos and any DMU with power:trailer car ratio of less than 1:1.   Nothing ever had a problem on the main line gradients other than a couple of older lightweight locos with pancake motors, traction tyres and when asked to deal with more than load 4.  The pancake motored HST set had fishing lead added as ballast and with deft driving was able to negotiate the layout in full 2+8 formation when hauling but not if the power car was pushing.  

 

Keep it as flat as you can and test, test and test again before fixing anything down and going past the point of no return.  

[royaloak]

4 hours ago, Reorte said:

 

So one engine was like the lazy one on a tandem?

Exactly.

I  was genuinly concerned about reachin the top but we got there eventually.

[DCB]

Peach James makes a good point about accidental grades.  Our flat bit through the main station is now about 1 in 100, It may be subsidence but I doubt it ever was level.  Working to accuracy of better than 1% isn't really what woodworkers do let alone modellers. Our loco depot had to be lifted about 40mm where the shed floor sagged even after the shed corner was packed up and raised.  Our house has gradients around 1 in 50 even on the concrete floor,   Add in warping if you use new rather than reclaimed seasoned timber and you carefully calculated grades soon become impassable.  In in 50 on a 1 in 100 baseline becomes 1 in 33?   You need a very accurate spirit level and check everything twice by reversing the spirit level to even out the errors if working close to your max gradient. Or you can rush it and spend the next 30 years tweaking it.   

[Denbridge]

Do as the real railways did and avoid gradients at all costs unless absolutely necessary. Most modern rtr steam locos will struggle with a reasonable load on all but the shallowest of grades.

Nothing makes a layout look toy like than unprototypical gradients and sharp curves that are more like corners.

[jonny777]

On 09/12/2019 at 18:34, Siberian Snooper said:

 

And bankers used to shove a lot of the trains up the hill,  typically the Z class 0-8-0s, but also smaller engines,  sometimes in multiples.

 

 

 

And in later years, the W 2-6-4T locos, as seen here in 1963

 

 

[Trog]

If you do have to make a gradient try and hand it so that up hill is the outside (largest radius) curve and use steel rail as the locos seem to grip that better than nickel silver.