Camber and level changes

[Lodgie_]

I have a G scale garden layout laid on picon, a hard, black and sharp crushed volcanic rock. I'm getting random derailments and uncouplings on various stock and loco's but it's more prevalent on bogie stock The track looks pretty good but nowhere near perfect so I need to get out with a spirit level and a straight edge

A quick couple of questions...
Camber: on R4 curves how much camber can I get away with before it?
Levels: Change in level should be gradual - but how gradual, there are some "humps" in the pw that don't appear to cause problems but perfectly innocent looking changes of grade cause chaos.

A couple of "rules of thumb" would be greatly appreciated.

 

[Paul M]

Could the random problems be caused by the combination of a dip and a slant across the track?

 

[The Tinker]

Also Check your track gauge especiallyon grade changes. Due to the shape of the rail it can roll in or out changing the gauge. This is something I found, It is wise on hot days

 

[Lodgie_]

Could the random problems be caused by the combination of a dip and a slant across the track?

That Paul is highly likely, its so hard to see the defects (dodgy eyesight) so its a hands and knees job with a straight edge and level. Not looking forward to that!

Also Check your track gauge especiallyon grade changes. Due to the shape of the rail it can roll in or out changing the gauge. This is something I found, It is wise on hot days

That's an interesting point. The temperature here on the ground has been at 40c+ this last week. I have left a lot of expansion joints so no buckling but I had no idea it rolled out - better start measuring...

 

[Paul M]

That Paul is highly likely, its so hard to see the defects (dodgy eyesight) so its a hands and knees job with a straight edge and level. Not looking forward to that!

It's not too difficult, get a flat or open wagon with a small spirit level on it, and push it round. A tripod level would be best, the sort with a circle and round bubble

 

[dunnyrail]

Have shown this before but it is a very usable device, with a coupling a loco could trundle it round slowly - no getting down on your knees. Axle guards and wheels are by tenmille, but others may work or even on an LGB Flat Wagon.

 

[Rhinochugger]

I am an expert in rough track - after all, I laid all mine myself

OK, let's start at the beginning.

What wheels are you using and what wheel standards (back-to-back)? The simplest way to check the wheel back-to-back (distance between the flanges) is to use an LGB wheelset that is pretty well new and unmolested and the Mark 1 eyeball to check the alignment of the flanges on your wheelsets.

I can say with confidence that you do not not to get the spirit level out; I have quite a bit of Accucraft Fn3 rolling stock with finer wheel flanges and they negotiate my rough and cambered track (with the odd bit cambered the wrong way) without a problem.

The fact that the problem is more prevalent with bogie stock is really interesting - I would suggest that the bogie pivots are maybe screwed too tight, and not allowing enough free movement. The other thing is to check that the journals / bearings / axle boxes are well lubricated. You need the axles to be able to move freely from side to side as well as running freely.

 

[Gavin Sowry]

Camber, as you call it, is not a problem, per se. It is the 'rate of change' of camber that is the issue.
Rule of thumb for the full size stuff that I work on, is 1 to 1000, that is a 1 mm change in cant, per metre of track. Tolerance of down to 1 to 500, with an absolute of 1 to 333.

 

[Lodgie_]

I am an expert in rough track - after all, I laid all mine myself

OK, let's start at the beginning.

What wheels are you using and what wheel standards (back-to-back)? The simplest way to check the wheel back-to-back (distance between the flanges) is to use an LGB wheelset that is pretty well new and unmolested and the Mark 1 eyeball to check the alignment of the flanges on your wheelsets.

I can say with confidence that you do not not to get the spirit level out; I have quite a bit of Accucraft Fn3 rolling stock with finer wheel flanges and they negotiate my rough and cambered track (with the odd bit cambered the wrong way) without a problem.

The fact that the problem is more prevalent with bogie stock is really interesting - I would suggest that the bogie pivots are maybe screwed too tight, and not allowing enough free movement. The other thing is to check that the journals / bearings / axle boxes are well lubricated. You need the axles to be able to move freely from side to side as well as running freely.

There's a wide variety of wheelsets from various manufacturers, plastic and metal, oddly enough the cheap stuff is more stable but not perfect. The old heavy LGB bogie stock is giving the most grief - long wheel base which after Pauls's comment it could be a combination of dip and slant. None of the bogie stock is overtightened - it's time for a really thorough inspection with lift and retamping : This is going make things ache.

 

[Lodgie_]

Camber, as you call it, is not a problem, per se. It is the 'rate of change' of camber that is the issue.
Rule of thumb for the full size stuff that I work on, is 1 to 1000, that is a 1 mm change in cant, per metre of track. Tolerance of down to 1 to 500, with an absolute of 1 to 333.

Blimey! 1mm per metre. That is pretty tight.

 

[Lodgie_]

It's not too difficult, get a flat or open wagon with a small spirit level on it, and push it round. A tripod level would be best, the sort with a circle and round bubble

Ta - off to hardware shop, this could be interesting. Anything level in the Canaries is purely accidental.

 

[Rhinochugger]

There's a wide variety of wheelsets from various manufacturers, plastic and metal, oddly enough the cheap stuff is more stable but not perfect. The old heavy LGB bogie stock is giving the most grief - long wheel base which after Pauls's comment it could be a combination of dip and slant. None of the bogie stock is overtightened - it's time for a really thorough inspection with lift and retamping : This is going make things ache.

What sort of track radius are you using? - something's not adding up.

Normally, bogies would survive on rough track where 4w stuff (troublesome trucks) would come off

 

[Gavin Sowry]

Blimey! 1mm per metre. That is pretty tight.

No, that means that if you have 70mm of cant, camber, slant, crossfall, superelevation (or any other word in the English language that means the same), then the run off from the point where there is none of the aforesaid mentioned, to the point where it reaches the 70mm measurement should be 70 metres. The track radius should also change from straight to full curve radius over that same transition length. The transition takes the shape of a cubic spiral...... think that's starting to get confusing, try having to do it in feet and inches, with chains and links thrown in for good measure, the way I learned it all 49 years ago (actually, I'd mastered all that when still at high school, after a mate loaned me a copy of the PWI book British Railway Track. I have long been a member of that august organisation).

 

[Lodgie_]

What sort of track radius are you using? - something's not adding up.

Normally, bogies would survive on rough track where 4w stuff (troublesome trucks) would come off

It's all R4 and better. I've just been out there I think I've found the problem, using a glass of water and a bogie good wagon - where there's a change in level and a change in camber on the curve, it is causing the leading wheels to lift above the rail height, it's no much but probably enough. The rear set of bogies are fine but the front set on one side only are slightly above rail level. This would explain the better behaviour of the troublesome trucks.

Now to identify and rectify.

 

[JimmyB]

The problem with gradient and curve is you cannot get 4 wheels to touch the rail at the same time (all round the curve) unless the are compensated, and the longer the wheel base the worse this becomes, this is basic maths. The front outside (to the curve) wheel is constantly trying to jump the rail, this can be seem with a long flat piece of wood on your curve, it will "rock". So, either no curved gradients, or no long wheel base without compensation, or use a check rail.
I have fallen foul of this, I have 1:25 gradient that none traction type loco struggle up (wheel slip) with more than 4 axles being pulled, and I have to be careful on the curves to ensure the rail at the outside leading point is the correct relative height, and occasionally still have issues.

 

[Rhinochugger]

Yep, now my knowledge of LGB is limited but their USA archbar trucks (bogies) do have some compensation built in.

If, as you say, the leading wheel is mounting the rail, being a devout bodger, I just pack up the outer rail with ballast a bit - usually works for me.

How much is a bit? Well, until the wagon stops de-railing

 

[AlanL]

JimmyB mentions using a plank of wood and it reminded me that I've found it very useful in the past for identifying uneven track.

Derailments on long wagons is caused by twist in the track.

Cut a piece of wood cut the same length as your longest wagon and place it on the track where you are having derailments. You should have minimal rock (less than a mm?) corner to corner. Put a spirit level on top if you want to be precise but as Rhinochugger has been said true level is not essential and if your track base isn't level anywhere it will be hard work.

Doesn't need to be technical. Pack the track at the low points and run your trains,

Alan

 

[PhilP]

One other thing that can 'throw' a bogie (truck)..

If you have a joint between two pieces of track, and it forms a 'kink', on one or both rails. - You can see this by looking along the track. you see the transition, rather than a smooth connection..

This can be if 'things were tight' when you laid the track, and forced the sections apart to fit. - Disconnect the two offending pieces of track, then lift/jiggle the track slightly, about a metre-back from the bad-joint..
This will 'relax' the formation. You may-well have an overlap of track-ends, or the alignment will be 'out' between the two pieces of track..

Go back a little further, and either ease things to fit, or cut-to-fit, if too long.. - Expansion (you did say the Canaries?) can exacerbate this, as well.

PhilP.

Oh! If it is so bad, as the track gets hot, the curves lift into the air out of the ballast.. Then you definitely have a problem!

 

[Lodgie_]

When I laid it I did take great care not to force connect anything, I have some very odd length, hand fashioned connecting pieces, the shortest being 1.5"! I also used joiners and left gaps for expansion joints which appears to have worked.

 

[Lodgie_]

​

I've found a level checking solution inspired by Dunnyrail but using an old smartphone with a level checking app - mainly because it saved me a 20 mile drive to the nearest hardware shop and it actually found a use for the vile orange LGB hopper and an old phone. Note the pine packing pieces and cunningly crafted wedges. There are a few level apps available, I've tried three so far and they are all remarkably accurate - this of course depends on how good the phone sensors are (this is a venerable Sony).

I've run it around the worst affected area and discovered some nasty lateral variations which I have repacked and levelled. They didn't look that bad to the Mk1 eyeball but one of which, over the course of half a metre had a 5 degree change left to right on a rising gradient. Then the sun came out and I couldn't read the screen so that's it until the clouds return. Obviously a bit of a design flaw. Perhaps a cocktail umbrella? Only another 200ft to check...

Thank you everyone for the help, suggestions and pointing me in the right direction.