Track and ballast

Ballasting the polder branch, and the two attempts that lifted

Ballast looked like the easy part of the job, which is how it came to take three goes. The version on the layout now is the third, and it is the only one that has not lifted, cracked or dried the colour of wet cement.

At the bench: 18 April 2024
Typed up: 9 June 2024
Reading: 7 minutes
Room: Amsterdam, Noord-Holland

Stop 01 of 04

Choosing a colour that is not grey

Branch-line ballast in the period we model is not the crisp grey of a main line. It is a mixture of local stone, old slag and a great deal of dirt washed down off the fields, and after twenty years of very little traffic the top of it sits closer to the colour of the soil around it than to anything that ever arrived in a wagon. I spent two evenings holding sample tubes against photographs before admitting the obvious problem: every photograph was taken in bright sun, and our attic is lit by four strips under a sloping ceiling.

The mixture we settled on is three parts of a fine buff grit, one part of something considerably darker, and a small part of ordinary sieved soil out of a plant pot, baked dry in the oven at its lowest setting. Under the attic strips it reads as a warm grey with variation in it, which is the whole point, because single-colour ballast looks like a product however neatly it is laid. I wrote the ratio on the underside of the board. The first two batches were mixed by eye and neither can ever be repeated.

Written on the card

Judge the colour under the light the layout actually has
Use at least three components; one is always too uniform
Bake anything that came out of a garden before it goes near track
Write the ratio somewhere attached to the layout itself

Stop 02 of 04

Laying it dry, and then leaving it alone

Dry laying is the slow half and it decides everything that follows. Ballast goes on in small heaps along the sleepers, a soft brush moves it out into the shoulders, and a fingernail tapped on the rail settles the grains down between the sleeper ends. The part I kept getting wrong was the height. Ballast brought up level with the sleeper tops looks like gravel tipped on a plank, and on a line like ours, half-buried and barely maintained, the sleepers should stand proud while the shoulders slump away, because nobody has been along with a tamper since the sixties.

Points get less of everything. We keep ballast away from the tie bar completely, and the area under the blades is brushed clean twice before any glue appears, which takes roughly as long as all the plain track either side of it. Two of our earliest points were ballasted properly, in the sense that the ballast went absolutely everywhere, and both have since been lifted and replaced. A point that half-throws on a damp evening will spoil an operating session more reliably than anything else on this layout.

Stop 03 of 04

Glue, water and a week of vinegar

The wet stage is where the first two attempts failed. Dilute white glue run in with a pipette after a wetting agent is the standard method and it does work, but in a cold attic in February the water sits on the surface instead of sinking, and when it finally does sink it carries the ballast with it in small rivers. Our second attempt cracked along the board edges a fortnight later as the timber took up damp and moved. The third waited for a dry week in May, hatch open, a fan heater running at the far end of the room.

The mixture is one part white glue to four parts water with a few drops of washing-up liquid stirred in gently enough not to foam. Wet the ballast first with the same water and detergent, then feed the glue in at the sleeper ends and let capillary action carry it sideways. It smells faintly of vinegar for three or four days; the attic smelled of it for a week and somebody carried it home in a coat. Do not try to hurry the drying with heat placed directly over the track, which is how our first shoulders crazed.

Written on the card

Wet first, glue second, never the other way round
One part glue to four parts water, fed in at the sleeper ends
Work in a dry week rather than a cold damp one
Let it dry slowly; direct heat crazes the surface

Stop 04 of 04

What the ballast did to the running

Ballast changes how a layout sounds, which nobody warns you about. Bare cork under our track gave a dull thump; with ballast glued hard onto it every wagon is louder and a little sharper, and the terminus throat now has a rattle that tells you which road a train has taken without anyone looking up. Some modellers put a layer of foam between cork and board specifically to kill that. We rather like it, and one of the things a slow shunting move gains from is a sound that alters when the wheels cross a joint.

It also found faults. Glue that has crept up the inside face of a rail will stop a short-wheelbase shunter dead at the speeds we like, and two of those took most of a season to find because each only showed itself in one direction of travel. A scalpel blade run along the rail web, followed by a hardwood block along the railhead, cleared both. The rule at the benches since then is that nothing gets ballasted until it has been run over slowly, both ways, by the worst-picking-up locomotive we own.