Ballasting without tears?.on thin sleeper track

[gordon s]

Gordon

 

Barry Norman recommends Febond PVA which he says is thicker than the Resin W type variety sold in the DIY sheds. Is that the type of PVA that you use and do you dilute it?

 

I need to "ballast" my goods yard and I'm using chinchilla sand because I also want the track (other than the rails) and non-track surfaces in the yard to have a similar end texture. Like you I have tried the usual dodge of laying the sand, misting it with water + detergent and then dripping diluted PVA over it. The sand is so light that it makes a run for it as soon as the water hits it and the result after the diluted PVA is added and dried is that the surface is very rough.

 

I think that after my third different attempt at covering the non-track surface with the sand that I have got that sorted but your approach to "ballasting" the track looks to be my only option with the sand unless I lift the track and put the glue down first, something I'm loathe to do.

 

I use Evo Stik Resin W wood adhesive, Ray.  I pour some into a dish, (the sort of things about 3" in diameter that hold peanuts or olives at a party) about 5-10mm deep.  To this I add the merest drop of water, just enough to thin the glue, but not so much it doesn't hold the ballast.  The mix is probably 97/3, certainly not much more.  If I were doing track in a yard area, I would use a mix of ballast and sand on the track and just sand on the surrounds.

 

For ground areas, I use a brown emulsion paint and whilst still wet, sprinkle on kiln dried sand which you can get from a builders merchant.  It's type of stuff they use for brushing into the gaps on block paving driveways.  Very fine and consistent size.  Although it is natural colour sand, I then airbrush afterwards.  This is the type of finish you get.  Just looking at the pic, the ballast may have been a mix of medium and fine from an earlier ET. I'm only using the fine variant on the current version.

 

Edited February 19, 2014 by gordon s

[PGC]

I use Evo Stik Resin W wood adhesive, Ray.  I pour some into a dish, (the sort of things about 3" in diameter that hold peanuts or olives at a party) about 5-10mm deep.  To this I add the merest drop of water, just enough to thin the glue, but not so much it doesn't hold the ballast.  The mix is probably 97/3, certainly not much more.  If I were doing track in a yard area, I would use a mix of ballast and sand on the track and just sand on the surrounds.

 

For ground areas, I use a brown emulsion paint and whilst still wet, sprinkle on kiln dried sand which you can get from a builders merchant.  It's type of stuff they use for brushing into the gaps on block paving driveways.  Very fine and consistent size.  Although it is natural colour sand, I then airbrush afterwards.  This is the type of finish you get.  Just looking at the pic, the ballast may have been a mix of medium and fine from an earlier ET.    I'm only using the fine variant on the current version.

 

IMG_8457.jpg

 

At first I thought this was a photo of real track and then Gordon tells us this is his way of modelling.Wow!!!!!!

 

Phil

[Ian Abel]

This is a BRILLIANT thread - and found it by accident as it came up in "New Content"!

 

I've been panicing about the choices/options/suggestions for ballasting and was about to commit myself   now I feel like there is a method I will try without ending up totally crazy!

[Ray H]

Gordon

 

Many thanks for the additional information.

 

One further clarification point if I may ask. The inference from your original post is that the glue fills the space between the sleepers - I'm using unmodified Peco - and to all intents and purposes dries as a lump (of glue) with a sprinkling of ballast impregnated on top rather than the ballast absorbing the glue as the glue dries/evaporates. Is this correct or if you could remove the track would you simply see "lumps" of ballast? I'm trying to get some idea of how much ballast is used (i.e. how much to sprinkle over the glue). Do you tamp it down or just let it lie on the glue until the vac comes along and removes the excess?

[fender]

great thread. thanks again Gordon.

 

I'm also interested in Ray's post above to clarify as I'm coming to ballasting soon on my own layout.

[gordon s]

Gordon

 

Many thanks for the additional information.

 

One further clarification point if I may ask. The inference from your original post is that the glue fills the space between the sleepers - I'm using unmodified Peco - and to all intents and purposes dries as a lump (of glue) with a sprinkling of ballast impregnated on top rather than the ballast absorbing the glue as the glue dries/evaporates. Is this correct or if you could remove the track would you simply see "lumps" of ballast? I'm trying to get some idea of how much ballast is used (i.e. how much to sprinkle over the glue). Do you tamp it down or just let it lie on the glue until the vac comes along and removes the excess?

 

I think I understand your question, but not sure I can give you a definitive answer.  The big difference is that I'm using SMP track and the sleeper thickness is just 0.9mm, whereas I'm guessing Peco is probably 2.0mm?

 

I have removed track before and yes, it does leave a block of ballast as it were.  The PVA does shrink down though, as you can see from my pics earlier in the thread.  It's going to be worth some experimentation, but personally I would perhaps use a slightly thinner mix maybe 95/5.  It has to remain 'sticky' but not watery as I always hoover up the excess immediately using a vacuum with a double layer of a handkerchief or similar pushed into the hose to form a pocket around 3-4" deep.  If the suction is too fierce you'll pull the ballast out of the glue or even lift the glue out.  Too soft and it won't pick up the excess.  This all sounds a bit hit and miss, but is far easier to do than write about...

 

I never tamp the ballast down as that to me would introduce other problems and break the surface of the PVA.  Just sprinkle on sufficient ballast to fill every nook and cranny.  It doesn't matter if you add too much though as you will hoover up the excess anyway.  I would not recommend you leave it to dry before hoovering as if the glue mix is too thin it will be sucked up into the ballast and form lumps through capillary action.  It was that process that caused me to rip up my favourite double junction on an earlier version of ET.  It ended up looking like this..

 

 

Try some things and let me know how you get on.  I'd experiment myself, but all my track is SMP...

[47137]

I have used a similar technique, but I applied the glue using a cocktail stick. I dipped the stick into the glue, then used it to draw a box between the sleepers. This was Peco OO9. Reloaded the stick for the next gap, and so on. This was slow but it did put down a consistent amount of glue along the track.

 

I was using a thinner mix of glue than is descibed here, more single cream than double, but it seemed to work. A thicker mix of glue would be more controllable, and I'll try this next time.

 

- Richard.

[alangdance]

Would it be possible to use a pin point applicator instead of a brush or cocktail stick. Maybe a bit more control over the glue. Any comments?

 

Alan

[Ray H]

As someone on TV might say "Dis-ast-er"!

 

And before we go any further apologies to Gordon and others for deviating from his excellent thread.

 

The gap between the sleepers is too narrow and as Gordon suspected, the sleeper thickness too deep with the consequence that even only tackling a couple of sleepers at a time the vac picked up all but the merest layer, possibly because the sand particles are so small in size and light in weight.

 

I tried using a pipette and a brush, the former being a little more successful at avoiding the sleepers (although I doubt that an even smaller brush would have been much different as it would be too small to carry the quantity of glue required).

 

There is some good news. The principle of spreading glue onto the base, patting it down, leaving it a short while and then using the vac, has provided a good result for the non trackwork area.

 

I think that I have some ballast that's finer than what I used on the running lines and heavier than the sand. I could try that.

 

Presumably there's no alternative to trying to mist the area with water prior to applying dilute PVA? Would using more washing up liquid (or IPA) when diluting the PVA avoid the need for misting?

[gordon s]

I did try using a PVA needle point applicator  http://www.finetip.co.uk  but that gave me the problem as per the pics.  You had to thin the PVA so much to get it through the needle point, capillary action took over and glued the ballast in clumps, so that's where I abandoned that method and went back to thicker PVA and a paint brush.  It works perfectly with SMP sleepers.  

 

Whilst searching around upstairs this evening I found a length of Tillig track with 2.0mm sleepers.  I'll have a play around on Friday and report back.

[47137]

I get the feeling that part of the success comes from painting the track before starting the ballasting. Paint doesn't stick to Peco track very well (I expect SMP is better for this) but paint does help seal the sub-base and make a consistent surface for the glue to stick to.

 

- Richard.

[DavidR]

Many years ago, I built a layout on the cheap using dried tea leaves for ballast (I was on a v. tight budget!). I painted all the baseboards, (which were blockboard obtained for free), in blackboard paint to seal them. Anything and everything stuck to it well(ballast, buildings, scenery, etc.) and you never had those embarrasing gaps, with natural wood showing. Sealing the surfaces beforehand is a godsend.

[gordon s]

I get the feeling that part of the success comes from painting the track before starting the ballasting. Paint doesn't stick to Peco track very well (I expect SMP is better for this) but paint does help seal the sub-base and make a consistent surface for the glue to stick to.

 

- Richard.

 

Good point, which I had taken for read.  I would always recommend that you paint the track before ballasting.  I add all the electrical connections first and then spray the track with a primer and track colour with an airbrush.  Takes just a few minutes and job done.

[Ray H]

Gordon

 

Apologies for continuing to hijack your excellent thread.

 

My track and sleepers had been painted. I think the main problem is that the chinchilla sand is such small particles and therefore quite light. There's not enough weight to resist the impact of the misting (using the diluted PVA method) and not enough body (i.e. size) to allow an accumulation to gather between sleepers (using Gordon's method).

 

I was wondering if I could get a similar effect by using plaster instead of glue (to fill the void between sleepers) and spreading the sand on top of that whilst it was still wet.

[gordon s]

Well, as promised, I experimented this morning with this short length of Tillig track with deep sleepers.  I used my same mix of PVA, with just a drop of water.  

 

Impressions?  The gaps needed a bigger fill and if every there was an advantage to UK sleeper spacing, this is it!  Certainly more gaps to fill which might make the job a little slower, but overall the result is the same at the moment.  

 

Painted on the PVA and then used the same vacuum cleaner to suck up the excess immediately.  Still throttled back the vacuum using a double layer of a cotton handkerchief to reduce the suction efficiency and prevent the glued ballast coming out.

 

So far, so good, but if I'm honest, there could be an issue with shrinkage on the PVA as it dries out.  This pic was taken immediately afterwards and shows it works, but I will take another in a few hours time to see if there are any changes.

 

It does work perfectly with thin sleeper track such as SMP.  I accept the jury may still be out with Peco or thick sleepered track.

 

 

 

 

 

[gordon s]

Sad to say the PVA has shrunk back and the results would suggest more experimentation needs to be done.  I'm sure you will be able to get good results and had I more time I'd happily continue playing around with this method.

 

I have now amended the title and added an edit to the first post to clarify what is successful and what needs further work.

[gordon s]

Just thinking about it, it might be worth experimenting with this method on Peco track, only this time allowing the PVA to rise via capillary action by not hoovering straight away.  It might be that a little capillary action would be the answer on Peco track.  If you look at my SMP track in post 31, the ballast has stuck together above the 1mm sleepers.  This may not be an issue with 2mm sleepers as the greater depth would accommodate any capillary action.  

 

Certainly food for thought…..

Edited February 21, 2014 by gordon s

[Ray H]

I tried again this evening on a micro section of track.

 

I applied the PVA and then the sand which I left for about 2 hours. I then gently brushed the excess off with make-up brush - not mine I hasten to add. I probably didn't leave it quite long enough but the early signs are that it is a vast improvement and more in line with what I was looking for.

 

I'll try again tomorrow covering a larger area and leaving it longer before attempting to remove the excess and report back.

[Ray H]

Ugh!

 

I've been a tad diverted this morning and had just decided to tackle the bigger experimental area on the layout. Alas, when I got to what I did yesterday I find that the glue has effectively solidified albeit having shrunk slightly and is visible through the thin layer of sand. I was about to give up when . . . . .

 

I recalled that I said yesterday that I thought that I might have brushed the excess off too soon and this may account for the thin layer that is now apparent. However, even allowing for that I'm not sure that there has been any capillary action but I shall now return to the layout and try the bigger area as previously indicated and see what that turns out like if I then don't disturb it until tomorrow morning.

[gordon s]

I know I said I was pushed for time, but I love a good challenge and couldn't let this one go as I know there is a solution in there somewhere...

 

Here's the first test piece which saw me vacuum up the excess immediately.  The PVA has shrunk back and there are gaps in the ballast.

 

 

Then I started thinking about capillary action and whether it was better to leave the whole thing to dry before vacuuming off the excess.  This result is a little better, but still not 100% as you still have shrinkage although the gaps are less obvious.  Perhaps a thinner glue mix will do the trick, so I will try it with a watery mix and my fine tip PVA applicator to see if that works...

 

 

Just to prove this method works, here's a pic from this morning on SMP track.  This wins on two counts for me.  UK spacing and the thin sleeper base which makes it easy to ballast.  I'll get back to you with the results of the next test...

 

[Ray H]

Gordon.

 

Many thanks for the time you're taking on this.

 

I've tried two scenarios. The first was to tamp down and then try a reduced suction from the vac. I probably got too close but any further away and the sand didn't seem to want to know. It was an improvement on yesterday but still far from perfect.

 

I've since done as I've said that I would do and left a section with the sand on the glue until the morning. One thing that did make me wonder is whether the sand is too fine & light to break the surface tension unless it is tapped down. I may find that answer out in the morning when I brush off what I just did as I haven't done anything other than sprinkle the sand over the glue (and a not excessive amount at that).

[Ray H]

 

The camera can lie! This is the result of the most recent test and the camera all but gives the impression that the "ballasted" area is reasonably level. It isn't.

 

The glue appears to have dried into blocks, some with dimples and some without with the result that the ground between the sleepers would do a good job of emulating the Sahara sand dunes in places.

 

The picture also suggests that the sand has taken consistently. Again, it hasn't. There are numerous patches where all that is visible is the dried glue; the glue hasn't grabbed any sand.

 

What is equally baffling is that the yard area away from the track where I'd spread a reasonably thick layer of Febond (thick) PVA and then covered it with sand, is showing signs of the glue not holding the sand despite it being dry for all but a week and the sand having been tamped down whilst the glue was still wet. From time to time I pass the hand-held (not that efficient) battery vac over the layout to pick up all the excesses of the sand which seem to get everywhere and oddments that shouldn't be there. And each time a little more (of what apparently isn't) glued down sand disappears from the surface exposing patches of bare card below.

 

Finally, the reasonably smooth texture that I was hoping to achieve - for I reason that other than ruts and general unevenness of the ground, the area would look reasonably smooth when viewed at a scale distance away - is looking quite coarse.

 

I fear the time has come to re-think the concept - I still wonder if a fairly runny mix of plaster would produce the desired effect - and thank Gordon for allowing me to raid his excellent series of postings and consider using a different thread.

[trisonic]

Maybe success, or not, depends on the what media the ballast really is?

 

Best, Pete.

[Ray H]

Maybe success, or not, depends on the what media the ballast really is?

 

Best, Pete.

 

It certainly appears that way from my experiments.

[nick_bastable]

thanks for this its working for me  in  2mm