Eastwood Town - A tribute to Gordon's modelling.

[Pete Harvey]

Gordon

 

What have you done with the old station & walling?

 

Will you be using them again.

 

Pete

[gordon s]

Hi Pete, yes I will still be using the station buildings, but will need to make new platforms as the curvature of the platform roads will change.  That's no big deal and I'm happy to do that.  I'm working through all various levels and trackbed widths right now to ascertain where embankments and retaining walls will placed and will hope to cover that in more detail as the build progresses.

 

I still have mixed feelings about which product to use on retaining walls.  Even though the Slaters 7mm dressed stone is oversize, there is something about it that looks substantial and really solid.  The Wills stone product is much nearer to scale, but somehow reminds me of ornamental walling that you might see in a garden, rather than in a major civil engineering project.  There are other aspects to consider though.  The Slaters sheet is larger in length so can be used in longer runs without joins, but brings it's own problems when forming edges around support pillars.

 

The Wills sheet is easier to use on pillars as it is a moulded product rather than vacuum formed, but does need joins to form larger panels.

 

The third alternative is to use brick paper, but you lose the three dimensional effect of embossed plastic.  I think brick paper looks fine in close up shots, but as bricks are so small in 4mm scale, as soon as you step back, you lose that definition.  This is one of those things that suffers from scaling.  Your eyesight can pick out individual bricks in a real wall from some distance, but once you reduce to 1/76 of the size, it is much harder to see the individual bricks from a similar scaled distance.

 

To my mind that is where moulded sheets come into their own, particularly the Slaters 7mm dressed stone.  

 

I'm sure I'll revisit the subject many times in the next few months, but right now I want to push on with cutting and laying the trackbed.  Hopefully in the next few days I'll be able to complete this first corner and then work my way around. 

[coachmann]

Your previous Stone walls looked just right due to the way you coloured them (sooty grey finish). The 7mm scale stones didn't seem to matter. In addition, the straight mortar courses were correct for many retaining walls. Your work on walls influenced me into trying Slaters out on the retaining walls at Greenfield but I think mine failed due to being lower in height than yours.

[hayfield]

Gordon

 

I agree with Coachman, your previous builds look terrific. I think to somewhat the scenic section rather than being absolute scale should be a bit like the theatre scenery where its slightly overstated on the painting front. mostly the viewing is from a distance so details have to be emphasized a bit.  

 

In other words more of the same please

 

By the way I need a new shaft for my driver, do I get it repaired or wait for the January sales ?

[cheesysmith]

As to the stone walls, stick to what works, namely the 7mm sized ones. I have to agree that the 4mm sized ones look too small for the big stone built retaining walls.

[gordon s]

Hi Guys, I haven't forgotten you, just struggling again with gradient construction.  A straightforward up and down is no problem at all, but when you have a 40-50' run of 1:100 gradient all sorts of things come into play.  In the old days, I would set the baseboard level and then use risers working from a datum level of the baseboard frame.  It was quite easy to cut risers at various lengths, but when they were all in place, running a straight edge down the track bed showed it was a series of rise and falls over the risers themselves.  I tried everything possible, but once you had to move boards about across a room, the levels would change.  Of course we're only talking a few mm here and there, but nevertheless the holy grail of a flat bed was hard to achieve.

 

Using a spirit level is a very simple answer but that doesn't work either as you can create compound errors over a 60' run. Starting from one point around a continuous loop of 64' it is possible to have everything level as you work around the room, only to find as you get back to the beginning you have a rise or fall of 10-20mm between the trackbed ends and then have to fudge the join with an incline or decline.

 

Now technology has come into play and I've tried using a laser level and instead of constructing the gradient from the baseboard frame up to the trackbed, I thought I'd try reversing the process and establishing a flat trackbed and then working down from that datum using a laser level set up on a tripod.  That was fine until I realised just moving around the room could impact the laser itself.  This is a purpose build room and you would think with five huge RSJ's it would be solid as a rock, but not so.  I weigh around 13.5 stone and I can see the deflection on the laser as I move about.

 

The problem with any multilayer design is that you have various levels from the same trackbed as it winds around itself crossing over and under the same gradient over it's 50' run. You have to have a minimum of 73mm clearance to allow a loco, cork trackbed and the height of rail between two levels.  I have the space to do this and have the engineering background to do it, but somehow can't find a simple way to achieve it.  I will get there and once I get one section done it will all be plain sailing.

 

The key challenges for me are the gradient should be 1:100 or greater.  Box ticked.

 

Minimum clearance to be 73mm.  Box ticked.

 

The trackbed should be in one plane and not a series of up and downs as it crosses risers along the slope.  Getting there.

 

Starting at one point you must be able to run 60' plus around the room on a flat and level trackbed and find when you get back to the beginning the ends of the boards match up perfectly.  Nearly there....

 

I have every confidence I will get there as there will be a lightbulb moment very soon, but I've now had five attempts at getting it right over 25% of the layout and am still falling short of my goals.

 

Right now, I'm surrounded by drawings and spreadsheets show the various dimensions required to achieve the holy grail. I'm off to golf tomorrow and the break will do me good, but if anyone has any ideas how to meet these goals, then I'm all ears.

 

Simple up and down theories work fine on straight or short gradients.  This one is proving far more difficult with continual curves, the length of gradient and the shallow angle of 1:100.

Edited November 11, 2015 by gordon s

[Jock67B]

Gordon,

I've just looked back and so I must side with the lads above, the 7mm stones looked absolutely superb, thanks in no small measure to your excellent weathering! Good luck with the gradient, it will be very interesting to see how you solve it!

Kind regards,

Jock.

Edited November 11, 2015 by Jock67B

[martin_wynne]

Simple up and down theories work fine on straight or short gradients.  This one is proving far more difficult with continual curves, the length of gradient and the shallow angle of 1:100.

 

Hi Gordon,

 

When we were building Adavoyle Junction (30 years ago now, difficult to believe), we found a local firm of ecclesiastical joiners who said they were happy to cut 8ft strips of 1/8" hardboard to a taper of 1:96, i.e 1" wider at one end than the other. In the end they were happy to do any taper if we specified the end widths over 8ft. The results were extremely accurate with two straight edges. We assembled them side by side with spacer blocks and attached the roadbed along the top edges. There was a bit of a problem with drumming noise, which we cured by drilling and injecting cavity foam. Doing it again blocks of expanded foam could be included during construction. 

 

A bit of calculation is needed to allow for the difference in radius/length of the two sides, and in the event we introduced a little cant (superelevation) at the same time. We were a bit cautious there and always wished we had added a bit more -- it made a significant effect to trains running round the transition curves as they leaned into the curve. It's important not to overdo it because track twists can cause derailments and there is already a track twist in any helix.

 

Here are a few pics from 1985, sorry about the quality. The cant angle is just visible in the board-joint formers in the first pic.

 

 

 

 

regards,

 

Martin.

Edited November 11, 2015 by martin_wynne

[chris p bacon]

Interesting to see you buy Whisky by the boxfull.

[trisonic]

You’re right, Martin.

Your layout looks somewhat like the inside of an acoustic instrument, like a guitar. Whilst foam works to a degree its use in anechoic chambers (for example) cuts reverberation to zero - somewhat surprisingly most anechoic chambers are not soundproof. You can hear what is going on inside, outside, although it is heavily attenuated.  It is: “Mass that kills sound”.

 

If it is still annoyingly noisy Gordon  I might be able to come up with a few ideas.

 

Best, Pete.

Edited November 12, 2015 by trisonic

[Ian_H]

Hi Guys, I haven't forgotten you, just struggling again with gradient construction.  A straightforward up and down is no problem at all, but when you have a 40-50' run of 1:100 gradient all sorts of things come into play.  In the old days, I would set the baseboard level and then use risers working from a datum level of the baseboard frame.  It was quite easy to cut risers at various lengths, but when they were all in place, running a straight edge down the track bed showed it was a series of rise and falls over the risers themselves.  I tried everything possible, but once you had to move boards about across a room, the levels would change.  Of course we're only talking a few mm here and there, but nevertheless the holy grail of a flat bed was hard to achieve.

 

Using a spirit level is a very simple answer but that doesn't work either as you can create compound errors over a 60' run. Starting from one point around a continuous loop of 64' it is possible to have everything level as you work around the room, only to find as you get back to the beginning you have a rise or fall of 10-20mm between the trackbed ends and then have to fudge the join with an incline or decline.

 

Now technology has come into play and I've tried using a laser level and instead of constructing the gradient from the baseboard frame up to the trackbed, I thought I'd try reversing the process and establishing a flat trackbed and then working down from that datum using a laser level set up on a tripod.  That was fine until I realised just moving around the room could impact the laser itself.  This is a purpose build room and you would think with five huge RSJ's it would be solid as a rock, but not so.  I weigh around 13.5 stone and I can see the deflection on the laser as I move about.

 

The problem with any multilayer design is that you have various levels from the same trackbed as it winds around itself crossing over and under the same gradient over it's 50' run. You have to have a minimum of 73mm clearance to allow a loco, cork trackbed and the height of rail between two levels.  I have the space to do this and have the engineering background to do it, but somehow can't find a simple way to achieve it.  I will get there and once I get one section done it will all be plain sailing.

 

The key challenges for me are the gradient should be 1:100 or greater.  Box ticked.

 

Minimum clearance to be 73mm.  Box ticked.

 

The trackbed should be in one plane and not a series of up and downs as it crosses risers along the slope.  Getting there.

 

Starting at one point you must be able to run 60' plus around the room on a flat and level trackbed and find when you get back to the beginning the ends of the boards match up perfectly.  Nearly there....

 

I have every confidence I will get there as there will be a lightbulb moment very soon, but I've now had five attempts at getting it right over 25% of the layout and am still falling short of my goals.

 

Right now, I'm surrounded by drawings and spreadsheets show the various dimensions required to achieve the holy grail. I'm off to golf tomorrow and the break will do me good, but if anyone has any ideas how to meet these goals, then I'm all ears.

 

Simple up and down theories work fine on straight or short gradients.  This one is proving far more difficult with continual curves, the length of gradient and the shallow angle of 1:100.

Hi Gordon,

 

I don’t know if it’s a stupid idea but how about cutting lengths of 2.4m ply at the required gradient and fitting them under the rail bed, if 1:100, then the ply would be zero (datum) at one end and 24mm greater at the 2.4m mark. Next piece of ply would be 24mm at one end and 48mm at the other end. On straight sections then both sides of the rail bed would be identical but on a curve it will get a little tricky. Perhaps okay on simple curves, you will have the circumference/length on the inside and outside of the curve but on transitions I’m not too sure, may have to think on that a bit.

 

Edit - Opps, You could just follow Martins lead which had it all planned out thirty years ago ... also, to Martin,why are you drinking Scotch while building an Irish railway?;-)

Edited November 11, 2015 by Ian_H

[gordon s]

Thanks for all your suggestions once again.  I sat down for a total break, consumed a couple of glasses of red over dinner and then had another go.  

 

First error I found was a schoolboy one in as much I'd missed one board out completely in the levelling process and couldn't fathom out how I was left with an incorrect measurement from the baseboard frame to the underside of the trackbed at one point even though the laser dimension downwards was correct.  Once I had levelled this board it all made sense and the results I'm getting are now in line with my spreadsheet calculations.

 

I also found that moving the laser tripod away from where I was working stopped the laser bounce/deflection as I was moving around the room.  Both of these are fundamental errors, but somehow when you get very familiar with any job, you are always looking for the hidden answer and miss something bloomin' obvious in front of you.  After two days of continuous work I can now go to bed a happy man....

 

I guess the moral of the story is never do something complex sober.....

 

So it's golf tomorrow, which will be a welcome break and then back into ET over the weekend.

[gordon s]

Hi Gordon,

 

I don’t know if it’s a stupid idea but how about cutting lengths of 2.4m ply at the required gradient and fitting them under the rail bed, if 1:100, then the ply would be zero (datum) at one end and 24mm greater at the 2.4m mark. Next piece of ply would be 24mm at one end and 48mm at the other end. On straight sections then both sides of the rail bed would be identical but on a curve it will get a little tricky. Perhaps okay on simple curves, you will have the circumference/length on the inside and outside of the curve but on transitions I’m not too sure, may have to think on that a bit.

 

Ian

 

I love those pictures of Adavolye, Martin.  Some superb carpentry.  Hopefully I get some way towards that standard of workmanship.

 

Thanks Ian.  A great idea, but part of the problem is that ET track layout is always on the curve.  I have a hang up about perfectly straight track, so every piece has a curve to it, even if it's 30-50' radius.  Laying perfectly straight track is always difficult and it's straightness simply highlights every error along the way, so I always add some curvature to every long length of track where possible.

 

I have a metre flexible rule which I use to form along a curved trackbed to test if it stays on plane, but somehow I get the impression that the 'plane' changes as you form the rule into a curve and the outcome is not actually correct.  I'm not a maths expert, but no doubt someone will be able to confirm whether or not that is true.  

 

I recall a discussion in this thread a while back about how gradients changed as the radius reduced.  If you can imagine a flat piece of plywood with three parallel curved tracks on it and tipped it sideways but maintained a gradient of 1:100, then the inner curve will have a gradient greater than 1:100 as the length of track over the same gradient will be less that those of the outer curve for the same vertical height gain.

 

At least I feel a bit better tonight, so will push on over the weekend.  Funny how you struggle for days, take a break to post a question on here and then, if by some miracle, all becomes clear.

[islandbridgejct]

... struggling again with gradient construction.  A straightforward up and down is no problem at all, but when you have a 40-50' run of 1:100 gradient all sorts of things come into play. ...  Of course we're only talking a few mm here and there, but nevertheless the holy grail of a flat bed was hard to achieve.

 

Using a spirit level is a very simple answer but that doesn't work either as you can create compound errors over a 60' run. Starting from one point around a continuous loop of 64' it is possible to have everything level as you work around the room, only to find as you get back to the beginning you have a rise or fall of 10-20mm between the trackbed ends and then have to fudge the join with an incline or decline.

 

...

 

The key challenges for me are the gradient should be 1:100 or greater.  Box ticked.

 

Minimum clearance to be 73mm.  Box ticked.

 

The trackbed should be in one plane and not a series of up and downs as it crosses risers along the slope.  Getting there.

 

Starting at one point you must be able to run 60' plus around the room on a flat and level trackbed and find when you get back to the beginning the ends of the boards match up perfectly.  Nearly there....

 

I have every confidence I will get there as there will be a lightbulb moment very soon, but I've now had five attempts at getting it right over 25% of the layout and am still falling short of my goals.

 

Right now, I'm surrounded by drawings and spreadsheets show the various dimensions required to achieve the holy grail. I'm off to golf tomorrow and the break will do me good, but if anyone has any ideas how to meet these goals, then I'm all ears.

 

Simple up and down theories work fine on straight or short gradients.  This one is proving far more difficult with continual curves, the length of gradient and the shallow angle of 1:100.

 

Hi Gordon,

 

64' = 19392mm (approx! ie 303 x 64)

 

So in a run of 19392mm you have a variation in gradient of 20mm max.

 

20mm variation in 19392mm = 1:969.6

 

Effectively, your variation is a gradient of 1:1000, transforming your 1:100 into a 1:100.1 or 1:99.9 depending, so I would suggest (armchair thinking here!) that it is practically insignificant and of no concern. You are probably engineering your baseboards to a standard the prototype does not meet.

 

Whilst perfection is nice to aim for, it may not always be desirable, and I fear you might be in danger of putting yourself off again. Somebody more practical than I am may be able to confirm that, at this level of accuracy, you really have nothing to worry about, and should just make sure there are no major deviations within your 19,392mm of main line.

 

If you stand in Sallins station looking towards Cork, the line is dead straight for a considerable distance. An approaching train actually disappears at one point into a hollow before reappearing much closer to you. It's quite something to watch, but that's Irish bogs for you. If this sort of roller coaster bothers you, slap a speed restriction on it, but it is not unprototypical, and is unlikely to affect performance significantly.

 

(Somebody tell me if I've got my maths wrong on this.)

 

Alan

[gordon s]

Hi Alan, thanks for you contribution.  Perhaps I've not made myself clear.  The main loops are double track and they run around the room at the same level at a height of 140mm above the board frames.  The run around the room is about 60' which as you say is 18288mm.  The problem is that using a spirit level is that you can set up one metre of track on the risers and doing by eye could be just 1mm out of level.  As you progress around the room, you then compound the error by just 1mm each time and when you return to the start you suddenly find the end of the run is 18mm lower or higher than the start of the run.

 

I think I have now overcome this using a laser level on a tripod, but even then you have to be careful how you move around the room as the beam can move as your weight moves across the floor.  I've found now the way to deal with this is to set up the laser some way away from the area you are working on and then wait for the laser to stabilise and find a true level before fixing anything in place.

 

The gradient issue was how to effectively measure the rise and fall from this datum position 140mm above the baseboard frame and then fix the risers into place so that the run is flat across the whole length albeit on a gradient on 1:100.  Again I feel I cracked this aspect last night.  I put together a spreadsheet showing the drop at each foot of run and that is measured from the underside of the trackbed set at 140mm.  Where I was struggling was that if the drop was 50mm that should then leave you a distance of 90mm to the frame. The problem I had was that I hadn't set up one of the boards to it's own level and so where I was expecting to see 90mm left, the measurement was actually over 100mm as the frame itself was out of true level by 10mm.  Try as hard as I could, I couldn't fathom out what was wrong until the penny dropped yesterday. Now that I've picked that one up, everything is OK.

 

An even bigger problem mentally is where the track on a descending gradient loops around and goes under itself.  Now you have to deduct and additional trackbed thickness of 12mm and still maintain a minimum clearance of 73mm to allow a cork trackbed, the height of the rail plus the minimum I have set myself of 70mm.

 

All of those factors made life very difficult and the only way I could resolve it was to raise the double loops from their original datum of 110mm to 140mm.  Of course this is further compounded by the rise up to ET terminus which you have yet to see.  I have the whole plan here and ET terminus itself brought another set of challenges where the rising track to the terminus has to go under the terminus itself.  That involved raising the terminus to get over the minimum clearance dimensions which in turn lengthened the gradient to stay at 1:100.  Having seen steam loco's plus 7/8 coaches struggle up long gradients on the curve before, that is one compromise I wasn't prepared to make, hence the length of time to get a workable plan.

 

Tonight has seen some more work underway and I'm hopeful to have half of the trackbed down over the weekend.  Once it's all down and I see the gradients are fine then track will go down and something may run before Christmas!......,

Edited November 12, 2015 by gordon s

[BoD]

Have you ever thought about making Eastwood a 6x4 oval?

[cheesysmith]

Come on Gordon, you don't have to prove your superiority over the rest of us mere mortals when it comes to the woodwork. We know you are better than us, so stop trying for perfection, when you get it perfect it is just showing off

[skipepsi]

It is his layout and he will make it how he wants to, or a song something like that.

[gordon s]

Have you ever thought about making Eastwood a 6x4 oval?

 

It is.  6m x 4m.....(give or take a bit).......

[cheesysmith]

you forgot to add eventually

[islandbridgejct]

Sorry, Gordon. I thought you were concerned about the amount of variation in height, rather than the difficulty of making the two ends join up. I'd probably use a piece of string and a small spirit level stretched from one side of the room to various other points to establish a base height at different points, but only because I wouldn't know how to go about it mathematically or logically, and it would no doubt be considerably less accurate. Please keep on explaining - I'm storing up knowledge from threads like yours for the day I can get building again.

 

Alan

[gordon s]

Morning all....building a model railway seems so trivial this morning after the events in Paris last night.  Truly dreadful and my thoughts go out to all the families impacted  by this tragedy.

 

I have made some progress during the week and here's the latest state of play.  Having finally got to grips with gradients and the building of the slopy bits, I now feel relatively comfortable about pushing on. I'm taking one section at a time and printing off the trackbed templates and pasting them onto 12mm ply sheets to cut out the trackbed with a jigsaw.  The sharp eyed amongst you may see that one template has no track on it.  No, I haven't lost my marbles, it's on the other side as that was the only way I could get these two templates out of one sheet.

 

This specific template is the one with the two turnouts that will connect to the reverse loops you can see at the lower levels.  The whole template was too big for one sheet, so I have cut the template in two and will then join the two pieces after routing out the edges to take biscuit joiners.  I've deliberately used the cut edges to form the joins to ensure they will align properly once they are set up correctly.  

 

Having now set up the first few boards, I am having some thoughts about the overall height of the layout.  Originally the height of the base level boards was dictated by clearance of a multi drawer plan cabinet that is 945mm high.  The baseboard frame adds another 95mm plus 5mm clearance taking it up to 1045mm.  Once I've added another 324mm to take it up to ET terminus level, I have a track bed of 1366mm or just under 4'6" in old money.

 

Now I'm 6'3" so it's a great height for me, but I am wondering if it is going to be too high overall.  I now realise there is space to put the cabinet under ET terminus where there is much greater clearance, so could drop the whole layout around 4" to sit onto the stair balustrade.   None of this is an issue as I can easily cut 4" off the legs as I am working on each board.  Right now, I'm not going to change anything, but will review it as construction progresses.

 

I quite like the idea of boards at 4"6" as that will put trains at eye level and change the whole perspective overall.  I may have to provide boxes for operators who are vertically challenged, but at least we can have a good laugh about it.  Maybe I'll have to have one of those signs you see at theme park rides with qualifying height limits....:-)

 

 

 

 

 

 

 

 

[NWJ]

 

... Having now set up the first few boards, I am having some thoughts about the overall height of the layout.  Originally the height of the base level boards was dictated by clearance of a multi drawer plan cabinet that is 945mm high.  The baseboard frame adds another 95mm plus 5mm clearance taking it up to 1045mm.  Once I've added another 324mm to take it up to ET terminus level, I have a track bed of 1366mm or just under 4'6" in old money.

 

Now I'm 6'3" so it's a great height for me, but I am wondering if it is going to be too high overall.  ,,,

 

Gordon

 

I am also 6'3" and am currently building a layout where the main trackbed will be 1310mm off the ground. It is a great height to work at, easy on the back! However, it does have issues with reach, at this sort of operating height, I would not be able reach as far over the board as I could on something lower, but as the boards are only 600mm wide this is not a problem. What I am pleased with it, when operating while sitting on a stool, everything is at eye level. and I get a more realistic view than the usual helicopter view I often have when looking at many lower layouts.

 

Neil

Edited November 14, 2015 by NWJ

[colin penfold]

Hi Gordon,

 

When you talk about quality handbuilt track I have nothing useful to contribute as it is way beyond my capability, but I can contribute to your thinking about baseboard heights. My boards are on two levels, set at standing eye level (58") and sitting eye level (44") at between 2'6 and 3' baseboard width.

 

Tests before I built the boards showed that I can comfortably operate the layout at these heights, including manual uncoupling. What I can't comfortably do is work on the tracklaying and wiring etc so it's a boon for me that I can remove boards one or two at a time and work with them at a better level. I would suggest this for you for any complicated track areas.

 

Just thought it was worth mentioning this - hope it helps

[gordon s]

Thanks Neil.  The good thing about this plan is that it is only ET terminus that is at that height and as it runs right down the centre of the room it is accessible from both sides.  Whilst all the operational stuff is on one side of the room, having the boards that high also means it will be relatively easy to duck under if need be.  

 

I had always steered away from building a board down the middle of the room, but once I realised it would solve the design failings of earlier plans I sat down and considered how it could be achieved with the minimum need to go under the access the other side.  With two operators, the option is there to have one each side, but even operating on your own will be straightforward as ET terminus is only 700mm wide in the station area and the shed is all easily accessible.

 

The majority of all the running lines on the other side is plain track with just two turnouts to access the return loop and a crossover to take trains from the main lines into the goods access loop.

 

If I can construct flat baseboards and make the 00-SF pointwork correctly, then all should be fine.  Fingers crossed...:-)