An Anniversary Parallel Boiler Royal Scot with CSB
- Thread starter andrewb
- Start date
andrewb
Western Thunderer
This was a fun fret to create. It aimed primarily to give me the ability to mount the springs in a way that allowed removal, plus catering for the important matter of mounting the double coil springs under the front driving axle. I found that the kits original motion brackets were the wrong shape so redid them too, with a correctly rounded profile and re-jigged to take the more protypically accurate Ragstone slide bars. Then, being a bit nerdy about such things, decided to design a kind of ‘cradle’ to mount the DCC chip above the motor, taking advantage of this types cavernous fire box. It’s the funny shape almost exactly half way down the right hand side.
The expansion link in the original kit is frankly a little disappointing when compared with contemporaneous photos, so I worked those up too. In theory I should be able to have working reversing gear... in theory!!
andrewb
Western Thunderer
By designing my own horn guides I was able to significantly simplify how I would actually mount the aforementioned coil springs: the stays shown in this fret are the obvious solution. Certainly better than the funny little space invader-shaped brackets I’d explored in the preceding etch...!
andrewb
Western Thunderer
... and of all the little ancillary bits and bobs that needed a tweak by widening the frames, these little brackets are the stays for the sanding pipes.
I’m not selling CSB very effectively if I appear to be implying that you need to pretty much rework the entire kit - please don’t forget that most of this work is solely to do with changing the width of the frame spacers. The design and etching of the the CSB-specific items took no time at all! Honest!
Last edited:
A (Very) Brief Exposition on CSB
andrewb
Western Thunderer
So - onto the business of fitting Continuous Springy Beam (CSB) to a Parallel Boiler Royal Scot in 7mm scale. But before start, we ought to run over a few basics. That's all I'm qualified to do, being neither an engineer nor teacher, but it could be useful to explain a little of how I approach the subject. For in-depth detail you can't go wrong on the CLAG site at Continuous springy beams
Some useful definitions/explanations (in Andrew-speak!):
CSB: A picture paints a thousand words...
The 'beam' itself would typically be made from steel guitar string or, more likely for the size and mass of larger 7mm models, thin piano or harpsichord wire. Steel would normally be used due to the predictability of its elastic modulus (whatever that means!) Frame fulcrum points can be provided in a number of ways - either designed into the frame from the outset or, more commonly, short handrail knobs from suppliers such as Markits are used. It might be useful to know that their short knobs measure 2mm from the hole centreline to the 'shoulder', meaning a consistent distance of CSB from the frame can be maintained. It's quite important to keep the CSB as straight as possible, longitudinally. Whatever is done to provide the fulcrum point on the axle blocks needs to account for this. In theory, its possible to simply lay the CSB over the top of the axles or a suitable rounded collar on the axle blocks, but this will limit exactly where vertically the CSB must run. Most modellers create some kind of fixing which extends vertically from the block by a given distance - that's what I've done for this trial (see my penultimate photo above). It's worth noting that the CSB could run under the block too, and this can often be useful in keeping it out of sight on more 'open top' frames.
Deflection: As the diagram above shows, once the mass of the loco (NB not the unsprung mass, so don't include the wheels, axles, gearbox and motor) is resting on its wheels, the CSB will deflect between each frame fulcrum point. In the 4mm world, a deflection of 0.5mm has been settled on. I've not been able to find an equivalent convention for 7mm models. A balance needs to be struck: there's no absolute right or wrong figure and, in theory, one could have the static deflection at any figure. The likely masses of 7mm models, the capacity of the handrail knob with regard to wire diameter, the limits of vertical movement of the axle blocks within the guides and the desirability of having suspension that is neither too soft or too hard suggest to me that a static deflection of 1.5mm is probably about right. This will allow me to use wire from 22-28 thou to maintain correct ride height from 850g to 2225g. As I anticipate adding additional mass to the model for traction purposes, I expect to use wires at the thicker end of the spectrum so it shouldn't end up too sloppy. To be honest, I just don't know, so this'll be my chance to experiment.
Why is deciding on a static deflection important? I have created fixings for the axle blocks that measure 7.5mm from axle c/l to its CSB mounting hole c/l. If I want to use a 1.5mm deflection I need to mount my frame fulcrums at 6mm above axle c/l (7.5-1.5mm). It's worth noting that if you are using CSB below the axle blocks, you'd add the deflection.
It's also important to note that we are not trying to mimic prototype spring deflection limits - static deflection is purely what puts the loco at the correct ride height. The actual amount by which the CSB can act as suspension is dependent on the physical limits of axle block movement.
Some useful definitions/explanations (in Andrew-speak!):
CSB: A picture paints a thousand words...
The 'beam' itself would typically be made from steel guitar string or, more likely for the size and mass of larger 7mm models, thin piano or harpsichord wire. Steel would normally be used due to the predictability of its elastic modulus (whatever that means!) Frame fulcrum points can be provided in a number of ways - either designed into the frame from the outset or, more commonly, short handrail knobs from suppliers such as Markits are used. It might be useful to know that their short knobs measure 2mm from the hole centreline to the 'shoulder', meaning a consistent distance of CSB from the frame can be maintained. It's quite important to keep the CSB as straight as possible, longitudinally. Whatever is done to provide the fulcrum point on the axle blocks needs to account for this. In theory, its possible to simply lay the CSB over the top of the axles or a suitable rounded collar on the axle blocks, but this will limit exactly where vertically the CSB must run. Most modellers create some kind of fixing which extends vertically from the block by a given distance - that's what I've done for this trial (see my penultimate photo above). It's worth noting that the CSB could run under the block too, and this can often be useful in keeping it out of sight on more 'open top' frames.
Deflection: As the diagram above shows, once the mass of the loco (NB not the unsprung mass, so don't include the wheels, axles, gearbox and motor) is resting on its wheels, the CSB will deflect between each frame fulcrum point. In the 4mm world, a deflection of 0.5mm has been settled on. I've not been able to find an equivalent convention for 7mm models. A balance needs to be struck: there's no absolute right or wrong figure and, in theory, one could have the static deflection at any figure. The likely masses of 7mm models, the capacity of the handrail knob with regard to wire diameter, the limits of vertical movement of the axle blocks within the guides and the desirability of having suspension that is neither too soft or too hard suggest to me that a static deflection of 1.5mm is probably about right. This will allow me to use wire from 22-28 thou to maintain correct ride height from 850g to 2225g. As I anticipate adding additional mass to the model for traction purposes, I expect to use wires at the thicker end of the spectrum so it shouldn't end up too sloppy. To be honest, I just don't know, so this'll be my chance to experiment.
Why is deciding on a static deflection important? I have created fixings for the axle blocks that measure 7.5mm from axle c/l to its CSB mounting hole c/l. If I want to use a 1.5mm deflection I need to mount my frame fulcrums at 6mm above axle c/l (7.5-1.5mm). It's worth noting that if you are using CSB below the axle blocks, you'd add the deflection.
It's also important to note that we are not trying to mimic prototype spring deflection limits - static deflection is purely what puts the loco at the correct ride height. The actual amount by which the CSB can act as suspension is dependent on the physical limits of axle block movement.
Last edited:
adrian
Flying Squad
very impressive - if only for the fact that you still have all the drawn items. Given how fine some of those tabs are, I'm surprised a few components didn't end up in the bottom of the etch tank!
andrewb
Western Thunderer
So why CSB in the first place? CSB combines the best of individual springing and compensation into one method. Taken in isolation, the way an axle is suspended between 2 frame fulcrums on a bendy wire is just like any other method of springing, but the beauty is that as one axle is compressed upwards the adjacent axles are influenced downwards. This is because the wire is a loose fit in the fulcrums so can slide a little. So there is also longitudinal compensations going on, with the spring action dissipating the effect of irregularities in the track. Consequently, it ticks a lot of boxes: good electrical contact is maintained, it's self-levelling, the smoothness of ride is unsurpassed by any other method, it is quieter and just somehow looks good!
Working out the Frame Fulcrum position: This is done with a spreadsheet from the CLAG website. I use the Wyatt/Lichfield version which seems to an old art college graduate like me the most intuitive process. It is an iterative methodology, which is time-consuming, but the accuracy has always seemed good. I've attached a pdf of the plot for 1.5mm deflection with 25 thou spring wire.
Working out the Frame Fulcrum position: This is done with a spreadsheet from the CLAG website. I use the Wyatt/Lichfield version which seems to an old art college graduate like me the most intuitive process. It is an iterative methodology, which is time-consuming, but the accuracy has always seemed good. I've attached a pdf of the plot for 1.5mm deflection with 25 thou spring wire.
Last edited:
andrewb
Western Thunderer
The spreadsheet is not as complicated as it looks. Under 'Input Values' you enter the distance in mm between the drivers. By habit, I always imagine the loco facing to the left, so 'p' and 'q' are the distances between lead and intermediate, and intermediate and trailing, driver axles respectively. I usually start by getting it to auto-calculate, then adjust the values to avoid surface details, holes, and other artefacts that would be unfortunate or impossible to drill into. I reckon I can work to 0.1mm so don't ask it to round to the nearest 0.5mm.
What you're aiming to get in the 'Outputs' part is a gradient as close to zero as you can, the deflections Y1-3 to be as close to equal at the deflection you have decided on, with the centre axle bias reading a small % figure (the box colour automatically changes from red to green when you're about right). You definitely do NOT want Y2 to be less than Y1 & Y3, as this would mean that for a given axle weight the deflection is less in this axle than its neighbours (in other words the suspension firmer) which could induce porpoising. But you don't want it too much greater either or you lose the traction on that axle, and the associated problems that causes with the other axles.
This is where the iterative bit kicks in. I reworked the figures over and over for different spring size, weight, and the fulcrum positions (bearing in mind where they absolutely couldn't go) and finally achieved a 1.501mm deflection under 1414g unsprung mass with 25 thou spring wire, and a chassis gradient of 1 in 566896 and a +4% centre axle bias. I could probably have stopped the reworking earlier and settled for a much less stringent gradient - I reckon 1 in 1000 would probably be fine, as the final levelling will have to be done on completion with the actual weights and CofG realised. As it is, it looks like the CofG is about 1.5mm aft of the intermediate wheel: ideally it should be about the same distance in front of it. But these little tweaks will be needed unless one knows from the outset what the actual weight distribution is - the spreadsheet assumes equal weight on each axle.
As you can see from the ‘Fulcrum Offset’ results, they will be placed 29mm in front of the lead axle, 30.4mm in front of, and 33.2mm behind, the intermediate axle, and 24.5mm behind the trailing axle. Then I was able to rework the figures with differing weights and wire thickness to make sure the 1.5mm deflection - and therefore the correct ride height - was maintained. They are:
What you're aiming to get in the 'Outputs' part is a gradient as close to zero as you can, the deflections Y1-3 to be as close to equal at the deflection you have decided on, with the centre axle bias reading a small % figure (the box colour automatically changes from red to green when you're about right). You definitely do NOT want Y2 to be less than Y1 & Y3, as this would mean that for a given axle weight the deflection is less in this axle than its neighbours (in other words the suspension firmer) which could induce porpoising. But you don't want it too much greater either or you lose the traction on that axle, and the associated problems that causes with the other axles.
This is where the iterative bit kicks in. I reworked the figures over and over for different spring size, weight, and the fulcrum positions (bearing in mind where they absolutely couldn't go) and finally achieved a 1.501mm deflection under 1414g unsprung mass with 25 thou spring wire, and a chassis gradient of 1 in 566896 and a +4% centre axle bias. I could probably have stopped the reworking earlier and settled for a much less stringent gradient - I reckon 1 in 1000 would probably be fine, as the final levelling will have to be done on completion with the actual weights and CofG realised. As it is, it looks like the CofG is about 1.5mm aft of the intermediate wheel: ideally it should be about the same distance in front of it. But these little tweaks will be needed unless one knows from the outset what the actual weight distribution is - the spreadsheet assumes equal weight on each axle.
As you can see from the ‘Fulcrum Offset’ results, they will be placed 29mm in front of the lead axle, 30.4mm in front of, and 33.2mm behind, the intermediate axle, and 24.5mm behind the trailing axle. Then I was able to rework the figures with differing weights and wire thickness to make sure the 1.5mm deflection - and therefore the correct ride height - was maintained. They are:
- 847g (pretty much the finished model's basic weight) - with 22 thou wire
- 1200g - with 24 thou wire
- 1414g - with 25 thou wire
- 1653g - with 26 thou wire
- 2225g - with 28 thou wire
- 238g with 16 thou guitar string
- and so on.
Last edited:
Len Cattley
Western Thunderer
It's nearly a year now from the last post, any update from the last post?
Len
Len