A CSB Source book

billbedford

Re: A CSB Source book

Postby billbedford » Tue Mar 04, 2014 10:53 am

davidb wrote:The 'How to . . . ' I could do with concerns working out where the pivot points are meant to go. The rest is straightforward.

The spreadsheets are completely beyond me and I know this puts other people off CSB. How necessary are they? If I were to put a pivot half way between wheels with one at each end to match, would that be 'wrong' or mean it won't work? Can the placing of pivot points be simplified?


If you have a symmetrical wheelbase there is a simple rule of thumb. If the intermediate anchors half way between the axles, the outer anchors can be placed a distance 30% of the wheelbase beyond the outer axles. This relationship will hold true for any number of axles.

On a 6' x 6' wheelbase the distance between the axles and the outer anchors will be 8mm, for 7' x 7' it's 9.3mm and for 8' x 8' it's 10.7 mm (to the nearest 0.1mm)

On asymmetric wheelbases it is more complicated because there are many solutions that will give equal weights on the axles. However knowing of this rule of thumb gives you a starting point for using the spreadsheet. Enter in the the four anchor points as above, i.e. half way between the axles and 30% of each inter axle distance, and then adjust TWO of the anchors, either the two intermediate points or the two outer points, until you have a satisfactory solution.

Alan Turner
Posts: 643
Joined: Sun Jul 20, 2008 4:24 pm

Re: A CSB Source book

Postby Alan Turner » Tue Mar 04, 2014 10:53 am

When I wrote my spread sheet I tried to make it as intuitive a possible. All data is entered using scroll bars (yes I know they work opposite to how you expect but that's Microsoft's fault not mine) except for the basic data of wheel base, where the Centre of Gravity of the model is and weight, and you get an instant picture of wheel loads and Centre of Action/Gravity.

As Will says, all the maths is hidden and you have no need to understand how it works.

I would be happy to provide an appendix of the maths if you really want one.

regards

Alan

Terry Bendall
Forum Team
Posts: 2420
Joined: Sun Jul 27, 2008 7:46 am

Re: A CSB Source book

Postby Terry Bendall » Wed Mar 05, 2014 7:46 am

Noel wrote:computers and workbenches are not very compatible.


Might depend on the state of your bench Noel. :) I use a lap top computer for all my IT work and it regularly migrates between the desk that I use and the workbench. One use of the computer is as a source of reference pictures for whatever it is that I happen to be working on.


billbedford wrote:What is needed are succinct, easy to follow 'How-To' booklets that can lead people through the process of building working loco frames.


Thet sounds like a useful idea. All it needs is someone to write them. Are you going to offer Bill? :)

Terry Bendall

User avatar
David B
Posts: 1509
Joined: Sun Jul 20, 2008 5:30 pm

Re: A CSB Source book

Postby David B » Wed Mar 05, 2014 8:52 am

Thank you for the 'rule of thumb', Bill.

At Will's request I am re-reading the entries in the Source Book. I have not finished by some way, but I spent a couple of hours last night reading the CLAG pages and Will's earlier postings. I think I might be able to explain to someone what CSB is, but not how one applies it to any model. This is so far a completely beyond me and I am, as yet, no nearer to knowing how I would deal with any spreadsheet other than entering the wheelbase.

I will continue with my reading but I am reminded of the time I was doing the Science Foundation course with the OU in the 1970s. The first unit (week's work) was fine but the second and third units were on relativity and I very nearly gave up the whole course. In the end I just had to leave it and move on - I never did understand anything of relativity. It may be that I have to do the same with CSB but I have not finished trying yet.

User avatar
Noel
Posts: 1975
Joined: Wed Jun 23, 2010 1:04 pm

Re: A CSB Source book

Postby Noel » Wed Mar 05, 2014 11:45 am

Terry Bendall wrote:
Noel wrote:computers and workbenches are not very compatible.


Might depend on the state of your bench Noel.


Cluttered. It obviously depends on what you are doing, but I wouldn't have a computer anywhere near glues, paints or soldering fluxes in case of accidents. If you can be sure it's protected from these risks then fair enough.

Noel
Regards
Noel

dal-t
Posts: 654
Joined: Mon Jan 11, 2010 8:06 pm

Re: A CSB Source book

Postby dal-t » Wed Mar 05, 2014 12:23 pm

Noel wrote:I wouldn't have a computer anywhere near glues, paints or soldering fluxes in case of accidents


Safe enough, providing you're using something like this. Pity it's a Windows (7) box.
David L-T

David Thorpe

Re: A CSB Source book

Postby David Thorpe » Wed Mar 05, 2014 12:44 pm

Russ Elliott wrote:
Noel wrote:This might present problems for those who prefer a printed page or pages for reference whilst working, rather than a screen. Apart from readability issues, computers and workbenches are not very compatible.

In general, Noel, I would agree, but in the particular case of 41.0, whose primary purpose is to try and get ideas and principles across ('digest'), I find it difficult to imagine anyone having that on their workbench as an adjunct in actual construction. It's not like a set of kit instructions, is it?

But not everyone wants to read things on their computer especially if, like me, you don't have a laptop or tablet. Frankly, if it's not printable, it loses a lot of its usefulness as far as I'm concerned. It's a bit like magazines - they're not kit instructions, but the paper versions are far more versatile and readable than the online ones.

DT

User avatar
Guy Rixon
Posts: 909
Joined: Sun Mar 27, 2011 6:40 pm

Re: A CSB Source book

Postby Guy Rixon » Wed Mar 05, 2014 12:49 pm

davidb wrote:I will continue with my reading but I am reminded of the time I was doing the Science Foundation course with the OU in the 1970s. The first unit (week's work) was fine but the second and third units were on relativity and I very nearly gave up the whole course. In the end I just had to leave it and move on - I never did understand anything of relativity. It may be that I have to do the same with CSB but I have not finished trying yet.


Pro-tip: one can use general relativity, even in professional astronomy, without actually handling the tensor maths. Many common, useful cases in cosmology have simple equations that can be looked up and applied.

Similarly in P4, I find I can easily apply the CSB spreadsheets whereas if I try to use the springing equations (Russ Elliott's pages on the CLAG site) directly, I get bizarre and inconsistent results (suggesting I'm coding them wrongly, not that Russ stated the equations wrongly).

In each case, one needs some contextual understanding of when to apply the computational tools and what measurements to feed them to get meaningful results. I.e., one needs to understand the procedure, not the maths.

User avatar
Paul Willis
Forum Team
Posts: 3033
Joined: Sun Jul 20, 2008 6:00 pm

Re: A CSB Source book

Postby Paul Willis » Wed Mar 05, 2014 7:20 pm

David Thorpe wrote:But not everyone wants to read things on their computer especially if, like me, you don't have a laptop or tablet. Frankly, if it's not printable, it loses a lot of its usefulness as far as I'm concerned. It's a bit like magazines - they're not kit instructions, but the paper versions are far more versatile and readable than the online ones.

It is a bit time and place... Like many things in the hobby, not having a fixed approach is handy.

A couple of Missenden's ago, I turned up on Friday for the weekend with a High Level kit. In fact the Coffeepot that you may have seen elsewhere on here.

I started off on Friday night building it, but by about stage three, I was questioning what I was doing. In a very rare mistake, my kit for the GER variant had been packed with the generic Neilson industrial loco instructions.

So on Saturday morning, I called Chris Gibbon to check that they wouldn't work :-(

Almost instantly, he emailed the correct version to my laptop, and I happily spent the rest of the weekend with them on display, balanced on top of my toolbox. And yes, I did print them off to use on paper as soon as I was home! But without an electronic option, I'd have been stuffed for the weekend...

Cheers
Flymo
Beware of Trains - occasional modelling in progress!
www.5522models.co.uk

User avatar
Russ Elliott
Posts: 930
Joined: Thu Jun 02, 2011 6:38 pm

Re: A CSB Source book

Postby Russ Elliott » Thu Mar 06, 2014 1:24 am

Guy Rixon wrote:Similarly in P4, I find I can easily apply the CSB spreadsheets whereas if I try to use the springing equations (Russ Elliott's pages on the CLAG site) directly, I get bizarre and inconsistent results (suggesting I'm coding them wrongly, not that Russ stated the equations wrongly).

Ah, if I'm correct in assuming what you are doing, I think we have an application misunderstanding.

The deflection of any beam is founded on the equation of the curve of each segment of the beam, and the curve equation of each segment of the beam will depend on what are called the 'boundary conditions' at each support/load point. These boundary conditions constitute the moments on the beam, and these moments determine the reaction loads at the support/load points.

Where the load is midway between two supports in a non-CSB situation, for example, like a normal wagon sprung W-iron, the reaction load at each support is half the counteracting axle load. With a CSB situation, where multiple supports/loads are involved, the magnitudes of these reaction loads are very different, because the moments about the midframe point(s) are non-zero. In other words, the equations given for the simple non-CSB situations are not applicable.

Varying the position of the axle with respect to the frame points will produce a different set of reaction loads, but in all cases, the vertical loads are in balance - the sum of the 'ups' is equal to the sum of the 'downs'.

reaction-load-differences.png
You do not have the required permissions to view the files attached to this post.

User avatar
Russ Elliott
Posts: 930
Joined: Thu Jun 02, 2011 6:38 pm

Re: A CSB Source book

Postby Russ Elliott » Thu Mar 13, 2014 1:01 pm

Powered by the Roger Wyatt engine, I've made an 'entry level' 3-axle version in response to davidb's cri de coeur ("The spreadsheets are completely beyond me"). The visual presentation is hopefully more accessible, and the method and the notes have been kept minimal. It deliberately does not carry the bells and whistles of Will's spreadsheet (nor the rigour of Alan Turner's), which users can of course progress to if they want to explore more.

Screenshot:
You do not have the required permissions to view the files attached to this post.

billbedford

Re: A CSB Source book

Postby billbedford » Sun Mar 16, 2014 2:34 pm

For my own amusement I've reworked Russ's diagram using the input criteria I suggested in a earlier post, and here are the results. They were calculated using Alan's spreadsheet which gives the gives the proportion of the total weight carried by each axle. For comparison Russ's plot will give 34.30 : 32.54 : 33.16 percent on each axle respectively.

First using the centre anchors fixed to the wheel base centres:

simple-CSB-3-axle-spreadsheet-fixed-centres.png



and now with the outer anchors at approximately 0.3 times the adjacent wheelbase:

simple-CSB-3-axle-spreadsheet-fixed-outers.png


Note that on both plots the outer anchors are closer to the axles, and therefore less likely to collide with other parts of the frames, than on Russ's plot.
You do not have the required permissions to view the files attached to this post.

Alan Turner
Posts: 643
Joined: Sun Jul 20, 2008 4:24 pm

Re: A CSB Source book

Postby Alan Turner » Sun Mar 16, 2014 5:34 pm

And if you use my spread sheet you get:

Six coupled chassis C.png


Six coupled chassis B.png



regards

Alan

Edit to show both of Bill's examples
You do not have the required permissions to view the files attached to this post.
Last edited by Alan Turner on Mon Mar 17, 2014 8:43 am, edited 5 times in total.

User avatar
Will L
Posts: 2516
Joined: Sun Jul 20, 2008 3:54 pm

Re: A CSB Source book

Postby Will L » Sun Mar 16, 2014 6:43 pm

Russ Elliott wrote:Powered by the Roger Wyatt engine, I've made an 'entry level' 3-axle version in response to davidb's cri de coeur


Interesting, but given that yours and mine are both powered by Rogers basic calaculation, why do we get different deflections?

billbedford wrote:For my own amusement I've reworked Russ's diagram using the input criteria I suggested in a earlier post.


Interesting but just how did you make the translational from deflection to weight distribution? I'm surprised Alan let that pass. We decide some some ago that the Wyatt process, while producing accurate results when correctly handled, it doesn't tell you anything which can be reliably translated into weight distribution. The discussions that led to this conclusion are buried in the Abstruse CSB theory thread somewhere

Alan Turner
Posts: 643
Joined: Sun Jul 20, 2008 4:24 pm

Re: A CSB Source book

Postby Alan Turner » Mon Mar 17, 2014 8:22 am

Will L wrote:
Interesting but just how did you make the translational from deflection to weight distribution? I'm surprised Alan let that pass.


I had just returned from laying ballast (that is 12" to the foot) so was a bit tired and didn't pick up on what Bill was apparently doing.

Not wanting to go over what was said 3 years ago (how long ago? - time flies) wheel loads are an output. If you start with wheel loads then you have to adjust them so that all deflections are equal for them to be a valid solution. The underlying assumption in doing that is the Centre of Action is coincident with the Centre of Gravity.

This version gives a slightly softer spring to the centre axle. You will also notice that the Centre of wheel base/Centre of Gravity/Centre of Action are more coincident.

Six coupled chassis D.png


regards

Alan
You do not have the required permissions to view the files attached to this post.

David Thorpe

Re: A CSB Source book

Postby David Thorpe » Mon Mar 17, 2014 9:52 am

This is all becoming over complicated again. I've now made some chassis (loco and tender) using CSBs. All I've done is to use one or other of the CSB calculators (Will's or Alan's or Russ's, I'm not sure which), put in the wheelbase dimensions, guessed at the weight, and used the resulting information to position the fulcrum points - I haven't attempted to change them at all as they have always been within the limits of the chassis. The results seem to run well, though I've no way, or need given the limitations of my layout, to test haulage capacity. I'm now concerned that I'm not doing enough!

DT

User avatar
David B
Posts: 1509
Joined: Sun Jul 20, 2008 5:30 pm

Re: A CSB Source book

Postby David B » Mon Mar 17, 2014 10:19 am

David Thorpe wrote:This is all becoming over complicated again.


This was the thrust of my earlier posting. If CSB is to be used more, and I think it THE way to spring a locomotive, then it has got to be presented in a simpler way for the 'average' modeller.

Russ' latest sheet is a step forwards for which I am grateful. It takes out bits that Mr Average just does not need to know and he gives simpler instruction on what to do. Bill gave a 'rule of thumb' earlier in this thread and from conversations I have had with other people, little more appears to be needed to get a good, running chassis.

David Thorpe refers to haulage capacity and he is right in that for his and I suspect many layouts it is not an issue, especially when pinpoint bearings are so widely used. Will gave a good explanation elsewhere on wheel slipping. If one has a layout that does test locos to the limits of their pulling power, then more care and work needs to go in to setting out the CSB but for most modellers it is not going to be necessary.

Let us simplify this and make sure it is accessible to the majority of modellers.

User avatar
Will L
Posts: 2516
Joined: Sun Jul 20, 2008 3:54 pm

Re: A CSB Source book

Postby Will L » Mon Mar 17, 2014 11:47 am

David Thorpe wrote:This is all becoming over complicated again. I've now made some chassis (loco and tender) using CSBs. All I've done is to use one or other of the CSB calculators (Will's or Alan's or Russ's, I'm not sure which), put in the wheelbase dimensions, guessed at the weight, and used the resulting information to position the fulcrum points - I haven't attempted to change them at all as they have always been within the limits of the chassis. The results seem to run well, though I've no way, or need given the limitations of my layout, to test haulage capacity. I'm now concerned that I'm not doing enough!


Don't let us put you off. In the end our basic purposes is to determining where the fulcrum points on a chassis should be. All that's actually needed is to chose a spread sheet, enter the wheel base, juggle the fulcrum points till you get an suitable answer that fits the locos chassis, and weight the the loco so the CofG turns up where the spreadsheet says it should be. The result will be a loco that performs well and makes best use of the adhesive weight available. This you've done and I can't see any need to do any more, unless you have a particular wish to understand why it works. I think you can safely ignore anything which seems to suggest you haven't done enough.

I've queried Russ's new spread sheet because it shows minor difference in the detail results from mine and I don't understand where these have come from. The size of the difference isn't sufficient to worry the casual user. Bill has, I'm afraid, committed a heresy which was at the heart of the long discussion about this that went on the appropriately named "abstruse CSB theory" thread. The net result of that discussion was that Alan produce a spreadsheet that worked by a totally different method, hence avoided the heresy, but rather satisfyingly gives the same final answers in terms of the fulcrum point placement. Again the general user shouldn't let these esoteric concerns worry them.

What I do get worried about is when people start thinking they can guess where the fulcrum points should be. This is because the performance of the springy beam where it passes through the fulcrums in the centre of the chassis is significantly different from it performance with the fulcrum points at either end. The result is that fulcrum point placement isn't intuitive, unless perhaps for those who alreday have lots of previous experience gained from using a spread sheet. Guessing is likely to produce odd result and the last thing we need is somebody slagging off the whole concept because they guessed wrong.

Chris Mitton
Posts: 258
Joined: Sun Aug 31, 2008 1:18 pm

Re: A CSB Source book

Postby Chris Mitton » Tue Mar 18, 2014 12:25 am

David Thorpe wrote:The results seem to run well.... I'm now concerned that I'm not doing enough!

You've surely answered your own questions - if they seem to run well (meaning they do run well enough for them to do what you want them to!) ergo you've done enough, since good running is the object of the exercise....
Regards
Chris

billbedford

Re: A CSB Source book

Postby billbedford » Tue Mar 18, 2014 10:08 am

davidb wrote:Let us simplify this and make sure it is accessible to the majority of modellers.


That's the problem, there just is no simple solution. In fact there are a myriad of simple solutions. You could produce plots where two of the anchors were placed at any random point in the relevant section of the wheelbase and the other two points calculated from them. Almost all such plots would give a solution that was as valid as any other.

The question is about how this calculation is presented to people who are not interested in the mechanics of the calculation, should we produce 'definitive' plots, as Russ has done, even though they are quite arbitrary, and he can't explain the criteria for choosing the initial points? Or should we come up with simple forms of words that guide people into using the spreadsheets with the minimum of effort?

Such form of words could be, but are not limited to, these:

Placing the inner anchors at the mid points of the wheelbase will always give a solution with 2 or 3 axles.
Placing the outer anchors at 0.3 (or one third) of the adjacent wheelbase from the outer axles always give a solution.
Place any two of the anchors a specific distance from the position of two brake hangers always give a solution.
Etc, etc.

billbedford

Re: A CSB Source book

Postby billbedford » Tue Mar 18, 2014 10:29 am

Will L wrote:
billbedford wrote:For my own amusement I've reworked Russ's diagram using the input criteria I suggested in a earlier post.


Interesting but just how did you make the translational from deflection to weight distribution? I'm surprised Alan let that pass. We decide some some ago that the Wyatt process, while producing accurate results when correctly handled, it doesn't tell you anything which can be reliably translated into weight distribution. The discussions that led to this conclusion are buried in the Abstruse CSB theory thread somewhere


I pasted results from Alan's spreadsheet into Russ's diagram.

In the real world on a balanced loco all the spring deflections are equal and the axle weights differ. Continuing to present this information in any other way just adds another layer of confusion to the mathematically challenged.

Alan Turner
Posts: 643
Joined: Sun Jul 20, 2008 4:24 pm

Re: A CSB Source book

Postby Alan Turner » Tue Mar 18, 2014 10:42 am

billbedford wrote:
In the real world on a balanced loco all the spring deflections are equal and the axle weights differ.


At last a light has finally illuminated.

regards

Alan

User avatar
Russ Elliott
Posts: 930
Joined: Thu Jun 02, 2011 6:38 pm

Re: A CSB Source book

Postby Russ Elliott » Tue Mar 18, 2014 9:16 pm

Will L wrote:Interesting, but given that yours and mine are both powered by Rogers basic calaculation, why do we get different deflections?

We don't. The 0.33mm beam (13 thou) used in the example in my previous message gives the following deflections for all the Wyatt-engine spreadsheets: 0.467mm, 0.548mm, 0.464mm (subject to rounding of the last decimal place, such rounding arising either from the use of a metric value for the beam diameter or a conversion of imperial to decimal) when using an elastic modulus of 205Gpa. Will - your spreadsheet uses 200GPA.

The elastic modulus of spring steel is usually quoted at somewhere between 200 and 210GPa, so 205GPa was used in the original Wyatt spreadsheet as a typical value. I don't think anyone is going to be worried about a 2% difference in deflection value, and besides which no one really knows what the value is for a bit of plated Ernie Ball E string.

Moving on to the debate concerning the difference between approaches of the various spreadsheets, I don't think I can add much to what Will has already said. In all cases, the percentage loads are in direct proportion to the actual loads when all the deflections are equal. If, when using a Wyatt sheet, they are not initially equal, they can be made equal most easily by altering the load value on the middle axle. When all the deflections are equalised, the Wyatt engine acts exactly the same way as Alan's spreadsheet. Alan's merely uses equal deflection as a precondition, but does not allow, as the Wyatt sheet does, an examination (if one wishes to do so) of the sensitivity of the loads to a nominated value of track irregularity.

I think we need a sense of perspective. Worrying about the third place of decimals on deflection values or the first place of decimals in the load percentages, or trying to place the CofG to an accuracy of greater than 0.5mm, is just plain silly.

If anyone feels in need of worrying about something significant, think about the assumption of all the spreadsheets that frame fulcrum points are of zero width.

User avatar
Will L
Posts: 2516
Joined: Sun Jul 20, 2008 3:54 pm

Re: A CSB Source book

Postby Will L » Tue Mar 18, 2014 9:56 pm

Russ Elliott wrote:
Will L wrote:Interesting, but given that yours and mine are both powered by Rogers basic calaculation, why do we get different deflections?

We don't. The 0.33mm beam (13 thou) used in the example in my previous message gives the following deflections for all the Wyatt-engine spreadsheets: 0.467mm, 0.548mm, 0.464mm (subject to rounding of the last decimal place, such rounding arising either from the use of a metric value for the beam diameter or a conversion of imperial to decimal) when using an elastic modulus of 205Gpa. Will - your spreadsheet uses 200GPA.

The elastic modulus of spring steel is usually quoted at somewhere between 200 and 210GPa, so 205GPa was used in the original Wyatt spreadsheet as a typical value. I don't think anyone is going to be worried about a 2% difference in deflection value, and besides which no one really knows what the value is for a bit of plated Ernie Ball E string.


That explains it, thank you. I agree the differences weren't significant, but I am used to getting exactly the same answers from different incarnations of the Wyatt engine so I was worried when we didn't.

User avatar
Will L
Posts: 2516
Joined: Sun Jul 20, 2008 3:54 pm

Re: A CSB Source book

Postby Will L » Tue Mar 18, 2014 11:59 pm

billbedford wrote:
davidb wrote:Let us simplify this and make sure it is accessible to the majority of modellers.


That's the problem, there just is no simple solution. In fact there are a myriad of simple solutions. You could produce plots where two of the anchors were placed at any random point in the relevant section of the wheelbase and the other two points calculated from them. Almost all such plots would give a solution that was as valid as any other.

The question is about how this calculation is presented to people who are not interested in the mechanics of the calculation, should we produce 'definitive' plots, as Russ has done, even though they are quite arbitrary, and he can't explain the criteria for choosing the initial points? Or should we come up with simple forms of words that guide people into using the spreadsheets with the minimum of effort?

Such form of words could be, but are not limited to, these:

Placing the inner anchors at the mid points of the wheelbase will always give a solution with 2 or 3 axles.
Placing the outer anchors at 0.3 (or one third) of the adjacent wheelbase from the outer axles always give a solution.
Place any two of the anchors a specific distance from the position of two brake hangers always give a solution.
Etc, etc.


That is true, and I agree that some guidance as to where to look for good solutions is of value. If you look back at this post from August 2010 I was trying to provide exactly that.

Also my version of the spread sheet has an "autocalulate" option that means all you need enter is the wheelbase and it will give you an answer based on calculating all fulcrum points based on the overall wheel base. This is also documented here as a paper and pencil method, thus avoid the spread sheet altogether. Answers should always be acceptable but asymmetrical wheelbases(where the middle wheel is significantly closer to one end than the other) may be improved by trying other values of the constant used to calculate the outer fulcrum points (shown as the "End Factor" value on the spread sheet) within a very limited range. There are a couple of boxes on the spread sheet which are green if the results are acceptable and red if they aren't

To obtain the required fulcrum points
csb exl 01anotated 2.jpg


The results produce an answer which uses a relatively short thin wire. Russ's hand crafted results usually use a longer thicker wire, functionally they come to the same thing, what fits you loco best is the right answer.

When the loco is finished, to weight the loco and select the right size of CSB wire, run the spread sheet again but this time...
csb exl 01anotated 3.jpg
You do not have the required permissions to view the files attached to this post.


Return to “Will L”

Who is online

Users browsing this forum: No registered users and 0 guests