Atomic Systems

For those who have been folllowing the earlier posts in the General MSTS section, where it probably got a little technical for most, I have just realized something that will be an issue it getting the best values.

The problem is tank engines:

Since the input data for Joe's FCalc 2.0 requires number of axles, weight, frontal area, drag coefficient and speed, the results for plain bearing tank engines are somewhat distorted.

For example a Great Western 51XX 2-6-2T is more or less a 43XX 2-6-0 with an extra pair of wheels and extra weight. (I'm aware that Churchward designed the proginator of the 51XX first, and the 43XX was developed from it)

Now all this is well and good as far as the basic friction calculation is concerned, however for the added motion/cylinder values, the tank engine figures, being based primarily on weight will be higher than they should be.

It seems that the solution may be to calculate the basic values for both the tank engine and a similar, but not necessarily identical, tender engine and add the tender values for cylinder and motion friction to the basic bearing resistance of the tank engine.

Um..... yes - I think... it's a bit early to get my poor old brain around this, but I can see a small problem.

Many of the earlier UK railways, my pet GER being a great example, had many classes of tank engine without any 'matching' tender loco (mostly smaller ones like the J65/6/7/8/9 0-6-0Ts and the F3/4/5/6 2-4-2T suburban tanks) How would you recommend we deal with that, Ian? Bearing in mind that we're trying to get a semblance of reality here as absolute truth-to-life is impossible with the restrictions of MSTS, might it be possible to find a workable 'tank-engine factor'?

Barry,

I agree with what you say,

There usually are a few locos that match reasonably closely, For a J69, you could use the smallest 0-6-0, for the F series you could use a small 2-4-0

What I intend to do is exactly what you suggested, using the tender engines that have similar cylinders. size and boiler pressure to establish a tank engine factor.

I'm not sure there is a problem, a 2-6-2T will be heaver than a similar 2-6-2 and so more friction to take into account - but remember it has no tender to pull.

The extra weight on the axles of tank engines did mean some had a lesser route classification than a similar tender loco - the GWR 94XX 0-6-0T being a good example, C instead of the B of the 2251 tender class. Therefore, you should expect higher friction values.

I do not think you can truly replicate a tank engine or even a tender
one as the fuel is like the postman's bag it gets lighter as you travel on
your way........well know fact tank engines would hold there feet when
fully loaded different story when water mainly was used up, 2000 gals
= 9 tons near enough not that you would let it get that low but 3 quarters down would be enough to effect traction on the wrong type of rail with use
of coal as well friction values use half water and half coal that's my policy anyway..........Rick

A slight misunderstanding here.

The tank engines were heavier and often had more wheels than the corresponding similar sized tender engine.

The point I was making is that while the wheel bearings and wind friction will work as in Joe's formula, the additional weight of the tank engine will distort the cylinder and motion addition which will be too high, being based on a much larger total weight, even with the fuel and water at half levels.

For example an LMS 4MT 2-6-4T will appear to have total friction closer to a large 4-6-0 or small 4-6-2, than say the Ivatt 4MT 2-6-0 which is the closer tender type.

What is the result if you apply roller bearing figures to a tank?

Mervyn

This thread should be read in conjunction with and as a continuation of the post on Friction Bearings in the General MSTS Discussion forum, where the last post was September 13.

This will make it clearer what this is about

Now I am at home and have access to my full documents I can reply in detail.

The recomended values from the post of September 7 0351 BST were as follows:

For roller bearing locomotives the value should be ((Roller bearing value)+(Roller bearing Value * 14.26))

For Plain Oil Bearing locomotives, the value should be ((Roller bearing value * 1.157)+(Roller bearing Value * 14.26))

For Plain Grease Bearing locomotives, the value should be ((Roller bearing value * 1.471)+(Roller bearing Value * 14.26))

I would stick to them for tender engines, however for tank engines, for the present I would use:

For roller bearing locomotives the value should be ((Roller bearing value)+(Roller bearing Value * 9.84))

For Plain Oil Bearing locomotives, the value should be ((Roller bearing value * 1.157)+(Roller bearing Value * 9.84))

For Plain Grease Bearing locomotives, the value should be ((Roller bearing value * 1.471)+(Roller bearing Value * 9.84))

This should bring it closer to the intent of Rick's post

Incidentally, I am using a half full value on the consumables for this.

I am working with some of Joe's data to bring this automatically into the .eng file calculator version 2.0.

I've been playing with Joe's Fcalc and have had a couple of thoughts about starting resistance for UK steam loco's:

I wonder if anyone has access to info about the smallest steam chest pressure required to start real locos? Presumably this figure could then be applied in the formula for tractive effort to give an idea of the starting resistance. From my own experience, our 35 ton 0-6-0 saddle tank engines won't move on much less than 20psi. It would be really helpful if I could remember their tractive effort figures of course...

On loco's with steam chest pressure gauges it should be relatively easy to measure, though I suspect there aren't a huge number about. Oddly enough, Linda on the Ffestiniog has one IIRC!

Another option which would give us some figures, is for me to dig out Bristol Harbour Railway's 1 ton spring balance when I next get the chance (probably not till March sadly) and see what force is required to start one of the engines rolling (we tow then out of the shed with a mechanical loader shovel).

My approach with Joe's Fcalc on UK steam loco's has been to use the friction bearing steam loco outputs, but modifying the -0.25 to -0.1 to stop the bouncing and putting this revised figure into the relevant Excel spreadsheet to recalc the other parameters to suit. I applied this approach to Paul Gausden's lovely model of GWR 1369 and it just starts to move with 19psi on the steam chest pressure gauge, which feels about right. Of course I don't know how MSTS uses the steam chest pressure to calculate tractive force, so this could be completely wrong!

HTH

Cheers

Richard

Now someone else has noticed that it is the -0.25 in FCalc2 that makes things bounce. I have been experimenting with the .eng file for my HR Passenger tank and found that I have to take this all the way down to -0.05 to stop the bouncing altogether. Since this is an exponent to a small number (speed up to 5 or 6 mph) in the formula the change doesn't make an enormous difference and still allows you to keep a high starting resistance for the solid bearings. (I should have thought to plug it into the spreadsheet though!) There is hope for me to use FCalc2 data yet.

I would advocate nothing more than -0.11 after experimentation, however it should be OK to use -0.25 on the locomotive only, but use -0.11 elsewhere in the train.

I had some very unfortunate incidents surging backwards through a stop signal behind the train using -0.25 throughout the train.

Are you also adjusting the coupler values or sticking to defaults? It might be that and not the f1 value in FRICTION. FWIW it's very hard to get starting friction close to correct w/o pushing f1 below -0.2

Definately a blast from the past.

The last posts were before the long series of posts in the physics forum discussing couplers, so the F1 value may need some revisiting to tie in with this topic, when I have time (Probably when I retire) I'll revisit this but at present there have been no changes, although further testing is definately required.


Ian