oldnorton
Statesman
5" gauge LMS enthusiast
Posts: 688
|
Post by oldnorton on Oct 8, 2021 16:14:43 GMT
I cannot find the thread that I thought I saw suggesting 0.045" end float for the centre axle on a three axle, 5" gauge tender. But, I am pleased I saw it because it caused me to realise I had not checked my Modelworks Britannia tender frame/axlebox float and without that float things are going to derail.
Sure enough, there was no end float whatsoever, well perhaps a few thou only in all three axles because the axle boxes were nice and free in the horns.
Before I machined the centre axle flanges to create the necessary end float, I attempted the maths to work out what radius of track would need how much end float on a middle axle. I did it using sines and cosines on a scientific calculator, but I suspect it should have been a job for radians, but my maths does not stretch to those.
The first and third axles on the Britannia tender are 15" apart (approximately), and my scanning of the web suggested reasonable minimum track radii of 50ft for a raised track and perhaps 18ft for a portable track. Now here I would appreciate some comment regarding what a radius might be negotiating switches (points).
Anyway, my elementary maths came up with needing an end float of 0.047" at 50ft radius and 0.130" at 18ft. My concern is whether my sums are correct and what radii might be encountered?
Has anyone done these sums themselves?
Norm
|
|
|
Post by keith1500 on Oct 8, 2021 16:38:54 GMT
|
|
|
Post by Deleted on Oct 8, 2021 17:02:59 GMT
Hi Norm
I can't help with your 3 axle Brit tender but can pass on the details for my Gresley 4 axle as stated by Don. The intermediate axles have 1/8th side play and 3/16 tilt ability.
Kind regards
Pete
|
|
jj
Active Member
Posts: 47
|
Post by jj on Oct 9, 2021 15:56:00 GMT
Hi Norm
I haven't done the calculation for the tender axles, but I did look into the sideways movement of the front loco bogie in response to a query from another Modelworks Brit builder, who was concerned that his bogie was going to hit the drain cock linkage. My reply to him was as follows:
The rear bogie axle is about 350mm from the centre driver, and the front one 520mm. So if we assume a minimum curve radius of 12 metres [ie about 40ft], using Pythagoras, the rear bogie axle if not displaced would be sqrt(12*12 + 0.35*0.35) = 12.005 metres from the centre of the curve – so it has to move in by 5mm to follow the curve. Similarly the front bogie axle is sqrt(12*12+.52*.52) =12.011, so needs to move in by 11mm. So the bogie pivot would slide across by (5+11)/2 = 8mm, and the bogie would pivot just very slightly to give the 3mm difference at each axle ... In practice the movement may be even less because the bogie will tend to pull the loco around the curve, causing the drivers to skew in the track (the track on a tight bend will be slightly more than 5” gauge to accommodate the fixed wheelbase of the drivers).
A similar calculation could be done for the tender axles.
Kind regards
John
|
|
oldnorton
Statesman
5" gauge LMS enthusiast
Posts: 688
|
Post by oldnorton on Oct 11, 2021 11:36:44 GMT
Thank you Keith for finding that very recent thread. With the Sagitta calculation I can develop this a bit more:
To answer the first of my problems (have I calculated correctly using sines and cosines?), the good answer is yes. For a 15" overall, three axle separation and 50ft radius curve, using the Sagitta calculation quoted by Keith I come up with a difference between the chord (straight line across flanges) and the point on the track curve of 0.047" (rounded 3DP), and my previous calculation gave an identical figure.
However, I have had a lot of thinking about all this. I was making a false assumption in my first posting that the axle end play was essential to help negotiate curves. I think I was quite wrong.
If you have three axles (rigidly fixed in-line), with the first and third set 15" apart, what is the distance between the centre axle flange and the track inside as the track takes on an increasing curve? The calculations above tell us that the track deviation from a straight line is 0.047" at 50ft radius. But our flanges are not set at exactly 5" apart. Using GL5 standards with 4 11/16" back-to-back and 3/32" flanges the width, at the point where the taper tread radius rises to meet the flange, is 4.875", i.e. a clearance of 0.125". If you use a smaller back-to-back, as indicated on some drawings, of 4 5/8" then that clearance increases to 0.187".
Therefore our tender with no end float in any of the axles will not derail at 5" gauge and 50ft curve radius, but the wheels will be running off the tapered portion of their tread and onto the radii before the flange. The track radius could in fact reduce to 18ft (garden railway territory!) when the Sagitta distance will be 0.130" and the GL5 flanges are now tight on the track (but not those flanges built to a more generous back-to-back).
But this all ignores a common solution in track laying, when the gauge is increased from 5" to perhaps 5 1/8" on tight curves (whatever tight means?). Now there will be an adequate 5.125" track separation, less the 18ft Sagitta reduction of 0.130" to give 4.995", and our flanges are 4.875", thus 0.120" clearance.
So, if the wheel design and track widening on tight curves is giving us adequate clearance do we need end float in axles, and who recommends it?
I have looked at the Don Young Black 5 design for the reason that Don was a railway employee and he perhaps had a understanding of what might be better design. I am part way through a Black 5 build so have the drawings and have previously built the loco axles. Don says nothing about this in his writings in LLAS on the Black 5 design, but if you analyse the drawings and use a calculator it shows that he allows 0.049" end float in each of the three tender axles, and for the locomotive three main axles there is 0.036" float in each one.
Now, if Don Young thought this necessary what about Perrier for the 5" Britannia and Les Warnett for the 92220 tender (used by most Britannia builders). Again, by analysing the drawings I see that Perrier allows an end float of 0.005" only on the locomotive, and Warnett shows 6.9255" width over the bearings sitting between 6.9260" across the axle boxes in the horns, i.e. 0.0005" or effectively nothing.
I now see why my Modelworks tender had no end float in its axles as the design was copying what had been drawn. So should there be some end float? I am thinking that Don Young says 0.049" on each axle and Ed suggested 0.025" as a maximum on the middle axle. Personally I think there should be some float, and 0.047" on the middle axle would let those treads come off the radii if the tender is at speed on a 50ft curve. It should ride better I think.
Thinking about the locomotives, is it sensible that Perrier has no end float specified for the three main axles? Don Young says 0.036" for his Black 5 so surely 0.020 - 0.030" would be sensible for the Britannia? I know that my rebuilt Modelworks has some play, but I failed to measure it when I fitted new springs. Next time it is upside down I shall use a DTI and see what it is.
This whole subject must surely have been written about by somebody in ME in the past? Whether my thoughts are of interest to anyone here I really don't mind as I have written up these notes for my own sake to help understand what is the correct policy on end play. I am grateful to original poster 'Britannia' asking the question about his Black 5 tender.
Norm
|
|
don9f
Statesman
Les Warnett 9F, Martin Evans “Jinty”, a part built “Austin 7” and now a part built Springbok B1.
Posts: 960
|
Post by don9f on Oct 11, 2021 15:52:03 GMT
Hi Norm, during the recent overhaul of my 9F and its tender, I discovered that there was virtually no sideplay in the tender wheelsets...as I think you found from the drawings. This had never been any apparent problem during the 7 years running the loco had seen, but I re-machined the axle ends/wheelbosses to create a total nominal 1/32nd inch sideplay on each wheelset....as this just seemed a sensible thing to do! Also to take into account is that "ideally", the gauge face of the rails should have a radius, not a square corner, so this would help a bit where the sideplay may otherwise be a bit lacking. Cheers. Don
|
|
|
Post by ettingtonliam on Oct 11, 2021 20:02:05 GMT
Your comprehensive calculations assume (I think) that the track is exactly to gauge, or at least not tight to gauge. This might not be the case, so wouldn't it be sensible to allow some end play? Just in case? Can't do any harm can it, provided its not taken to excess. Dons figure seems reasonable.
|
|
|
Post by keith1500 on Oct 11, 2021 20:05:37 GMT
I know the deliberations I had when building my LMS milk tank over the amount of end float. It got me thinking while traveling home on the tube and I remembered a book I had. RAILWAY MECHANICAL ENGINEERING. Here is an extract from that which might help? Three axle six wheel vehicle by GL5Keith1500, on Flickr
|
|
|
Post by doubletop on Oct 13, 2021 7:03:20 GMT
In the past I have considered some sort of model, even in Excel, that could work this out. The controlling factors are; track guage, number of axles, end float of each, axle spacing, wheel diameter, flange height, back to back spacing, flange root thickness, flange profile (but assuming square would give a bit of allowance when they aren't' and track gauge varaiation (but that could be assumed to be correct). The minimum track radius the loco could work one could then be calculated. I did have a go at doing it in Excel at one point and then my head exploded so I gave up. One of you maths wizzes out there should be able to crack it. Input from keith1500 book may be a help but I've not had time to read it yet. Pete
|
|
|
Post by Deleted on Oct 13, 2021 7:48:20 GMT
I recall reading a great article about the new P2 design study and tests with Tornado. There's lots of interesting info in it and the results of the 'Vampire' computer model used. This link gives some of the info although I recall there being more including a computer simulation. www.p2steam.com/design-study/Pete
|
|
Neale
Part of the e-furniture
5" Black 5 just started
Posts: 276
|
Post by Neale on Oct 14, 2021 7:54:05 GMT
Interesting (to me, anyway - I'm also building one!) that DY shows less end float on the loco axles than the tender. Is this partly because too much float on loco axles means that connecting and coupling rod bearings must be that much slacker? And how much clearance over a "good running fit" should one allow to cope with axle movement?
|
|
oldnorton
Statesman
5" gauge LMS enthusiast
Posts: 688
|
Post by oldnorton on Oct 14, 2021 9:11:44 GMT
Neale
I would guess that there are a good number of model engines and tenders running with no end float whatsoever. I can only give as examples the Perrier Britannia and Warnett BR1 tender, but I would think there must be more. I wonder how many builders do analyse the clearances in the drawings and then check the build. And these engines seem to run around raised tracks ok, but we don't know if they would have problems on tight ground level tracks that might have been avoided with some end float. Going around on big raised tracks is a completely different environment.
Don Young had an approach to this that I am thinking might have been based on some full size practice perhaps? His 0.049" clearance on the Black 5 tender axles (odd decimal because 3mm plates used when everything else is fractional!) is perhaps more than enough, and others, and I would agree, feel that or a bit less is enough on the centre axle alone.
Too much Loco end float will cause problems with parts of the motion now catching; you often only have 0.040" or so clearance on revolving parts anyway. I would try not to exceed the DY drawing figure of 0.036" per axle, even 0.020" would be perhaps ok. I found my Black 5 to be 0.030" on the leading and trailing and 0.040" on the centre - it just came out like that making everything to the drawing. The coupling rods go around nicely but I have not got the valve motion on yet so cannot report that it will not foul.
Norm
|
|
Gary L
Elder Statesman
Posts: 1,208
|
Post by Gary L on Oct 24, 2021 10:59:47 GMT
Interesting (to me, anyway - I'm also building one!) that DY shows less end float on the loco axles than the tender. Is this partly because too much float on loco axles means that connecting and coupling rod bearings must be that much slacker? And how much clearance over a "good running fit" should one allow to cope with axle movement? It is often forgotten that in full-size, it was normal for the flanges of centre driving wheels on a 6- or 8-wheeler (ignoring the carrying wheels) to be thinned. This would reduce or eliminate the need for end float, which is very inconvenient on coupled axles. The ten-wheeled 9F had thinned flanges on the second and fourth driving axles, and no flanges at all on the centre axle. Gary
|
|