Buffer locking

[MikeWilliams]

Ever since reading our Chairman's article in the latest Newsletter I can't stop thinking about buffer locking.  In 7mm I used to find there was no problem on even reverse curves over about 6ft radius, so do other people suffer from it in Gauge 3?

Maybe somebody more clever than me could tell us the theoretical radius of a reverse curve at which it would happen?  Carriages have larger or oval buffers, so taking something like a 16ft or 18ft wagon on 9ft or 10ft wheelbase and with 1mm sideplay on each axle (i.e. 0.5mm each side) and scale 13in heads, what should be our minimum curve radius?  The mathematics is beyond me, but I'd be interested in the result!

Mike

[cabbage]

Mike,

The maths is not beyond you -you are just rusty!!! This is a variation on the "planks cladding a pillar problem" from "O" level schoolboy maths. Give me the width of the wagon, the length over buffers and the wheelbase. I hated Calculus but I aced "Trig"!!! 

regards

ralph

[MikeWilliams]

Ralph - I thought you might rise to the bait!  Yes, I enjoyed maths at school, but very rusty since.

If we say wheelbase 9ft, length over buffers 20ft, buffer heads 13in diameter, all at 13.5mm/1ft.  The 0.5mm sideplay either side on each axle will be a factor, as will the play between wheel and rail on G3 standards which I think is considerable and a factor also in Kelvin's question elsewhere about gauge widening.

Width of the wagon?  Don't quite see the relevance but say 8ft and distance between buffers about 5ft 8in - and shall we assume all vehicles are the same?

I know this is theory, but its a starting point.  I guess also the discrepancy between buffer heights of adjacent vehicles would have a big difference, but should perhaps be ignored at this stage?

Mike

[AllWight]

The issue of buffer locking happened on the full size wagons as well so it is not just restricted to those who model the smaller scales. The Americans had the right idea by using a central coupling/buffer assembly in the form of a buckeye.

Mark

[cabbage]

The answer is 1695mm radius using non GW track.

Taking the distance from the outside edge of the buffer to the inside edge of the rail as "O" then "A" = half the length of the wagon. Then Tan (?) = O/A = 4.6 degrees. Dividing 4.6 into 360 gives a (rough) 78 sided polygon. a2 +b2 =c2 gives the length of the side of the polygon 136.6mm . Thus 78 x 136.6 / 2 / TT = 1695.

I told you that you were simply rusty!!!!

regards

ralph

[MikeWilliams]

Thanks Ralph - I knew I could rely on you.  So that's about 5ft 6in in old money which is very tight indeed for Gauge 3.  Even allowing for extra slop in some vehicles I suggest in that case buffer locking shouldn't be a big problem for most of us?

And as for the practical reality do people suffer this in our scale - Mark?  Dave? ...  and if so on what radius?

Ted - your article prompted my thoughts on this - what radius caused the problem for you?

Mike

[Andy B]

Ralph,

Your calc is for a simple curve, I believe?
A quick lunchtime dabble in CAD shows something of what happens on a reverse curve.

The model is set up with:
Curve centreline radius 1700mm
Reverse curve offset on centrelines 152mm
Wagon Wheelbase 121.5mm (9ft prototype)
Length over buffers (20ft prototype)
Buffer heads 13.5mm (12" prototype)
I've allowed 1mm - opposite offset from centreline on both axles. This is probably optimistic, as half the (gauge - over-flange-width) = 0.85mm on nominal. Add in axle end-float and tolerances, and you're likely nearer 1.5mm.

This gives the result that the buffer edges only have 0.7mm of overlap!

It really wants checking on a series of scenarios - for the same overall length, a shorter wheelbase gives less impact but the end-throw increases, so making it worse again.
It also takes no account of the vertical curvature, track twist and variance in buffer height between wagons. (this is a 3D problem!)
Also need to condider the shape of buffer heads, off-centre buffers (my Slaters covered goods were visiblyoff-centre), and other building tolerances. Plus what happens with sprung vs unsprung buffers, or sprung buffers where the spring rates are very different on, or stick-slip effect of buffers in their housing caused by the angular contact forces?

Andy

[Andy B]

Just to reinforce my point on vertical alignment:
If the buffer centres on 2 touching buffers are 2mm out of line vertically, then their effective contact width is reduced from 13.5mm to 12.9mm - that's our slight overlap above used up.

2mm vertical misalignment is easily achieved due to different running heights and / or track vertical aligment and twist.

So next time anyone experiences a buffer-locking issue for real, can we have some forensic evidence, please, so that we can evaluate the contributing factors?

Andy

[AllWight]

Well on blackgang I am on GRS 8' radius points. While my engines are fine to shunt, the GRS Prairie will buffer lock against other propelled wagons through the crossover. Given hind sight I wish I had used 10' radius points but I am not going to change it now. On my 00 layout bembridge the wagons are mostly of a similar overall length but I took the precaution of fitting the same slightly oversized buffers to all the wagons. I now do not suffer with buffer lock on the small layout.

Mark

[cabbage]

Andy,

Yes I used a simple curve! I also did it on my slide rule which was my 16th Birthday present. My father would have worked it all out in his head, faster, and to a far greater degree of accuracy....

regards

ralph

[MikeWilliams]

can we have some forensic evidence, please, so that we can evaluate the contributing factors?

I like your style Andy!!

Mike

[blagdon]

Speaking from my own experience, at Chew Magna I can propel my GRS autotrailer with the GWR 2021 (0-6-0T) through a pair of 14' 9'' turnouts opposite each other. I have not tried the slightly longer GWR 8750 as they were not auto-fitted! Oh, and the main problem is that with the long buffers on the autotrailer the screw coupling has to opened out to it's maximum length.

Ian the Gauge '3' Pirate

[454]

All the theoretical measurements in the world are all well & good.
Spent a lifetime doing it so I know. But in retirement I suck it & see. Life is exciting that way. I don't bring work home, I play trains.

However, there are circumstances during general running, not necessary propelling, when buffer locking occurs.

My own experience is during forward running at higher speeds. Upon slowing, the train tries to catch up with the loco.
Result is a vehicle rides up onto the buffers of the vehicle in front of it. This can sometimes be un-noticed until the train hits the first curve. Then calamity. Likely as our models do not have brakes.

Other experience is it does not matter how well one makes a kit, how much springing & travel is incorporated into a model rolling stock vehicle, after many hours of use, things happen.  The coupling may jam up & not stretch as it should. The buffers become sticky on their spindles or completely seize up. I know that when all this happens it is down to oneself & general maintenance, but how many of us get home from a GTG a bit tired. Or from a weekend at an exhibition & say to ourselves, Oh I'll leave that freeing up of the buffers & a bit of oiling to later on. The wagons stay in their boxes & don't see the light of day until the next run. Sadly slightly damp & the steel buffer spindle mildly oxidizing in the humidity of a closed environment.

My worst performers are poorly maintained long suffering vehicles.

My indoor line is 8 feet point rad. Buffer locking can be avoided once the performance of various vehicle combinations has been experienced. Then operating circumstances (probably just like the real thing) take over & the vehicles are run in a manner which will avoid buffer locking. With the exception of the following circumstance.

Like on an exhibition layout, when someone brings along a visiting loco with a bogie auto coach, you say to the person run it conventionally & not in auto mode, i.e. run around at the terminus & they ignore you then the consequences are obvious. It is an "I told you so" situation.

Match the vehicles to the track. OK model the vehicles that interest you, we all do that. Then run them with consideration of what might happen then find ways of avoiding it.

Just increasing the point radius from 8 feet to 10 feet or 12 feet is not a solution.

I run my Brit through 8 feet radius points! OK laugh. Solution it drives on the straight bit of the point then reverses back. Never do I send around the tight rad.

My last words on the matter.
After all BR did not run Brits on the Culm Valley Branch.
Nor did they run auto coaches on it. In the latter days of passenger operation the 14XX tank loco hauled a non WR coach & did the run around at the terminus. They would have auto-coached it if they could, but didn't. (or does someone know differently)

It is horses for courses.

Dave
454

[John Candy]

As Dave has mentioned, buffer-lock can happen during braking in normal operation.

I remember reading of a test carried out by BR in the late 60's or early 70's, following a lot of puzzling goods train derailments.

They used a diesel loco (from memory think it was a class 47) plus a non-fitted train of 16t BR standard minerals and vans.

The test took place near Cheltenham on a section of the already closed GWR line via Cheltenham Racecourse station.

The test showed that the momentum of the unbraked train forced the buffers of the leading wagons to override and buffer-lock, if the loco slowed too rapidly (had not been a problem with the relatively weak braking on steam locos but was a problem with the powerful air brakes on modern locos.

John.

[Peaky 556]

Re the over-riding under braking, we can avoid this using either big flywheels on our motors (especially if using worm drives) or lots of weight in the loco to keep the deceleration gentle (lead-acid cells are good!!), or just use steam!
Tim