I haven't so far been able to find the Steve Hall article so far that you refer to, Phil - which MRJ was it do you happen to know?
What you say about the Exactoscale FastTrack I find very interesting - the +0.2mm for whatever radius curve is the same idea as I quoted from the Protofour Manual - it is the maximum gauge widening "allowed" in the P4 standards for the tightest curve, so therefore, I guess (?), much more "lax" than Keith says we really need for say 60mph curves...?
I am trying to get my head round the various issues here still. One problem is the various measurements all being in different units, as in the P4 Digest: ""prototype gauge widening at 10 chains radius is 0.25in, at 7 chains radius is 0.5in, and at 5½ chains radius is 0.75in maximum" , while earlier the maximum GW is given as +0.22 at 528mm radius.
This is put onto graph form in Russ Elliott's "Easing the Traverse through the Switches" on the Clag website, here:http://www.clag.org.uk/switch-traverse.html
Here is the graph, though confusingly we still aren't in metric everywhere, we have chains and inches too. The subject of the article is specifically about the switch of the turnout, obviously a related but separate matter.
My head aches trying to do maths and I may be going wrong now.....?
1 chain = 66 feet.
1 foot = 304.8 mm
4mm scale = 1: 76.2
Translate the prototype chains into our scale and inches and millimetres, I get:
0.25" gauge widening at 10 chains = 660ft = 7920" = [to 4mm scale roughly 104" ]= 2640mm
0.5" gauge widening at 7 chains = 1848mm
0.75" gauge widening at 5.5 chains = 1452mm
Reading from the graph:
0.1mm widening for curves of approx 105" = 2667mm
0.2 widening for curves of approx 63" = 1600mm
So here I am now again surprised, that prototype gauge widening according to this graph equivalent to our +0.2mm would be at near enough 6 chains = scale approx 63" = 1600mm.
Yet our standards specify +0.22 widening at 528mm...I don't get that, why is it so different?
Keith says the increased tolerance has the effect of reducing the need for gauge widening. Now correct me if I am mistaken here, but isn't the tolerance an allowance for inaccuracy in B2B setting and wheel wobble inevitable if the hobby is to be affordable or practical for heathens like me who take fright at the sight of a lathe? - and that that inaccuracy may well take my B2B to nearly the prototype width in some instances. So while some wheelsets may not need prototypical GW, others very nearly may do.
It may not be quite appropriate to quote this from one of Russ's posts without the full context, which is at:viewtopic.php?f=20&t=3749&hilit=B2B+envelope&start=5
Where I came in on this whole subject was regarding making turnouts: not only the question of how much gauge widening is needed in pointwork, but even whether it should be added at all, hence I was looking on the Forum about the whole topic of gauge widening.
An A6 turnout is not outside fairly average P4 modelling is it, in yard areas for example? According to the template information at the top of it, the "closer radius" is 1240mm (or - thanks Google for doing the sum for me! - 49 inches). According to the above graph, that would require at least +0.2mm GW, probably more, prototypically. A four foot radius curve is well inside the average modelling envelope, I think.
From the same information, a B8 has a closer radius of 2452mm, so needing +0.1 at least.
What I wonder is, what the triangular gauge gives for these two examples.
The other thing that is often said is that GW is not needed for short wheelbase trucks, but are such vehicles are less prone to derailment than others on a typical exhibition layout? Of course there are other issues there, and I personally think that in P4 to leave any vehicle without some sort of suspension is potentially a problem. But maybe gauge widening would not help...?
So, going back to my original question and Russ's statement, is he saying that the triangular gauge gives half of the prototype GW, but that this is the Society's standard GW? - I think so. But if so that means that our decision to take prototypical GW over and above the triangular tool might be appropriate at quite a normal size curve...doesn't it? By the way the quote is from:viewtopic.php?f=5&t=2021&hilit=gauge+widening+at+the+crossing
As before, if I am talking rubbish please say!