What has happened up to this point, and as revealed by submissions made to our Journal, are ad-hoc efforts to deal with one or another isolated aspect of the whole subject of Proto:87. However, and notwithstanding my blurb on the equalizing truck from Red Caboose on Page 8, before we worry about standard axle length, or which couplers would be most appropriate, or most certainly before we suggest changes in the standard for back-to-back dimensions of a wheelset, shouldn't we instead regroup and begin where the MRSG began _ by studying the relationship between wheel profile and rail?
I'm pretty sure most know that the P4 standard (hence P87 also) varies from "dead scale" in order (i) to accommodate dimensional tolerances for fabrication and assembly that were deemed necessary to permit the commercial production of components at a reasonable price, and (ii) to at least bow in the direction of accommodating the sharper than prototype curves typically found on model railroads. How many of us know which dimensions were "eased" and by how much? How much of the variation from "dead scale" is attributable to which of the above cited reasons [See Pages 4 and 10 - Ed.]? What assumptions regarding machining techniques and tooling wear underlie the tolerances assigned to the fabrication and assembly category? Are these assumptions still valid some 20 or more years later?
Since we must assume that the MRSG would not have built in any greater variations from "dead scale" than were felt to be absolutely necessary, and since the derivative P87 standard is, for the most part, simply a scaled down version of P4, has the MRSG left us with proportionately tighter tolerances which might prove prohibitively expensive for a manufacturer to maintain? If not, has the proportional reduction from 4mm scale to 3.5mm scale resulted in the entire tolerance being consumed by the fabrication and assembly process, with none left over for accommodating sharper than prototype curves?
There is another aspect of this subject of tolerances that can create mischief. It is that the prototype itself assigned permissible ranges to certain dimensions. For example, "The standard distance between the backs of flanges of car, engine truck, driver and tender wheels at base line of tread shall be not less than 4 ft. 5 3/32 in., nor more than 4 ft. 5 1/2 in." [Car Builders Cyc, 1940] When you pencil those numbers out to 3.5mm scale equivalent, you get a difference of 0.12mm (0.0047in)! Do we, as [Reichert 96] suggests, really want to stipulate a "tolerance" range for back-to-back of 0.03mm when the prototype routinely allowed a variation 4 times greater? Does everyone understand why the prototype could tolerate (indeed, insist upon) this magnitude of variation?
Enough, for now. I hope I've made my point. I'm sure I know the answers to a few of the questions I've just posed. I think I know the answers to a few others. And, I know I don't know the answers to most of them, nor to a whole raft of others which have occurred to me, but aren't set forth here. However, I'm bound and determined that, before long, I'm going to learn the answers to all of them.
Our search for answers need not be lengthy or arduous. After all, we have the benefit of the detailed and careful work already accomplished by the MRSG, and we have Alan Gibson's wheels and Gourley's cobbled together diesels to use in our experiments. Indeed, I fully expect that when the journey of research and discovery is over, I will find myself in favor of recommending the MRSG dimensions without alteration (Well, except for that darn 16.50mm track gauge instead of the spot-on figure of 16.48mm. - Yes, a change would affect other dimensions - But, it bugs me.) Adoption of the MRSG dimensions would be good because it would mean the existence of a single standard of truly international scope. That, in turn, would have to be impressive to those whom we hope will provide commercial support. The difference of course, between then and now, is that by then we will all thoroughly understand the tradeoffs it took to get there. We will have confidence in our recommendation because we will know why it works!
For my next contribution to our Journal, I intend to use Autodesk's AutoCAD LT to make drawings of the wheel and rail sections specified in the P4 standard, as well as wheel and rail sections typical of North American practice, circa 1940, and show each combination in a way which will illustrate the effective flange dimension. I'll also try to throw in a drawing of the gage actually used by the prototype to set wheel spacing when pressing wheels onto an axle, together with the instructions on how the gage was used. This will be a tall order for me because I'm a neophyte at using CAD. I can draw the stuff alright, though sometimes it's a push, but it's merry hell making the dimensions appear the way I want.
In the meantime, haven't you learned something that the rest of us might like to know? Won't you please fire up the wordprocessor, or whatever, and join in the dialog. Ren will accept just about anything, assuming it isn't written or printed on something that smells bad. Even a question for which you have no answer is still a contribution. And, above all, don't be bashful! Every question, no matter how basic, is still legitimate until we all know and understand the answer.
