>
> > Your dimensions below indicate different dimensions between each
> > standard. Other prototypes vary as well. [...]
> > I notice you are mixing 2 standards to broaden your tolerances
> slightly. It is correct to only
> > use one or the other.
>
> I have only mentioned older DSB /Danish State Railways), newer DSB
> and/or UIC where there are differences. Where I mention only UIC values
> they agree with both older and newer DSB values. In general, I have
> based my Proto:45 on DSB 1930 or earlier.
>
> > The prototype clearance scaled down between the minimum wheel back to
> > back and track and check face to check face (UIC scaled to P-45 is
> > only 0.022mm). This is tight and will limit your models to scale > > curves only.
>
> Minimum wheel set back-to-back is 30.15 mm. Absolute maximum check rail
> face to wing rail face is 30.11 mm. This gives an absolute minimum
> clearence of 0.04 mm which is tight.
>
Then you do realise scaling down exact prototype values is impractical.
It is, however, very unlikely that check rail face to wing rail face
> should reach 30.11 mm. The probable maximum can be calculated:
>
> Measurements are expressed as nominal values plus/minus a symmetrical
> tolerance. If total tolerance on the flangeways are 0.10 mm, we have: >
> Track gauge 31.965 +/-0.145
> Check rail flangeway 0.96 +/-0.05
> Crossing flangeway 1.14 +/-0.05
>
> The nominal check rail face to wing rail face is 31.965 - 0,96 - 1.14 =
> 29.865 mm, and the probable symmetrical tolerance is SQR(0.145^2 +
> 0.05^2 + 0.05^2) = 0.161 mm.
>
> Hence the probable maximum check rail face to wing rail face is 29.865 +
> 0,161 = 30.03 mm and the probable minimum clearance is 30.15 - 30.03 = > 0.12 mm.
>
Unfortunately your probable dimensions rely on accurate measurements and gauges made to exact dimensions. Gauges have dimension errors, and therefore depending on the gauge size, your probable track dimensions will vary from your dimensions. Also construction methods will result in a small difference. What is important is the standards limits. In your case the clearance between the minimum wheel back to back and maximum check face to check face will restrict you to scale curves.
> For example using UIC numbers, maximum
> > flange width 33mm, check gauge minimum 1393mm, leaves a back to back
> > maximum of 1360mm. Minimum back to back allowed is 1357mm. This
> > leaves only 3mm total variation for back to back and flange width,
> > that scales down to 0.067mm. If we machine flanges to a tolerance of
> > 0.03mm that leaves a maximum back to back variation of
> > 0.037mm. Not easy to do.
>
> This is one reason that I have set 0.68 +/-0.03 mm for flange width so
> that the maximum is 0.71 mm, a bit narrower than the maximum prototype
> 33 mm (0.73). The absolute maximum check gauge is 30.29 + 0.71 = 31.00 > mm.
>
> The maximum check gauge for the wheel set (31.00 mm) may be a little
> larger than the minimum check gauge for the crossing/check rail (30.91
> mm) in Proto:45 just like on the prototype.
Fortunately many prototypes do not do this. It's OK for slow speed low load track. For small scale models it's unnecessary and results in rougher running if the extreme case exists. Because your standard allows this prototype extreme, it is possible for incompatibility between different modellers using the standard due to small errors in measurement, for example a micrometer is out by 0.003mm. These errors are taken care of if you make the track check gauge limit match the wheel check gauge limit. Belt and braces approach.
I have thoroughly tested that this works in practice, and I feel that
> the reason is the much narrower flanges and check and wing rail
> flangeways in Proto:45. In a 1:7.5 point where the curve continues
> through the crossing and with the check rails removed, I could not
> derail a short-wheelbase wagon with standard wagon wheel sets neither in
> the straight nor the curved road.
>
> NEM standards are not that forgiving, I can tell you.
I know from my experience using HO finescale the same is often achieved. However the HO wheel is smaller, and there is less safety margin. You are correct, the narrow flangeway at the crossing V with scale clearances is what counts, but without a check rail, the flange profile becomes critical. Considering you cannot see the difference between a 0.91mm wide flange way and a 1.19mm flange way, why not simply specify your track standard so the crossing V flange way is the same dimension as the check rail flangeway. Doing this results in only needing one track drawing and less dimensions to define your standard. The dimensions directly transfer to track gauges, no calculations required, and you only need one set of gauges for all types of trackwork. Here are the numbers I suggested before.
Track gauge 32.05mm to 32.1mm Flange way gap 1.1mm to 1.05mm
Minimum clearance between check faces and wheel back to back is 0.15mm, using minimum scale minimum back to back, about double what is proven to work in HO for scale curves. That's it. Decrease the maximum check gauge of your wheels and if you like make the back to back 30.15mm to 30.24, you can still use your existing flange dimensions. Using these dimensions will result in less fine tuning when building track and wheels. You could double your tolerance for track to make it easier to build, though I doubt you would want to do this.
> --
> Best regards
> Erik Olsen
>
formatting link