A question about 'Finescale OO'

[martin_wynne]

9 hours ago, Andy Reichert said:

They just switched to using standard engineering practice of writing tolerances from a preferred nominal value.

 

Hi Andy,

 

That's the whole point. There isn't a nominal value for track and wheel dimensions.

 

Each quoted dimension is either a maximum. Or a minimum. The tolerance arises from the interdependence of two separate dimensions.

 

So for example check gauge is minimum. There isn't a maximum. It mustn't be less than the specified dimension, but it can exceed that by any amount. That doesn't mean the check rail can be anywhere you like, because the check span dimension is a maximum. There isn't a minimum for that one.

 

Giving everything a nominal value and a tolerance either way means that everything is triple-dimensioned. That definitely isn't "standard engineering practice".

 

cheers,

 

Martin. (former toolmaker).

[Bernard Lamb]

13 hours ago, Andy Reichert said:

 

The NMRA Standard HO dimensions and tolerance ranges (max-min) didn't change. They just switched to using standard engineering practice of writing tolerances from a preferred nominal value.   

It is most definitely not standard engineering practice to express tolerances from a nominal value.

Let us take a simply case. A window frame and a piece of glass.

One must have an all plus tolerance while the other must have an all minus tolerance.

Any thing other than that and it might not work.

Bernard

 

 

[Enterprisingwestern]

1 hour ago, Bernard Lamb said:

It is most definitely not standard engineering practice to express tolerances from a nominal value.

Let us take a simply case. A window frame and a piece of glass.

One must have an all plus tolerance while the other must have an all minus tolerance.

Any thing other than that and it might not work.

Bernard

 

 

 

I personally wouldn't call window frames and glass "engineering", just measurements.

 

Mike.

[hayfield]

11 hours ago, Torper said:

Finescale OO is an OO layout that you think is EM when you see it.

 

DT

 

 

I totally agree many of the finest 00 gauge layouts can be quoted as finescale, its an attitude towards a certain style of modelling/ modelmaking, nothing to do with either scale and or gauge

 

Edited December 13, 2019 by hayfield

[Il Grifone]

It's just that the tolerance is zero in one direction.

 

 

[34theletterbetweenB&D]

4 hours ago, Bernard Lamb said:

It is most definitely not standard engineering practice to express tolerances from a nominal value...

After longer in metrology than most folks, I would not dare to venture such a statement. In my experience every applied technology has its own 'custom and practise' and that directly effects how dimensions are expressed by the practitioners. What is standard in one industry is not so in another.

[PenrithBeacon]

6 hours ago, Bernard Lamb said:

It is most definitely not standard engineering practice to express tolerances from a nominal value.

...

Bernard

 

 

Well it certainly used to be the case but times change and different firms have different ways of doing things

Cheers

[Andy Reichert]

When  you are manufacturing something with a modern tool that is ten times more precise than the tolerance asked for, you have to position the tool somewhere. If that position isn't defined otherwise, each production batch, even from the same company,  can end up quite different.

 

That's basically what has happened with the track gauge,  gauge widening and wheel back-back.

 

Andy

[martin_wynne]

35 minutes ago, Andy Reichert said:

When  you are manufacturing something with a modern tool that is ten times more precise than the tolerance asked for, you have to position the tool somewhere.

 

Yes Andy. But that is not the standard.

 

The standard sets the limits on the finished work, it's not a recipe for the manufacturer. In setting a tool, you set it at the position which produces the desired result. You don't look at someone else's idea of where that should be.

 

The NMRA standards have mixed up the finished track standard with the manufacturing process. I very much doubt if any manufacturer would dream of looking at the NMRA docs in setting up their equipment. I wouldn't.

 

cheers,

 

Martin.

[Andy Reichert]

1 hour ago, martin_wynne said:

 

Yes Andy. But that is not the standard.

 

The standard sets the limits on the finished work, it's not a recipe for the manufacturer. In setting a tool, you set it at the position which produces the desired result. You don't look at someone else's idea of where that should be.

 

The NMRA standards have mixed up the finished track standard with the manufacturing process. I very much doubt if any manufacturer would dream of looking at the NMRA docs in setting up their equipment. I wouldn't.

 

cheers,

 

Martin.

 

From PECO's website . . .

 

 

Andy

[martin_wynne]

38 minutes ago, Andy Reichert said:

From PECO's website . . .

 

Andy, you missed my point. I didn't mean no UK manufacturer would make NMRA-compliant products.

 

I meant that Peco's CNC machine operator does not set his tool positions by reference to the NMRA docs. He has his own production drawings and methods. Having produced a component he tests it against the NMRA standard, and adjusts the machine settings accordingly. The triple-dimensioning in the standard is completely unnecessary and confusing.

 

I'm giving up on this argument because it has gone way beyond what the OP was asking about.

 

Martin.

Edited December 13, 2019 by martin_wynne

[Andy Reichert]

23 hours ago, martin_wynne said:

 

Hi Andy,

 

No reason to be concerned. No-one running British 00 models is much bothered whether they comply with the current NMRA H0 standard, or not.

 

NMRA changed their system of dimensioning about 15 years ago. British 00 still complies with the old dimensions -- as I suspect still does a lot of current H0 production.

 

I'm pleased about that, because it all works fine. Whereas the newer "targeted" system of dimensioning is flawed in my view, and it is easy to find anomalies in the published dimensions.

 

cheers,

 

Martin.

 

You haven't substantiated this claim either yet. PECO doesn't seem to have found it wanting either.

 

Andy

[spamcan61]

4 hours ago, PenrithBeacon said:

Well it certainly used to be the case but times change and different firms have different ways of doing things

Cheers

Speaking as somebody who occasionally creates / reviews drawings for machined parts which are manufactured in UK/USA/Far East then tolerancing from a nominal value is the norm for us. Can't recall what specific ISO we quote for the tolerancing.

[martin_wynne]

39 minutes ago, spamcan61 said:

Speaking as somebody who occasionally creates / reviews drawings for machined parts which are manufactured in UK/USA/Far East then tolerancing from a nominal value is the norm for us. Can't recall what specific ISO we quote for the tolerancing.

 

But that's a drawing, not a standard.

 

Martin.

[Andy Reichert]

 

?

 

Andy

[martin_wynne]

That's a drawing, not a standard. A standard has MAX or MIN (or sometimes both), but not NOMINAL dimensions.

 

A standard must create a boolean result. You cannot test something about whether it complies with a nominal dimension.

 

For example, that drawing shows the check gauge as 15.3mm +/-0.05mm.

 

That is three dimensions. 15.25mm. 15.30mm. 15.35mm. If you measure some track, you can say only whether it complies with two of them. More than the first one, yes or no. Less than the third one, yes or no. But it is impossible to say whether it complies with the 2nd one.  You can't have a dimension in a standard which is impossible to test.

 

Martin.

Edited December 13, 2019 by martin_wynne

[Compound2632]

The drawing implies that the check gauge must lie in the range 15.25 mm - 15.35 mm; the centre of that range, 15.30 mm, is a redundant piece of information. It is therefore more economical to state the more critical of the two limits, which in this case would be the maximum value, 15.35 mm, and the maximum permitted deviation from this, -0.10 mm.

 

The track gauge is specified in this way: minimum 16.5 mm, with a maximum permitted deviation of +0.3 mm (to allow for gauge widening on curves). 

 

The flangeway clearance and distance between check rails are redundant information and may in fact over-constrain the specification. Only four numbers are required for the complete specification: the minimum track gauge and its permitted deviation; and the maximum check gauge and its permitted deviation.

 

That's my interpretation as a physicist and former metrologist; I'm not an engineer and was only ever an amateur draughtsman.

Edited December 13, 2019 by Compound2632

[martin_wynne]

1 hour ago, Compound2632 said:

The drawing implies that the check gauge must lie in the range 15.25 mm - 15.35 mm; the centre of that range, 15.30 mm, is a redundant piece of information. It is therefore more economical to state the more critical of the two limits, which in this case would be the maximum value, 15.35 mm, and the maximum permitted deviation from this, -0.10 mm.

 

Hi,

 

It is the minimum track check gauge which is the critical dimension, 15.25mm here for DOGA Intermediate.

 

No other dimension is needed for the check gauge, because the maximum position for the check rail is governed by the check span, where the maximum 14.2mm is likewise the only dimension required for that one.

 

As you say, the drawing contains a lot of redundant information, some of which is impossible to measure on a finished piece of track.

 

cheers,

 

Martin.

[Andy Reichert]

It's labeled a standard. It is the only document provided regarding the Intermediate Track standard for track. The drawing is obviously provided to anchor the position of the critical dimensions to the lay modellers. All drawings contain redundant information if you look at them long enough   But the rest of the drawing isn't dimensioned, or stated to be to scale. So why discuss it?

 

If you have discovered a dimension conflict in the NMRA standards, it's important that you publish it here ASAP. Providing accurate and complete information is what you said you wanted to do here.

 

Andy

 

 

[The Bigbee Line]

I try and keep out of these debates....   but....

 

Working on the real railways things are manufactured to do a job.

 

In most cases, If a drawing is produced the manufacturer will go to great lengths to make sure there is enough information missing to prevent the component being manufactured by a competitor...  Sound crazy, but there it is.

 

If you are lucky enough to have all the dimensions and tolerances, you then need to know what it's made of.... Another nightmare.....

 

 

THEN....

 

 

When you've made the component and fitted it, the next stage is to know when it is worn out...

 

So you need a dimension at which the component needs to be replaced..

 

You have 3 dimensions:  New, permitted limit in service permitted limit at overhaul (part worn but ok to remain in service.

 

 

I deal a lot with wheelsets, so find the modelling world quite amusing...  One particular wheelset could be between 923mm - 860mm diameter.   We limit to 13mm difference in a bogie and 24mm over the whole wagon.  Flange limit is machined to 32.5mm an can run to 24mm, With numerous other dimensions to check.

 

The roller bearings also have age criteria, Same with the grease...

 

On top of that there is a whole range of non destructive testing that is carried out at key stages....

 

That is only the wheel.  The other part of the interface, the rail has a similar plethora of standards...

 

Finally there is the dynamic properties to consider, my wheelset might be loaded to 22.5 tonnes axle load running at 75 mph,  Everything has to work....

[Andy Reichert]

17 hours ago, martin_wynne said:

That's a drawing, not a standard. A standard has MAX or MIN (or sometimes both), but not NOMINAL dimensions.

 

A standard must create a boolean result. You cannot test something about whether it complies with a nominal dimension.

 

For example, that drawing shows the check gauge as 15.3mm +/-0.05mm.

 

That is three dimensions. 15.25mm. 15.30mm. 15.35mm. If you measure some track, you can say only whether it complies with two of them. More than the first one, yes or no. Less than the third one, yes or no. But it is impossible to say whether it complies with the 2nd one.  You can't have a dimension in a standard which is impossible to test.

 

Martin.

 

From my formal Engineering classes, and my five year engineering apprenticeship at ICL, Letchworth, which included a fair amount of  designing, drawing, prototype and production line machining, x +/- y defines a continuous RANGE, from x-y continuously through to x+y, which any value within is valid and acceptable. That also matches the manufacturing problem of setting a cutter, or a mould edge, or whatever, to an ideal or nominal place in between, to end up with a part that always has the finished dimension within the x +/- y range.

 

To suggest that PECO has only two places (gauge max and gauge min) to validly measure and position the stock rails on a turnout, because the Range is specified as an x +/- value, is not logical.  You cannot place or test to a boolean value, because it implies infinite accuracy.  That's why the ideal position is referred to as "nominal".  Just as when you go to purchase a nominal 5mm nut and bolt. You don't ask for a 5mm +0/-0.01 bolt and  a 5mm +0.01/-0 nut.  And it would be impossible to get either to those the stated absolutely accurate min and max limits.

 

Andy

[martin_wynne]

8 minutes ago, Andy Reichert said:

Just as when you go to purchase a nominal 5mm nut and bolt. You don't ask for a 5mm +0/-0.01 bolt and  a 5mm +0.01/-0 nut.

 

You ask for a bolt which complies with the standard for a 5mm bolt. In other words, one in which the diameter is not greater than the standard specifies as the maximum. And is not less than the standard specifies as the minimum.

 

What diameter is regarded as the "ideal", or what "nominal" diameter the manufacture specifies on his production drawing, is irrelevant, unknown, and unknowable, from the bolt in your hand.

 

Martin.

 

 

 

 

[Ian J.]

1 hour ago, Andy Reichert said:

...

To suggest that PECO has only two places (gauge max and gauge min) to validly measure and position the stock rails on a turnout, because the Range is specified as an x +/- value, is not logical.

...

 

But that's not what's being measured. The two measurements are of different things, but related and overlapping. The check gauge has a minimum, with a plus tolerance, while the check span has a maximum, with a minus tolerance. The two work together and must be considered together to produce the final plus/minus tolerance.

[Andy Reichert]

3 minutes ago, Ian J. said:

 

But that's not what's being measured. The two measurements are of different things, but related and overlapping. The check gauge has a minimum, with a plus tolerance, while the check span has a maximum, with a minus tolerance. The two work together and must be considered together to produce the final plus/minus tolerance.

 

I've never mentioned the check gauge. But I'm happy to discuss issues arising from shrinking it later.

 

I'm saying that there is no difference from a standards setting point of view between (say)  10 mm +2/-0 and `11 mm +1/-1.   If the standards authority wants to define it that way, it's their right and the interpretation is allowed to be the same either way. But whether it's referred to as nominal 10 mm or nominal 11mm  is a language matter.

 

Something that seems to have been overlooked is that the NMRA Standards are intended for manufacturers. Not hand layers. The whole point was to make different manufactures products inter-operate and prevent the free for all of multiple and changing standards that has occurred in the UK. 

 

Which is why the OP asked the question.

 

Andy

 

 

[martin_wynne]

27 minutes ago, Ian J. said:

But that's not what's being measured. The two measurements are of different things, but related and overlapping. The check gauge has a minimum, with a plus tolerance, while the check span has a maximum, with a minus tolerance.

 

Hi Ian,

 

My red.

 

The check gauge does not have a tolerance. The specified dimension is a minimum. No other information is needed.

 

Likewise the check span does not have a tolerance. The specified dimension is a maximum. No other information is needed.

 

Suppose you are fixing a check rail. You can move it inwards until the minimum check gauge is infringed. Or you can move it outwards until the maximum check span is infringed. Those two dimensions are sufficient to control its position and no others are needed.

 

If you were to continue to move it outwards, in order not to exceed the maximum check span you would need to move the wing rail outwards also. That would widen the crossing flangeway, which has a maximum limit, preventing such movement.

 

All the rail positions are interdependent, and don't need double or triple dimensioning to define their positions. Each dimension is either a maximum or a minimum, and nothing else is needed.

 

The specified track gauge is also a minimum, so that it can go wider on sharp curves where needed.

 

cheers,

 

Martin.