Accurate Iron Minks

[Quarryscapes]

Maybe a complete drawing won't be necessary, though there are so many issues with the one in All about GWR Iron Minks I'm not sure how much of it will prove reliable. 

 

What I do need however is the correct spacing of the stanchions, and it's also be nice to get confirmation what the actual width and height of an Iron mink body is! 

 

The drawing shows the spacings as 8" + 2' 5 3/8" +   2' 5 3/8" + 5' 3 1/2" +  2' 5 3/8" +  2' 5 3/8" + 8" = 16', but it doesn't!  Substitute the door width with the clear width of 4' 10 1/2" and it does, but then that does't give me the correct spacing of stanchions! 

 

The Heights and widths quoted don't seem to work either. At 7'6" wide inside with 1/8" sheeting gives you external width of 7' 6 1/4". With a roof radius of 7'3" that then gives you a centre height of 6' 9 1/8" without the roof, which doesn't tally with the drawing which shows 6' 8 1/2" with the roof. 

 

Bring the internal width down to the 7'5" it's suggested they were actually built to, and you then get a centre height of 6' 8 5/16" Which is closer but still gives the roof negative thickness! 

 

This of course assumes the height at the eaves really was 5' 8 1/2" ... 

 

Someone please check my maths and tell me I haven't lost the plot! 

[buffalo]

...The drawing shows the spacings as 8" + 2' 5 3/8" +   2' 5 3/8" + 5' 3 1/2" +  2' 5 3/8" +  2' 5 3/8" + 8" = 16', but it doesn't!  Substitute the door width with the clear width of 4' 10 1/2" and it does, but then that does't give me the correct spacing of stanchions! ...

Not sure what you think the problem is here. Clearly the dimension arrow from second stanchion to the door is in the wrong place, so you should be using the "in clear" measurement, not the overall width of the door. Why does this mean the stanchions are wronly spaced? If you look at photos you'll see the panel next to the door is always a little narrower than the outer one -- count the rivets at the base of the panel and you'll see that usually one is masked by the door (the actual number varies between early and later lots). Incidentally, this error is repeated in the drawing in Atkins et al.

 

As to the height, it all depends on where you think the centre of the radius is located. On the drawing, it looks to be a couple of inches below the centre of the coupling backplate. It is certainly not at the centre of the plate. Looking at other GWR van and coach drawings, the actual position of this radius is rarely specified and there is a wide variation in position.

 

Nick

 

ps. the drawing also says "All plates 11BWG", that's a little less than 1/8" but hardly enough to matter in most small scales.

[Quarryscapes]

Not sure what you think the problem is here. Clearly the dimension arrow from second stanchion to the door is in the wrong place, so you should be using the "in clear" measurement, not the overall width of the door. Why does this mean the stanchions are wronly spaced? If you look at photos you'll see the panel next to the door is always a little narrower than the outer one -- count the rivets at the base of the panel and you'll see that usually one is masked by the door (the actual number varies between early and later lots). Incidentally, this error is repeated in the drawing in Atkins et al.

 

As to the height, it all depends on where you think the centre of the radius is located. On the drawing, it looks to be a couple of inches below the centre of the coupling backplate. It is certainly not at the centre of the plate. Looking at other GWR van and coach drawings, the actual position of this radius is rarely specified and there is a wide variation in position.

 

Nick

 

ps. the drawing also says "All plates 11BWG", that's a little less than 1/8" but hardly enough to matter in most small scales.

 

But it isn't clear is it, it could be wrong placement or it could be the wrong dimension to start with, if all I have is this drawing I'm not going to know which it is am I. 

 

The heights would indeed depend on the position of the centre of the radius, but that doesn't affect the problem here, which is that one, or both the values given are in error. Lets assume the distance from the base of the body to the underside of the roof at the centre is actually the 6' 8 1/2" stated. That would mean for a body width of 7'5" inside, or 7'5 1/4" outside we would get heights at the side of 5' 8 3/16". If we assume instead that the centre height is over the roof, which we will assume to also be 1/8" thick then we get side height of 5' 8 13/64" 

 

Basically what I'm getting at is a 7'3" radius curve will not join those 3 points! (it will, but not at the quoted values for width)

 

Since the Ratio kit is constantly criticised for it's width and height issues I need to make damn sure anything I come up with is 100% rivet counter proof. 

 

Whilst we're here, what's the radius of the corner anyway? I've assumed it to be the same as the absorbed ones at 4.5"... 

[buffalo]

...That would mean for a body width of 7'5" inside, or 7'5 1/4" outside we would get heights at the side of 5' 8 3/16". If we assume instead that the centre height is over the roof, which we will assume to also be 1/8" thick then we get side height of 5' 8 13/64"...

Have you considered what the difference between 3/16" and 13/64" is at your scale, or what the original tolerances in cutting the sheets might have been?

 

Nick

[hmrspaul]

Isn't the "Drawing" in Atkins a reproduction from the diagram book (indeed my diagram book if it is what was used for the 1st edition) and not a drawing at all? I know that the GWR diagrams are more accurate than those of any other pre-nat company, but they were never intended to be an ultra accurate reproduction of what the workshops produced. They are diagrams for use by the revenue earning departments.

 

It is not possible to draw, at 4 mm (or even 12mm) to the foot to the accuracy of 1/64th of an inch, because there is a line width that has to be taken into account. Personally, having measured hundreds of wagons, that have been drawn up by friends, I/we rarely measured closer to 1/4 inch. OK we did occasionally go to an 1/8th but we couldn't reproduce this in the 12mm foot drawings. Rivets especially varied considerably in pitch one to another.

 

And then we are talking about a GWR wagon built of a metal. Even the BR versions of Tunny, Ling and Starfish I measured were clearly made by a smith. None of them were built with any accuracy, all of the thicker steel work appeared to have been cut using smithy tools so they were at angles instead of square cut. The plate work also was slightly out of square.

 

Please go and measure an iron mink in conservation if 1/64th worries you so much. It could be life model changing.

 

Paul Bartlett

[Quarryscapes]

Have you considered what the difference between 3/16" and 13/64" is at your scale, or what the original tolerances in cutting the sheets might have been?

 

Nick

 

At scale it's negligible, but full size it matters. The full size sheets won't have had too much tolerance, they certainly aren't going to be interested in wasting material. Without seeing the actual proper drawing for the sheets (not the GA) I couldn't tell you. I'd love to see that drawing, as it's make a model much much easier to make. 

 

Isn't the "Drawing" in Atkins a reproduction from the diagram book (indeed my diagram book if it is what was used for the 1st edition) and not a drawing at all? I know that the GWR diagrams are more accurate than those of any other pre-nat company, but they were never intended to be an ultra accurate reproduction of what the workshops produced. They are diagrams for use by the revenue earning departments.

 

It is not possible to draw, at 4 mm (or even 12mm) to the foot to the accuracy of 1/64th of an inch, because there is a line width that has to be taken into account. Personally, having measured hundreds of wagons, that have been drawn up by friends, I/we rarely measured closer to 1/4 inch. OK we did occasionally go to an 1/8th but we couldn't reproduce this in the 12mm foot drawings. Rivets especially varied considerably in pitch one to another.

 

And then we are talking about a GWR wagon built of a metal. Even the BR versions of Tunny, Ling and Starfish I measured were clearly made by a smith. None of them were built with any accuracy, all of the thicker steel work appeared to have been cut using smithy tools so they were at angles instead of square cut. The plate work also was slightly out of square.

 

Please go and measure an iron mink in conservation if 1/64th worries you so much. It could be life model changing.

 

Paul Bartlett

 

Where did I say I'm bothered about 1/64th of an inch? I'm bothered by the fact that drawing is basically wrong, and since I don't have a mink in my back garden to measure, or access to Swindon's proper drawings, I can't say by how much. 

[Siberian Snooper]

I've been involved in ship repair and I can tell you if it was closer than + or - 2inches to the drawing, even a so called "as fitted drawing" was very accurate. A compartment could be as much as + or - 6inches and zometimes quite a bit more.

The bulkheads on type 21 frigates were supposed to be no more than + or - 3/4inch difference between each ship, I think they were having a laugh, bearing in mind that the burning machines used to cut the plate were somewhat more accurate 30 odd years ago than some bloke with a bit of chalk, a length of steel bar and a gas axe was 100 odd years ago.

 

I would think the basic dimensions for something built then, would be as accurate as the chassis was built and every thing else would be cut and fitted to suit.

 

Having replaced panels on various cars that I had through the '70s, it was amazing how much fettling was required to fit replacement panels to what was more or less jig built '60s car, where they should all be identical.

 

I would go by the maxim if it looks right it is right. If you're exhibiting your stock, I doubt that anyone stood 2foot or so awy is going to run a mic or vernier over it.

[MikeOxon]

I scanned the relevant drawing at high resolution to see exactly where the dimension lines terminate and, from this close examination, I believe that your apparent 'discrepancy' has arisen from the way the drawing was constructed. 

 

The stanchions are placed at a regular horizontal spacing of 2' 5 3/8" between their centre lines.  But there are no stanchions alongside the doors, so this pitch has been brought to a 'virtual line' inside the outside edges of the doors.  The dimension of 4' 10 1/2" represents the distance between these 'virtual lines' and thus the overall length becomes 4 x 2' 5 3/8" + 2 x 8" + 4' 10 1/2", which is exactly 16 feet.  I don't know why the draughtsman did it this way - possibly a convention with regular-panelled vehicles - but he did, and it does add up correctly!

 

I found that the drawings in Atkins were not reproduced quite to scale, and I had to stretch the end view very slightly in a vertical direction to make all the dimensions match the quoted figures.  I also noted, from Atkins, that the first batches from 1886 were only 7' 1" width 'in clear', whereas later batches were 7' 4".   I have shown these dimensions relative to the Ratio 4mm scale model in my post at http://www.rmweb.co.uk/community/index.php?/topic/35398-gwr-iron-mink-questions/?p=1561806

 

Mike

[buffalo]

I've been involved in ship repair and I can tell you if it was closer than + or - 2inches to the drawing, even a so called "as fitted drawing" was very accurate. A compartment could be as much as + or - 6inches and zometimes quite a bit more.

The bulkheads on type 21 frigates were supposed to be no more than + or - 3/4inch difference between each ship, I think they were having a laugh, bearing in mind that the burning machines used to cut the plate were somewhat more accurate 30 odd years ago than some bloke with a bit of chalk, a length of steel bar and a gas axe was 100 odd years ago....

 

A salutory lesson for those many modellers with little understanding of engineering or drawing. Nineteenth century wagon building was much more in the realm of blacksmithing than precision engineering. Though I suspect your 60's car example is closer to the tolerances used and variations seen in the more mass-produced V6s, there were, after all, well over four thousand of them built in less than fifteen years. There is certainly no point in obsessing about sixty-fourths, and I would not be surprised to find variations of an inch or more in overall dimensions across a sample of any vehicle of this period.

 

I scanned the relevant drawing at high resolution to see exactly where the dimension lines terminate and, from this close examination, I believe that your apparent 'discrepancy' has arisen from the way the drawing was constructed. 

 

The stanchions are placed at a regular horizontal spacing of 2' 5 3/8" between their centre lines.  But there are no stanchions alongside the doors, so this pitch has been brought to a 'virtual line' inside the outside edges of the doors.  The dimension of 4' 10 1/2" represents the distance between these 'virtual lines' and thus the overall length becomes 4 x 2' 5 3/8" + 2 x 8" + 4' 10 1/2", which is exactly 16 feet.  I don't know why the draughtsman did it this way - possibly a convention with regular-panelled vehicles - but he did, and it does add up correctly!...

 

I assume by "relevant drawing" you mean drawing No 2 on P18 of the All About Iron Minks book? This is the one referred to in the OP. In which case, I'm not sure what the point of scanning at high resolution was. On the other hand, the Tourret version in Atkins et al. may well benefit from enlargement. The mis-placed dimension arrow is clearly visible on the original printed drawing and the sum of lengths adding to 16' was agreed in the first two posts. As to "virtual lines" I assume you mean the dashed lines representing the edge of the iron sheet at the door opening.

 

Drawings like these need to be treated like any other historical document. With great care. We need to ask who drew them, when, why and what sources they used. Unfortunately, much of this information is not always available. The drawing in "all about" is by  one of the authors M.E.M. Lloyd amd is dated June 1958. It is not clear whether any Swindon drawings were used to create it or whether the dimensions come from a measured survey.

 

Atkins et al. offer several drawings. Fig 351 is from the GWR diagram book, so is after 1905, and only provides the basic overall and "in clear" dimensions as would be expected. These drawings were for people using and identifying the vehicles, not for construction purposes. . Fig 352 is a Tourret drawing of the earlier 16'6" type, dated 25/8/95 and giving only underframe dimensions. Fig 353 is also by Tourret and carries the same date. I'm fairly certain it is derived from Lloyd's earlier drawing as it includes most, though not all, of the dimensions given by Lloyd and includes the same error with the misplaced dimension arrow.

 

Fig 354 is perhaps the most interesting. It is a larger, fully dimensioned drawing and is referred to in the text as an "engineering general arrangement drawing". Text on the drawing suggests that it comes from the Railway Engineer, but no date is given. Such drawings were usually very accurate and were based on information from Swindon. Most importantly, it confirms the 2' 5 3/8" measurement between edge of doorway and centre of stanchion.

 

...I found that the drawings in Atkins were not reproduced quite to scale, and I had to stretch the end view very slightly in a vertical direction to make all the dimensions match the quoted figures.  I also noted, from Atkins, that the first batches from 1886 were only 7' 1" width 'in clear', whereas later batches were 7' 4".   I have shown these dimensions relative to the Ratio 4mm scale model in my post at http://www.rmweb.co.uk/community/index.php?/topic/35398-gwr-iron-mink-questions/?p=1561806

 

Mike

Ah, the modellers' obsession with "scale drawings"! All very well if you have nothing else, but the only measurements you can trust are the dimensions given on a drawing, whether drawn for production purposes or from measured prototypes.

 

As to the width, both the current Atkins et al. and the "all about" volume give the width of pre 1888 minks as the same as later versions. The main difference was that they were 16'6" long with 9'6" wheel base. Perhaps you are confusing them with some of the Rhymney, Taff Vale or Spillers PO versions which were 7'1" inside?

 

Nick

[TheSignalEngineer]

I remember a gang blacksmith who when we requested a special part, say "two foot four and five eights" long, would say "do you want big eights or little eights".

[MikeOxon]

 

I assume by "relevant drawing" you mean drawing No 2 on P18 of the All About Iron Minks book? This is the one referred to in the OP. In which case, I'm not sure what the point of scanning at high resolution was. On the other hand, the Tourret version in Atkins et al. may well benefit from enlargement. The mis-placed dimension arrow is clearly visible on the original printed drawing and the sum of lengths adding to 16' was agreed in the first two posts. As to "virtual lines" I assume you mean the dashed lines representing the edge of the iron sheet at the door opening.

 

............................................

As to the width, both the current Atkins et al. and the "all about" volume give the width of pre 1888 minks as the same as later versions. The main difference was that they were 16'6" long with 9'6" wheel base. Perhaps you are confusing them with some of the Rhymney, Taff Vale or Spillers PO versions which were 7'1" inside?

 

Nick

I apologise for causing confusion - posts above mine in the thread had referred to the Atkins 'drawing' and that is what I looked at and found that I needed to enlarge it, to see where the dimension lines actually ended.  My copy of Atkins is the combined volume dated 1986 and the figure to which I referred is on p.177. 

 

The OP seemed concerned that written dimensions did not seem to add up correctly and I believe that his difficulty arose from the fact that these dimensions were not all given between externally visible features (such as edges of doors) but from the internal structure of the panels.  My term 'virtual lines' was an attempt to indicate this. I don't think the OP was about accuracy of the drawing itself but only about adding up the written figures.

 

My edition of Atkins refers to 7' 1" width for earlier vans but, perhaps, later research may have changed this view.

[corneliuslundie]

As an example of why not to measure drawings and to rely only on the dimensions given therein, I have a copy of a Metropolitan Wagon Company drawing for a Rhymney Railway iron ore and rail wagon from the 1870s. It has the wheelbase given as 9 ft, but this is crossed out and 8ft 9in written in. But the drawing (which scales at just about 9ft) was never changed. For the purpose for which it was intended, it didn't need to be.

 

Jonathan David

[Il Grifone]

I read somewhere that a test for apprentices was to take a inch off thick plating* using a cold chisel (sooner him than me!). From this it would seem that anything more accurate than about 1/2" is irrelevant (since this is less than 0.2mm in 00, it's rather beyond my skills anyway!).

 

Normal drawing room practice was to make the working drawings on paper and then later (or much later) to trace this onto linen for archiving. Obviously a source of errors which might or might not have been corrected, even if discovered.

 

Management might have been concerned about waste, but it is doubtful whether the 'man on the job' was. Above all with with 12 hour working days for a few shillings a week.

 

*Loco main frames IIRC

 

Weight diagrams are notorious for being inaccurate. I believe I'm right in saying that the K's autocoach has an extra length of beading through following one.

[Quarryscapes]

As an example of why not to measure drawings and to rely only on the dimensions given therein, I have a copy of a Metropolitan Wagon Company drawing for a Rhymney Railway iron ore and rail wagon from the 1870s. It has the wheelbase given as 9 ft, but this is crossed out and 8ft 9in written in. But the drawing (which scales at just about 9ft) was never changed. For the purpose for which it was intended, it didn't need to be.

 

Jonathan David

 

Precisely so, and in the case of the drawing I'm on about, it's the dimensions that are wrong! 

 

Someone above mentioned 7'4" inside - where did that figure come from? 

[buffalo]

...Someone above mentioned 7'4" inside - where did that figure come from?

As with the 7'1" dimension, it comes from the 1986 and perhaps earlier editions of Atkins et al. It is not mentioned in the later editions. The much more detailed drawing from the Railway Engineer that I mentioned earlier appears in both editions and has 7'4³/₄". I'd suggest you may have more luck with this drawing than the one in "all about" for the reasons discussed earlier.

 

Nick

[Quarryscapes]

As with the 7'1" dimension, it comes from the 1986 and perhaps earlier editions of Atkins et al. It is not mentioned in the later editions. The much more detailed drawing from the Railway Engineer that I mentioned earlier appears in both editions and has 7'4³/₄". I'd suggest you may have more luck with this drawing than the one in "all about" for the reasons discussed earlier.

 

Nick

 

Then I shall hunt it out. Thanks.