How Good Are Drawings ?

[relaxinghobby]

This year my resolution as far as modelling is concerned is to make models from drawings because they will be more accurate than my usual stick odd bits from different kits together and hope the result are true and railway like. Does not always turn out so.

 

Often the drawings are not reliable for instance, the small cattle truck from the LNWR society website. Many dimensions are given but it is slightly rotated and not square on the page and it does not conform to it's own printed dimension arrows.

 

So before modelling can proceed the hopefully accurate modeller must sort out and correct the drawing as best as he can.

 

Computer Aided Design methods offer a good way of going about this and at the moment my favourite computer drawing tool is Inkscape, an open-source version of Corel Draw I believe.

 

It is all new to me, so I am slowly understanding how it does things and learning to manipulate images on it.

 

For example this work in preogress.

 

 

So far the weight diagram is scanned, rotated and reduced in size, the colours represent my tracing and Iam trying to reconsiliate the size of the ends and size.

The roof is too high which is right the side view of the end view?

Edited April 2, 2018 by relaxinghobby

[Londontram]

I to draw on Inkscape often after impoting drawings from a variety of sources and like you frequently find the drawing and known measurement to be out against the drawing and can take an awful lot of adjusting to get right.

 

I got caught out a few times when drawing for the silhouette cutter but now check drawings more carefully which means l can quickly establish the correct proportions befor going for a cut of parts.

The morel is dont just asume but check.

[Steamport Southport]

It all depends on what the drawing is and it's purpose.

 

I doubt that weight diagrams are going to be accurate enough to build anything. They were drawn so that the operating department knew whether a wagon, coach or loco was suitable for operating on certain lines where there were restrictions in height and weight.

 

Many of them probably weren't to scale as they didn't need to be.

 

 

 

Jason

[Steve_Beattie]

Many on-line drawings are scanned in and the X - Y dimensions are not accurate. Often the two elevations end up being of different scale. 

 

Unless you have the original works drawings you can never reproduce to absolute accuracy. 

[Spitfire2865]

Dont trust any drawing thats not for works use.  Line drawings are notoriously iffy as has been posted above.  However proper professionally drafted drawings should be taken as true unless they have been reproduced by photograph rather than flatbed scanning.  Ive turned down offered drawings in the past due to them being photographed.  No use in it for me if I cant scale off of it.  

[Dagworth]

Treat the PDH drawings as rough guides and not as fact.

 

http://www.rmweb.co.uk/community/index.php?/topic/40228-excellent-loco-drawings/

 

Andi

Edited April 2, 2018 by Dagworth

[Ravenser]

This year my resolution as far as modelling is concerned is to make models from drawings because they will be more accurate than my usual stick odd bits from different kits together and hope the result are true and railway like. Does not always turn out so.

 

Often the drawings are not reliable for instance, the small cattle truck from the LNWR society website. Many dimensions are given but it is slightly rotated and not square on the page and it does not conform to it's own printed dimension arrows.

 

So before modelling can proceed the hopefully accurate modeller must sort out and correct the drawing as best as he can.

 

Computer Aided Design methods offer a good way of going about this and at the moment my favourite computer drawing tool is Inkscape, an open-source version of Corel Draw I believe.

 

It is all new to me, so I am slowly understanding how it does things and learning to manipulate images on it.

 

For example this work in preogress.

 

CattleSmallLNWR.png

 

So far the weight diagram is scanned, rotated and reduced in size, the colours represent my tracing and Iam trying to reconsiliate the size of the ends and size.

The roof is too high which is right the side view of the end view?

 

As a fundamental principle, trust the quoted dimensions rather than the drawing.

 

In your case , remember there is a reason that these things are called "diagrams" not "works drawings" - you may well need to redraw the diagram as a scale drawing using the quoted dimensions

[Compound2632]

As a fundamental principle, trust the quoted dimensions rather than the drawing.

 

In your case , remember there is a reason that these things are called "diagrams" not "works drawings" - you may well need to redraw the diagram as a scale drawing using the quoted dimensions

 

... and what was drawn was what was intended to be built. Photographs...

[Bob Reid]

Treat the PDH drawings as rough guides and not as fact.

 

http://www.rmweb.co.uk/community/index.php?/topic/40228-excellent-loco-drawings/

 

Andi

Yes Andy, very rough - completely R.A.F.

[Spitfire2865]

... and what was drawn was what was intended to be built. Photographs...

And be sure you watch for revisions. I have a works drawing for an L&Y tintab that was drawn before increasing the height 3" but the dimensions were changed to add those 3".

Took me over an hour to figure out why my calculations werent matching up.

[Horsetan]

Treat the PDH drawings as rough guides figments of imagination and not as fact.

 

FTFY

[Compound2632]

Going back to relaxinghobby's Diagram 20 small cattle wagon, the drawing and photo on the LNWR Society website also appear in LNWR Wagons Vol. 1 (Wild Swan, 2001), although without the attribution to The Model Engineer and Electrician. The only other illustration is the actual diagram from the Diagram Book. This gives less visual detail but more dimensions. This volume contains plenty of works drawings of later designs of cattle wagon, from which details of this early design can be inferred. See also airnimal's Scale 7 build, though his is an even earlier version with dumb buffers at one end.

 

Wagon books of the standard of the LNWR ones are always worth buying.

Edited April 3, 2018 by Compound2632

[Guy Rixon]

In general, a scan of a GA drawing from the works won't be right for making cutting templates and won't let you scale off dimensions accurately. By the time that a drawing on linen has been stored for 100+ years and then dragged through a scanner it can have all sorts of unequal shrinkage that distorts things. These drawings are not designed to be scaled that way.

 

What a GA does tell is how the vehicle fits together from 3D pieces, and this allow one to infer missing dimensions. In particular, it will show where wagon carpentry diverges from conventional patterns.

 

As an example of mutant carpentry, consider how the solebars, headstocks, floor planks and curb rails fit together in an open wagon. Typically, the solebars are less tall than the headstocks by the thickness of the floor planks. The floor lies directly on the top of the solebars (and middle bearers etc.) and butts up to the headstocks at the ends; the floor is as long as the distance between the inside faces of the headstocks and the tops of the headstocks are level with the top of the floor. The curb rails abut the ends of the floor planks and the top of the curb rails is level with the top of the floor. The LNWR did it differently and had the end floor-planks overlapping the headstocks; these planks were rebated to fit. The SER (in the D1327 round-ended opens) did their own thing and made both solebars and headstocks taller than those of other railways (12" solebars and 15" headstocks IIRC) while keeping the curb-rails low; on these wagons, the bottom of the floor planks was clear of the top of the curb rails!

 

The moral is that unless you understand the carpentry, you can easily build in errors of 3" or so.

[bécasse]

Works drawings generally tell you how the draughtsman (who was often the actual designer) thought a loco (or item of rolling stock) ought to go together, although general arrangement drawings were often only produced after the item had been built, showing how it had been put together (but not necessarily accurately, particularly in respect of details, because these drawings were produced by junior draughtsmen or even apprentices).

 

Drawings were hardly ever used on the shop floor in a works (except for experimental work), foremen noted what needed to be done, selected suitable material from the stores and marked it with dimensions. The men then worked with that, cutting or shaping accurately where necessary but generally using their experience and skills to make something that worked and would fit but which frequently differed from the precise detail of the drawing. To give one example, a rivet line would be marked out with some precision but the actual position of the drilled holes for individual rivets was set out using a pair of dividers - so one man's line might have, say, 23 rivets and another man's 24, although since the men would good at avoiding unnecessary work, 23 would have been far more common than 24. So long as both were adequate for the task, either would have been acceptable to the foreman (and the drawing may well have shown 25!). Another simple example of avoiding unnecessary work occurred with wagon floors - a draughtsman would set out the planking using a pair of dividers from the centre line, so ending up with part-width planks at both ends of wagon, but they were never built like that, the planking was always started from one end so that any odd width planks occurred only at one end, less sawing along the grain and less wastage.

 

Until machinery driven by CAD technology started to appear on the shop floor, no two items were ever precisely identical except by luck.

[micknich2003]

Works drawings generally tell you how the draughtsman (who was often the actual designer) thought a loco (or item of rolling stock) ought to go together, although general arrangement drawings were often only produced after the item had been built, showing how it had been put together (but not necessarily accurately, particularly in respect of details, because these drawings were produced by junior draughtsmen or even apprentices).

 

Drawings were hardly ever used on the shop floor in a works (except for experimental work), foremen noted what needed to be done, selected suitable material from the stores and marked it with dimensions. The men then worked with that, cutting or shaping accurately where necessary but generally using their experience and skills to make something that worked and would fit but which frequently differed from the precise detail of the drawing. To give one example, a rivet line would be marked out with some precision but the actual position of the drilled holes for individual rivets was set out using a pair of dividers - so one man's line might have, say, 23 rivets and another man's 24, although since the men would good at avoiding unnecessary work, 23 would have been far more common than 24. So long as both were adequate for the task, either would have been acceptable to the foreman (and the drawing may well have shown 25!). Another simple example of avoiding unnecessary work occurred with wagon floors - a draughtsman would set out the planking using a pair of dividers from the centre line, so ending up with part-width planks at both ends of wagon, but they were never built like that, the planking was always started from one end so that any odd width planks occurred only at one end, less sawing along the grain and less wastage.

 

Until machinery driven by CAD technology started to appear on the shop floor, no two items were ever precisely identical except by luck.

 Here's a typical example "5/8 Bolts about 81/8 pitch." NERly Class L, later J73.

[Compound2632]

Works drawings generally tell you how the draughtsman (who was often the actual designer) thought a loco (or item of rolling stock) ought to go together, although general arrangement drawings were often only produced after the item had been built, showing how it had been put together (but not necessarily accurately, particularly in respect of details, because these drawings were produced by junior draughtsmen or even apprentices).

 

Drawings were hardly ever used on the shop floor in a works (except for experimental work), foremen noted what needed to be done, selected suitable material from the stores and marked it with dimensions. The men then worked with that, cutting or shaping accurately where necessary but generally using their experience and skills to make something that worked and would fit but which frequently differed from the precise detail of the drawing. To give one example, a rivet line would be marked out with some precision but the actual position of the drilled holes for individual rivets was set out using a pair of dividers - so one man's line might have, say, 23 rivets and another man's 24, although since the men would good at avoiding unnecessary work, 23 would have been far more common than 24. So long as both were adequate for the task, either would have been acceptable to the foreman (and the drawing may well have shown 25!). Another simple example of avoiding unnecessary work occurred with wagon floors - a draughtsman would set out the planking using a pair of dividers from the centre line, so ending up with part-width planks at both ends of wagon, but they were never built like that, the planking was always started from one end so that any odd width planks occurred only at one end, less sawing along the grain and less wastage.

 

Until machinery driven by CAD technology started to appear on the shop floor, no two items were ever precisely identical except by luck.

 

I'm not convinced by parts of the above. At the larger wagon works there must have been mass-production methods - probably including the use of templates and jigs - rather than individual marking-out of components. Take for example the Midland's 5-plank open wagons, of which 62,000 were built over a 20-year period, with the same underframe design being used on another 30,000 or so wagons; there's evidence that large quantities of standard components - solebars, headstocks, frame members as well as ironwork - were prepared in advance so that assembly was more like putting a kit together than scratch-building. The same went for Earlestown - that D2 wagon constructed in 1 hour 41 minutes wasn't marked out and cut from the raw timber; likewise Stratford's 9 hours 47 minutes to put together a Y14 0-6-0. 

 

No two items were precisely identical but very large numbers of items were identical within acceptable tolerances. That wasn't luck but judgement.

 

But I agree that the function of a GA was to record how things had gone together - spend a while studying the annotations to this version of a drawing originally from 1882 but updated for wagons built in 1917. 

 

(Incidentally this drawing illustrates that the Midland's standard method of constructing an open wagon was for the solebars and headstocks to be the same height - 11" - with the floor boards extending to the end of the wagon and retained in place by the end pillars - though I note that in this case the draughsman has drawn a 2 1/2" square piece of timber at the very end. Pace Guy, I think this arrangement is more typical of construction methods generally, though some builders used a wider planks for the end, so that the edge of the lower plank sat on the headstock, with the floor plank butting up against it.) 

[Spitfire2865]

I'm not convinced by parts of the above. At the larger wagon works there must have been mass-production methods - probably including the use of templates and jigs - rather than individual marking-out of components. Take for example the Midland's 5-plank open wagons, of which 62,000 were built over a 20-year period, with the same underframe design being used on another 30,000 or so wagons; there's evidence that large quantities of standard components - solebars, headstocks, frame members as well as ironwork - were prepared in advance so that assembly was more like putting a kit together than scratch-building. The same went for Earlestown - that D2 wagon constructed in 1 hour 41 minutes wasn't marked out and cut from the raw timber; likewise Stratford's 9 hours 47 minutes to put together a Y14 0-6-0. 

 

No two items were precisely identical but very large numbers of items were identical within acceptable tolerances. That wasn't luck but judgement.

 

But I agree that the function of a GA was to record how things had gone together - spend a while studying the annotations to this version of a drawing originally from 1882 but updated for wagons built in 1917. 

 

(Incidentally this drawing illustrates that the Midland's standard method of constructing an open wagon was for the solebars and headstocks to be the same height - 11" - with the floor boards extending to the end of the wagon and retained in place by the end pillars - though I note that in this case the draughsman has drawn a 2 1/2" square piece of timber at the very end. Pace Guy, I think this arrangement is more typical of construction methods generally, though some builders used a wider planks for the end, so that the edge of the lower plank sat on the headstock, with the floor plank butting up against it.) 

That is precisely the drawing I built my D299 to.  

Of what Ive seen from drawings and prototype photos, the floor was not built to be a part of the body but to sit within the body snugly.  Which explains the 2-1/2" plank at the bottom of the ends.  Therefore, the entire weight of the end of the wagon was bearing on the end planks rather than the floor or end pillar bolts. 

The key detail for any wagon is the relationship between solebar and headstock and where the curbrail is placed.  

[Compound2632]

That is precisely the drawing I built my D299 to.  

 

No you didn't, you interpreted this drawing in the light of your extensive knowledge of Midland practice! You gave your model Elliss 10A grease axleboxes, rather than the oil ones shown on the drawing, and the short brake lever, not the long one drawn. Although that drawing is drawing 550, it's not the 1882 original: it's been re-drawn to depict (as it says) the last 242 wagons of Lot 513 (the last Lot of mass-production) and then amended for Lot 919, the random batch of 1,000 built in 1917, presumably get some much-needed extra wagons in service by using up stockpiled spare parts.

[relaxinghobby]

Picture of cardboard wagon P1010015a

 

Before computer drawing the old fashioned way to scratch a wagon was to find a drawing of your prototype and photocopy it several times, glue the photocopies to some cardboard and cut out to give you some sides and end.

 

This model was made using intermediate technology, using a computer to scan and copy sections of the original drawing then printing it out onto a single piece of thin card, 0.6mm in this case.

 

Some drawbacks with this method is that you can't check the drawings for accuracy easily and running thicker materials through the printer can give you a distorted print as the printer's rollers struggle with the thicker material. Another drawback is sometimes the original drawing gives you an in accurate proportioned model like here. This one was too wide, when checked with the dimension arrow on the drawung. I've had to pull the side off so I can cut a slither from the floor and ends and reattach them again.

 

Because the printer was an ink jet type the ink can run or smudge so I cello-taped it to a piece of flat board and varnished it to protect the surface.

 

Another problem is the thinness of the material and it's relative weakness, a box van can be reinforced inside, out of sight but not but open wagons can get distorted, Maybe this can be taken advantage of to reproduce prototypically bowed out sides due to them holding heavy loads of loose materials.

 

Sides and ends and floor are two card thinknesses and the insides are blank, with CAD I could have drawn up inside detail.

Edited April 5, 2018 by relaxinghobby

[Compound2632]

I've toyed with the idea of printing wagon-sides in full colour - from computer-generated artwork, CorelDraw in my case - and then gluing on the ironwork cut from thin black card. But I have to admit I've not yet tried it out. I fear one would end up with prominent white lines where one scribed the planks, but perhaps weathering would take care of this.

 

As to bowing of the sides, for a coal wagon a block of wood supporting a coal load should fix the problem - if it is a problem? 

 

EDIT: Don't forget to pack out those dumb buffers so they are of square section end-on.

Edited April 5, 2018 by Compound2632

[bécasse]

I've toyed with the idea of printing wagon-sides in full colour - from computer-generated artwork, CorelDraw in my case - and then gluing on the ironwork cut from thin black card. But I have to admit I've not yet tried it out. I fear one would end up with prominent white lines where one scribed the planks, but perhaps weathering would take care of this.

 

 

Ah, Peco Wonderful Wagons, you mean. You will be making the springs out of soft nylon next so that they really do work like springs!

[Compound2632]

Ah, Peco Wonderful Wagons, you mean. You will be making the springs out of soft nylon next so that they really do work like springs!

 

Yes indeed - though did those have embossed ironwork? I used the underframes a couple of times on some early scratchbuilding attempts - a step up from the hacked Airfix 16 ton mineral u/f of my very first attempts!

 

But the point of home-printed sides would be to do complex PO wagon liveries that aren't available from e.g. POWSides. The alternative route which some truly dedicated types have gone down is to commission POWSides to make transfers for the wagons they want. These then go into the POWSides catalogue, so they're doing a public good.

[34theletterbetweenB&D]

I'm not convinced by parts of the above. At the larger wagon works there must have been mass-production methods - probably including the use of templates and jigs - rather than individual marking-out of components...

No two items were precisely identical but very large numbers of items were identical within acceptable tolerances. That wasn't luck but judgement....

 Every item thus produced will have fallen within the range of the applicable standard error. I recall very well the first time I heard the term 'scant' used in an assembly shop. This was of some standardised cut timber sections that were to be used in frame assembly* (of lightweight vehicles designed to be towed by underpowered cars and thus to create slowly movng tailbacks all over the road network in the summer months). Cut poorly seasoned timber on the best jigs and machines and the shrinkage will do that... And all machines wear, end stops move, temperature varies on a well known annual cycle, timber's qualities and treatment will jigger up the best laid plans, and the well - or not-so-well - meaning make 'adjustments'. Endless causes of variation, random and systematic would be all over something like wagon assembly. A glance at the range of tare weights of notionally identical vehicles coming off the line over the course of a year will give a clue to the scale of variation.

 

* The old boys knew what to do and the scheduled forty of the popular Arquebusier model were assembled that day. They may have been fractionally smaller than drawn, and significantly weaker than designed, but they got built, job dione.

[relaxinghobby]

In reply to Compound2662, can you get coloured card say from an art materials shop? If the interior is colourd a darkish tint it will not show up when scored, or at least not be very noticeable.  I've had this problem building Alphagrafixs card kits are they are usually printed on white postcard material, and in the end I've ended up painting them, loosing and printed lettering under the paint. I've just had a thought, if one can get some very fine pointed felt tip pens from the same art shop they could be used to colour in the white scores?

This leads into hand made drawings;

 

A bit like an old treasure map this one, all dog eared and tatty and perhaps leading the anxious seeker after dimensions like the seeker after black beards treasure astray too.

 

This is even more pre-CAD than the wagon printout above, a hand made drawing. I prefer lined writing paper as the lines can be used as a guide to help drawing things square. We did a year of technical drawing when I was at school. Using set squares, drawing boards and T-square rulers to try and get a square drawing, but it was all about neatness of drawing what technique to use nothing to do with design really.

 

This paper exercise was to enable a build an 00 model of the first incarnation of the little Sharp Stewart tanks used by the old Cambrian Railways on a Mainly Trains etched chassis. I had side elevations of the later version and the Great Westernised 20^th century engines to work from but no front or end views.

 

It seems the GWR rebuild had a much wider footplate than it's previous version and when I got as far as cutting plastic it did not look right in three dimensions. Looking at photos I had to work out the width by squinting at the distance from the end of the buffer beam and the start of the buffer fitting, then see it that seemed to give a good proportion for the front of the tanks. Different days gave different results. I ended up with a narrower model than the old white metal kit from GEM but it looked right to me.

 

Edited April 5, 2018 by relaxinghobby

[Compound2632]

In reply to Compound2463 can you get coloured card say from the art materials shop? If the interior is colourd a darkish tint it will not show up when scored, or at least not be very noticable.  I've had this problem building Alphagrafixs card kits are they are usually printed on white postcard material, and in the end I've ended up painting them, loosing and printed lettering under the paint. I've just had a thought if one can get some very fine pointed felt tip pens from the same art shop they could be used to colourin th ewhite scores?

 

The problem with coloured card is the same as the problem with making one's own transfers - printing an opaque white. I know it can be done but the equipment is outside the usual domestic price range. However, the felt pen idea is interesting - and is indeed what I use on the exposed folds of Metcalfe kits.