Designs published in ME

The latest ME has an interesting letter from Geoffrey Johnson, who has found out the hard way that just because a design is in print doesn't mean it can be built to work!

As a professional engineer he knows how to design and draw having spent a long time learning how to do it properly. Most model designers are amateurs who want to make their favourites in miniature. I am one of these, and a few mistakes were evident when modellers started to build from the Q1 series in ME.

By building two locos to prove the drawings I also found a few mistakes myself, so was able to update the drawings. Anyone purchasing drawings from Polly models will get the latest revision, and this will be shown with the revision date on the drawing number. This really should be standard practice.

But for future series in ME it really would help to know if at least one working model had been built to the author's drawings.

The main problem with errors is letting future builders know when the drawings cannot be amended owing to cost.. Even when such discoveries are mentioned in ME there is the problem of finding the references. It would help if every drawing set budgetted in an A4 sheet of cross references to any letters or articles on the model. This would still be an ongoing cost but far less than updating the drawing.

Mistakes happen, mis-interpretatons happen, and even re-drawing for publication can cause problems.

Whilst it would be nice to be able to purchase digital versions of drawings, just having a list of drawing versions and corrections would be nice - eg, if Polly or ME had a simple list of Nick's Q1 drawing numbers and versions and change notes, one could verify that the set one has purchased is as up to date as required for construction.

I have (and, still do) have software distributed by Apple and Google/Android. It is most certainly not perfect, but at least someone who has plunked down the 99p for my Android app will get an automatic update without charge, until the end of time.

I wish drawings could be handled in the same way, but I don't know how to stop piracy of the design in that case.

Another JohnS.

Having produce several sets of drawings for ME I think there are some important considerations to bear in mind.

Those for simple devices, like a drawbar or hole punch, are relatively easy to check, something like the QCTP was challenging as many fits had to be checked, but those for a complex model like the crane of the Norden steam engine are impossible to check completely accurately except by building another comlete model. Even then there are two problems - one is of simply as doing things instinctively without reference to the drawing (such as machining non-critical dimensions or making to parts to fit each other), the second is that most of us make small changes as we go along to 'improve things' or use materials to hand - these have knock on effects that might disguise a drawing error.

I must say I felt awful about the error in the canal crame frames. The original drawing spaced the gears at their outside diameters, I corrected the model as I went along to use the PCD, but among other changes I omitted to make this change on the drawing. This surely shows ho remaking amodel won't neccesarily find the errors.

Plain and simply, if ME and MEW are goinmg to rely on amateurs to produce the drawings, then there will be some errors and idisyncracies. That isn't because people don't check or don't care - its mainly because we build AND design AND draw the darrn things. If they went to a professional draftsman, and then to a factory floor then, yes the faults would be found - hopefully - look at the (re)building a Britannia series with its catalogue of errors that Doug and his colleagues had to correct. That was a 'professional' design.

That said, I have just made a gadget for my lathe. The drawings were by an ex-toolmaker and impeccable. I had to work hard to intrioduce my own errors. Even then, there was a significant change I had to make because of an unexpected difference in the arrangementy of my machine.

I do think it is best if a model has actually been made and works, because that at least proves the basic feasibility and throws up any machining issues. For me it's the 'design process' and the 'problem solving' that are most interesting (I look forward to more of the Derwent series for this reason).

I also think that series should, wherever, possible, be regular and uninterrupted (but monthly is fine) so builders don't get 'left in the lurch'. The monthly split is good because it means greater variety, but still with certainty.

Finally back to those errors. When mistakes re found (and they always will be) they should be published in the magazine and on this website. Original drawings should be amended and used whenever supplied in the future. A change log is a good idea, but not essential for simpler models.

But readers must be prepared to use a little brain power - it isn't reasonable to put every fit on a drawing when the fit required is blindingly obvious (I keep using the examples of a wheel on an axle and an axle in its bearing). Similarly some non-essential dimensions can be left to the judgement of the builder - or inferred from another drawing (can YOU see where to get the position of the horizontal groove on Curly's slide valve this week?)

This in itself is an interesting issue - if two parts mate, should we only specify the size of one of them? Older drawings often do this and look less cluttered, less room for errors too. But some folks are desparate for every single dimension to be listed.

I hope that in future I will take on more and better new designs. I hope they will be of interest to others, and I hope my drawings will be better. Two things will help me with this - people letting me know when they find a mistake, so I know what to watch outfor next time, the second is constructive criticism so I know what I could do better in future.

In teh end, only two things are certain:

1) You can't please everyone.

2) A realistic allowance has to me made for abilities of contributors - unless those who knock other's work are prepared to submit their own for criticism their achievement will be to cause the supply of new ideas and articles to dry up completely.

Neil

The use of 3D CAD can make some errors much more obvious early on, because the virtual model is assembled from actual sized parts. Checking for interferences becomes a simple exercise (as long as the Designer remembers to do it!), and gaps should be fairly obvious. Since 2D views and dimensions are generated automatically from the 3D model, the scope for simple typographical errors in final drawings is much reduced.

Even with 3D CAD it is still possible to make errors of design or of final presentation which might leave the builder head scratching. It is quite easy to leave a dimension off the drawing!

I should declare that my view on this is likely to be biased, as I earn some of my income from supporting a 3D CAD product.

I look at it this way, I'v got a collection of bits, I'v got a rough idea of the finished item, if all else fails look at the drawings, in other words I do my own measurements. I don't normally build scale models, but design and build small hot air engines. The Stuart S9 was the first of that type of thing, and I just took the measurements off the job. Ian S C

A list of known issues with published designs would help. Doesn't affect copywrite as no changes are made to drawings but highlights stuff to watch out for. Most people will sort out their own work-arounds. Could be as simple as 'Clearance issue on side plates on sheet 2". You never know could be an additional money spinner for publishers.

Martin

3D CAD can throw up a lot of howlers, though - I've seen it myself - when we came to assemble some bits it was physically impossible to assemble them, but if you could have sawed them in half and put them together again it would have fitted beautifully! Really need a combination of both 3D CAD and an independent maker of the bits to prove a design. "There is always one more mistake in any design...."