Toleranced and open dimensions on drawings

[Tony MartyrParticipant @tonymartyr14488]

I served a Student Apprenticeship (1960-65) with WH Allen who made turbines, large diesels, pumps and gearing: every component drawing issued was dimensioned in one of three ways: Toleranced, open or calculated.

(Remembering that these were 'imperial' units)

Toleranced dimensions always took the form of the Decimal 'maximum metal condition' dimension + 000, – the chosen tolerance. Bores often had their 'fit' designation as in Size, tollerance, H6.

Machining open dimentions were usually +/- 10 thou and open fitting dimensions were given in fractional figures.

Calculated dimensions were just that but finally set by the fitter's final adjustment.

Why do none of the model engineering designs for locomotives or stationary engines that I have ever seen have toleranced dimensions?

I am currently building a modified version of the Bolton/Bertinat Triple Expansion Engine and without my knowledge of full size marine engines it would be quite difficult to know which of the dimensions are vital in such a composite build up of component assembies.

When I started to build a 3 1/2" William some years ago I knew very little of loco chassis construction and without key dimensions being toleranced I had to learn the hard way. I was told 'don't be too 'tool-room' about the dimensions, as locos need a lot of slack to run' – which I found to be only true in parts!

Have the esteemed designers of the model engineering classics never used industrial practice or guide tolerances in their drawings and if I ever used a modern metric design would I find that the practice is used nowadays?

[Tony MartyrParticipant @tonymartyr14488]

[JasonBModerator @jasonb]

I'm happy if I get a dimension and it is corect.

I think a lot depends on where you look for your model drawings, I have taken MEB from the start and you generally get 20 odd sides of A3 drawings with that all of which have a tollerance usually base on the number of decimal places used for the imperial size.

Eg a part dimensioned at 0.5" will not be so critical as one dimensioned at 0.500"

I have also seen the question raised on a number of forums "what does H6 or h7 mean" so either way the beginner is at a disadvantage.

You also nead to bear in mind that we are not making parts that need to fit an assembly on the other side of teh world, we have both parts in our hands and can just fit one to the other though it does help to know what sort of fit to aim at.

J

[Stewart HartParticipant @stewarthart90345]

Tony

Like you a served an apprenticeship attended day release and night school and was taught all about tolerancing, min and max metal conditions ect, but that was for industry where trained engineers were talking to trained engineers. In the model engineering environment you are talking to all sorts of people with a different knowledge base, tolerancing drawings I'm sure would leave many people baffled and bewildered and looking for another hobby, far better to leave them off, and let people work out for themselves where you need a running fit, tight fit, alignment, squareness etc, most people with a little common since can sus these things out.

As Jason said you're not looking for interchangeability of part, its your model you can make it any size you want, the important thing is to get the fit correct for it to function correctly

Stew

[Roderick JenkinsParticipant @roderickjenkins93242]

Posted by Tony Martyr on 21/01/2013 17:46:45:

Have the esteemed designers of the model engineering classics never used industrial practice or guide tolerances in their drawings and if I ever used a modern metric design would I find that the practice is used nowadays?

No, because they know that we amateurs who are their audience have no idea what H fits are, whereas we understand what a "press fit" or "sliding" fit is. Forgive me, but I sometimes think that people forget that this is the MODEL engineers forum. While I find the comments from those involved in industry and full size practice interesting and informative I think there is a danger of putting readers off by suggeting that full scale practise (and equipment) is essential to make models.

cheers,

Rod

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

Having proper tolerances on drawings makes making things easier not more difficult .

Most of the esteemed designers – particularly of locomotives – were nearer to being illiterate than being the geniuses they are always made out to be and the dreadful design and drawings practices associated with them have left a poor legacy for new comers who are trying to 'learn' model engineering .

Most of the esteemed designers seem to either not bother with tolerances at all or go overboard with largely meaningless sets of numbers attached randomly to dimensions .

One of the problems is that none of them had any concept of relational and associative tolerances – that is the tolerances required on a set of individual parts so that a complete complex assembly will go together and work . Relational and associative tolerances are a stage up from simple part to part tolerances but much more relevant to (eg) model locomotives .

To take one example which everyone is familiar with – the locomotive chassis and associated component groups in it .

Spacing of axles needs to match spacing of bearing in coupling rods for free running of chassis .

Positioning of axles is controlled by setting out of slots in frame , width of horn slides and centering of axle hole in axleboxs (we'll ignore minor considerations like running clearances for now) .

Slots , hornslides and axleboxes can each have tolerances to give required accuracy and coupling rods can have tolerances to make them match the resulting axle centres . Sounds professional and impressive doesn't it ??

Actually its a load of nonsense – there's no fixed reference point , nothing is fixed dimensionally , the whole of the dimension set is in fact floating around in statistical space and an average assembly will be very much looser than needs be .

What do you think would be a more meaningful way of tolerancing the parts ??? There is no stipulation that it need be conventional practice .

Notes : Tolerancing can be descriptive or numerical – either can be very effective .

Michael Williams .

[SpringbokParticipant @springbok]

well said rod

[JasonBModerator @jasonb]

Posted by MICHAEL WILLIAMS on 21/01/2013 19:30:11:

To take one example which everyone is familiar with – the locomotive chassis and associated component groups in it .

Thats assuming all model engineers are interested in making Locos, a lot of people comming into the hobby would rather make stationary or IC engines or make bits for their RC models. Or are like me and have never built a loco.

J

[Sub MandrelParticipant @submandrel]

Tolerancing is only useful where parts have to be interchangeable or in the rare cases where absolute accuracy is necessary to match third party components or for function (e.g. a measuring device).

For model engineers what typically matters is the class of fit.

Neil

[Mike AngeloParticipant @mikeangelo94956]

It would make sense if the desired clearance ( or interference) on some parts was specified . Thus a bearing would have the radial clearance specified e.g. 0.001" for a 1.5" shaft. This would give the machinist some idea of what to aim for, even if the actual size was not spot on.

[Phil PParticipant @philp]

I am of the opinion that when a model is described in a magazine article, the drawings are often lacking in a lot of detail. It is only by reading the associated article that a builder can understand what he should be trying to make and the types of clearances he should be aiming for.

I am from a mechanical engineering design background and produce drawings all day every day to earn a living. I do not expect our machinists who make parts to my drawings to read an article about it before they proceed, everything should be on the drawing and fully understandable.

As far as tolerances go, then we normally have a general tolerance that applies to the whole drawing, and only apply a specific tighter tolerance when absolutely neccesary such as a bearing fit in a bored out housing etc. In industry tolerances cost money and should oly be applied when it cannot be avoided especially if a job is to be sub contracted out to someone else. As soon as they see a lot of tolerances "ker-ching" up goes the price.

This argument has done the rounds many times before, and I don't think there is one answer that will suit everyone.

As far as I am concerned, I have a rule to never cut metal until I have drawn something myself. That way I cannot blame anyone else if it is wrong, but again this method will not suit everyone as some people like to be spoon fed in the hope that the drawings are correct, but sadly that is rarely the case.

Just my 2p worth

Phil

[nigel jones 5Participant @nigeljones5]

I too served an apprenticeship…but that said, i try to work to the exact dimension, im under no pressure to do it quickly or even do it right first time in many cases. But you can always work to this: If it wont move its too tight, if its tight it will loosen and if it rattles it needs rebushing! Works for me and my steamers.

[TakeawayParticipant @takeaway]

Hello Tony – I know exactly how you feel. Having worked in the industry for many years I became used to the disciplines (as you were by the sounds of it) that were imposed on us by toleranced drawings and eagle eyed inspectors. If you did it by the book all was OK.

But the way I see it, model engineering covers such a vast selection of potential projects that expert drawing office detailing is impossible for most of it – and a good job too. Why? Because IMHO not being under the cosh of the drawing tolerances means we can all do our own thing and that means the discipline has gone and we can all enjoy this stuff as a hobby and not as a means of earning a crust.

Highly detailed and toleranced drawings would be just the ticket for some blokes who like to craft scale models. But go to other end of the scale (where I live) and enjoyment of the hobby is in scratch building own designs (or modified designs of other blokes) from whatever materials are clanking around in the scrap box. Freddy Mercury (he was born the same day as me) once professed to being a musical prostitute – I feel as laid back about the hobby as he felt about his calling. I regard myself as artist rather than artisan (ok, I suppose the whisky helps)!

Anyway Tony, I was brought up as a kid in Queens Park, Bedford in the 'fifties and remember WH Allens in Ford End road very well. I never worked there but I can tell you that on a rare visit from her home in a small Wiltshire village in the early 'fifties my granny was tickled pink one day when we watched the factory disgorge its workforce at around 5 p.m. teatime. She'd never seen so many bicycles in her life in one go. Hundreds of men – mostly in dirty brown raincoats and cloth caps riding their bikes in line abreast over the train bridge and into Bedford.

There was a lot of engineering being done back then and a lot of clever blokes doing it – permission to feel sad sir?

[Jens Eirik Skogstad 1Participant @jenseirikskogstad1]

Tolerance is suitable for mass production of the same product due we need spare parts who will fit into the product as a replacement of broken part. And we as modelengineer find there is a requirement for a specific adaptation of a part that should go into the second part to work best as we can find for example in a model engine without piston rings. The spareparts in our own product is a "try and fail" fit to replace the broken part.

[Phil PParticipant @philp]

Sounds like you are working to approx H7 – g6.

H7 being approx normal for a reamed hole, note the capital H for hole.

g6 being around the size that gives a nice sliding fit with not much play, note the lower case g for a shaft.

It is that simple, all you need is some easy to understand charts.

http://www.roymech.co.uk/Useful_Tables/ISO_Tolerances/ISO_LIMITS.htm

Phil

[Anonymous]

Tolerancing didn't do W. H. Allens any good; the factory is now a housing estate.

What's left of the company is camped out in what used to be A Block of the grammer school in Bedford wot I went to, before it closed. Come to think of it, all of the schools I went to in Bedford have been closed and/or knocked down. Hmmm, may be there's a pattern there?

Regards,

Andrew

[Mark CParticipant @markc]

Regards "H7 being approx normal for a reamed hole, note the capital H for hole" the H in the H7 is the tolerance and the upper case "capital" letter indicates it is a hole not the letter H standing for "H"ole.

There exist two systems in general use – Hole basis and Shaft basis (hole basis being the common). In hole basis the hole remains a standard tolerance and the shaft is adjusted to give the "class" of fit. Unsurprisingly, in shaft basis fits, the shaft remains a fixed tolerance and the hole tolerance (size) is matched to the shaft to give the fit.

In the ISO system there are three types of fit – Clearance, Transition and Interference with four preferred tolerance bands for either the hole or shaft (depending on whether it is hole or shaft basis). The class of fit is then achieved by combining the 4 preferred tolerance bands with the required mating tolerance to give the myriad fits within the basic 3 (clear, transition & interference).

I hope I have expanded this in a meaningful way for those who have never seen them or tried applying them! Do not expect a clear definitive answer when it comes to applying them as there are frequently arguments about how they should be applied to give sensible drawings – the best approach, as mentioned earlier, is to avoid them at all cost unless they are absolutely necessary and try and design in a way that reduces sensitivity to tolerance issues!

Mark

 

Edited By Mark C on 22/01/2013 14:39:26

Edited By Mark C on 22/01/2013 14:39:57

[Tony MartyrParticipant @tonymartyr14488]

Having read the interesting responses to my initial query I am now confirmed in the view that most of the sets of drawing that I have used over 50 years of model engineering are badly draughted, in that, through lack of any dimensional tolerancing, there is no differentiation between important and 'open' dimensions.

Tolerancing makes the interpretation of an engine design easier to understand particularly if, like me, you build multi-cylinder steam or IC engines where the build-up of arbitrary errors can lead to a disastrous misalignments between cylinder and entablature components.

I find it interesting that so many model engineers are using machines fitted with DRO, which gives them the capability of easily measuring to fine limits, yet have no guidance in most set of (imperial) drawing sets of classic designs as to when and where to use such accuracy.

I dislike the concept expressed here that 'model' engineering is a lesser form of engineering and that practitioners simply bodge things to make them fit. I have only made one 'model' in my life (31/2" William done as I wanted to test my boiler and painting skills) all my other work has been miniature working engines in which running clearances, inter-bearing and cylinder dimensional tolerances important, whereas many other dimensions are not.

Looking at the present set of drawings I am using I have come to the conclusion that they were not draughted well and that without the supporting ME articles most of us would be hard put to work out a sensible order in which to build the components and will need considerable experience to judge the relative precision of their dimensions. Its not a good tradition nor a good guidance for young people starting the hobby

Tony

[Roderick JenkinsParticipant @roderickjenkins93242]

Tony,

Most drawings do have implied tolerances. For example: fractions imply a tolerance of+/- 1/64" and I think it is fairly clear a shaft to fit in a 1/2 " ball race needs to be 0.500 +0 -.001. However, I suppose this is part of the language of model engineering. Maybe these things need to be expressed on the drawings. I rather resent your implication that anything made to existing drawings is a bodge. Perhaps you could write an article for ME explaining how to do it "properly" and give us a design that we don't have to bodge.

Respectfully yours,

Rod

[Francois MeunierParticipant @francoismeunier96697]

I can cope with imperfect drawings.
numerous model engineer build great engines that work very well, without any drawings, at most a few sketches on the back of an envelope, and some do blue prints afterwards, which is not the proper way to proceed in industry, but they allow others mates to copy their model, I should therefore like to thank them.
Moreover, these drawings are not set in stone, and are generally adapted to take account of tools and materials available in our workshops (ie Imperials or metric taps and dies ..).

Cheers

Zephyrin

[Tony MartyrParticipant @tonymartyr14488]

No need for articles Rod – the proper way of producing a drawing can be found written in any basic level Technical Drawing training course from City & Guilds etc. With the increasing use by model engineers of DRO and CAD/CAM the historical designs are increasing anachronistic and were always, in my view, poor compared with their contemporaneous industrial practice

Tony

[Roderick JenkinsParticipant @roderickjenkins93242]

I don't think I want to do a City and Guilds before I can become model engineer.

cheers,

Rod

[Sub MandrelParticipant @submandrel]

Tony…

You don't mean to say my B-grade in Geometric and Engineering Drawing O=level is next to useless

Neil

[Stewart HartParticipant @stewarthart90345]

In the end Model Engineering is what you want to make of it, you're working away in you own little shed making something for you're own enjoyment, if you want to work with industrail quality drawing correct to BS308 with everything correctly toleranced thats fine, if you want to work off the back of a fag packet making things to fit as you go along thats fine as well, both methods will result in wonderful models, that the builders can quite rightly be proud off.

You can be a Mr Maticulas or a Mr Bodger

Whats realy important that we all work safely and enjoy our modeling.

Stew

[Roderick JenkinsParticipant @roderickjenkins93242]

I agree with Stew that what we do in the privacy of our own sheds is entirely up to the individual. Tony's point is the quality of the drawings that we share with others. He has a valid point that the drawings that we purchase should be stand alone and not need the words and music, which may not be available in the future. I only disagree with him on the standards that we use. My view is that the drawing should be intelligible to the relatively inexperienced and the modeller should not have to learn a new language before starting to cut metal. In order to achieve this though, the originator must be clear what is intended, using copious notes if necessary.

It's a useful discussion. As I have mentioned before, It would be nice if we could come to some sort of consensus on drawing guidelines for publication in the magazines.

cheers,

Rod

[Author]

Posts