By continuing to use this site, you agree to our use of cookies. Find out more
Forum sponsored by:
Forum sponsored by Forum House Ad Zone

What’s the tolerance??

All Topics | Latest Posts

Search for:  in Thread Title in  
Bob Wild03/07/2020 23:50:23
86 forum posts
63 photos

Many years ago I were an ingineer. All drawings had a tolerance on each dimension. But now I’m building a steam engine where there are no tolerances at all on the drawings. It’s quite complicated as it is a twin cylinder and each is driving a water pump. My big concern is how to get them all lined up. Obviously the greater the clearance between the pistons and their cylinders the easier it is. So my big question is “How much clearance should there be between the piston and its cylinder”. Incidentally the bore is 3/4 in. Any suggestions from the experts would be most appreciated.


Nigel Graham 204/07/2020 00:57:53
2133 forum posts
29 photos

Model-engineering drawings and any accompanying constructional articles very rarely if ever come with tolerances. They might become more common on items such as locomotive frame-plates bought laser-cut from a materials stockist whose system needs such tolerances. Otherwise there are no formal dimension controls in the hobby; just the principle of working to the best of our workshops and abilities.

If you are making something that does need a good-quality fit such as a piston and cylinder I can only suggest being guided by the industrial standard classes of fit summarised in publications like the Zeus books.

I don't quite follow the logic as you have phrased it though. I am guessing you are actually building a Weir-type pump. Your text seems to ask if the alignment is a function of the fit of piston within cylinder, but I am sure you do not mean that. A piston should be a smooth sliding fit for the full length of its stroke, and both it and its cylinder should be as closely cylindrical as possible.

If the machine has a rigid rod connecting two pistons in separate, in-line cylinders, the quality of alignment should be controlled by the surrounding metalwork and the concentricity of cylinders, pistons and rods; not by the fit of the pistons in the bores. As I am sure you realised, you cannot correct misalignment here by turning the piston down - that would not work at all.

The closeness of the bore to the drawing is of secondary importance to these, within reason measured in "thous", as long as the piston hugs the cylinder wall as it should.


There are no tolerances for two reasons.

Firstly, we are not normally making interchangeable parts in the way vital to industry. It would not greatly matter if someone else making the same engine as you, machined his cylinder bores 0.005 " larger than yours as long you both made the pistons to suit your own engines. Yet obviously if you swapped the pistons with each other, that mis-match would give an engine kit that cannot be assembled, and one that can be assembled but would not run.

Secondly, model-engineering has always worked on the principle that as most of us are not professionally-trained machinist/ fitters or designers using full industrial facilities, we have to follow the drawings as closely as we can with the abilities and facilities we have, so we tend to make mating parts fit each other by specific example.

You do see some notes like "A must be a close sliding fit in B", or on some older drawings, the strange term "bare" after the dimension. Tolerances though; no. You bore the cylinder as closely as you can to the design, and turn the piston to a slightly firm, smooth and even, sliding fit.

Hopper04/07/2020 02:30:01
6393 forum posts
334 photos

It used to often be dimensions specified to 3 decimal places were + or - 1 thou. Two decimal places 10 thou and 1 decimal place 100 thou. I tend to just make everything to the nearest thou.

Piston clearance etc were often left to the maker to achieve by feel what was known as a "neat sliding fit".

Paul Lousick04/07/2020 03:21:50
2043 forum posts
722 photos

All manufacturing drawings should have a tolerance for parts on drawings or at lease the type of fit of mating parts. There are industrial standards for these. BUT most of the detail drawings which are supplied for models seems to have them missing.

This I beleive is because they have been created "on the cheap" and not by experienced draftsmen who have the required qualifications. (One of my pet rants as I have worked as a professional design draftsman/engineer for 40 years and expect a better standard on drawings).

On the other hand, if the drawings were made by a professional, they would cost $1000's instead of $100's to buy. Ya get wot ya pay for ! dont know


Edited By Paul Lousick on 04/07/2020 03:22:53

JasonB04/07/2020 07:24:25
22747 forum posts
2653 photos
1 articles

On the other hand putting an H7 tolerance on a hole for say a Stuart 10 would have most beginners wondering what it meant and even when they found out they would not have the equipment to measure it.

You sometimes get a piston shown as a thou less than the bore but that is about as much as most ME drawings though some of the German ones do have hole, shaft and surface tolerances shown.

Once I get close to finished size on a piston by measureing with callipers I will then turn until I get the fit that feels right, could not tell you what it actually measures but all the engines run. It's just a case of building up the experience to know when 2thou will do and when 2tenths is needed. Hopefully as an engineer you should have a good feel for what is right unlike hobbists like me.

Philip Burley04/07/2020 07:43:24
198 forum posts
1 photos

Some of the drawings I have been working to , have all dimensions in fractions . A part made to 3/16 inch can have quite a wide variation when miked , So its just parts made to fit isn't it


Howard Lewis04/07/2020 08:20:12
6104 forum posts
14 photos

At one time, if the dimension was given as a fraction of an inch, it was assumed that the tolerance was 1/64

In high volume manufacturing industry, often an untoleranced dimension would be taken to be +/- o.005".

Almost always there was a note on the drawing template, stating this..

The parts made had to be interchangeable with others to the same drawing.

The manner in which dimensions were shown could cause problems. A series of holes dimensioned from centre to centre could result in a tolerance build up meaning that the distance between first and last holes could vary greatly, although all hole centres were within tolerance. Imagine ten holes all to +/- 0.005 centres,, for the studs to secure a sump to an engine block The first and last holes could differ by as much as 0.100" and yet be to drawing!

The better way was to dimension each hole from the same datum point, so that no tolerance build up could prevent a matching part from fitting.

Tolerances will be important if you are working on a Club project, where parts that you make have to fit to parts made by someone else, But where you are making the matching parts they are not important, as long as the bore does not break through when it should be blind or have a minimum wall thickness.


Andy_G04/07/2020 10:36:36
173 forum posts

Traditional model plans seem to rely on the dying art of 'Fitting', rather than manufacturing to numbers - it is expected that the modeller had an understanding of how the parts function and fit together and applies their judgement and skill to making sure that this is achieved - if something is required to match a proprietary part (such as a bearing race, etc.) then a tolerance might be called for, otherwise, it is only required that the parts in the model work with each other;

If one were to bore a cylinder slightly over-size, would one scrap the part (manufacturing)? Or would one just make the piston slightly bigger to suit (fitting)? I realise this is entirely down to personal choice, and the ultimate goal of the modeller - a working model, or the satisfaction in producing a 'blueprinted' example.

I've recently watched a lot of machining videos on Youtube, and the trend of doing everything by DRO bemuses me - e.g. 3 accurate set-ups to drill 4 holes to hold a steam chest and its cover onto a cylinder block. Aside from true scale modelling, the position of the holes doesn't really matter, as long as they don't go into thin air - mark and drill the holes on the cover so they look" right" and then spot the holes through to subsequent parts - everything lines up, no tolerances necessary, no mill necessary, no DRO necessary. (If people have the kit, and want to do it this way I'm not knocking it, but it isn't the only way, and probably wasn't the method that the originator of the plans had in mind when they were drawn.)

Are we building models or are we manufacturing parts that someone else has determined can be assembled into a model?

Anyway, that's my rant! cheeky

(Part of my job is producing manufacturing drawings, btw.)

Dave Halford04/07/2020 10:41:51
2050 forum posts
23 photos

A lot of ME drawings are not drawn to any one scale or projection and are more a collection of 'scrap views'. Bolt holes may not have the intended thread mentioned, just the centre location.

If your engine was originally serialised in a magazine, then assembly, marking out and lining up was dealt with in the write up. Hughes tells you how to align the engine mounted on the boiler in his book on Minnie, which will work for any traction engine.

Known mistakes are not corrected, the author having 'moved on' to the next project. Omissions from the magazine write up remain omitted from any subsequent drawings. Haining omits the steering gear detail on his 16hp single ploughing engine all you get is the chain drum.

It's all guidelines Bob, just guidelines.

Alan Jackson04/07/2020 10:50:49
258 forum posts
146 photos

The benefit that model engineers have is that they are the only one working on their project and as such can make any part near enough to the dimensioned size on the drawing. The mating part can then be made to match, but the art is knowing how the mating part has to fit and work to suit the required operations. If drawings were provided with machining tolerances, it would force the maker to comply with rigid requirements that would make the part much harder to produce, as well as require remaking in some instances to achieve a size which is often really only a dimension put on the drawing to show the general outlines etc. Tolerances are required for components made by various outside sources in order that other parts from other sources can fit together and comply with the design requirements.

Imaging making the drawings for a clock, say, with toleranced dimensions. It would force the designer to now legally comply with ensuring that every clock made to these drawings will run like clockwerk, sorry could not resist it. Have you ever tried to go through a design ensuring that the tolerance build ups (that is adding and subtracting the variations in a design due to component tolerance extremes) to ensure it can still function when the worse variations of the mating components come together?


Edited By Alan Jackson on 04/07/2020 10:53:43

Nick Clarke 304/07/2020 11:03:44
1425 forum posts
63 photos
Posted by Paul Lousick on 04/07/2020 03:21:50:

This I beleive is because they have been created "on the cheap" and not by experienced draftsmen who have the required qualifications. (One of my pet rants as I have worked as a professional design draftsman/engineer for 40 years and expect a better standard on drawings).

On the other hand, if the drawings were made by a professional, they would cost $1000's instead of $100's to buy. Ya get wot ya pay for ! dont know

I don't think you are being fair to the people who over the years have created many successful designs. Martin Evans wasn't a draftsman and neither were LBSC or W.J. Hughes, which is not to say they didn't have drafting skills, but most of the drawings published in magazines were redrawn from pencil originals for publication. I suspect that there was a deliberate policy to not include tolerances as this would only open up the way for amateurs to fail if they did not make things plus or minus a thou. Look at almost any published drawings and the design is often a long way from the purported original in order to make it suitable for a non professional to have a chance of successfully completing it.

Machine tools, right up to the second world war were often not provided with graduated dials and many of these were still in use years later. Micrometers and verniers were not always found in the amateur workshop, callipers and rule were the norm for many.

In my own case I can with confidence say that I can make a piston fit a cylinder, but I doubt it is ever exactly as drawn, but it has always worked.

Tolerances are only one way of saying Pass/Fail or Good/No good - the parts fitting together and working is an equally valid, and far more achievable, way of saying the same thing. If you look at the articles for a published design they will add to what the drawings have on them - eg turn down the end of an axle until it will just enter the wheel and similar.

Finally commenting on your question of the cost of drawings, designers have frequently ceded their copyright to the drawing to suppliers so what you are paying for is prints, not drawings.

Edited By Nick Clarke 3 on 04/07/2020 11:08:25

Paul Lousick04/07/2020 11:52:44
2043 forum posts
722 photos

Hi Nick,

My reference to the standard of drawings was directed at those sold by commercial suppliers of model engines, which should be to a better standard. These drawings have been re-engineered for making a model version of the original and normally have a copyright warning on them.

Yes, drawings in magazines are often created by non-trained draftsmen (sorry, draft person), but should indicate the type of fit required. eg. Interference, slide fit, running fit, etc. to give an indication of what to make. Which is why this thread was originally posted as there was no reference on the drawing.


Mick B104/07/2020 11:52:46
2191 forum posts
120 photos

When I was working as a Tool Designer, the drawing sheet blanks (some kind of coated film that had enough friction to take good pencil markings) all had standard printed boxes bearing 'tooling Req. no.', 'drawn by', 'checked by', projection, title, tool no. etc., alongside a list of General Tolerances. These were (IIRC) no decimal +/-0,25mm, 1 decimal 0,1mm, 2 decimal 0,025mm, 3 decimal as specified. That meant that if a particular precise value was required - for example detailing punch and die dimensions for press tools - the draughter should normally specify top and bottom limits fully. But there were still often grey areas, and not only where some Industrial Engineer came along and smudged your drawing by carelessly pointing out some feature with a grubby finger ! angry

I think that for most ME applications, it wouldn't be practical or sensible to specify every dimension exhaustively, so what you usually get is more or less what's needed, although many force the builder to work out values that could helpfully be noted as '(Ref.)' dimensions. A general guide as to piston/cylinder and shaft/bearing clearances would be useful on many.

Edited By Mick B1 on 04/07/2020 11:53:31

Nick Clarke 304/07/2020 12:18:33
1425 forum posts
63 photos

I think the whole concept of drawings published in magazines as engineering drawings and as such containing all the information required to build a finished model is erroneous.

Magazines publish articles written by people who are, in this context, journalists and not engineers.

To create a model you may well need the text as published in the magazine as well as the drawings (what LBSC called the words and music) Unfortunately you may also need the subsequent copies of the magazine where corrections were noted. These may or may not have been included in the plans you buy from a supplier, which in any case do not include the text. Basically when you buy plans you only have the illustrative sketches provided in a magazine article. Knowing this is half the battle.

Of course this may not be the case if you are buying a design that was not published in a magazine originally, but the conventions laid down by 120 years of articles are followed in many of these as well.

Edited By Nick Clarke 3 on 04/07/2020 12:18:48

Bob Wild04/07/2020 12:43:51
86 forum posts
63 photos

Well I didn't expect such a furore of responses, but it does make interesting reading. Thanks to everyone for their replies.

Nigel, I should have been more explicit about my application. It is a Merryweather Fire King self propelled fire engine. It is described in my thread Merryweather Fire King

Dave "If your engine was originally serialised in a magazine, then assembly, marking out and lining up was dealt with in the write up." There is a reference to the 1908 article in my thread, and I was astonished by the lack of guidance from the author, not to mention the lack of dimensions!

All Topics | Latest Posts

Please login to post a reply.

Magazine Locator

Want the latest issue of Model Engineer or Model Engineers' Workshop? Use our magazine locator links to find your nearest stockist!

Find Model Engineer & Model Engineers' Workshop

Support Our Partners
Eccentric July 5 2018
Rapid RC
Eccentric Engineering
Subscription Offer

Latest "For Sale" Ads
Latest "Wanted" Ads
Get In Touch!

Do you want to contact the Model Engineer and Model Engineers' Workshop team?

You can contact us by phone, mail or email about the magazines including becoming a contributor, submitting reader's letters or making queries about articles. You can also get in touch about this website, advertising or other general issues.

Click THIS LINK for full contact details.

For subscription issues please see THIS LINK.

Digital Back Issues

Social Media online

'Like' us on Facebook
Follow us on Facebook

Follow us on Twitter
 Twitter Logo

Pin us on Pinterest