Lets Talk about Engine Drawings

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.


Well-Known Member
Staff member
HMEM Supporting Member
Global Moderator
Jul 2, 2021
Reaction score
MidSouth, USA
There are some recent discussions about engine drawings.

I hate to admit it, but I am ignorant of almost all drawing terms, except "front view", "side view", "bottom/top view", and "isometric".

I have heard some mention that dimensions should not be repeated.
I don't really follow this line of thinking.
I generally put the dimensions in an arrangement that makes it easy to take them off of a drawing, when making manual patterns.
I often have a string of dimensions, and then an overall dimension, which is redundant, but I don't want to have to pull out a calculator while I am reading drawings.

Here are the drawings I made for the green twin.
I did not follow anyone's rules about making engine drawings, and am basically ignorant of any and all engine drawing rules.

I just do my own thing with drawings; I have always ignored almost every rule I was ever taught in drafting class.

So somebody should look at my green twin drawings, and tell me what you like or don't like about them, or how to make them better.
Are there some gaffs in the green twin drawings?

I never have problems using my own drawings.
I guess this is prove that ignorance is bliss.

I guess I somewhat mimic the style of Kozo Hiraoka, but it is not an intentional thing, but more of just a desire to make the drawings very clear and concise looking.

I am pretty sure I don't use the standard callouts for holes and things, bascially because I am ignorant of those standards, and I don't think that affects the usability or readability of engine drawings.

Have I repeated dimensions in the green twin drawings?
Honestly, I have no idea; I just make the drawings, I don't critique my own work.

Last edited:
I posted this recently in a related thread, so I will copy it here for continuity.

I have been drawing things for many years, but I must confess I am pretty ignorant of most of the terms, such as "third angle projection", "first angle projection", etc.

I guess I have just been ignoring all that and drawing things like I like them.

But now I am wondering how I do my own drawings.
I never really thought much about it; I just draw engines.

I will go back and review some of my drawings.

I do recall making isometric drawings by hand, and later in 2D CAD, and I remember that being rather tedious, and not very accurate.
I was very pleased to find out that Solidworks allows a view to be dragged out at a 45 degree angle, into what I call an isometric view.
I have no idea if "isometric view" is even the correct term.

I guess I am one of those folks who doesn't really know the names for all the drawing features, but I know what I like when I see it.

I do like to line up parts on a 2D sheet, as if they were an exploded view.
Randomly placing parts of an engine all over multilple sheets drives me nuts.

I have to put connecting pieces adjacent to each other; I guess its an obsessive compulsive thing.

I don't like to crowd a sheet either, and this also drives me nuts when I see others pack a sheet completely full of parts and dimensions.

To each their own I guess, and as long as the drawings are useful, and relatively intuitive, then pick your own style and go with it.

Regarding your string of dimensions, I generally try to dimension from a datum rather than a string firstly because you need to start adding up the string to get to where you want to be and secondly there can be rounding errors that will add up with the more dimensions that are in the string.

The lower line of the mounting hole son the base being an good example where you would have to add them up as you go along. More so if using a DRO it would be better to just have them all set out from a 0,0 datum which I would probably place along the ctr line you show in Y and the ctr line of the crankshaft in X

However the downside to this method is that it often takes up more paper space so can be a problem if you want to keep the page count down or work to sizes of sheet to fit a magazine etc where if you are not carful the scale has to be smaller to fit everything in.

I have done a lot of work for the automotive industry and I just trolled through several block and head drawings from several manufacturers - I would like to post one but I would be violating lots of rules if I did so.

I can say that the styles are all over the place and have changed considerably over time with greater use of ISO and tolerancing standards.

Another thing is that they are usually extremely cluttered and nowadays most refer to the 3D model as being the only source of "truth"

Some use string dimension or more commonly dimensions to datum where a dimension is indicated with only an arrow point vis :-

Usually the datum point is a reamed / precision tooling hole and/or face. In the above case 400mm to the left and 200mm below the leftmost hole.

Regards, Ken I
The datum lines reference sounds like a good idea.

I literally just drew the green twin drawings with no thought at all about how engine drawings are generally done, since I have no idea how engine drawings are generally made.

This is some good feedback.

You jumped in & I was going to edit in more :-

Generally you need to avoid double dimensioning because of differing tollerance stack up (none in examples) - however you might , as you stated want to know these values without having to add them up - you can get around that like this :-

Where you can resolve the duplication by calling it a "reference" dimension (Ref. for short) which means it is untolleranced and simply for reference .

It could be that the 40's are important and then the 120 would simply be for reference etc. etc. (you obviously wouldn't bother with such a simple example - but a string of complex numbers and decimals ??)

I've been designing machinery for 50 years and I do it any way I please - particularly since its mostly me or my staff using them - but you always have to bear in mind it still needs to be clear to an idiot (customer) and all the adherence to styles, rules and geometric tollerances won't save you from that.

For modelling it generally only need to be clear to you - that said - recently a customer ordered a new machine I designed and built 22 years ago - looking at the drawing I wondered what the hell I'd been smoking at the time.

Regards, Ken I
The convention that I learned is that reference dimensions are simply placed in brackets (parenthesis).
Ideally, the drawing should allow the part to be machined ( and inspected) - without recourse to calculations, tables, inferences or guesswork. This involves thinking about how you would make and measure it. For example, I like to put tapping drill sizes on the drawing.
When you include draft angle in the 3D model, that can complicate the dimensions too.

"I've been designing machinery for 50 years and I do it any way I please...."

I like that attitude.
I have also been told many times "this is the way you should do things", only to discover many times that there is a much better way to do things than what many folks will tell you.

That being said, I do value the feedback, and it makes good sense.

I always feel there is room for improvement.

Last edited:
Another thing I have noticed with my engine work is that I can get so immersed in the project that it becomes difficult for me to see what is going on (the proverbial "can't see the forest because all the trees are in the way"), whereas someone who is not familiar with the build may spot all sorts of things very quickly.

My main objective is to be able to produce the part without having to use a calculator to determine the position. That frequently violates the rules about double dimensioning. The use of DRO makes many dimensions incorrect according to the rules and if I am going to be using the DRO I frequently use ordinate diminutions because that is what I am going to be using. Frequently having an additional dimension is also useful as when a slot may start at x=2 and ends at x=6 but I may also want to know that the slot is 4" long. That being said we are producing drawings for our own use and other hobbyist, not NASA or even GM standards. Frequently I will produce a CAD drawing using the correct method and then make a secondary drawing which I only use during manufacturing.
You have to understand that some drawing conventions were adopted when all drawings were done by hand, triangles, T-squares etc.i served my apprenticeship and learned drafting when the conventions were still in place. Double dimensioning was only used, occasionally, when working on a large multi-sheet drawing and it took time to go back and find the original. Double dimensioning took more time to draw. Other conventions like hidden line use changed when electronic drafting came about, CAD. Ther were reasons for all the drafting rules. I worked for Ford and our drawings were done on standard rolls of frosted acetate which were 40 inches wide and sometimes up to 6-8 foot long. Engineering drawings that came from Detroit were 40 by whatever length it took to complete the drawing. Each panel was a standard size and labeled as sheet 1-2-3 etc. When I make my 2d drawings in Autocad I spend the time to use the proper conventions because I'm not working for someone else.
I have a friend with a machine shop and he works to other people's drawings and some of them are appalling, in terms of being correctly drawn. It's easy to miss dimensions in CAD because you just draw the part to a dimension but forget to add it.
I would say your drawings are very satisfactory except for 'stacking dimensions.' In manufacturing 2 place and 3 place dimensions have implied tolerances and stacked dimensions with low tolerances could lead to overall size discrepancies.
Before ordinate dimensioning a part drawing might have multiple extension lines all stacked one over another but with ordinate or datum dimensioning the drafting time was reduced and the drawing was easier to read. Not to mention it fit modern machining setups.
Last edited:
@GreenTwin , I like the look of your drawings and enjoyed them in LSOR as well. They are excellent.

I follow some basic rules when I'm drafting for model engineering. None of these are hard and fast, but may give some insight into how someone else "does it"
1) Assume the person reading the drawing has no formal training in reading engineering drawings.
2) Assume the person reading the drawing has no formal training in machining.
3) The drawing should be clear, organized, and pleasing to the eye.
4) Include more information on the drawing than would be expected in "industry". For example, I may list a hole as "Drill #50 (0.070), 0.25 deep. Tap 2-56." rather than "2-56 0.25DP"
5) Dimension a part the same way it will be machined. (Typically starting from the far-left edge to the bottom right so that the dials on the mill do not have to be reversed, always turning toward the positive.)
6) Never over-dimension. For a string of dimensions, do not add an overall dimension unless the last (and preferably least important) dimension is left off.
7) Tolerance appropriately. I use the number of decmal places shown method (0.0, 0.00, 0.000) unless there is a specific tolerance needed.

Clarity to the intended audience is the key and I think you nailed it. I believe ASME Y14.5 uses parentheses for ref dimensions but most people don't read y14.5 and most can read "ref". I often redo drawings to include what Todd described, particularly steps 4 and 5.

So for the example below, I'm making the assumption that you've prioritized the part in the following way:
1. With a ground blank, hole edge distances are the most critical feature.
2. Overall length is more critical than distance between Ø1/8 centers.
3. Overall length is not as critical as hole edge distances since it goes to two decimal places.

I'd double-check the part's function in the assembly to make sure my assumption is correct, and that I can machine it indicated from only one end. This exercise carried out a few times helps me get a read on the designer on what their intent is when they dimension. Maybe they aren't too worried, and that's okay. Not right or wrong, just gives me a feel of where I need to pay better attention.
(ok I'm back from the drawings)

Looking at the assembly, I might prioritize distance between hole centers after I fit check the bearing supports and ensure the cylinder axes can be aligned close enough. I think the design is very forgiving given all 8 bearing supports can be adjusted along/parallel the crank axis, so it may not matter. I like that everything is in one place, rather than going back to sheet 1 to get a BOM quantity. There's also plenty of room for my own chicken scratches.
The requirements of a model engine drawing to be built by hobby machinists have different requirements than do production drawings. Typical mechanical drawings are dimensioned with the tolerances in mind. For a shaft that typically means that either all the segments of the shaft are dimensioned and the overall length is either left off, or is in parenthesis to indicate that it is a reference dimension, or the overall length of the shaft is dimensioned along with all but 1 section of the shaft.

For a one off hobby engine the requirements are not as strict, and you can make adjustments as you go to make things fit.
I uploaded a .pdf of the original drawings of A.W. Ray of the famous Coles Corliss in a previous post. the drawings (14 or them) are certainly a work of art, perseverence and love of what he was doing. However . . . . The modern standards were not put in place till 1972--almost seems incredulous?! I mean, the war certainly had to have enough standardization that differnt companies could read the same prints and know what they meant.

I tried to find out more about A.W. but could find very little. I thimpfks he lived in Chicago and his wife, Eloda, put a copyright on the drawings. Anyway, regardless of the fact that this work of art is such a wonder, it is absolutely HORRIBLE as a useful drawing. Most of the pages are cramped (done on paper), and way too many have not only double dimmed, but triple and even quadruple! this makes the drawing difficult to decipher, I mean VERY difficult. Some of the drawings so squeeze other drawings that the dims interfere and one dosen't know where the dim goes.

As far as double dimming goes, I do it if I find it is easier to follow a drawing. I also ALWAYs put in an overall dim--that is for the machinist so he/she can cut off a piece of material for the part. I have some disagreements with the standards -- that being one.

Also, I thimpfk to HELL with the g^&dA**ed reference bit. I have one drawing program that positively will not allow that last dim without a lot of trouble then it puts in (ref).

Another thing is that in machining, the machinist is supposed to read the notes before proceeding (I've made several misteaks because I didn't), and often the notes will tell you some titbit that yhou needs to know.

As far as GreenTwins drawings go, they are better than mine (which honestly, I only make for myself even tho' I share them), they are clear, not crampt, and generally look like they are made for someone to follow. If I lookt hard enough I could probably spot some tiny misteak but who cares about a tiny error if you can produce the part?

Notice that in the older mags a ruler was printed on the page and you took measurments using that and about half the dims simply were not there. I have been very frustrated by that as I prefer my drawings fully dimmed so I don't have to do calculations and measurements.

One last thing, most machinists that I have known, write in a lot of information before they proceed, that is, they do the calculations and write the dims in in pencil. So what is the use of a draftsman if the machinist has to do all that? It is MY Not so Humble Opinion that drafting SHOULD put in anything that the machinist might want to know. That is what a draftsman is for, IMNSHO.
Thanks for all the kind comments and helpful feedback from everyone.

I did not know really what I was getting into when I started the green twin, and I sort of learned as I went, and did the best I could with very limited knowledge of model building and casting work.

Pattern making is a whole new ball of wax, and "pattern" drawings have to be made for hand-made patterns, and all of those dimensions are larger than the "as-machined" dimensions.

I know that some of the dimensioning methods I use are directly related to how I make patterns manually.
With 3D printing, one does not need to see dimensions on a sheet of paper, but instead one just needs to know the appropriate scale factor.

And some (most?) pattern making (with the exception of patterns for things like gears) is not very exact.
You use approximate dimensions, and approximate metal shrinkage factors, and you make the castings such that they fall into a usable range (tolerance) of dimensions.

Machining allowances have to be added to surfaces that have to be machined, and these are also an approximate value.

I recall being very nervous about the green twin drawings being published, since this was my first attempt at a complete engine build, and basically everything was new to me.

The publisher insisted on a running engine built from the plans, and so after two successful builds from the green twin plans (the green and gray oscillators), I was confident about having them published, and confident that they were accurate enough for anyone to use for this engine build.

Unfortunately my ignorance of drafting terms is vast, and so I am looking up the definition of "stacked dimensions", LOL, and trying to understand the significance of them, as opposed to "non-stacked" dimensions.

My 2D green twin drawings were derived from the 3D models, and so things like draft angle show up in the 2D drawings.
And so I recall wondering from which surface should I dimension, ie: dimension the top of the engine support bosses, or the bottom?, etc.

There was also the question of do I make 2D drawings as if they were going to be a barstock build, or make them assuming castings would be used? I ended up making the drawings assuming castings would be used, but I have seen one individual make a barstock build from them.

I am still learning, and I have much to learn, but that makes it fun, and I enjoy the hobby.

Pat J
Here is an example I found online of stacked dimensions (not my image).

So thinking back about how I made the green twin drawings, obviously there was a method to the madness, but I am hard pressed to state exactly what that method was/is.

I do recall that I don't like to have to use a calculator to derive dimensions from other dimensions that are on the drawing, especially when making patterns.

I am not sure if I used consistent dimensioning methods from part-to-part.
Sadly I will have to go back and study my own drawings to figure out what I did; I just worked merrily along in my own little bubble, and still do to some extent.

As far as duplication of dimensions, that use to be a very big no-no in the old days of manual drawing, and with 2D engine drawings.

These days, everything I do is generally derived from a 3D model, and so dimensions are basically a convenience thing; the accuracy is never in question. Regardless of how many dimensions I use, or where I use them, they are always correct since they come from the 3D model.
I use reduncant dimensions for convenience, but I understand there may be some issues with this in the greater design world.

Tolerances are a whole different ball of wax, and I basically leave that as a task for the model builder to handle during the construction phase.


As far as decimal places, I used a chart of standard fastener and bar stock sizes, and tried to adhere to standard sizes.
Where dimensions were close to a whole number, I always round to the nearest whole number, so I don't have some sort of odd length.
Sometimes an odd dimension is necessary in order to make all the assembled parts fit correctly.
Last edited:
As I try to think back about how and why I dimensioned things, some of it comes from barstock building methods, ie: showing an overall length on a part allows one to go cut off a piece of barstock material, to begin machining.

I don't like to have to add up a string of dimensions just to cut off a length of rod, and there is the danger of making a mistake when adding a string of numbers, which is pretty typical for me.

Another topic is "fully defined" sketches in 3D modeling.

I have always ignored "fully defined" in my 3D work.

Some say I fully define things by default, just because of how I do 3D modeling, but if that is true, it is not an intentional thing, and not a consciencious thing.

I am self-taught with 3D modeling, and so I learned the things that affected the design, and I ignore the things that have never affected my design, such as fully-defining a sketch.

Everyone has their methods I guess.

As a general statement, I guess I like the looks of stacked dimensions, but that is a very unscientific approach, and is more like an emotional/feelings thing.

I look at drawings sort of like I look at old engine designs; ie: I want the drawings and engine build to look "visually correct", with "visually correct" being defined as looking visually correct to me.

Different strokes for different folks I guess.


Latest posts