What's new
What's new

I need steam engine crank design info


Sep 8, 2017

I'm designing a horizontal / mill stationary steam engine from the ground up, and wanted to know if anyone can recommend a method or a book etc on how to calculate specifications for a single cylinder crank - 200 RPM speed, 4" dia piston, double acting , 6" stroke.

I found some bits of info from old books, but nothing definitive.

Any thoughts very welcome.....

Thank you.
Last edited:
The loading on your crankshaft is going to be so low compared to the strength of material that you make it out of , just draw something up that looks good and you will be fine.
If you can get up close to an old steam engine, measure the parts. Keep your crank dimensions in the same ratio to the bore that you find on the old engine......as long as you will be running a steam pressure similar to the engine you measure (thanks johnO)
single cylinder engine can get stuck at ends of stroke and not be self starting. You need to be able to bar it over center.
Bill D
There is practically nothing to it. 12.6 square inches of piston times 150 psi is 1890 lbs on big end. 1 square inch of good bearing bronze will easily be good for 3500 psi at those low speeds assuming lube in place, so a projected area of 1 square inch of crank pin would be over kill as far as big end bearing loads.

(Great chapter in MHB on the subject)

That is a 1 X 1 crank pin

Babbitt instead of bronze? Shoot for 1000 psi max, which means bigger crank pin

As far as the cantilevered load on such a crank pin? - there is a word for it - NEGLIGIBLE
Just a comment

Though I applaud your wanting to build an engine of your own design, I can pretty well bet that copying ANY of the multitude od designs that have preceded you will result in a better end.

The adage "If you wish to learn something new, Read and old book" applies to steam power more than just about any subject, baring religion.
I'm guessing from the OPs first post that you are interested in building a model of an engine, rather than an entirely new full scale engine design. In that case you probably would want to retain some visual similarity to a full sized mill engine. Not everything scales properly of course, but you would not go too far wrong in adopting the same general proportions as a full sized model.

If I am wrong, and you are wanting to design an new working engine with a 4" bore and 6" stroke and need to work out likely output etc then you will need old and new engineering books and some exposure to thermodynamics. In this case resemblance to an old large mill engine is not important and may actually lead you to poor design choices. We now have better materials, and a much wider choice of bearing types. A high speed engine might worth considering, with full enclosure and a proper lubricating system for example. I applaud you for the ambition in wanting to design something like this from scratch - a most interesting project if that is what you have in mind.

If a model of an old engine, Johnoder has made the point that bearing size is not likley to be a problem, and that is the case for many other aspects of a mill engine design. Mill engines were generally extremely conservative designs that were expected to run for 40-50 years at a minimum. Most of them were routinely overloaded because the mills that they drove kept getting bigger. Our museum has 20+ mill engines like this that still run well and some of them are 150 years old. We constantly have visitors who wish to make models of the engines. They can spend a long time taking photographs and making sketches with all the critical dimensions. Some of the models they have made are truly fabulous, and of course we have some superb models of our own which are on show.

Give us an idea what your goals are and I'm sure you will get plenty of advice.
We have a 5" bore x 6" stroke Orr & Sembower vertical steam engine at Hanford Mills. We ran the steam plant today, and I was called upon to start and look after the O & S 5" x 6" engine while it drove some lineshaft driven woodworking machinery. A new fellow was with me, so I explained the O & S engine to him and got a chance to look at it up close (I usually am up running the3 bigger 10" x 12" horizontal engine).

A few ideas: the O & S engine is a vertical "bottle frame" engine. It's crankshaft and crankpin journals are a good 1 1/2" diameter. The crankpin journal is "square", diameter = length. The crankshaft is counterweighted. It is what is known as a "center crank" engine. This little engine runs at about 300 rpm, using a throttling governor. If you are designing a steam engine from scratch, you need to decide what type of crankshaft you will use: side crank (like the crank on a pencil sharpener or the pedal on a bicycle; or center crank (the kind of crankshaft used on gasoline engines). Each type of crankshaft has its pros and cons.

As Billmac notes, the old steam engines were heavily overdesigned. In part his was due to conservative design practices due to inconsistencies in materials (such as cast irons of varying analysis, different alloys of babbit and varying qualities of bearing due to how they were poured and fitted).

As Billmac also0 notes. give us details of what kind of engine you are wanting to build and what you want it to do. So you want an engine which replicates an old-time engine and looks old, or do you want a working engine to be put to real work with modern design practices used in its design ? What are your capabilities for building this engine ? Are you going to fabricate the major parts from steel bar stock and plate using welding, or are you going to utilize ready made castings for things like the cylinder ? Some very simple engines in the 4" x 6" range have been built using ball bearing pillow blocks for the main bearings, and a sealed double-row ball bearing for the crankpin bearing, con rod milled out of aluminum.

Rather than re-invent the wheel, I can suggest you look up HasBrouck's book on building steam engines without any castings. The late Mr. HasBrouck used to exhibit his engines at Hanford Mills when there were events. He had some good designs using bar stock to make the parts, and one design using a cylinder liner from a small diesel engine if I recall correctly. You may find a design of Mr. HasBrouck that can be adapted to a horizontal engine and works for your purposes. I believe Reliable Steam Engine (google them) has cylinder block castings, as has Rapahannock Boat Works in the range you are looking to build. Using a block casting will save you a lot of work. The rest of the engine, if you are not a purist going for making a new engine look like something from 'way back when, can then be fabricated from steel using welding followed by machining. Even something like a piston need not be an iron casting. What works well for a small steam engine of this type is what is known as a "pancake piston"- a simple disc piston. Rather than make an iron casting, the piston can be made from plate steel, turned about 0.100" undersize of the cylinder bore. The outer circumference of the steel piston is built up with bronze brazing (oxyacetylene torch to lay on the bronze brazing), then the overlay of bronze is machined to final diameter and the ring grooves cut. Piston rings can be purchased ready made from several suppliers. The bronze overlay on the piston provides a good surface to ride on the cylinder walls and will give a very long service life. This was a standard method of rebuilding worn pistons on steam engines run on saturated steam, and I've used it on both a new engine (with a steel piston) and to rebuild pistons on existing engines.

Lots to discuss, and the more details you furnish, the better. Let us know your level of skills and shop capabilities as these will have a bearing on the design of the engine as well. Plenty of good vertical steam engines have been built by people who used the crankcases and crankshafts from old air compressors. Other horizontal steam engines have been built by people who adapted piston type shallow well pumps. Get hold of a large capacity shallow well pump and you will have the crankshaft, crankcase, mainframe , conrod and crosshead guide. Reisdential piston type shallow well pumps are usually on the order of maybe a 3" stroke, so finding a 6" stroke pump may be a losing proposition.

200 rpm on a 6" stroke steam engine is really not that fast a piston speed. However any design should have some counterweighting on the crankshaft and a sufficiently rigid mainframe to tie the cylinder, crosshead guides and main bearings together. This is why a mainframe made as a one piece casting, or fabricated by welding steel together (then stress relieving and machining) is almost mandatory.

Your choice of engine size is such that you can make full-size drawings to check how the parts will fit together and for interferences if some of the moving parts do not clear other parts). Make drawings, starting with sketches done freehand, play with the dimensions, do the math to be sure the dimensions add up, then make full size drawings. After that, start designing the parts. I'd say again: get hold of Mr. HasBrouck's book and it will solve a lot of your issues in designing a simple steam engine.
The last thing I heard that Skinner engine was pursuing before
going out of business is a bit off the wall, but might work here.
They were working on taking Detroit diesel engines, (the 2 stroke series) and replacing the head with a new design
to admit steam.

The Detroit Diesel liner idea would work for a Unaflow type of steam engine. While this seems simple at first glance, the reality is the Unaflow design only works when steam is exhausted from the cylinder to a partial vacuum- i.e.- into a condenser. The only way Unaflow engines really worked when running against back pressure or even against atmospheric pressure (free exhaust to the atmosphere) is by adding what were known as "auxiliary exhaust valves". This complicated things.

Skinner had pockets and ports cast into their Unaflow engine cylinders. If a customer ordered the engine to be run condensing, these ports were blanked off. If the customer spec'd exhausting against back pressure, then the auxiliary exhaust valves with additional valve gear was fitted. On Unaflow engines for dual service (such as running condensing under some situations, and against back pressure at other times), Skinner had control valving to throw the auxiliary exhaust valves in, or keep them closed. It was a complication of what started out as a simple concept- the Unaflow engine being envisioned to have the steam move in one direction thru the cylinder, maintain a higher cylinder temperature, and higher overall efficiency. The piston in the Unaflow engine is quite long, as it has to cover the "belt" of exhaust ports, doing double duty as a piston and exhaust valve. Unaflow pistons were made hollow to reduce their mass.

What the late Mr. HasBrouck came up with was a steam engine of the "counterflow" type- the common design, using a sleeve liner from a smaller automobile diesel engine cylinder. Counterflow engines can use a wide variety of valves, but the most common- particularly for people like the Original Poster in this thread- are the slide valve or piston valve types.

The OP, from his opening post stating he wants to build a "mill" engine. In its most basic form a "mill engine" is horizontal engine with a cylinder having a slide valve, usually a side crank, and the crosshead runs on flat guides on the mainframe. This is about the easiest and most basic design to build. How the OP goes about designing and building a 4" x 6" mill engine from scratch is going to be interesting to follow along with. It is a function of what the OP has for expertise, skills, and resources (welding equipment, brazing equipment, machine tools, etc). 4" x 6" is an engine of big enough proportions to get beyond the "model" size, and is an engine capable of real work. Whether the engine can actually do "real work" is a function of its design. Is it built to pull full load for sustained periods and come on and off load suddenly ? Is the cylinder's steam passages, valve ports, and the valve as well as valve motion (eccentric, eccentric rod, etc) designed to be reasonably efficient in the use of steam ? The fact the OP wants to build a 4" x 6" engine tells me that something like "real work" is the ultimate purpose of this engine- maybe belted to a generator ?

A very practical and sound mill type engine can be designed using steel and welded fabrication for the major parts. It would be a utilitarian engine, not particularly aesthetic nor classic in its appearance, but a solid runner that would be good for real work and long periods of running at full load.
If the OP wants to build a practical steam engine that has some of the appearance of much bigger mill engines, then it might be worth while looking at some examples of portable workshop engines. These were very common and used when a semi-portable source of power was needed for a job. A good example of the use of this type of engine was to drive line boring or cylinder boring machinery to clean up the wear in a full sized mil engine. The cylinders of such an engine could easily be many tons and moving them to a workshop to do a job like this would be extremely costly in time and money - so they were often machined in-situ. In the early days of mill engines there was no electricity supply and the use of electric motors was in any case very rare. You used a little workshop engine to power your boring rig. You coupled the engine up to a nearby steam main in whatever way you could, ran a belt over the flywheel and away you went.

Engines like this were not sophisticated. In fact they were made as simple and cheap as possible. Typically no governor, just a manual steam valve to control the engine. Usually slide valve, with some means of adjusting the valve timing to suit the job and allow the engine to run in either direction. Rudimentary total-loss lubrication - messy engines. You could consider them as being the 'universal motor' of their day and just like the universal motor in a typical hand tool, you didn't care about efficiency, smoothness or long life - just the provision of power to get a job done.

We have an example of such an engine in our museum. We run it at every steam day but it doesn't attract much comment - its just a very simple little engine, which happens to have roughly the bore and stroke that the OP is interested in. Here is a short video of it running.

We have lots of smaller engines and... - Bolton Steam Museum

There have been several models built of this engine, see this thread which gives a complete description of the build process:


Although this thread describes a model, you will find some useful ideas about how you might build a full sized version.

This type of engine is definitely not what I and I think Joe had in mind when we said that it would be interesting to build a small useful engine with modern ideas and materials. If that is what you have in mind a different mindset is needed to come up with a really good design of that kind.
Last edited:

Everyone has been extremely helpful and generous with info and advice - thank you!

I have 2 goals :

1 - build a biomass flash steam boiler to run this engine ( separate project )
2 - build the engine itself.

Basically the plan for the engine is to be run at 200 RPM and to run quietly pretty much all day 24x7 if needed. I live in an urban area so excessive noise is not desirable. I also wanted long term durability and a long time between lube periods.

It will be a fully modern engine, unless there is a significant case for using bronze/babbit bushings instead of roller bearings, I will go that way. As a bit of fun, I considered trying to have a hint of art deco, we'll see how we go.....

I do have questions about making a crank - I dont have the ability to forge a crank, but could make one but dont know exactly how...any suggestions on method? I considered making it out of webs and pins and silver soldering it together but dont know what the limits are for such methods , or should I mig weld everything ( then have to machine it later once its relaxed again)?

I also considered making the cylinder head out of a bit of 6" ID pipe and then silver solder the steam chest to the outside. My biggest concern is getting the right methods to make it. I have a 9 x 20 lathe and a Seig Super X3 mill.

I did modify a 2 stroke lawn mower engine into a uniflow steam engine by filling the transfer ports with high temp epoxy, then making a separate steam chest and valve gear and it ran OK, it was a large oversquare engine and sounded a bit Harley-like when it ran.

Any thoughts welcome.....


A crankshaft can be built from pieces, but accuracy has to be right on the mark. A typical way to build a crankshaft rather than forging it would be to machine the main sections with a reduced diameter and shoulder. This reduced diameter must have 0.000" total indicated runout relative to the outer (main journal) diameter. Turning in a 4 jaw chuck and indicating each time you setup a piece of the crankshaft in your lathe is the only way to go about this. If the crank is to be a center-crankshaft, make both crank webs together. I.E., make each crank web to shape, and drill and ream the hole where the main journal's shouldered portion is to fit. Do not think you can get by with drilling. Either reaming or boring. Once you have established this bore, turn the reduced diameter on the main journal section for a shrink fit. The sections of this crankshaft are going to be shrink fitted and pinned once assembled.

The sections with main journals must have a dead-nuts accurate center at the outer ends, as you will need this to true the crank once it is assembled.

For example, if your main & crankpin journals are 1 1/2" nominal diameter, the shouldered sections would be 1 7/16" (1.4375") nominal. You ream and bore the holes for the main sections in each crank web, then make a "dummy pin" or dowel that is a light press fit in this hole. You dowel both crank webs together temporarily, set them in your mill, indicate off the dowel pin, and then move the table to the get the crank throw distance. You then drill and line bore or line ream the hole for the throw through both crank webs as a single unit.

You turn the shouldered diameters to a shrink fit in the holes in the crank webs, and do not use a dial caliper for this measurement. Telescoping gauges and a micrometer are what it takes. The coefficient of expansion of steel is 0.0000056 inches per inch-degree Fahrenheit. On a 1 1/2" nominal sized fit, and maybe a 400 degree F temperature difference, we are talking about 0.0029" for the amount the 1.500" bore will grow when heated to 400 degrees F in a kitchen range oven.
A dial caliper is not the instrument to take these kinds of measurements for this class of work.

To insure the crank webs are in line with each other, you will also need to turn a longer dowel pin to a "wringing fit" in the bores where the main journal shouldered sections will ultimately go. You shrink-fit one end of the crank pin into one web, then set the dowel pin in the center hole, and get the other crank web hot.
Chilling the pin on some dry ice is not a bad idea and buys you more time to play with things. Once the crank pin and first web have normalized in temperature, the pin will be locked into that crank web. You then take the heated second crank web and drop it onto the center dowel, and onto the opposite end of the crankpin.

Once things have normalized, you heat the whole assembled set of crank webs and crankpin as a single unit, chill the main journals and fit them to the center bores of the crank webs. Once that has all normalized, you set the crankshaft up between lathe centers. See how it runs as you turn it over by hand. Do not be surprised if you have to bump one of the crank webs in relation to the crankpin to bring the whole shaft into true. It is the same process as truing a Harley Davidson built-up crankshaft.

Once you have the crankshaft trued- and you will need it as close to 0,000" total indicated runout as possible if you intend to use antifriction (ball bearings) for the mains- you drill thru the sides of the crankwebs and crank pin as well as the main journal sections. You then ream each hole with a hand reamer (preferably a taper pin reamer), and drive in a tapered dowel pin. File this off flush and stake in place with a couple of punch marks.

No silver soldering, no Loctite, this is how to build a built-up crank without having to turn the journals to true it up.

If you put heat into a crankshaft to join it- such as silver soldering, welding, brazing- you can count on the crank looking like a crankshaft but rotating like a pretzel.

Forget MIG welding for this work. It is TIG, SMAW, or oxy-fuel brazing to build up an engine from pieces of steel. Solid Wire MIG welding is the most over-used and mis-used welding process on this planet. To build up a cylinder from pieces of steel and tube, you need to machine the parts to good fits, machine bevels for weld preps, and do carefully sequenced welds to control post weld stress. I do a lot of this kind of work, and I usually keep an air needle scaler handy and peen welds as I go to relax and relieve things. I also leave a generous machining allowance as welded parts can be counted upon to have plenty of distortion following welding. My other recommendation is to do some post weld stress relieving. My "backwoods method" is to simply heat complex welded parts in a bed of wood coals or in the coal bed of my coal fired home heating boiler. When the part is at a good red heat I keep an eye on it to avoid overheating and burning of the steel, and using the rule of 1 hour per inch of thickness, I pull the part after the appropriate "soak time" and bury it in a bucket of dry coal ash and cinders. A day or two later, I pull the part out and clean it up to start machining it.

A dial caliper and light duty machine tools and a MIG welder are not going to do the job when you are talking about a 4" x 6" engine for real work. Flash boilers are a good idea in theory, but the reality is another matter. A flash boiler is built on the premise that the steam demand (steam used by your engine) will equal the steaming rate of the boiler. IOW, what goes in one end as feed water comes out as steam in just the flowrate the engine needs. It is hard to get this kind of exact balance.
A boiler with some kind of steam drum to ride through transient conditions such as load swings or coming off line with your engine is what is needed. Flash boilers are more suited to liquid or gaseous fuel firing rather than solid fuel firing. solid fuel firing will not respond to quick changes in steam/water flows that would happen in a flash type (monotube) steam "generator". A flash boiler is technically more of a "steam generator". A simple water tube design of boiler with a reasonable size steam drum would be a better idea to pursue. In that way, you have a firebox with some grate area and a flowpath for hot flue gasses to get some good heat transfer on the watertubes. A watertube design also allows for natural convection to provide circulation within the boiler, and the drum provides a good steam releasing area. Flash boilers must have a good steam separator as there is usually carryover of water slugs with the steam.

Biomass is not a fuel with a high BTU content per unit mass. You need plenty of grate area to burn enough of it to make steam for a 4" x 6" engine, and a flash boiler is about the worst match for this type of fuel and application.

Get a good text on boiler design as well as on engine design. Compute the steam flow the engine will need. Solid fuel boilers, hand fired, typically will produce about 7 lbs of steam per square foot of heating surface (this is a rule of thumb for firetube boilers, if I recall correctly). Then, once you have the required heating surface, add a fudge factor to cover load swings and additional loads like a feedwater pump (either a steam pump or a positive displacement pump driven off your engine's crankshaft). Once you have some idea as to steam flow, you work backwards based on your fuel, type of draft, and get the approximate grate area.... and you throw in another fudge factor....

Neither the design of the engine and more especially, the design of a boiler, is anything that you do by reading a few articles and jumping headlong into it. The engine is the least of it. In the worst case, the engine uses too much steam for what it is, runs roughly, or does not run at all. The boiler or steam generator is the real concern. Boiler require some serious engineering effort to design, and a high standard of workmanship and known materials to build. I speak from many years' experience, and it makes no difference if a boiler is relatively small in size or large- the principals that cause steam to be made and the resulting stresses on the boiler's parts are the same, big or small. Workmanship and materials are the rest of the equation. MIG welding on a boiler is off the table. Unless you produce some working drawings and run calculations as to developed stresses, design things with a minimum factor of safety of 5, and choose materials that are right for the temperatures and conditions, and design your seams and joints properly, and then use the best workmanship to build the boiler, you are playing with something that can be worse than fire itself.
It will be a fully modern engine, unless there is a significant case for using bronze/babbit bushings instead of roller bearings, I will go that way.
If you want it to be "fully modern" then you'll drop that idea of roller bearings. Roller bearings suck.
Wokka7777 -

I think you have some contradictions in your initial requirements.

Summarising :
You want a durable engine that can run 24/7 quietly in a domestic environment. It is to be fuelled with biomass and you wish to use a flash steam boiler.

I will guess from those requirements that you also want a complete installation that does not require full-time attendance - i.e. does not require constant hand stoking, boiler control and engine attention.

The engine design is an interesting problem but I would say definitely feasible. You do not say what level of output you require but assuming that it is realistically small I do not think there are any problems that will require any major innovations in design.

The boiler is an entirely different matter. I am not aware of any biomass fuelled flash steam boilers. I would consider a boiler such as this to require a major research effort and with no guarantee of success. To run 24/7 unattended will require an automatic stoker and a control system that can respond quickly to achieve stable operation. It will also require some means of ash handling. Biomass fuel (you don't say what kind?) is likely to vary greatly in calorific value and possibly will need drying and crushing to be consistent in burning. Superheated steam can be used to dry biomass and also to prepare it for burning but that would add another large complication to the project. I think it would be extremely difficult to burn biomass cleanly enough on a small scale to avoid obvious air pollution, but I guess that would depend on the source of the biomass. It will also require some means of ash handling, of which there may be significant volumes.

If you are determined to build a flash steam boiler that can run 24/7 then I think you must go for an oil or gas fuelled system. There are many flash steam boilers that use these fuels and you have a reasonable chance of getting something that works, but perhaps not very well because there are still some quite tricky control problems inherent in the design.

As Joe suggests, you would be much better off going for a conventional boiler. In any case you may well find that when you have done your calculations that the boiler needs to be physically larger than you are currently anticipating. A boiler like this (even if you just bought a ready made model) needs to be regularly inspected and fully insured. This is particularly the case for a boiler running 24/7 in an urban environment. A boiler is just a big bomb that hasn't yet exploded. A boiler that you build to your own design should have a thorough design review from an experienced engineer before you build it, and should be constructed by someone with all the skills and certifications required. That means a certified pressure vessel welder etc.

I'm sorry to be so negative about your idea for your boiler, but I think it would be worse to get into the project having spent time and money to find that your goal is not going to be reached.