First serious steam engine build.

This forum is dedicated to the Live Steam Hobbyist Community.

Moderators: cbrew, Harold_V

zimirken
Posts: 44
Joined: Mon Feb 04, 2013 1:11 pm
Location: West MI
Contact:

First serious steam engine build.

Post by zimirken »

The time has finally come for me to make a real steam engine. I've made several of various types and sizes before, but I've never really run any of them on live steam. I did a little bit of test running on steam with a solenoid valve engine I built several years ago, but there were drawbacks with the valves that made it not run very well, and it wasn't worth the effort to work around them at the time. I've put lots of hours into modeling steam engine parts, but I've always had trouble with decision paralysis, since there are a million different ways to make a steam engine.
https://youtu.be/Etam5XgpWWc
https://youtu.be/3zNtibG-I2I
https://youtu.be/ZXpuH0ZFAvI

Now that I've got access to a CNC, it's time. So far I've decided on a vertical engine using piston valves and viton X rings. The bore is 1.5in, stroke is about 46mm, and the piston valve is 13/16in diameter. Forgive my mixing of units, as I think and design in metric, but mostly have access to standard sized parts in America. I'm leaning towards the hackworth valvegear, since it's simple and it worked well on the very first steam engine I ever built. I'll be using ball bearings on everything, as oiling things does not appeal to me. Ball bearings are dirt cheap these days, so the only reasons not to use them are things like aesthetics or authenticity.

According to my calculations, this should give me approximately 250 watts of power with one cylinder, 60psi steam, and 70% cutoff at 500rpm. This design should also be easy to modify into a two cylinder engine as well. Of course these calculations don't take into account inefficiencies, but usually losses mostly increase steam usage, not so much reduce horsepower.
Image

So I decided that the place to start was the cylinder and valve block. Those are the most demanding part of the engine, as they require the tightest tolerances, and are the part that are exposed to steam temperatures. Everything else can be mocked up using printed parts. It's pretty much all going to be aluminum, since that's what I have laying around. This should be fine since I'm using x rings. Now, the HAAS CNC at work is basically my personal toy, but the lathe in the machine shop is usually always busy. While it's easy to start the CNC and then go work on things, I'd have to use the lathe during my lunch break. So I'm trying to use the CNC and not the lathe as much as possible.
https://i.imgur.com/8D0IPOY.mp4

So I started with the cylinder. I did the cylinder bore on the CNC by using a large carbide insert end mill and spiraling down the bore. I was expecting to have to finish it with a boring bar or on the lathe somehow, but it came out perfectly. I continue to be amazed by the capabilities of this machine, as the bore is very round, with no faceting that I can see, and it has a mirror finish. This is plenty good enough for running x rings, and better than I've seen inside industrial air cylinders. I may see if I can try to anodize it at a later time, as it is a bit on the soft side.
Image

Since the cylinder block turned out so well, I did the valve block next. I made the valve passages small slots, which should work well with the x rings. The finish isn't quite as good on this, but it should be good enough for now. I can always make it better, if it wears out the x rings.
Image

Now the piston valve. This seems like the perfect thing to make on the lathe, but that would be getting uncomfortably close to putting in actual effort, so I gave a go at cutting it on the CNC using a slitting saw. It turned out good enough for testing. I'll likely remake it later.
Image

Image
This is as far as I've gotten with the design so far. I wanted to make sure I could make the cylinder and valve block first. If I couldn't do that, the rest would just be a waste of time. Also, making them first means I can't change their design, so no choice paralysis.

Image
This is how I feel now. I've been into steam engines since I was a child, but I've never really had the time, money, and tools at my disposal to make a real go of it until now.
User avatar
Bill Shields
Posts: 10459
Joined: Fri Dec 21, 2007 4:57 am
Location: 39.367, -75.765
Contact:

Re: First serious steam engine build.

Post by Bill Shields »

you are way out of the velocity range for X rings for the power pistons...unless this thing is only going to tick over in very slow motion...

strongly suggest you think again...
Too many things going on to bother listing them.
User avatar
NP317
Posts: 4557
Joined: Tue Jun 24, 2014 2:57 pm
Location: Northern Oregon, USA

Re: First serious steam engine build.

Post by NP317 »

Interesting work. Thanks of sharing with us.

As for design paralysis, I usually look at historic works to see what has been done before.
No need to change functioning designs that have worked well for many decades.
Just sayin'.
RussN
zimirken
Posts: 44
Joined: Mon Feb 04, 2013 1:11 pm
Location: West MI
Contact:

Re: First serious steam engine build.

Post by zimirken »

Information online as far as reciprocating velocities for o rings and x rings seems to be a bit sparse. The only things I could find are "up to 1.5fps" for x rings and that o rings suffer spiral failures under 1fps. As an automation engineer I'm surrounded by o rings in aluminum bore air cylinders that have millions of cycles on them and they do fine at all sorts of different movement speeds. Of course this is with air at room temperature.

In any case, all these rings are cheap enough that I am able to order different types of rings for testing with plenty of spares. I'm ordering fkm x rings, 90 durometer fkm o rings, and regular 75 durometer fkm o rings. I ordered a bunch of everything in order to meet the $25 minimum for free shipping. I'll start with the x rings and we will see what works best.

My decision paralysis was mostly trying to pick what would work best and still be easy to make. There are many choices that are all historically successful. slide valve, piston valve, poppet valve, eccentric driven linkages, or cam operated. Large diameter short stroke, or long stroke with smaller diameter. The slide valve seems like it would be easier to make on a mill, but the piston valve ended up bieng easier for me. Since the cylinder block was aluminum, I would have had to make a steel plate for the slide valve to run on, and that means the chest walls would be separate parts. That added a bunch of machining complexity compared to the way I made the piston valve block, which was just two setups. Plus using o or x rings, means I don't have to manufacture piston rings or worry as much about tolerances on the valve vs bore diameter, which further simplifies construction.
User avatar
Bill Shields
Posts: 10459
Joined: Fri Dec 21, 2007 4:57 am
Location: 39.367, -75.765
Contact:

Re: First serious steam engine build.

Post by Bill Shields »

sorry...not talking about X vs Metal Rings -> looking at X vs O-rings both of elastimer.

i have tried the X rings and they have been total failures, where O rings seem to last a lot linger -> but both with significantly lower 'squeeze' than shown in manuals.

if it is just a matter of swapping out X for O -> then why not?...

HOWEVER if you need a different groove dimension for X than O -> well....that may be a horse of a different color
Too many things going on to bother listing them.
zimirken
Posts: 44
Joined: Mon Feb 04, 2013 1:11 pm
Location: West MI
Contact:

Re: First serious steam engine build.

Post by zimirken »

Okay, I have the design mostly done. The dockstrader valve gear simulation programs are a great help for figuring out dimensions. The only issue is that there is no piston valve hackworth program. So I have to design the piston valve dimensions with a stevenson valve gear and then use the hackworth outside admission program to figure out the other dimensions. The only thing I couldn't really use a ball bearing for was the eccentric. I couldn't find one the right size at the right price, but that's not really surprising. I have a rod of this weird moly impregnated plastic that will work as a bearing. The hackworth valve gear uses a 5mm rod and linear bearing to replace that square sliding block.

I don't like the way I take off the exhaust on the lower part of the piston valve. I'll probably redesign that later. I had a hard time finding reference material online to see what was done on other engines. The base plate is also unfinished. I'll probably print the first one for testing and then finalize it later for easy mounting and easy manufacturability.
Attachments
stage 2.PNG
User avatar
Bill Shields
Posts: 10459
Joined: Fri Dec 21, 2007 4:57 am
Location: 39.367, -75.765
Contact:

Re: First serious steam engine build.

Post by Bill Shields »

been using that weird moly impregnated plastic for best part of 50 years..

will let you know when the first one wears out (if I live that long).

the name: Nylatron seems to come to mind...

is my favorite for crosshead (wrist-pin) type of bushings. Delrin AF is also up there on the list...

you may want to be careful of rocking with your piston, since you really do not have much of a guide there.

Yes...there is the sleeve through which the piston rod travels -> but all things considered...not something that I would do.

If you don't want a guide at the bottom...you may consider extending the piston rod out the top and putting a guide on the other end of the piston in the head (which would be typical of how it is done in Europe)....
Too many things going on to bother listing them.
zimirken
Posts: 44
Joined: Mon Feb 04, 2013 1:11 pm
Location: West MI
Contact:

Re: First serious steam engine build.

Post by zimirken »

Actually, the piston rod is running through a LM12UU linear ball bearing that's inside that green part. Screwing that part down also sandwiches 2 o rings into a slot inside the cylinder mounting plate. The only thing is that steam leakage out this gland will want to travel into the linear bearing. Now that I think about it, I may add a horizontal escape hole and another o ring below it to allow leakage to escape outwards instead of through the linear bearing.

I remembered a chamfer tool I have that allows me to cut the piston valve chamfers on the CNC instead of having to finish it in the lathe. Between that and the slitting saw, I can now go from stock to a fully completed piston valve in only one CNC setup. The run takes less than two hours. This means I'll be able to rapidly make new piston valves if I need to test changes. I happened to have some o rings laying around of the right OD but with a different thickness than the ones I originally planned on, so I made a new piston valve to accommodate them.
20220422_084610.jpg
Stock goes in, parts come out.
User avatar
Bill Shields
Posts: 10459
Joined: Fri Dec 21, 2007 4:57 am
Location: 39.367, -75.765
Contact:

Re: First serious steam engine build.

Post by Bill Shields »

ok...you appear to have the side thrust under control (at least on paper).

linear bearing at the bottom of a steam cylinder...going to be trouble since the O-ring will never keep all the moisture out.

drain hole a move in the correct direction.

stainless bearing?
stainless shaft?

otherwise...issa gonna rust....

how flush water out when done running?
Too many things going on to bother listing them.
zimirken
Posts: 44
Joined: Mon Feb 04, 2013 1:11 pm
Location: West MI
Contact:

Re: First serious steam engine build.

Post by zimirken »

The piston rod is case hardened chromed tool steel. The bearing it doesn't say. It's either going to be stainless or tool steel. It has it's own rubber seal too. This is one of those parts where I had to compromise. I was originally going to use a 1/2in piston rod, but I couldn't find a 1/2in linear bearing for less than $20. Meanwhile 12mm linear bearings are 4 for $10. I couldn't find any reasonably priced 12mm stainless rods though. Ah well, if something doesn't last as long as it should, I can always change the design in the future.

Here's a picture of the current valvegear design. I'll be building the bottom end using 3d printed parts at first to verify the design. One extra advantage of the hackworth valvegear is that all of the parts are off to the side of the crank, which makes a cylinder very compact longitudinally. This makes it easier to add more cylinders in a future design.
3.PNG
User avatar
Bill Shields
Posts: 10459
Joined: Fri Dec 21, 2007 4:57 am
Location: 39.367, -75.765
Contact:

Re: First serious steam engine build.

Post by Bill Shields »

Why do you think you need a linear bearing?

Piece of bronze with a bit of packing has been more than satisfactory for 150+ years.
Too many things going on to bother listing them.
zimirken
Posts: 44
Joined: Mon Feb 04, 2013 1:11 pm
Location: West MI
Contact:

Re: First serious steam engine build.

Post by zimirken »

It lives! It's rough, loud, unbalanced, and angry, the perfect thing to pull me around. I had some issues with the original valve gear design, but luckily it was super easy to print up new designs to test. I haven't gotten my linear bearings in the mail yet, but I printed some plain bushings for temporary use that work well enough. I had enough parts together that I just had to test it, even without the piston rod bearing or even any gaskets beyond the o rings.

https://i.imgur.com/wD92567.mp4

This is with about 10psi I think. It even runs in reverse, and I can turn up the air and notch up and it still runs. The flywheel could be bigger for benchtop running, but it was enough to make it run. It's got a bunch of holes in it for tungsten inserts to add mass around the outside, as I have a bunch of little scrap tungsten chunks. The flywheel was actually from another project and I just threw it on because it wouldn't run without it.
20220429_134005.jpg
You can see that the piston rod bearing is missing. Adding that will likely make it run smoother as the piston rod rubs on the aluminum without it. The piston rod clevis is bigger that it's supposed to be due to a design error I didn't see until after machining it. It doesn't seem to affect the running, but I'd like to remake it so there isn't play in the future. The valve rod just rides in a 5mm long hole in the bottom valve cap. I may add a seal and/or bearing in the future. The valve rod is a little chewed up at the bottom because I thought I'd use the threads to adjust the valve timing, but I don't like how that turned out. I've got plans to make the valve con rod adjustable instead. That's why the valve con rod looks bent, as I heated it up and shortened it.
20220429_134014.jpg
I had to make a few revisions to the valve gear to improve stiffness, as a little bit of bending here has a large effect on valve timing. It probably would have worked if the parts were metal, but that would just be compensating for subpar design. The eccentric is also one of the crankshaft throws, which makes this engine more compact. There is a separate inner bearing that's lightly pressed on the aluminum eccentric, for easy replacement. This is the only spot that doesn't have ball bearings. The PLA+ seems to work well in initial tests without suffering from heat buildup that softens standard PLA. I'll leave it until/unless it fails, then probably try nylon or a hard wood. The eccentric uses set screws to hold on to the crank rods, but the other crank throw is a pinch bolt design. This seems to work very well, while also making it easy to disassemble. There is no counterweight on this revision, but that can always be changed later. You can see a little gray bushing where the linear bearing will be going.
20220429_134018.jpg
I didn't even wait to put gaskets on the cylinder parts before testing! I only made a bottom valve cap, but I will be making a similar part for the top. I want to collect the exhaust, so I'll have to make the upper cap and some adapters before I can really run it on live steam. All the o rings seem to be sealing, although the valve o rings seem to be an extra loose fit, so I may need to remake the valve. The cylinder and valve surface finish is as-is straight off the CNC, with no honing or anything. I currently have buna-n o rings in it, but I'll be switching to viton before any live steam testing. I've got hard and soft viton o rings as well as x rings in my cart.
20220429_134030.jpg
You can see the top of the valve here. This engine was designed to breathe well, for high rpms. According to some valve passage equations I found in a hundred year old engineering handbook, I have enough passage area to go well over 1000rpms.
20220429_134037.jpg
Here's an underneath shot of the crankshaft. The threaded rods I'm using to connect the cylinder and base parts are a little curvy, so I will be replacing them with something else. I might use solid 12mm rods with the ends threaded 8mm, so it's held on the shoulders. I'm thinking about putting springs on the bolts holding the eccentric strap together.
20220429_134044.jpg
You can see the linkage I made to control the valve gear. It's very important to have hard stops to limit the travel on the hackworth. If it rotates too far things will break. This is set up to give me 55 degrees in either direction.

Now that I know it works, I can slow down a bit and work on finalizing the design. The next step is to start replacing most of the printed parts with machined parts, along with prepping it for live steam testing.

Once it's tested successfully on live steam, and all the designs are finalized, I plan on adding a second cylinder. This should be super easy, as everything was designed with this goal in mind. It shouldn't be much more complicated than making a duplicate of most everything and slapping them together with a new cylinder plate and crank base design.
Post Reply