36x60x54 Twin Tandem Mill Engine

[Kawka777]

that’s good advise JasonB. Very sad for that guy that you mentioned. I know a few of my dad’s friends that retired only to be gone a few months later.

I tend to get super motivated for a project so putting it off is not an option. If I start something, it is the only thing I want to do in my spare time.

I was checking to see what drawings I have of the starting engine. Bethlehem steel prints can be pretty convoluted sometimes. They are set up as tracings as the original, with Bethlehem drawing numbers. I do not think the starting engine was made by Southwark because the only reference I have is a pinion gear. None of the Southwark drawings have it. Likely if it was made by a different company, it would have not been in this drawing set. I went through the archive but haven’t turned up anything. Unfortunately not all drawings were transferred into the archive, especially if they were obviously obsolete when they were scanned.

I think I will start with the HP cylinder. It had the smallest valve and cross drills. If the valve stems and eccentric scale down okay to whatever scale I take it to, i am thinking it should work.

More to come shortly!

 

[mikehinz]

You can get all the parts in an assembly to move easy enough in alibre.

Jason, may I ask which version of Alibre you use? I'm asking as I have a paid version of Fusion 360 now and I may want to switch when my subscription expires. I've looked through the version descriptions and I THINK I need at least Design Pro vs Atom to have something close to the functionality I have with Fusion. I'm thinking of changing mostly due to the pricing. Any advise you could give on this is most appreciated! Mike

 

[Jasonb]

I use Pro as it has a few things that I tend to use like Draft for patterns. Atom should allow you do move an assembly around like that too. You don't get the ability to have say timing gears running 2:1 in atom but you can in pro.

You can download the Atom or pro trial and have a play for a few weeks to see how you like it.

 

[mikehinz]

I use Pro as it has a few things that I tend to use like Draft for patterns. Atom should allow you do move an assembly around like that too. You don't get the ability to have say timing gears running 2:1 in atom but you can in pro.

You can download the Atom or pro trial and have a play for a few weeks to see how you like it.

Jason, thanks. What about holes and threads? Those are features I use in Fusion quite a lot and that gives you hole call-outs on the drawings, plus ordinate dimensions on drawings. Do you have to have Pro to do these functions? Also thanks for your suggestion of doing the trials. I'll download and try them out. Mike.

 

[Jasonb]

Pro is needed to automatically add hole call outs and to show then as threaded on a 2D drawing. You have the option of the double lines or can have simulated threads, full library of threads to choose from. It will also show a hole with the conical end left by a drill. Though as a lot of mine is just for me I tend to do most holes as sketched circles and then cut-extrude them.

Not sure if Atom has Ordinate, Pro does

There is an Alibre forum if you want to ask anything there. Also one of the Alibre tech help guys will answer on Model Engineer Forum, just make sure you put Alibre into the title of the thread so he sees it. He runs Atom and Pro on his machine so can usually tell you what does what quite quickly.

 

[mikehinz]

Pro is needed to automatically add hole call outs and to show then as threaded on a 2D drawing. You have the option of the double lines or can have simulated threads, full library of threads to choose from. It will also show a hole with the conical end left by a drill. Though as a lot of mine is just for me I tend to do most holes as sketched circles and then cut-extrude them.

Not sure if Atom has Ordinate, Pro does

There is an Alibre forum if you want to ask anything there. Also one of the Alibre tech help guys will answer on Model Engineer Forum, just make sure you put Alibre into the title of the thread so he sees it. He runs Atom and Pro on his machine so can usually tell you what does what quite quickly.

Jason, again thanks for the info! I'll check out the Alibre forms and will try our both using the trail versions. Mike

 

[Kawka777]

Hi all,
Finally got Alibre and started some drawing. It’s not bad. There is a bit of a learning curve for me. It took me forever to figure out there’s two different button for additive and subtractive extrusions!

It became apparent pretty quick, starting with the HP cylinder, that I need to decide/design for either a fabricated, machined, or casted part.

I have some resources that can cast aluminum, but none that can do cast iron. Is there any companies out there that would cast a low volume set of parts?

Other options would be to fabricate from brass and silver solder. No matter if I cast aluminum or fabricate from brass, I have to steel sleeve the bores.

Mike

 

[Richard Hed]

Hi all,
Finally got Alibre and started some drawing. It’s not bad. There is a bit of a learning curve for me. It took me forever to figure out there’s two different button for additive and subtractive extrusions!

It became apparent pretty quick, starting with the HP cylinder, that I need to decide/design for either a fabricated, machined, or casted part.

I have some resources that can cast aluminum, but none that can do cast iron. Is there any companies out there that would cast a low volume set of parts?

Other options would be to fabricate from brass and silver solder. No matter if I cast aluminum or fabricate from brass, I have to steel sleeve the bores.

Mike
View attachment 157832

Yes, I had trouble too, but after I figured outy what the problem was, I fount it to be so trivial, it was a wonder I had the troubler in the first place.

 

[Jasonb]

If you go down the casting route are you able to make the patterns and core boxes? If not then forget casting.

If you decide to fabricate then a mix or brass and bronze will allow you to make it without a sleeve. Using hollow bearing bronze for the cylinder saves on material costs and machining time as well as giving a perfectly usable cylinder.

Have you decided on a scale yet? I see the smaller flange uses 1" studs at 1/16" scale they will be #0-80 is smaller available to you and can you actually tap it? Metric M1 would work at 1/24th scale but it is fiddly and I'm sure there will be smaller fasteners than that somewhere on the engine. What is the flywheel diameter and what will fit your lathe?

 

[GreenTwin]

The way I treat 3D models that I intend to cast is to model the part(s) "as-machined".

Once I assemble the parts, and test the motion/travel, then I add a bit of machining allowance to the surfaces to be machined.

And I add draft angle to surfaces that need it.
Generally I try to build draft angle into the 3D model as I build it, and not add it to the model later, but some times a draft angle has to be added, such as at the exterior of a flywheel rim.

Your cylinder has many curves to it, and so perhaps not much draft angle (if any) would be required.

For coreboxes, I 3D print pattern halves, and add coreprints at the ends.

The valve cages appear to be pressed into the valve chamber from either end, I guess to allow replacement, and to allow machining too.

I can cast gray iron perhaps down to 1/4" thick.
I prefer gray iron to any other material due to how well it machines, drills, taps, wears, etc.

You would have to 3D print the patterns, or at least get the 3D models ready for me to print it.
This part would not be too difficult if resin-bound sand is used for the molds and cores.

I make castings as a hobby only; no commercial work.

Edit:
As I mentioned, if I have to trade off something when scaling an engine down, I trade off internal dimensions, to maintain the exterior shape/scale/appearance.

If the scaled down cylinder or valve walls are too thin when scaled down, then I add material to the interior, and reduce the piston and valve diameter a bit.

.

 

[Jasonb]

That cylinder might be round but no way is it going to be an easy pull. It will need many loose pieces (retracts) as there are a lot of features that stick out such as the large side mounting, flanges to connect the chest, various bosses and the webs around the flange for starters.

You may be lucky casting some of those external webs but they are only going to be 0.072" wide.

 

[Jasonb]

Going to be a fun one to sand the 3D prints smooth on and then work out how many tiny loose pieces will be needed. A lot of the detail still to add.

Though debatable if all those webs are needed on the model. Assuming teh finished engine would have had insulation between the webs then sheet cladding you could get away with a solid cylinder. Though it would need to be smooth* to represent the cladding so maybe not ideal to cast. Doing it solid would also save those tiny less than 0.020" dia screws to hold the cladding on.

If it is a textured cladding as was common on American engines then you have more work to do, this sort of thing

 

[GreenTwin]

Your longitudinal stiffeners are too high on the upwards passage.
They should be at the base of passage.

Most of the cylinder exterior will be hidden by cladding, so not too much 3D print cleanup/smoothing would be required.

If the bore is small, then a lost PLA casting would probably make a lot of sense.
For a larger bore, such as 2", you could get into some bound sand molds.

With all those flanges and such, I would be tempted to try a lost PLA casting regardless of the bore size, if you go the casting route.

Casting anything with a thickness less than about 1/4" could be a problem.
Casting thin aluminum sections is easier than casting thin iron castings, since it is easy to superheat aluminum, but difficult to get much if any superheat with an oil burner furnace.
.

 

[Jasonb]

I only said it was likely to have been clad, without seeing the rest of the drawings and no signs of fixing hole son this one it could go either way.

What stiffeners are you saying are too high? They all run to the edge of the 50" dia end flange.Radially they need a bit of adjustment to suit the 28 hole pattern

Bores at 1/24th scale will be 1.5" HP and 2.5" LP

 

[Kawka777]

HI All,
JasonB, I did go 1/24 scale. There is not a fly wheel on this engine, so I have no issues on swing. This engine is set up like this Weirton Mill Engine that was scrapped a few years ago. I did go with #0 fasteners for most of the below design.

I read alot about making plate patterns for casting last night. It makes alot of sense to have multiple pieces on a plate. I would have to cast the LP cylinder, HP cylinder, the caps, and the crank bed. If I was to have 3 plates in the end to take to some foundry, that would be pretty cool. I designed last night assuming I would be using plate style patterns. I have access to a ton of friend's 3D printers (some with 3ft work spaces) that I can make patterns.

I agree with all of you about webs. There is a cladding drawing that would theoretically cover that all up. It came out so thin I assumed to leave it out. If I went a fabricated route, it would be easy enough to silver solder the webs in... but I really want to go cast iron for the look.

Here is what I came up with. I fear I simplified it too much, so let me know what you all think. I will add fillets in the end once I figure out if I did this in a feasible way.

The flange connecting the cylinder to the valve body I assumed would be impossible to mate due to space restrictions, just no way to fasteners on that. This would also cause issues on alignment of the bores, so made this one body. I decreased the number of arrayed holes on all connections to fit the nuts for the #0 fasteners.

I assumed using a core for the steam passage into the cylinder was almost impossible for modern casting, so I added two holes that will be plugged to drill straight through. Interestingly, these features are on the actual one, so it works out. I should add a flat boss where the holes are to make it look intentional.

I kept the exhaust square flanges with a decreased number of fasteners. The actual exhaust connecting pipe is square, so i could either have 3d metal printed, or more likely, fabricate one from brass.

I make the steam HP inlet and steam exhaust to the LP cylinder the same for ease of making flange fittings. They were pretty close on the drawing that it didnt make sense.

The spool valve bore will be a straight bore to accept the balanced spool. I changed the bore to be 3/4" nominal so i can get a reamer that size.

As far as actual casting, Do you think this would be possible using split plates?? Think I just need to add some draft angles on the protrusions and excess materail on the faces to be machined?

I watched a bunch of BlondiHacks videos on her making her engines and it seemed there was a issue with castings with not enough material. In order to get the wall thickness for casting and the having enough material, i would going to go well undersized on the core. This model has 0.25" on the cylinder wall when finished.

 

[GreenTwin]

You have the stiffener up on the side of the passage.
It should be at the base of the passage.
You have it at the red mark.

Edit:
Looks like it varies from side-to-side.

.

 

[Jasonb]

that is why I said they needed adjusting radially to suit bolt holes. It os equal side to side but spaced for either 3 or 4 holes

But academic now as it looks like it is going to be a simplified engine "in the style of" rather than a full scale replica.

As you have drawn it now it would be a relatively easy silver solder fabrication, where would not be that much bare iron to see, cladding would have gone right upto the edges of the cylinder flanges so there are very few machined surfaces on show. End covers may have a bit more on show but could be turned from CI bar.

If you go with match plates then you will have to do all the runners and gating and also make sure your plates fit the flasks that your chosen foundry uses.

No flywheel will give you more scope for size, some of our rolling mill engines have large gears which take the place of a flywheel.

 

[Kawka777]

JasonB,
Yeah, I am definitely not ready for a scale exact model. I am going with a model to fit in form and function to capture the essence.

It is almost a shame to put the lagging on this, it covers all connections and forms a box rather than over the profile. Here’s the lagging drawing for the HP cylinder.

You are correct. This engine has a 5’8” pinion on the crank.

For now, I think I will keep with castings only because I have never designed plate castings. I’m making this a learning experience on so many levels haha.

 

[GreenTwin]

I assumed using a core for the steam passage into the cylinder was almost impossible for modern casting,

The drilled holes are not a bad idea, but keep in mind that while passage cores may be difficult, they are not necessarily impossible.
Generally if the passages get too small, then either an investment or lost-PLA process is used.
If you go with matchplates, then the drilled holes idea is a practical solution.

There is a lot to the pattern making/mold-core making process, and you won't learn it overnight.
You may have to play around with the patterns/runners/risers, etc. to find a configuration that works well.

And you need an exact amount of 75% ferrosilicon added to gray iron to avoid hard spots in the casting, and let the castings cool as slow as possible in the sand mold, generally overnight (between 0.06 and 0.12 oz/lb of iron, low/high range).
Too much ferro and you will have excessive shrinkage and hot tears.
Too little ferro and you will have chills (hard spots) in the thinner sections.

I like gray iron castings a LOT, and would never consider going back to any other metal.

Aluminum 356 castings work for many, and is much easier than iron work.
Aluminum castings are sort of a "foundry-lite" method, and they work pretty well, especially when tempered to T-6.
And you can always use a cast iron sleeve in an aluminum cylinder, which is a pretty good mix between aluminum and iron use.

If you know a foundry/foundry person, than can save you a lot of time, and that can make the difference between failure and success.

A fellow named Rainer Radow has done some complex compound steam engine casting work in iron for a full sized launch steam engine.
Very complex resin-bound molds and cores.

Resin-bound sand is incredibly versatile for casting engine parts, and it is designed specifically for use with iron and steel, although it works well with aluminum too. You can sand/drill/carve/modify/glue resin-bound sand molds and cores after they have hardened.

https://steamboating.de/kaiser/dampfmaschine-kaiserliche-marine-03-09.php

https://steamboating.de/kaiser/dampfmaschine-kaiserliche-marine-03-10.php

https://steamboating.de/kaiser/dampfmaschine-kaiserliche-marine-03-11.php

I think Rainer made the patterns, and had a commercial foundry make his casting.
It took a couple of attempts even using a good commercial foundry, to get a solid casting (as I recall reading his thread).
Somewhat similar to what you are attempting.
.

 

[Jasonb]

Thats considerably larger than what we are talking about here. That cylinder ends up 3.9" long

Cores will be very small section and delicate

Cores and loose pieces (retracts) add a lot to the cost if you are not casting yourself as you are paying for the labour, the iron cost is insignificant. As an example one of teh companies I do patterns for has several engines in their range yet one of the smallest is the most expensive. The reason for this is that the main casting needs several cores and loose pieces which puts the price up close to 10 times what just casting a simple part would.

You would also have to make all the core boxes which like pattern making needs a reasonable understanding of the process.

I don't see a need to core anything on the cylinder except the main bore, It really only has the two slots in the side which can be milled. The valve chest could probably make better use of cores as well as some of the connecting pipework