The Birth of a Swan

[maury]

In the not too distant past I was given a special opportunity to develop a set of castings for a very special engine. I discovered through an associate/friend/customer of mine, he had measured and had partially completed drawings for the John Swan 25hp oil field engine. At that time I had decided to make a project of developing castings for this engine, but another project under way caused some delays. Now it's time to get it on the drawing table.

The 25hp Swan is on display in the Coolspring Power Museum at present, and that is the one which was measured. By the way, if you have not visited this museum, you have not seen the
"Grand Canyon" of engines.

In this post I plan to share some of my experiences in engineering a new casting set, pattern making, and various other things one runs across when doing a project like this.

The reason this engine is so special is that it combines elements of a very early 4 stroke engine design with elements of a more modern design. The design is a large frame horizontal side shaft engine with a long stroke relative to the bore. It uses a flyball governor for speed control, and has a cross slide. Unusually, it has both exhaust and intake valves operated by cams off the side shaft, contrasted with a hot tube ignition.

One of these engines has been restored by Craig Pucha. who has done an excellent job documenting every step of his restoration in a web site.
http://www.antique-engine.com/swan/swan25.htm
I don't know if his is the engine at Coolspring, but it seems there are not many of these engines around.

A lot of engineering and some patterns have been completed at this point, but there is a long and interesting road ahead for this project.

I want to give my thanks to Ynze Miedema for the drawings and information he has given me and has worked so hard to produce.

maury
www.lonestarengineworks.com

 

[ShedBoy]

I will be pulling up a chair for this one. Are you planning on doing the casting yourself or are you going to farm that out? Either way I would love to see the pattern making process.
Brock

 

[steamer]

As would I....

Dave

 

[Harold Lee]

Maury - You sure have my attention. This is a process I have only read about at a very high level. To watch it up close is going to be very interesting. Please post often.

Harold

 

[maury]

To start out with when I scale an engine, I look at the big picture. Mainly, what size do I want my final model to be. Then I look at the implications cast on building the model. Since it's a foregone conclusion the model will be made of iron and bronze castings, there are some foundry considerations, for example castings thinner than 3/16" are possible but almost always cause problems. Governors don't scale well, so there are trade offs to make here. The castings have to fit on my machines. Hot tubes may not work well with bore/displacement which is too small....

A lot of these questions were already somewhat resolved in the drawings I received, I just had to look over them to verify the model was buildable.

Beginning, I looked at the side shaft. This is the single most troublesome area in the design, as all moving parts on the engine interface with it except the piston train. Getting the placement of the side shaft relative to the centerline is critical, so that was the first step. Going through the design, making dimensioned sketches of the head, rockers, side shaft bearings, timing gears... I came up with a placement. I designed a set of helical gears to fit. By the way, this engine has right handed helicals, as opposed to most other side shaft engines I'm familiar with having left handed helicals. So then it's off to the races making patterns.

First I went with making the side shaft bearing patterns. There are 3, using the same cap on each. These will be cast in bronze.

 

[maury]

Then I worked on the intake and exhaust cages. THese are similar, so I made one main pattern and modified it to get both shapes. I make rubber master molds from my master patterns. These are then used to make the final production molds.

maury
www.lonestarengineworks.com

 

[cfellows]

Very interesting, Maury. Do you have a sense of how big your model will be, bore and stroke?

Chuck

 

[maury]

I have an exact sense of the stats on my model at this time,
2" bore, 3" stroke, 11 3/8" dia flywheels. The fuel will be propane.

To answer a previous question about doing my own casting, I use a foundry.
It is too difficult to get professional results with aluminum in a home foundry, besides,
it is not an appropriate material for this style of engine. Iron is too difficult and dangerous.
That said, I do cast bronze parts on a very limited basis, and sometimes make prototype
parts from bronze.

maury
www.lonestarengineworks.com

 

[maury]

There are about 25 to 30 castings planned for this engine. The way I make my patterns is to make a master pattern which accounts for shrinkage out of some appropriate material. Mostly depends on the pattern and what tools I will use to make it. THen I make what I call a master mold out of silicon rubber. This mold is then used to make the urethane pattern pieces for the foundry. There is no way to predict what a foundry will do to your pattern, so these individual patterns need to be replacable.

These are the next master patterns I worked on

maury
www.lonestarengineworks.com

 

[maury]

Well, it seems not all my pictures were posted, guess I need to learn how to use this thing.

Here in the master molds you can see impressions of the master patterns. These rubber molds are more or less permenant if properly taken care of. It's a good way to preserve the design.

maury
www.lonestarengineworks.com

 

[maury]

I took a break from pattern making to play with my gear hobber. Wanted to check ou tmy math on making the helical gears. THey are 32dp, 15 tooth on the pinion and 30 teeth on the timing gear. I made them at 63deg. 20 min. helix for the pinion and 26 deg 40 min for the gear so they could use the same diameter blank. Fun stuff.

maury
www.lonestarengineworks.com

 

[maury]

Progress keeps marching on with the Swan. Lately there have been some threads about making flywheels, here's my version.

I made six wedges on my table saw, 60 deg triangles. Then I set them up in a circle, and with a compass I made two circles outlining the rim with some material to spare. After band sawing out the arcs, I glued them to a piece of plywood which had 60 deg lines plotted on it. I then mounted the plywood to and cut out the rim with an RT on my mill. I made the spoke stock with my router, cut them out and glued them in place. THen I added stock for the hub and back to the RT. A little bit of clean up and... enjoy the pictures.

I used cherry wood and a tapered end mill to get the draft. Cherry mills nice and leaves a good finish if you climb mill.

The blue part is the actual pattern which will be mounted to the air set box and sent to the foundry.

maury
www.lonestarengineworks.com

 

[1Kenny]

Hi Maury,

I have been watching and have a seat in the bleachers for this too. Craig Puchas' story on restoring his Swan was enjoyable reading. Your molds look very nice. When you say rubber molds, is it a two part compound? Second is what is an air set box?

Thanks for including us with your build.

Kenny

 

[nsfr1206]

Is that pink mold smooth-on mold max 30?

 

[maury]

All, indeed, the pink stuff is Moldmax 30. I use it for all of my patterns. It is a 2 part silicon rubber, mixed 10:1 by weight. It has excellent definition and release characteristics.

An airset box is what the foundry uses to make my castings. There are numerous molding practices. on the old days all that was available was green sand. That is still used a lot, and is the least expensive way to produce castings. Air set is a method where the sand is mixed with a 2 part binder that sets and binds the sand to a very hard mold. The method requires a machine to mix the sand as it is being used, it hardens completely in 20 min, and does not flow well in just a few minutes. It generates a fair amount of heat, requiring the sand to be dumped out of the mold as soon is it is rammed. I use airset with my molds because it gives a better finish on the castings.

Attached is a picture of an airset box (drag side). It includes the patterns which I make from the urethane, the gating and a place for a ceramic filter. Also, you can see the sides make a box for the sand to create a mold. At the base there is an "upset" board where the patterns are mounted. This allows the mold to be registered very accurately, and prevents glue from entering the mold cavity when the cope and drag are glued together.

maury
www.lonestarengineworks.com

 

[1Kenny]

Boy howdy, Maury,

This is a 1st class education on quality pattern and mold making. Everyone should be following this. I need another soda pop and more popcorn.

Kenny

 

[maury]

Well, all that work on the patterns is starting to pay off a little bit. I did a bronze pour this morning, and made some prototype castings for checking out my drawings and the correctness of the patterns. There are 5 castings for the governor, 6 for the side shaft bearings, and I added the mixer and intake valve cage so the butterfly can be checked out. One of the side shaft castings failed, it was too far from the sprue, will have to work on my gating a bit. I made a casting for the 1/8 NPT street ell, it seems these are getting difficult to find, easy to make.

maury
www.lonestarengineworks.com

 

[maury]

Well, I know there has been a bit of a dry spell since my last post here. I've been busy in the shop and coming up to speed with AutoCad. I have finally completed the first pattern set to be sent to the foundry for iron castings. It includes the bed plate, main bearing caps, 5 castings for the cross slide, head, valve cages and a hot tube chimney. THis will give me enough castings to verify the design for the crank& rod fit, the side shaft fit ( lots of stuff here) the rocker arms, and I can play with the hot tube a bit.

On the Auto Cad side, first pass drawings for the side shaft assembly, governor, mixer, & valve cages as well as the bed plate are done. There is a good bit of work to do on all of these yet, but they are detailed enough to build a prototype.

Now to clean up the shop and prepare for the castings. Machining the first engine will be the fun part.

Pictures are of the first air set boxes and the 5 core boxes. Believe it or not, this represents a lot of work.

 

[maury]

With the pattern at the foundry and waiting for castings, I thought I'd start with some of the parts I have. My first goal is to verify the fitting of the side Shaft and all of the associated parts. This is probably the most critical part. The side shaft is supported by three bearing blocks which must be aligned in all 3 dimensions for non binding operation. Additionally, the design of the bearing blocks makes for precarious setup and machining. I decided to make a jig to aid in the critical alignment parts. The jig is simple, to use in a vice which has been squared up on the milling table. Additionally, the initial machining is aided by the use of two precision vices.

First, (first Pic)I hold the casting in the small vice a precisely as possible to get it aligned with the dimensions, and clean up a surface. This will be the reference surface, almost always a must have when working with castings. The bearing cap, already cleaned up. is then super glued ( I cheat too) on to the bearing block. Bolt holes drilled & tapped.

The small vise is then re-positioned, note keeping things square, then the bearing hole is reamed. (2nd pict)

Now for the fun part ( 3rd pict). Removing the bearing cap and installing the bearing casting on the jig, and installing the jig in the vise at 45 deg. the rear side of the casting is cut and the mounting holes drilled. What the jig does is to hold the part with the reference surface aligned parallel to the x,y on the mill, but also aligns the bore of the bearing with the mills x,y.

4'th pict, the finished part.

The jig, Pict 5. Just square up a piece of stock, ream a hole square with one of the sides the size of the shaft, and mill the end down to 1/2 the diameter of the hole. Make a couple of clamps and add the hold down screws, easy as pie.

Hope this long winded explination wasn't too boring.

maury
www.lonestarengineworks.com

 

[maury]

Got limited out, here's the 5'th pic.
maury