Category Archives: CK Holliday build

The swing casting is finished. With it placed in the assembly, one can see just how exactly the swing casting system works:

[youtube http://www.youtube.com/watch?v=clPFcE4FYKc&w=560&h=315]

Here’s a general overview. The swing casting is connect by 4 links, forming something similar to a “swing set” found on kid’s playgrounds. The casting is highlighted in blue.

The cosmetic components (pins, bolts, etc.) have not been added to the model yet, but the mechanical definitions are there, which make the above animation possible.

Thanks to the discussion at Burnsland, I was able to refine the pilot truck a bit more:

The equalizer bars were made to be thinner to allow the bottom of the leaf spring to show. Transverse bar and longitudinal bar remodeled, etc.

Interesting that the pilot only has three longitudinal bars. The side with two bars faces the front… I think.

There is now a discussion on Burnsland about the pilot truck, with rare pictures of the truck detached from the engine.

Some observations I took from the thread:

• Disney used almost no bolted connections on the frame itself! Everything appears to be (heavily) welded. In one way it makes some sense given the scale of the truck, but I wonder how easy it is to perform maintenance. For example, if one of the braces crack?
• The transverse bar has a different profile–more trapezoidal than rectangular. That’s easy to fix, but it also appears to be “doubled”–having an inside and outside pair.
• I can finally see the center swing casting–but unfortunately the photos aren’t clear enough to show how it’s assembled. However, I think I have enough to make a good guess.
• I will revise the equalizer bars to be thinner.
• A good point was brought up that I give a bit of thought to previously. In actual pictures of the engine you will see the bottom of the leaf spring, and in the side profile of the truck I made, you do not. I believe this is attributed to the actual weight of the engine flattening the spring. The spring I model is a static solid body, so it cannot compress.

I’ve never drafted suspension system before and I’ll have to think about which state to show: compressed or uncompressed. The purpose of the drawing should govern first, but there’s a question of faithfulness representation of the engine “as seen” by the user.

As there is so little material out there on the Holliday’s pilot truck, what I’ve made is pretty much a guess of what it would look like based on contemporary sources and a few known major dimensions.

The major shapes and look, I think, are at least correct: the flat equalizing bars, the leaf springs, the axle boxes, the squared pedestals. These parts were drawn from the original Holliday drawing. The truck’s profile shows these components:

Back from vacation now. The lifting shaft (blue) is finished and is now connected to the links via the lifting links.

Now, time for something a bit different. Here’s start of the truck assembly, with the equalizing bars connected to the axle boxes.

In working and viewing the model in solidworks, which shows an isometric view of the model, one can forget the size and depth of the model being built. Today I did a little photo-real rendering test, which gives the image accurate perspective and lighting.

The mid-tone gray coloring is commonly used in rendering tests.

Here’s detailing of the valve gear and link.

Several things to note that are artifacts of rendering: the faceting on circular solids making them appear rough, and the corners that are “too sharp”. In reality, there is no such thing as a perfect corner so these “computer perfect” generated images tend to give that “fake” look. Once the modeling is completely done, I would have to go in and chamfer these corners to allow the “light” to reflect in a more realistic way.

Not much of an update. With the major parts of the valve gear made–slide valve, rocker, link, eccentrics and rods–I am now trying to fit it all together and tweaking the components. It’s tough–add half an inch here, take 3/16 of an inch there, to get the exactly right amount of slide valve motion and lead.

Here’s how the eccentric rods look connecting to the link. Below it is the reference photo from the same angle.

I was wrong! If you look closely when I completed the cylinder and saddle casting, you can see the 1-degree incline known to the Rogers locomotives, of which the DLR #1 and #2 are based on. On the closer look, you can see that although the cylinder is inclined, the saddle is not. I did this intentionally because it made more sense at the time to assemble a flat-decked saddle onto the frame and the boiler on it.

Before:

I just recently came across a drawing showing that the Rogers really did have inclined saddle deck. This was a nightmare trying to incline the saddle with all the ports into, along with figuring out how to cut horizontal surfaces for the frames and the boiler.

After:

One day later, the steam chest and valve rod are also inclined. This results in the rocker arm being slightly shorter than previously designed (6.62in).

Just a quick update today. The steam chest and its packing are in place.

I think the bolts make it look like the Pantheon, with Roman columns surrounding the perimeter.

With transparency turned on, this is a great opportunity to see the inner workings of the steam chest. Here, the 2 valve stem collars at the gland packing are highlighted in blue. The horizontal bolts are used to tighten the exterior collar into the packing, creating a seal at the valve stem.

(The actual packing is not modeled).

Another interesting finding today. Meyer suggests that the pin joining the rocker arm and the valve rod should be tapered to allow the pin to be removed later. But, the CK Holliday photos also show that the Disney engineers use a bolt-cotterpin-slotted nut assembly here, which generally are not tapered. It’s certainly possible to manufacturer’s own tapered bolt, then machine thread the end to receive the slotted nut. But, wouldn’t that make replacement and repairs time consuming?

Anyway, here is the connection. I’ve decided to go with tapered bolt, but also with cotter pin assembly.

In case of a locomotive with short valve rod, the traditional practice is to place a knuckle joint in the rod to allow for additional flexibility. The diagram below shows that the valve rod must be able to move in vertical direction, because it’s connected to the rocker arm which moves in an arc.

So indeed the original CK Holliday drawing shows a knuckle joint just aft of the cylinder.

But in looking at the pictures of the “contemporary” CK Holliday, we see a different connection:

It looks like the flexible knuckle joint has been replaced with a rigid, male-female thread type connection; the male valve stem is screwed into the female valve rod, with a hex nut to tighten the connection. And it appears to be adjustable. Could this be used to allow the engineers to finely tune the valve slide’s travel, by shortening or lengthening the valve rod?

And is it even possible to have a rigid connection here?

Below is a graph of the rocker’s motion through 45 degrees of travel derived through geometry. Red graph is horizontal projection of the rocker’s arm, and the blue is the vertical.

From the valve slide motion study, the maximum amount that the valve needs to travel, in one direction, to allow full exposure of the steam port is 1.13in.

Using the graph, or directly solving, the rocker will need to rotate, in one direction, 7.84 degrees to allow the valve slide to travel the distance it needs. This also results in a vertical deflection of the rod in amount of 0.077in. That’s about 6.88% of the rod’s diameter, or L/401.

So, it certainly works, as the induced stress on the rod and the rocker is very minimal. But, still, it seems like an unnecessary stress.

The computer model needs that rotation degree of freedom to function properly. The above shows that the model is rigid.

Besides, I wanted to do the CK Holliday as it was intended anyway. This will bring the model much closer to the 1955 build. So, back to the knuckle joint!