Project Planet - Notoriously Fragile Axles

[What About The Bee]

This is a two part post about Planet. First, the current state of the CAD model, which shows those parts which I intend to be on my OO Planet. Second, I investigate the rear axle, which was notoriously prone to breaking. 

Current State of CAD

The primary focus since the last update has been the rear axle and how it is connected to the frame.  

The wheels and nominal axles were sketched. As commercially available components will be used for these, no details are needed. Every effort was made to capture detail for the the rear axle spring assemblies and axle boxes. These are highly visible components.

The axle boxes are captured in all four corners by the wheel frames (yellow). The travel of the axle box is thereby constrained to be vertical, while all four ends of the wheel plates are rigidly tied together by the load transfer link (green) at the bottom.

At the top of the axle box is the oil well, for the total loss lubrication system. There are two tiny holes leading from the well to the bearing, but this will be lost when I go to OO, so it is not represented. The well should be visible, and therefore is represented.

The spring assembly (blue) and riding rod go thru the wood part of the sandwich frame. The spring assembly was a ridiculously difficult part to create in CAD. Each leaf at a different radius and progressive arc length, with an elliptical shaped end, at the angle of the terminus of the leaf. I think the effort worth it. It is a highly visible component and will absolutely draw the eye in.

Next up will be the front axle spring assembly, as we can see in this image of the reproduction Planet. Note that the builders of the reproduction likely used genuine spring steel for the leaf springs, not wrought iron, and thus required far less material than Stephenson.

Stephenson's Fragile Axle

When Stephenson went from outside pistons to inside pistons (Rocket-Class to Planet-Class) the plain driving axle was changed to a crank axle. Given the state of metallurgy and machining in the 1830s, it is no wonder that these axles were prone to fracturing. There were also other reasons for breakage, as you will see. Stephenson, in response to this problem, added several other axle supports, such that when the axle fractured, the locomotive would still be supported.

There are SIX, count them, SIX support bearings on the rear axle. The two outer bearings, green arrows, are shown in my CAD model. The yellow arrows point to a pair of bearings around the left crank, the red arrows point to the bearings around the right crank. These comprise part of an internal frame, running between the piston assembly and the firebox. Not only do these members support the rear axles via bearings, but also each crosshead. 

From this, we can infer that the breakage generally occurred at the cranks, leaving the wheels supported by the outside green/yellow and red/green bearing pairs.  

Stephenson's solution comes with knock on problems. All six bearing centerlines must be co-linear. Any imprecision in co-linearity will create an eccentric of that bearing. This will result in lots of friction as it wears in, just before the bearing wears out. All six bearings are free to move only in the vertical dimension, any other direction results in wear and stress.

Further, the axis of rotation for the crank bearings must be parallel to the axle axis of rotation, otherwise stress will be placed on the crank as the driving rods twist left and right to account for a non-parallel seat.

The rear axle is a large piece of metal. It is well over the track guage long, as the axles extend beyond the frames. To machine this object in a lathe, without causing axial deflection due to tool pressure, will be difficult, particularly near the center of the axle . Then, to turn each of the crank bearings, the part must be shifted in the lathe, once for each crank. Maintaining parallelism during these shifts will also be difficult.  

Overall, this is a very significant machining problem on a very large piece of stock. Even today, this represents a very large lathe.

And now the cherry.  

Look at the axle boxes depicted in the CAD and in the Armengaud axle drawing. Axle motion, from side to side, is constrained by the rounded ends of the bearing seat. The axle, as shown by Armengaud, has rounded surfaces to match. By this, we can see that the axle cannot slide to the left or to the right. Each end of the axle is trapped. Each end of the axle has an independent spring. Therefore, each end of the axle can move vertically and do so independently.  With one side up, and the other down, a triangle is formed. This, in effect, tries to lengthen the axle!!! Yikes! All the components along the axle are stressed. The weakest part of the axle are the cranks. As the axle is stretched, the stress will be concentrated right at the bend for the crank. Right where I think it breaks. Right where Stephenson tried to protect the system.

For the Stephenson system to work, the entirety of the axle must move only vertically, whilst maintaining parallelism to the frame. Any angular deviation from parallel results in stress. Stress leads to breakage.  

Having inside pistons had benefits. A cranked axle is not one of them. Just my opinion of course, your mileage may differ.

Bee