Star Ford

Essays on lots of things since 1989.

PRT switch design

on 1999 January 25

This paper presents a PRT switch design and notes on a black box design methodology.

Reasons to defer design choices

The best PRT system will be one that is flexible and changeable, allowing many different types of vehicles, and making it possible to redesign parts of the system piecemeal, while maintaining compatibility with older components. For example, someone ought to be able to invent a better vehicle and run it on the existing tracks. Or, someone ought to be able to invent a better track that can carry the existing fleet of vehicles.

In any system, the interfaces between parts must be the most stable specification, while the parts themselves can be independently improved over time. The most important interface in PRT is the interface between the guideway and the vehicle. This makes the guideway geometry the most important design element in the whole system. The guideway geometry must be carefully thought out, in order to have a design that is durable over generations of possible changes in the engineering of the other parts.

In fact, it may be possible to develop and publish a fully engineered guideway geometry while deferring major decisions about how the other parts of the system work. I would propose that a consortium of US states and/or cities fund such a design initiative, on an even per-capita funding basis, such as $1 per person. This could raise about $1 million. The results would be placed in the public domain and this small public initiative could spur the development of detailed propulsion, guideway construction, and vehicle construction efforts by private companies.

The flexibility of the car & road system is one of the main reasons it has done so well compared to other forms of transportation. A car can be completely redesigned without upsetting the car/road interface, because the interface is so simple. The car/road interface for a highway specification only requires certain features like lane width and grade and roughness to be within certain parameters, but it does not specify any power source or ownership structure or speed limit: all these can be changed. The same goal of flexibility should be built into PRT by deferring design choices to the extent possible, and keeping the interface as simple as possible. The design elements that ought to be deferred include:

  • power source
  • propulsion method
  • vehicle size/weight limits
  • number of bogies (wheel sets)
  • control system
  • curve radius and speed

There will be needs for variations in all of these over time, and even at the same time in different places. For example, in an urban area, the transit system would use a linear induction motor, with publicly-owned “dumb” vehicles with a 1-ton weight limit. This would allow close spacing and would optimize the price by making the vehicles as cheap as possible. At the same time, a guideway might be built over a mountain (where road building would be too expensive), and this guideway would be un-powered, and use smart self-propelled, mostly privately owned, vehicles. In this situation, optimizing the price favors making the guideway as cheap as possible since the traffic would be sparse. In a third environment, a large industrial park might decide to use freight pallets with an 8-ton weight limit that are controlled by drivers, and use a linear motor.

In these scenarios, not all vehicles are allowed on all guideways, but some vehicles can use multiple guideway types, making the system connected. For example, the rural vehicle could be equipped to turn off its own power and switch to city mode, making it a more versatile vehicle.

Requirements of the vehicle/guideway interface

The remainder of this paper contains some design ideas for the guideway geometry of a suspended PRT system. The aim of this design is to demonstrate that all the necessary elements of the geometry can be designed without needing to nail down any of the choices that should be deferred. As I see it, geometry must do the following things passively, by virtue of its shape:

  • prevent the vehicle from falling out of the track
  • have fail-safe switching (i.e. on control failure, the vehicle passively goes either left or right but does not crash)
  • allow for banking

The main design problem is the switching, specifically how to allow the right-side wheels to pass over the gap during a left turn, and vice versa. This is not as hard a problem to solve if you set limits to the curve radius and speed in curves. For example, if you specify that curves out of switches must be fairly tight, then you can design the wheel width and axle-to-axle distance such that at least one wheel is always on solid ground while the others are passing over the gap. However, if you defer the choice of the curve radius and speed as I argue above, then you need a more general way to solve the problem. In the case where both branches of a switch are very slight curves designed for high speed, the gap widens over several vehicle-lengths, making it impossible to guarantee that any of the right-side wheels would be on a running surface in the switch. Therefore the guideway has to be designed so that it can passively support the vehicle from only one side, and while so doing, the vehicle cannot fall out under any load conditions, including side and upward loads.

Sketches of a design that meets these requirements

This sketch shows a design that I came up with to meet the requirements outlined above, while deferring as many other choices as possible.

If a public initiative were to be undertaken to engineer a flexible guideway geometry, they could begin with input like this one, and ideas from others, as a starting point for a fully engineered design. This design is only an idea supported by intuitive analysis.

The guideway cross-section is barn-shaped with a gap at the bottom. The vehicle is attached to the wheels through the gap in the guideway. The vehicle weight is carried on the two large wheels, which run on the horizontal surfaces on either side of the gap. The bogie (shaded region) might be about 12-18 inches long – on my monitor it is displayed at about half its true size. The size should be minimized, but its lower limit is constrained by the durability of wheel materials and heat dissipating capacity in general.

There would be at least two bogies per vehicle, which are vertically stiff, but can freely swivel left and right independently of each other. The vehicle can also freely swing left and right – although this motion should be damped with a shock-absorber – and the swing hinge axis has to be below the swivel mechanism to maintain vertical rigitity.

The switch guidance wheel is mounted higher than the four fixed guidance wheels. The switch guidance wheel is mounted on a rod that rocks to a left or right position.

This drawing shows the rod in the left position, and the dotted ghost of the rod in the right position. You are looking at a cross section of the track on the right, and the vehicle bogie on the left.

In the spirit of deferring choices, the switch rod could either be actuated by the vehicle, and/or be thrown by small arms obstructing its path in the guideway. In the latter case, a small arm would be mounted on both sides of the guideway in advance of each switch. If the guideway intended for the next vehicle to go left, the right arm would extend into the path of the switch wheel and the left arm would be retracted. If the vehicle’s switch rod happened to already be in the left position, nothing would happen, but if it was in the right position, it would strike the arm (or “ramp”) and be thrown to the left position.

Sketch S1 shows the arm in the left position in a part of the guideway before the switch starts. The switch rod and wheel should be at rest on the correct side, probably several feet before the switch starts, at a minimum.

The football-shaped plate on top of the switch wheel ensures that if the rod gets stuck in the exact center, it will get thrown one way or the other when it runs into the point of divergence in the guideway. This is not meant to happen, but it could, and since we are aiming for passive safety here, we need a passive solution if it does happen. This is shown in S2.

The ceiling of the guideway divides into two channels before any divergence of the walls. Once this happens, the vehicle is locked into a choice of left or right. In sketch S3, the switch guidance wheel is locked into the left position.

By the way, these cross-sectional sketches do not show the four fixed guidance wheels, which are mounted so that they run on the walls at about the height of the top of the load-bearing wheels.

In S3, the vehicle has moved into the very beginning of the switch. S4 shows the situation a few feet later when the walls have diverged and the gap has widened. In S4, the load is carried entirely by the left wheels and the switch wheel. In fact, if there is a lateral load swinging the vehicle very far to the left, (such as a high wind), then there could be a large load on the switch wheel.

A few feet later, as shown in S5, the guideway has fully diverged into two guideways, and the two branches are banked for the curve. The vehicle continues to be supported by only the left wheels and the switch wheel. Note that in S5 an unusually large load upwards or to the right could cause the central block of the bogie to hit the “ceiling”. If the central block did not stick up so high (very close to the ceiling), then it would be possible to pull the vehicle entirely out of the track. Having the bock shaped as shown is therefore necessary to prevent this.

After the two branches of the switch are far enough apart, everything happens in the reverse order, and the vehicle is back in a complete guideway as shown in S6.

It is assumed that the switches are all designed to be navigated in all directions. So, a merge is just a switch navigated the other way.



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: