NOTE: All APT vehicles TILT - Gaps between vehicles and their bogies are large to allow for tilt movement.
THIS GAP CAN
CHANGE TO THAT
If you have to work on a
bogie use a safety chock.
Otherwise KEEP OUT OF THE GAP.
Driving Trailer Second 48106 tilting at Crewe Heritage Centre, August 2018.
If a vehicle travels faster than the recommended "balance" speed (based on rail "cant" angle and curvation) it is said to be travelling with cant deficiency. If travelling slower than the "designed for" speed at a given rail cant it is said to have cant excess. If the body of the vehicle is designed to tilt and to assume cant angles in excess of the designed rail cant it will be able to negotiate curved track at greater speeds whilst retaining passenger comfort i.e. with large amounts of cant deficiency.
The figure shows clearly the "cant" produced in the vehicle body relative to earth i.e. –
Facilitating the rate of progress around curved sections of track, along with passenger comfort at such high speeds, spirit-level sensors on the bogies controlled a tilt mechanism which enabled each vehicle to move up to nine degrees.
The tilt transducer is a bubble type "accelerometer" which contains a glass vial similar to that contained in a spirit-level. Three electrical contacts measure the variation in liquid resistances as the bubble moves under the influence of the forces imparted by the track contour. The transducer is electrically supplied by the electronic system within the tilt pack and gives back the appropriate signal to the pack for amplification.
The basic problem of the tilt system, as with any closed-loop control system, is to achieve fast response whilst retaining adequate stability margins. Also, unwanted disturbances caused by random track roughness must be rejected to safeguard ride quality and minimise power consumption. In solving these problems, the APT-P tilt system, has undergone three phases of development -
Key To Following Diagrams
θi ‑ Tilt Demand, θo ‑ Tilt Angle, A ‑ Accelerometer, LPF ‑ Low‑pass Filter, HPF ‑ High‑pass Filter, G ‑ Gain Stage, HMS ‑ Hydromechanical System, I ‑ Integrator, DFB ‑ Displacement Feedback, VFB ‑ Velocity Feedback, CFB ‑ Complex Feedback.
In the original Mk I prototype system, the accelerometer was mounted on the tilting bolster within the bogie. This provided a mechanical feedback loop to null the accelerometer. A low-pass filter was used to filter out track irregularities. It was found, however, that phase lags were sufficient to degrade stability margins. Hence, gain had to be reduced, so limiting the response capability of the system.
The Mk II system aimed to improve performance by separating out two conflicting problems, that of responding quickly to low frequency tilt demand signals and that of rejecting higher frequency track irregularity 'noise'. A second loop was added to the control system by fitting a tilt displacement transducer and a high-pass filter. The high and low-pass filters together formed a complementary pair, so that phase lags could be counterbalanced. the second loop minimised high frequency movements of the tilt jacks; and a third loop, with velocity feedback, stiffened up the response of the hydro-mechanical system. The Mk II system gave some improvement in performance. However, the lags inherent in the complex feedback function of the main loop caused the stability margins to still be inadequate for acceptable tilt response rates.
The elusiveness of success using a full closed-loop system led to the MK III tilt system. In this system, the accelerometer is mounted on the bogie frame, rather than on the tilting bolster, so that it, together with the low-pass filter, is external to the control loop. Tilt feedback is provided via a displacement transducer. Having removed several phase lags from the control loop, the gain of the system can be greatly increased without prejudicing stability margins. Thus, the required tilt rates can readily be achieved.
As the accelerometer is no longer nulled, advantage has been taken of anticipating changes in cant deficiency. The accelerometer for each vehicle is therefore positioned on the leading bogie of the preceding vehicle (except for the driving vehicle). Tilt system performance is optimised for 200 km/h running, which gives roughly ½ second anticipation. With this advance signal, the system responds with exceptionally close matching between the tilt angle and the ideal.
Comparative track testing of the three tilt systems on APT-P showed that the Mk III system was excellent and by far the superior. The tilt action was smooth and positive, with minimal tilt deficiency through transition curves. Following the tests, all vehicles were modified to take the Mk III tilt system.
Inevitably, an active tilt system gives rise to various failure modes. The design approach for APT has been to revert a vehicle to its upright position if a system failure occurs.
When the Mk III system was adopted, system duplication was partially impaired because of practical difficulties in modifying existing vehicles. Thus, the chances of a 'hard-over' (9°) tilt failure were increased. To overcome this, a supervisory system for detecting tilt failure was devised. A detection unit mounted in each vehicle body continuously monitors tilt deficiency. If the tilt is erroneous by a threshold angle for a set period, the vehicle is actively uprighted. A tilt locking device mounted on the bogie is then engaged, so retaining the vehicle in its upright attitude.