Machines that are subject to harsh or prolonged vibration present challenges for many components – none more so than position and speed sensors. This article lists 10 simple rules for design engineers when selecting position and speed sensors that must cope with shock or vibration.
There are many examples of harsh shock and vibration environments: off-road vehicles, airborne avionics and mining equipment. There are also some less obvious examples such as pumps and refrigeration plant, where the vibration is less extreme but persists over many years. Of course, characteristics will vary from application to application but generally all environments with vibration or shock can present significant problems for electrical control systems, particularly position and speed sensors.
The following 10 simple rules should help design engineers select position or speed sensors that won’t fail once installed in the field:
1. Use non-contact sensors
Potentiometers are by far the most common form of position sensor but are generally not suitable for environments with either extreme or prolonged vibration. This is because a potentiometer’s sliding contacts wear and so they have a finite lifetime. If we consider a potentiometer with a lifetime of 1 million cycles, for example, this is likely to be fine for a benign application which cycles perhaps 100 times per day because this equates to 10,000 days (or 27 years). However, place the same potentiometer in an application which is vibrating at 20Hz such as an engine or a pump, and the same potentiometer is likely to fail in less than a day. This is because the potentiometer’s contacts will see each vibration as a cycle on a microscopic scale. If the potentiometer is normally positioned at a particular point, the wear effect is accelerated and the potentiometer is likely to fail even more quickly.
Figure 1 – Potentiometers are generally not suitable for harsh vibration applications
2. Damp the sensor electrical output
By definition, the position or speed being measured is likely to be changing at the vibrating frequency (or some function of the frequency). A sensor with undamped electronics will output the measured position and so its output will appear to bounce along at the vibration frequency. However, if the sensor is damped, the sensor’s output becomes the average of its measured position. In some sensors the length of time over which the output is averaged can be varied – from a fraction of a second or many seconds – to suit the application.
If a switch or solenoid is to be activated at some point in the measurement cycle (e.g. a pump switch when a tank is full) then it may be that the switching point is made repeatedly over a short period due to the cyclic nature of the vibration. This will cause the switch or solenoid to open or close rapidly, in turn causing rapid start-stop of the host system. This can be solved by either damping the sensor or by introducing some hysteresis, which will only allow it to switch after a set time.
3. Measure position or speed directly not indirectly
If position or speed is to be measured in a vibrating system, it is likely that different components within the system will be vibrating at various frequencies and amplitudes. Accordingly, it is more important in vibrating environments to measure the position of the actual elements whose position or speed is to be measured directly. This is opposed to measuring position indirectly – say at the end of a gear train or multi-link mechanism. Without direct measurement, measurement accuracy will be degraded.
4. Avoid glass scales for optical sensors
Optical position sensors often use a glass scale through which they shine and measure their light path. In benign environments most optical sensors will perform satisfactorily – provided that there is no foreign matter to interfere with the optical path. However, any glass scale is susceptible to fracture in environments with heavy shock or vibration. Of course this results in catastrophic failure of the optical sensor, with little or no warning.
5. Minimise the weight of the sensors
An often overlooked phenomenon is that damage imparted to sensors is usually not directly due to the vibration itself but rather as a result of the momentum of the sensor’s own components. Minimising weight will minimise momentum and hence minimises the potential for damage. Lightweight sensors are generally less susceptible in harsh vibration environments.
6. Use heavy duty connectors – or preferably no connectors
The single largest cause of electrical failure in hash vibration environments is cables and connectors. Harsh vibration environments are no place for the flimsy connectors normally used on consumer electronics. Instead, connectors should be heavy duty – such as military standard 38999 (shell types) – or at least include jackscrews to bind the connector’s male and female elements. If possible, connectors should be eradicated and electrical interconnections made by direct wiring or flying leads.
7. Potting and encapsulant
An excellent way to mitigate problems due to vibration is to pot sensors and cables into position. There are a wide variety of 2-part epoxies used for electronic encapsulation and these are an excellent method of securing position sensors into the host equipment. This has added advantages of providing a barrier against contaminants and improving heat dissipation at elevated temperatures.
8. Stress relieve connecting wires
Wires and cables tend to be forgotten in stress and vibration analyses but a moving cable is a sure way to generate problems from conductors or electrical joints cracking due to fatigue. Potting is the preferred method of eradicating such problems but alternatives include cable wraps, potting or tightly fitting conduit.
9. Lock any fasteners
This may seem an obvious step but nevertheless it is one that is often forgotten. Fasteners that secure position sensors should be bonded into position with thread lock or, preferably, an anti-rotation fastener such as a tab washer to prevent hex-headed screws from turning and becoming loose.
10. Use caution with magnetic sensors
If a magnetic sensor is to be used then extreme care should be exercised in its selection. Firstly, modern rare earth magnets (notably NdFeB types) are extremely brittle and subject to catastrophic failure if subject to shock. Magnetic reed switches are prone to fatigue over prolonged periods where the vibration causes the magnetic switching vane to vibrate and hence fatigue quickly. Magnetostrictive sensors rely on delicate and precise location of (easily damaged) amorphous crystal ribbons in wave guides. The fixture of the ribbon in the wave guide is susceptible under conditions of either shock or vibration.