This website uses cookies primarily for visitor analytics. Certain pages will ask you to fill in contact details to receive additional information. On these pages you have the option of having the site log your details for future visits. Indicating you want the site to remember your details will place a cookie on your device. To view our full cookie policy, please click here. You can also view it at any time by going to our Contact Us page.

Test bench design tips

04 January 2017

HBM looks at the critical variables in the design of new test benches and explains how to cost-effectively optimise test routines and shorten test times.

With today’s focus on engine development, rolling resistance and energy conversion efficiency, test bench design is critical to bolstering energy and cutting carbon emissions, as this will reduce costs and support manufacturing innovation.
So, when it comes to designing the test bench, what steps can be taken? Firstly, it must be possible to implement test structures quickly. This can be achieved with intelligent sensors and measuring amplifier systems that communicate with each other and exchange data configuration, for example by using Transducer Electronic Data Sheet (TEDS) sensor data detection which offers a standardised method of storing transducer identification, calibration, correction data, and manufacturer-related information.

In addition, the measuring amplifier and control system must be capable of further processing the measurement data in real time so that the test bench can be regulated. It is also essential to capture measurement data at a high resolution for analysis and to be able to save it.

Designing torque sensors
When it comes to design, modern torque transducers must work with digitalised data and at high sampling rates, in order to meet the requirements of the function tests.  Ideally, available output signals should include not only torque, but also rotational speed and angle of rotation, as these measurements are important to calculating power and energy conversion efficiency.

As measurements are often taken under harsh ambient conditions, it should also be possible to convert signals into frequency signals, to ensure noise-free transmission.

When it comes to optimising torque data, a measuring amplifier system must include a series of internal computing channels which are specifically designed for operation and use of torque transducers.

This should include a 21-point linearization of the characteristic curve of the transducer, which will improve the raw signal and enable it to be further processed to increase the measurement quality of the test bench.  
Other ways of scaling is the use of polynomials and straight pitches, especially with the use of polynomials scaling, which can increase efficiency as they represent the sensor characteristic with greater accuracy.

To increase the accessible accuracy of torque sensors, calibration equipment can also be used to capture the behaviour of the sensor under various loads, which include dynamic right and left rotation, as well as highly accurate measurement in partial ranges, which is necessary to capture the residual breaking torque, in which different applications are measured during the calibration of the sensor.  

Parallel to this, another important function to consider is independent processing of raw measurement value – such as filtering – which makes it possible to adapt the signals for regulation and automation of the test bench.

Due to the work cycle with compression and expansion in the individual cylinders and the corresponding fluctuations in combustion, the torque generated by an engine exhibits highly dynamic behaviour and, in many measurement systems often appears as ‘noise’. In order to eliminate this, a Centre for Advanced Studies in Measurement and Assessment (CASMA) filter angle – synchronously can be used.

Additional functions to bear in mind include the ability to determine peak values or mean values of measurement signals to verify and document test limits. These control values can then be monitored in turn with limit values or tolerance bands in real time, making it possible to control the test bench.

It is also worth noting that if the raw values of the torque measurement with torque and speed are available, they can be used to calculate and output the application of torque in real time, using mathematical computing channels.

Finally, test signals make it possible to test signals and system states, as well as functionality capability during start up, without even having to place the test bench completely in operation. Simply done on the sensor side, by activating a simulated sensor called a shunt signal, this enables the torque transducer to emit 50% of its nominal (rated) signal, enabling the function to be tested in a dry run.

When it comes to performance features of a data acquisition and automation system, the range of measurement signals to acquire is extensive and runs from simple signals acquired at a low frequency, to complex measurement data which has to be simultaneously measured at a high measurement frequency. Decisive factors here include not only durable and accurate sensors, but also robust and accurate measurement acquisition.

Both should be in the same accuracy class and should be at least 0.1%, or better still 0.01%. The sampling rate of the signals is as important as measurement accuracy. It should be high enough so that fast or small partial changes can still be reliably resolved and displayed. In order to cover acquisition of peak values, computing speed and regulating quality, all measuring and computing channels must be sampled in parallel at a rate of at least 20 kHz, which is equivalent to a measurement and calculating grid of 50 microseconds.

PC-based –v- embedded 
There is a basic distinction between PC-based data acquisition systems and embedded control systems within the measurement system. This applies to acquisition of measurement data control, regulation and visualisation. If regulation requires high real-time deterministics, embedded systems are a better choice because data, while quite small, is also time critical. However, regulation in real-time cannot be implemented on PC-based systems, mainly because resources are distributed uniformly over all components of the PC and control tasks are not processed in real time and must wait in some cases before they can be executed.  

In terms of storing results, if only the end results of the test need to be logged or stored, a measurement device with an embedded acquisition system will suffice.  However, PC systems are able to store and manage larger amounts of data due to bulk storage options and, in this case, data acquisition software can record the data on a PC in parallel to measurement and control operation. 

It’s safe to ascertain that the ability to measure torque, rotational speed, angle of rotation and the qualities derived from these variables is increasingly critical in the design of new test benches for use in industrial environments. Likewise, when selecting and designing test benches, higher requirements for accuracy and speed, automation, and efficient operation must be considered.  

Overall, the general trend seems to suggest that conventional measurement technology systems and automation solutions are moving ever closer together.  Thanks to modern and powerful torque sensors, which can be combined with embedded systems that feature open communication interfaces, systems are now suitable for high-quality measurement and regulation tasks.  In addition to controlling the measurement sequence themselves, this type of modern system can also control machines and implement modern, forward-looking and innovative test benches, which in turn, can result in valuable cost and time savings.

Print this page | E-mail this page