TechArchive

Original article date: July 2000

Automotive engineers concentrate on developing efficient engines. MILES DADSON, HBM UK’s automotive engineering specialist, discloses some unexpected areas of development

In order to reduce the fuel consumption of car engines, engineers within the automotive industry are trying reduce friction in the cylinder head. Valve gear is of particular interest to the energy efficiency lobby because this part of the internal combustion engine is a major contributor to engine friction at low speeds. In order to compare different designs and various combinations of materials, the cylinder head is separated from the engine and driven by an electric motor. Friction is then calculated by measuring the torque required from the electric motor to operate the valve gear.

In a series of experiments carried out in Germany, at the University of Heilbronn in conjunction with the Steinbeis Transfer Centre, a number of issues have been raised about conventional cylinder head testing configurations. Test rigs that use torsion shafts and supported couplings have shown their limitations over new flange-based transducers.

The first concern is a conventional test rig’s susceptibility to damage by unforeseen loads. The forces exerted by the valve springs give rise to very high alternating torques in the camshaft. In a typical set-up, when testing over the whole cylinder head, the average operating torque will range between 1 and 25Nm, but peak torques of over +/-150Nm can occur due to the valve springs. When the University technicians took measurements with conventional transducers, frequent damage occurred to both the shaft and the curved tooth couplings because the peak torque loading was well outside the shaft’s specification.

To overcome the break-downs, the technicians tried fitting a centrifugal mass on the cylinder head to even out the dynamic loading on the valve gear. This addition smoothed the torque peaks but created more problems. The large rotating weight posed an additional risk to those in the test environment and changed the quadratic response of the measuring equipment.

Even with the centrifugal mass, damage repeatedly occurred to both the curved tooth coupling and transducer shaft. To eliminate this maintenance burden, the technicians needed to develop a tougher set up that could handle the torque peaks.

Simply using larger specification shafts and couplings is not sufficient. While larger units would be able to handle the transient torques the device will be less accurate at measuring smaller changes in torque that are required in different tests. The unit would also be physically larger and may require test bed modification.

As well as monitoring the performance of the complete cylinder head, testing technicians can employ strip measurement, where certain components are removed from the cylinder head. The friction contribution of the removed components can then be inferred by comparing the stripped data to the data collected from the complete test rig. For example, if the camshaft only is measured, without actuation of the valves, then torque values occur only in the range 0.5 to 1Nm, and there are no torque peaks. In these tests the equipment needs to be sensitive to smaller changes in torque to obtain meaningful test data.

The University’s technicians were presented with a dilemma. How to provide a torque measuring device that could withstand the rigours of testing the complete cylinder head at the same time as having something sensitive enough to carry out strip measurements, without changing the transducer between tests? They turned to a new type of flange-based transducer from HBM. The device, a T10F, has a very short overall length, allowing the flange to be mounted conveniently without making lengthy and expensive modifications to the test bed during a test programme.

The torque, speed and position-measuring flange consists of an inner and outer ring connected by four ‘I’-section spokes and housed inside a free standing split-ring stator. The spokes make the device up to 30 times stiffer than the conventional transducers previously used by the University. The T10F can withstand 160% of the test rig’s nominal torque, compared to the 70% offered by conventional transducers.

• HBM

July 2000