With the advent of Europe's NCAP and other automotive legislative testing protocols, pedestrian safety is now an increasingly important aspect of new car design and safety.

With the advent of Europe's NCAP (New Car Assessment Programme) and other automotive legislative testing protocols, pedestrian safety is now an increasingly important aspect of new car design and safety. Jaguar Cars in Coventry has invested heavily in this area, with new cars now addressing these growing concerns. The NCAP legislation falls into two main areas: lower leg injuries, concentrated around the knee joint, and head injuries that focus on the brain.

This legislation assesses the injury levels using mechanical testing of bio-mechanic models on vehicles.

The lower leg testing calculates damage to the knee joint by measuring deceleration pulses, knee shear and knee bend, with limits applied to critical levels of all three variables.

The head testing calculates soft tissue damage by measuring resultant deceleration of the head on first contact with the vehicle, with limits applied to the resultant variable.

As part of a continual design process, Jaguar Cars has a 'Pedestrian Test Facility' at its Engineering Centre in Coventry.

As part of the testing legislation there, bio-mechanic models are propelled towards a static test car at pre-defined speeds to replicate real-world impact conditions.

As Miles Dadson, project engineer at Jaguar Cars in Coventry explains: 'The impact speed must be within a very tight tolerance band.

Non-attainment of the required speed results in incorrect contact energies and therefore incorrect injury calculations.

If the tolerances are not achieved, the implications on vehicle design can be enormous.' Dadson says that testing standards recommend that the speed is measured as 'the speed prior to first contact, whilst in free flight after disconnect from the propulsion device, using two lasers mounted in parallel, 50mm apart, acting as light gates, with a known distance apart using the formula: speed equals distance over time.' However, according to Dadson, this arrangement has several drawbacks.

'First, there are difficulties in laser positioning for true parallel lines.

Second, there are difficulties in laser positioning at first point of contact.

Then there's the susceptibility of lasers to airborne vibration and they are vulnerable to impact damage from bio-mechanic models.

Also, inaccuracies are created by measuring very high velocities [up to 40kph] over very small displacements [50mm].' An operating criteria consideration for choosing a system is that it must be able to measure faster than 1,000Hz.

As an example, an object launched at 24kph (15mph) travels 6.7mm every millisecond.

Therefore, in trying to measure the free flight speed of an object over 50mm, there would only be seven measurements.

This would create an uncertainty in the measurement equal to 30%.

'As all impact displacements must be within a window of +/- 0.1m/s, a two-beam laser measuring system would not be able to define our measurements accurately,' says Dadson.

In practice, most pedestrian test laboratories and certifying agencies use accelerometers to measure the acceleration pulse from the propulsion device, then integrate the resulting acceleration to give speed.

The accelerometer is used as the preferred measuring tool, despite known measuring inaccuracies as the projectile must be measured in free flight.

At this point, several errors are introduced: acceleration pulse; accelerometer drift; accelerometer damping errors; single and compounded double integration errors; and software integration algorithm differences.

At present, continues Dadson, the only way of verifying an accelerometer-integrated velocity, is to use a reference accelerometer.

Other alternatives, he says, include conventional contacting displacement transducers, but generally these become very inaccurate or suffer 'dropouts' above 5kph.

In an effort to increase the impact speed measurement accuracy, alternative measurement principles were explored at Jaguar Cars, with preference being given to a non-contact measurement principle.

A suitable device was identified, namely a Micro-Epsilon displacement sensor, which uses a laser-optical sensor that operates using the triangulation measurement principle to determine displacement, with post-test differentiation analysis to determine speed.

The Micro-Epsilon laser triangulation sensor was the only device in the marketplace that could meet the criteria laid down by Jaguar Cars.

The sensor operating principle is as follows: The reflected laser spot is focused onto a CCD element, where the focal position determines the displacement, enabling an accuracy of 0.01% FSD.

This principle enabled Jaguar to take a measurement along the axis of travel as a linear analogue rather than as a digital measurement, normal to the axis of travel in the case of a two-beam laser system.

The sensor type selected by Jaguar Cars was Micro-Epsilon's long distance sensor, the 'ILD 1800-500', with a displacement of 500mm, operating at +/- 250mm from a central reference point.

The laser device has a +/- 5VDC output at +/- full-scale output, and a frequency response of 2,500Hz.

'At the time, this was the fastest frequency response laser available on the market for the displacement range we required,' states Dadson.

Chris Jones, managing director at Micro-Epsilon UK explains how the sensor works: 'Our non-contact displacement ILD 1800-500 sensor uses optical triangulation as the measuring principle.

The visible laser used is rated as Class 2.

Our optical displacement sensors measure with a large reference distance and a very small measuring spot diameter.

A digital CCD array is used as the position-sensitive measuring element.' As part of the guidelines observed by vehicle safety testing, SAE J211/1, all data has to be collected at 10,000Hz.

The triangulation principle for laser sensor system sample rate across all channels is synchronous.

Data recording is triggered upon rig fire command from the rig control system at Jaguar; this enables a time history to include the acceleration, free flight and impact phases of the test.

Initially, explains Dadson, the laser sensor was tested at Coventry using a linear quasi-static ram, with a conventional, linear potentiometer mounted to provide ram displacement and the laser sensor mounted in parallel.

The linear potentiometer was calibrated at Jaguar Instrumentation Services and given an accuracy of < 0.5%.

The ram was cycled to produce a saw tooth output, with a rate of <10mm/s.

'Correlation between the laser and potentiometer was good,' says Dadson.

He continues: 'To test the suitability of the laser for dynamic testing, the pedestrian facility ram was setup with a target located on the planar end of the ram, normal to the axial beam of the laser, with the laser located at the forward end of the ram stroke, effectively firing the ram at the laser.

The actuator was set to deploy at 11.11m/s and a sample of 30 deployments carried out over three test programmes.

The actual ram deployment speed was recorded as a single attribute value from the rig control system, whilst the laser displacement was recorded as a time history triggered from rig deployment.' The speed was produced by differentiating the displacement.

One benefit of using differentiated speed, explains Dadson, is that only a few values at the end of the ram stroke are required, whereas an integrated speed would require the complete time history.

'The peak laser speed measurement showed good repeatability,' he says.

When filtered with a two-pole, Butterworth 250Hz low pass filter, over the three test programmes, all distributions from the ram accelerometer were mirrored by the laser sensor response.

According to Dadson, the laser sensor measured the displacement of the ram as 36.1cm during fire phase; the actual ram displacement was 36.0cm.

From the tests carried out, the error indicated by the laser was 0.28% from desired deployment speed, the error indicated by the ram control system being 4.17%.

Over the three test programmes, the calculated difference between the ram control system and Micro-Epsilon's laser sensor was 96% of ram value, the ram overstating the propulsion speed by 4%, whereas the actual error was < 0.5%.

When using an adult, lower leg form of mass 13.8kg, an error of 4% with an overstated impact speed of 11.11m/s, results in an undershoot error of 66.8J.

'This energy undershoot,' explains Dadson, 'is compounded by the energy loss as the biomechanic part de-latches from the ram magnets that hold the part to the ram during test setup.

This de-latching loss accounts for a further 0.1m/s.' Whilst the laser sensor showed the ram deployed slightly lower than the required tolerance (ie.

11.11m/s, +/- 0.1m/s), the feedback from the ram control system would induce the operator to reduce the deployment speed still further by a nominal 4%, thereby significantly under achieving the desired speed.

'Historical data from biomechanical impactors show a double integrated ram displacement less than 36cm, the actual displacement is dependant upon impactor used.

This confirms a control system adjustment under-achieves the desired impact speed.

Now that these operational errors are known and quantified, the test protocol has been modified to account for them, and testing is carried out with a high degree of confidence in the accuracy of the free flight speed,' concludes Dadson.

'It is not intended that the laser sensor is used for all deployments, but rather as a preventative maintenance tool used every 20 deployments to check the ram and any drift in the accelerometer control loop.

More importantly, it serves as an independent verification tool for both accelerometers and for accelerometer-derived velocity algorithms.

This will be used between launch rigs, not only for pedestrian impact testing, but also as interior head impact testing for effective impact damping of interior vehicle components.'