Fast but riskyTo avoid these delays, some companies have turned to an alternative responseonly TPA approach, also known as operational path analysis (OPA). Instead of NTFs, this method characterizes the transfer of vibration and noise energy as transmissibility – the ratio of accelerations or sound pressures at the noise source to that of the receiver. Structural measurements are made using accelerometers at attachment points between the engine and the vehicle body. Under-hood acoustic measurements are taken with microphones mounted next to the engine. Microphones at the receiver location (a vehicle cabin, for example) measure the response noise of these combined structural and airborne components of the sound.
The primary advantage of this approach is that it’s fast. All measurements are operational – that is, taken with the engine under load in one or a number of sequential test runs where no mount stiffness data is needed. Response-only TPA measurements are accelerations and pressures of the vibrations and sound levels. As a result, an entire testing cycle usually can be completed in a day – far quicker than the two weeks needed by conventional TPA.
“On the downside, response-only TPA is risky and unreliable,” noted Gielen. “Cross-couplings between accelerations may occur, throwing off some of the attachment point measurements. Moreover, transmissibilities vary significantly according to where the physical loads are acting. As a result, response-only TPA is much less accurate than conventional TPA and individual vibration paths are often falsely identified as critical or missed entirely.”
The biggest drawback of the responseonly method is that – unlike conventional TPA – it does not allow for detailed what-if path analysis to find appropriate noise remedies. Rather, individual components and mounts and rigid connections on the vehicle must be modified and further tests performed until a solution is reached – if ever. As a result, although each response-only TPA test itself takes just a day, the multiple prototype cycles needed in some cases take months to complete and may still be suboptimal starting from a faulty diagnosis.
Best of both worldsTo address the limitations of both methods, LMS International developed the new approach LMS Test.Lab OPAX – a tool with almost the same speed of response-only TPA and nearly the high accuracy of conventional TPA.
“OPAX achieves higher levels of accuracy than response-only TPA because, as with conventional TPA, it uses on NTFs and physical loads in evaluating individual vibration paths to accurately identify the root cause of problems,” explained Gielen. “Likewise, these NTFs and loads also enable engineers using OPAX to study potential remedies by modifying component characteristics in a CAEdriven process.”
Unlike conventional TPA, however, OPAX measurements to determine the NTFs can be made with the engine and mounts intact in the vehicle. The method uses reciprocal measurements in which the target is excited – by loudspeakers inside a vehicle cabin, for example – and responses measured by sets of vibration sensors located on the engine as well as microphones around it.
These measurements can be made straight after completing the operational test run – just like response-only TPA and thus hardly affecting the overall time required. Also, mount stiffness data is not needed with OPAX, which instead computes the information using breakthrough software models to compute stiffnesses entirely from operational data. So-called band estimator models are mostly for representing rigid structural connections – bolted or welded joints, for example – while soft mounts such as rubber bushings are characterized with SDOF (single degree of freedom) models.
“Because mount stiffnesses are included in the OPAX calculations rather than obtained separately – and since all measurements are made with the engine and mounts still in the vehicle – OPAX takes approximately the same time as response-only TPA: about a day,” reported Gielen.
OPAX is not intended to replace conventional TPA. When assessing noise and vibration performance at component level rather than at full system level, conventional TPA is the only option. Also in the case of airborne radiation from panels in reverent environments such as vehicle passenger compartments or machine enclosures, conventional TPA for airborne source quantification is by far the most accurate method.
OPAX in action at Kia Motors
To benchmark OPAX against the other methods, LMS International implemented this approach at Kia Motors by studying a 280 Hz booming noise in the vehicle cabin fitted with a 6-cylinder, 4-liter engine running at wide-open-throttle. LMS engineers were called in by the Korean carmaker because previous attempts by another company using response-only TPA were unsuccessful in reducing the noise level. After three refinement cycles using this approach over a period of four months, the company’s advice to Kia was to redesign the engine – an option which was out of the question because it would cause further costly delays in the vehicle launch schedule.
In addressing the noise issue using OPAX, LMS engineers positioned two loudspeakers in the cabin interior (at the location of the driver’s right ear where the noise was most pronounced) and measured the under-hood response with eight vibration sensors and six microphones.
Color maps of results – with the individual contributions of structural as well as airborne paths – clearly showed
the culprit noise source: an underhood acoustic resonance excited by the engine’s 3rd order unidentified by previous response-only TPA tests. So in just one day with a single test run, OPAX straightforwardly identified the problem – much to the astonishment of Kia engineers.
Indeed, the booming noise was entirely eliminated by a rather simple adjustment to the prototype vehicle – a relatively inexpensive remedy that Kia eventually implemented in their production vehicle.
“Benchmark activity with LMS Test. Lab OPAX turned out to be extremely effective,” said Dr. Ji Hyeon Ki , NVH
Manager of the Kia Function Test Team 3. “Based on the measurements and initial analysis, the fundamental problem became so clear. We were able to design countermeasures and a modified prototype was in place even before the formal presentation of the detailed analysis results was complete.”
“Because of the detailed nature of NTFs in representing these vibration paths and the ability of conventional TPA to analyze and rank the impact of all these contributing paths, the approach is one of the most accurate NVH testing methodsin identifying the root cause of problems,” said Ludo Gielen, LMS Product Management Director. “Moreover, engineers can quickly evaluate potential remedies with computer models, thereby avoiding multiple trial-and-error cycles of modifying and retesting prototype vehicles.”
