Bias Demonstrates Vehicle Test System for Accelerated Durability Testing

Issue 40 - February 2013

Turkey’s land platform variety is expanding everyday with main vehicle manufacturers and their local suppliers. Durable components and systems are required for difficult tasks. On the other hand parts must be optimized for weight, stability and vibration. Development of defense systems whether it is a simple bracket or complete vehicle takes a lot of design iterations and testing. Until final field testing, laboratory tests are performed for optimum design verifications. We want the test to be done in shortest time yet still be as realistic as possible. To keep the test loading realistic, best method is to replicate field loads. This is called road load data collection. It could be acceleration input at the attachment points of the subsystems or it could be loads applied to the wheels or track. Signal processing is necessary to calculate the damage of the road load data and remove the non-damaging sections. Therefore, taking only most damaging sections of the road load data test can be generally between 2 to 10 times. This same technique is also used in air platform components. The difference between land platforms and air platforms is the way to undertake the test. Air platform testing can be done in electrodynamics shaker in frequency domain as the loading is mainly stochastic in nature. However, land platform loading is more deterministic and requires time domain testing. Some tests include vehicle suspension and the vehicle is reguired to be vibrate as it is in the field. This is best achieved by time domain replication of real road load data.

Due to low frequency content and heavy weight of the land platform test articles, servo hydraulic test systems are often used. Bias has developed advanced controller for time replication of road load data on its servo test systems. System calculates servo drive signals by doing system identification. Then drive signals are improved in an iteration process trying to achieve real signals as close as possible.