Fidelis’ FEA services team has helped customers with a whole host of different engineering challenges, utilizing simulation to provide invaluable design direction and reduce product development costs. Here, we take a look at just one example of that, with the goal of highlighting our capabilities, process and, ultimately, the value we provided to the customer.
We all know that simulation is capable of providing insights into a large number of design iterations quickly and efficiently. This analysis exemplifies exactly that – and how the results allow selection of optimum designs prior to any physical testing. Of course, we can’t include the real results of the analyses – NDAs are important – but a simplified version shows what the results looked like.
Problem Statement
The customer (a prominent electronics manufacturer) was designing an integrated electronics box composed of an aluminum chassis and multiple PCBs and electronics chips. They wanted to verify that the package would meet derated temperature requirements for all the chips as well as end-customer shock and vibration specs. The customer provided CAD plus the power output for all thermally relevant chips and the shock/vibration inputs.
Methodology
For thermal analysis, Fidelis utilized a steady-state thermal analysis in ABAQUS. We manually converted the power output of the chips into a body flux based on the volume of each chip. We also included thermal contact conductance based on the junction to board values, which we approximated based on industry standard values if we could not find the specific value. The final challenge was calculating the orthotropic thermal conductivity properties of the PCBs, as they had a much larger in-plane thermal conductivity than through their thickness. Temperature boundary conditions were applied to the base. Convection and radiation did not need to be considered due to customer requirements.
In terms of structural analysis, linear dynamic analysis was utilized, with vibration being applied in terms of a power spectral density (PSD) and shock by way of shock response spectrum (SRS). PSD analysis allows RMS stresses to be generated, which can be compared with the material’s fatigue limit (often at 3RMS, which accounts for 99.73% of stress peaks). Shock response analysis provides an unsigned, positive envelope of peak component stressesrather than a true simultaneous physical stress state, which can then be compared with the material’s yield or ultimate strength.
Results and Conclusions – Our Value
The structural analysis in this case passed just fine. However, after running the initial thermal model, we found that one of the modules had many chips exceeding the derated temperature. We worked with the customer to increase the number and change type of thermal insulating materials to allow the heat to dissipate from the chips into the chassis, which allowed the chips to meet their derated temperature requirements and meet end-customer needs.
TL;DR
Fidelis structurally and thermally simulated an integrated electronics box and was able to help improve the thermal performance without the need for costly prototyping.