Drop survivability is a critical reliability requirement in many industries. It is also a good way to evaluate if the product meets the compliance standard. A reasonable survivable height is also critical in defining the reputation of robustness. Taking the cellphone/tablet drop for example, typical risks include,
- Housing catastrophic fracturing, cracking or dent
- Cover glass cracking
- Adhesive/fastener failure
- Sealing failure
- Internal module failure due to shock
- PCB deflection caused BGA failure
- Internal system clearance closure
- Camera permanent deformation
Our experts at Apex CAE have extensive experience in drop simulations and offer analysis to revolutionize design starting from architecture planning to successful shipping, feel free to schedule a no-obligation discussion/demo if you have the need.
This article is to provide an overview of typical cover glass failures and analysis methods. Although PSA and LED failure are also common in cellphone drop, cover glass cracking is the most typical failure mode. Accurate prediction of the glass cracking generally includes following steps,
- Characterization of the cracking criteria
- Identification of the critical drop orientations
- Accurate prediction of stress/strain
1. Characterization of cracking criteria
We have published another article (link) on the glass characterization. I highly recommend reading it if you haven’t. Generally, the glass strength can be evaluated based on the statics of surface/edge flaws, or through ring-on-ring, 4 points bend, and ball drop test. Glass is a brittle material, typically it is not strain rate dependent, therefore the quasi-static test results are applicable to interpret drop/impact/shock type of simulations.
2. Identification of the critical drop orientations
It is impossible to simulate all possible drop orientations, because they are infinite. Therefore, we have to focus on high-risk orientations. For cell phone glass cracking, the most sever failure is usually drop at a corner or ‘slap’ (‘slap’ is also called ‘face down drop’). Drop at a corner leads to highly concentrated stress at a small impact area, which would likely cause glass cracking or PSA failure. The failure modes also highly depend on the mounting design. Although it is structurally favorable to mount the glass on a mid frame with a lid, edge to edge mounting is usually preferred from industrial design perspective. Unfortunately, edge to edge mounting significantly increases the chance of hitting the ground directly, so the failure mechanism will switch from deflection to impact.
‘Slap’ is a facing down drop with slight tilt. In physical test or actual drop, a facing down orientation will never be perfect and is usually tilt slightly. The slight angle led to conversion of translation to rotation. As a result, the cover glass experiences multiple impacts with a combination of shock and deflection. Physically testing a ‘slap’ orientation is often challenging as it is almost impossible to create two identical test conditions. The simulation provides a repeatable measure to the mitigation solutions.
3. Accurate prediction of stress/strain
The basis of predicting glass cracking is to accurately capture the stress/strain. The accuracy of a FEA model may be affected by many factors. Generally, to ensure the fidelity of stress prediction, we need to make sure,
- the material properties of components are reliable. That includes not only the glass itself, but also other components in the system. For example, the plasticity and viscoelasticity of the mounting adhesive might affect the load transformation.
- we have good quality of mesh. Typically, I would recommend 5 layers of solid elements across thickness and adding a thin layer of shell element on the surface so that the max stress can be captured. Alternatively, one can also use continuum shell (also called thick shell) which may help save the meshing and computational time. In terms of mesh size in the glass plane, I would simulate and correlate the physical glass characterization condition (e.g. 4 points bend test), and use the same size in the form factor simulation. Overly refining the mesh is not recommended, because the glass is usually defined as elastic material and singularity may lead to over estimation of stress/strain.
- the boundary conditions and interactions need to be defined based on actual mounting. If available, it is recommend to define a failure mechanism (cohesive zone model) for the adhesive rather than letting them rigidly tied together.
Cover glass failure is a complicated subject matter and this article is not intended to cover everything, if you have any comments or questions, please contact us and schedule a no-obligation call.