HORK Enterprises
3221 Quick Road
Holly, MI 48442, USA
 +1 (248) 328-0231
 consult@hork.com

Analysis based illustratives

Even when recorded at 5000 frames per second and replayed at 30 fps, an airbag deployment only lasts 5 to 15 seconds. That can be too fast for an untrained eye to spot subtle details. Slowing the video down even further, makes that the dynamics of the event are lost. The overall deployment is easy enough to comprehend. The module door opens, the cushion gets pushed out by the pressure, unfolds and pressurizes. Gases exhaust from the main vent holes, evidenced by the plume of white smoke that obviously poures out of the vent.
This airbag is special, however, in that it has another set of vents, visible in the foreground. These vents are initially open, but are closed during deployment. Close observation shows that the vent is initially also exhausing some white smoke, but that this stops when the opening is cinched. If the process is observed even closer, one can see that while the opening cinches, a small hole remains. Just like when you try to close a "draw string" garbage bag and it just won't close entirely. That is, of course, because the bunched up material, around the opening, prevents that. Even though the hole is now much smaller, there is no reason why some gas and smoke shouldn't still be pouring out, right? But, no matter how many times you watch the video, the flow of smoke out that cinched opening clearly stops at some point. If only we could look inside the cushion to see what is going on! And that is where simulation analysis can help, not only to slow things down further, if necessary, but also to view the deployment from all sides. Even from inside the cushion. This simulation analysis is focussed purely on the collapse of that closing vent. On the outside we see the same pattern as observed in the test video. The tube cinches, but a small hole remains open in the top. Then the simulation results show what the test video cannot.
The string that cinches the tube, runs between two layers of fabric that make up the tube wall. On the other end the string is attached to the cushion panel that faces the occupant. As the cushion deploys, the string gets pulled into the cushion by the deployment forces. As it does so, the inner wall separates from the outer wall and gets pulled into the interior of the cushion. That closes off the path for the gas to flow out of the vent. With that it meets the essential criteria of . Imagine you paid extra so your car would be equipped with an optional side airbag and then when the car gets into a serious side impact, the airbag doesn't inflate. Imagine if that happened to your loved one and she was fatally injured. Yes, you are going to want some answers!
In icy weather the vehicle had careened off the road and down an embankment. There it met with a telephone pole. The angle of impact was determined to be 25 degrees rearward of a side impact. A rare angle indeed. The vast majority of side impacts are biased towards the front, between 10 to 30 degrees forward of lateral. The sensitive axis of the side impact crash sensor is aligned with that. That and considerations of inadvertent deployment, explained why the airbag had not deployed. The question then becomes: would it have made a difference, if the side airbag had indeed deployed? That is where crash simulation analysis can provide some insights. Our analyses, with and without side airbag, convincingly showed that the type and severity of the impact was such that the presence of a side airbag would not have significantly affected the outcome of the accident. The hard truth is that with all the advances made in structural design and protective systems, we unfortunately still can't save everybody. VirtualCrash.com combines a fully capable explicit finite element solver, dummy models, vehicle frontal environments, and restraint systems with an easy to use web interface to provide the capabilities, available in the crash-safety departments, to the rest of the world. The portal follows the software as a service (SaaS) architecture, in that solver and models are hosted by VirtualCrash.com and accessed across the internet with no other tools needed than a web browser. There are no upfront costs, there is no software to license or install, and no specialized computer hardware to acquire or maintain.
Building an analysis model with the application is alike to filling out a test request form. The request starts by selecting a vehicle platform, like one would select a buck for sled testing. The selected driver or passenger environment is automatically dressed up with instrument panel, seats, and steering column. A crash pulse can be selected from library or designed with the pulse editor utility. Each crash pulse has predetermined trigger times for belt pretensioner and airbag stages, based on an internal algorithm. The defaults are overridden, simply, by specifying alternates. This theme repeats for other sub-systems. If a study is about whether or not to have a belt pretensioner, the details of the airbag and sensor systems can well be left at the default settings. This keeps the input data set remarkably small and thereby easy to convey to other interested parties, so they may confirm the findings for themselves or use the data to set up alternate scenarios for further investigation.
A case utilizing the VirtualCrash.com analysis tools was described in our .