Documentation:FIB book/Biofidelic Rear Impact Dummy (BioRID)

From UBC Wiki

Introduction

BioRID II[1]

Due to the complex biomechanical behavior of humans during car accidents, the development of multiple anthropometric test dummies (ATDs), such as the Hybrid III, Biofidelic Side Impact Dummy (BioSID), and Biofidelic Rear Impact Dummy (BioRID) have been necessary. Each ATD works to provide a biofidelic (measure of how accurately something mimics human body responses in a crash) response under a specific crash scenario. A particular area of interest is low speed rear impacts, as due to the viscoelasticity of human tissue, the neck behaves differently than during high speed frontal impacts and therefore not be accurately studied when using the Hybrid III dummy. Low speed rear impact accidents are of particular interest as rear end crashes are the most frequently occurring type of collision[2] and is where the majority of whiplash injury occurs[3]. Additionally, the mechanism of injury for whiplash remains poorly understood and the injury can have significant long term consequences[4]. In order to improve safety mechanisms to reduce incidence of injuries in rear end collisions, the development of ATDs that can accurately represent human behavior in these types of accidents is essential. The BioRID was designed specifically to behave biofidelically under low speed, rear impact collisions, and addresses concerns with the biofidelity of other crash test dummies when loaded in this manner. In this review, the design of the BioRID I, BioRID P3, BioRID II, and EvaRID will be discussed and key differences in comparison to the Hybrid III highlighted. The biofidelity of BioRID models will also be covered in the context of low speed rear impact collisions and compared to volunteer data, cadaver data, and the Hybrid III dummy.

Design of BioRID Models

BioRID I

The BioRID I was the first model of rear impact dummy made and is essentially a modified version of the Hybrid III dummy. It is constructed with the use of Hybrid III arms, legs, head, and slightly modified pelvis. Crucial differences between the BioRID I and Hybrid III dummy exist in the cervical spine and thoracic region[5]. While the Hybrid III dummy has five vertebral elements in its cervical spine, the BioRID I has seven. The main goal of the Hybrid III neck was to replicate overall biomechanical response whereas the BioRID I is more concerned with specific motion. This is why the design features seven cervical vertebral elements, the same number of cervical vertebrae in humans. Similar changes between the Hybrid III and BioRID I are also made in the thoracic and lumbar spine. The BioRID I consists of 12 thoracic and 5 lumbar vertebrae, whereas the Hybrid III thoracic spine is welded and the lumbar spine is made of a polyacrylate elastomer[6]. With the thoracic spine of the Hybrid III welded, flexion and extension motion of this region is not possible.  Flexion and extension of both the thoracic and lumbar spine is possible in the BioRID I as each vertebra is connected with a pin joint. The pin joint also restricts angular motion of the spine to the sagittal plane[7].

Polyurethane rubber placed between all vertebrae is used to control joint stiffness in the spine of the BioRID I dummy. Two separate blocks are included in the cervical vertebrae whereas only one is used in the thoracic and lumbar spine. The first block of polyurethane is used to characterize overall joint behavior and the second to characterize hyperflexion and hyperextension. Additionally, neck motion is further controlled in the BioRID I dummy using wires starting at the skull and running through all seven cervical vertebrae, ending at T1. The wires were used to maintain resistance to flexion and extension while decreasing resistance to retraction motion[5]. As mentioned above, each vertebra is connected using a pin joint. Adjustment to the pin joint angle is possible and allows for the initial posture of the BioRID I to be manipulated. In combination with a silicon thorax, this also makes the BioRID I more conducive to out-of-position testing[7].

BioRID P3

Diagram of BioRID P3[8]

The BioRID P3 is prototype preceding the BioRID II and offers improved biofidelity than that of the BioRID I through updated neck muscle substitutes, softer thoracic spine, softer torso, and a redesigned interface between the spine and torso[8]. Retraction of the neck during impact was further refined in the BioRID P3 to improve biomechanical response. The wires that ran from the neck of the BioRID I and connected to the T1 vertebra were extended to attach to the T3 vertebra. Similarly, the cables extend through all seven cervical vertebrae and through T1, T2 and T3. Other components, such as the Hybrid III arms, legs, head, and modified pelvis continue to be used in this model of rear impact dummy.

BioRID II

BioRID II and EvaRID Spines[9]

The BioRID II is the most advanced rear impact dummy[1]. Differences between the BioRID P3 and BioRID II include further updates to the cables running through the cervical spine. These adjustments were made to make the BioRID II dummy more user friendly and improve its durability[10]. Rather than connecting to the T3 vertebrae, as seen in the BioRID P3, one cable is thread from the skull, through the cervical spine, T1, T2, T3, and around a damper at T4 before connecting back to the skull. Two other cables are also included in the design, which originate at the skull and end at a tensioning device on the right hand side of the dummy’s torso.

Table 1: BioRID II compared to EvaRID and 50th Percentile Female[9]

BioRID P50F and EvaRID FE

The BioRID impact dummy is based on the 50th percentile male[1]. As this is only representative of a small portion of the general population. The first attempt to make a new rear impact dummy fit for a female (50th percentile) population resulted in a virtual model: the EvaRID LS-Dyna Model by Humanetics[11] . This dummy was released by the developers of the BioRID II dummy and is a scaled down version of the BioRID II virtual model to match the dimensions of a 50th percentile female, as described by the University of Michigan Transport Research Institute with a height of 161.6 cm and mass of 62.3 kg[9][12]. The most recent version of this virtual model is the EvaRID FE, with scaled dimensions, but the same stiffness and damping material and discrete element properties as the BioRID male model.

The virtual female model was a good first step, but a biofidelic physical female dummy would still be a very valuable addition in the world of injury biomechanics. Therefore, a group of researchers set out to develop a physical rear impact dummy representative of the 50th percentile female[9]. This resulted in the BioRID P50F, a prototype of a rear impact crash test dummy with a mass and dimensions similar to that of the 50th percentile female. Development of the BioRID P50F was completed through modifications to the original BioRID II. To reduce the seated height of the BioRID P50F, both the L4 and L5 vertebrae from the BioRID II were removed and the height of the sacral vertebrae was reduced. As a result, the angular range of motion of the lumbar spine in the sagittal plane of the BioRID P50F was reduced when compared to the BioRID II. Furthermore, the torso was made smaller, both in width and height and it was attempted to reduce the neck and spine stiffness and damping properties to 70% of the original value[9]. A comprehensive outline of all the specifications and adjustments of the BioRID P50F can be found in Carlsson et al., (2021)[9]. These modifications were based on the EvaRID LS-Dyna Model, by Humanetics, which was as mentioned based on the dimensions of Schneider (1983) [12]. Table 1, from Carlsson et al., (2021)[9], further breaks down segment sizes of the BioRID P50F and EvaRID model and compares it to that of the 50th percentile female and BioRID II dummy. The biofidelity of the EvaRID P50F is further discussed below.

Biofidelity

BioRID P3

In a study conducted by Linder et al. (2018)[13], they compared the responses of the Hybrid III, BioRID P3, and cadaver data during pendulum impacts to the back. In their study, when measuring horizontal head displacement they found that the BioRID P3 stays within the corridors from cadaver testing approximately four times longer than the Hybrid III. When analyzing head horizontal, vertical, and angular displacement relative to the T1 vertebrae they found that overall the BioRID P3 behaved more biofidelically and aligned closer with cadaver data than the Hybrid III dummy. Results of their study support the improved biofidelity of the BioRID P3 dummy over the Hybrid III in low speed rear end impact testing.

While the BioRID P3 shows an improved biofidelic response to rear impact collisions than the preceding model, the BioRID I, further revision of the dummy is still necessary. The thoracic spine of the BioRID P3 was changed to be softer than the BioRID I. This revision was made to increase upward motion of the T1 vertebrae to better align with volunteer data during rear end collision testing. However, during sled testing in a standard seat the upward motion of the T1 vertebrae in the BioRID P3 dummy was still found to be too small[8].

BioRID II

When comparing the biofidelity of head, neck, and torso response of the human volunteers, post-mortem human subject (PMHS) and crash test  dummies, specifically the BioRID II, Hybrid III, Thor-NT, and Thor-FT, Yaguchi et al. (2006)[14] concluded that the BioRID II was the most similar to human volunteers and PMHS during low speed rear impacts. Similar to the BioRID P3, they found the horizontal head displacement of the BioRID II fell within cadaver corridors while the Hybrid II was not after 70ms. While the BioRID II may have the most biofidelic response in low speed rear impacts, improvements can still be made. Yaguchi et al. (2006)[14] found that while the BioRID II measured the lowest peak force, of 6.3 kN, this was still 1.5 times the peak force measured in post mortem human subjects. However, this discrepancy may be due to inherent limitations with cadaver testing as post mortem human subjects PMHS do not exhibit any muscle tension and requires further investigation.

BioRID P50F

As can be seen in Table 1, both the EvaRID model and the BioRID P50F overall have smaller dimensions than the BioRID II. This makes both models more representative of the 50th percent female than the BioRID II dummy. While the BioRID P50F may be smaller in size than the BioRID II, one limitation is that it still uses parts for the BioRID II dummy based on a 50th percentile male. For example, the BioRID P50F spine is made through removing the L4 and L5 vertebrae from the BioRID II dummy. The BioRID II dummy was intentionally constructed with 24 vertebrae to resemble a human spine, thus removal of the L4 and L5 vertebrae calls to question the biofidelity of the BioRID P50F’s response under loading. The virtual model does not have this problem, since it is made by adjusting each segment while maintaining the general composition. Though this conserves the biofidelity, the fact that the model is only virtual does limit the testing abilities.

While the goal of the BioRID P50F is to represent the biomechanical response of a 50th percentile female, its biofidelity requires further investigation. In the study conducted by Carlsson et al. (2021)[9] investigating the performance of 50th percentile female rear impact dummies in comparison to volunteer data, they found that the horizontal displacement of T1, horizontal acceleration of T1, and amplitude of horizontal head acceleration in the BioRID P0F to be significantly different than that of volunteers. The thoracic and lumbar pin joint in the BioRID P50F dummy is the same as used in the BioRID I dummy and has a stiffness representative of the 50th percentile male rather than 50th percentile female. When developing the BioRID P50F, the stiffness was not scaled down, impacting the biofidelity of the BioRID P50F dummy. Specifically, when impacted rearward, the pin joint stiffness caused the angular and horizontal displacements of the T1 vertebrae to be less than the female volunteers[9].

While many improvements to the BioRID P50F have yet to be made, the initial development of a 50th percentile dummy still demonstrates work being done to improve diversity and inclusivity within automotive testing. It is also important to note that issues in terms of biofidelity with the BioRID P50F may not be specific to the female dummy but rather carried over from the BioRID II. For example, both the BioRID P50F and BioRID II had different displacement of the T1 vertebrae when compared to volunteer data in rear impact testing. As the BioRID P50F is constructed based on the BioRID II, this discrepancy may point to necessary further research and development into improving the biofidelity of T1 motion across all low-speed rear impact dummies.

BioRID vs Hybrid III

For frontal crash testing, the Hybrid III remains the most widely used crash test dummy around the world[15]. The Hybrid III dummy is also used in low speed, rear impact collisions with specific neck pieces, such as the RID-neck and TRID-neck, to address differences in neck biomechanics under low and high-speed loading. Both the BioRID and Hybrid III crash test dummies serve distinct purposes within automotive safety testing, each offering differing levels of biofidelity.

The BioRID is tailored for evaluating the effects of rear-end collisions, a common cause of whiplash injuries. Its sophisticated articulation, especially in the spine and neck, provides high biofidelity for simulating human movement during low-speed rear impacts. The BioRID's spinal column is particularly human-like with its S-shaped curve, designed to accurately reflect the human spine's behavior under such crash conditions[10]. Consequently, it is an invaluable tool for assessing the protection offered by vehicle seats and head restraints against whiplash.

In contrast, the Hybrid III is the industry standard for a broader range of crash tests, primarily focusing on frontal impacts. While it does offer a biofidelic response in those scenarios, its design is less specialized, resulting in a less accurate reproduction of the human body's reaction to rear-end collisions when compared to the BioRID. The Hybrid III's versatility makes it indispensable for a wide array of safety testing and regulatory compliance, despite its generalized biofidelity[15].

Each dummy's deployment is determined by the specific type of crash test. For studies of whiplash and similar injuries in rear-end collisions, the BioRID's high biofidelity for neck and spine movements is essential. The BioRID and Hybrid III dummies showed clear differences in their response to a rear-end collision, and the BioRID showed higher biofidelity than the Hybrid III[16]. Meanwhile, the Hybrid III remains the standard for evaluating impacts where the body's frontal response is the primary concern. This specialization ensures that safety assessments are as accurate and informative as possible for each unique crash scenario.

Limitations

Accident test dummies, crucial tools in automotive safety testing, inherently come with limitations and controversies. Despite technological advancements, these dummies still struggle to perfectly replicate the nuanced responses of the human body in varied real-world crash scenarios. Each dummy is designed to mimic human reactions to specific types of impacts, yet human bodies exhibit a vast range of biomechanical responses depending on numerous factors like age, gender, body type, and pre-existing conditions[17]. Moreover, the high cost and complexity of these sophisticated dummies can significantly inflate the expenses associated with testing and development. This financial barrier can impact the frequency and diversity of safety testing conducted, potentially limiting the breadth of data collected[17]. These limitations and the controversies they spawn are critical because they touch upon the core purpose of crash testing—to ensure that vehicles are as safe as possible for every potential occupant. The ongoing challenge for the industry is to balance the need for comprehensive, biofidelic testing with the practicalities of cost and technological capabilities.

BioRID

The BioRID I dummy, a pioneering effort in the pursuit of more biofidelic crash test dummies for assessing whiplash protection, faced several limitations and controversies. One primary limitation was its initial lack of sophistication, which meant it could not capture the full range of human neck movements with high fidelity, potentially leading to less accurate predictions of injury in real-world collisions. Moreover, controversies often arose from its one-size-fits-all approach, which did not account for the wide variability in human body types, including size, weight, and gender differences. Such limitations raised concerns about the generalizability of the test results. The introduction of BioRID I and its use in regulatory testing underscored the ongoing challenge within automotive safety research to create dummies that accurately reflect the diversity of the population and the complexities of human anatomy and injury mechanisms[17].

The BioRID II, while an advancement over its predecessor, faced quite similar limitations and associated controversies. These again regarded biofidelity, as discussed in more detail above, as well as representation of the population. Moreover, there is controversy surrounding the adoption of the BioRID II in regulatory standards, as some stakeholders in the automotive industry may claim that these standards may encourage the development of seat and head restraint systems that perform well in tests with the BioRID II but do not offer equivalent protection in varied, real-world scenarios. This controversy is significant because it highlights the tension between designing vehicles that meet specific regulatory test criteria and those that provide the highest possible protection in the full spectrum of real-life accidents.

EvaRID

The EvaRID, a crash test dummy developed to represent the average female body, addresses a critical gap in automotive safety testing by providing data on how crash impacts affect women, who have historically been underrepresented in this field[18]. However, limitations and controversies accompany its introduction and usage. A notable limitation is that the EvaRID FE is not as extensively used or as thoroughly researched as male-centric dummies, potentially leading to a smaller data set from which to draw conclusive safety insights[19]. Additionally, the integration of EvaRID into the standard suite of crash test dummies has been slow, raising concerns about the ongoing gender disparity in vehicle safety research. The controversy here is centred on the urgency and importance of including female dummies like EvaRID FE in regulatory testing to ensure that safety features are optimised for both sexes. This issue is particularly significant as studies have shown that women face a higher risk of certain injuries in automobile accidents, underscoring the need for inclusive testing that leads to universally safer vehicles[20]. Furthermore, the EvaRID FE has the limitations of limited testing abilities since it is only a virtual model and tests have shown multiple areas to improve on[21]. This limits implications as it is only a simulation of a potential real life scenario and does not do a great job at that just yet. The BioRID P50F, seemingly the only current physical prototype of a female dummy version, has been made experimentally by a group of researchers[9]. This means they had to work with the existing BioRID II parts to build a primitive and downsized prototype, which then inevitably has a limited biofidelity both in shape and material.

Future Work

Future research priorities with the BioRID models involve a few key areas that need further exploration and development and align with future work required for all model of ATDs.

Enhanced Biofidelity

Continuing to improve the biofidelity of dummies to more accurately simulate human body responses across different genders, ages, and physical characteristics is crucial. Research could focus on better understanding the biomechanical variations in the population and developing more biofidelic materials to be used in dummy construction.

Material Science Innovations

Advancements in materials used to construct dummies could lead to better simulation of human tissue behavior and more accurate injury prediction. Research into novel materials that closely mimic the mechanical properties of human tissues could be prioritized. Currently, some biomaterials are being developed to more closely mimic human tissue behavior. For instance, researchers are working on combining silk with tropoelastin, a highly elastic and dynamic structural protein, to create materials that mimic the elasticity of diverse tissue structures. This approach is particularly aimed at controlling stem cell fate and function, which could be critical in replicating human tissue responses in test dummies[22].

Advanced Sensing and Data Analysis

Implementing advanced sensors in ATDs to collect more granular data at higher resolutions is a significant step towards gaining deeper insights into injury mechanisms. Advanced sensors can provide detailed information on various impact forces and the body's response in different crash scenarios. When coupled with sophisticated data analysis techniques like machine learning, this approach can help uncover patterns and injury mechanisms that are currently not well understood.

One relevant example is the research conducted by Wake Forest University, where thousands of virtual crash simulations were run through a supercomputer using real-world crash data. This approach allowed the team to study the effects of vehicle design parameters and safety features on occupant factors, proposing solutions to prevent and mitigate occupant injury. Anthropomorphic test devices in these studies provide data on around 20 points on the body, but digital simulations enable a much more detailed examination, testing various body shapes, sizes, and positions at the moment of impact. The Wake Forest model can quantify the risk of bone fractures and damage to soft tissue and organs, offering insights beyond the capabilities of traditional crash test dummies[23].

Real-World Scenario Simulation

Utilizing real-world crash data can significantly enhance the predictive capabilities of crash test results. By analyzing data from actual accidents, researchers can identify common injury patterns and crash scenarios not currently covered in standard tests, leading to the development of dummies that better mimic human injuries in those conditions. Additionally, there is a need for longitudinal studies that correlate data from crash test dummies with actual post-crash medical outcomes. Such studies would provide critical insights into the accuracy of dummy data in predicting real human injuries. This can lead to improvements in dummy design and usage, making them more biofidelic.

Cost Reduction Strategies

Investigating methods to reduce the cost of dummies without compromising on their biofidelic properties would allow for more comprehensive testing and make safety testing more accessible.

A physical crash test dummy can cost over $200,000, made from a variety of materials to mimic human dimensions and loaded with sensors to record extensive data during crashes[24]. Using FEA for simulating crash dummy performance is significantly more cost-effective than physical testing. FEA allows repeated virtual tests at a much lower cost than physical tests. This simulation helps in optimising the performance of safety devices and finding potential issues early in the design process, thus reducing the risk of expensive failures in actual crash tests​​. FEA also enables detailed testing of individual dummy components and full-body simulations under various standards, offering comprehensive insights that can inform design improvements

Running a full-scale physical crash test can cost around $50,000, a significant investment for each test[25]. Using simulation models substantially reduces this risk and cost. The first model costs roughly the same as a physical test, but repeated simulations can be done at a fraction of the cost​​​​. Simulation offers the flexibility to optimize device performance by adjusting variables like material choices and configurations, contributing to overall cost savings.

References

  1. 1.0 1.1 1.2 "Humanetics BioRID-II".
  2. Lee, Suzanne; Llaneras, Eddy; Klauer, Sheila; Sudweeks, Jeremy (October 2007). "Analyses of Rear-End Crashes and Near-Crashed in the 100-Car Naturalistic Driving Study to Support Rear-Signaling Countermeasure Development" (PDF).
  3. Astrup, Jens; Gyntelberg, Finn (2022-03-12). "The Whiplash Disease Reconsidered". Front. Neurol.
  4. Eck, Jason C; Hodges, Scott D; Humphreys, S. Craig (2001-06-01). "Whiplash: a review of a commonly misunderstood injury". The American Journal of Medicine. 110: 651–656.
  5. 5.0 5.1 Linder, Astrid; Svensson, Mats Y; Davidsson, Johan; Flogård, Anders; Håland, Yngve; Jakobsson, Lotta; Lövsund, Per; Wiklund, Kristina (1998). "The New Neck Design for the Rear-End Impact Dummy, BioRID I". Annu Proc Assoc Adv Automot Med. 42: 179–192.
  6. Foster, J. King; Kortge, James O.; Wolanin, Michael J. (1977). "Hybrid III—A Biomechanically-Based Crash Test Dummy". SAE Transactions. 86: 3268–3283.
  7. 7.0 7.1 Davidsson, Johan; Svensson, Mats; Folgard, Anders; Håland, Yngve; Jakobsson, Lotta; Linder, Astrid; Lövsund, Per; Wiklund, Kristina (1998). "BioRID l - A New Biofidelic Rear Impact Dummy" (PDF). IRCOBI Conference.
  8. 8.0 8.1 8.2 Davidsson, Johan; Lövsund, Per; Ono, Koshiro; Svensson, Mats Y.; Inami, Satoshi (2001-01-01). "A Comparison of Volunteer, BioRID P3 and Hybrid III Performance in Rear Impacts". Journal of Crash Prevention and Injury Control. 2: 2003–220.
  9. 9.0 9.1 9.2 9.3 9.4 9.5 9.6 9.7 9.8 9.9 Carlsson, Anna; Davidsson, Johan; Linder, Astrid; Svensson, Mats Y. (2021). "Design and Evaluation of the Initial 50th Percentile Female Prototype Rear Impact Dummy, BioRID P50F – Indications for the Need of an Additional Dummy Size". Frontiers in Bioengineering and Biotechnology.
  10. 10.0 10.1 Davidsson, Johan (1999). "BioRID II final report". Crash Safety Division Department of Machine and Vehicle Design.
  11. "Humanetics EvaRID FE".
  12. 12.0 12.1 Schneider, Lawrence W. (1983). "Development of anthropometrically based design specifications for an advanced adult anthropomorphic dummy family, volume 1. Final report". Anthropometry of motor vehicle occupants. Volume 1 - procedures, summary findings and appendices.
  13. Linder, Astrid; Holmqvist, Kristian; Svensson, Mats, Y (1 May 2018). "Average male and female virtual dummy model (BioRID and EvaRID) simulations with two seat concepts in the Euro NCAP low severity rear impact test configuration". Accident Analysis and Prevention. 114: 62–70.
  14. 14.0 14.1 Yaguchi, Masayuki; Ono, Koshiro; Kubota, Masami; Matsuoka, Fumio (September 2006). "Comparison of Biofidelic Responses to Rear Impact of the Head/Neck/Torso among Human Volunteers, PMHS, and Dummies" (PDF). IRCOBI.
  15. 15.0 15.1 "Humanetics Hybrid III 50th Male".
  16. Gotou, Tsukasa; Ono, Koshiro; Masahiro, Ito; Matuoka, Fumio (2001-06-04). "A Comparison Between Biorid and Hybrid Iii Head/Neck/Torso Response in Middle Speed Sled Rear Impact Tests". SAE Technical Paper.
  17. 17.0 17.1 17.2 Xu, Tao; Sheng, Xiaoming; Zhang, Tianyi; Liu, Huan; Liang, Xiao; Ding, Ao (2018-11-13). "Development and Validation of Dummies and Human Models Used in Crash Test". https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6257900/. 2018. External link in |journal= (help)
  18. Bose, Dipan; Segui-Gomez, ScD, Maria; Crandall, Jeff R. (2011). "Vulnerability of Female Drivers Involved in Motor Vehicle Crashes: An Analysis of US Population at Risk". Am J Public Health.
  19. Chang, Fred; Carlsson, Anna; Lemmen, Paul; Svensson, Mats; Davidsson, Johan; Schmitt, Kai-Uwe; Zhu, Fuchun; Linder, Astrid (2010). "EvaRID: a dummy model representing females in rear end impacts". Whiplash 2010: Neck Pain in Car Crashes.
  20. Storvik, Steven G.; Stemper, Brian D.; Yoganandan, Narayan; Pintar, Frank A. (2009). "Population-based estimates of whiplash injury using nass cds data - biomed 2009". Biomed Sci Instrum.
  21. Carlsson, Anna; Change, Chang; Lemmen, Paul; Kullgren, Andres; Schmitt, Kai-Uwe; Linder, Astrid; Svensson, Mats Y. "Anthropometric Specifications, Development, and Evaluation of EvaRID—A 50th Percentile Female Rear Impact Finite Element Dummy Model". Traffic Injury Prevention.
  22. "Biomaterials".
  23. "A Smarter Kind of Crash Test Dummy".
  24. "Simulating Crash-dummy Performance with FEA".
  25. "Simulated Crash Testing Saves Time and Money".


External Links