Effectiveness of load legs on rear-facing only child restraint systems (CRS) in frontal sled tests

1 Introduction

Child restraint systems (CRS) are effective at reducing the risk of death and injury for children in motor vehicle crashes (MVCs) (Kahane, 2015). All CRS must pass regulatory crash testing standards to be sold on the United States (US) market. The protocols are outlined in the Federal Motor Vehicle Safety Standards (FMVSS) No. 213 and its accompanying test procedures document (National Highway Traffic Safety Administration (NHTSA), 2023a and 2023b). This process ensures a baseline level of safety for all products. Current FMVSS 213 protocols require CRS to pass a regulatory frontal crash test without the use of a load leg, which braces the CRS against the floor of the vehicle structure. NHSTA has asserted that this practice will ensure quality performance of every CRS whether the load leg is used or not (National Highway Traffic Safety Administration, 2023a).

Emerging work is documenting the additional benefits of the load leg feature. Patton et al. (2020), Patton et al. (2021) conducted frontal impact tests in the Consumer Reports buck at 35 g and 56 kph. They report reductions in head resultant acceleration (3 m clip) and HIC15 with use of the load leg on rear-facing only (RFO) infant CRS. The trend was also true for larger occupants (Q1.5, Q3, and Q6) in different types of convertible and extended-use CRS (Patton et al., 2020). Reductions in head injury metrics with the load leg were likely due to reduction in forward rotation into the blocker plate–a surrogate structure that represents the back of the front row vehicle seat (Patton et al., 2020). Further testing included both frontal and frontal-oblique impacts at 30° with two different models of RFO infant CRS (Patton et al., 2021). The authors report head contact against the blocker plate for all tests without the load leg. Using the load leg prevented head contact in all cases and significantly reduced head injury metrics. A similar study of frontal-oblique (30°) and oblique (60°) impacts continued to find reductions in head injury metrics when RFO infant CRS were installed with the load leg compared to without (Patton et al., 2023). Head contact against a side door panel was prevented when the load leg was used in the 60° impacts. In a study of full vehicle tests of CRS without load legs, Tylko et al. also reported elevated head accelerations when the RF CRS or anthropomorphic test device (ATD) head contacted the back of the front row seat or center console (Tylko, 2011). Real-world crash data confirms that head is the most commonly injured body region for children in RF CRS (Arbogast and Durbin, 2013). Therefore, the ability of the load leg to reduce the risk of head contact against vehicle interior structures offers potentially significant safety benefits to children.

The physical responses of load leg assemblies have been documented briefly in a few studies. In a series of vertical drop tower loading on a sample of isolated load legs, one load leg model compressed under loading by 23 mm (Patton et al., 2025). The authors note that it was unclear if the load leg either partially failed or if the plastic deformation was an intended energy-absorbing mechanism. Partial failures of load legs are also reported in two sled tests where the adjustment pin of the load leg sheared during frontal impact loading (Patton et al., 2020; 2021). In one of these cases the collapsed load leg resulted in head injury metrics similar to tests without a load leg, indicating no benefit to the occupant. In the second case, the collapsed load leg resulted in a partial reduction of head injury metrics compared to corresponding tests without a load leg. Therefore, while collapse is not ideal, it appears that presence of a load leg might, in some cases, provide some benefit to the occupant even if it fails at some point during loading. More study is needed on this occurrence and its overall effect on child occupant safety.

This study expands on previous work to explore how RFO CRS from various manufacturers perform in frontal impacts with and without the load leg engaged. Additionally, this study evaluates an experimental load leg with increased stiffness at its connection point to the CRS base compared to a different experimental version that dampens the forces with controlled motion. The goal of this study is to establish the efficacy of various designs of load legs in frontal impacts, including cases of load leg collapse, compared to identical tests without load legs.

2 Materials and methods

2.1 Test setup and matrix

A test buck was built to the specifications of the Consumer Reports CRS test buck (Consumer Reports, 2020). Briefly, a rigidized seat from a 2010 Ford Flex was positioned in front of a blocker plate to represent the front row seat (Figure 1).

The Consumer Reports frontal pulse (35 g and 56 kph) was used as the target pulse for these frontal impacts. Actual sled acceleration was 34.5 ± 0.6 g (minimum: 33.5 g, maximum: 36.3 g) measured from accelerometers mounted on the sled buck (Figure 2). Actual velocities were 62.1 ± 1.7 kph (minimum: 58.3 kph, maximum: 65.7 kph), calculated from accelerometer data.

A convenience sample of six different models of RFO CRS were selected. One CRS, the Romer Juni, included two different experimental load leg designs: one with a shear pin designed to allow controlled deformation at the hinge joint (“dampening” design) and the other with a reinforced stiff connection at the hinge joint (“stiff” design). All other CRS were purchased off-the-shelf with their production load legs. Each CRS was tested with and without its load leg for at least three repetitions each. The CRS was replaced after each test. A total of 55 tests were run (Table 1). Unequal numbers of tests were run across CRS models because tests were repeated if key data were lost (e.g., a faulty chest z-accelerometer). However, the remainder of the accessible data were included from these tests to increase sample size.

TABLE 1. Test matrix. The series included six CRS models tested with and without their load legs. The Romer Juni included load legs of two different designs (dampening and stiff).

Each CRS was installed with its base using the lower anchors. Two CRS models had flexible webbing lower anchor assemblies (Evenflo LiteMax and Peg Perego Primo Viaggio) while the other four had rigid lower anchor assemblies. Load legs were positioned against the floor of the test buck according to manufacturers’ instructions or were folded into their storage position underneath the base.

The Child Restraint and Airbag Interaction (CRABI) 12-month-old anthropomorphic test device (ATD) was used for all tests. This is the standard ATD used for regulatory tests of RFO CRS in the US (NHTSA, 2023b). The ATD was positioned and harnessed into each CRS according to FMVSS 213 protocol (NHTSA, 2023b).

2.2 Data collection and analysis

High speed video was collected at 1,000 fps from the right lateral view and oblique overhead view. Head and chest acceleration data were collected along x, y, and z-axes according to FMVSS 213 protocol (NHTSA, 2023b). The following metrics were calculated: head injury criterion across a 15 m interval (“HIC15”), peak head resultant acceleration over a 3 m clip (“head acceleration”), and peak chest resultant acceleration over a 3 m clip (“chest acceleration”). ATD head contact against the blocker plate was determined via conductive tape. CRS shell contact with the blocker plate was determined via video review. The change in angle from the CRS initial position to peak point of rotation was calculated by video review.

To understand the effects of load leg design, the three different load leg conditions for the Romer June (none, dampening, stiff) were evaluated using one-way Analysis of Variance (ANOVA) with post hoc Tukey-Kramer tests for continuous metrics (change in CRS angle, HIC15, head resultant acceleration, and chest resultant acceleration) and Fisher’s Exact Tests for categorical metrics (CRS shell contact and head contact).

To evaluate the effect of load leg presence across the full sample of six CRS models, two-way ANOVA tests including interaction effects (presence of load leg, CRS model) were conducted for the continuous metrics. Cochran-Mantel-Haenszel (CMH) tests were used to evaluate the association between two categorical variables (presence of load leg (Yes/No) and contact against the blocker plate (Yes/No)) while controlling for the effects of a third categorical variable (CRS model). Separate CMH tests were run to evaluate CRS shell contact and ATD head contact against the blocker plate.

To better understand the implications of load leg collapse, the CMH tests and two-way ANOVAs were repeated on the subset of CRS models whose load legs consistently collapsed. JMP Statistical Discovery (Student Edition 19.0.5) was used for all statistical analyses with alpha level set to 0.05.

3 Results

One representative test of each CRS in its initial position and at peak displacement is shown in Supplementary Appendix Figures A1-A6.

3.1 Dampening vs. stiff load leg design

For the three load leg conditions on the Romer Juni (none, dampening and stiff), the occurrence of CRS shell and ATD head contact against the blocker plate are shown in Table 2. CRS shell contact was significantly different across load leg conditions (p = 0.0103, Fisher’s Exact Test). Video was lost for one test in the no load leg condition. The singular stiff load leg test that collapsed was the test that resulted in CRS shell contact with the blocker plate. The ATD head did not contact the blocker plate in any of the tests regardless of load leg condition. Head contact was determined via conductive tape, not video footage, so all tests were available for analysis.

TABLE 2. Occurrence of CRS shell and ATD head contact against the blocker plate for the Romer Juni, with Fisher’s Exact tests as applicable.

Video footage was lost for one test, so percentages are calculated from n = 2 tests.

CRS, shell contact occurred when no load leg was used and during one stiff load leg test. ATD, head contact did not occur in any Romer Juni test.

For the Romer Juni, HIC15, head acceleration, and change in CRS angle were significantly different across load leg conditions (p = 0.0004, 0.0016, <0.0001, respectively, one-way ANOVA tests, Supplementary Appendix Tables B1, B2, B4). Specifically, outcomes for the no-load leg condition were significantly higher than those of both the dampening and stiff load leg; however, the dampening and stiff load legs were not significantly different from each other (Tukey-Kramer post hoc tests, Table 3). Chest acceleration was not significantly different across any load leg conditions (p = 0.1831, one-way ANOVA test in Supplementary Appendix Table B3 and post hoc tests in Table 3).

TABLE 3. Tukey-Kramer connecting letters report for the Romer Juni.

Levels not connected by the same letter are significantly different. The performances of the two load leg designs were not significantly different from each other for any of the injury metrics recorded. Both load leg designs significantly reduced HIC15, chest acceleration, and change in CRS, angle compared to no load leg.

Further analyses in this study consider only the stiff load leg design for the Romer Juni, as this design is more similar to most production load legs.

3.2 Load leg vs. none

Considering all six CRS models included in this study, presence of the load leg significantly reduced the occurrence of CRS shell contact and ATD head contact against the blocker plate. Table 4 shows outcomes across all CRS models combined. Supplementary Appendix Table C1 shows detailed outcomes within each CRS model. Nearly all of the CRS shells contacted the blocker plate both with and without load legs except for the Romer Juni, for which the load leg mostly eliminated CRS shell contact. Head contact was reduced with load leg presence across a wider variety of CRS models. CMH tests control for the effects of CRS model as a categorical variable.

TABLE 4. Cochran-Mantel-Haenszel (CMH) tests for CRS shell and ATD head contact against blocker plate.

The results indicate significant reductions in both types of contact when a load leg is used compared to tests without load legs.

Across the full sample of CRS, usage of load legs reduced the change in rearward rotation angle of the CRS by an average of 10.3° or 47.9% (Figure 3; Supplementary Appendix Table D1). Use of load leg and CRS model are both significant as main effects (p < 0.0001 and p < 0.0001, respectively) and their interaction effect is also significant (p < 0.0001, two-way ANOVA; Supplementary Appendix Table D2).

Across the full sample of CRS, usage of load legs reduced HIC15 by an average of 1,339.7 or 71.2% (Figure 4; Supplementary Appendix Table E1). Use of load leg and CRS model are both significant as main effects (p < 0.0001 and p < 0.0001, respectively) and their interaction effect is also significant (p < 0.0001, two-way ANOVA; Supplementary Appendix Table E2).

Across the full sample of CRS, usage of load legs reduced head acceleration by an average of 69.3 g or 48.7% (Figure 5; Supplementary Appendix Table F1). Use of load leg and CRS model are both significant as main effects (p < 0.0001 and p < 0.0001, respectively) and their interaction effect is also significant (p < 0.0001, two-way ANOVA; Supplementary Appendix Table F2).

Across the full sample of CRS, usage of load legs reduced chest acceleration by an average of 14.6 g or 19.8% (Figure 6; Supplementary Appendix Table G1). Use of load leg and CRS model are both significant as main effects (p < 0.0001 and p < 0.0001, respectively) and their interaction effect is also significant (p = 0.0121, two-way ANOVA; Supplementary Appendix Table G2).

3.3 Collapsed load legs

Crash forces caused several of the load legs to collapse under loading. The load leg collapsed on all repetitions of the Clek Liing, Cybex Cloud G Lux, and Nuna Pipa. The load leg collapsed on one test of the Romer Juni (stiff design) perhaps due to a positioning error but did not collapse on the other six repetitions of the stiff design nor any of the eight repetitions of the dampening design.

To better understand the implications of load leg collapse, the CMH tests and two-way ANOVAs were repeated on the subset of CRS models whose load legs consistently collapsed (Clek Liing, Cybex Cloud G Lux, and Nuna Pipa; Supplementary Appendix Table H1). Even though all the load legs in this subset collapsed, there were still significant reductions in the occurrence of head contact against the blocker plate (p = 0.0020), HIC15 (p < 0.0001), head acceleration (p < 0.0001), and chest acceleration (p = 0.0096) compared to corresponding tests without a load leg. However, collapsed load legs were not significantly different from tests without load leg regarding occurrence of CRS contact against the blocker plate (p = 0.3173) and change in CRS angle (p = 0.1157). Therefore, collapsed load legs do not seem to offer kinematic benefits compared to CRS without load legs (i.e., the physical movements of the CRS were similar). However, the significant reductions in HIC15, head acceleration, and chest acceleration suggest that the collapsed load legs managed crash energy (by directly absorbing and/or redirecting to other CRS or vehicle seat components) and rendered any subsequent contacts against the blocker plate less forceful. The smaller sample size likely reduced the significance of CRS model for some variables.

4 Discussion

The use of load legs in these frontal impacts established several significant benefits over the same set of CRS without load legs. Across the full sample of six CRS models, load legs significantly reduced the occurrence of CRS shell and ATD head contact against the blocker plate. On average, load legs reduced the change in CRS angle by 10.3° (47.9%, p < 0.0001) HIC15 by 1,339.7 (73.3%, p < 0.0001), head acceleration by 69.3 g (50.8%, p < 0.0001), chest acceleration by 14.6 g (19.8%, p < 0.0001). When comparing the performance of two experimental load legs (stiff vs. dampening design), performance outcomes were similar to one another. Both load legs significantly reduced the change in CRS rotation angle, HIC15, and head acceleration compared to corresponding tests without a load leg. Chest acceleration was not significantly affected by any condition.

This dataset offers a novel investigation of the effects of load leg collapse. The load leg collapsed on all repetitions of the Clek Liing, Cybex Cloud G Lux, and Nuna Pipa. All these cases of collapse were caused by shearing of the pin that holds the load leg at the desired length. The load leg collapsed on one test of the Romer Juni (stiff design) but did not collapse on the other six repetitions of the stiff design nor any of the eight repetitions of the dampening design. It is suspected that an initial positioning error (i.e., the pin was not fully engaged in the length setting) was responsible for this singular collapse because no post-test evidence of pin shearing was found in this case nor among the other repetitions that did not collapse.

The timing of each of these collapse events appears to be near the middle of the loading process as the CRS rotates toward the blocker plate. Exact timing could not be determined due to lack of time stamps on the video footage. The CMH and two-way ANOVA tests on the subset of CRS that experienced load leg collapse identified significant reductions in head contact against the blocker plate, HIC15, head acceleration, and chest acceleration compared to corresponding tests without a load leg (Supplementary Appendix Table H1). This indicates that substantial amounts of energy were successfully absorbed or redirected by the load legs prior to their collapse and the occupants in these conditions still partially benefited from having a collapsed load leg vs. none. However, the collapsed load legs did not significantly reduce the CRS rotation angle compared to corresponding tests without load leg (Supplementary Appendix Table H1). Additionally, the singular Romer Juni test with collapsed load leg was the only “with load leg” test from that CRS model that resulted in CRS shell contact against the blocker plate. Therefore, it appears that collapsing load legs are not effective at reducing the magnitude of CRS rotation or occurrence of CRS shell strikes against the vehicle interior, but might still reduce the peak forces experienced during these kinematic trajectories. Real-world crash data indicate that the head is the most commonly injured body region for children in RF CRS (Arbogast and Durbin, 2013). Therefore, the ability of the load leg to reduce the risk of head contact or CRS shell contact against vehicle structures, ideally by maintaining its structural integrity, should be a priority for this feature. Previous studies with smaller samples of load leg collapse concluded that load leg integrity is important to reduce head injury metrics (Patton et al., 2020; 2021). This evidence combined with the current series suggests that maintaining load leg structural integrity is ideal to reduce CRS rotation and associated injury metrics, but a load leg that collapses can still be beneficial compared to no load leg at all.

FMVSS 213 includes structural integrity and position retention requirements for CRS. However, it is important to note that FMVSS 213 requires CRS to be tested without their load leg so instances of load leg collapse may not be fully evaluated under FMVSS 213 guidelines. Additionally, the tests in this series were conducted at a higher severity (34.5 g, 62.1 kph) compared to the standard FMVSS 213 frontal pulse (maximum of 25 g, 48 kph). Therefore, the sheared pins and load leg collapses observed in this study are not representative of outcomes for current test regulations. The pulse in this study represents a very high severity crash, as only about 0.6% of real-world crash occupants experience a delta-V greater than 60 kph (Heller et al., 2010).

The presence of a load leg demonstrated a significant influence over the kinematics and kinetic responses observed here. The two-way ANOVA tests also showed significant effects of CRS model and the interaction effect of load leg and CRS model. These outcomes indicate that features other than just the load leg were significant to each CRS model’s performance. For example, the Cybex Cloud G Lux has a flexible back that seemed to flex into a more reclined position during tests with the load leg (Supplementary Appendix Figure A2). The Romer Juni base includes a curved aluminum bar which may have absorbed or redirected forces as the CRS rotated over the front edge of the vehicle seat. Another important CRS-based feature is the type of lower anchor system: rigid or flexible. The rigid lower anchor systems in this series appeared to create a more well-defined fulcrum point around which the CRS rotated with minimal forward translation of the base away from the seat bight (see Supplementary Appendix Figures A1, A2, A4, A6). In contrast, CRS with flexible lower anchor systems displayed more forward translation away from the seat bight (Supplementary Appendix Figures A3, A5) and, in one case, less rotation about the lower anchor fulcrum (Evenflo LiteMax, see Figure 3; Supplementary Appendix Figure A3). The effects of lower anchor type could not be isolated in the current study because of other confounding variables that were also CRS model dependent, such as load leg collapse and flexibility of different CRS components.

All of the tests in the current study were set up with a gap between the back of the CRS shell and the blocker plate. Recent work has explored the role of this spacing between RF CRS and front row seat structures. Maltese et al. (2025) report that reducing the amount of space or bracing the RF CRS against the front row is beneficial in reducing HIC15, neck tension, and in some cases chest acceleration. Similar conclusions are reported by Patton et al. (2022) for a Q3 occupant in a RF convertible CRS. However, Patton et al. (2022) report a Q1.5 occupant in a RFO CRS having the highest head injury metrics in the braced condition due to head contact against the front row seat. Use of a load leg significantly reduced most injury metrics in that study with front row vehicle seats (Patton et al., 2022) and a subsequent study with front row center consoles (Patton et al., 2024). Considering the wide variance in vehicle interior designs and space availability between rows, load legs could offer a reliable way to reduce the movement of the CRS to make it less likely to contact vehicle interior structures and therefore make the CRS installation less sensitive to variations in vehicle seat and environment design.

A recent load leg usability study indicated that most caregivers of young children in the US are unaccustomed to this safety feature. Practice installations of a rear-facing only (RFO) CRS indicated that only 54% of study participants noticed and used the load leg without being prompted by the research team (Mansfield et al., 2026). Only 56% of participants said they liked the load leg feature on their first impression of it. Nearly all participants (96%) correctly inferred that the load leg’s purpose might be to improve “safety,” or “stability,” of the CRS, but many expressed concerns over the amount of floor space it occupied, whether it would function during a crash, and whether it would stay in place during normal driving, ingress, and egress of the vehicle. Caregivers’ hesitancy to embrace this safety feature indicates a need for further education and accessibility of information about its safety benefits. Such data would be valuable to caregivers, safety advocates, and industry professionals to demonstrate the value of load legs as a safety feature in the US market.

This study includes several limitations. The Consumer Reports test bench uses a rigidized second row vehicle seat from the 2010 Ford Flex. This seat may not be representative of modern vehicle seats. The sled velocity of the current study was slightly above the Consumer Reports target range, which may create differences between these results and similar literature. Further study is needed in vehicle environments with varying interior features. Some vehicles prohibit the use of load legs due to hollow storage space underneath the floor. Further study is needed to verify safe and reliable usage scenarios for this safety feature. The CRABI 12mo is near the upper range of occupant size allowable in most RFO CRS, due to most requiring the top of the occupant’s head to be at least one inch below the top of the CRS shell. However, most RFO CRS can accommodate occupants up to 30 or 35 lbs, while the CRABI weighs only 22 lbs. Higher weight occupants might present a higher risk of load leg collapse, increased CRS rotation, and risk of contacting the blocker plate. Neck injury metrics are not well validated in the CRABI 12mo and were not collected in this study.

Load legs provide significant benefits to child occupants in RF CRS, including reductions in the occurrence of head/CRS contact against front row vehicle structures, CRS rotation, HIC15, head acceleration, and chest acceleration. Load legs that maintained their structural integrity throughout the crash event were most effective. However, these novel findings suggest that even in cases when the load leg collapsed due to pin shear at this high severity pulse, benefits to the occupant were still significant compared to no load leg. Use of this safety feature in family vehicles should be encouraged when possible.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmech.2026.1917681/full#supplementary-material

Supplementary material Supplementary material at the source

Footnotes

Edited by: Costin D Untaroiu, Virginia Tech, United States

Reviewed by: Rodney Rudd, National Highway Traffic Safety Administration, United States

Juliet Haarbauer-Krupa, Emory University, United States