SEAT BELT PERFORMANCE AND AFrER-MARKET WEB-LOCKING DEVICES: AN EXPERIMENTAL STUDY

F. Zuppichini

Emergency Surgery Dept.,Verona University Medical School (Dir. Prof. Claudio Cordiano)

ABSTRACT Thisexperimentalresearch isaimed to evaluate the consequencesof theslack offeredby after-market web-locking components ("comfort clips") on the perfonnance of three-point seat belts. The tests were performed by simulation of a frontal impact at km/h, recording ehest and pelvis 50 displacement of the dummy and variations in belt loading. We tested all the deviceson salein ltaly up to September 1989: each device has beentested five times, with a configuration of the restraint system according to Tipo. The laboratory results show a very bad perfonnance of the devices under the test conditions, with a mean increase of 15-20% for ehest and 40% for pelvis displacement. The most worrying finding was the breakage of the belt,in casesover other cases,there was 7 35; in 3 severe belt incision. All these laboratory data demonstrate danger the use of these web-locking components. in

BACKGROUND belt manifacturers that some four million There is no question about the effective­ Italian circulating cars are still lacking in seat ness of seat belts in preventing or reducing belts. 7 injuries to car occupants6. • According to a study from an ltalian Despite the worldwide acceptance of insurance company 5, only 47% of the car these car restraint systems, was the only occupantsinvolved in acddentsin the months European country not to provide a compul­ following the new law were using their seat sory rule about their use. Up to 1987, it was belts (61.% in Northern Italy, decreasing to even possible to buy cars without front seat 12% in the South of the country). belts: hundred of thousands of Fiat Panda, Because of absence of information, seat Ritmo, Regata, 126, 127, A 112, belt usage was misunderstood and adversed and even a special 1050 cc version of Volks­ even by authorities. wagen Golf were sold without seat belts. By summer 1989, a lot of Italian dealers After a long gestation,a seat belt law is were marketing devices designed to clarnp operating from 26April 1989,for all the drivers the web and pull it to a comfortable distance and front passengers of cars registered after from the body of the car occupant. These 1978; the car is missing of front seatbelts, it if articles found a great success and it has been is compulsory to retrofit them. Forcars before estimated that their selllngs exceeded one 1978, the installation (and use) of front seat million of pieces. is due from 26 October 1989, fitting belts if Similar articles were found to be on sale holes are present on the car. also in France and in the United Kingdom. Italian car users had a poor compliance experimental research is aimed to This to the law, and after the first days the fastening evaluate theconsequencesof the slack offered very low. Even the retrofitting rate ranked by these "comfort clips" on the performance rule was ignored: it has been estimated by of automatic three-point seat belts.

- 15 - MATERIALS AND METIIOD Dataabout chest and pelvisdisplacement We collected any type of "comfort-clip" after the impact and variationsin belt loading available on the Italian market in sumrner were collected and storedin a database;some They can be divided into two groups tests were high-speed filmed. The obtained 1989. (Picture 1 ): data were compared with the current A. "Hook" type, to be inserted between the omologation rules. web and the loop at the pillar; Each device has been tested five times, B. "Pincers" type, to be inserted afterthe loop, replacing both the comfort clip and the seat near the shoulder. belt. The reference group, with the sametype The tested devices are seven, four of of seat beltsand the sameconfiguration, com­ which of the "hook" type; their comrnercial prises 40 sled tests. name is reported in Table 1. The tests were performed in Autoliv RESULTS laboratories in (ltaly), by simulation of a frontal impact at (±0,2) km/h with an A summary of the results is shown SO in anthropomorphic durnmy. Table I, for each crash test with "comfort The devices were tightly wom to the clips" and for the reference group. d y and then retracted to obtain mm The worst finding was the breakage of umrn 100 of slack, equally distributed to the thoracic the seat belt (Picture 2), in 7 cases over 35, equally distributed among "hook" and and abdominal hamesses. By comparative tests, this amount of slack on the dummy was "pincers" types;in other 3 cases,the web was assurned to be pair to 3-5 cm on a normally severely twisted and lacerated. all findings of belt breakage, a clear dressed person. In The configuration of the restraint system incision on the plastic coverage of the loop is was according to Fiat 160(''Tipo.") saloon presentat the rearward-facing angle. 5- Nearly the remainingtests show a net doors, with an altemate choice of the belt all retractor at the left or at the right side, simul­ worsening of belt perforrnance, with values ating the position of the driver or the front around + 20% forchestdisplacement and over right passenger. +40% for pelvis displacement (Picture 3).

B

t of "comfortclips". Picture 1: Diff� kinds A: "Hook" type,to be the web insertedwithin and the B-pillar; B: type,tobe locked e thoracic "Pincers" in th harness ,near to thebelt loop.

- 16 - Commercial name Test 1 Test2 Test3 Test4 Tests . 290.00 310.00 Chest Clip Confort Breakage Breakage Breakage 177.00 179.00 Pelvis 292.00 267.00 299.00 281.00 Tap-Plast Chest Breakage 180.00 153.00 168.00. 173.00 Pelvis 281.00 306.00 305.00 277.00 287.00 Free Belt Chest 171.00 165.00 180.00 162.00 176.00 Pelvis 258.00 296.00 290.00 312.00 Cliper Chest Breakage 157.00 180.00 174.00 173.00 Pelvis 280.00 292.00 300.00 298.00 Joliy Chest Breakage 172.00 165.00 176.00 182.00 Pelvis 298.00 302.00 266.00 293.00 Lampa Clips Chest Breakage 175.00 176.00 157.00 164.00 Pelvis 300.00 312.00 272.00 296.00 306.00 Rapid Chest Lux 163.00 174.00 157.00 169.00 181.00 Pelvis (Min 202; Max 276) 250.17 .• .• Chest Standard belts (mean values of40 tests) (Min. 97; Max 173) 118.82 • .• Pelvis

Results of the crash tests. Values (in millimetres) for ehest arid pelvis displacement, referred to each Table I: single test with clips, and to the mean of tests with standard three-point seat belts. 40 Configuration according to Fiat "Tipo." 5-doorsaloon.

In nearly half of the cases (16 over35, that is not surprising that adding some centimetres is 45,7%) the recorded values were out of ECE of extra slack worsens the situation. The mechanisrns by which the belt 44 omologation criteria (admitted values: is 100-300 mm for ehest displacement; 80-200 broken seem strictly related with the per­ manent clamping of the web. A constant find­ mm for pelvis displacement). The high-speed film examination ing is the narrowing of the belt in one comer evidenced, in some cases, severe submarining of the B-pillar loop, with various degrees of and body displacement during impact. tearing; the cases in which the web is only The instrumentation set for the tests could incided show that early deterioration is due not give precise infonnations on ehest and to the narrowing in the rearward-facing comer shoulder loading during the impact. of the loop. This can be explained by supposing that the comfort clip excludes the small recall DISCUSSION if All the laboratory results clearly demon­ force of the retractor coil, then the loop is no strate a very bad perfonnance of the after­ more subject to moderate pressure from the market "comfort-clips" under the test con­ web, but is free to rotate with its bolt as a ditions. pivot. During the impact, the belt will work If we consider that a nonnally buckled with a badly positioned loop and will twist seat belt is already exposed to some 10-15 cm itself with possible tearing (Picture 4). of slack because of spool effect, delay in The increased ehest displacement could give a greater rate of head contacts with the retractor locking and web fibers stressing 3.s, it

- 17 - Web Picture 2: brcakasea _!.ter the crash lest.

steering wheel or the dashboard; since a head lt is important to observe that the mani­ trauma is always an "all or nothing" event, a facturer was identifiable only in two packages few centimetres may determine the difference of clips. bet\'.·een an unhurt and a wounded driver or In some cases, there were concise instruc­ passenger. Theslackat theabdomen is highly tions to the use, where any responsibility was dangerous, since it can allow the belt to slip denied for eventual damages due to belt slack over the pelvis with severe submarining. exceeding one-three centimetres. These two effects may considerably Conclusively, the "comfort clips" ex­ reduce the performance of belts and cause change a feeling of comfort (or psychological derangements in its action, with possible side freedom) with a real danger of belt malfunc­

lesions 1.2.4 (seat belt syndrome). tioning during an impact. Theseobjects should In some car models, the installation of a be removed from the market, and their use thick plasticcomponent between the coverage should be discouraged or forbidden. of the B-pillar and the web may cause friction, From a practical point of view, it does or actually lock the retraction of the seat belt, not seem profitable to install more and more allowing the belt to hang down. sophisticated and expensive restraint systems

, .,,.. Rapid Lux -6J • Chest v D Pelvls Lampa Clips 1 VJ / Jolly -61 „ Cliper 19 / Free bell t9 „ A Tap-Plast 1)' il A Clip Confort 1 i9 / 0 5 10 15 20 25 30 35 40 45 50 55 60 % Variation vs Standard 3·p seat belts

Comparison of ehest (dark columns) and pelvis (light columns) displacement with "comfort clips" Picture 3: i:·s the mean values obtained with standard three-point seat belts. The tests with breaking of belts are not considered in this diagram.

- 18 - c

Possible mechanism of belt brcaking during the deceleration. From the normal position the bell Picture 4: ( A), loop can tilt so that the web is forced in a corner and can be severely damaged by sharp plastic coverages or (8) by the metallic structure of the /oop (C).

on new cars (for instance, belt pretensioners), REFERENCES if cheap and sometimes raw components like 1. CHRISTOPH! C, McDERMOTT IT,McVEY 1. the tested ones can easily jeopardize the efforts HUCHES ESR: Scat bcltinduccd trauma to the of technology. small bowel. World J Surg, 9:794-797, 1985. lt is difficult to figure how a prevention 2. GARRETI JW, BRAUNSTEIN PW: The seat bclt syndromc. J Trauma, 1962, 2:220-237. of this sabotage to seat belts can be worked 3. HALAND SKANI3ERG T: A mechanical buckle out: even without the described devices, the Y, pretensioner to improve a thrce-point scat bclt. clamping ofseat belts is possible with clothes­ 12th ESV Conference, Göteborg, Paper 89-513-0- pins, paper-clips, or by forcing something 001, 1989. within the web and the loop. 4. MACKAY GM, HILL J. SMITH M, P ARKI:-.J S: The While a "foolproof" restraint system is limits of seat belt performance in crashes. 12th still beyond the horizon, it is clear that users' ESV Conference, Göteborg, Paper 89-53-0-003, 1989. education still represents a central topic in the 5. NEGRIN D: In: "Prevenzione degli incidenti improvement of seat belt effectiveness. stradali. Valutazione del danno fisico e formalizzazione de! linguaggio in traumatologia dclla strada". Paper, Milane, Hotel Executive, 23.4.1990.

6. NYCREN A: lnjuries to car occupants - Some aspccts on the intcrior safcty of cars. Acta Oto­ Laryngol (Suppl.), 395-1:164, 1984. 7. WALZ FH: Unfalluntersuchung Autoinsassen. EidgenössischesJustiz- und Polizeidepartement, Bern, 1982.

8. ZUPPICHINI F, HA LAND Y: Effectiveness of prctcnsioncrs on the performance of seat belts. XXIII Fisita Congress, Paper 905139, Procccdings (2):153-158. Turin, 1990.

- 19 -