Bicycle Helmet Safety Institute
The Helmet UpdateVol. 8, No. 2 - July, 1990
Consumer Reports Publishes a Helmet Article!As scheduled, the May issue of Consumer Reports contained a full-blown article on bicycle helmets, complete with impact testing results. This is the first article on bicycle helmets they have done, and it is providing a boost to helmet promotion campaigns.
CU will not permit us to send you a copy of the article, but you can get one directly from them. (See attached yellow slip, which we are now including in our Consumer's Guide pamphlet.)
When you read it, you will find some first rate test results and a few disappointments. The first-rate part is CU's ranking of helmets, primarily based on impact test results, and their judgments on the effectiveness of retention systems. The retention system test was particularly good, since it used a human test panel and emphasized convenience as well as performance. It also used an "anatomically correct" headform, which would tend to favor some helmets, since no single headform can be considered "anatomically correct" for all of the various shapes of human heads. The top four helmets were the Bell Quest, Performance Enduro (made by LT), Paramount Team Issue (made by MPA/Vetta)(See Schwinn's correction in the next newsletter), and the Specialized from Specialized Bicycle Components. Rated "unacceptable" were the Avenir (retention), Avenir Advantage (retention), Monarch Aero-Jet (impact, retention and visor) and Spaulding Youth (coverage).
The article has a good explanation of the Snell and ANSI standards. The overall rankings were done after careful study of a great deal of lab test data. But there were some disappointments. CU decided that hard shell helmets are too heavy to recommend, ignoring the evidence of hundreds of thousands of hard shell users over fifteen years in the field. Voigt Hodgson's research on the sliding resistance of soft shells was dismissed in a single sentence, and all of the top-rated helmets are soft shells. The article includes a penetration test, which is essentially irrelevant in a world without spears where people are injured by hitting flat or rounded things, without a word in the text on the significance of penetration. The text also leaves ambivalent whether or not the penetration test was applied to vents, (it was not) or how much of the space around a vent was avoided. CU subjected each helmet to a hot trunk test -- at 180° fahrenheit. We think that your gasoline would be boiling in your tank if your trunk were that hot, and you could not start your car. There is no explanation of why a soft shell helmet should stay in one piece after the first impact, or the effectiveness of reinforcing in holding an all-foam helmet together in a crash. There is no indication that a helmet may be too heavy for a very small child's neck structure, or that a pediatrician should be consulted if there is any question on that score. Finally, the article fell victim to the publication lags that beset all magazines. The helmets were procured in 1989, so the article misses out on all the new, modified and withdrawn models since then. CU expects that their next bicycle helmet article will not be sooner than five years from now. We think that illustrates why the bicycle helmet field still needs an ongoing testing program with accessible and frequently disseminated results.
Readers were alerted by the article that CU did not use the ANSI and Snell standards approach using a threshold of injury of 300 g's and instead used a "softest landing" criterion. The threshold approach is based on the assumption that if six helmets all prevent you from incurring a serious injury you do not care that much which of them provided the softest landing at a given drop height. You really want to know which of them will keep the maximum acceleration below the injury level at the highest drop height, representing the worst crash. The CU approach has merit if you do not hit harder than their test drops, or you are a senior citizen. It requires such accurate and consistent test data that it ignores normal sample-to-sample variations, and it would be difficult to verify or replicate at another site. (A Canadian researcher once flunked a Snell-certified helmet on Canadian national television to illustrate this point.)
After those little disappointments, this was still a terrific article, and a real landmark for consumers. It will reach people who will never see any other article on helmet brands, and will be the first place many consumers will look when they think about buying bicycle helmets. It carries the stamp of authority that only CU brings to the consumer information field. It required masses of data and the research was expensive. We learned a lot from the CU approach and testing, and would like to apply some of it eventually to our own test program. The article points up the need for a rolloff test in the ANSI and Snell standards, and challenges manufacturers to consider other design criteria than just keeping acceleration below 300 g at a given drop height. If you have not seen it yet, order it today.
CU's comments on this review are in the next newsletter.
Wayne State Study on Sliding Resistance of Hard Shells vs. Soft Shells Concludes that Soft Shells can Contribute to Neck Injuries but Hard Shells May Have Their Own ProblemsThe study of hard vs. soft shell helmets conducted by Dr. Voigt Hodgson of Wayne State University in Detroit has finally been released. At least two reporters jumped the publication date to rush early conclusions to you, but they had apparently not actually seen the study. A complete copy is attached to this Update.
Hodgson concludes that soft shells do have higher sliding resistance and can "hang up" on the pavement under certain specific circumstances, "predicting more risk of neck injury than in the no-shell helmets." Under other conditions he finds them not much different, and under some conditions he believes that using hard shell helmets may lead to facial injuries and rotational injury of the brain. He finds that "either type would be much safer than riding a bicycle without a helmet..."
There is a lot to digest in this study. Hodgson conducted impact tests using his own "humanoid" headform, which differs substantially from the metal or resin headforms used for the ANSI and Snell standard. His strap test also is an original. He uses the Gadd Severity Index, which not all researchers in the field find helpful, and there are no waveform traces to judge whether a given helmet's g number is close to its ultimate capacity to perform or there is lots of remaining capacity to handle additional energy. For those who are accustomed to judging helmets by the ANSI or Snell benchmark, it is significant that only one of the helmets tested permitted the headform to be subjected to more than 300 g's, assumed by each standard to be the threshold of injury. Using the Severity Index as a criterion for injury, however, a head in a Bell Brava which was subjected to a maximum of 196 g's was judged to represent risk of serious injury to 35% of the population. Clearly either the Severity Index approach or the approach used by the two standards is off.
Hodgson's photographs illustrate what happens in his lab when the angle is just right and sliding resistance is high: the head "sticks" to the pavement, bending the neck under, then suddenly jerks loose as the dummy keeps moving. Much the same action occurs with a hard shell, except that the sticking lasts for only about 35 milliseconds, rather than approximately 175 milliseconds with the soft shell. Angular acceleration (rotational jerk) is higher with the hard shell as the neck jerks the head back up. With the soft shell this component is lower, and of course if the neck breaks it would be lower still. In the lab the face of the dummy then hits the pavement in the hard shell helmet. Hodgson recommends a clear plastic face shield to prevent this, a device unfortunately not suitable for active bicycling.
A complete copy of Hodgson's report is appended because it raises as many questions as it answers. Only one surface roughness was tested, and that was considerably smoother than normal asphalt. We had noted differences in sliding resistance depending on surface roughness when we scrubbed a hard shell and a soft shell helmet around on different surfaces. Based on that, we suspect the no-shells' performance will be worse on rougher surfaces. No Snell-approved helmets were included in the small sample of hard shells, although there were five Snell helmets among the soft shells. The Snell helmets performed better than the non-Snell helmets, so the best hard shells were not represented. The tests used small size helmets, which sometimes are considerably different in liner thickness than adult sizes, and may show different performance characteristics because the helmet surface is closer to the center of gravity of the head. Slippage of the helmet on the relatively sticky rubber of the Hybrid III dummy is not comparable to a sweaty human head, nor is the impact of the Hybrid III's face on the pavement likely to produce the same result as a human face, since humans break bones and lose skin. We would be interested in any reactions to this research from any reader. We have already seen the Bell Helmet response, which tries to ignore the study's main conclusion about neck injury.
This report is the second landmark study of 1990, and will be discussed at every bicycle rally this year. Dr. Hodgson will be conducting further studies this summer. We should all be grateful to him for continuing, since we are not aware of anybody else who is working on this problem in a systematic way, despite the talk about it.
Full-sized copies of the attached report with clearer photographs are available from: Health Surveillance Section, Michigan Department of Public Health, P. O. Box 30195, Lansing, MI 48909. NO ACTION ON ANSI STANDARD REVISION
The revised ANSI Z90.4 Bicycle Helmet standard is still under revision. Committee members have commented on the September, 1989, draft. A "final" draft is still in the works. As we have noted, the biggest changes will be an increase in the flat anvil drop height from 1.0 to 1.5 meters and testing for strap strength after the impact tests.
Bell Helmets Provides Market EstimatesBell Helmets has been good enough to share with us their market estimates for total industry sales of bicycle helmets for 1989-1990. We are grateful to have these numbers from Bell, since they are not available to us elsewhere. Bell cautions that the usage rates, which they refer to as Usage Penetration rates, are approximate.
Bell: Helmet Sales and Usage Rates
Industry-wide sales (millions)
Usage Rate (all bicyclists)
Usage Rate (regular bicyclists)
Howard County, Maryland, Passes Mandatory Helmet LawIn May, Howard Coujty, Maryland, adopted a helmet law requiring all bicyclists over 16 to wear helmets. We are attaching a copy of the law, which takes effect in October. By next spring we should have some idea of how well such a law works in at least one setting. In Australia the State of Victoria's mandatory law took effect on July 1, and a law for the State of New South Wales will take effect January 1. Although these laws are probably premature while helmet usage is still so low, testing them in a few locations should provide interesting results.
BHSI Grants Top $4,000 for Year
A $500 grant from the Fund for Southern Communities brought our total funds raised this year over the $4,000 mark. The rest of our funds have come from generous individuals. This is encouraging, but with no major funding source for our program and test laboratory we are beginning to run out of full-time volunteers. Susan Matson, our chief fund-raiser, had to go back to paid work some months ago. I will probably follow her in the fall. BHSI will continue to operate with part time people and full time ambition!