Should Harnesses Be “Fail Safe”
Recent events and the data received from them, i.e.: the fatal harness failure in Europe and the Type 17/Mini-Ring problem, should make us look into the structural integrity of our harnesses.
As a leading manufacturer of harness/container systems we feel that it is imperative that we share the results of our investigations with the public and the industry if it is for the common good of the sport.
The best way to tell the story is anecdotally with factual substantiation, and it begins in Zephyrhills in the early 70’s. John Sherman had designed and built a full step-in harness for himself and a few friends. The harness was built using one long continuous piece of webbing which was wrapped around the body and held in place with a few stitches. This harness had no chest strap, as it was for a chest reserve, for which it had “D” Rings on the top main lift web. The hip junction was a 90 degree crossing of the horizontal and the main lift web. Here was used a 4 point Double “W” the width of the webbing with no backing. This stitch failed under a hard ram air opening. Fortunately, the jumper suffered no discomfort and landed without further incident.
Sherman then did an investigation and redesign of that area and came to the configuration that he uses today. He learned “a bunch” about harness structure and over the years shared that information with other rig manufacturers as they shared with him. We have learned that the stress on that joint comes at a diagonal to the stitch line and that much efficiency is lost as a result. John’s approach was to try to deflect the stress back into line by using a longer stitch with a stiffener. Another approach was to realign the stitches to the force line. After testing, we decided on the first method. The difference in the strength was minimal. They both failed in the area of 2500 lbs. to 3000 lbs. load. Bill (Booth) has indicated that this is about the same result he has achieved and we would venture to say that this is true for just about every harness out there.
The specification for harnesses originated with the round parachute. The round parachute, just like the square, began its life without any opening shock attenu- ation. Just out there and “Bang”. The first TSO standard was written with two levels of security, one for low speed 3000 lbs, and one for high speed 5000 lbs. The low speed required 2″ Block letters indicating restriction to use in aircraft under 150 MPH. Sherman did not want to put 2 inch block letters on his nice, new piggyback design so he certified his harness to the 5000 lb. high speed category, most of the other guys followed suit. Fifty percent of 5000 lbs., the strength requirement of one side of the harness, was more than enough to pass the structural tests as the round parachute generally produced close to a balanced load. Now comes the square. The square underwent considerable development to get where it is today. Without shock attenuation the square is only good for deployment at very low speeds and even then it opens hard. We have learned over the years that openings can be greatly one sided. Para-Flite reports measurement of 80/20 distribution and projects differentials as much as 90/10. Fact: an unreefed ram air parachute at terminal can destroy a parachute harness. It has happened. A harness has failed. A man is dead.
The type 17/mini-ring problem has been redefined as an opening shock problem and so it is. However, it has brought to light a number of facts. The maximum load capacity of the mini 3-ring is about 3600 lbs. We were failing risers in the field with forces as low as 2000 lbs. and as high as what is to be believed as 3000 lbs. These failures may have actually saved lives by providing a weak or fusible link to the system. The point is that we have definition of opening shock loads, from the field, that may fail harnesses.
A little more about the harness failure in Europe. The harness was not certified under the TSO system and it was not manufactured by a U.S. firm. The design is one in common use, however. The risers were type 8. The webbing according, to Hans Ostermunchner the investigator, was mil spec. type 7 and the thread was mil spec. 5 cord. Both were in spec. as determined by the investigation. Other sources report that the other side failed at 1800 lbs. – a differential from design of about 700 lbs., maybe 25%. This is not a surprise and could be explained by differences in the dynamometer mandrel diameter or testing methods. Whether the number is 1800 or 2800 is academic. We are in the range of potential failure, and shock load predictions and measurements – and field failures demonstrate this fact.
Mike Fury made the statement at a somewhat recent PIA meeting that people should come apart before harnesses fail. He is absolutely right! However, the state of the art in harness design prevent that goal from being achieved in reality. Before we get confused let us define what we think Mike is saying. That means, define the word “fail”. To a harness maker it means no failure of any stitch at any time. Realistically that is not possible with today’s ram-airs and the information discussed earlier bears this out. Ideally, the way to frame what Mike is saying is to say that it must “FAIL SAFE”. That is, that if a complete primary stitch pattern fails, the user must not come out of the harness. The body retention parameter must be maintained. Just like John Shermans’ friend in Zephyrhills who didn’t come out of the harness when the stitch failed no jumper should ever come out of any harness as the result of a stitch failure.
The TSO tests don’t allow for stitch failure during the tests and we are confident that all of the designs out there have passed in the proper manner. But the TSO tests have not provided for the asymmetries that have been recently defined. At this point we’re still unsure how to write such a test when and if we sit down again to try to cover this problem.
The way we go about solving this problem is to educate the jumping public about packing of ram airs and about the potential of hard openings. We must additionally, adopt resolutions and or regulations that require future designs to be “Fail Safe” and we must phase out the old non fail safe designs. Retrofit packages might be developed to save the old designs. By using the education first approach we will probably solve the immediate problem with out a major financial impact on the industry. The industry will be spurred to advancing the technology of safety in harnesses by forcing the phase out of the old less safe designs. Remember we are dealing with a market driven economy in a life saving business – a conflict in terms if viewed in the best light.