The following information is provided
for your consideration. Opentracking.com assumes no responsibility for its
accuracy. Review our legal disclaimer at the bottom of this page.
--------
Notes On Race Car Harnesses - Design and Installation
©Michael Henderson. Written for ‘The Greasy Rag’, an Australian vintage
racing publication. Reprinted in its entirety and with the author’s
permission from CSRG News, a monthly publication of the Classic Sports
Racing Group.
During private practice at Amaroo recently, a very quick Sports 1300
went straight on over the hill and down into the earth-filled tire wall
outside the sweeper. The car was badly damaged, with most chassis rails
bent and the engine/gearbox shifted substantially forwards.
The driver, David Williamson, was extensively injured:
• fractures and dislocations both feet and ankles, necessitating a
wheelchair after discharge from hospital;
• undisplaced fracture of cervical vertebra;
• fractured ribs, punctured lung;
• bruised spleen;
• fractured left lower arm.
Most of these injuries, if only slightly worse, could have been fatal or
caused permanent disability. David has readily agreed to my telling his
story.
None of us like to think about being hurt in a race car. But while
exploring the edge of control is part of the kick we get from racing, it
seems reasonable and socially responsible to minimize the risk of being
hurt if we crash. I first started analyzing race crashes in 1966, and
successfully convinced a skeptical world that it was better to be belted
into an open car than to be thrown out of it. My development of the
six-point harness with the GQ Parachute Company, taken over later by
“Dumbo” Willans, was part of this work. I still analyze crashes, but
road cars these days. Knowing my background, David asked me to examine
his car and his injuries, to see whether they might have been prevented.
This process was so instructive I thought it was worth writing down some
of the lessons in the context of what new research is showing about race
car harnesses.
- First, we worked out what happened to him in the crash. The actual
speed of impact would be an estimate, but is likely to have been in the
order of 100 km/h. The impact was slightly angled, so the speed might
have been higher than that. The car was effectively crushed about
two-thirds of a meter, and the relatively stiff tire wall probably
compressed about one-third of a meter. The total stopping distance of a
meter then gives the deceleration force on the car at around 40 g. -
Now we know from the black box crash recorders in Indy cars that drivers
can ride out 40 g crashes with no more than bruising (the limit of human
tolerance is being approached at about 50 g). Why not in David’s case?
The first reason is that David “submarined”. Basically, he slid partly
underneath the lap belt. As it rode up his stomach to his chest it
bruised his spleen, then it went up and broke some ribs, which in turn
punctured a lung. His feet were forced down to the footwell, with
forces being directly transferred into a collection of vulnerable small
bones and joints.
Hanging on to the wheel, his arm was broken as he instinctively tried to
stop himself going forward. This was not enough to stop him being
violently flexed over the lap belt so far that he hit his helmet hard on
the steering wheel, bending the rim about 30 degrees forward. The
impact was enough to break a bone in his neck, just as it was at full
stretch.
Because problems with the harness and its installation and use could
have contributed to these injuries, we studied it carefully. It was a
six-point belt, with two-inch webbing. The shoulder belts had been
routed over a transverse chassis rail behind his shoulders and down to a
lower rail at the bottom of the car. The crotch straps were joined at a
central single latch plate. David confirmed that they were quite loose,
and could be clipped into the buckle very easily.
The submarining happened because there was little to stop it. Crotch
straps are there for two reasons. Not only do they have a direct effect
in preventing sliding underneath the lap belt, they also stabilize the
whole system. Unlike the tree-point belts in a passenger car, the
buckle of a race harness is in the center. This means as soon as the
shoulder belts are loaded, they pull the lap belt upwards and the lower
part of the body tries to shoot underneath. This killed Jochen Rindt,
who told me in 1969 that although he had come to accept a harness in the
Lotus, he would never wear crotch straps. In his crash at Monza his lap
belt ended up near his neck, rupturing his liver on the way. Australian
child car seats, which also have central buckles, have crotch straps for
exactly the same reason.
The excessive flexion of the upper part of David’s body started when he
finally got held up by the loose crotch straps, by which time his feet
were mashed in the footwell. The flexion was allowed by the
geometrically loosened shoulder belts and increased by the stretching
allowed by the long length of the straps.
General Motors has been doing Indy car crash simulation (sled) testing,
using dummies, for about five years. We now know far more than we did
about the details of what happens in this kind of crash. It turns that
what we thought was right in the beginning, was later proved to be
pretty right all the time.
At impact, the whole body moves forwards until lap and crotch straps
restrain the hips. Then the torso rotates 30 degrees until the shoulder
belts stop the rotation. Most of this movement is due to changing belt
geometry and shape changes to the body within the belt, even with the
mounting points just behind the shoulder. High-speed movies of a
correctly-restrained dummy reclining at 45 degrees show forward shoulder
movement of about 250 mm in a 40 g crash. About 20% of this forward
motion is due to belt stretch, working out to about 50 mm.
This confirms two things for us. First, to restrain body movement
within the confines of the shoulder straps, the belts must be as tight
as possible throughout the crash sequence. And that means tightness in
both the shoulder and crotch straps, which balance each other. Second,
the shorter the shoulder belts, the less the total stretch. In David’s
Sports 1300, fibers in the shoulder belts were melted as the webbing
stretched over the chassis rail.
Not only do slack belts allow more movement - or “excursion”, as we call
it in the trade - but they also directly increase forces on the neck and
chest because the body’s deceleration is more violent. The body slams
into the belts at the pre-crash speed of the car, instead of riding the
crash with the car as it collapses. Loose shoulder belts are a threat
to the neck.
It’s very important to get the shoulder belts loaded as soon as possible
in the crash. The best way to arrange this is to place the mounting
points so that the belts leave the shoulders at about 90 degrees to the
spine. If you sit bolt upright, take the belts straight backwards; but
when reclining, the mounting points should be below the shoulder but not
so as to take the belts back along a line 40 degrees below the
horizontal. When the diver is reclining, horizontal shoulder belts don’t
get fully loaded until well into the crash, by which time the shoulders
will have moved well forward.
Whatever the car, lap belts should be anchored near vertically (say, 80
to 90 degrees), with the webbing passing over or in front of the hip
bones.
In a 40 g crash, an 80 kg driver will be loading the belt system with a
3,200 kg force, about twice the weight of a fully-laden Falcon. The more
widely that load is distributed, the lower the risk of belt-induced
injury. That’s why three-inch (75 mm) belts, with reasonably stiff
webbing, are far better than belts with two-inch straps. They'll soon be
mandated by the FIA.
Crotch straps can’t be made of three-inch webbing, but in any event they
should bear on the bones of the pelvis between the legs. This allows
them to be really tight without discomfort, and puts crash loadings into
strong (and hard) body parts rather than weak (and soft). Crutch straps
take a heavy beating in a crash. A friend recently stuffed a March very
hard into an earth bank at Oran Park, and two days later - without other
injuries - had two jet-black bruise lines across the top fronts of his
legs.
To get the right geometry the crotch straps must be widely separated as
they approach the lap belt, just like a parachute harness. Easily the
best way to do this is to take them up through D-rings on the lap
belts. Every Formula One car has this kink of system. Taking the twin
crotch straps (or, worse, a single one) via a single latch plate to the
bottom of the buckle is a compromise aimed at cutting cost and adding
convenience, mainly for Sedan cars.
Only a few race harnesses comply with my criteria: three-inch webbing,
D-ring crotch straps, a central rotary buckle of course, and details
such as spring-loaded anti-slip adjusting buckles. They include the
following, and I readily concede there may be more I don’t know about:
• Willans 3" Silverstone 6;
• TRW/Sabelt 3" Professional 6-Point;
• Stand 21 STH-36SS 3";
• Simpson 3" 6-point.
The best harnesses don’t cost a lot more that the worst, and less than a
couple of tires these days. Fitting and wearing them properly is free.
As David told me when I was writing his story, if only one driver’s pain
is prevented by better understanding, then it makes his own pain worth
while.
BIOGRAPHICAL NOTE
Michael Henderson M.D. is a physician and internationally-known motor
vehicle crash researcher. He has also been an auto racer since 1960.
He was the author of the seminal “Motor Racing in Safety” in 1968, and
influential in turning round Formula One and other open-wheel racers to
acceptance of full harnesses. His first prototype harness became the
Willans, now used world-wide. He currently drives a Lola T560 Formula
Atlantic in vintage races, and also has a Lotus Eleven (Dick McGovern’s
old car), Elva-BMW and a Ralt RT4. in June, Michael was involved in a
major crash with the Elva-BMW, the car was a write-off, but Michael came
through it with “limited” injuries, and will be racing again soon.
