Don't worry, the
guy who welded this fixture (that's me) is not the person who
welds our chassis and suspension components. As with sausage making
and legislation, not everybody wants to see how structural testing
is done, but when us citizens spend billions on, for example,
developing a B-2 bomber, some of that gets spent load testing
the pedals--and those test fixtures aren't much more sophisticated
than this one.
Load test fixtures
for individual parts have to meet two needs: they need to support
and stress the part in a realistic simulation of the part's actual
use, and they need to be strong enough to test the part, rather
than test the fixture. When designing a test fixture, weight doesn't
matter and pretty doesn't matter, but time is money so one gets
points for expedient, quick and functional, and this one fills
The pedal is mounted
in our pedal bracket, which is welded to the fixture frame, and
pivots on the lowest quality hardware we think it's likely to
encounter—a Grade 2 coarse thread bolt with no nut. The
brake pushrod, with pushrod clevis from a first generation Miata,
fits into a rounded pocket on the fixture, which allowed us to
test pedals drilled for a pressure ratio of 4/1 to 6/1 (need #1:
realistic simulation) without mounting it to a master cylinder
(need #2: test the part, not the fixture—using a master
cylinder would have tested if its seals could handle over seven
tons psi* but might not have tested the pedal). The load (sandbags
on a wooden platform--not shown in this photo) hangs from a length
of seat belt webbing (not placed on the pedal in this photo) and
the fixture is lifted by an engine hoist until the load leaves
the ground. The sandbags were weighed on a Pelouze Model 4010
digital platform scale.
In aviation the
standard for any metal structural component is a demonstrated
ultimate strength 1.5 times the design requirements, times 2 again
for castings. The X2 casting factor is based on centuries of experience
with structural castings (though only a century of that includes
airplanes), and while there is some variance in strength between
parts cast from the same mold, the worst is at least half as strong
as the best.
I don't know what
the SAE calls for in brake pedal design strength (any card-carrying
automotive engineers out there that can help me out?), for Light
Sport Aircraft it's 125 pounds for pedals (rudder or brake)--times
1.5 times 2 is 375, so a 375 pound ultimate static test would
qualify these pedals for an LSA, suitable for FAA certification
to carry passengers and fly a mile above your house. But since
automotive standards for strength of critical controls may be
greater than the FAA's requirements for equally critical controls,
let's double that to 750 pounds.
We tested four samples.
I'll admit I breathed a sigh of relief when the first one reached
750 pounds, but I sure wasn't surprised. At 1000 pounds I wasn't
surprised either—it was designed to be good for 1000 pounds,
plus a tiny fudge factor—but I was keeping my faceshield
down and my padding in place and it wouldn't have been a shock
if it had failed at that point and half a pedal had been hurled
about the room at Civil War artillery speeds**.
it didn't. Here's a photo of a pedal holding up 1072 pounds of sandbags
and support structure.
at 1072, and all four pedals survived without yeilding (bending), much
less failing (breaking). The pushrod clevis did yeild, but never failed,
so I'm not going to complain to Mazda.
Here's a closer
look at the fixture, still with 1072 pounds on board. I set it up with
the pedal horizontal (at right angles to the load) because that is worst
case, and used the seat belt webbing to distribute the load over the
footpad. The seat belt webbing is about 1/2" off center because
I didn't want it to slip off, but oh well, close enough.
loading at 1072 pounds. Our next available sandbag was an 81 pounder
and the hoist is rated at 1000 pounds in this setting; I was willing
to bet that the manufacturer has a 10% fudge factor built in to that
figure, but that's as high as I'll go.
Also, at 1153
pounds (the next step up) I'd expect a pedal to fail, and I didn't care
to replace a window just so I could say "I told you so." 750
pounds was all I needed to be happy, going over 1000 was just plain
...or back to Pedals
*Using a 6/1 pressure ratio
at the pushrod, a 1072 pound load, and a master cylinder with a 3/8"
radius (3/4" diameter) piston, we get (6 x 1072) / (pi x ((3/8)squared))
= 6432 / .4418 = 14,559, which is a bunch. It's about 1000 times atmospheric
pressure, it's the pressure at the sea floor about six miles down...I
don't think most brake seals could take it.
**Not a joke, by the way.
Doing tests of this sort, I wear a full coverage helmet with face shield
<and> safety goggles, gloves, a leather jacket over a Mr. Rodgers
comfy sweater, and other strategic padding. The energy stored in the
stretch of the seat belt webbing and rope, the flex of the engine hoist
beam, the bow of the 4 by 4s and plywood under the sandbags...at 1000
pounds load they're capable of tossing a quarter pound of aluminum faster
and harder than any crossbow bolt of the Middle Ages, and armor didn't