Control Arm Tests


IMPORTANT NOTE: These tests may satisfy us, but that doesn't mean they have to satisfy you. If you're building yourself a car, you're going to have to decide for yourself what is good enough and what is not. We'll give you honest test results and hope they'll help you make an informed decision, but the bottom line is, if a suspension member falls off your car, you and/or others may be injured or killed. So take what you're doing seriously, and in your search for information, consider how it relates to your particular project.

Because our latest upper control arms are different from "standard Locost parts" (as if there were such a thing) we felt they needed structural testing before they hit the streets. Our concern was the tab-and-clevis junction between the front and rear tubes all the other components have a long history in LocostLand and though the connection looked good on paper, we're belt-and-suspenders folks and like to back up our calculations with physical tests.

Here's a right control arm, as seen from the bottom.

&and here's a closer look.

The tab is laser cut from 1/4" mild steel plate. Hole diameter is .377" (lasers aren't perfect), and at the thinnest point, the edge of the hole is .310" from the edge of the part.

The ball joint shown is a Moog ES323R, a component I don't currently recommend for reasons you'll soon see.

The pneumatic ram (above, left) has a 7 square inch piston, which is hoisted by regulated air a maximum of 840 pounds ram force at an air pressure of 120 pounds per square inch.

The upper arm (see closeup below) has four lifting points, which use leverage to increase the ram force by a factor of 3, 4, 5 or 6, and convert the force from compression to tension.

The chain above is experiencing 5000 pounds of tensile force; 120 pounds psi times 7 square inches times 6/1 leverage factor...okay, that's 5040 pounds, but there are some friction losses so we've rounded it down to 5000.

By the way, at 4500 pounds tension (I back the air pressure off 10% before I get near the thing) that 3/8" chain feels like an andiron.

So that's the tensionizer. The structure is made from 1-1/2" .095 mild steel square tubing with 1-1/2" .125 mild steel plates at the fulcrum.

Load carriers are made as needed, from 1/8" mild steel, 3/8" hardened chain, and 1/2" grade 2 bolts (which did dent some during the 5000 pound chain pull test).

Here's our first pass at a control arm carrier. Essentially it simulates the frame, and is made of the same materials as the tension generator.

The variety of holes on the lower arm of the fixture allow us to pull the control arm from a variety of angles. The settings above simulate a combination of braking and cornering forces; roughly 75% braking and 25% cornering, with a 2400 pound load. The 1" square tubes along the sides are guides to keep the control arm from flopping over to one side as the loads go up. Note that the 1-1/2" square tube that is the base of the control arm fixture is bent. After the control arm passed the 2400 pound test, I tried for 3200 pounds and the base tube yeilded--I couldn't get the load over 2800 pounds.

That was the most interesting and surprising result of the tests. I hadn't bothered calculating the load and strain on the fixture because...

Well, because Locosts are made out of 1 x 1 x .065 and that base tube is three times (a horseback estimate, but close enough) stronger and stiffer than the chassis tubes, so why even check, right? The fixture kept bending until the piston topped out at 2800 pounds.

I <think> it was staring to yeild at 2400 but I can't say for sure because I didn't check it that pass...let's see; to get a 2400 pound load straight ahead, figure a 1600 pound car (1350 plus driver and fuel) with a weight shift on braking of 75% front wheels, 25% rear wheels is 600 pounds per front wheel, so you'd get a 2400 pound load from roughly 4 Gs braking force...that's a lot but you wouldn't expect permanent frame deformation, would you?

Anyway, these preliminary results imply that the chassis may be the weak point if one is, for example, screwing around on the street and finds an unexpected speed bump in the way and hits the brakes hard at the wrong time.

So we bult a new control arm carrier, using 1-1/2" x 2-1/2" .083 rectangular tubing nothing engineered, just the strongest stuff we could find at the steel store and ran it up to 3200 pounds. Also we traded the steel rear for a Rod End Supply swaged aluminum radius rod 'cause I've been curious what it would take to pull the threads out of those. In practice, 3200 pounds is an impact load for a Locost, and unless you're building yours with a Chevy Rat or a Ford Windsor under the hood, you'll have to hit something solid to put this sort of load into your upper control arm. Still it's nice to know there is ample safety margin in the tab-and-clevis hookup.

However, 3200 pounds pretty well uses up the safety margin of the 5/8" rod end on the spindle side of the control arm. It had shown the teensiest symptom of yeild from an earlier pass (the jam nut, which had been put on finger-tight, felt slightly draggy to the touch after 2400 pounds), but it's unmistakeable here. This particular ball joint is an EM10 from RES's "economy" line; since this was expected to be the first yeild point I figured I'd bend a cheap one instead of an LXM10.

The tab yeilded a bit too. The 3/8" hole elongated .007" and flared the steel on the load side about .002" which was enough to make it hard to disassemble but not something I see as a failure risk.

If I ever hit an unyeilding object (a pothole, a fireplug, a Chevy Suburban) I'll probably look here first to see if the upper balljoint/rod end/track rod bent. And I think I won't use ES323 track rod ends any more, because they're a half inch longer than a standard 5/8" ball joint, and thus have a longer cantilever and thus more stress where they spigot into the control arm.

The Moog ES312RL is the same length as the ball joint, so that's my standard for the time being. Its only drawback is it's only available in left thread. Oh yes, and you need a tapered reamer to get its stud to fit full length into the Miata spindle.

I'd be interested to know if the history of minor Locost accidents (and there's gotta be a few by now) shows the upper outboard rod end yeilds before the chassis in real life. I'd wager the chassis bends first.

Above is that 1-1/2" x 2-1/2" .083 rectangular tubing after the 3200 pound test. It's bent. After the load came off, it stayed bent. The load/distance on this tube is about double what it is on the book frame's FU1 and FU2, but still, this is a pretty sturdy tube compared to 1 x 1 x .065".

Since we have CMC Locost frames stacked up like cordwood around here, I'm sorely tempted to reproduce these tests on an actual frame, but I'm pretty darn sure the result will be permanent deformation forward of the firewall. But that's another test; this was a test of control arm strength, and I credit it as a Pass With Honors. Well, maybe a B+ for the outboard rod end, but an A for the arm itself.

It appears that the clevis-and-tab arrangement is sufficient to the needs of this control arm, so as of 10/28/06 the clevises and tabs are available for purchase. The clevises are $4.75 and the tabs are a whopping ten bucks; at which price the first 30 cars worth will cover our expenses drafting, converting the drawing to code to drive the laser, materials and cutting charge, fixtures and test samples and the next 30 will reimburse me for my time on this particular project, at roughly minimum wage. After that I'll be rolling in dough, but for now I'm just lucky I enjoy this work.

For those who are even lower on the wage scale than Yours Truly, we're providing a tab drawing (see below) so you can make your own tabs for your own car.

For those who are higher on said scale, we'll deburr the clevis, sand the tab down to a snug fit, and package them together for $5.25 a job you can easily do yourself with some 100 grit sandpaper, a sanding block, and a popsicle stick.

For more fabrication information regarding these control arms, click here. That will put you in the middle of an earlier [Sage advice>Works in Progress] page about Quick Adjust Control Arms.

&or more &or back to the top