Control Arms

[Wally]

Just looking for a consensus here to get an idea, what tube sizes/wall thickness/material for Upper & Lower CA's is everyone using?

[Wally]

Just looking for a consensus here to get an idea, what tube sizes/wall thickness/material for Upper & Lower CA's is everyone using?

[Travis R]

1" x .095" Inconel of course. http://fsae.com/groupee_common/emoticons/icon_wink.gif

[BStoney]

I see that the winter months of fabrication are now setting in...

Man that is great....

[-]

you're using tubes.. we'll have to throw out those 1" solid steel bars we picked up the other day for our control arms..

[Denny Trimble]

Nah, keep 'em. Just get a really really long drill bit...

[Cement Legs]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by -:
you're using tubes.. we'll have to throw out those 1" solid steel bars we picked up the other day for our control arms.. <HR></BLOCKQUOTE>

Dont throw them away... we are looking for something to mount our brake lines to...http://fsae.com/groupee_common/emoticons/icon_wink.gif

[Frank]

5/8 * .035 seems standard

up to 18" long

[Wally]

thanks!

[MikeWaggoner at UW]

You might want to do careful analysis.....

http://students.washington.edu/dennyt/fsae/a-arm_failure/ http://students.washington.edu/dennyt/fsae/a-arm_failure/

[Travis R]

Interesting.
It looks like it failed at the weld behind the insert.
We've run as small as .625 x .029 with out any problems.
Good luck

[Frank]

.029 is considerably more difficult to weld although

we actually discourage the use of .029 on our car altogether

(but very much encourage the use of .035 everywhere possible)

[Wally]

I'm working on the analysis, but my fea skills are in their infancy. What loading conditions should I apply to the simulation?

[Denny Trimble]

It's all from the tires. Skidpad results in Detroit show 1.3 lateral g's, but that's usually on cold tires. So, design for 1.5 or more g's, which means an approximate tire coefficient of friction of 1.5. This means the car can (approximately) pull 1.5 g's in cornering and braking, less in forward acceleration unless you're running AWD. So, assume braking and cornering accelerations (alone and combined), plus some bump acceleration, find the worst case normal load on the tire from load transfer, the lateral and longitudinal loads (worst case normal * 1.5), and resolve them into balljoint forces. The combined loading case can be found by assuming a "total g vector magnitude" of 1.5, with lateral and longitudinal g's less than 1.5. Piece of cake... http://fsae.com/groupee_common/emoticons/icon_smile.gif

Don't forget the factor of safety for stress, and remember to calculate your camber and toe stiffnesses, to compare with physical testing. There's no sense in going crazy with your kinematics for that extra 0.1 deg/deg on your camber curve, when your structure is giving up .2 degrees due to flex and balljoint compliance.

For FEA, you can either model a spherical contact surface and restrain it radially, or taper your tube down to a solid point, and restrain translations but not rotations. The first case will be more accurate but harder to mesh, the second will be easy to mesh but give you artificial stress peaks at the points. Don't forget to restrain the outer balljoint vertically at one node to prevent rigid body motion of the part.

[-]

I have seen a few a-arm failures due to teams unfortunatly using to small diameter and too thin wall for their a-arms.

The one of note was a team that used 1/2 thin wall tubes (don't know the wall, but it was really thin from a glance). Needless to say while they were driving their endurance event, they hit a cone, and the cone sufficiently dented the tube so that when the a-arm was loaded it again it buckled like a pretzel.

Take this as warning, would you rather sleep well at night knowning you won't have your a-arms fall if they are damaged even a little bit or, or save a little bit of weight on your car?

[Big D]

Our 2004 car had 3/4" tube all around (sorry I don't know the wall thickness off the top of my head). This stuff seems fine on the front of our car, but not the rear. We (Saskatchewan) had one of the widest tracks, and narrowest rear sections at the comp (i.e. really long arms). We haven't buckled any, but when a driver launches with decent grip, but some tire spin, the rear wheels will shimmy forward and back quite severely. http://fsae.com/groupee_common/emoticons/icon_eek.gif

So that's something else to consider, especially if your arms are long. Even if it will survive, it should be stiff enough not to whip around like a wet noodle. Whipping around like that might also contribute bending into your rod ends that shouldn't be in bending. (See the other thread on rod ends... but personally I think that a rod end on anything but a single straight link is a horrible idea)

[Lash]

Wally,

0.25 X 0.028 is wwwaaaayyy to thin. Any kind of bending (hiting a cone!!), breaking/accel, will cause bucking prety easy. DId you FEA what Denny was saying? I'm sure you'll get yielding at the least. I like the previous reply that mentioned being safe rather then being sorry. Save alittle and sleep at night rather then saving a whole lot and risk breaking.

P.S. Has anyone ever heard of using 1/4 X 0.028 wall on the upper control arm? Thats like puting a pencil on there!!!

[Nate Notta]

Lash, why don't you just talk to Wally about this tonight at the shop? http://fsae.com/groupee_common/emoticons/icon_razz.gif

[Mark Bacchetti]

Our last suspension guy designed for the worst case loading... rookies picking the car up by the A-arms.

We still keep them away from the arms when the car is in transit. It is heavy but has saved us many times from a "whoops".

Mark
-Cal Poly Pomona

[dartmouth01]

I saw this great idea in a picture somewhere on this board that I think I'm going to have implemented this year. On the control arms, stickers were placed on them that said "Do not grab here". So simple yet so effective.....