Way heavy roll bar discussion

Discussion in '1969-1972 K5 Blazer | Truck | Suburban' started by jeffer, Jun 29, 2001.

1. jeffer1/2 ton status

Joined:
Nov 6, 2000
Posts:
113
0
Location:
I have been looking at getting a rollbar for my '72 K5. I have looked at different places. Most offer the standard hoop 4 pt design. I had some questions about tubing size and thickness and wondered what is the best to go with. Just a little background here, I am an electrical engineer and asked a couple of the mechanical engineers that are buddies/coworkers of mine about this from a technical standpoint. I will list below my question and the response. It's pretty theoretical, but I found it interesting nonetheless and thought I would pass it along. I hope that someone is interested in this babble.

---------- My note the to the mechies --------
Question: If I were to determine which of two tubes was a better support structure, how would I go about it? I'll just cut to the chase. I'm looking at roll bars. Bar #1 is 2.00" OD with a wall thickness of 0.120". Bar #2 is 1.75" OD with a wall thickness of 0.134". Now Bar #1 has a cross-sectional area of 0.7087 in^2, while Bar #2 has a cross sectional area of 0.6803 in^2.
Now lets assume that the material is identical. Is it better to have a thicker tube (albeit with less material overall due to the smaller cross-sectional area) or the bigger diameter? Or, is there no real way to determine this empirically?

Thanks,

Jeff.

---------- The Response -----------

Tubes supporting a structure:
Failure mode: column buckling.

Euler Column Buckling equation:

Pcr =
(pi)^2*E*I
----------
(Le)^2

where
E = Modulus of Elasticity (fixed by material)
I = moment of inertia of the cross section wrt the buckling-bending axis
Le = equivalent length depending on end conditions (sliding, hinged, fixed, etc.)
Pcr = critical force to failure
pi = delicious

I is the important [censored], for you, young EE. More I means it can handle more force (note, [censored]'s different if you want to design for stress, but I don't think you do).

~*~*~*~

I = A(rho)^2

where
A = cross-sectional area

rho(slender rod) = sqrt((Dod^2+Did^2)/16)
A1 (OD=2", t=0.12") = 0.7087 in^2
rho1 = sqrt((2^2+1.76^2)/16) = 0.666 in
I1 = 0.314 in^4

A2 = 0.6802 in^2
rho2 = 0.573 in
I2 = 0.224 in^4

If my math's correct (and it rarely is), then go big and go home.

~N

(I had to pull out Juvinall. I'm not enough of a nerd to remember the one day of Euler Buckling off the top of my head.)

------------- The weanie ----------

I didn't realize that thicker wall tubing is not necessarily better. OD counts for quite a bit in structural rigidity. Time to go have beer before my brain explodes.
Jeff.

2. Burt4x43/4 ton status

Joined:
Feb 17, 2000
Posts:
6,854
0
Location:
California (Modesto area)
wo dude, like I skiped that day in school!
don't you know this is an english speeking message board?
heheh JK
I hope that all did something for you. My backwards ass brain sure the hell got corn fuzed!!

72K5Led ZeppelinRock ON!
<A target="_blank" HREF=http://www.geocities.com/baja/5099>http://www.geocities.com/baja/5099</A>

3. arq1/2 ton status

Joined:
Jan 17, 2001
Posts:
2,017
0
Location:
Imperial Valley/Maricopa
I would think that the tube with a larger OD would provide more stability, like a wider car. but then again it ain't that strong.

This is more of a Steve Chin question.

ARQ.

1-72 4x4 CST Blazer
2-71 4x4 CST Blazer

Joined:
May 29, 2001
Posts:
214
0
Location:
5. Steve_Chin1/2 ton status

Joined:
Mar 6, 2001
Posts:
1,126
0
Location:
Foster City, CA, USA
I dabble in the mechanical end of things (design engineer for mechanical and compliance issues) and I've done a few cages in race cars. The material with the larger OD will normally be stronger than the one with the smaller OD, given the larger OD material also having at least the same cross-sectional area (to steal a phrase: Go big and go home).

Besides, on a vehicle as big as a Blazer, the 1.750" OD stuff looks spindly and the 2" stuff looks at home.

Oh, a couple of other notes: No mandrel bends! Mandrel bending thins wall thickness at the outer radius of the bend. Make sure the bending is done on a machine that will make smooth bends without kinks. Placement of the pickup points on the floor boards is important, as is the size and shape of those pickups (which really has to be determined by where the bars will intersect the floor). Make sure that the sheetmetal is supported on both sides of the floor sheetmetal and that the support on the under side of the metal is larger than the support on the upper side. If you can tie the supports on the under side of the floor to the frame, that is good, but it is not absolutely necessary at all points. The supports should be at least 6" x 6" in size, and at least 0.250" thick.

6. tshikoRegistered Member

Joined:
Apr 13, 2001
Posts:
26
0
Just like Steve says, watch for you support points. Chances are the tubes are a lot stronger than your floor board. If you have the least kink in your tube that is where it will collapse and at a fraction of what you cal Pcr. But I agree with all you calcs so far. If I were to design a roll bar I'd make sure at least the two main pillars bolt to the rails of the frame. Let us know when you have figured what size welds, bolts, etc. to use ....

7. Triaged1/2 ton status

Joined:
Feb 21, 2001
Posts:
3,808
0
Location:
CA (LA/OC area)
Wow. I actually get to use some of the stuff I learned in school! I just had that in strengths of materials class (I'm a Mech. Eng. major at Cal Poly Pomona). The other thing to look at in that equation is Le. This has to do with triangulation. If tringulated properly and the floor mounts are sturdy than Le will be smaller making a stronger roll bar.

'71 Blazer CST with alot of rust

8. tshikoRegistered Member

Joined:
Apr 13, 2001
Posts:
26
0
Yeah Mechanics of Materials is a lot of fun,it was one of my favorite. I am a structural engineer and I agree with you that the effective length factor would be less than one but I would watch for bending in your roll bar.

Keep in mind that it is not a column, but it is more like a frame. What you have described is only the axial part of the problem, you need to estimate a bending moment in the bar when your truck hits the ground upside down, then you can find the stress using (M/S)

I don't know what you call triangulation, the method seems too analytical and I suggest you draw a free-body diagram with the forces and the buckled shape of the bar. Looking at the buckled shape, the inflection point will help you guess-timate the effective factor (be creative).

I would guess that k=0.8 for the pillars, but I would keep it as one for the horizontal bar. Make sure you include all foreseeable loads on your FB diagram, and keep in mind that if the horizontal bar buckles all the others will give way at the knees and you become the road kill!

9. jeffer1/2 ton status

Joined:
Nov 6, 2000
Posts:
113
0
Location:
Hey, thanks for all of the extra input guys. I'm not sure that I'll go to the extreme of doing a free body diagram or doing a cad simulation of the roll cage, but I have a lot more ammunition when selecting/installing a cage. Steve, I'm sure that my buddy Neal will be delighted that his 'Go big and go home' phrase worked for a good summary.

Joined:
Apr 22, 2001
Posts:
654