[grouch]

If the purpose of boxing a frame is to make it more rigid, torsionally or otherwise, why not weld the extra piece in from one inside corner to the diagonally opposite outside edge of the typical C channel frame? Wouldn't that provide more rigidity for the same amount of steel added?

It would make for more work in curved sections of a frame, but it looks to me like it would result in a very rigid structure.

[powerrodsmike]

When you box a channel type frame, it increases the torsional rigidity of the beams themselves, and reduces the tendancy to bend at the ends, (or sag in the middle).. The cross bracing adds to the torsional rigidity of the frame as a whole, but adds little to the resistance to bend.
Also, all of the pickup points for the suspension and motor can all benefit from being attached to a boxed frame member.

Being as how there is alot going on in the front of the frame, as well as the back, why not box the whole thing?

That's what I think , anyway.

We really need some REAL structural engineers to answer these kinds of posts.

I know there are a few on the board.

Mikey

[deuce_454]

the object of boxing is to increase the moment of inertia.. which is a geometric, calculatad constant for the beam section you are working on.. it determins the torsional rigidity allong with the modulus of elasticity of the material of the beam.. and it increases with the distance from the center of the beam to the material... so in laymans terms, the bigger the diameter of the tube, the more tortionally rigid it will be... if you were to weld the boxing plate from corner to corner instead of closing the c-chanell you would decrease the "diameter" of the tube and it would be less rigid.. it would be a smaller tube , with a "wing of material" not adding to the torsional rigidity, only weight ... leaving you with less stiffness and the same weight (in fact the weight would be a little higher as the distance from corner to corner is longer than from edge to edge)....

[grouch]

Ok, given a C channel of A, B, A' dimensions, typical boxing closes the open side, to yield A, B, A', B'.

If, instead of that 2nd B side, you put in a web from one corner, say the B-A corner to the upper outside, it would be sqrt(A^2 + B^2) wide. That triangular tube would be weaker than the ABA'B' tube? The shape doesn't figure into its strength?

[home brew]

I'm no engineer but I think that your new web piece (call it C) would be transferring loads from A to B just as it does when it is not boxed by B'. In the boxed situation the loads on A are shared by B and B'.

[grouch]

Yes, loading is shared among A, B, C. The bottom flange, A' is no stronger than before C was added.

Imagine clamping the frame in a vise at about mid-way along one side and then bouncing the opposite side up and down. Wouldn't a rectangular box shape distort more readily than a triangular one? Is this the type of flexing that boxing is aimed at?

Clamp it at one suspension mount point and bounce the frame at the diagonally opposite corner. Would the triangular cross-section resist that flexing more than a rectangular cross-section?

[BillyShope]

grouch, my CAD software calculates the polar inertia for me, so all I had to do was describe a "for instance." With a 4 X 2 "C" fully boxed (0.090 walls), the polar inertia is only about 10% greater than with the triangulation. So, if you have some need to triangulate a section, there's not a whole lotta difference. Note, however, that the fully boxed is the better choice. And, it's lighter.

[grouch]

Does that calculation take into consideration the flexing at corners or is it based upon an ideal, theoretical, unflexing shape? I would be surprised if that CAD software was doing anything more than plugging numbers into the formulas and spitting out the results. Those results don't cover everything to be considered.

I just don't think that polar moment of inertia is the whole picture. That only tells you how much the object resists. It doesn't tell what happens to the object if it is subjected to continuously varying torsion over a long period of time. It doesn't tell what happens to the steel along the edges where the planes meet after that boxed frame has been bounced down the highway a few years.

It looks to me like the greatest stress will occur at the corners. A box can flex at the corners at levels of stress far below the failure point. Isn't that what causes metal fatigue? Especially with sharp corners, isn't boxing setting the frame up to fail from little cracks, sometime down the road?

[BillyShope]

Your list of "what if's" are the sort which require "finite element analysis" and, now that I'm retired, I don't have access to the software with that capability. I modeled with sharp corners, which certainly would be avoided. But, I believe I can safely say that, after the FEA and after every effort is made to remove stress concentrations, a structural engineer would still go with the fully boxed configuration. But, again, if you're looking for space to neatly tuck away fuel lines, for instance, the triangulation would be quite adequate.

[powerrodsmike]

Originally Posted by grouch

Does that calculation take into consideration the flexing at corners or is it based upon an ideal, theoretical, unflexing shape? I would be surprised if that CAD software was doing anything more than plugging numbers into the formulas and spitting out the results. Those results don't cover everything to be considered.

I just don't think that polar moment of inertia is the whole picture. That only tells you how much the object resists. It doesn't tell what happens to the object if it is subjected to continuously varying torsion over a long period of time. It doesn't tell what happens to the steel along the edges where the planes meet after that boxed frame has been bounced down the highway a few years.

It looks to me like the greatest stress will occur at the corners. A box can flex at the corners at levels of stress far below the failure point. Isn't that what causes metal fatigue? Especially with sharp corners, isn't boxing setting the frame up to fail from little cracks, sometime down the road?

How long do you need it to last? I would venture that either design will outlast all of us.

Metal fatigue doesn't usually occur in structures that are never subjected to stresses that are less than 1/2 their yield point. I would think that the open leg of a channel would be subject to developing stress risers at the edge way before the same channel that was boxed. Even figuring in a crappy weld. Also, the opposite corner would be sharper of an angle, and would have less ability to resist development of stress risers than a 90 degree corner that you'd have with a fully boxed beam.

And making joints at an angled side of a beam would make a fabrication nightmare, Although there would be more of an interface area to spread the stresses out.



Do you have any examples of structures that use a triangular cross section type beam instead of a rectangle or square?

I'd think if it was stronger it would be in much greater use today.


I think many times metal fatigue occurs because it was crappy metal to begin with, and joint design and misapplication of the end use causes failure.
It's easy to say that metal fatigue was the cause of failure, when you've exceeded the design limits of the frame many times over by putting more weight, power and subjecting the frame to higher stresses than it was ever intended to resist..

BTW, I misunderstood your first post. I thought you were talking about the frame structure as a whole, not just the beam.

Mikey

[grouch]

Originally Posted by powerrodsmike

Do you have any examples of structures that use a triangular cross section type beam instead of a rectangle or square?

Only similar: The three tubes making up a typical antenna tower or the space station arms or frames for large (sports arena) domes.

While I was looking around trying to find an answer before opening this thread, I ran across something that looks promising. The trouble is that some of it looks like math and some of it looks like some kind of programming language and almost all of it flitters by right over my head:
http://documents.wolfram.com/applica...lAnalysis.html

[bet on black]

maybe what he is trying to say is overall flex in the frame, not just one rail of it. while i think we can all agree that a boxed rail would be stronger and less prone to flex than an open c-channel, the cross members may still flex and let the two channels no longer be parallel in side view.

it seems like there are two different situations being discussed here. i don't have any way to prove it, but a triangulated cross member would have to be less prone to flex as a simple square one would, right?

[chieftain]

This is a good thread for me right now. I removed the rusted old x-frame and left with open c-channels. The front section is all boxed in and middle to rear is open 2 x 5 c-channel and towards the back 2 x 3". I'm going to box mine in and like you i'm concerned about the extra weight. My solution along the 2 x 5 section anyway is to make a nice row of 3" or so holes like a swiss cheese frame. After it's boxed I was going to add in a new trans, rear shock crossmember and another crossmember like one for a ladder bar in front of my front leaf spring mounts and add in a 2x2" ? square tubing at a 45 degree from the ladder bar crossmember to the boxed in siderails.

[runn141]

Originally Posted by grouch

If the purpose of boxing a frame is to make it more rigid, torsionally or otherwise, why not weld the extra piece in from one inside corner to the diagonally opposite outside edge of the typical C channel frame? Wouldn't that provide more rigidity for the same amount of steel added?

It would make for more work in curved sections of a frame, but it looks to me like it would result in a very rigid structure.

i had an s10 that i welded the frame diagonally behind where the boxed part of the frame ends. if you look at my photos you will see where i totaled it and because i had done that i saved my motor and trans and trans crossmember including surviving with minor injuries. i think that if i had just boxed the frame it would have caved in more on impact. every s10 i build now gets that upgrade.

[deuce_454]

you wont save more than 5-10 pounds by swiss-cheesing the frame.. and im sure you will accumulate that weight in road crud in the frame prety quickly.. also water wont be able to run back out.. ive attached some pics of what i did on my malibu, it added about 20 pounds and made a world of difference