substituting aluminum for steel - is there a rough guideline?

[josephmartins]

I'm building an automobile rotisserie capable of holding up to 5000 lbs. I'd like to use aluminum rather than steel. The steel verson weighs nearly 500 lbs.

This is what the finished product looks like:

http://www.harwoodperformance.bizlan...torial_20k.htm

I'm not sure how to go about selecting the right aluminum for the job. I've been told 6061-T6, or 6063-T52 or something similar would work best. That's great, but what about the dimensions?

For example, let's say I have a vertical support post made of 3"x3" square 3/16" walled steel tubing. If I want to use 6061-T6 square aluminum tubing instead, what should the size and thickness be? Is there some sort of rough conversion I can do to ballpark the correct sizing? Is there an online resource where I can learn what I need to do the calculations myself?

Any help/insight would be appreciated.

Kind regards,

Joe Martins

[JMcDonald]

1) Avoid making anything structural out of welded aluminum. Welding puts it back into its annealed state, which is VASTLY weaker. If you can, at least assemble everything with large bolts.

2) 6061 Aluminum is about as strong is typical Hot Rolled mild steel (its actually a big stronger in terms of bending). 6063, however, is only about half as strong, and is mainly used when dimensional accuracy and appearance is of greater concern than strength (like if you were making drawer guides, and you wanted it smooth and strait). 6063 is produced with a machined finish that is smooth, with sharp corners.

3) While aluminum can easily be as strong as steel, its weak point is its fatigue life. If you made your item out of steel, it can flex a bit when loaded, and for the most part be fine after you remove that load (to a degree). However, due to Aluminum's brittleness, everytime it flexes, some of its bonds are broken without reforming (as opposed to the steel's greater ability to "stretch" its bonds). Basically, everytime its flexed, it gets weaker. For example, to break a paperclip by bending it back and forth my take 20 bends. For an aluminum one with comparible strength, however, it might break in the first one or two bends.

In a nutshell, to build something out of aluminum means more thought must be put into keeping the assembly as rigid as possible, to avoid this fatigue. Be generous with triangles and you would be fine.

You could use the same dimensions of aluminum (maybe step up to .25" thick instead of .1875", in more critical places), but use bolts wherever possible, instead of welding.

Like here, for example. Instead of welding the top piece to the bottom. Pass 3 or 4 bolts (at least 1/2", grade 8, since the stress would be all tensile) vertically through both tubes:


Also, on the supports for the main column, would would more likely want to have them go all the way from the ends of the legs to the top of the column (or at least as far as possible, but at the same time try to stay close to a 45* angle), instead of just part way down each axis like here:


However, even those would probably be ok how they are. But again, instead of welding them, attach them to the legs and colums with a flat piece of probably .25" thick 6061 Aluminum bar that extends from the sides of the angled supports out, to straddle the side of the main column so bolts could be passed through them.

The largest problem by far, however, is the lack of support on the rear of the column. You would need to extend the base rail out past the rear of the column. A foot extra would probably do. Then, just add another support like on the sides.

Here, Ill make a few images real quick.

*edit*

Here is the view from the front. It is not really to scale, especially the bolts heh. Note that the rear support is mounted higher on the column (as close to the top as reasonably possible) than the side supports, because it will have the greatest impact on preventing forward flexing when the load is applied. Also note how the spindle part passes THROUGH the column, rather than over the top and welded. You could pass it through the column, then weld it, and it would still have the same strength of the surrounding metal.


Here is a view from the top. The same applies, but you can also see how the rear support would be roughly the same as the side ones, just longer.

[Switcher]

Why not stay with steel?

It looks like it's a good design that you could break down into 2 parts (250#) each, & it's on casters.

http://www.harwoodperformance.bizlan...erieFinal1.jpg

.

[josephmartins]

John, thank you very much. This is valuable insight. I originally thought I'd only need to change the dimensions of the components (i.e. buy thicker parts) in order to use aluminum in the design. But it appears I'll need to make substantial changes to Mark's original design if I want to construct it in aluminum. I had no idea welding did that to aluminum...why the heck are all these bike companies and such building aluminum frames? Or are they retreating the aluminum to give it back some of the strength lost during welding?

I understand the reinforcing suggestions, though I'm now wondering how I would attach angular braces to legs and uprights using only bolts with no welding at all. I'll have to think about that some more because I'm having a hard time visualizing your suggestion with the aluminum bar...unless you mean that I should pass the bar through an opening in the angled brace and anchor it downward.

I'll be out for the next several hours, but I will check the thread when I return this evening. Thank you very much for your insight. I appreciate it.

[JMcDonald]

And I would also more likely use steel. Still, however, I would incorporate my above additions into the design (primarily bolted construction, rather than welded; added rear column support, etc).


A friend and I have been tossing around the idea of a semi-portable automotive jacking system, and some of my designs have been similar to this, heh.

[josephmartins]

This forum is like an addiction. I was just about to open my door to head out when I heard the "ding" of the notification in my inbox.

I like the original design, but it just seems like manipulating those supports even on wheels, would be pretty tough if it is built using steel. The chassis already weighs several hundred pounds stripped. I need to be able to easily roll them over to a vehicle, stow them away, and even push/pull them when I want to load the vehicle onto a flatbed to head to the sandblaster or painter.

I suppose I could just deal with it. I was just hoping to drop some of the weight using aluminum. What if I used smaller steel tubing (same thickness, but say 2.5 x 2.5 rather than 3 x 3) and some triangular steel plate reinforcements instead? I'll have to run some numbers when I get back later.

Thanks for the input and the drawing suggestions.

[DSL PWR]

If you want to make it easier to move, put larger casters on it. That will also be cheaper than aluminum, and easier than redesigning it.

[JMcDonald]

It is indeed addicting!

I can't say much now, but I definately understand your concern about the weight, especially if you will be working alot by yourself.

I will draw up some more designs later, but I gotta run now! However I assure you it would be quite simple to get that whole design to work like the other one, using little to no welding.

BTW, the unit would almost definately be stronger by using larger diameter tubing, but with thinner walls. You might not lose much weight after the increase in size, but you would gain alot in stiffness. As long as the walls were not thin enough to buckle under the weight (a drastic and quick failure). Id say you could use .12" walled tubing (roughly 1/8") instead of the .1875" tubing used in the example, and still get roughly the same strength.

*edit*

Last quick point:
In a nutshell, you could probably use the same dimensions in aluminum as he did with steel, but add the triangular supports on the side and rear of the column, and have a unit that is about 1/2-2/3 the weight, Id guess (so, 250-350lbs, rather than 500).

[josephmartins]

Thank you all for providing insight.

DSL,

I'm definitely going to mount the unit on casters. Outrigger-style 10" pneumatic tires (much easier to transport than metal casters and solid rubber casters). This is another reason to save weight. The 10" tires I have are rated at about 500lbs a piece. Minimizing the rotisserie weight will maximize the available capacity.

John,

The reason I'm so concerned about weight is because this is just the first phase of the design. Once this is built I'm going to construct a special jig which will attach to the rotisserie. The jig will allow me to fix and align chassis components in preparation for welding. In other words, I'll be able to construct an entire unibody undercarriage on the rotisserie. It's never been done before. But, if it works, it'll make my hobby of reconstructing and restoring cars a whole lot easier. The first candidate is a 1968 Mustang fastback that I'm working on.

So, rather than end up with a 700+ pound monstrosity, I'd like to pare weight wherever I can by substituting aluminum wherever possible.

By the way, I had a metal supply company run some numbers for me this afternoon. They estimated the steel alone would weigh 486 lbs, and the aluminum (same dimensions) would weigh about 176 lbs. Over 300lbs shaved off...even with your suggestions I bet the design wouldn't exceed 210lbs.

I could go steel...but I want to do this in aluminum. Geek and gearhead that I am, I enjoy a good challenge.

[JMcDonald]

Sounds good!

Oh, I just realized that with the rear column support, you wouldnt need that triangular support in the front anyways. Although it could be done using the same method that the side and rear supports are attached.

But basically you use all tubing that has the same dimensions. Say, 3" square Aluminum tubing. For the angled supports, you would line them up in the same place as you would with a steel structure. Except, instead of welding them to the column and base rail, you could use something like .25" thick x 3" wide Aluminum bar, to make "straps" to pass bolts through. Each "strap" bar would be about 6-12" long, and it would be oriented parallel to the angled support, but extending down over the sides of the base rails. Then you would pass bolts through the straps, with some going through the base rails and some going through the column support. Now you have a rigid mechanical connection with no welding. Even if you used only one bolt in each end of the strap, the supports would still do their job completely adequately, as the only forces that would act on the supports from the column trying to deflect would be tensile or compressional.

Does my first pic make more sense now?

[neilw20]

Originally Posted by JMcDonald It is indeed addicting! I can't say much now, but I definately understand your concern about the weight, especially if you will be working alot by yourself. I will draw up some more designs later, but I gotta run now! However I assure you it would be quite simple to get that whole design to work like the other one, using little to no welding. BTW, the unit would almost definately be stronger by using larger diameter tubing, but with thinner walls. You might not lose much weight after the increase in size, but you would gain alot in stiffness. As long as the walls were not thin enough to buckle under the weight (a drastic and quick failure). Id say you could use .12" walled tubing (roughly 1/8") instead of the .1875" tubing used in the example, and still get roughly the same strength. *edit* Last quick point: In a nutshell, you could probably use the same dimensions in aluminum as he did with steel, but add the triangular supports on the side and rear of the column, and have a unit that is about 1/2-2/3 the weight, Id guess (so, 250-350lbs, rather than 500).

Aluminum is about three times "springier" than steel.
Get a piece of length of steel, and a length of aluminum the same size.
Hang the same weight of each and see how far it deflects.

With your existing frame, see what is almost too springy now.
Aluminum parts in shear will be fine, but things that have bending loads will need attention. Like has been said. More triangles.

Youngs modulus? Learn about it here...
http://en.wikipedia.org/wiki/Youngs_modulus