smart51
|
| posted on 3/6/11 at 01:02 PM |
|
|
How do you calculate torsion strength?
I want to make a U shaped profile as stiff in torsion as a flat sided oval tube by making the wall thickness bigger. How do I calculate the
stiffness, so I can work out how thick it has to be?
[Edited on 3-6-2011 by smart51]
|
|
|
|
|
britishtrident
|
| posted on 3/6/11 at 01:48 PM |
|
|
Without using CAE either.
Parallel axis theorem and principle of super position.
Or just look it up in Roarks Formulas for Stress & Strain.
Be aware that in torsion prismatic bars with cross section shapes such U C and Z usually give nasty out of axis deflections
|
|
|
Fred W B
|
| posted on 3/6/11 at 02:08 PM |
|
|
Do you mean a "D" shaped tube, or a "U" shaped pressing. It can't be a tube with one side open?
Cheers
Fred W B
You can do it quickly. You can do it cheap. You can do it right. – Pick any two.
|
|
|
smart51
|
| posted on 3/6/11 at 02:32 PM |
|
|
Fred, I've corrected the original post. I want to compare a U shaped profile with a flat sided oval tube of the same overall width and height.
|
|
|
Liam
|
| posted on 3/6/11 at 02:33 PM |
|
|
quote:
Originally posted by britishtrident
Without using CAE either.
Parallel axis theorem and principle of super position.
Isn't that methods of obtaining second moment of funny shapes for bending stiffness calcs?
I thought torsional stiffness was more scary voodoo sort of levels of nasty (for funny shapes) usually meaning FEA. Unless you're right and
someone's already done it and you can look it up!
Off the top of my head though - torsional rigidity of a U section compared to a full tube - alot alot alot alot less!
|
|
|
mcerd1
|
| posted on 3/6/11 at 02:46 PM |
|
|
open sections are not good in torsion (channels, angle, 'H' sections) - tubes are much, much stonger
but if you want to do the math then have a look at this: linky
(ISBN 1870004442 - its the one I use at work)
[Edited on 3/6/2011 by mcerd1]
[Edited on 3/6/2011 by mcerd1]
-
|
|
|
smart51
|
| posted on 3/6/11 at 03:09 PM |
|
|
Hmm. That sounds a bit harder than I was anticipating.
The piece I'm making is an underfloor chassis brace for each end of a transmission tunnel of a coupe. The two floor pans slide relative to each
other and twist (one goes up at the rear and the other up at the front). The braces stops them from moving. The load is partly tension and partly
diagonal shear. As open sections are worst in torsion, I thought I'd do that calculation. perhaps it's not the best idea.
|
|
|
Liam
|
| posted on 3/6/11 at 03:22 PM |
|
|
Any reason it has to be an open section? (I guess so or you wouldn't be asking  ) Maybe that's not the right choice of member for the
application?
I said member - huh huh
[Edited on 3/6/11 by Liam]
|
|
|
mcerd1
|
| posted on 3/6/11 at 03:24 PM |
|
|
the book in my link above will take you throught the basics - its from structural steel design, but it should still do what you need it to
(its designed to be calculated by hand and its got worked examples in it too   )
-
|
|
|
smart51
|
| posted on 3/6/11 at 04:02 PM |
|
|
quote:
Originally posted by Liam
Any reason it has to be an open section?
Not really. The two ends where it bolts to the car have to be 11mm proud of the rest of the beam. I was thinking of cutting a RHS tube so it would
be open for most of its length. If the tubes could be cut by water jet then it would be the least amount of work to make the required shape. The
current design needs fishmouth cuts and lots of welding. Fine for a one off but the owners club want me to make several.
|
|
|
rachaeljf
|
| posted on 3/6/11 at 05:50 PM |
|
|
Could you show us a sketch? It sounds like you just need a flat plate shear brace at each end of the tunnel, but I would need to see why you need such
awkward connections.
Cheers R
|
|
|
smart51
|
| posted on 3/6/11 at 09:36 PM |
|
|
This is a quick sketch of the car, with the top hat shaped things the sub frame. The blue is the shape of the brace needed to use the subframe
mounting bolts. The one I have made has a penny washer welded to a short length of round tube at each end to bolt to the car and a piece of flat oval
with a fish mouth at each end welded to the two short tubes. Making one is fun. Making several is a PITA.
|
|
|
smart51
|
| posted on 3/6/11 at 09:37 PM |
|
|
The car thing on all the pictures is really annoying. Is there a reason for it?
|
|
|
DIY Si
|
| posted on 4/6/11 at 05:03 PM |
|
|
Could you get a bit of round tube bent to shape, then weld a circle of steel in at either end with a captive bolt welded in from behind it? Or is this
more something to knock up in the garage when you have time?
“Let your plans be dark and as impenetratable as night, and when you move, fall like a thunderbolt.”
Sun Tzu, The Art of War
My new blog: http://spritecave.blogspot.co.uk/
|
|
|
smart51
|
| posted on 4/6/11 at 06:01 PM |
|
|
The sub frame is held on with bolts and the brace just piggybacks onto the bolts, so no welded studs. There's not much gap between the bolt and
the subframe either, so have to be fairly clever with the design. The idea is to make them in the shed, though perhaps with some water jet cutting to
save work.
|
|
|
rachaeljf
|
| posted on 4/6/11 at 06:10 PM |
|
|
If you have only one bolt at each end there is a lot of "prying" (hate the Americanism) on the bolt to generate the torsional resistance
you are after.
What would be better is to make the whole tunnel act as a torsion member by shear bracing it at each end. Imagine a cross country skier doing that
funny sliding walk they do, the skis are doing what your two floorpans are doing, allowing the chassis to twist. Shear braces to stop that fore and
aft shear motion will give you the desired overall torsional resistance.
A top hat shape shear brace could be bent up from say, 2 mm plate. It needs two bolts at each end, ideally 100 - 150 mm apart, aligned along the
car's long axis. The width of the brace needs to be about 150 to 200 mm. You can get away with one bolt if you can jam the vertical sides of the
top hat against the sides of the subframe top hats as you have drawn them.
However, if you want to go with your torsion brace, a length of CHS or RHS butt welded to a short length of angle iron at each end would do the job.
This would be best done again using two bolts in line along the long axis of the car at each end of the brace; this makes the two bolts work in
tension only. A single bolt at each end will suffer prying effects as I mentioned before.
Cheers R
|
|
|
smart51
|
| posted on 4/6/11 at 06:49 PM |
|
|
Thanks Rachel. You've got the problem spot on. Sadly, there's only one bolt hole and I don't want to modify the car. The brace I
made works very well but I wouldn't want to make many that way. Your idea of angle iron is interesting.. wouldn't it flex, being un
triangulated?
|
|
|
rachaeljf
|
| posted on 4/6/11 at 08:52 PM |
|
|
I really need some dimensions to be sure what sizes to use, but I think it will be fine.
I suggest you use 50 x 30 x 3.0 RHS and 50 x 50 x 4 cold formed angle. I presume the bolt is M12, use a 40 SQ x 3 mm flat washer under the bolt head
too.
Assuming you have enough flat area around the subframe bolt, try and make the end angle brackets about 150 mm long. Make the end brackets triangular
in profile so you have full height where the RHS joins, and tapered down to say, 10 mm at the ends. You can also taper them in plan too (say 50 mm
down to 25 mm) if you really want to save weight.
Cheers R
|
|
|
smart51
|
| posted on 4/6/11 at 09:18 PM |
|
|
Thanks Rachael. That is probably the simplest way to do it.
|
|
|
