craig1410
|
| posted on 9/7/03 at 10:29 PM |
|
|
Cymtriks,
Ah, now I know what you are talking about. (You have to excuse my slowness of uptake...)
You are concerned that the ends of the de-dion tube may bend and thus collapse due to excessive bending loads introduced by the "ears" of
the axle from the leverage of the road on the wheel. (At least I hope I have understood you correctly...)
I'm not sure how you convert bending torque into pressure though. I'm not a mech engineer, only a mere elec engineer, could you please
explain the maths if you get a moment?
Cheers,
Craig.
|
|
|
|
|
blueshift
|
| posted on 10/7/03 at 12:08 AM |
|
|
cymtriks, does your model assume that the bar is loaded at one point, on an end face as it were?
I wonder if it makes a difference that the hub carrier attaches to the dedion bar by two ears spaced some distance apart (around 30mm I think on the
plans I have).
Intuitively, again, to me it seems that the further apart those ears are the less.. whatever. twist? on the bar. I'm not sure what the right
word would be, torque, load in a specific area, or what. but it would be more spread out over length, if you see what I mean.
|
|
|
Alan B
|
| posted on 10/7/03 at 01:11 AM |
|
|
Guys,
Let's put this in perspective...
MK have made many cars with and without DeDion rears.....
Caterham also...
Myself and Rorty between us have actually built in real life many cars and raced many of them...
I am aware of no failures of a serious design related nature at all.......
With all respect to Mr Cymtriks when he shows us some real life examples of his work I'll take his warnings more seriously than the combined
experience of the above.....
I'll state again...I do not dispute the figures only the assumptions and parameters on which they are based....especially when they are in
conflict with real life experience.....
I don't say this as an old guy knocking new technology...I have a mech eng. degree and have used FEA etc. ......but, in the real world an ounce
of practice is worth a ton of theory....I know this because I've done both....
Again, no offence intended to Cymtriks....but as an "old guy" I'll always go with what is known to work....
<flameproof suit on.......>...
|
|
|
blueshift
|
| posted on 10/7/03 at 09:28 AM |
|
|
So alan, what size / thickness tube do you think one could get away with for locost dedion?
we want to stick 200+lbft of torque through our diff if that makes any difference.
|
|
|
Alan B
|
| posted on 10/7/03 at 12:21 PM |
|
|
Personally, I'd go with 3" 16g...as Steve did......if you can't fit 3" in then I would consider thicker
wall............Cymtriks list is a useful guide of their relative strengths in a bending application..the lower the stress figure the better.....
As I said earlier I'm not disputing the numbers....I'm just saying don't ignore what is already proven to work..
|
|
|
cymtriks
|
| posted on 10/7/03 at 08:15 PM |
|
|
Where did it all go wrong!
Guys,
I know my results indicate a tube built on the solid end of the scale but I'm genuinely interested in where the figures went wrong.
The tube stresses are right for the load given. The results I posted above were from basic text book beam equations. However I have made an f.e. model
today and it gave virtually identical stresses. What the f.e. does show is that the maximum stress is distributed over a narrow strip going from the
loaded end to where the lateral restraint is applied. My model resembles a Caterham design so this is at the centre of the tube. I hope to be able to
post this as a picture soon.
The wrong part of the analysis that leads to big tubes may be sought out as follows-
11.5 inches distance from the road to the beam. This an assumed value but should be close for most applications.
3g lateral. This is a measured value but was measured inside a car. What the suspension sees may be different.
385lbs corner weight. This is an estimated value but should be close enough. Using the corner weight may not be appropriate for lateral forces.
plus 50% for luck. A complete guess but this is lower than typical safety factors. The lowness is my attempt to reduce the tank like answers.
50000psi yield strength for basic steel. This figure is published in plenty of texts but so are a whole range of yeild strengths from 45000 to 220000
for different types of steel.
If anyone has any thoughts on the above then this might help indicate the actual load on a suspension. With that my table of stresses could be used
as a look up table for design; i.e. estimate the weight on the rear axle and look up the sizes of tube suitable in standard tube and for chrome
moly.
No more guesswork and no more overweight designs just to be safe.
A final thought. Most of the experince indicates tubes with stresses about the same or lower than my suggestion.
|
|
|
blueshift
|
| posted on 10/7/03 at 09:01 PM |
|
|
I'm curious to know what model of a dedion bar you're using. I don't know how a caterham bar compares with the ones I have plans
for.
I am CADing up my bar at the moment and could send you the model / pictures if they'd be of use, just to see if I'm on the same page as
you.
|
|
|
craig1410
|
| posted on 10/7/03 at 09:20 PM |
|
|
Yeah I was thinking the same thing.
Does anyone have a picture of the Caterham design so we can see roughly how it goes together?
Cheers,
Craig.
|
|
|
timf
|
| posted on 11/7/03 at 09:46 AM |
|
|
quote:
Originally posted by Syd Bridgeif the corty and escort diffs worked in general use, and racing, then this is the size tube to use?
Just a thought, anyway.
Cheers,
Syd.
the rally escorts ran baby atlas diffs (shortened capri style diffs) and they suffered from tube breakage so the standard approch was to make a dural
brace that was held by the rear diff plate and large 'U' clamps down the axle tubes
or for the escort axles a bar was bent over the top of the housing and welded in place. so in effect the thickness for used wasn't quite up to
spec.
|
|
|
timf
|
| posted on 11/7/03 at 11:35 AM |
|
|
true but i was just pointing out that the general racing comment could be mistaken.
and the way ford recomended the upgrade.
BTW the capri 'BRSCC' cars had the same mods as the rally kit but then they were also running cossie developed 3.4 v6's
|
|
|
craig1410
|
| posted on 11/7/03 at 12:19 PM |
|
|
Hang on though, on a de-dion setup wouldn't the bending force be localised around the point where the "ears" attach to the outboard
ends of the de-dion tube? If I understand you correctly here Tim you are meaning that the escort axle used to break near the diff? If so then I think
this is different altogether. A de-dion system wouldn't place significant loads in the middle of the tube IMHO. Would it? Hmm?
Also, isn't the material used in a typical live axle inferior to CDS tubing?
Cheers,
Craig.
|
|
|
timf
|
| posted on 11/7/03 at 01:00 PM |
|
|
it didn't break as such the axle tubes used to bend so the half shafts didn't run concentric.
and the material used was very inferior to cds tubing. the loads on the de-deon would be at the greatest at the hub location points so very good
welding would be required.
|
|
|
blueshift
|
| posted on 11/7/03 at 02:33 PM |
|
|
De Dion bar that Craig might recognise at
http://www.leetfighter.com/gallery/view_photo.php?set_albumName=solidworks&id=dedion
No measurements on that picture, but you can see where the trailing arms link to the ears, so that's from your chassis.. and the sierra hub
carriers bolt on the outside, so that dimension is whatever it is.
the bar is 2" OD 2mm, carrier plates 1/2", carrier tubes 4" OD 2mm, everything else 3mm plate, all mild steel
don't read too much into the relative positions in that picture, they were just guess-fits.
No panhard rod / watts linkage on there yet either.
That the kind of thing you're modelling, cymtriks?
|
|
|
Alan B
|
| posted on 11/7/03 at 03:06 PM |
|
|
Blueshift, I just checked out your other pics....good stuff..
Picture 12 is erm..?..........interesting...
|
|
|
blueshift
|
| posted on 11/7/03 at 03:38 PM |
|
|
A slight photoshop job on a picture of my co-builder mackie there. Bit of an in-joke
|
|
|
cymtriks
|
| posted on 11/7/03 at 04:02 PM |
|
|
The tube I've modeled is basically the Caterham design. I hope to post some actual pictures of this analysis soon. I need to get a picture off
my works computer onto a disk and then take it home to post.
I modeled the Caterham as it has a simple straight tube with a central A frame and two upper trailing links making it a very "clean"
design i.e. easy to model in a restricted time. It is also proven and I have actually measured one as having 2.5 inch OD.
In order to simplify things the ears are represented by a force distributor which transfers the lateral force from the tyre contact patch directly
onto the tube ends. This saves having to include the tyres, wheels, hubs, etc... However it does mean that the ears are not included, just the tube
itself.
The results for lateral cornering forces show that the compressive stresses are concentrated in a long thin strip along the side of the tube nearest
the contact patch of the tyre. This is exactly what simple beam theory predicts but is interesting to see it visually.
This may be why the very high stress that I calculated in earlier posts does not seem to be a problem in practice. It may be high but it is acting on
a small area so its effect is not as much as you'd think. I know from real compressive buckling tests that I have done on Aerospace parts that
sometimes the maximum stress, if acting on a small area, can be less important than a smaller stress applied over a large area.
Another reason why the high stresses may not be problem is that I have assumed an over the top load in the first place as a lot of assumptions are
included in its calculation. If anyone knows what loads actually apply or if anyone knows what the specification of off the shelf tube is likley to be
then let me know.
|
|
|
blueshift
|
| posted on 11/7/03 at 04:21 PM |
|
|
An A-frame? sounds like the caterham tube is quite different to the plans I have. Did you see the picture I posted?
|
|
|
craig1410
|
| posted on 11/7/03 at 04:45 PM |
|
|
Timf,
Regarding the "very good welding" you mentioned. I agree 100% and that's why I have only tack welded my axle bits in position so far
and intend to get them welded professionally.
That raises a slightly off topic but relevant question: What type of welding is strongest for an application such as this? Tig or Mig (or other?) I
was guessing Tig but it might be easier to find someone who is a "good" mig welder which I guess will be yield a stronger weld than an
"average" Tig welder?
Blueshift, yes that de-dion tube does ring a bell... 
I think from what has been said I would build a de-dion tube using 2.5" 2mm wall if I did it all again but I'm certainly not worried too
much about my existing axle design. As Syd said, I'm not planning to Rally it or jump off ramps so it should cope fine.
Cheers,
Craig.
|
|
|
