silex
|
| posted on 21/9/04 at 07:03 AM |
|
|
Chassis tube thickness
Gents,
Having seen a number of discussions about chassis stiffness, etc. I intend to attempt a complete a complete analysis of the book chassis strength and
tortional stiffness.
This will hopefully not only include the bear chassis which has been done before, but the aluminium body panels as well (in rivited and bonded forms).
This will take a bit of time but please be patient.
I need a little help though. I want to know the thickness of your chassis tubes - Please dont give me a guage size, measure it if you can. Most stock
is normally 2, 2.5 or 3mm. Also, if possible the thickness of aluminium sheets used for the body panels as well.
Thanks in advance,
Silex.
Murphy's 2 laws
1. If it can go wrong it will
2. In case of emergency - refer to rule 1.
|
|
|
|
|
JoelP
|
| posted on 21/9/04 at 07:12 AM |
|
|
16g is the commonest size used i think, this is 1.6mm. Some people use 2mm. i havent heard of anyone using more than this.
i think the gauges are just fractions of an inch, ie 16g is 1/16 thick, or 25/16mm, which is 1.6.
|
|
|
James
|
| posted on 21/9/04 at 07:29 AM |
|
|
1.6mm here too.
HTH,
James
|
|
|
David Jenkins
|
| posted on 21/9/04 at 07:35 AM |
|
|
I'm with James - mine was 1.6mm, but I believe that 1.5mm is getting to be the standard - varies with the supplier. This isn't just for
our benefit - this tube is used by all sorts of light industry.
Wire gauges are hard to convert without a reference book - sometimes it's the number of thicknesses that make up an inch, other times (like the
shotgun) it's the number of round lead balls that make up a pound! For example, the 12-bore will take a ball that's 1/12 pound of lead.
There's Standard Wire Gauge (SWG) in the UK, American Wire Gauge, different wire gauges according to the material (steel, brass, etc.). Thank
God for metric!
rgds,
David
|
|
|
stressy
|
| posted on 21/9/04 at 07:59 AM |
|
|
Silex,
What method are you using for analysis, manual spreadsheets or FEA? If fea what solver? How are you intending to integrate the panels?
just asking to be nosy really
Cheers
WHO DARES SPINS
|
|
|
MikeRJ
|
| posted on 21/9/04 at 10:03 AM |
|
|
quote:
Originally posted by JoelP
16g is the commonest size used i think, this is 1.6mm. Some people use 2mm. i havent heard of anyone using more than this.
i think the gauges are just fractions of an inch, ie 16g is 1/16 thick, or 25/16mm, which is 1.6.
Unfortunately not as sensible as that!
code:
Gauge Imperial sheet/mm Metric sheet/mm
10 3.25 3.0
12 2.64 2.5
14 2.03 2.0
16 1.63 1.5
18 1.22 1.2
20 0.91 0.9
22 0.71 0.7
24 0.56 0.6
26 0.46 0.5
[Edited on 21/9/04 by MikeRJ]
|
|
|
David Jenkins
|
| posted on 21/9/04 at 10:37 AM |
|
|
Coming back to the original theme of this post,  - there is another aspect that needs to be considered - the thickness and grade of aluminium used
to panel the chassis.
I used 1.6mm "half-hard" (but it felt like "bloody hard and unbendable" ) for the sides, and a fairly soft 1.2mm for the rear
panel. My tunnel sides and footwell fronts are panelled in 1.6mm steel, just to confuse the calculations!
cheers,
David
|
|
|
silex
|
| posted on 21/9/04 at 11:29 AM |
|
|
Stressy - I have started to draw up the chassis, etc. within Pro Engineer. This will then get full FEA step by step within Pro Mechanica.
All - thanks for the info. I will base the chassis on 1.6mm thick 25x25mm steel. I will set the steel grade as S275 as this is usually whats given
unless specified when purchasing. I will base the aluminium on a similar standard grade to get as close to the "As Book Built"
performance.
Also, there has only been one answer to the thickness of aluminium - 1.6mm half Hard and 1.2mm Soft. Is it safe for me to assume that everyone has
pretty much the same thickness of 1.2mm soft ?
After I have a result for the "Book" build I hope to go on and hopefully suggest some way of improving the design without a weight
disadvantage - probably in a number of steps so that the builder can choose the mods depending on what the car is used for.
Cheers,
Silex.
[Edited on 21/9/04 by silex]
Murphy's 2 laws
1. If it can go wrong it will
2. In case of emergency - refer to rule 1.
|
|
|
blueshift
|
| posted on 21/9/04 at 01:20 PM |
|
|
I assume you have seen cymtriks' work already?
|
|
|
silex
|
| posted on 21/9/04 at 04:36 PM |
|
|
I have seen at least some of cymtriks work. His analysis of the frame and then the improved tunnel and a couple of other items. He may have gone
further than this, but if so I have not seen it. If that is all the work carried out so far, I am sure he also understands that its not quite the full
and true picture as yet.
I am not dismissing his work, in fact I will be looking closely to see if my results tie up with his before progressing.
But, as far as I am aware no one has ever run the analysis on a completed tub i.e. steel frame with aluminium panels rivited in place,etc. Until this
is done, the actual strength of the book chassis is an unknown.
[Edited on 21/9/04 by silex]
Murphy's 2 laws
1. If it can go wrong it will
2. In case of emergency - refer to rule 1.
|
|
|
stressy
|
| posted on 21/9/04 at 08:18 PM |
|
|
Chances are cymtriks made the assumption that the normal fixing method for ally panels to the chassis (antivibration mastic and pop rivits) combined
with a generally low grade material results in the skins being not alot more than windbreaks.
In reality this form of construction will have fairly high performance degredation and so working with a conservative "they do nothing"
anaysis is always a good starting point.
From the point of view of correlation, if it helps, i got values within around 5% (from memory) of cymtriks when i did a quick analysis of the book
chassis, that was a good 6 years ago now, how time flys....
WHO DARES SPINS
|
|
|
craig1410
|
| posted on 21/9/04 at 09:01 PM |
|
|
Hi,
I used 2mm thick steel for my 1" and 0.75" square chassis tubes. This was more to do with what I had available rather than what I actually
wanted. However, I could have got 1.5mm at the time but didn't want to go any lower than 1.6mm. The 2mm steel is easier for the amateur (that
would be me!) to weld as you can use more power without blowing it apart.
I also used 1.2mm ally for the panels. the alloy is 5251-H22 which I think is "half hard". It's quite a nice compromise between hard
and soft I think as it is difficult to bend but holds a bend nicely once bent. It is also hard enough to resist dents and should be durable.
Do any of you FEA guys know of a freeware/shareware/trial FEA package that I could play with. I'm keen to mess about with structural designs to
get a better understanding of good and bad designs but don't want to spend hundreds of pounds for a commercial package.
Cheers,
Craig.
[Edited on 21/9/2004 by craig1410]
|
|
|
limy john
|
| posted on 22/9/04 at 02:01 AM |
|
|
material thickness
most frames were originally built with 16 swg, that's .064''
1.5mm is now the accepted replacement or 2mm
interestingly enough, rectangular section welded tube when purchased in 2mm wall thickness has a larger radius to the outside corners.
i considered this when the s.v.a. were getting funny about any radii in the driver area, the weight is equivalent to a whole chassis made of 22mmx.5mm
wall.
now i was in a position to have two chassis at the same time and the 25% extra weight was noticable.
welding the thicker chassis was also different and the ends of all the tubes had to be bevelled to propmote penetration.
aluminium sheet, if you had n. s.4. half hard, you should aneal whith a torch before bending as it might just drop in two at the bending point, to do
this with an oxy set, light the torch with no oxygen and allow the black smoke tocover the area to be bent with soot then turn onthe oxygen and heat
the panel untill the soot burns away, then you can bend it safely.
this material is far more resistant to dents and the trapaning effect of loose pop rivets, be carefull which rivets you use though as accelerated
corrosion may break out
|
|
|
stressy
|
| posted on 22/9/04 at 06:44 AM |
|
|
Craig,
If your trying to get an understanding of spaceframe structures it would be well worth an our in the local library. Have a look in applied mathematic
and basic mechanical / structural engineering books. In here you will find some information about frames, and how to calculate the forces in each
member. From that you can calculate the change in length of each member and hence the deform stiffness of the frame. This is not too complex. You will
also see the limitations of frames with regards to no bending.
Then when you pick u an basic FEA package you can build a frame as per the one youve calculated (2D) and compare. This should give you some confidence
that you are doing things corectly, afetrall fe will do essentially the same calculations just much quicker than we can.
I will see if i can lay my hands ona demo or shareware for you later.
Cheers
C
WHO DARES SPINS
|
|
|
silex
|
| posted on 22/9/04 at 06:48 AM |
|
|
Thank you all for the input so far.
Cheers.
Murphy's 2 laws
1. If it can go wrong it will
2. In case of emergency - refer to rule 1.
|
|
|
jcduroc
|
| posted on 22/9/04 at 02:00 PM |
|
|
quote:
Originally posted by JoelP
i think the gauges are just fractions of an inch, ie 16g is 1/16 thick, or 25/16mm, which is 1.6.
nOOO.
Gauge 16 is
1.587 mm in B.G. (Birmingham Gauge),
1.651 mm in B.M.G. (Birmingham Wire Gauge),
1.63 in I.S.W.G. (Imperial Standard Wire Gauge) which is, AFAIK, what we refer to,
1.519 in U.S.G. (Revised United States Gauge).
Similarly 18 is
1.257 mm in B.G. (Birmingham Gauge),
1.245 mm in B.M.G. (Birmingham Wire Gauge),
1.22 in I.S.W.G. (Imperial Standard Wire Gauge) which is, AFAIK, what we refer to,
1.214 in U.S.G. (Revised United States Gauge).
In Portugal (and most EU countries AFAIK) we use square 25mm x 1.5 mm wall and 20x1.5 instead of 1" 16 ISWG and 3/4" 16 ISWG respectively.
JCM
|
|
|
kiwirex
|
| posted on 23/9/04 at 08:29 AM |
|
|
The other factor that might be worth giving some thought to is tube size.
That is if you go up to 30mm square, instead of 25mm but stay at the same wall thickness.
I thought that you'd get more additional strength for the additional weight with bigger tube rather than thicker wall.
Could be wrong.
- Greg H
|
|
|
cymtriks
|
| posted on 2/10/04 at 09:05 PM |
|
|
Silex and Stressy,
I assumed that the aluminium panels do nothing. This is, as pointed out, a fair assumption to make for the locost.
I have written up my analysis and posted it on the files section of http://locost7.info/ This file covers more than just Locost chassis, ladder frames
are covered to.
I can e-mail you a Nastran .dat file of my analysis if you'd like.
It's nice to know that someone else agrees very closely with my numbers. A couple of guys at work did their own analysis without them knowing
anything about what I'd done. They got very close to my numbers as well.
What do you two do by the way. I work at RR.
|
|
|
TTK77
|
| posted on 4/10/04 at 01:06 PM |
|
|
Is 2mm thickness better for racing chassis than 1,5mm? 1,5 would lighter but for racing use 2,0mm might be more reliable..
[Edited on 4/10/04 by TTK77]
|
|
|
