Student Dissertation on Ultima chassis stiffness improvements
an interesting read
Good amount of work in that MSc.
I now know that building a chassis with carbon fibre is very expensive.
But then I already knew that. I won't be sitting any engineering degree exams anytime soon.
There are two ways of viewing work like this:
The easy way is to sit back and say how obvious the findings are; look at a chassis find rectangle openings and either turn them into triangles or
cover the opening with a plate!
But that’s not what his work set out to do, like us, he already knew that was the approach to take (possibly he had seen the other Ultima build site,
so what!), what he did set out to do was to test this approach via cad modelling and computer testing. Basically to put figures to these
modifications.
What I would have liked is to see is a comparison of single diagonal braces against the “X” bracing that were chosen to see if the additional weight
of the “X” was justified. Also I’d have liked to see the standard chassis put on a stress jig to check the correlation of the model, likewise if the
modifications were actually done have the modified one likewise tested.
Thanks for posting the link which was interesting.
The low cost (Canadian $50) option is LISA Finite Element Technologies from http://www.lisa-fet.com/ that I have played with and can recommend.
Lots of interesting looking software here
http://os.cqu.edu.au/oswinsdvd/doc/engineering.html
quote:
hat I would have liked is to see is a comparison of single diagonal braces against the “X” bracing that were chosen to see if the additional weight of the “X” was justified.
quote:
Originally posted by coyoteboy
quote:
hat I would have liked is to see is a comparison of single diagonal braces against the “X” bracing that were chosen to see if the additional weight of the “X” was justified.
According the chassis design books I've read, the answer is "not really but application specific" - diminishing returns and increasing weight.
[Edited on 6/1/12 by coyoteboy]
Indeed, hence the application specific - it doesn't make much difference but should you not wish to compromise in some locations then take it to that extreme. As a general chassis triangulation it's OTT, for a safety cell it may help (or hinder, too stiff and you'll massively increase impact forces on the driver of course).
Here in the states I remember them doing a 35-mph crash test of a NASCAR racecar. As I recall, the car only got dented in about a foot (and was able
to drive away), but the datalogs showed that the driver would have died due to the high deceleration... It's a fine line we walk...
[Edited on 1/7/12 by kb58]
Really? A Nascar driver would die in pretty much any crash they have? I must have read that wrong...
quote:
Originally posted by kb58
Here in the states I remember them doing a 35-mph crash test of a NASCAR racecar. As I recall, the car only got dented in about a foot (and was able to drive away), but the datalogs showed that the driver would have died due to the high deceleration... It's a fine line we walk...
[Edited on 1/7/12 by kb58]
Make the vehicle stiff enough and you're effectively just throwing the occupants at the object you crashed into at the impact speed, ive no idea of figures of stIffness of a nascar in frontal impact though, so can't comment on that.
IMHO full finite analysis is a waste of time for a tube chassis, it’s much faster to use a structural analysis program such as this one
(Framework):
http://members.ziggo.nl/wolsink/
Which is what I used myself, this is the type of analysis that done when doing basic analysis of structures such as steel framed buildings or power
pylons. Basically the whole PITA step of meshing is skipped as you have members of fixed cross sections going from node to node.
Even in buildings with cast in floors (which as in a similar fashion to sheet metal stiffening a chassis) Structural Engineer often still do not go to
full FEA but put in ‘fake’ elements to act as the equivalent of the sheet body.
On X bracing v other types, the length of a tube in compression will determine how much load it can take before buckling, and there’s a square
relationship so reduce the length of the tube and the potential load it will take will go up a lot. By X bracing you effectively halve the length of
the tube being considered in compression.
quote:
Originally posted by Doug68
IMHO full finite analysis is a waste of time for a tube chassis, it’s much faster to use a structural analysis program such as this one (Framework):
By X bracing you effectively halve the length of the tube being considered in compression.
quote:
Originally posted by wylliezx9r
quote:
Originally posted by kb58
Here in the states I remember them doing a 35-mph crash test of a NASCAR racecar. As I recall, the car only got dented in about a foot (and was able to drive away), but the datalogs showed that the driver would have died due to the high deceleration... It's a fine line we walk...
[Edited on 1/7/12 by kb58]
35 mph ? Or 135 mph ?