Jump to content


Photo

Structural optimisation


  • Please log in to reply
16 replies to this topic

#1 Greg Locock

Greg Locock
  • Member

  • 6,353 posts
  • Joined: March 03

Posted 28 April 2012 - 06:37

This is a bit of a sales blurb http://insider.altai...ember_110896633

I've found it very handy for designing castings for suspension arms, but I have never got to the point of actually making and testing an optimised part. I have used it for arranging tubes in a spaceframe, and did look at an improved ladder frame as well.

The shape of the optimised beam in figure 3 strikes me as being a bit odd, I'd have expected more of a traditional triangular shape than this rather curvy thing.



Advertisement

#2 Tony Matthews

Tony Matthews
  • Member

  • 17,519 posts
  • Joined: September 08

Posted 28 April 2012 - 07:27

Some of the examples look very like the work of the architect Zaha Hadid, or the internal structure of avian long bones.

#3 rachael

rachael
  • Member

  • 118 posts
  • Joined: October 10

Posted 28 April 2012 - 08:14

This is a bit of a sales blurb http://insider.altai...ember_110896633

I've found it very handy for designing castings for suspension arms, but I have never got to the point of actually making and testing an optimised part. I have used it for arranging tubes in a spaceframe, and did look at an improved ladder frame as well.

The shape of the optimised beam in figure 3 strikes me as being a bit odd, I'd have expected more of a traditional triangular shape than this rather curvy thing.


Figure 3 looks similar to this classic railway bridge; Crescent bridge

I've used this software alot - often the optimisation comes up with organic designs and it's hard to decide whether they are truly optimal or not.



#4 gruntguru

gruntguru
  • Member

  • 7,637 posts
  • Joined: January 09

Posted 28 April 2012 - 09:59

The shape of the optimised beam in figure 3 strikes me as being a bit odd, I'd have expected more of a traditional triangular shape than this rather curvy thing.

A couple of interesting observations
- The structure is mostly triangulated
- The members in compression have a lower L/R than those in tension
- The almost vestigial, upper "necklace" structure seems to have no function other than to tie four compression members together and prevent buckling.

It helps to imagine the structure as a single, simple triangle with the two sloping sides replaced by the two fish-shaped props meeting at the apex.

#5 Magoo

Magoo
  • Member

  • 3,703 posts
  • Joined: October 10

Posted 28 April 2012 - 10:03

Some of the examples look very like the work of the architect Zaha Hadid, or the internal structure of avian long bones.


The bonelike examples make me think they are really onto something, though just what might not be clear. There are no mistakes in biology, only processes we can't totally understand.


#6 NeilR

NeilR
  • Member

  • 623 posts
  • Joined: October 09

Posted 28 April 2012 - 10:50

There are no mistakes in biology...


Oh...I can think of a few people...


#7 desmo

desmo
  • Tech Forum Host

  • 29,394 posts
  • Joined: January 00

Posted 29 April 2012 - 13:44

The bonelike examples make me think they are really onto something, though just what might not be clear. There are no mistakes in biology, only processes we can't totally understand.


Indeed. Evolutionary biology is iterative development at the highest level. If the output of a design process doesn't look biological in complexity, it probably is optimized more for ease of conception and construction using conventional components than function. Of course designing organic looking structures with no real underlying understanding of why they are the way they are in nature is worse.




#8 MatsNorway

MatsNorway
  • Member

  • 2,822 posts
  • Joined: December 09

Posted 29 April 2012 - 21:27

There is little co-operation in nature, it is a evolutionary race.. and it has been that way for millions of years.. But the design process is dumb. Thats why nature got no birds that does mach 1.



#9 gruntguru

gruntguru
  • Member

  • 7,637 posts
  • Joined: January 09

Posted 29 April 2012 - 23:36

There is little co-operation in nature, it is a evolutionary race.. and it has been that way for millions of years.. But the design process is dumb. Thats why nature got no birds that does mach 1.

More to do with the fact that there is no survival advantage for a bird to go that fast.

#10 pugfan

pugfan
  • Member

  • 177 posts
  • Joined: August 09

Posted 30 April 2012 - 00:45

The beam case is very trivial, when would a beam ever be loaded like that?

As the part becomes more optimised, it will be less tolerant of non designed for loads I would have thought.

Safety factors wouldn't necessarily compensate for this increased 'fragility' because the safety factors will only be applied to the designed load cases.

#11 bigleagueslider

bigleagueslider
  • Member

  • 1,235 posts
  • Joined: March 11

Posted 30 April 2012 - 00:46

This is a bit of a sales blurb http://insider.altai...ember_110896633

I've found it very handy for designing castings for suspension arms, but I have never got to the point of actually making and testing an optimised part. I have used it for arranging tubes in a spaceframe, and did look at an improved ladder frame as well.


Greg Locock,

Where I work the aero guys use a similar application called Sculptor for optimizing aero surfaces.

http://www.optimalso...tions/index.php

slider


#12 gruntguru

gruntguru
  • Member

  • 7,637 posts
  • Joined: January 09

Posted 30 April 2012 - 03:01

Evolutionary biology is iterative development at the highest level.

Albeit a little slow.

#13 Greg Locock

Greg Locock
  • Member

  • 6,353 posts
  • Joined: March 03

Posted 30 April 2012 - 03:04

The beam case is very trivial, when would a beam ever be loaded like that?

As the part becomes more optimised, it will be less tolerant of non designed for loads I would have thought.

Safety factors wouldn't necessarily compensate for this increased 'fragility' because the safety factors will only be applied to the designed load cases.

1)it's just an example, but if you were designing a bridge or other beam-like part it might give you some ideas.
2)Obviously
3)Obviously.

It seems odd to criticise a tool for exhibiting the same failings as the real world. The more you know about the loadings the lighter you can make it, but if there are loadcases that you haven't considered then they may cause a failure.




#14 bigleagueslider

bigleagueslider
  • Member

  • 1,235 posts
  • Joined: March 11

Posted 30 April 2012 - 03:58

It seems odd to criticise a tool for exhibiting the same failings as the real world. The more you know about the loadings the lighter you can make it, but if there are loadcases that you haven't considered then they may cause a failure.


That's a very interesting point about how these optimization softwares iterate a solution versus how biological evolution works in theory. The softwares converge on a solution based on some fixed targets and input constraints for a single example. Nature on the other hand, uses numerous examples exposed to lots of different changing variables, with some random unpredictable effects thrown in for good measure.

In nature, similar inputs can result in different results. For example, why don't bats have hollow bones like birds?

Regards,
slider

#15 Tony Matthews

Tony Matthews
  • Member

  • 17,519 posts
  • Joined: September 08

Posted 30 April 2012 - 06:44

In nature, similar inputs can result in different results. For example, why don't bats have hollow bones like birds?

A case of not starting from the same point, I suppose, in the same way that cetaecean tails move up and down, fish tails move side to side. If only F1 displayed the same differences from team to team...

#16 MatsNorway

MatsNorway
  • Member

  • 2,822 posts
  • Joined: December 09

Posted 30 April 2012 - 13:57

More to do with the fact that there is no survival advantage for a bird to go that fast.


Now your just being argumentative. if a cheeta evolved to go 200k the prey would have been evolved to go that too or found some other way to survice . Its a race..

#17 pugfan

pugfan
  • Member

  • 177 posts
  • Joined: August 09

Posted 02 May 2012 - 07:04

It seems odd to criticise a tool for exhibiting the same failings as the real world. The more you know about the loadings the lighter you can make it, but if there are loadcases that you haven't considered then they may cause a failure.


I guess the magic bullet flavour to the speil put me off a bit.